Cognitive Approaches to Pedagogical Grammar: A Volume in Honour of RenA© Dirven (Applications of Cognitive Linguistics)

Cognitive Approaches to Pedagogical Grammar ≥ Applications of Cognitive Linguistics 9 Editors Gitte Kristiansen Mic...

Author: De Knop | Sabine

112 downloads 3769 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Cognitive Approaches to Pedagogical Grammar

≥

Applications of Cognitive Linguistics 9

Editors Gitte Kristiansen Michel Achard Rene´ Dirven Francisco J. Ruiz de Mendoza Iba´n˜ez

Mouton de Gruyter Berlin · New York

Cognitive Approaches to Pedagogical Grammar A Volume in Honour of Rene´ Dirven

Edited by Sabine De Knop Teun De Rycker

Mouton de Gruyter Berlin · New York

Mouton de Gruyter (formerly Mouton, The Hague) is a Division of Walter de Gruyter GmbH & Co. KG, Berlin

앝 Printed on acid-free paper 앪

which falls within the guidelines of the ANSI to ensure permanence and durability.

Library of Congress Cataloging-in-Publication Data Cognitive approaches to pedagogical grammar / edited by Sabine De Knop, Teun De Rycker. p. cm. ⫺ (Applications of cognitive linguistics ; 9) Includes bibliographical references and index. ISBN 978-3-11-019595-8 (hardcover : alk. paper) 1. Language and languages ⫺ Study and teaching. 2. Cognitive grammar. 3. Corpora (Linguistics) I. Knop, Sabine De. II. Rycker, Teun De. P53.C568 2008 407⫺dc22 2007052034

Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de.

ISBN 978-3-11-019595-8 ISSN 1861-4078 쑔 Copyright 2008 by Walter de Gruyter GmbH & Co. KG, D-10785 Berlin All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Printed in Germany

Acknowledgements It is only fitting that someone who has made such a substantial and lasting contribution to the field of linguistics over the past fifty years should be celebrated with a volume in his honour. When René Dirven turned sixty in 1992, a Festschrift was published to mark the occasion, reflecting in its contributions from all over the world his wide-ranging academic interests, passions as well as his numerous friendships and connections. His retirement from full-time professorship five years later has not slowed, however, his appetite for language research. Over the past decade René has been involved in an impressive number of book projects, to which can be added his usual output as a researcher, writer and inquisitive thinker. More than anything else perhaps we feel it is his continued support and encouragement of others – in their own (cognitive) exploration of language and linguistics – that has been invaluable. When René contacted us in 2006 to start work on this volume, we had no idea of the journey ahead, neither in terms of the workload or what it would mean to collaborate so closely with someone of his intellectual caliber, with someone who is so au fait not just with the literature but also with newly emerging trends. We would like to thank René sincerely for his extensive and insightful guidance during the production of this volume, and it is with pleasure that we dedicate it to him. It is largely thanks to his critical but constructive comments on many of the earlier drafts that the final articles are of such a high quality. Talking of these articles, we really enjoyed working together with the eighteen authors who so generously contributed their research. We would like to express our gratitude to them for their professionalism, commitment and patience. Our thanks also go to the anonymous reviewer whose detailed assessment has helped us edit and fine-tune the manuscript in its final stages. We would also like to thank Birgitta Meex and Tanja Mortelmans, who were both involved in getting the project on the rails but who had to leave the editorial team due to other professional obligations. Finally, we greatly appreciate the help received from Birgit Sievert, managing editor of the Applications of Cognitive Linguistics series, as well as our other Mouton de Gruyter contacts and especially editor-in-chief, Anke Beck, for her belief in our project’s potential.

Table of contents

Acknowledgements List of contributors

..................................... .....................................

v ix

By way of introduction .................................. Sabine De Knop and Teun De Rycker

1

Part I: Cognition and usage: Defining grammar, rules, models and corpora The relevance of Cognitive Grammar for language pedagogy Ronald W. Langacker Some pedagogical implications of cognitive linguistics John R. Taylor

....

7

........

37

Cognitive linguistic theories of grammar and grammar teaching Cristiano Broccias

..

67

Corpora, cognition and pedagogical grammars: An account of convergences and divergences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fanny Meunier

91

Part II: Tools for conceptual teaching: Contrastive and error analysis Cross-linguistic analysis, second language teaching and cognitive semantics: The case of Spanish diminutives and reflexive constructions .......................................... Francisco José Ruiz de Mendoza Ibáñez Spanish middle syntax: A usage-based proposal for grammar teaching .............................................. Ricardo Maldonado

121

155

viii

Table of contents

What can language learners tell us about constructions? ........ Javier Valenzuela Manzanares and Ana María Rojo López

197

Conceptual errors in second-language learning Marcel Danesi

231

................

Part III: Conceptual learning: Construal of motion, temporal structure, and dynamic action Motion events in Danish and Spanish: A focus on form pedagogical approach ............................................. Teresa Cadierno

259

Motion and location events in German, French and English: A typological, contrastive and pedagogical approach . . . . . . . . . . . . . Sabine De Knop and René Dirven

295

Making progress simpler? Applying cognitive grammar to tenseaspect teaching in the German EFL classroom ................ Susanne Niemeier and Monika Reif

325

Aspectual concepts across languages: Some considerations for second language learning ................................ Barbara Schmiedtová and Monique Flecken

357

The use of passives and alternatives in English by Chinese speakers .............................................. Liang Chen and John W. Oller, Jr.

385

Author index Subject index

417 427

.......................................... ..........................................

List of contributors

Cristiano Broccias Department of Linguistic and Cultural Communication Università di Genova Italy [email protected] Teresa Cadierno Institute of Language and Communication University of Southern Denmark Odense Denmark [email protected] Liang Chen Communication Sciences and Special Education College of Education University of Georgia Athens, GA USA [email protected] Marcel Danesi University of Toronto Ontario Canada [email protected] Sabine De Knop Department of Germanic Languages and Linguistics Facultés universitaires SaintLouis

Brussels Belgium [email protected] Teun De Rycker Lessius Hogeschool Antwerpen Belgium [email protected] René Dirven Universität Duisburg-Essen Germany and Mechelen Belgium [email protected] Monique Flecken Seminar für Deutsch als Fremsprachenphilologie, University of Heidelberg Germany [email protected] Ronald W. Langacker Department of Linguistics University of California San Diego, CA USA [email protected] Ricardo Maldonado Universidad Nacional Autónoma de México and Universidad Autónoma de Querétaro Mexico

x

List of contributors

[email protected] Fanny Meunier Centre for English Corpus Linguistics Université catholique de Louvain Louvain-la-Neuve Belgium [email protected]

Francisco José Ruiz de Mendoza Ibáñez Department of Modern Languages Faculty of Humanities and Education University of La Rioja Spain [email protected]

Susanne Niemeier University Koblenz-Landau Campus Koblenz English Department Koblenz Germany [email protected]

John R. Taylor Department of English University of Otago New Zealand [email protected]. nz

John W. Oller, Jr. Department of Communicative Disorders University of Louisiana Lafayette USA [email protected]

Barbara Schmiedtová Seminar für Deutsch als Fremdsprachenphilologie University of Heidelberg Germany [email protected]

Monika Reif University Koblenz-Landau Campus Landau English Department Landau Germany [email protected]

Javier Valenzuela Manzanares Department of English Philology University of Murcia Spain [email protected]

Ana María Rojo López Department of English Philology University of Murcia Spain [email protected]

By way of introduction Sabine De Knop and Teun De Rycker The notion of pedagogical grammar is ambiguous, being used grammatically either as a mass noun or as a count noun. The mass noun stands metonymically for a process, i.e., research into grammar pedagogy, and as a count noun it refers to the concrete outcome of that research in the form of a pedagogical grammar package for a specific age and proficiency level learning a given target language. In this volume, the term is mostly used in the first sense, i.e., it most of all refers to the research process. Of course, if continued systematically, some studies in the present volume may eventually lead to a partial pedagogical grammar in a certain area of the language concerned. It may be important to emphasize that a pedagogical grammar, as the term package suggests, is different from a didactic grammar or school grammar in the sense that it can never be equated with, nor be intended as, a practical manual for teachers, let alone learners. Put in the simplest way, a pedagogical grammar is both an inventory of all the formmeaning units of the target language, and a didactic approach to their acquisition. Therefore a pedagogical grammar will always consist of two interwoven layers. The first layer is a representation of the main units of a language, which according to the theoretical viewpoint taken can be lexical, grammatical or constructional in nature, and possibly also representations of more abstract metacategories such as subject, noun or verb, or representations of abstract schemas and rules. This will form the main topic of the more theoretical papers in Part I. Interwoven with a survey of all the form-meaning units is the development of an optimal teaching and learning infrastructure to facilitate their acquisition. In fact, these two layers of a pedagogical grammar come in most research programs as two or more separate phases. Indeed, one of the first necessary steps is the examination of the contribution that theoretical, descriptive and contrastive linguistics can make towards the inventory of items to be learned and the facilitation of their learning and teaching in guided learning situations. Consequently, pedagogical grammar research can be practiced in a number of different ways, some of which are focused upon in this volume.

2

Sabine De Knop and Teun De Rycker

— Providing the repertory of all the essential linguistic units considered relevant by theoretical and descriptive linguistics in a given language. The discussion of what these units are is the central theme of Part I. — Pointing out the areas of overlap and contrast between a target language and the learner’s native language, and analysing interferences between those languages in the growth of the learner’s interim grammar, i.e., the grammar that he or she develops in attaining some degree of “conceptual fluency” in the target language (to borrow the term used by Danesi in this volume). These research areas are traditionally known as contrastive analysis and error analysis, and form the focus of Part II. — Delineating learning problems by means of certain conceptual areas, either differently represented in the two languages or differently matched in the pairing of forms and meanings. Such descriptive and teaching- or learning-oriented analyses can directly or indirectly serve the purpose of preparing teaching materials and exploiting these in L2/FL instruction. These areas of pedagogical grammar are further structured and explored in each of the volume’s main parts as follows. Part I, Cognition and usage: Defining grammar, rules, models and corpora, tries to answer the question of what the “basic form-meaning units” of language are – an essential condition for setting up a repertory of the target language units. Within Cognitive Linguistics there is, in spite of the numerous common assumptions about language, a great divergence of opinion on what these basic units are. Are they mainly lexical units with grammatical dependencies or arguments, as assumed in Hudson’s (1991) Word Grammar model? Or do they, on top of lexical and grammatical categories, also include metatheoretical abstract categories for word classes and sentence constituents, as claimed by Langacker? Do basic units also include abstract schemas or “rules”, as Taylor suggests? Or do they mainly contain “constructions” – i.e., fixed, even idiomatic sequences at phrase level or sentence level – as claimed in Goldberg’s (1995) Construction Grammar and in Croft’s (2001) Radical Construction Grammar, both surveyed in Broccias’ contribution? Before one can even start pedagogical grammar research, one has to come to grips with these basic questions. In applying the usage-based model that Cognitive Linguistics claims to be, it will also be necessary to define what counts as admissible data, that is, the corpora one can use in linguistics and in

By way of introduction

3

language pedagogy. This is explored by Meunier, who compares the various meeting grounds between corpus linguistics and cognitive linguistics. Part II, Tools for conceptual teaching: Contrastive and error analysis, examines what a cognitive approach to these two traditional areas of pedagogical grammar can contribute to the field of conceptual learning. Contrastive analysis plays an important role in Ruiz de Mendoza’s and Maldonado’s contributions. Both look at Spanish as the target language. Thus the former explores the diminutive with its very rich conceptual differentiation in Spanish vs. its minimal significance in English. The latter explores se constructions in Spanish mainly as variants of the ubiquitous middle construction, in comparison with which the se reflexive is only a minor area for learners of Spanish as a foreign language. Whereas the data in Maldonado’s analysis are based on error analysis as a data base, this becomes the main topic of Valenzuela and Rojo’s and Danesi’s papers. Valenzuela and Rojo base their error analysis of the use of the English ditransitive construction on data from the Spanish component of the International Learner Corpus of English and show that constructions exist in the English data of L2 learners, who acquire certain constructional configurations in an exemplar-based way, just as Tomasello (2000, 2003) did for L1 acquisition. Danesi, who replaces the Chomskyan notion of a disembodied “grammatical competence” by the more person- and interactionoriented notion of “conceptual fluency” and that of “formal-linguistic error” by “conceptual error”, examines the errors and lack of metaphorical or other associations in the use of color terms by English learners of Italian. Part III, Conceptual learning: Construal of motion events, temporal structure, and dynamic action offers analyses of learning problems in these three areas. As Talmy (1985) has shown, motion events are conceptually structured quite differently in Romance and Germanic languages. The construal of motion events is explored from the viewpoint of Spanish speakers learning Danish by Cadierno and of French speakers learning German or English by De Knop and Dirven. The construal of external temporal structure (tense) and internal temporal structure (aspect) by German learners of English is described by Niemeier and Reif, who also devise a teaching and learning experiment based on these learning difficulties. Schmiedtová and Flecken concentrate on the fundamental differences between the two Slavic aspects (perfective vs. imperfective) and the progressive aspect in English and Dutch. Chen and Oller explore the different ways of construing dynamic action by the English passive constructions and its many variants and manage to unveil the very restricted use of them by Chinese learners of English. All in all, most papers in the volume amply demonstrate the array

4

Sabine De Knop and Teun De Rycker

of potential choices to construe a scene and the extremely restricted use made of them by foreign learners. Conceptual learning and teaching is the first point on the future research agendas of language pedagogy in general and of pedagogical grammar in particular. Like any other comprehensive theory of language, CL is faced with the problem of turning a rich, specialized and emerging body of applied cognitive linguistic research into a practical guide for foreign-language teachers, course designers and materials writers. Language pedagogy is of necessity interdisciplinary in character, crossing over into and closely collaborating with, among others, psycholinguistics and educational psychology. What CL brings to this crowded and rather confusing field of study – more than any other contemporary form of linguistics – is a powerful conceptual unity. The contributions to this volume have shown that this unity in theoretical assumptions, basic insights and constructs produces highly relevant research. We hope that they may also offer rewarding avenues for further exploration of what it means “to think before you speak” in a foreign language.

References Croft, William 2001 Radical Construction Grammar: Syntactic Theory in Typological Perspective. Oxford: Oxford University Press. Goldberg, Adele E. 1995 Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press. Hudson, Richard A. 1991 Word Grammar. Oxford: Blackwell. Talmy, Len 1985 Lexicalization patterns: Semantic structure in lexical forms. In Language Typology and Syntactic Description, Vol. 3: Grammatical Categories and the Lexicon, Timothy Shopen (ed.), 36–149. Cambridge: Cambridge University Press. Tomasello, Michael 2000 First steps toward a usage-based theory of language acquisition. Cognitive Linguistics 11: 61–82. 2003 Constructing a Language: A Usage-based Theory of Language Acquisition. Harvard: Harvard University Press.

Part I Cognition and usage: Defining grammar, rules, models and corpora

The relevance of Cognitive Grammar for language pedagogy1 Ronald W. Langacker

Abstract Cognitive Grammar offers a natural and promising basis for language instruction. It advances a conceptual account of linguistic meaning which, by showing how alternate expressions construe the same situation in subtly different ways, renders comprehensible the varied means of expression a language provides. This conceptual semantics is not confined to lexicon but also extends to grammar: every grammatical element or construction imposes a particular construal on the situation being described. Grammar can thus be presented as an array of meaningful options whose ranges of application are in large measure predictable. Also important is the usagebased nature of Cognitive Grammar. Language structure emerges by abstraction from usage events, embracing all dimensions of how expressions are understood by interlocutors in the social, cultural, and discourse context. Implications for language learning include the importance of non-descriptive modes of speech, the need to produce and understand appropriate expressions in a natural context, and the dependence of fluent speech on mastery of a vast array of conventional expressions and phraseology. Keywords: cognitive grammar; language instruction; centrality of meaning; meaningfulness of grammar; units; constructions; composition; constructional schema; schematization; conceptual substrate; usage-based perspective; mental construction; metaphorical construction; fictive motion; mental space; pedagogical implications

1. Introduction Few would maintain that language instruction is easy. Nor can the advice of linguists always be counted on to make it any easier. Unless they are themselves experienced language teachers, the advice of linguists on language pedagogy is likely to be of no more practical value than the advice of theoretical physicists on how to teach pole vaulting. What they can offer, qua linguists, is insight into the structure of particular languages and

8

Ronald W. Langacker

the properties of language in general. But even when limited in this fashion, the input of linguists cannot necessarily be trusted. They quarrel with one another about the most fundamental issues, suggesting that some of them (at least) must be fundamentally wrong. It is therefore unsurprising that the impact of linguistic theory on language pedagogy has been less than miraculous and sometimes less than helpful. It remains to be seen whether language teaching will fare any better when guided by notions from cognitive linguistics. There are however grounds for being optimistic. Compared to other approaches, cognitive linguistics offers an account of language structure that – just from the linguistic standpoint – is arguably more comprehensive, revealing, and descriptively adequate (certainly I have argued this, e.g., in Langacker 1995a). More to the point, the present discussion will focus on three basic features of Cognitive Grammar (Langacker 1987a, 1990, 1991, 1999a) that suggest its potential utility as a basis for language instruction: the centrality of meaning, the meaningfulness of grammar, and its usage-based nature. Although extensive pedagogical application remains a long-term goal, I regard its effectiveness in language teaching to be an important empirical test for the framework.

2. The centrality of meaning If generative linguistics views syntax as being central to language, cognitive linguistics accords this honor to meaning. The latter seems far more natural from the perspective of language users. When ordinary people speak and listen, it is not for the sheer pleasure of manipulating syntactic form – their concern is with the meanings expressed. This does not of course imply that grammar is unimportant in language or in language teaching. It is however helpful to realize that grammar subserves meaning rather than being an end in itself. The centrality of meaning is reflected in a fundamental claim of Cognitive Grammar (henceforth CG), namely that lexicon and grammar form a continuum consisting solely in assemblies of symbolic structures. A symbolic structure is nothing more than the pairing of a semantic structure and a phonological structure. It follows from this claim that grammar itself is meaningful, just as lexical items are. Grammatical meanings are generally more abstract than lexical meanings. This is however a matter of degree, so there is no clear line between lexicon and grammar.

The relevance of Cognitive Grammar for language pedagogy

9

Over the years, progress in describing meanings was greatly impeded by certain assumptions which, from the standpoint of cognitive semantics, were simply gratuitous. First is the notion that a given element, such as a lexical item, has just a single linguistic meaning. A glance at any comprehensive dictionary makes this seem quite dubious. Although many details have yet to be resolved (Allwood 2003; Langacker 2006; Sandra and Rice 1995; Zlatev 2003), the basic point that lexical items are frequently polysemous – having multiple, related senses – has been established through numerous case studies (e.g., Brugman 1981; Lindner 1982; Tuggy 2003; Tyler and Evans 2003). The polysemy of lexical items is a special case of the general cognitive linguistic claim, also well established, that linguistic categories are usually complex: their full description takes the form of a network of related variants (Lakoff 1987; Langacker 1987a; Taylor 2004). A second standard but gratuitous assumption is that a lexical item’s meaning is circumscribed and distinct from general knowledge. Metaphorically, it is like a dictionary entry: a short definition in a special format, capturing everything speakers know about the entities denoted just by virtue of knowing the language. The problem, pace Wierzbicka (1995), is the absence of any non-arbitrary way to draw the line (Haiman 1980). Instead of being distinct from general knowledge, lexical meanings recruit and exploit it, representing particular ways of viewing it and making it accessible for linguistic purposes. Though some specifications are clearly more central and frequently accessed than others, virtually any aspect of our knowledge of the entities denoted can be invoked for linguistic purposes (Langacker 1987a, 2003). More basic is the assumption that meaning resides in correspondences with the world: the set of entities a word denotes, or the conditions under which a sentence is true. In stark contrast to the objectivist tradition, where the mind is left out of the loop, cognitive semantics views meaning as a mental phenomenon. It resides in conceptualizing activity, whereby we engage the world at many levels: physical, mental, social, cultural, emotional, and imaginative. Crucially, we have the capacity to conceive and portray the same situation in alternate ways. While there may be defaults, there is no completely neutral way to describe situations – expressions necessarily construe them in a certain manner. An expression’s meaning is therefore only partly determined by objective properties of the situation described (if, indeed, it has any objective existence at all). One dimension of construal is specificity, i.e., degree of precision and detail. It is reflected in lexical hierarchies like thing > creature > animal >

10

Ronald W. Langacker

dog > poodle, or do > act > move > run > sprint. Within such a hierarchy, there is usually a basic level (in this case dog or run) at which we tend to operate lacking any reason to be more specific or more schematic (Rosch 1978). For a given lexical choice, greater specificity can always be achieved through modifiers and more elaborate descriptions. One measure of success in language learning is facility in moving away from basic level descriptions and going into sub- and superordinate levels of categorization; this has been dubbed “conceptual fluency” (Danesi 1995, 1999; Kecskes and Cuenca 2005). A second dimension of construal is prominence, of which there are various sorts. Two are especially important: profiling and the focal prominence of relational participants. An expression’s profile is what it designates, i.e., its referent within the array of conceptual content it evokes as the basis for its meaning. For instance, roof evokes the conception of a building (primarily as seen from the outside), within which it profiles the structural portion that covers it on top. It contrasts with ceiling, which is seen from the inside and covers part or all of it. It is predictable that many languages do not make this distinction, using the same term for both concepts. Similarly, it is a language-specific matter that brother-in-law can designate either a wife’s brother or a sister’s husband. The word key invokes the conception of locks and how they function, profiling an object characterized most essentially by its role in their operation. In the diagrams of Figure 1 the heavy lines indicate profiling; the less heavy lines in diagram (a) indicate the base. In diagrams (b)–(d) the base – an indefinite spatial expanse – is not separately indicated. (a) roof

(b) on

(c) move (intrans.)

tr

tr

(d) move (trans.) tr

lm

lm

Figure 1. Schemas for roof, on, intransitive move and transitive move

As Figure 1 shows, expressions profile either things, as in (a), or relationships, as in (b)–(d), assuming very general definitions of those terms (Langacker 1987b). In the case of relational expressions, a second kind of prominence comes into play: the degree of salience conferred on the participants in the profiled relationship. There is generally a primary focal


11

participant, called the trajector (tr), and often a secondary focal participant, called the landmark (lm). The trajector is the entity being described, located, or otherwise characterized. As shown in Figure 1(b), on profiles a relationship where (prototypically) the trajector is in contact with the upper surface of the landmark, which supports it (double-headed arrow). The verb move has two basic senses, sketched in diagrams (c) and (d). The intransitive move profiles a relationship wherein the self-moving trajector successively occupies a series of locations (single arrow). As a transitive verb, its trajector is characterized as somehow exerting force (double arrow), acting on the landmark and thereby causing its motion. Let us note just one more dimension of construal, namely perspective, which is also multifaceted. Among its facets are vantage point and orientation, implicated in the various interpretations of Jack was sitting to the left of Jill. For instance, if the speaker was facing both Jack and Jill, the sentence may indicate either that Jack was to the speaker’s left (and Jill’s right), or else that Jack was to Jill’s left (hence the speaker’s right). Another facet is the contrast between a local and a global perspective. For example, sentence (1a) is the sort of thing one would say while actually travelling along the road. What counts as this road is then the portion one can see at a given moment. Use of the progressive (is winding) indicates that the road (so characterized) changes position through time vis-à-vis the mountains (Langacker 1987b). On the other hand, (1b) is the sort of thing one would say when the entire configuration of road and mountain is apprehended as a single gestalt (e.g., in looking at a map). Use of the simple present tense (winds) indicates that their relationship is stable through time. (1)

a. b.

This road is winding through the mountains. [local perspective] This road winds through the mountains. [global perspective]

The pervasive importance of construal shows clearly that linguistic meaning does not reside in the objective nature of the situation described, but is crucially dependent on how the situation is apprehended. Indeed, the situation in question is very often a mental construction which has no objective existence in the first place. Much of what we express linguistically is imaginative in nature, even in talking about actual occurrences. The sentences in (1), for instance, can both be used in reference to an actual configuration of road and mountain. They nonetheless involve what is known

12

Ronald W. Langacker

as fictive (or virtual) motion: although the road is really stationary, we describe it with expressions (like wind and through the mountains) normally used for movement along a path: (2)

A snake was winding through the grass. [actual motion]

In (1a), we construct a sense of motion by imagining distinct road segments (the portions successively in view) as being a single moving entity. In (1b), we do so by mentally scanning along the road’s extension (the same scanning we do when conceptualizing something actually moving along it). Fictive motion is varied, extremely common, and psychologically real (Langacker 1986; Matlock 2001, 2004; Matlock and Richardson 2004; Matsumoto 1996a; Talmy 1996). It is a special case of fictive change (Dapremont 2001; Matsumoto 1996b; Sweetser 1997). Some further examples of fictive change are given in (3). (3)

a. b.

The company’s president keeps getting younger. broken line (cf. broken pencil); scattered villages (cf. scattered marbles); sunken bathtub (cf. sunken ship)

The president who defies the laws of nature is a fictive entity analogous to the road in (1a). We mentally construct this person by treating different instantiations of the role as if they were a single individual. Past participles like broken, scattered, and sunken normally designate the situation resulting from a change-of-state process, e.g., a broken pencil is one that has undergone the process of breaking. The uses cited, however, do not involve any actual change – scattered villages have never been clustered together, nor has a sunken bathtub ever actually sunk. The change implied is only virtual, representing the conceived departure of the situation described from the one regarded as canonical. Everyday language is replete with references to fictive entities, often invoked for describing actual situations, as in (4). (4)

a. b. c. d.

He doesn’t have a sister. If she writes a novel, she will try to publish it. A kitten likes to chase its tail. Each boy was holding a frog.

If we take the sentences in (4) as truthful statements of what the world is actually like, the nominals in bold nevertheless have referents which are


13

only virtual – they are “conjured up” as parts of mental constructions that relate to actuality in various ways. In sentence (4a), a sister is conjured up precisely in order to specify what is not the case. The novel referred to in (4b) exists only in the hypothetical situation introduced by if. A generic statement like (4c) represents a generalization about the world’s essential structure. As such, it does not refer to any actual kitten or its tail, but rather to fictive instances of these types which correspond to open-ended sets of actual ones. Similarly, (4d) makes a generalization about a delimited set of actual boys. The entities directly expressed linguistically are only virtual, however: each boy is not any actual boy, nor does a frog designate a specific frog; these virtual entities are invoked to describe a virtual relationship taken as corresponding to a contextually determined set of actual ones. Despite their fictivity, these nominals all have referents in the linguistically relevant sense. Observe that the pronoun it refers back to the novel in (4b), and to the kitten in (4c). Much of what we talk about is constructed metaphorically (Kövecses 2005; Lakoff and Johnson 1980; Lakoff and Núñez 2000; Turner 1987). In (5), for instance, achieving a goal is metaphorically construed as hitting a target (aim at), drug dependency as captivity (free from), and the lure of drugs as fishing (hooked). Even the preposition on is metaphorical: to be on drugs is to be in contact with them and require them for support. (5)

The therapy is aimed at freeing him from the drugs he is hooked on.

Metaphor is a special case of blending (Fauconnier and Turner 2002), where selected elements of two conceptions are projected and integrated to form a third, which is often purely imaginative but nonetheless real as an object of thought. One such case is a cartoon character, e.g., a dog that behaves like a person and thinks in English. Brunch is a blend of breakfast and lunch (both conceptually and phonologically). Blending in turn is a special case of mental space configurations (Fauconnier 1985, 1997; Fauconnier and Sweetser 1996). In both thought and discourse, we divide our mental world into separate representational “spaces,” which are connected to one another in particular ways but nonetheless retain a measure of autonomy. In (4b), if introduces a hypothetical space distinct from reality. Generic statements like (4c) pertain to a space representing the world’s essential nature (Goldsmith and Woisetschlaeger 1982; Langacker 1999b). In (5), aim at implies a goal and thus a person who entertains that goal. The

14

Ronald W. Langacker

goal is a special mental space within the space representing that person’s thoughts. Though generally implicit, these various kinds of mental constructions are crucial to both the form and meaning of expressions. They are facets of an elaborate conceptual substrate that supports and makes coherent the notions overtly expressed. This substrate gives us the freedom to be selective in what we explicitly code linguistically, knowing that the listener can fill in the gaps. It also gives us the freedom to focus on what is salient or easily expressed, at the expense of full accuracy. This is the basis for metonymy, in which some entity is invoked by overtly mentioning another, associated entity which calls it to mind (Kövecses and Radden 1998; Langacker 1993; Panther and Radden 2004). Metonymy is utterly pervasive in language. A few examples are given in (6). (6)

a. b. c. d. e.

I’m parked down the street. [I Æ my car] Chicago made the playoffs. [Chicago Æ team from Chicago] She heard a truck. [a truck Æ sound emitted by a truck] The omelet is impatient. [the omelet Æ the customer who ordered the omelet] I phoned my lawyer. [phoned Æ talked to on the phone]

Since language is all about meaning, the findings of cognitive semantics are clearly relevant to language teaching. Most broadly, they show the blatant inadequacy of the conventional CONDUIT metaphor (Reddy 1979) employed in thinking and talking about language itself: meaning is conceived metaphorically as a substance (meaning is a mass noun); words are containers (cf. empty words) which hold only limited quantities of this substance; expressions holding ideas are conveyed from the speaker to the hearer (get the idea across); so understanding an expression is just a matter of opening the containers and combining the thoughts they hold. But it is simply not true that an expression’s meaning is “in” its words. Actually, words are merely prompts for an elaborate process of meaning construction that draws on the full range of our mental resources. An appreciation of the richness and flexibility of these resources would seem essential for effective language instruction, especially at advanced levels. It is also a worthy educational goal, even for first language instruction.


15

More specific pedagogical implications are limited only by the availability of cogent and detailed analyses. For example, Kövecses (2001) has suggested – and to some extent shown empirically – that the learning of idioms is facilitated by apprehension of their metaphorical motivation. Along the same lines, Kurtyka (2001) reports positive results in using cognitive semantic descriptions as a basis for teaching phrasal verbs. It stands to reason that the teaching and/or learning of these verb + particle combinations would be aided by the realization that the choice of particle, rather than being arbitrary, virtually always has a semantic rationale (Lindner 1982; Rudzka-Ostyn 2003). It is likewise pertinent to realize that a particular element does not have a single, fixed meaning, but rather an array of senses related in principled ways to its prototypical value. Consider on. Prototypically it designates a physical relationship in which the trajector is supported by the landmark and in contact with its upper surface (e.g., the cat on the mat). Other uses are then obtained by suspending one or more of its specifications: that the surface be an upper one (the painting on the wall); the notion of support (the spots on that cow); or that of contact (the accent marks on these vowels). Still other uses can be seen as metaphorical applications of these same features (e.g., contact and support for on drugs). The basic point is that conventional usage almost always has conceptual motivation. Though it has to be learned, it represents a particular way of construing the situation described. With proper instruction, the learning of a usage is thus a matter of grasping the semantic “spin” it imposes, a far more natural and enjoyable process than sheer memorization. The pedagogical challenge is then to determine the optimal means of leading students to this understanding (see Boers and Lindstromberg 2006).

3. The meaningfulness of grammar A conceptualist semantics that properly accommodates construal makes possible a symbolic account of grammar. Like lexicon, with which it forms a gradation, grammar reduces to form-meaning pairings. All the elements correctly invoked in grammatical description should thus have semantic import, however schematic they might be in terms of conceptual content. The meaningfulness of grammar must obviously be recognized in devising strategies for teaching it effectively. From the CG perspective, the first order of business in analyzing grammar is to ascertain the meanings of grammatical structures and the elements invoked to describe them. These include both general descriptive

16

Ronald W. Langacker

notions (e.g., noun, verb, subject, object, and clause) and the grammatical formatives (markers or function words) of particular languages. Semantic characterizations are possible when it is recognized that such elements are often polysemous, that their meanings are usually quite abstract, and that meaning resides in how we conceptualize the world (not simply in correspondences to it). These points invalidate standard arguments for the doctrine – found in every linguistics textbook – that basic grammatical notions like noun, verb, subject, and object cannot be semantically characterized. It is presupposed that the only possible characterizations would be notions like “(physical) thing,” “event,” “agent,” and “patient,” pertaining to objective properties of the entities involved. The arguments then consist in showing that such definitions are inappropriate for many instances, e.g., a noun like explosion (which refers to an event) or the subject of a passive (which is nonagentive). But this is comparable to looking for your car keys under the street light instead of where you dropped them. Any general definition would have to be considerably more abstract, and would have to be based on conceptual factors rather than objective properties. The characterizations proposed in CG are based on factors (e.g., the directing of attention) justified independently as being necessary for semantic description. Briefly, an expression’s basic category depends on the nature of the profile it imposes on evoked scenarios (not its overall conceptual content). A verb profiles a process, characterized schematically as a relationship followed sequentially in its evolution through time. Let’s have a look at the following figure: (a) move (verb) tr

lm

(b) mover (noun) tr

lm

Figure 2. Schemas for move and mover

Move is thus a verb, as it profiles the change through time in a spatial relationship. On the other hand, mover is a noun, even though it evokes the same conceptual content. It is a noun because it profiles a thing, specifically the trajector of the verb it derives from.


17

A thing is defined abstractly as any product of two fundamental conceptual operations, namely grouping and reification. (Many nouns designate groups: team, stack, herd, set, convoy, etc. Reification is the treatment of a group as a unitary entity for higher-level cognitive purposes – e.g., in assessing a herd as large, even when all the animals constituting it are small. In the case of physical objects, the prototype for nouns, these operations occur automatically at a low level of processing, leaving us unaware of constitutive entities.) Likewise, the correlation of subject and object with the semantic roles agent and patient, or subject with discourse topic, is at best only prototypical. Schematic characterizations valid for all instances have to be independent of specific conceptual content and discourse status. From the CG standpoint, they find a natural basis in the focal prominence of relational participants: a subject specifies the trajector of a profiled relationship, and an object specifies the landmark. It is further suggested that this primary and secondary focal prominence resides in the participants being invoked as initial and subsequent reference points in building up to the full conception of the profiled relation (Langacker 1993, 1999c, 2001a, 2001b). This dynamic characterization meshes with Chafe’s description of a subject as the “starting point” for apprehending a clause (Chafe 1994).2 The conceptual definability of basic grammatical notions makes evident the meaningfulness of grammatical formatives. Among these are derivational elements, such as -er, which combines with a verb to derive a noun. Since these categories are meaningful, so are elements effecting a change of category membership. What -er contributes semantically to a form like mover is an aspect of construal: the specification that it profiles the verb stem’s trajector – a thing – rather than the overall process it designates. Similarly, any elements affecting the choice of subject or object are meaningful by virtue of conferring focal prominence on particular relational participants. For instance, a passive marker confers trajector status on what would otherwise be a clausal landmark. Conceptual characterizations have been offered in CG for numerous parade examples of “purely grammatical” markers, e.g., do, of, and Spanish se. The auxiliary verb do is semantically equivalent to the schematic description of the verb class: it profiles a process, i.e., a relationship tracked through time. Since it does not specify any particular kind of process, it is used as a “pro form” for clauses, referring back to a previously mentioned process (just as a pronoun refers back to a nominal referent): She does; They did. The preposition of indicates that the relationship between its trajector and landmark is somehow intrinsic rather than contingent (Lan-

18

Ronald W. Langacker

gacker 1992). Thus it is used for part-whole relations (the sole of my foot), for identity (the month of January), for intrinsic functions (the father of the bride as against the bride’s dress), to specify the participants in a reified process (the shooting of the hunters), etc. Note that we say the color of the lawn but the brown spot in (*of) my lawn, the difference being that the spot is not supposed to be there. Another case is the (non-reflexive) se that occurs with many Spanish verbs: sentarse ‘sit,’ caerse ‘fall,’ lavarse ‘wash,’ enojarse ‘get mad,’ ahogarse ‘drown,’ etc. Though polysemous, it has been shown by Maldonado (1988, 1999) to be consistently meaningful. Its various senses are natural extensions from a prototype that is intermediate between a prototypical transitive and a prototypical intransitive (hence the term “middle voice”). While the trajector has patient-like properties, there is also a notion of force or agentivity, without however invoking an agent distinct from the trajector. Grammar consists primarily in patterns for combining simpler expressions into more complex ones. Complex expressions are called constructions, and the patterns they instantiate are constructional schemas. In CG, a construction is simply an assembly of symbolic structures linked by correspondences. Consider expressions where a prepositional phrase modifies a noun: the table by the window, a cabin in the woods, the roof on that house, etc. The schema for such expressions is sketched in Figure 3. N P+N X

N

Y

tr

X

Y

tr

lm P

Figure 3. Schema for the roof on that house

lm

P+N

Y N


19

Two levels of organization are shown. At the lower level, a preposition (P) combines with a nominal expression (N) to form a prepositional phrase (P+N); P and N are component structures at this level, while P+N is the composite structure resulting from their combination. At the higher level of organization, the component structures N and P+N combine to yield the composite structure N P+N. The two nominal expressions profile things (represented as circles), with X and Y abbreviating their additional semantic content. The preposition designates a relationship (shown as an arrow) between two things characterized only schematically. Correspondences, given as dotted lines, indicate how component structures at each level are conceptually integrated to form the composite structure. At the lower level, the profile of the nominal (that house) corresponds to the preposition’s landmark. The composite structure (on that house) results from merging the specifications of the corresponding elements while preserving the profile of the preposition (on). Because it inherits its profile from the preposition (which is thus the head at this level), the composite expression is a prepositional phrase (rather than a nominal). This prepositional phrase then modifies the other nominal element (roof) at the higher level of organization. Here the nominal profile corresponds to the schematic trajector of the prepositional relation. At this level it is the nominal that imposes its profile (making it the head), so the overall expression is nominal rather than relational. Taken as a whole, for example, the roof on that house profiles the roof. The various aspects of construal needed for describing lexical meanings also figure in the description of complex expressions. In particular, profiling and trajector/landmark organization are pivotal in grammatical constructions. On that house is a prepositional phrase because it profiles a relationship rather than a thing, and that house is the prepositional object because its profile corresponds to on’s landmark. Similarly, the overall expression is nominal rather than prepositional because it profiles a thing, and the head noun roof is the subject (as broadly defined in CG) with respect to the prepositional phrase because its profile corresponds to the latter’s trajector. The profiling and correspondences in Figure 3 are specified by the constructional schema describing this grammatical pattern, hence they are characteristic of all the expressions which instantiate it. They constitute the constructional meaning of the pattern, which particular instantiations elaborate based on the lexical elements chosen. Constructional schemas can also be thought of as patterns of semantic composition, since they specify how the meanings of constitutive elements combine to form the meaning of the whole. As viewed in CG, however,

20

Ronald W. Langacker

semantic structure is only partially compositional, since we draw on many resources in arriving at the full composite meaning – it is not true that an expression’s meaning is “in” its words. This is so even for seemingly straightforward examples like the roof on that house. On the most likely interpretation, it is not precisely parallel to the vase on that table, since the vase and the table are distinct objects (hence we cannot say *the vase of that table), whereas the roof and the house are not distinct entities. Based on our knowledge of typical scenarios in everyday life, we presume that the roof is an inherent part of the house, not that the roof from some other building is resting on the house’s own roof. (Given a proper context, e.g., in describing the aftermath of a hurricane, this default assumption may of course be overridden.) Just because the trajector and landmark are nondistinct, this usage of on is non-prototypical. We would normally say the roof of the house, recognizing the intrinsic nature of their part-whole relationship. By using on instead, we highlight the role of the house in supporting the roof, as well as suggesting that their relationship is non-intrinsic (i.e., the roof in question is not the only one the house might have had – it is considered in relation to other possible sorts of roof). On the CG account, therefore, grammar consists in patterns for assembling not just complex expressions, but also complex meanings. It incorporates particular ways of construing conceptual content and symbolizing the construal imposed on it. Besides the sorts of prominence already indicated, the construal embodied in grammar includes the related factors of perspective and sequentiality in the evocation of conceived entities. A local vs. a global perspective was exemplified in (1) [is winding vs. winds]. Another difference in perspective is whether a situation is viewed in terms of interacting participants or in terms of how the scene is sequentially accessed. Illustrating the latter are numerous constructions that first introduce a setting or location and then indicate what is found there. In some of these constructions the setting or location is focused as trajector, making it the grammatical subject: (7)

a. b. c. d.

The streets were lined with spectators. [cf. Spectators lined the streets.] This year has seen some big events. [cf. Some big events have occurred this year.] It appears that she’s quite smart. [cf. She appears to be quite smart.] There are problems with that theory. [cf. That theory has problems.]


21

The setting in (7b) is temporal rather than spatial. The it in (7c) is an abstract setting, something like the scope of awareness for the following judgment (Langacker 2004). In (7d), there is an abstract realm of existence. If not the grammatical subject, a setting or location can nonetheless be initial in terms of word order: (8)

On the mat was a lazy cat. [cf. A lazy cat was on the mat.]

This represents the discourse strategy of first directing attention to a known location as a way of introducing a new participant found there. It exemplifies sequentiality as a dimension of construal. Another, more striking example is the contrast in (9): (9)

a. b.

The cookies are in the pantry, on the bottom shelf, in a plastic container. The cookies are in a plastic container, on the bottom shelf, in the pantry.

Both sentences use the same locative phrases to describe exactly the same spatial configuration. They are nevertheless semantically distinct by virtue of imposing different ways of mentally accessing it: the conceptual experience is that of “zooming in” vs. “zooming out.” This is just one of many grammatical phenomena whose essential import resides in sequence of mental access (Langacker 1993, 2001b). These brief examples merely hint at the conceptual richness embodied and reflected in grammatical structure. The fact that grammar is meaningful, not an autonomous formal system, creates the potential for new and different approaches in teaching and learning it. Being conventionally determined, differently in each language, the proper form of expression does of course require instruction. Learning grammar does not however have to be the soulless internalization of arbitrary restrictions. If properly analyzed, every grammatical element makes a semantic contribution and every grammatical distinction has conceptual import. Awareness of these factors offers a basis for effective language instruction aimed at their full exploitation in thought and communication. How can this be achieved? Not being qualified to make specific pedagogical recommendations, I will limit myself to some general considerations. A point not yet mentioned is the extensive iconicity of grammar (Givón 1991; Haiman 1983, 1985). The symbolic nature of grammar does not imply that the form-meaning pairings are simply arbitrary. On the contrary,

22

Ronald W. Langacker

the symbolization of grammatical notions invariably has some kind or degree of conceptual motivation. Usually, for example, the more tenuous meanings of grammatical elements are paralleled by their more tenuous phonological realizations: the latter tend to be shorter than lexical items, and are often affixes or inflections rather than independent words. A wellknown iconic principle is that elements which belong together semantically generally occur together phonologically. To state this more precisely in CG terms, the conceptual integration effected by grammatical constructions tends to be symbolized by phonological adjacency. With respect to Figure 3, the semantic relationship between a preposition and its object is symbolized by placing the object nominal directly after the preposition. Likewise, a prepositional phrase is placed directly after the nominal expression it modifies to symbolize that relationship. A further dimension of iconicity is the tendency for the sequencing of words to mirror the sequence of events or some other conceptual ordering (Langacker 2001b). In (8), word order implements the discourse strategy of using a known location to anchor the introduction of a new participant. The contrasting orders in (9) represent alternate natural paths for arriving at the full conception of a complex spatial configuration. However it might be exploited pedagogically, the iconicity of grammar is an important factor in its learnability (Langacker 1987a: Section 9.3.2). Cognitive linguistic notions are also potentially helpful in showing connections among the varied uses of a grammatical element or construction. Why should possessives, for instance, be used in expressions like Lincoln’s assassination (since Lincoln was hardly the “owner” of this event, even metaphorically)? The answer lies in the conceptual characterization of ’s and the possessive construction (Langacker 1995b, 2001a; Taylor 1996). Their wide range of conventional usage, exemplified in (10), covers far more than ownership, part-whole, and kinship relationships, which all have some claim to prototypicality: (10)

Sam’s house; my neck; the girl’s mother; our town; the cat’s fleas; your bus; the store’s location; their anxiety; Zelda’s problem; his height; the year’s biggest event; my driving; anyone’s guess; Booth’s assassination [of Lincoln]; Lincoln’s assassination [by Booth]

The prototypical senses are special cases of a schematic characterization based on a mental operation rather than any specific conceptual content. Possessives manifest our capacity for invoking one conceived entity as a


23

reference point in establishing mental contact with another, i.e., for mentally accessing a target via and in relation to the reference point. Out of all the houses in the world, Sam’s house directs attention to a particular instance of that type, namely the one mentally accessible via Sam. This would typically be the house he lives in, but given the proper context it might also be the house he owns and rents out, the one he dreams of owning someday, the one he designed, the one he is scheduled to paint next week, etc. Recall now that the focal prominence of trajector and landmark was ascribed to their functioning as conceptual reference points for purposes of building up to the full conception of a profiled relationship. On this account both the agent and the patient of assassinate are reference points for purposes of conceptualizing (mentally accessing) the profiled event. For this reason either one can be coded as possessor of the derived noun assassination, obtained from assassinate by conceptual reification: Booth’s assassination; Lincoln’s assassination. More generally, it is helpful to realize that grammatical elements of any sort are likely to be polysemous, having a prototypical meaning as well as an array of other, less central values possibly susceptible to schematic characterization. It may not be true that the possessive construction carries the meaning of ownership, but neither is it appropriate to ignore the centrality of this particular kind of reference point relationship (nor that of part-whole and kinship relations). Indeed, full mastery of the construction implies awareness of a wide range of usage possibilities like those in (10), since there is no assurance that every type of usage that conforms to the schematic meaning is conventionally exploited in the language. Similarly, while it is important to realize that nouns are not simply the names for physical objects, and that subjects are agents only prototypically, it is also useful to know that these notions are indeed central to the categories, a conceptual core which is grounded in physical experience and provides a basis for extension to more abstract or less typical values. Finally, grammar is less mysterious when it is recognized that the form of expressions is shaped by an elaborate, multifaceted conceptual substrate. Grammatical peculiarities are commonly the visible trace of tacit, often imaginative mental constructions which are readily grasped by native speakers even when they fail to cross the threshold of explicit awareness. A simple example concerns the prepositional phrase in (11): (11)

It’s pretty through that valley.

24

Ronald W. Langacker

Normally a through-phrase modifies a motion verb and specifies the path traversed (as in (2)). But (11) has no such verb – indeed, it’s pretty describes a static situation. So what does through the valley modify? Instead of modifying an overt element, it itself evokes the conception of someone travelling through the valley and observing its appearance. Moreover, this travel scenario may be entirely fictive. It cannot be identified with ongoing movement by the current speaker (as it might if that were changed to this). Nor is it necessarily based on memory or knowledge of any specific journey. It may be just an imagined journey, on the part of an imagined representative viewer, conjured up as a way of describing the valley. The general point is that an expression’s semantic and grammatical coherence is often dependent on imagined scenarios or other tacit conceptions. As one more example, this is the key to understanding non-present uses of the English present tense (Langacker 2001c), as in (12): (12)

a. b. c.

I come home last night and a stranger opens the door. The children leave for camp next week. A kitten is born with blue eyes.

Inter alia, the so-called present tense is used for past events (the “historical present”), anticipated future occurrences, and “timeless” statements (such as generics). The label is not really a misnomer, however. These uses are based on particular mental constructions in the context of which an instance of the profiled process coincides with the time of speaking. The coincident process is not however actual but rather virtual – a fictive occurrence that serves as a representation of one or more actual ones. In (12a), the profiled events belong to a mental “replay” of a previous episode. Sentence (12b) does not refer directly to the actual event of leaving (as would be the case with the future will); it amounts instead to “reading off” an entry on a conceptual plan or schedule, where each entry represents a planned future occurrence. Invoked in (12c) is something like a “blueprint,” a supposed representation of the world’s essential nature. In this generic expression, the kitten and the process of its being born are virtual entities corresponding to an open-ended set of actual instances. These sentences are in the present tense because the profiled virtual occurrences are part of a mental construction – the replay, the schedule, or the blueprint – immediately available at the time of speaking.


25

4. A usage-based approach A final aspect of CG that makes it relevant for language teaching is its usage-based nature (Barlow and Kemmer 2000; Langacker 2000). A language comprises an immense inventory of conventionally established elements (lexical items, formatives, grammatical constructions, sound patterns, etc.) which fluent speakers learn as units, i.e., they are thoroughly mastered and can thus be employed in largely automatic fashion. Conventional units are abstracted from usage events: actual instances of language use, in their full phonetic detail and contextual understanding. A crucial factor in this process is the reinforcing of features that recur across a sufficient number of such events. Since every usage event is unique at the level of fine-grained detail, the recurring commonalities are apparent only at a certain level of abstraction. To some extent, therefore, all linguistic units – even the most specific – are schematic relative to the usage events in which they figure. Fine-grained differences fail to be reinforced and are therefore filtered out as units emerge. The emphasis is thus on actual learning. Whatever might be its innate basis, mastering a language requires the specific, usage-based learning of a vast array of conventional units. This itself has pedagogical implications (which may seem obvious, but are not so in every linguistic theory). It suggests the importance of providing the learner with sufficient exposure to representative uses of a given unit. Ideally, moreover, this exposure should occur in the context of meaningful exchanges approximating socially and culturally normal usage events. In this respect the usage-based approach resonates with the natural approach to language teaching (Achard 2004). From a usage-based perspective, a basic question is the degree of abstraction relative to usage events. Given that all units derive from such events by schematization, it remains to determine just how schematic they might be. There is no real doubt that linguistic units run the gamut from highly specific to highly schematic. We have noted this in regard to the meanings of lexical items (e.g., do > act > move > go > run > sprint). In CG, grammatical structures are schematic relative to instantiating expressions; the constructional schema in Figure 3, for example, is a schematized representation of particular expressions like a cabin in the woods, the roof on that house, etc. Phonologists find good reason to characterize sounds at different levels of abstraction, for instance [i] < [HIGH VOWEL] < [VOWEL] < [SEGMENT]. In the case of grammar, degree of abstraction figures in several contentious theoretical issues. The first is whether – as claimed in CG – gram-

26

Ronald W. Langacker

matical structures are in fact simply schematized representations of complex expressions. On this account they consist in assemblies of symbolic structures fully reducible to form-meaning pairings. With certain qualifications (Langacker 2005a, 2005b), CG shares this outlook with Construction Grammar (Fillmore, Kay, and O’Connor 1988; Goldberg 1995). It is not however standard linguistic doctrine. A second issue is the relative importance of higher- versus lower-level schemas, and a third is whether we learn both schemas and specific instantiations of them. These latter issues pertain to the abstractness and hence the generality of linguistic knowledge. Theorists have a natural tendency to look for the broadest generalizations, corresponding to highly schematic descriptions. Broad generalizations should of course be sought. Still, it cannot be taken for granted that speakers have the same proclivity. One aspect of the usage-based approach is the notion that speakers rely extensively on low-level schemas and specific learned expressions, even when these conform to general patterns. The learning of specific forms is obviously necessary in cases of irregularity or limited productivity. At the extreme we find the student’s nightmare of complex morphological paradigms that simply have to be memorized, e.g., the conjugations of irregular verbs. For this CG and the usagebased approach have no magic cure. They do however suggest the reasonableness of what usually occurs in practice. What typically does not occur is that a student thoroughly learns all the forms of a complex paradigm – e.g., all the person, number, tense, and mood inflections of an irregular verb – and instantaneously retrieves them as needed in actual use. Instead, students tend to do what children presumably do in learning a language natively: the forms they learn first and learn best are those which occur most frequently. Aiding the learning process are many low-level generalizations (e.g., several verbs forming their past tense in analogous fashion). Eventually an awareness develops of all the dimensions represented in paradigms, patterns (morphological schemas) emerge for constructing any desired regular form, and a large number of irregular forms are learned with different degrees of thoroughness. This is not to say that paradigms should never be studied as such. But that should not be thought of as the primary means of learning what is needed for fluent speech. The opposite extreme, that of full productivity and regularity, is far less prevalent than theoretical attitudes would lead one to expect. Much of our knowledge of grammar resides in intermediate cases: patterns usable with multiple lexical items, perhaps even an open-ended set, but not with every member of a basic category. A well-known construction of this sort is the English ditransitive, where a verb has two object-like complements, e.g.,


27

She gave her brother a watch. Naturally, the verbs occurring in this pattern are limited to those involving three central participants. A further restriction is that the verbal action has a causal relationship in regard to the situation of the first object possessing or having access to the second (e.g., the giving results in her brother having a watch). Observe that this causal relationship may be negative rather than positive (She denied them permission to interview me), and instead of actually bringing about the situation it may represent only a commitment to do so (She promised her brother a watch). Quite a number of verbs satisfy this schematic characterization. Yet not every verb which does is conventionally used in ditransitives. For instance, while get appears in this construction (She got her brother a watch), obtain does not (*She obtained her brother a watch). Also excluded are provide, supply, donate, contribute, deprive, etc. Hence the distributional facts do not lend themselves to a single, fully predictive generalization. But neither is the picture one of randomness or complete idiosyncrasy. If the construction cannot be adequately described by means of a global generalization, promising absolute predictability, it nonetheless exhibits considerable regularity in the form of local generalizations, which offer degrees of motivation for the observed distribution (Goldberg 1992, 1995). In other words, the ditransitive construction represents a complex category whose full description consists in a network of related variants centered on a prototype. The prototypical pattern is for the verb to profile an act of transfer from agent to recipient (e.g., give, send, hand, mail, throw, bring). In another basic pattern, the agent creates an object with the intent of the recipient having access to it (cook, bake, knit, build, make, write). Somewhat more peripheral are verbs of commitment, either positive (promise, owe, permit, allow, guarantee) or negative (refuse, deny). Alternate groupings may of course be proposed, as well as various subgroups. The point, however, is that the distribution is anchored by particular, fairly frequent verbs well-established in the construction. These verbs form clusters on the basis of semantic similarity, and certain clusters give rise to constructional subschemas capable of being used productively. This organization provides the basis for devising pedagogical strategy. The obvious suggestion is to start with the prototype and then move on to other major clusters, in each case focusing initially on the most frequent and basic verbs. Full mastery of the construction, with native-like knowledge of the conventional range of usage, will come about only gradually through long-term practice with the language. But the same is true of its learning by native speakers.

28

Ronald W. Langacker

The ditransitive is not atypical of the pedagogical challenge posed by grammar overall. Fully general and exceptionless rules are themselves the exception. Instead of being monolithic, most constructions exist in multiple variants, with schemas abstracted to represent what is common to different sets of them. At the lowest level are constructional variants that incorporate a particular lexical item well-established in the pattern. Such units thus include both specific and schematic elements, providing one argument against any sharp distinction between lexicon and grammar. Thorough knowledge of a construction resides in the entire network of variants (rather than any single unit), and lower-level structures are often the most important in determining conventional usage. This leads to a final but possibly crucial pedagogical implication of the usage-based approach. A substantial proportion of what is needed to speak a language fluently tends to be ignored because it is part of neither lexicon nor grammar as these are traditionally conceived. What I have in mind are the countless units representing normal ways of saying things. Native speakers control an immense inventory of conventional expressions and patterns of expression enabling them to handle a continuous flow of rapid speech (Langacker 1987a: 35–36, 2001d). While they can certainly be included, I am not referring to lexical items of the sort found in dictionaries, nor even to recognized idioms. At issue instead are particular ways of phrasing certain notions out of all the ways they could in principle be expressed in accordance with the lexicon and grammar of the language. These units can be of any size, ranging from standard collocations to large chunks of boiler-plate language. They can be fully specific or partially schematic, allowing options in certain positions. In fluent speech, piecing together these prefabricated chunks is at least as important as productively invoking lexical and grammatical units. Native speakers cannot avoid using them – here are some examples from the previous paragraph: leads to; a substantial proportion; speak a language fluently; tends to be ignored; as … traditionally conceived; what I have in mind; continuous flow; rapid speech; I am not referring to; nor even; at issue; out of all the …; in principle; in accordance with; of any size; ranging from … to …; large chunks of …; boiler-plate language. Given the prevalence of conventional expressions, as well as their critical role in fluency and idiomaticity, finding effective ways to facilitate their learning would seem essential.


29

5. Conclusion Of necessity (given its source), this discussion has been long on theory and short on practical recommendations. Nor have the sample descriptions been anything more than fragmentary. If we nevertheless suppose that CG concepts and descriptions are relevant for language instruction, a basic question remains to be addressed: Just who are they relevant to? The main possibilities are the student, the instructor, and those responsible for designing curricula or developing teaching materials. Assuming their validity, explicit awareness of CG notions would seem most clearly helpful for the latter. With effective materials and a sensible curriculum, an instructor with lesser awareness of CG insights can nonetheless still exploit them. And students would hopefully benefit even if they are never exposed to theoretical concepts or explicit analyses. I cannot help thinking, however, that the cognitive linguistic view of language is a matter of universal interest, and that its conceptual descriptions of linguistic phenomena are sufficiently natural and revealing to be widely appreciated. In some form, I can imagine these ideas being an integral part of general education or first language instruction. I can further imagine them as being useful in second language learning, especially at more advanced levels.

Notes 1.

2.

This is a slightly changed version of the paper called “Cognitive Grammar as a basis for language instruction” published in 2007 by Lawrence Erlbaum Associates in Handbook of Cognitive Linguistics and Second Language Acquisition, edited by Nick Ellis and Peter Robinson. Used with permission by Lawrence Erlbaum Associates. This characterization of the notions “subject” and “object” is claimed to be valid for all languages. This universalist definition of Cognitive Grammar is contrasted to a language-specific definition in Croft’s Radical Construction Grammar by Broccias in this volume.

References Achard, Michel 2004 Grammatical instruction in the natural approach: A cognitive grammar view. In Cognitive Linguistics, Second Language Acquisition,

30

Ronald W. Langacker

and Foreign Language Teaching, Michel Achard and Susanne Niemeier (eds.), 165–194. (Studies on Language Acquisition 18) Berlin/New York: Mouton de Gruyter. Allwood, Jens 2003 Meaning potentials and context: Some consequences for the analysis of variation in meaning. In Cognitive Approaches to Lexical Semantics, Hubert Cuyckens, René Dirven, and John R. Taylor (eds.), 29– 65. (Cognitive Linguistics Research 23) Berlin/New York: Mouton de Gruyter. Barlow, Michael, and Suzanne Kemmer (eds.) 2000 Usage-based Models of Language. Stanford, California: Center for the Study of Language and Information Publications. Boers, Frank, and Seth Lindstromberg 2006 Cognitive Linguistic applications in second or foreign language instruction: Rationale, proposals and evaluation. In Cognitive Linguistics: Current Applications and Future Perspectives, Gitte Kristiansen, Michel Achard, René Dirven, and Francisco Ruiz de Mendoza Ibáñez (eds.), 305–358. (Applications of Cognitive Linguistics 1) Berlin/New York: Mouton de Gruyter. Brugman, Claudia 1981 The story of over. MA thesis, Department of Linguistics, University of California, Berkeley. Chafe, Wallace 1994 Discourse, Consciousness, and Time: The Flow and Displacement of Conscious Experience in Speaking and Writing. Chicago/London: University of Chicago Press. Danesi, Marcel 1995 Learning and teaching languages: The role of “conceptual fluency.” International Journal of Applied Linguistics 5 (1): 3–20. 1999 Expanding conceptual fluency theory for second-language teaching. Mosaic 6 (4): 16–21. Dapremont, Elena M. 2001 Assembled data, mounting evidence: Conceptual contrasts between past and present participle based change. Paper presented at the 7th International Cognitive Linguistics Conference, Santa Barbara. Ellis, Nick, and Peter Robinson (eds.) 2007 Handbook of Cognitive Linguistics and Second Language Acquisition. Mahwah, New Jersey: Laurence Erlbaum. Fauconnier, Gilles 1985 Mental Spaces: Aspects of Meaning Construction in Natural Language. Cambridge, MA/London: Massachusetts Institute of Technology Press.

The relevance of Cognitive Grammar for language pedagogy 1997

31

Mappings in Thought and Language. Cambridge: Cambridge University Press. Fauconnier, Gilles, and Eve Sweetser (eds.) 1996 Spaces, Worlds, and Grammar. Chicago/London: University of Chicago Press. Fauconnier, Gilles, and Mark Turner 2002 The Way We Think: Conceptual Blending and The Mind’s Hidden Complexities. New York: Basic Books. Fillmore, Charles J., Paul Kay, and Mary C. O’Connor 1988 Regularity and idiomaticity in grammatical constructions: The case of let alone. Language 64 (3): 501–538. Givón, Talmy 1991 Isomorphism in the grammatical code: Cognitive and biological considerations. Studies in Language 15: 85–114. Goldberg, Adele E. 1992 The inherent semantics of argument structure: The case of the English ditransitive construction. Cognitive Linguistics 3: 37–74. Goldberg, Adele E. 1995 Constructions: A Construction Grammar Approach to Argument Structure. Chicago/London: University of Chicago Press. Goldsmith, John, and Erich Woisetschlaeger 1982 The logic of the English progressive. Linguistic Inquiry 13: 79–89. Haiman, John 1980 Dictionaries and encyclopedias. Lingua 50: 329–357. 1983 Iconic and economic motivation. Language 59: 781–819. Haiman, John (ed.) 1985 Iconicity in Syntax. (Typological Studies in Language 6) Amsterdam/Philadelphia: John Benjamins. Kecskes, Istvan, and Isabel M. Cuenca 2005 Lexical choice as a reflection of conceptual fluency. International Journal of Bilingualism 9 (1): 49–67. Kövecses, Zoltán 2001 A cognitive linguistic view of learning idioms in an FLT context. In Applied Cognitive Linguistics II: Language Pedagogy, Martin Pütz, Susanne Niemeier, and René Dirven (eds.), 87–115. (Cognitive Linguistics Research 19.2) Berlin/New York: Mouton de Gruyter. 2005 Metaphor in Culture: Universality and Variation. Cambridge: Cambridge University Press. Kövecses, Zoltán, and Günter Radden 1998 Metonymy: Developing a cognitive linguistic view. Cognitive Linguistics 9: 37–77.

32

Ronald W. Langacker

Kurtyka, Andrzej 2001 Teaching English phrasal verbs: A cognitive approach. In Applied Cognitive Linguistics II: Language Pedagogy, Martin Pütz, Susanne Niemeier, and René Dirven (eds.), 29–54. (Cognitive Linguistics Research 19.2) Berlin/New York: Mouton de Gruyter. Lakoff, George 1987 Women, Fire, and Dangerous things: What Categories Reveal about the Mind. Chicago/London: University of Chicago Press. Lakoff, George, and Mark Johnson 1980 Metaphors We Live By. Chicago/London: University of Chicago Press. Lakoff, George, and Rafael E. Núñez 2000 Where Mathematics Comes From: How the Embodied Mind Brings Mathematics into Being. New York: Basic Books. Langacker, Ronald W. 1986 Abstract motion. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 12: 455–471. 1987a Foundations of Cognitive Grammar. Vol. 1: Theoretical Prerequisites. Stanford, California: Stanford University Press. 1987b Nouns and verbs. Language 63: 53–94. 1990 Concept, Image, and Symbol: The Cognitive Basis of Grammar. (Cognitive Linguistics Research 1) Berlin/New York: Mouton de Gruyter. 1991 Foundations of Cognitive Grammar. Vol. 2, Descriptive Application. Stanford, California: Stanford University Press. 1992 The symbolic nature of cognitive grammar: The meaning of of and of of-periphrasis. In Thirty Years of Linguistic Evolution: Studies in Honour of René Dirven on the Occasion of His Sixtieth Birthday, Martin Pütz (ed.), 483–502. Philadelphia/Amsterdam: John Benjamins. 1993 Reference-point constructions. Cognitive Linguistics 4: 1–38. 1995a Raising and transparency. Language 71: 1–62. 1995b Possession and possessive constructions. In Language and the Cognitive Construal of the World, John R. Taylor and Robert E. MacLaury (eds.), 51–79. (Trends in Linguistics Studies and Monographs 82) Berlin/New York: Mouton de Gruyter. 1999a Grammar and Conceptualization. (Cognitive Linguistics Research 14) Berlin/New York: Mouton de Gruyter. 1999b Virtual reality. Studies in the Linguistic Sciences 29 (2): 77–103. 1999c Assessing the cognitive linguistic enterprise. In Cognitive Linguistics: Foundations, Scope, and Methodology, Theo Janssen and Gisela Redeker (eds.), 13–59. (Cognitive Linguistics Research 15) Berlin/New York: Mouton de Gruyter.

The relevance of Cognitive Grammar for language pedagogy 2000

33

A dynamic usage-based model. In Usage-based Models of Language, Michael Barlow, and Suzanne Kemmer (eds.), 1–63. Stanford, California: Center for the Study of Language and Information Publications. 2001a Topic, subject, and possessor. In A Cognitive Approach to The Verb: Morphological and Constructional Perspectives, Hanne G. Simonsen and Rolf T. Endresen (eds.), 11–48. (Cognitive Linguistics Research 16) Berlin/New York: Mouton de Gruyter. 2001b Dynamicity in grammar. Axiomathes 12: 7–33. 2001c The English present tense. English Language and Linguistics 5: 251–271. 2001d Cognitive linguistics, language pedagogy, and the English present tense. In Applied Cognitive Linguistics I: Theory and Language Acquisition, Martin Pütz, Susanne Niemeier, and René Dirven (eds.), 3– 39. (Cognitive Linguistics Research 19.1) Berlin/New York: Mouton de Gruyter. 2003 Context, cognition, and semantics: A unified dynamic approach. In Job 28: Cognition in Context, Ellen van Wolde (ed.), 179–230. (Biblical Interpretation Series 64) Leiden/Boston: Brill. 2004 Aspects of the grammar of finite clauses. In Language, Culture and Mind, Michel Achard, and Suzanne Kemmer (eds.), 535–577. Stanford, California: Center for the Study of Language and Information Publications. 2005a Construction grammars: Cognitive, radical, and less so. In Cognitive Linguistics: Internal Dynamics and Interdisciplinary Interaction, Francisco J. Ruiz de Mendoza Ibáñez, and Sandra Peña Cervel (eds.), 101–159. (Cognitive Linguistics Research 32) Berlin/New York: Mouton de Gruyter. 2005b Integration, grammaticization, and constructional meaning. In Grammatical Constructions: Back to the Roots, Mirjam Fried and Hans C. Boas (eds.), 157–189. (Constructional Approaches to Language 4) Amsterdam/Philadelphia: John Benjamins. 2006 On the continuous debate about discreteness. Cognitive Linguistics 17: 107–151. Lindner, Susan 1982 What goes up doesn’t necessarily come down: The ins and outs of opposites. Papers from the Regional Meeting of the Chicago Linguistic Society 18: 305–323. Maldonado, Ricardo 1988 Energetic reflexives in Spanish. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 14: 153–165. 1999 A media voz: Problemas conceptuales del clítico se. (Publicaciones del Centro de Lingüística Hispánica 46.) Mexico City: Universidad

34

Ronald W. Langacker

Nacional Autónoma de México, Instituto de Investigaciones Filológicas. Matlock, Teenie 2001 How real is fictive motion? PhD dissertation, Department of Psychology, University of California, Santa Cruz. 2004 Fictive motion as cognitive simulation. Memory and Cognition 32: 1389–1400. Matlock, Teenie, and Daniel C. Richardson 2004 Do eye movements go with fictive motion? Proceedings of the Annual Conference of the Cognitive Science Society 26: 909–914. Matsumoto, Yo 1996a Subjective motion and English and Japanese verbs. Cognitive Linguistics 7: 183–226. 1996b Subjective-change expressions in Japanese and their cognitive and linguistic bases. In Spaces, Worlds, and Grammar, Gilles Fauconnier, and Eve Sweetser (eds.), 124–156. Chicago/London: University of Chicago Press. Panther, Klaus-Uwe, and Günter Radden (eds.) 2004 Metonymy in Language and Thought. (Human Cognitive Processing 4) Amsterdam/Philadelphia: John Benjamins. Reddy, Michael J. 1979 The conduit metaphor: A case of frame conflict in our language about language. In Metaphor and Thought, Andrew Ortony (ed.), 284–324. Cambridge: Cambridge University Press. Rosch, Eleonor 1978 Principles of categorization. In Cognition and Categorization, Eleonor Rosch and Barbara B. Lloyd (eds.), 27–47. Hillsdale, NJ: Erlbaum. Rudzka-Ostyn, Brygida 2003 Word Power: Phrasal Verbs and Compounds. A Cognitive Approach. Berlin/New York: Mouton de Gruyter. Sandra, Dominiek, and Sally Rice 1995 Network analyses of prepositional meaning: Mirroring whose mind – the linguist’s or the language user’s? Cognitive Linguistics 6: 89– 130. Sweetser, Eve 1997 Role and individual interpretations of change predicates. In Language and Conceptualization, Jan Nuyts and Eric Pederson (eds.), 116–136. (Language, Culture and Cognition 1) Cambridge: Cambridge University Press. Talmy, Leonard 1996 Fictive motion in language and “ception.” In Language and Space, Paul Bloom, Mary A. Peterson, Lynn Nadel, and Merrill F. Garrett,


35

(eds.), 211–276. Cambridge, MA/London: Massachusetts Institute of Technology Press. Taylor, John R. 1996 Possessives in English: An Exploration in Cognitive Grammar. Oxford: Oxford University Press. 2004 Linguistic Categorization: Prototypes in Linguistic Theory. 3rd ed. Oxford: Oxford University Press. Tuggy, David 2003 The Nawatl verb kiisa: A case study in polysemy. In Cognitive Approaches to Lexical Semantics, Hubert Cuyckens, René Dirven, and John R. Taylor (eds.), 323–362. (Cognitive Linguistics Research 23) Berlin/New York: Mouton de Gruyter. Turner, Mark 1987 Death is the Mother of Beauty: Mind, Metaphor, Criticism. Chicago: University of Chicago Press. Tyler, Andrea, and Vyvyan Evans 2003 The Semantics of English Prepositions: Spatial Scenes, Embodied Meaning and Cognition. Cambridge: Cambridge University Press. Wierzbicka, Anna 1995 Dictionaries vs. encyclopaedias: How to draw the line. In Alternative Linguistics: Descriptive and Theoretical Modes, Philip W. Davis (ed.), 289–315. (Current Issues in Linguistic Theory 102) Amsterdam/Philadelphia: John Benjamins. Zlatev, Jordan 2003 Polysemy or generality? Mu. In Cognitive Approaches to Lexical Semantics, Hubert Cuyckens, René Dirven, and John R. Taylor (eds.), 447–494. (Cognitive Linguistics Research 23) Berlin/New York: Mouton de Gruyter.

Some pedagogical implications of cognitive linguistics1 John R. Taylor

Abstract This paper considers the possible contribution of cognitive linguistics – more specifically, cognitive grammar – to the contents of a pedagogical grammar. The nature and purpose of a pedagogical grammar is briefly discussed, and some salient aspects of cognitive linguistics are reviewed, with an eye on the potential relevance of these aspects to the pedagogical presentation of the grammar of a foreign language. Two areas of English grammar are considered from a cognitive-pedagogical perspective, namely: the choice of complement after a matrix verb, and the status of a noun as count or mass. These are both topics which are known to cause problems for many learners of English. The aim, in discussing these topics, is not to give a definitive account, but rather to highlight issues of principle and methodology, concerning the choices and decisions that a writer of a pedagogical grammar needs to make, and the criteria for making them. Keywords: pedagogical grammar; symbolic unit; schema; prototype; motivation; explanation; rules; units; grammatical constructions; schematicity; complementation; count noun; mass noun; imagery; conceptualization

1. Introduction Any major innovation in linguistic theory is bound, sooner or later, to have an impact on the language teaching profession. No doubt – and this is true especially if we take a wide-angle view of the history of linguistics – the innovative character of cognitive linguistics is less dramatic than some of its current exponents would have us believe. Even so, there is little doubt that the work of Lakoff, Langacker, Talmy and many others does represent a major break with the transformational-generative paradigm which has dominated mainstream academic linguistics during the past half century. Now, as the cognitive linguistics movement grows in strength and selfconfidence, it is only natural that scholars should be turning to the possible pedagogical applications of the approach.2

38

John R. Taylor

This paper had its origins in the author’s involvement in early work on a pedagogical grammar of English on cognitive linguistic principles,3 and addresses the question of how the insights of cognitive linguistics can be exploited in the presentation of grammar in foreign language pedagogy. First, I give a brief characterization of the nature of pedagogical grammar and its role in language pedagogy. I then go on to review some salient aspects of cognitive linguistics, with an eye on the potential relevance of these aspects to the pedagogical presentation of the grammar of a foreign language. Two areas of English grammar are considered from a cognitivepedagogical perspective, namely: the choice of complement after a matrix verb, and the status of a noun as count or mass. These are both topics which are known to cause problems for many learners of English. My account of these topics is in no way meant to be definitive or comprehensive. My aim, rather, in discussing these topics, is to focus on broader issues of principle and methodology, namely, the choices and decisions that a writer of a pedagogical grammar needs to make, and the criteria for making them.

2. Pedagogical grammar A pedagogical grammar may be characterized as a description of a language which is aimed at the foreign language learner and/or teacher, and whose purpose is to promote insight into, and thereby to facilitate the acquisition of, the foreign language.4 Given this characterization, it will be apparent that a pedagogical grammar differs fundamentally from a descriptive, or linguistic grammar (Dirven 1985). Linguistic grammars (or grammar fragments) are written by linguists, for fellow linguists, and are evaluated against the demands of linguistic theory. On the other hand, a pedagogical grammar is written to meet the needs of the language learner and/or teacher, and is evaluated by its success in promoting insight into, and acquisition of, the foreign language. In accordance with its purpose, a pedagogical grammar will differ from a linguistic grammar with regard to both content and presentation. Obviously, to reach its intended audience, a pedagogical grammar will make use only of concepts and terminology that are easily accessible to the linguistically naive reader. But a pedagogical grammar is not simply a linguistic grammar whose terminology has been simplified. Chomsky (1986: 6) observes that the concerns of linguistic and pedagogical grammar are “in a certain sense, complementary.” A pedagogical grammar will focus of necessity on learning problems, i.e., in the main, on what is “idiosyncratic” or

Some pedagogical implications of cognitive linguistics

39

“parochial” in a language, rather than on those aspects of general crosslinguistic validity. Yet a pedagogical grammar need not reduce to a listing of language-specific facts. On the contrary, a pedagogical grammar will strive to present even the idiosyncratic and language-particular as coherent and systematic. This is to be achieved, firstly, through the grading and judicious ordering of the different components of the grammar, and through the highlighting of interrelationships among the various subsystems. Of no less importance is the need to offer explanations – explanations that are at once succinct, readily comprehensible, and intuitively plausible – as to why the foreign language should be as it is. Explanations constitute a powerful promoter of insight, and without insight, learning can scarcely progress beyond rote memorization. Against this broad characterization of a pedagogical grammar, we may draw some further distinctions, according to the proposed context of use of the grammar (Greenbaum 1987). A pedagogical grammar might be aimed primarily at the teacher trainee, at the practising teacher, or at the course or syllabus writer. Alternatively, a pedagogical grammar might have as intended audience the learners themselves, i.e., the grammar might be an intrinsic component of a set of teaching materials, or it might serve as a reference work for the intermediate to advanced student. It has been suggested (Graustein and Leitner 1989) that the term “didactic grammar” should be reserved for this kind of pedagogical grammar. The remarks that follow concern pedagogical grammar in the broader sense, that is, as a description of the language on which language teaching professionals (whether teachers, materials writers, or others) may draw in their design of didactic materials.

3. The need for pedagogical grammar The role of “grammar teaching” in foreign language pedagogy is controversial. While little, probably, is to be achieved by the rote learning of grammar rules, there is a growing consensus that a “focus on form,” and the “consciousness raising” which it entails, does promote the learning process (Sharwood Smith 1981; Rutherford 1987; Ellis 2001). Doubts in this regard come from the theory of language acquisition proposed by Stephen Krashen in numerous publications (e.g., Krashen 1981, 1982; Dulay, Burt and Krashen 1982; Krashen and Terrell 1983). I will briefly consider these here.

40

John R. Taylor

Fundamental to Krashen’s theory is a distinction between conscious “learning” and unconscious “acquisition.” Acquisition, it is claimed, takes place under conditions of communicative language activity, and is driven by the operation of the “Language Acquisition Device,” with which both adults and children are said to be endowed. A further claim is that the course of acquisition is relatively constant across individuals, the only barrier to the attainment of full mastery being the “affective filter,” which for various reasons can block the “intake” of relevant aspects of language “input.” In this scheme of things, conscious learning has only a very limited role to play. In fact, conscious learning will only come to fruition in those rare circumstances in which a learner has the time and the incentive to “focus on form.” Crucially, it is claimed that the conscious application of a learned rule cannot cause learning to “turn into” acquisition. Consequently, the main, indeed the only duty of the language teacher is to set up conditions which will promote acquisition. If the teacher is successful, the need for learning falls away. So too does the need for pedagogical grammar, since the raison d’être of a pedagogical grammar is to raise the learner’s consciousness of the structural properties of the target language. Krashen’s theory needs to be evaluated with circumspection. Certain aspects are surely uncontroversial. No one, presumably, would query the notion that a good deal of language learning is unconscious, that affective barriers hinder learning, that a necessary (although not perhaps sufficient) precondition for learning is comprehensible input, and that important duties of a language teacher are to provide that input and to reduce affective barriers. But many of Krashen’s assumptions and theoretical constructs have not gone unchallenged (Munsell and Carr 1981; Gregg 1984). Consider the hypothesis that learning does not “turn into” acquisition. If true, this hypothesis would leave no room for the conscious study of grammatical structure (or, for that matter, for the conscious study of anything in a language, not even words and their meanings). It certainly happens – as Krashen notes in support of his hypothesis – that learners often continue to make errors in spite of conscious knowledge of the rules they are violating. Yet, equally, learners sometimes do achieve error-free mastery of a form subsequent to conscious study and practice. Even adult native speakers extend their competence in this way – by looking up unknown words in a dictionary! Indeed, there is now empirical evidence that directing the learner’s attention to matters of form can significantly enhance the learning process (Ellis 2001). One could, no doubt, argue that in such cases acquisition has taken place in spite of conscious learning. Yet such a rejoinder merely renders Krashen’s hypothesis unfalsifiable, and therefore vacuous.


41

And even if we were to accept the general thrust of Krashen’s position, it is by no means clear why the output of conscious rule applications, as well as the interrelationships highlighted by a pedagogical grammar and the arrays of illustrative examples with which grammar rules are explicated, should not themselves count as comprehensible input to the Language Acquisition Device.5 This kind of concentrated input is likely to be at a premium precisely in those situations where the learner has only limited access to foreign language data, e.g., in the foreign language classroom. Krashen’s distinction between learning and acquisition thus turns out to be anything but clear-cut. Equally suspect is the distinction between two kinds of learner activity, viz., “form-focused” activities and “meaningfocused” (or “content-focused”) activities, whereby focus on form typifies the conscious attention to learned rules, while focus on content characterizes the spontaneous, interactive, communicative use of language. The very fact that Krashen (and his sympathizers) can invoke such a dichotomy would appear to rest, not just on the postulated learning/acquisition distinction, but also on a highly impoverished understanding of what constitutes grammar. To judge from the repeated citing of “morpheme studies” (i.e., studies of the acquisition of third person singular verb inflection, and the like) and the (by their own admission “excruciatingly boring”) formfocused exercises reviewed in Krashen and Terrell (1983: 142), grammar, for Krashen, would seem to comprise little other than inflectional morphology (of which there is relatively little in English) and the formation of negative and interrogative sentences. Few linguists, of any theoretical persuasion, would subscribe to such a narrow view. Certainly, cognitive linguists would fully endorse the following remarks in Gregg’s critique of Krashen: “focusing on form is focusing on meaning” and “focus on form ... boils down to trying to say what one means to say” (Gregg 1984: 83). The import of these remarks to the cognitive linguistic program will become clear in Section 6 of this paper, where we examine some of the basic theses of cognitive grammar.

4. Explanation in grammar According to Wierzbicka (1988: 491), a dominant characteristic of linguistics in the last quarter of the twentieth century was “a new emphasis on the non-arbitrariness of grammar.” By “the non-arbitrariness of grammar” is meant, above all, the thesis that syntax is motivated by semantics. Among the exponents of this thesis are, in addition to Wierzbicka herself, such

42

John R. Taylor

linguists as Dwight Bolinger, John Haiman, and Robert Dixon. Bolinger (1977) argues that different wordings always correlate with semantic differences; for Haiman (1985) the surface form of a sentence iconically “diagrams” its semantics; while Dixon (1982: 8) states that his own work has been informed by “the assumption that the syntactic properties of a lexical item can largely be predicted from its semantic description.” Cognitive Grammar, as developed in the 1980’s by Ronald Langacker (1987a), George Lakoff (1987), and others (see, e.g., the contributions in RudzkaOstyn 1988; for more recent overviews, see Taylor 2002; Langacker 2004; Dirven 2005), needs to be seen in this context. Thus, Langacker (1987a: 12) writes that “grammar is simply the structuring and symbolization of semantic content,” while for Lakoff (1987: 491) one objective of the cognitive programme is to “show how aspects of form can follow from aspects of meaning.” The thesis of the non-arbitrariness of syntax is in polemical opposition to some major assumptions of Chomskyan linguistics, as well as to postBloomfieldian structuralism, out of which Chomskyan linguistics developed. The Chomskyan paradigm is built on the twin assumptions of modularity and autonomy. The language faculty is construed as a module of the mind independent of (though interacting with) other mental modules, e.g., conceptual knowledge and pragmatic competence. In other words, language competence is autonomous of a person’s other cognitive abilities and social skills. Further, knowledge of language is itself viewed as a system of autonomous modules. Thus, in the Aspects model (Chomsky 1965), syntax is autonomous of semantics and phonology, in the sense that syntactic rules operate on strings of symbols which lack phonological and semantic content. In subsequent years there was, it is true, a considerable shift of emphasis in Chomsky’s thought, in that syntax came to be seen as resulting from the “projection” of properties of lexical items (Chomsky 1986: 84ff). At first sight, the projection principle might be seen as a move towards the position espoused by Dixon. But this rapprochement is only apparent. Consider, as a case in point, some facts about verb complementation: (1)

a. b. c.

I believe that I am right *I believe to be right. *I believe being right.


(2)

a. b. c.

*I want that I leave now. I want to leave now. *I want leaving now.

(3)

a. b. c.

*I considered that I leave tonight. *I considered to leave tonight. I considered leaving tonight.

43

Within the Chomskyan paradigm, these grammaticality judgments follow from the fact that the three verbs believe, want, and consider subcategorize for different kinds of complement, believe for a that-clause, want for a (subjectless) to-infinitive, and consider for a gerundial -ing complement. Cook (1988: 10), in his introduction to Government and Binding theory, stresses that subcategorization frames constitute arbitrary syntactic facts about lexical items. In contrast, the thesis of the non-arbitrariness of syntax demands a semantic explanation for the data in (1)–(3). What form might such a semantic explanation take? And what should be our methodology for developing and evaluating such explanations? The approach that I would suggest involves the following steps: 1. identify the items whose co-occurrence restrictions are up for explanation; 2. propose semantic representations for the items in question; 3. assess the semantic (in)compatibility of items which are (in)eligible to combine, with a view to explaining (un)acceptable combinations in terms of the semantic (in)compatibility of the items; 4. return, if necessary, to 2, and refine the proposed semantic representations. In the case in point, we would need to propose semantic values for the three complementation patterns, namely a that-clause, a subjectless toinfinitive, and a gerundial. Inevitably, the semantic values attributed to these structures will be highly schematic, in that they will abstract away from the lexical material of any particular instantiation. Once we have attempted this, we would try to show how the grammaticality (or, as the case may be, the ungrammaticality) of a sentence follows from the compatibility (or incompatibility) of the complement meaning with the meaning of the main verb. Pursuing the matter further, we would test our proposals for the semantic values of the complementation patterns by considering a wider range of complement-taking verbs, in the hope that we

44

John R. Taylor

can identify the semantic communality of infinitive-taking verbs, clausetaking verbs, and gerundial-taking verbs. It will be apparent that the semantics-based approach has important implications for the pedagogical presentation of the facts in question, quite different from those of the autonomous syntax approach. The autonomous syntax position suggests that we simply make an exhaustive listing of verbal predicates according to their subcategorization patterns. According to the theory, these would constitute arbitrary facts about the language and would “just have to be learned.” The semantics-based approach, on the other hand, would attempt to provide a semantic characterization of the various categories, such that learners would be in a position, if not to predict, then at least to see some systematicity in the presented facts. A semantics-based approach is not without difficulties. The main hurdle consists, precisely, in finding schematic meanings for general syntactic patterns, such as an infinitival complement. As mentioned, the meanings are likely to be highly schematic, and not easily read off from individual examples. The matter is also complicated by the fact that the semantic value of a schematic unit may not be unitary, i.e., it might exhibit polysemy, to a greater or lesser extent. We should also not discount the possibility that a small residue of examples may be truly idiosyncratic, and fall outside the purview of any generalization. Nevertheless, there are compelling theoretical and empirical reasons why we should pursue such an approach. Perhaps the most important of these is that speakers often have a choice with respect to the phenomena under consideration. Many complement-taking verbs are in fact compatible with more than one complement type. The choice of one complement type rather than another goes with sometimes very subtle semantic differences in the resulting sentences. Compare: (4)

a. b.

I propose to leave tonight. I propose leaving tonight.

In (4a) the speaker expresses the intention to leave, while in (4b) s/he is putting forward, for consideration, the activity of leaving. Thus, we should probably infer from (4a) that it is the speaker who will leave, whereas in (4b) it is not clear who is to leave (perhaps it is the speaker, or a group which includes the speaker, or perhaps other people). A mere listing of predicates according to the complements that they take would be unable to accommodate the fact that the choice of complement may be driven by semantic considerations. As a matter of fact, minimal pair contrasts, such


45

as that in (4), are likely to be especially useful in the identification of the contrasting semantic values. Let us consider another example in light of the above remarks, namely, the status of, e.g., furniture and information as mass nouns in English.6 Here, we are not dealing with the syntagmatic combination of items, but with the status of an item vis-à-vis a morphosyntactic category. On the autonomous syntax position, membership in the categories [COUNT NOUN] and [MASS NOUN] is essentially a matter of arbitrary stipulation, that is, we would simply tag each noun in the lexicon with the feature [+COUNT] or [+MASS]. A semantic-based approach, on the other hand, would search for a schematic meaning of the two noun classes, and assess the status of any given noun with respect to its compatibility with the schematic values. As was the case with complementation patterns, a compelling reason to pursue the semantics-based approach is the fact that it is descriptively inadequate to categorize each and every noun as either count or mass. Most nouns can be used both as count nouns and as mass nouns, with accompanying semantic distinctions.

5. Some themes in cognitive grammar Having positioned cognitive linguistics within the context of some recent trends in linguistic theory, I now turn to some specific claims of cognitive grammar, drawing attention to their relevance to pedagogical issues.

5.1. Symbolic units The basic construct in cognitive grammar – just as it was for Saussure – is the linguistic sign, or “symbolic unit.” For Saussure, the linguistic sign was the association of a “concept” with an “acoustic image;” similarly, for Langacker (1987a: 11), the symbolic unit “associates a semantic representation of some kind with a phonological representation.” In fact, for Langacker, a language can be characterized simply as an open-ended (yet structured) inventory of symbolic units. We need to elaborate on several details of the above claim. First, what counts as a symbolic unit? Saussure himself had explicated his notion of linguistic sign on the example of the lexical item arbre ‘tree.’ And indeed, the Saussurian sign is often popularly identified with the morphemes and words of a language.7 Langacker proposes to extend the inventory of sym-

46

John R. Taylor

bolic units to comprise, not only morphemes and words, but also units larger than words (that is, constructions), as well as units which are schematic for words and morphemes, and for the manner in which they can combine. Symbolic unit status is therefore attributed to: — morphosyntactic categories, e.g., [PAST TENSE], [PLURAL]; — lexical categories, e.g., [NOUN], [COUNT NOUN], [VERB]; — syntactic constructions of varying sizes and complexity, e.g.,[TRANSITIVE CLAUSE], [THAT-CLAUSE], [NOUN PHRASE]; — instances of syntactic constructions, e.g., actual noun phrases, thatclauses; — idioms and formulaic phrases, of varying degrees of internal complexity and productivity. Symbolic units vary with respect to their abstractness, or, in Langacker’s terminology, their “schematicity.” A schema is described as “an abstract characterization that is fully compatible with all the members of the category it defines ...; it is an integrated structure that embodies the commonality of its members, which are conceptions of greater specificity and detail that elaborate the schema in contrasting ways” (Langacker 1987a: 371). Thus, the symbolic unit [TREE] is an instance of the symbolic unit [COUNT NOUN]; conversely, [COUNT NOUN] is schematic for the symbolic unit [TREE] (as it is for all other count nouns in the lexicon). Similarly, the symbolic unit [NOUN] is schematic for both [COUNT NOUN] and [MASS NOUN], while these latter instantiate the more abstract unit [NOUN]. Very importantly, grammatical constructions are regarded as schemas for integrating two or more simpler units into a more complex, composite unit. For example, the expression the tree instantiates the schema [DET N], which in turn is an instantiation of the symbolic unit [NOUN PHRASE]. [DET N] has the status of a noun-phrase construction, which provides a pattern for the syntagmatic combination of component signs to form a more elaborate structure. Schematicity is one important principle according to which the inventory of symbolic units of a language is structured. The inventory is open-ended as a consequence of the fact that grammatical constructions are “productive,” i.e., their instantiations do not have to be learned individually (although they may be, as in the case of certain frequently used expressions), but can be created from a knowledge of the construction schema and of the symbolic units that can serve as its constituents. It follows from the above account that a study of the forms of a language cannot legitimately be divorced from a study of the meanings which


47

they symbolize. For if we extend the notion of linguistic sign to comprise, not only the words and morphemes of a language, but also lexical categories and syntactic constructions, we are under the obligation to provide characterizations of the semantic content of these units. Obviously, the semantic content of these highly schematic units will itself be highly schematic. Consider again the examples mentioned in the previous section. Furniture is a mass noun in English, which is to say that furniture instantiates the symbolic unit [MASS NOUN], and so, at its semantic pole, elaborates with greater specificity the semantic content of [MASS NOUN]. Obviously, the account would be incomplete without a statement of the semantic content of [MASS NOUN]. Similarly, the inability of believe, as used in (1), to take a subjectless to-infinitive invites an explanation in terms of the semantic incompatibility of believe with this kind of complement structure. Again, this account places us under the obligation to spell out the semantic content of a to-infinitive, in contrast to, say, a gerundial or a that-clause.

5.2. Schemas and prototypes Alongside the relationship of schematicity, Langacker also recognizes the role of prototypes in the structuring of the symbolic units of a language. Whereas a schema is an abstract characterization that is fully compatible with all its instances, a prototype is a typical instance, and other elements get assimilated to a category in virtue of some kind of similarity to its prototype (Langacker 1987a: 371). Categorization by prototype and categorization by schema are not in principle incompatible. For instance, a prototype and its more peripheral exemplars might still be compatible with an abstract, schematic representation (Langacker 1987a: 136ff.). More complex categories (e.g., the “radial categories” discussed at length by Lakoff 1987 and Taylor 1989) will typically be structured by relationships of both schematicity and prototypicality. The representation of the prototype will itself be schematic to some degree, while an element associated, through similarity, with the prototype could itself function as a prototype to which further elements get associated. Alternatively, an element peripheral to the central prototype could have the status of a subschema vis-à-vis its more specific instantiations. Both schematic and prototype characterizations have their place in a pedagogical grammar. The value of a schematic characterization lies in the fact that it makes possible a concise statement of the conceptual unity of a

48

John R. Taylor

category. We may illustrate on the example of the to-infinitive. In Dirven (1989a; see also Taylor and Dirven 1991), it is proposed that the schematic meaning of a to-infinitive is to denote an instance, or series of instances, of a situation, in contrast with a gerundial complement, which denotes merely a kind of situation. These schematic characterizations make possible an explanation of: — the use of the infinitive after various kinds of predicates, e.g., predicates which denote a desire to bring about a new situation (want, intend, mean, etc.), predicates which denote an effort leading to an accomplishment (manage, try, strive), and predicates of influence and indirect causation (persuade, ask, get); — the inappropriateness of a to-infinitive with other kinds of predicate, e.g., those which denote a psychological experience (such as enjoy), or an attitude to a proposition (such as believe and doubt); — the semantic contrast between a to-infinitive and a gerundial after certain predicates. Consider, as illustration of the last point, the following examples with propose, be afraid, and like: (5)

a. b.

I propose to go there tomorrow. [A specific event is proposed.] I propose going there tomorrow. [A kind of event is proposed.]

(6)

a. b.

I’m afraid to go there alone. [on a particular occasion] I’m afraid of going there alone. [in general]

(7)

a.

I like to go to the cinema on Saturday evenings. [Reference is to an indefinite number of events.] I like going to the cinema on Saturday evenings. [Reference is to a kind of event.]

b.

In (5) and (6), the examples with an infinitive refer to a specific event, on a specific occasion, whereas the gerundial designates a more general situation. The contrast in (7) is more subtle. The form I like to go generalizes over an indefinite number of specific events, whereas I like going expresses an attitude to a kind of event. Note that the form I would like


49

strongly prefers8 the infinitive. This is because I would like expresses a preference for a specific activity, not a kind of activity. (8)

a. b.

I would like to go to the cinema tonight. *I would like going to the cinema tonight.

It is also worth observing that the proposed characterizations of the toinfinitive and the gerundial link up with other facts about the English language. It is not an accident that an infinitive makes use of a morpheme (to) which is homophonous with a directional preposition (as in I went to the bank) (the two are in fact historically related). The preposition denotes motion along a path which terminates at a specific goal. The infinitive complement likewise designates, in many cases, a psychological goal, that is, the intended result of a decision. Accordingly, verbs which imply, or entail, the achievement of a goal typically take a to-infinitive. Concerning the gerundial complement, it is to be noted that a V-ing form can also be used as an abstract noun, designating a kind of activity, as in Smoking is bad for your health. What is referred to here is not any particular event of smoking, but simply the activity as such. Hence, verbs which designate an attitude towards, or an assessment of, a kind of activity typically take a gerundial complement.9 The search for an overarching schematic characterization may sometimes be of more limited pedagogical value. Consider the case of mass and count nouns. Langacker (1987b) has proposed schematic characterizations of [MASS NOUN] and [COUNT NOUN]: a count noun designates a “bounded region in a domain” while a mass noun designates an “unbounded region in a domain.”10 These characterizations are of limited pedagogical value for two reasons. Firstly, the characterizations appeal to a technical concept (“region in a domain”) which can only be understood within the context of Langacker’s more general semantic theory, and are thus not likely to be transparent to the intended user of a pedagogical grammar. More seriously, perhaps, is the fact that the characterizations do not provide a basis for predicting membership in the two categories (Taylor 1990). Langacker’s schemas for [COUNT NOUN] and [MASS NOUN] would probably be valid, with little or no modification, for those languages of the world which make a grammatical distinction between count and mass nouns. Yet membership in the categories varies considerably from language to language.11 For example, the translation equivalent of information in many languages is a count noun, a consequence of this fact being that errors such as *an information are highly “fossilized” (Selinker 1972) elements in the speech even

50

John R. Taylor

of quite proficient learners of English. To account for the idiosyncratic specifics of category membership, we need to appeal to categorization by prototype. Thus, we may propose, as the prototypes of [COUNT NOUN] and [MASS NOUN], a three-dimensional, concrete “thing” on the one hand, and an internally homogeneous, divisible “substance” on the other. The schema of a concrete “thing” gets projected onto entities in other domains, such as units of time (minute, year), products of mental and creative activity (idea, poem, symphony), events (earthquake, football match), and so on. The schema for a completely homogeneous “substance” is likewise projected onto more abstract domains, such as emotional states (anger, love) and activities (dancing, research). The above account of the mass/count noun distinction needs to be extended by the postulation of additional subschemas for mass nouns. For example, there are a number of mass nouns (examples include furniture, fruit, traffic, luggage) which do not designate a homogeneous “substance,” but which rather have the status of superordinate terms which focus on some common property of different kinds of things. Especially interesting is the group of what might be called “plural mass nouns.” Examples include groceries, left-overs, clothes, dishes (as in wash the dishes),12 dregs, etc. These nouns, like the preceding group, are superordinate terms for things of different kinds. They also have the distinctive property that they cannot be used in the singular. One can not, for example, go to the store and buy “a grocery” or “a clothe.” What distinguishes these nouns, semantically, is the fact that they denote different kinds of things which are found together in a single place, or which have been brought together for a specific purpose (cf. Wierzbicka 1988).

5.3. Imagery The semantic representations which constitute the semantic pole of a linguistic sign are not to be understood in terms of truth conditions on the possibility of successful reference. Rather, semantic representations are equated with “conceptualizations,” of varying degrees of schematicity (Langacker 1988: 94). By “conceptualization” is meant mental experience, in the broadest sense. Indeed, Langacker goes so far as to state that a full description of a language would presuppose (and might even be partly coextensive with) “a full description of human cognition” (Langacker 1987a: 64). Lakoff (1987) – developing ideas in Lakoff and Johnson (1980) – has emphasized the role of bodily perceptions and of experiential gestalts in


51

cognition, and the importance of metaphor and metonymy in people’s attempts to understand their experiences. Langacker (1987a) has appealed more to the contribution of general aspects of cognitive processing to the manner in which a person “construes” a state of affairs: scanning, profiling, level of specificity, focal adjustment, perspective, figure-ground relations, and so on. These various construal mechanisms are brought together under the general heading of “imagery.” As suggested, there are many facets to imagery. Primarily, the term has to do with specifically visual images. These may involve the construing of a spatial situation, or the metaphorical construing of a non-spatial situation in spatial terms.13 Consider, for example, those elements of language that have to do with the conceptualization of spatial relations between entities, in particular the prepositions. The spatial meanings of the prepositions are most directly explicated, not by verbal description, but by schematic drawings of the kind used by Brugman (1981) in her study of the polysemy of over. The pedagogical possibilities of this technique have been systematically explored by Dirven (1989b). But alongside their spatial meanings, prepositions also have a vast range of non-spatial uses. To account for these, cognitive grammar appeals in particular to processes of metaphor, whereby non-spatial domains are understood in spatial terms. On this perspective, “idiomatic” uses of prepositions turn out to be subject to a high degree of semantic motivation. Over, in one of its spatial uses, designates the surmounting of an obstacle on a path (He jumped over the wall). Given that life itself is often conceptualized, metaphorically, as a journey along a path, it comes as no surprise that difficult episodes in one’s life can be conceptualized as “obstacles” which one tries to “get over.” Imagery can be understood in a more abstract sense, to denote the way a speaker “manipulates” the elements of a conceived situation. As an illustration, consider the contrast between a partitive expression of the kind three of the boys, and a quantifier-plus-noun expression like three boys. The principal semantic difference between the two expressions has to do with the set of boys from whom the three exemplars are to be selected. Three boys invites the hearer to create a set of individuals, by selecting, randomly, any three members of the category “boys.” Three of the boys, in contrast, presupposes a set of particular boys, from which three are to be selected. More subtle is the contrast between both boys and both of the boys. Both expressions presuppose a set of two known individuals, with the consequence that the two expressions might appear at first sight to be semantically equivalent. The same goes for each boy and each of the boys, both of which likewise presuppose a fairly small set of particular individu-

52

John R. Taylor

als. Nevertheless, the syntax of the expressions does suggest different “imagings” (in Langacker’s sense) of the situation. Both boys merely invokes the two members of a given group of boys. Both of the boys, on the other hand, explicitly instructs the hearer to select from the given set of two boys.

5.4. Conventionalized conceptualizations As noted above, Langacker equates the semantic pole of a linguistic expression with a conceptualization. It needs to be emphasized, however, that these conceptualizations are “conventionalized” (Langacker 1988: 94), in the sense that they are made available to speakers by the language system that they have acquired. For its linguistic expression, mental experience must first be organized in such a way as to conform with the semantic structures symbolized by the available symbolic units (Langacker 1976). This position has important consequences for foreign language pedagogy (as well as for the contrastive study of languages),14 viz., that formal differences between languages are indicative of differences in conceptualization. Thus, learning a foreign language will involve not only learning the forms of the language but also simultaneously learning the semantic structures associated with these forms. Not infrequently, the conceptualizations symbolized by the forms of the foreign language will not be isomorphic with those of the learner’s native language. Now, this point is readily accepted with regard to the lexical resources of a language. Anyone learning English not only has to learn the phonological form associated with the word tall, they also have to learn the highly idiosyncratic and language-specific categorization of vertical space symbolized by this word (cf. Dirven and Taylor 1988). The distinctive contribution of cognitive grammar is to approach other symbolic units – including lexical categories and syntactic structures – in the same way in which words are approached, i.e., not as forms whose distribution is governed by an autonomous syntax, but as forms which symbolize meanings.

6. Languages in contrast The nature and purpose of a pedagogical grammar requires that it focus on learning problems. In the final analysis, the identification of learning problems is a matter for empirical inquiry, and practising teachers are usually


53

very adept at anticipating the kinds of difficulties that their students will experience. A concern of applied linguistics, however, has been a search for principles which will predict learning problems. One approach that has enjoyed considerable currency over the years is the Contrastive Language Hypothesis, according to which the structures of the foreign language will be difficult to learn to the extent that they are non-isomorphic with the structures of the language(s) already known to the learner (Lado 1957). The only limited success of the contrastive hypothesis in predicting a learner’s errors led, from the 1970’s onwards, to a number of investigators appealing to other mechanisms, such as unrestrained rule-generalization on the part of the learner (Richards 1972), universal developmental sequences, or the idea that target language structures will constitute learning difficulties to the extent that they diverge from the principles of “Universal Grammar” (Rutherford 1987; Cook 1988). Cognitive grammar suggests yet another approach. Given the symbolic nature of syntax, we could expect that target language structures will be difficult to learn to the extent that they symbolize “idiosyncratic” conceptualizations, i.e., conceptual categories which are not found in the learner’s mother tongue, or which are not completely isomorphic with those of the learner’s mother tongue. The position, it will be noted, is not dissimilar to that of the contrastive hypothesis à la Lado – with the difference that the focus is now not on the distribution of formal elements per se, but on the meanings symbolized by these forms. It follows that a pedagogical grammar will need to be inherently contrastive, focusing on what is idiosyncratic in the target language vis-à-vis the learner’s native language. One might even speculate on the possibility of a universal metric of conceptual idiosyncrasy. This is a position which would have certain affinities with the distinction between “core grammar” and “the periphery” in Chomsky’s work, again with a focus on semantic content rather than on merely formal entities. Thus, one might hypothesize that certain conceptualizations display a high degree of cognitive “naturalness;” one would expect these conceptualizations to be symbolized in very many languages of the world, and they would be easily learnable by speakers of any language, as well as by children acquiring their mother tongue. Cognitive naturalness would presumably be a function of universal processes of concept formation, universal features of the human environment, and so on. Other categories, on the other hand, might be cognitively “marked;” one would not expect them to be symbolized in language after language (they might even be attested in only a single language), and their learning might require considerable cognitive effort. Such a possibility is suggested by Berlin and Kay’s

54

John R. Taylor

work on colour terminology. Berlin and Kay (1969), in their investigation of the colour terminology of some 98 languages, found that the category “red” was present in practically all the languages of their sample. This category presumably enjoys a very high degree of cognitive naturalness. Other colour categories, like “grey,” “brown” and “pink,” were attested only in languages with more elaborate colour terminologies. For the domain of colour, at least, it does seem plausible to establish a hierarchy of cognitive markedness. Dirven and Taylor (1988) likewise argued for the marked status of the category symbolized by English tall, adding that it would be very remarkable if another language were to have an exact translation equivalent of tall, which categorized the dimension of vertical space in precisely the same way as the English word, and which would therefore be a translation equivalent of tall in all contexts of its use. In light of the above remarks, consider again the categories [MASS NOUN] and [COUNT NOUN] in English. Presumably, the status of chair and water as, respectively, count and mass nouns is highly “natural,” given the nature of their referents. If a language makes a grammatical distinction between count and mass nouns, we should expect that the translation equivalents of chair and water would be categorized as in English. But what about the three nouns information, advice and news? Prima facie, one would expect them to be count nouns. To inform (someone of something), to advise (someone to do something), to report (something to someone) denote events. (Only on a habitual reading would He advised me what to do be taken to refer to an activity.) Consequently, what is conveyed by an act of informing, of advising or of reporting could plausibly be conceptualized as so many discrete “things,” and not as a divisible, replicable, and internally homogeneous “substance.” And yet the English conceptualization is not without its own logic. The nouns information, advice and news all invoke the conceptual domain of verbal communication. Information, advice and news denote some of the different kinds of things that can be communicated from one person to another by means of language. Reddy (1979) has documented in great detail the CONDUIT metaphor which underlies much of our talk, in English at any rate, about verbal communication. According to the metaphor, the mind is construed as a container. Communication involves sending the contents of the mind – suitably packaged in containers of another kind, i.e., words – from one mind to another along a conduit, i.e., a communication channel. Now, a container – a box or a bucket, for example – may be filled either with things (The box is full of books) or with a substance (The bucket is full of water). We find the same possibilities with regard to the metaphorical containers which are words


55

and the mind. These containers may be filled with non-individuated “substance,” e.g., states and qualities like knowledge, learning, experience, wisdom, and imagination. The metaphorical containers may also be filled with individuated “things.” These include facts, ideas, thoughts, hints, and suggestions. Most languages which display the count/mass distinction construe information, advice, and news as “things;” English categorizes these contents of communicative acts,15 somewhat idiosyncratically perhaps, as “substances,” analogous to knowledge, wisdom, and learning. As noted earlier, even highly proficient non-native speakers of English occasionally make errors, in spontaneous speech, on the three words in question. Possibly, the persistence of these and other fossilizations is symptomatic of precisely this low degree of naturalness of the grammatical phenomena in question. If this speculation is valid, the need for a pedagogical grammar which motivates syntactic and morphosyntactic phenomena, by focusing attention on idiosyncratic conceptualizations conventionalized in the grammatical system of a language, becomes even more pressing.

7. Rules A pedagogical grammar will need to offer the learner “rules” for correct usage. But these rules are not the categorial devices of formal linguistics, defining all and only the grammatical sentences of a language. A rule of grammar merely states a conventionalized pairing of semantic structure with a formal structure. As a consequence, the ungrammaticality of a sentence is to be explained in terms of the oddness, incongruity, or other kind of ill-formedness of the meaning that the sentence has, or would have, rather than in terms of the violation of some arbitrary rule of syntax. The facts of verb complementation provide many illustrations of this principle (Dirven 1990). Consider sentence (9): (9)

*I enjoyed that he played the piano.

This sentence is marginal, at best. If it is intended to mean that I got pleasure from listening to him playing the piano, it would be unacceptable. This is because enjoy denotes a psychological experience of a situation, while a that-clause denotes not a situation or the experience of a situation but a proposition, i.e., a mental representation (an “idea,” “thought” or “knowledge”) of a situation.16 Consider next the oddity of (10a) in contrast with (10b) and (10c):

56

(10)

John R. Taylor

a b. c.

?I hope (that) I’ll leave tonight. I hope to leave tonight. I hope I’ll be able to leave tonight.

The acceptability of (10c) shows that we cannot account for the oddity of (10a) by means of some rule stating that, in case the subject of the complement clause is identical with that of the main clause, hope must take a to-infinitive rather than a clausal complement. But if we examine the meanings of the two kinds of complement – and the slightly different meanings that hope acquires in the different contexts – then the oddity of (10a) ceases to be a mystery. Hope with an infinitive belongs to that class of predicates which denote an intention to bring about a new situation. In this case, it is assumed that the person who hopes has some control over the future situation. Alternatively, hope can denote a mental attitude towards the idea of some possible future situation. Here, there is no suggestion that the person is able to control the future. Thus (10c) may be roughly paraphrased: “I hope that circumstances over which I have no control will be such that I’ll be able to leave.” (10a) is odd precisely because it presents a future act intended by the speaker as one over which she has no control. While a pedagogical grammar will strive to offer semantic explanations for grammatical rules, the learner’s attention also needs to be drawn to the possibility of “breaking” rules, just in case a special, unusual, or even bizarre conceptualization is called for. To state that cat is a count noun does not exclude its use as a mass noun, as in After the accident, there was cat all over the road. Or consider certain uses of the progressive in English. It is frequently noted that the progressive cannot be used with stative predicates, like cost (*The book is costing $20). An explanation that is often offered for this restriction is along the lines that a state, by definition, continues through time; it is therefore redundant to mark continuity by means of the progressive aspect. This kind of explanation – to which even Langacker (1987b) appears to subscribe – can hardly be correct. Multiple redundancy – the conveying of information by more than one means – is endemic in human language, and while redundancy might in some cases be censured on the grounds of stylistic infelicity, it is inconceivable that redundantly conveyed information could result in such a strong constraint on grammaticality as the one under discussion. Rather, the explanation must lie with the semantic incompatibility of progressivity and stativity. One approach is that progressivity is only compatible with situations that are (a) incomplete and (b) changing, i.e., in the main, with activities. Yet, in ap-


57

parent contradiction to this principle, the progressive can be used with stative predicates, just in case the stative predicate has some connotations of change. For instance, the stative situation may be one of quite short duration, or of recent beginning (Meat is costing a lot these days).

8. Concluding remarks How far can, and should, one go in offering semantic explanations for the grammar of a language? Dixon (1982: 8), it will be recalled, claimed that “syntactic properties ... can largely be predicted from semantic description.” Noteworthy is that Dixon found it necessary to hedge his statement with the word largely. This suggests that some syntactic properties are not predictable from semantics, i.e., that there will always exist a certain residue of arbitrariness. Wierzbicka (1988), in contrast, appears to adopt an uncompromisingly “maximalist” position, according to which a residue of arbitrary syntactic phenomena would merely betray the linguist’s failure to come up with the appropriate semantic generalization. I am inclined to sympathize with Wierzbicka’s position, at least, as a research programme. That all syntactic phenomena are subject to a semantic explanation should at least constitute the researcher’s null hypothesis, to be modified only in the face of intractable evidence to the contrary. But we also need to bear in mind that a pedagogical grammar imposes its own constraints, including the constraint of ready accessibility to the unsophisticated reader, not to mention publishers’ constraints on volume length and submission dates. For practical reasons, therefore, it will often be necessary to accept an arbitrary residue. The function of a pedagogical grammar is to promote the learner’s insight into the foreign language system. In essence, promoting insight means reducing the perceived arbitrariness of the foreign language system. A person perceives something as arbitrary if he can see no reason why it should be as it is. For this reason, it is not enough to merely inform the learner that a particular element belongs to a given formal category and not to another, and thus behaves in this way rather than that, or to state that such-and-such an expression is grammatically correct while other wordings are grammatically incorrect. Rather, we need to explain to the learner why the foreign language should be as it is. A cognitively-based pedagogical grammar is motivated by belief that the insights of Cognitive Grammar – presented in a suitable pedagogical format – can contribute substantially to the attainment of this objective. If the basic assumptions of the cognitive program are

58

John R. Taylor

correct, it should be possible to realize the objective by bringing to the learner’s consciousness the conceptualizations conventionally associated with the structures of the foreign language. In order to be able to use the word tall correctly, the learner needs to grasp the meaning of tall, i.e., the language’s conceptualization of the vertical dimension conventionally symbolized by tall. Likewise, in order to be able to handle the aspect system of English, or verb complementation, the learner needs to know the meaning of the progressive, and the meaning of a to-infinitive and of a gerund. The challenge of applying cognitive linguistic insights to a pedagogical grammar lies precisely in searching for descriptively adequate, intuitively acceptable, and easily accessible formulations of these meanings.

Notes 1. This is a revised version of Taylor (1993). In revising the paper, I have endeavoured to remain true to the theoretical orientation of the original text. As a matter of fact, the background theory – Langacker’s Cognitive Grammar – has proved remarkably resilient over the past couple of decades. For example, the recent upsurge of interest in constructions (see, e.g., Tomasello 2003) merely elaborates on theoretical notions which were already clearly delineated in Langacker (1987a). My current views, however, diverge from those expressed almost 15 years ago in two respects. Firstly, I believe that the notion of “semantic explanation” needs to be replaced by the much broader notion of “motivation,” whereby an expression is motivated not only, or not simply, by the conceptualization which it symbolizes, but by the existence of multiple links between the expression and other conventionalized resources of the language (Taylor 2004). Secondly, I am now strongly of the view that an account of the target language needs to be based in corpus data, concerning, for example, the relative frequency of various (morpho-)syntactic choices, and their incidence in different text types (cf. Biber et al. 1998; Biber et al. 2002). In this respect, recent corpus-based studies develop Langacker’s notion of “entrenchment,” and the status of Cognitive Grammar as a “usage-based” theory. 2. On the historical background to Cognitive Linguistics, see Taylor (to appear). 3. The pedagogical grammar alluded to here will see publication as Radden and Dirven (2007). 4. I shall use the term “foreign language learner” to include certain kinds of second language learner. For a review of the very extensive literature on pedagogical grammar, see Dirven (1990).


59

5. A question of some importance is the nature, and the manner of operation, of the Language Acquisition Device. For Krashen, and others who appeal to it, it has the status of a mysterious black box, whose output we can observe, but not its internal operation. Tomasello (2003) can be read as an attempt at demystifying this concept. 6. The status of a noun as count or mass is, of course, important insofar as it has a bearing on numerous aspects of usage. For example, mass nouns (or, more precisely, nouns used as mass nouns) cannot co-occur with the indefinite determiner a(n) and they cannot be pluralized, whereas count nouns cannot occur (as singulars) without a determiner. 7. This popular view is not in strict accordance with Saussure’s thought, as recorded in the Cours (Taylor 1996: 58–59). Thus we read, in connection with the issue discussed here, that “ce qui est dit des mots s’applique à n’importe quel terme de la langue, par exemple aux entités grammaticales” [what is said of words is true of any other element of language, for example, grammatical entities] (Saussure 1964: 161). The brief remarks on syntax in the Cours suggest a view of the syntagmatic combination of signs which is highly compatible with Langacker’s more elaborate discussion of the issue. 8. A reviewer suggests that would like always takes an infinitive. This is not strictly true. Googling threw up quite a few examples, including the following: (i) im looking for a woman who wants to have fun without playing games someone who would like going out to the bar or enjoy a night at home cuddling on the couch. (ii) I was not sure, before I came, that I would like going to a women's college, but I quite enjoy it. Note that both examples refer to a generalized situation, not to a specific event. 9. As stated earlier, this account of the meanings of the complements is not meant to be definitive, or final. The matter, in fact, has been the subject of considerable debate. For semantically-oriented accounts of complement selection, see Duffley 1992, 1995; Duffley and Tremblay 1994; Egan 2003; Hamawand 2002; Smith and Escobedo 2003. 10. [MASS NOUN] and [COUNT NOUN] are instances of the still more schematic category [NOUN], whose semantic content is stated to be “region in a domain.” In proposing a semantic characterization of [NOUN] – and of other lexical categories – Langacker is departing from the received doctrine of the arbitrariness of lexical categories. Thus Jackendoff – whose more recent work adopts a perspective which is in some respects compatible with that of Cognitive Grammar – dogmatically states “it is well known that ... the grammatical category noun cannot be identified with any coherent semantic category” (Jackendoff 1983: 14).

60

John R. Taylor

11. Croft (2001) has argued against the universal status even of categories such as noun and verb, on the basis that membership in the categories, and the syntactic behaviour of the categories, are language-specific. Even so, Croft does acknowledge the existence of a universal conceptual space, which may be carved up slightly differently by different languages. Croft’s views reinforce the need to focus on the language-specific properties of the categories. 12. Dishes, here, does not refer to a plural number of items of the kind “dish,” but to all the things (pots, plates, cutlery, etc.) that are used in the preparation and eating of a meal, and which need to be washed afterwards. 13. It is well-established that learning can be enhanced through supplementary, or parallel, visual information (Paivio 1986). This is something which most teachers, of all disciplines, are no doubt aware of. A distinctive feature of Cognitive Grammar, however, is that visualizations are not merely expository aids, but are inherent to the theory itself. (Recall that the name “Cognitive Grammar” replaced the earlier designation “Space Grammar.”) The visualizations inherent to the theory, especially as these pertain to more “abstract” conceptualizations, provide a rich source of material for pedagogical exploitation. 14. For a contrastive study undertaken within a cognitive linguistic framework, see Taylor (1988). 15. And not only these. Note the mass noun status of council (as in to give someone council) and intelligence (as in military intelligence), nonsense, and old hat (What he said was nothing but old hat). The existence of two alternative “paradigms” for the conceptualization of what is conveyed through acts of communication – we might call these the “fact-paradigm” and the “knowledgeparadigm” – is brought into focus by the fluctuating use of the word data. Some speakers, aware of the status of this Latin borrowing in the donor language, may be inclined to use data as a plural noun, i.e.,, the word is assimilated to the fact-paradigm. On the other hand, given that the word lacks any native-English plural marking, data is also used as a singular mass noun, i.e., it is assimilated to the knowledge-paradigm. 16. In an earlier version of this paper, I had claimed that enjoy (the fact) that would be ungrammatical, because of the incompatibility of enjoy with a propositional complement. A reviewer points out the incorrectness of this assertion, citing the Googled example He enjoyed the fact that she was pretty clear-eyed. There are, in fact, 7 examples of enjoy(ed) the fact that in the BNC, including we all enjoyed the fact that there was no master around to tell us what to do. (These uses, however, make up less than 1% of the 11,231 instances of enjoy(ed) in the corpus.) Marginally, therefore, knowledge pertaining to a proposition can be the source of enjoyment, alongside the dominant usage.


61

References Berlin, Brent, and Paul Kay 1969 Basic Color Terms: Their Universality and Evolution. Berkeley: University of California Press. Biber, Douglas, Susan Conrad, and Randi Reppen 1998 Corpus Linguistics: Investigating Language Structure and Use. Cambridge: Cambridge University Press. Biber, Douglas, Susan Conrad, and Geoffrey Leech 2002 The Longman Student Grammar of Written and Spoken English. London: Longman. Bolinger, Dwight 1977 Meaning and Form. London: Longman. Brugman, Claudia 1981 Story of over. MA thesis, University of California, Berkeley. Chomsky, Noam 1965 Aspects of the Theory of Syntax. Cambridge: Massachusetts Institute of Technology Press. 1986 Knowledge of Language: Its Nature, Origin and Use. New York: Praeger. Cook, Vivian J. 1988 Chomsky’s Universal Grammar: An Introduction. Oxford: Blackwell. Croft, William 2001 Radical Construction Grammar. Oxford: Oxford University Press. Dirven, René 1985 Definition of a pedagogical grammar. ITL Review of Applied Linguistics 67/68: 43–67. 1989a A cognitive perspective on complementation. In Sentential Complementation and the Lexicon: Studies in Honor of Wim de Geest, Danny Jaspers, Wim Klooster, Yvan Putseys, and Pieter Seuren (eds.), 140–168. Dordrecht, the Netherlands: Foris. 1989b Space prepositions. In A User’s Grammar of English, René Dirven (ed.), 519–550. Frankfurt: Lang. 1990 Pedagogical grammar (State of the art article). Language Teaching: The International Abstracting Journal for Language Teachers and Applied Linguists 23: 1–18. 2005 Major strands in cognitive linguistics. In Cognitive Linguistics: Internal Dynamics and Interdisciplinary Interaction, Francisco J. Ruiz de Mendoza Ibáñez and Sandra Peña Cervel (eds.), 69–100. Berlin/New York: Mouton de Gruyter.

62

John R. Taylor

Dirven, René, and John R. Taylor 1988 The conceptualization of vertical space in English: The case of tall. In Topics in Cognitive Linguistics, Brygida Rudzka-Ostyn (ed.), 379–402. Amsterdam: Benjamins. Dixon, Robert M.W. 1982 Where Have All the Adjectives Gone? Berlin: de Gruyter. Duffley, Patrick J. 1992 The English Infinitive. London: Longman. 1995 Defining the potential meaning of the -ing form in a psychomechanical approach. Langues et Linguistique 21: 1–11. Duffley, Patrick J., and Rachel Tremblay 1994 The infinitive and the -ing as complements of verbs of effort. English Studies 75: 566–575. Dulay, Heidi, Marina Burt, and Stephen Krashen 1982 Language Two. Oxford: Oxford University Press. Egan, Thomas 2003 Distance and direction: A usage-based study of infinitive and -ing complement clauses in English. PhD dissertation, English Department, University of Oslo. Ellis, Rod 2001 Form-Focused Instruction and Second Language Learning. Oxford: Blackwell. Graustein, Gottfried, and Gerhard Leitner (eds.) 1989 Reference Grammars and Modern Linguistic Theory. Tübingen: Niemeyer. Greenbaum, Sidney 1987 Reference grammars and pedagogical grammars. World Englishes 6: 191–197. Gregg, Kevin 1984 Krashen’s monitor and Occam's razor. Applied Linguistics 5: 79– 100. Haiman, John 1985 Natural Syntax. Cambridge: Cambridge University Press. Hamawand, Zeki 2002 Atemporal Complement Clauses in English. A Cognitive Grammar Analysis. München: Lincom Europa. Jackendoff, Ray 1983 Semantics and Cognition. Cambridge: Massachusetts Institute of Technology Press. Krashen, Stephen 1981 Second Language Acquisition and Second Language Learning. Oxford: Pergamon.

Some pedagogical implications of cognitive linguistics 1982

63

Principles and Practice in Second Language Acquisition. Oxford: Pergamon. Krashen, Stephen, and Tracy Terrell 1983 The Natural Approach: Language Acquisition in the Classroom. Oxford: Pergamon. Lado, Robert 1957 Linguistics across cultures. Ann Arbor: University of Michigan Press. Lakoff, George 1987 Women, Fire and Dangerous Things: What Categories Reveal about the Mind. Chicago: Chicago University Press. Lakoff, George, and Mark Johnson 1980 Metaphors We Live By. Chicago: Chicago University Press. Langacker, Ronald W. 1976 Semantic representations and the linguistic relativity hypothesis. Foundations of Language 14: 307–357. 1987a Foundations of Cognitive Grammar. Vol. 1: Theoretical Prerequisites. Stanford, California: Stanford University Press. 1987b Nouns and verbs. Language 63: 53–94. 1988 The nature of grammatical valence. In Topics in Cognitive Linguistics, Brygida Rudzka-Ostyn (ed.), 91–125. Amsterdam: Benjamins. 2004 A visit to Cognitive Grammar: Interview with Ricardo Maldonado. Annual Review of Cognitive Linguistics 2: 305–319. Munsell, Paul, and Thomas Carr 1981 Monitoring the monitor: Review of Second Language Acquisition and Second Language Learning by Stephen Krashen (1981). Language Learning 31: 493–502. Paivio, Allan 1986 Mental Representation: A Dual-Coding Approach. Oxford: Oxford University Press. Radden, Günter, and René Dirven 2007 Cognitive English Grammar. Amsterdam: Benjamins. Reddy, Michael 1979 The conduit metaphor: A case of frame conflict in our language about language. In Metaphor and Thought, Andrew Ortony (ed.), 284–324. Cambridge: Cambridge University Press. Richards, Jack 1972 Error analysis and second language strategies. Language Sciences 17: 12–22. Rudzka-Ostyn, Brygida (ed.) 1988 Topics in Cognitive Linguistics. Amsterdam: Benjamins.

64

John R. Taylor

Rutherford, William E. 1987 Second Language Grammar: Learning and Teaching. London: Longman. de Saussure, Ferdinand 1964 Cours de Linguistique Générale. Paris: Payot. First published 1916. Selinker, Larry 1972 Inter-language. IRAL 10: 209–231. Sharwood Smith, Michael 1981 Consciousness-raising and the second language learner. Applied Linguistics 2: 159-168. Smith, Michael B., and Joyce Escobedo 2003 The Semantics of to-infinitival vs. -ing verb complement constructions in English. In Proceedings from the Main Session in the Chicago Linguistic Society’s Thirty-Seventh Meeting, Mary Andronis, Christopher Ball, Heidi Elston, and Sylvain Neuvel (eds.), 549–564. Chicago: Chicago Linguistic Society. Taylor, John R. 1988 Contrasting prepositional categories: English and Italian. In Topics in Cognitive Linguistics, Brygida Rudzka-Ostyn (ed.), 299–236, Amsterdam: Benjamins 1989 Linguistic Categorization: Prototypes in Linguistic Theory. Oxford: Oxford University Press. 3rd revised edition: 2003. 1990 Schemas, prototypes and models: In search of the unity of the sign. In Meaning and Prototypes: Studies in Linguistic Categorization, Savas L. Tsohatzidis (ed.), 521–534. London: Routledge. 1993 Some pedagogical implications of cognitive linguistics. In Conceptualizations and Mental Processing of Language, Richard A. Geiger and Brygida Rudzka-Ostyn (eds.), 201–223. Berlin: Mouton de Gruyter. 1996 Possessives in English: An Exploration in Cognitive Grammar. Oxford: Clarendon Press. 2002 Cognitive Grammar. Oxford: Oxford University Press. 2004 The ecology of constructions. In Studies in Linguistic Motivation, Günter Radden, and Klaus-Uwe Panther (eds.), 49–73. Berlin: Mouton de Gruyter. to appear Cognitive linguistics and autonomous linguistics. In: Handbook of Cognitive Linguistics, Dirk Geeraerts and Hubert Cuyckens (eds.). Oxford: Oxford University Press. Taylor, John R., and René Dirven 1991 Complementation. LAUD (=Linguistic Agency, University of Duisburg) Paper, Series A, 303.


65

Tomasello, Michael 2003 Constructing a Language: A Usage-based Theory of Language Acquisition. Harvard: Harvard University Press. VanPatten, Bill 1996 Input Processing and Grammar Instruction: Theory and Research. Norwood, New Jersey: Ablex. Wierzbicka, Anna 1988 The Semantics of Grammar. Amsterdam: Benjamins.

Cognitive linguistic theories of grammar and grammar teaching Cristiano Broccias

Abstract This paper first briefly examines some of the most important theoretical approaches to grammar within the cognitive linguistic paradigm, Langacker’s Cognitive Grammar, Goldberg’s Construction Grammar, Croft’s Radical Construction Grammar and Fauconnier and Turner’s Blending Theory. It points out that there is a striking similarity between the development of (theoretical) cognitive linguistics (against the background of generative grammar) and the recent history of language teaching. Cognitive approaches to grammar are converging towards a usagebased/network model of language (i.e., a bottom-up, maximalist approach to language), which contrasts with the decontextualized view of language (i.e., a topdown, minimalist approach) espoused by generative grammarians. Similarly, language teaching in general and the teaching of grammar in particular have moved from decontextualized drilling activities to more meaningful, communicative/context-based methods, i.e., a usage-based model of language teaching. The second part argues that one obvious drawback of the usage-based model, both in language theorizing and in language teaching, is “data overload”. For example, it is not possible to say how many senses prepositions – a staple topic in discussions about applied cognitive linguistics – have or, more in general, how many extensions constructions have. Networks are in part arbitrary constructs by the linguist rather than fine-grained reflections of the language user’s mind. Consequently, approaches which try to implement network analyses in grammar teaching (as is usually done in the case of prepositions) may in fact turn out to be detrimental. They impose on the non-native speaker the burden of having to also learn how (alleged) extensions came into being while no psycholinguistic and/or diachronic evidence for such extensions is offered in the first place. Hence, the third part of this paper claims that, rather than insisting on detailed elucidations of extensions from prototypes, cognitive linguistics-based grammar teaching should aim at raising language awareness by concentrating on prototypes (and their relation to basic cognitive abilities) and explicitly pointing out that non-prototypical cases may be (synchronically) motivated only to a degree. This point is illustrated for change constructions, simultaneity constructions and noun phrases against the background of the four theories briefly analysed in the first part.

68

Cristiano Broccias

Keywords: grammar; construction; Cognitive Grammar; Construction Grammar; Radical Construction Grammar; Blending Theory; pedagogical grammar; usage-based model; network model; syntax-lexicon continuum; construal; schema; extension; prototype; motivation; polysemy; change construction; simultaneity construction; caused-motion construction; noun phrase

1. What is grammar? This paper deals with the relation between cognitive linguistic theories of grammar and pedagogical grammar (mainly for foreign students). But before discussing this topic in any detail, it is essential to first (try to) clarify what the meaning of the term grammar is. Traditionally (i.e., descriptively and pedagogically), grammar is identified with (inflectional) morphology and syntax (e.g., Quirk et al. 1985: 12). Traditional grammar concerns itself with both the shape of words and how words (and phrases) can be combined together. This view is not too dissimilar, within the realm of theoretical linguistics, from that of the generative tradition where (formal) grammar is viewed as being made up of two components (e.g., Radford 2004: Ch. 1). One component is the lexicon, a list of unpredictable forms, and the other component is a system of productive rules (the syntax), which specifies how lexical items are to be combined together. The traditional and formal uses of the term “grammar” contrast with the view espoused by cognitive linguistics, where the term grammar tends to be used in a more general fashion. Still, “grammar” will have to be understood in its traditional sense when I discuss the relevance, if any, of cognitive linguistic models to pedagogical grammar. In what follows, I will in Section 2 briefly sketch out the main assumptions underlying some of the most important cognitive grammar approaches (for a fuller description and evaluation, see Broccias 2006c). In Section 3, I will comment on the relation between cognitive linguistics and contemporary teaching methodology stressing the importance of pedagogical grammar despite the growing consensus on a usage-based model for language analysis. Section 4 tries to evaluate what the most important aspects of cognitive grammar approaches are for pedagogical grammar (i.e., networks, prototypes, and attention to constructions). Section 5 draws the conclusions.

Cognitive linguistic theories of grammar and grammar teaching

69

2. Cognitive grammar models 2.1. Cognitive Grammar Langacker’s Cognitive Grammar views grammar as “a structured inventory of conventional linguistic units” (Langacker 1987: 37). Any linguistic expression is defined as an association between a semantic and a phonological structure, i.e., any linguistic expression is a symbol. For example, the lexical item cat is regarded as consisting of a semantic pole (i.e., what would traditionally be called both its denotative and its connotative meaning, abbreviated as [CAT]) and a phonological pole (abbreviated as [kæt]). A linguistic unit is a symbolic structure which has unit status, i.e., it is accessed in largely automatic fashion or, to put it differently, is entrenched. Further, a linguistic unit is conventional if it is shared by a substantial number of individuals. By the term “inventory” in his definition of grammar, Langacker means that a grammar is not a generative algorithm but, rather, a collection of conventional symbolic units. Crucially, such an inventory is not a list, but is structured (i.e., it is a network) because symbolic units, as schemas abstracted from usage events, are related by way of categorizing relationships. For example, the clause I love you is an instantiation or elaboration of the transitive construction; down meaning “sad,” is a metaphorical extension of down from the spatial domain to that of psychological states. It should be pointed out that the definition of grammar above does not impose any restrictions on the complexity and specificity of linguistic units. Linguistic units can be of any length. Grammar includes symbolic assemblies of any internal complexity rather than just atomic units like cat. Complex constructions like the ditransitive construction (e.g., He gave her a present) are part and parcel of grammar. Further, linguistic units can be of any degree of specificity. Grammar includes both very specific patterns (e.g., I love you, which can be considered an entrenched unit in the English language although its meaning is transparent, unlike idioms such as kick the bucket, give up the ghost, etc.) and general schemas that may subsume them (e.g., grammar includes the transitive construction, of which I love you is an elaboration or instantiation). From these two observations regarding complexity and specificity, it follows that there are no clear boundaries between what are traditionally referred to as the lexicon, morphology, and syntax (this is known as the Syntax-Lexicon Continuum hypothesis). These labels refer to symbolic assemblies differing from one another only in terms of structural complexity and schematicity (or specificity), as is illus-

70

Cristiano Broccias

trated in Table 1 below (adapted from Croft 2001: 17). Indeed, any entrenched symbolic assembly is regarded as a construction in Cognitive Grammar (as well as in Radical Construction Grammar, see Section 2.2.2) irrespective of either its internal complexity or its specificity. More generally, Cognitive Grammar constantly underlines that much in language is a matter of degree. Since grammar is an integral part of human cognition and since human cognition rests, among other abilities, on categorization (which operates by way of instantiation and extension relations rather than feature-listing), sharp boundaries should not be expected in grammar. Table 1. The syntax-lexicon continuum Construction type complex and (mostly) schematic

Traditional name syntax

complex and (mostly) specific complex but bound atomic and schematic

idiom morphology word class

atomic and specific

word/lexicon

Examples noun verb noun (i.e., the transitive construction), adjective noun (i.e., a noun phrase) I love you, black cat noun-s verb, adjective, noun, pronoun love, black, cat, I, you

Although the symbolic links between the semantic pole and the phonological pole of a construction may often be regarded as arbitrary (e.g., it is arbitrary that a doer is often signalled morphologically by way of the -er morpheme), Cognitive Grammar does not neglect the issue of motivation behind symbolic links. For example, the form of the prepositional dative construction (e.g., He sent a letter to Susan) is not arbitrary but emphasizes “the path traversed by the letter with Susan as a goal” (Langacker 1987: 39), as is signalled by the motion preposition to. Conversely, the double object construction (e.g., He sent Susan a letter) is said to emphasize the possessive relation between Susan and the letter by way of their “juxtaposition and linear order” (Langacker 1987: 39). Cognitive Grammar claims that the meaning of these two constructions is not necessarily identical because, among other things, they impose different construals (i.e., “views”) on a common conceptual content. The prepositional dative construction construes transfer of possession in motion terms while the double object construction construes transfer of possession in terms of its out-


71

come, i.e., the establishment of a relation of possession (or, more generally, control) between the indirect object and the direct object. Although Cognitive Grammar emphasizes the emergence of linguistic units out of specific usage events (i.e., it is a usage-based model of language), Langacker strives to offer a conceptual (i.e., form-independent) and universal (i.e., valid for all languages) definition of grammatical (or syntactic) functions (also known as grammatical/syntactic relations or roles) and word classes (also known as parts of speech or lexical categories). In Cognitive Grammar, “subject” for example is regarded as a universal notion ultimately based on the (perceptual) distinction between figure and ground, viz. trajector and landmark, configurations in a scene. As is wellknown from Gestalt psychology (for an overview, see Ungerer and Schmid 2006: Ch. 4),2 certain objects function as primary foci of attention (e.g., smaller and/or moving objects as opposed to larger, static objects). A sentence like The table is under the book (employed to convey the scene depicted by The book is on the table) is felt to be deviant because it reflects a non-canonical perception of the intended spatial configuration. A smaller object like a book is usually chosen as the primary focus of attention (or trajector, in Cognitive Grammar terminology) whereas a larger object like a table functions as a secondary focus of attention (or landmark, in Cognitive Grammar terminology). Cognitive Grammar proposes that subject is a universal category in that it can be defined schematically by characterizing it as a clause-level trajector. Similarly, object can be defined schematically as a clause-level secondary figure or landmark. In addition to its schematic conception of grammatical relations, Cognitive Grammar also provides a schematic characterization of word classes. Cognitive Grammar contends that parts-of-speech can not only be defined prototypically (e.g., a noun prototypically refers to a physical object) but can also be described in a manner which is valid for all its instantiations. Consider nouns and verbs. (The semantic pole of) a noun is characterized as a “thing,” a technical term in Cognitive Grammar referring to a set of interconnected entities. The noun team (see Langacker 1987: 197), for instance, profiles a set of entities rather than singling out any constitutive member. (The semantic pole of) a verb depicts a process (or temporal relation), i.e., a relation between two entities scanned sequentially. The verb kiss, for example, refers to the relation between a “kisser” and a “kissed” entity whose various parts (i.e., the constitutive parts into which the kissing event can be broken down) we conceptualize successively as in a motion picture (rather than apprehending as a single gestalt).

72

Cristiano Broccias

2.2. Construction grammars 2.2.1. Goldberg’s Construction Grammar Goldberg’s (1995) Construction Grammar originally posited constructions in grammar (i.e., in the language system) if and only if some aspect of their meaning was regarded as “not strictly predictable from [the construction’s] component parts or from other previously established constructions” (Goldberg 1995: 4). For example, a sentence like Sam slept the whole trip away is said to be unaccounted for by the syntactic rules of English (e.g., sleep a trip is not a possible string in English) and, hence, a construction with the syntax “Subject Verb Object (referring to time) away” and the meaning (roughly) “to spend the specified amount of time by being constantly engaged in the activity denoted by the verb” should be recognized as a unit in the grammar of English. More recently, however, Goldberg (e.g., Goldberg 2006) has opted for a more comprehensive definition of construction, i.e., one along the lines of both Cognitive Grammar and Radical Construction Grammar. Constructions should be posited in grammar even if they are fully predictable (e.g., I love you vs. kick the bucket) and all the types listed in Table 1 should be categorized as constructions. This in turn implies that, as in Cognitive Grammar and Radical Construction Grammar, no strict separation is postulated between what are traditionally called the lexicon and syntax. Rather, they are taken to form a continuum. Despite the use of a more inclusive notion of construction in recent work, Construction Grammar still draws a sharp distinction between constructional meaning and word meaning. As is pointed out by Langacker (2005), Construction Grammar aims at minimizing lexical polysemy in favour of constructional polysemy. Consider the verb slice, which can be used in a variety of constructions, as shown in (1) (from Goldberg 2006: 7): (1)

a. b. c. d. e.

He sliced the bread. [transitive] Pat sliced the carrots into the salad. [caused motion] Pat sliced Chris a piece of pie. [ditransitive] Emeril sliced and diced his way to stardom. [way construction] Pat sliced the box open. [resultative]


73

Construction Grammar argues against postulating five different senses for slice (as would be done in traditional generative approaches, where constructions are projected from a verb’s argument structure). Rather, slice, meaning simply “cut with a sharp instrument,” is claimed to combine with five different constructions (or argument structure constructions), all of which are independently stored in the grammar of English. For example, the verb’s participant roles, which can be dubbed “slicer” and “sliced,” fuse respectively with the arguments causer and theme of the caused motion construction in (1b). Further, the caused motion construction provides a goal argument, realized as the PP into the salad. It should also be pointed out that, as in Cognitive Grammar, grammatical knowledge is represented in networks in Construction Grammar. But, instead of categorizing relations, Construction Grammar makes use of inheritance links (as in computer science) to relate constructions to one another. Construction Grammar allows for multiple inheritances, i.e., a construction can be categorized by two (or more) independently established constructions. For example, the intransitive motion construction (e.g., The car screeched around the corner) inherits from both the intransitive construction (since the intransitive motion construction is quite trivially intransitive) and the caused-motion construction (since motion is predicated of one of the constructional elements, i.e., the subject). Four major types of inheritance links are recognized: polysemy links, metaphorical extension links, subparts links and instance links (see Goldberg 1995: 74–81). Polysemy links and metaphorical links roughly correspond to extension relations in Cognitive Grammar. Construction Grammar claims for example that the double object construction has a variety of senses which can be viewed as extensions (via polysemy links) from the central sense “[a]gent successfully causes recipient to receive patient” (see Goldberg 1995: Figure 2.2). A metaphorical link connects the caused motion construction and the resultative construction (see (1d) and (1e) above) in that the latter is said to have originated from the former via the metaphorical construal of states (e.g., crazy in Chris drove Pat crazy) as locations. A subpart link is posited when a construction is “a proper subpart of another construction” (Goldberg 1995: 78). The intransitive motion construction is linked to the caused-motion construction via such a link since the former is a proper subpart of the latter (the only missing element in the intransitive motion construction is the “cause” argument; for a different view see Broccias 2003: Ch. 5). Finally, instance links correspond to elaborative categorization in Cognitive Grammar. This obtains when a con-

74

Cristiano Broccias

struction specifies another construction in more detail (e.g., Chris drove Pat crazy instantiates the schematic resultative construction). The importance of constructions in language acquisition has been discussed in various works (e.g., Tomasello 2003; Goldberg 2006) and needs to be summarized here because it has obvious implications for pedagogical grammar. It has been observed that category formation (in general, i.e., including non-linguistic categories) is influenced by frequency and order: high token frequency (e.g., of a particular verb) facilitates category learning (e.g., of the argument structure that a particular verb realizes) and presenting core category members first (rather than interspersing them with other less-core members in no particular order) also plays an important role. Goldberg (2006) also addresses the issue of how overgeneralizations are blocked in language acquisition. For example, how do English speakers learn that only (2b) is well-formed? (2)

a. b.

*She explained me the story. She explained the story to me.

Goldberg argues that pre-emption and openness are two crucial factors. If the learner consistently hears, e.g., the pattern in (2b) where she would expect (2a) to be used, then she can conclude that (2a) is not appropriate (pre-emption is therefore an instance of indirect negative evidence). Secondly, learners seem to use a new verb in a certain argument structure construction only if the verb is close enough in meaning “to verbs they have already heard used in the pattern” (Goldberg 2006: 99). This is consonant with the view that children are conservative in making generalizations. Still, it has also been remarked that children are quick generalizers. Generalizations are of course needed in order to understand and produce new utterances. If we simply memorized the sentences we have heard, communication would not be possible. Goldberg (2006) shows that we rely on constructions because they are as good predictors of overall sentence meaning as verbs and they are much more available (than specific verbs). The reliance on constructions has also been confirmed in second-language acquisition at least in the case of advanced learners (of English), see e.g., Valenzuela and Rojo (this volume), Gries and Wulff (2005).1


75

2.2.2. Radical Construction Grammar In contrast to Goldberg’s Construction Grammar, which is partly indebted to Fillmore’s non-commitment to a fully cognitive linguistic model, Croft’s (2001) Radical Construction Grammar is claimed to be radical for four reasons. Firstly, grammatical categories (i.e., word classes and syntactic roles) are not viewed as primitives but are argued to be constructionspecific. One of the greatest merits of Radical Construction Grammar is to have exposed the circularity inherent in much of modern syntactic argumentation. Constructions are used to define categories (this is known as the distributional method) and then the categories so established are used to define (or categorize) other constructions. As a matter of illustration, consider how direct object status is evaluated in English (see Croft 2001: 35– 45). Croft points out that, traditionally, passivizability – the fact that the alleged direct object in an active sentence can become the subject of the corresponding passive sentence, see (3a) vs. (3a’) – is taken to be the defining criterion for direct object status. Consequently, 80 kg in (3b) may not be categorized as a direct object by some linguists: no corresponding passive sentence is possible, see (3b’). By contrast, objects of prepositions may sometimes be passivized, see (3c’). This would lead us to classify such objects as direct objects. The problem here is that the choice of passivization (over, for example, obligatory presence vs. absence of a preposition after the verb) as the defining criterion for direct object status is not theoretically motivated. Further, there is no a priori reason why we should group different cases into the same category. In fact, a Radical Construction Grammar analysis would claim that the three sentences at hand contain three different types of “objects” whose properties overlap only partially: John immediately follows the verb and can become the subject of a passive sentence; 80 kg cannot become the subject of a passive sentence but, like John, occurs immediately after the verb; this house, like John, can become the subject of a passive sentence but, unlike John and 80 kg, is separated from the verb by a preposition. (3)

a. a.’ b. b.’ c. c.’

A mule kicked John. John was kicked by a mule. John weighs 80 kg. *80 kg are weighed by John. Hemingway lived in this house. This house was lived in by Hemingway.

76

Cristiano Broccias

The second reason why Radical Construction Grammar is radical is because constructions are taken to be the basic units of syntactic representation. This assumption is also shared by Cognitive Grammar and Construction Grammar (at least in its latest implementations) and was illustrated in Table 1 above. Thirdly, Radical Construction Grammar is radical because syntactic relations are claimed not to exist. Croft argues that the term syntactic role should be used instead. He points out that the term “subject,” for example, can be used in two different ways (e.g., Croft 2001: 24; Croft and Cruse 2004: 285–286). “Subject” can be used to refer to the syntactic relation between an element of a construction and another element (the verb), e.g., we say that Heather is the subject of sings in Heather sings. “Subject” can also be used to refer to the role that an element has in a construction, i.e., the part-whole (or meronomic) relation of an element with the whole construction. For example, we say that Heather has the role subject in the intransitive construction Heather sings. The crucial point here is that Croft uses the term syntactic relation in a traditional (formal) sense, whereby a syntactic relation like subject is taken to be construction-independent by definition – cf. the Government and Binding Theory configurational definition of subject as the Specifier of an Inflectional Phrase or [Spec, IP] – and does not necessarily symbolize any semantic content. Since the only entities which should be posited in grammar according to Croft are constructions (as well as relations among them), it follows that syntactic relations (as construction-independent objects) must be dispensed with. Accordingly, Radical Construction Grammar only uses the term role, which refers to the relation between an element and the whole construction rather than the relation between an element and another element within the same construction. Finally, constructions are said to be language-specific: no two constructions can be assumed to be identical across languages. This assumption is also shared by Construction Grammar and Cognitive Grammar. In sum, Radical Construction Grammar can be described as a usagebased model: grammar is maximalist and bottom-up, rather than minimalist and top-down (like formal approaches). It consists of taxonomic hierarchies of constructions which include very specific constructions (e.g., kick the bucket, kick the habit) as well as more general schemas abstracting away their commonalities (e.g., a transitive schema for kick, a general transitive schema, etc.). A self-explanatory example of the kind of taxonomy assumed in Radical Construction Grammar is offered in Figure 1 below (adapted from Figure 1.15 in Croft 2001: 56).


Sbj

IntrSbj

Kick1

Ktb1

IntrVerb

Pred

TrSbj

kick

Kick2

kick

the bucket

77

Kiss1

Kth1

TrVerb

TrObj

kiss

Kiss2

kick

the habit

Figure 1. A construction taxonomy in Radical Construction Grammar

Although Radical Construction Grammar claims that grammatical categories are construction- and language-specific, Croft (2001) shows that syntactic roles can be compared across languages once semantic notions (e.g., actor event, undergoer event, etc.) are appealed to. In a similar vein, he argues that a universal characterization of parts of speech is possible (For reasons of space, I will only discuss parts-of-speech here). Croft’s analysis is based on the interaction between discourse functions and semantic classes. Croft (2001: 66) recognizes three discourse (or pragmatic/communicative) functions, namely the propositional acts of predication (e.g., This intelligent detective is young, where young is predicated of this intelligent detective), reference (e.g., this intelligent detective, which picks out a certain referent), and modification (e.g., intelligent detective, where intelligent modifies our “cognitive file” for the detective in question, i.e., it enriches the detective’s identity by adding a property to the “features” already stored in our mind concerning the detective). As these examples illustrate, we tend to use certain constructions to code such discourse functions (the constructions in question are called propositional act constructions). Croft hypothesizes that the typological prototypes of the

78

Cristiano Broccias

referring, attributive, and predicating constructions are the semantic classes of objects, properties, and actions, respectively, i.e., we prototypically refer to objects, modify cognitive files by attributing properties to entities and predicate actions of entities. It should be observed that these semantic classes are only a small subset of the semantic classes of words found in human languages (and are defined in terms of four semantic properties, namely relationality, stativity, transitoriness, gradability, see Croft 2001: 87). Parts-of-speech in a given language can therefore be represented onto a bi-dimensional space (a semantic map) defined by the dimensions “discourse functions” and “semantic classes.” Nouns, adjectives, and verbs are viewed as the (prototypical) pairings of reference/object, modification/property, and predication/action respectively. This is shown in Table 2 (adapted from Table 2.3 in Croft 2001: 88). “Unmarked” means that no derivational morphemes are employed, e.g., cat is an unmarked noun while happiness is a marked, deadjectival noun.2 Table 2. The two-dimensional space for English parts of speech

semantic class

discourse function reference

modification

predication

objects

UNMARKED NOUNS

genitive, adjectivalizations, PPs on nouns

predicate nominals, copulas

properties

deadjectival nouns

UNMARKED ADJECTIVES

predicate adjectives, copulas

actions

action nominals, complements, infinitives, gerunds

participles, relative clauses

UNMARKED VERBS


79

2.3. Blending Theory Fauconnier and Turner’s (2002) Blending Theory is not a theory of grammar as such. Rather, Fauconnier and Turner show that the basic conceptual operation of blending should be invoked to account for a variety of grammatical constructions, including “simple” syntactic patterns like adjectivenoun combinations (e.g., red pen, safe beach) and noun-noun combinations (e.g., land yacht). Structure from two (or more) input mental spaces is projected into a separate space, the blend, where it is integrated into a single conceptual unit. Crucially, the blend can develop structure of its own (i.e., structure which was not present in either input), known as emergent structure. Blending Theory points out that both in supposedly (semantically) simple cases like red pen and more exotic cases like land yacht (used to refer to a large, luxurious car) we resort to identical complex mental operations. Consider the noun-noun combination land yacht. The conceptual pole of this noun phrase results from the blending of two input spaces. One input involves water, a skipper, a course, a yacht, and a tycoon; the other space includes land, a driver, a road, a car, and an owner. These elements from the two inputs are put into correspondence with one another (so that water corresponds to land, skipper to driver, etc.) and some of them (e.g., tycoon/owner and car/yacht) are projected into the blend so as to create emergent structure: a land yacht names a new object, which is neither a generic car nor a yacht. Let us now analyse the adjective-noun combination red pen. This noun phrase is deceptively simple. It could mean a pen whose case is red, a pen whose ink is red, a pen whose cap is red, a pen “used to record the activities of a team dressed in red” (see Fauconnier and Turner 2002: 355), and so on. In other words, we integrate (as was the case with land yacht) two input spaces, one containing the colour red and the other containing the pen plus, if necessary, relevant frames (i.e., scenarios) in which the pen plays some role (e.g., a frame describing the use of the pen to record the activities of a team). The operation of conceptual integration takes place by virtue of the correspondences established between the two inputs (e.g., between the colour of the pen cap and the colour red) and by projecting the pen and the colour red into the blend. The fact that certain interpretations seem to be simpler and more accessible than others is simply due to the existence of strong defaults: “… simplicity of form, conventionality of vocabulary, and frequency of use [do not] indicate simplicity of underlying conceptual organization” (Fauconnier and Turner 2002: 365).

80

Cristiano Broccias

It should also be observed that conceptual integration can go hand in hand with formal integration. This is the case, for example, of the word Chunnel, which is used to refer to the underwater tunnel connecting England and France. Morphologically, tunnel and Channel (i.e., the English Channel) are blended into a single word by substituting the phoneme /æ/ of Channel with /8/ from tunnel. The formal integration mirrors the conceptual integration whereby the frame of a tunnel is put into correspondence with the frame of the specific stretch of water separating England and France. Fauconnier and Turner claim that blending can also occur between a construction and a diffuse (or unintegrated) input. They illustrate this point by analyzing the resultative construction (e.g., I boiled the pan dry) and the caused motion construction (i.e., Goldberg’s 1995 famous example He sneezed the napkin off the table). They regard specific instantiations of these two constructions as originating from the blending of an input space which contains the schematic construction (represented in a Goldbergian format as a pairing of semantics and syntax) and an input containing the actual event sequence of sneezing and the napkin’s falling off the table. Importantly, in the blend it is as if the subject referent acted directly on the napkin (or, in the resultative example, the pan), see Fauconnier and Turner (2002: 178–179). We compress the diffuse event sequence into a scenario where an agent acts on object thus causing its displacement (or change of state in a resultative example). In sum, one of the important lessons to be learnt from Blending Theory is that very complex conceptual operations can often (as shown for English in Section 4) be coded through apparently “simple” structures.

3. Do we need pedagogical grammar within a usage-based model of language? The development of (theoretical) cognitive linguistics, if set against the background of generative grammar, has a striking similarity with the recent history of language teaching. As was shown in the preceding sections, cognitive approaches to grammar are converging towards a usage-based network model of language, which is bottom-up (constructions or schemas emerge out of specific usage events as generalizations over them) and maximalist (specific instantiations as well as general schemas that subsume them are both postulated as being part of our knowledge of language). This model contrasts with the largely decontextualized view of language es-


81

poused by generative grammarians, where syntax is an autonomous component of the (autonomous) Language Faculty. The Chomskyan model is top-down (syntactic templates simply result out of the setting of languageuniversal parameters) and minimalist (redundancies are eliminated programmatically). Admittedly, however, even within the generative camp, researchers (e.g., Jackendoff and Culicover 2005) have stressed the impossibility of syntactic analyses which dispense with semantic/pragmatic considerations. That is, not only is a clear-cut distinction between lexicon and grammar suspicious (see also Culicover 1999) but the very notion of autonomous levels of linguistic analysis is probably untenable. In a similar vein, it is well-known that language teaching has moved from decontextualized drilling activities to more meaningful, communicative, and contextbased methods, i.e., what could be termed a usage-based model of language teaching (see, e.g., Harmer 2001 for a useful overview). There are further similarities between the development of cognitive linguistics and language teaching. For example, Willis’s lexical approach (e.g., Willis 1990; Lewis 1993, 1997) bears strong resemblance to constructivist approaches to grammar. In both cases, it is recognized that “language consists not of traditional grammar and vocabulary but often of multi-word prefabricated chunks” (Lewis 1997: 3), e.g., sentences like I’ll see what I can do, I’ll be back in a minute, etc., despite the fact that many of them are fully transparent in terms of their meaning (unlike idioms such as kick the bucket). Clearly, such “chunks” are located midway between what is traditionally assumed to be the lexicon (as a repository of unpredictable forms) and grammar (as a repository of highly schematic patterns). The development of cognitive approaches to grammar in terms of the usage-based model obviously raises the question of whether cognitive linguistics should therefore simply advocate “exposure to large quantities of natural speech in context … and significant immersion in the target language” (Chen 2004) in the case of second language acquisition. After all, massive exposure to language input is what first language acquisition depends on and it would be, at least counterintuitive, to think that this is not also relevant to second language acquisition (e.g., Ellis 2002), although fossilization (e.g., Selinker 1972) is a well-known phenomenon by which exposure to the “correct” input does not result in any changes in the nonnative’s output. In fact, the argument that explicit instruction may not be necessary (if not counterproductive) is also a well-known issue in the language teaching literature. For example, Allwright (1979: 170) famously claimed that if learners are involved “in solving communication problems in the target language, then language learning will take care of itself.”

82

Cristiano Broccias

It follows that the same objections that have been raised against the type of argument put forward by Allwright can be made in the case of cognitive approaches to grammar teaching. The discussion of language learning cannot be divorced from issues such as the age of the students, their level of proficiency, their motivation, their education, and the environment where a language is being learned (Harmer 2001: 72). Consequently, pedagogical grammar cannot easily be dispensed with. At the very least, pedagogical grammar may be conducive to what Schmidt (1990) calls “noticing,” i.e., the process of becoming aware of, e.g., constructions that differ importantly from language to language (see also Batstone 1994 and McCarthy and Carter 1995 on the importance of discovery in language learning). Having established that pedagogical grammar is unlikely to be dispensed with (even within a usage-based theory of language), we can now tackle the question of how the various cognitive approaches to grammar can be put into effect in pedagogical grammar teaching. In particular, I will discuss the notions of network, prototype, and construction and try to assess their relevance to pedagogical grammar. My main contention is that cognitive approaches to grammar may not necessarily result in a new methodology but may nevertheless highlight the importance of claims that have already appeared in the literature on language teaching in general and grammar teaching in particular.

4. Networks, prototypes and constructions It is usually argued that one of the major applications of cognitive grammar approaches to pedagogical grammar is the use of semantic networks (e.g., Tyler and Evans 2005). Langacker’s Cognitive Grammar and Goldberg’s Construction Grammar underline that constructions are related to one another by way of categorizing relationships or inheritance links. Hence, it is suggested that showing how, e.g., the various senses of prepositions (such as over) can be linked to a prototypical sense may help foreign learners of English, who would not simply have to rely on rote learning but could be aided by establishing conceptual links between apparently different uses. As was shown above, the use of networks is also common in the case of syntactic constructions (to be understood here in the traditional sense of combinations of syntactic phrases). Quite recently, for example, Goldberg (2006), relying on work carried out with Del Giudice (see Goldberg and Del Giudice 2005), has argued that so-called Subject-Auxiliary Inversion


83

constructions, see (4) below, form a natural category in that they can all be regarded as extensions from a prototype: (4)

a. b. c. d. e. f. g. h.

Did she go? Had she gone, they would be here by now. Seldom had she gone there … May a million fleas infest his armpits! Boy did she go! He was faster at it than was she. Neither do they vote. So does she.

Goldberg claims that all the types exemplified in (4) can be traced back to a prototype (for non-prototypical sentences) which is non-positive, nonpredicate focus, non-assertive, dependent, and non-declarative (for more details, see Goldberg 2006: Ch. 8). Although such attempts (both at word-level and at construction-level) at network building are most welcome (first of all from a theoretical point of view) because they show the non-arbitrariness of grammar, it remains to be established empirically whether they are indeed of crucial importance in second language acquisition. As is well known, Sandra and Rice (1995) have questioned network analyses (of prepositional meaning) in the sense that they observe that language users do make very general distinctions (e.g., temporal vs. spatial uses of prepositions) as well as fine-grained ones, but it is difficult to establish the exact level of granularity at which such distinctions are made. Further, the types of categorizing relations between the various senses of, e.g., prepositions may sometimes not stand up to closer scrutiny (as argued by Broccias 2005). This means that too much insistence on the use of network models in second language acquisition may not be warranted after all. We may impose a further burden on language learners, namely that of having to learn links (between senses) which may not be correct in the first place. A simple solution to this problem (the importance of inheritance links in our theoretical view of grammar and its actual relevance to language teaching) lies, I believe, in the very nature of Cognitive Linguistics. As was pointed out above, the model of language that emerges from the cognitive enterprise is one where both very specific schemas and general schemas coexist. Further, it has been shown in Section 2.2.1 that the creation of categories is influenced by factors such as frequency and order. From the point of view of pedagogical grammar, this means that learners (as is al-

84

Cristiano Broccias

ready the case in most contemporary teaching methods) should be exposed to a graded input, which first relies on numerous prototypical cases and then moves on to non-prototypical uses (thus enabling both the creation of low-level schemas for specific cases and general schemas capturing generalizations across the low-level schemas). There is no doubt, for example, that prototypical uses of prepositions (e.g., spatial meanings of over) or positive sentences (such as She went…) should be presented before nonprototypical cases. Still, we should not push too far the quest for motivation (for the existence of the non-prototypical cases) in language teaching. Rather, we should strive to point out non-prototypical uses to foreign learners (because foreign learners are most likely to get these wrong) reminding students that, although such uses may be motivated, it may be difficult to pinpoint the exact factors leading to their existence. This process of course would enhance students’ “attention” (and reduce students’ possible perception of the arbitrariness of language) but would not cancel the need for rote learning. As a matter of illustration, consider the uses of the preposition in in (5). We may well be able to motivate such uses, which usually cause problems for foreign learners since they are often realized with different prepositions in other languages (as indicated for Italian in parentheses). (5)

a. b. c. d.

I like walking in the rain. Italian: ‘sotto la pioggia,’ Lit. ‘under the rain’ Please fill in the form in ink. Italian: ‘a penna,’ Lit. ‘at (i.e., with) pen’ Who’s the girl in the pink top? Italian: ‘con la maglia rosa,’ Lit. ‘with the pink top’ He was talking in a stupid voice. Italian: ‘con una voce stupida,’ Lit. ‘with a stupid voice’

Still, even if motivation is provided for such cases (by noticing also that Italian often uses with where English has in), students still need to learn them by heart. That is, the real contribution of cognitive linguistics is here to focus students’ attention on troublesome patterns and to suggest that the patterns are motivated extensions from a prototype. But the exact details of the extension links are not of primary importance to foreign learners in that they should aim to access such uses automatically, i.e., without much constructive effort. Similarly, in the case of subject-auxiliary inversion, it may of course be of great help for students to recognize that various constructions exhibiting


85

inversion are related to each other to some degree but this does not mean that students will not have to be exposed to a great deal of examples to strengthen the entrenchment of cases like those illustrated in (4). To be sure, it should be pointed out that all the constructions in (4b)–(4h) have the same structure as interrogative sentences, see (4a), because this may help students remember the correct order of the elements in the inverted patterns. Indeed, in some cases, the notion of inheritance may prove of vital importance to help students come to terms with complex inversion structures like (6) below: (6)

The house was still. All London seemed still. Only, presently, did there rise, from the room below, the muffled throb of Mr Leonard’s murmur [emphasis in roman mine], which told her he was hard at work again … (Sarah Waters, 2006: 160)

The construction used in (6) results from the blending of two inversion constructions. One is of the type illustrated in (4), which requires subjectauxiliary inversion, and the other is the verb group-subject inversion investigated within a cognitive linguistic framework by, e.g., Chen (2003). The latter construction is usually used for presentational sentences (where a setting is introduced first and the primary figure is placed at the end of the sentence in order to foreground it), as is shown in (7): (7)

There rose from the room below the muffled throb …

Instead of a referential prepositional phrase at the beginning of the sentence (e.g., From the room below rose the muffled throb…), (7) contains “expletive” there (which could be said to refer to an abstract setting further specified later on in the sentence by from the room). This spatial-inversion construction is blended with the one illustrated in (4c) since it contains the restrictive adverb only. Consequently, an auxiliary (did) is placed after only. Interestingly, the inverted subject in (6) is there, i.e., there is treated structurally in the same way as in questions like Is there any sugar left?, although not all scholars may be comfortable with this analysis. Some may regard (at least semantically) the muffled throb as the subject (on the analysis of there as a subject see Langacker 1991: 351–355). In sum, cognitive models of grammar can contribute much to pedagogical grammar by leading learners to “noticing.” This is especially so in the case of English, where the existence of “idiosyncratic” constructions (traditionally understood) may pose more problems for learners than in other

86

Cristiano Broccias

languages (see also Goldberg 2006: 120). One more example will suffice. Consider (8) below, from the very beginning of J.K. Rowling’s Harry Potter and the Philosopher’s Stone (i.e., Harry Potter and the Sorcerer’s Stone in the American edition): (8)

… Mrs Dursley gossiped away happily as she wrestled screaming Dudley into his high chair. (Rowling 1997: 7–8)

(8) is a very interesting example because it blends various construction types in just two clauses. The first clause contains the verb-away construction which is used “with intransitive verbs … in order to indicate that the process or activity continues throughout a period of time” (see the Particles Index in Collins Cobuild Phrasal Verbs Dictionary (Moon 2002); examples are bang away, grind away, slog away, work away, talk away, etc.). The first clause also makes use of what could be termed the “verb -ly adverb construction,” which occurs very frequently in J.K. Rowling’s novels. It is a device which allows the author to describe in a very compact way what action is being carried out as well as the concurrent (esp. emotional) state in which its doer is (i.e., the manner in which the verbal event is carried out: Mrs Dursley seems to be happy while gossiping all the time, see e.g., Geuder 2000). The notion of simultaneity is also relevant to the use of the immediately following as-clause, which is one more pattern foreign learners of English may have problems with (as the author has noticed repeatedly, although informally, when teaching English to his students). Simultaneity between two events in English can be expressed in various ways (see Broccias 2006a, 2006b; Schmiedtová 2004), among which are the use of as-clauses and while-clauses. One of the striking facts about these two simultaneity clause types is that they tend to be used differently and that as-clauses tend to be more frequent (at least in novels) than while-clauses. In general, as and while-clauses reflect different construals of external reality. As-clauses are preferred with relatively short actions and/or when verbs of change (of either state or motion) are used. Since motion is evoked in the temporal clause, we “correctly” expect the use of as rather than while. Finally, (8) also contains a caused motion construction (within the as-clause). Wrestle is used as a verb of caused motion (because Dudley is caused to go into his high chair) although, on its own, wrestle may not necessarily be categorized as such. To conclude, examples like (6) and (8) clearly illustrate (in the case of the English language) the pedagogical grammar fallout of the attention paid by cognitive linguistics to “idiosyncratic” structures. The role of the


87

teacher here is that of a facilitator, i.e., s/he can highlight potentially problematic constructions to second language learners by pointing out how such constructions tie in with the grammar of English as a network of related constructions.

5. Conclusion It may be unlikely that cognitive linguistics will result in a radically new teaching methodology both in the specific case of pedagogical grammar and in language teaching in general. Rather, cognitive models of language highlight certain aspects of already existing language teaching methodologies which deserve further attention, namely: (i) the need for combining drilling activities (which are conducive to the entrenchment of prototypes) with communicative-based tasks when acquiring grammatical structures (this follows from Langacker's view of grammar as a structured inventory of conventional linguistic units, i.e., units shared by a sufficiently large number of speakers); (ii) the emphasis on the nature of language as a network (i.e., grammatical structures are related to each other); and (iii) the importance of motivation, construal, and blending in the shaping of grammar, which can be related to our basic physical/psychological experiences (although the exact details of how this is the case may be an open question). Having said this, we should however recognize that cognitive linguistics will be able to contribute enormously to pedagogical grammar (at least in the case of English) thanks to its attention to constructions, especially if they are understood traditionally as combinations of syntactic phrases. The proponents of the lexical approach have already commented on the importance of “word prefabricated chunks” in language learning. Cognitive linguistics provides theoretical and experimental evidence in support of this claim and, hence, we should expect future pedagogical grammars to include much more discussion of constructions (along the lines of the previous section, for example), rather than focussing, more traditionally, on words.

Notes 1.

A second reason why constructions emerge is that the use of a certain construction primes the use of the same construction (see Goldberg 2006: Ch. 6).

88 2.

Cristiano Broccias Table 2 is not the complete map of English part of speech constructions (see Figure 2.3 in Croft 2001: 99). But it will suffice for our present purposes.

References Allwright, Richard 1979 Language learning through communication practice. In The Communicative Approach to Language Teaching, Christopher Brumfit and Keith Johnson (eds.), 167–182. Oxford: Oxford University Press. Batstone, Rob 1994 Grammar. Oxford: Oxford University Press. Broccias, Cristiano 2003 The English Change Network: Forcing Changes into Schemas. Berlin/New York: Mouton de Gruyter. 2005 Review of Studies in Linguistic Motivation by Günter Radden and Klaus-Uwe Panther (eds.) (2004). LINGUIST List Vol-16-872. Tue Mar 22 2005. 2006a The construal of simultaneity in English with special reference to asclauses. Annual Review of Cognitive Linguistics 4: 97–133. 2006b The English simultaneity network: The case of as and while-clauses. LACUS Forum XXXII: 33–41. 2006c Cognitive approaches to grammar. In Cognitive Linguistics: Current Applications and Future Perspectives, Gitte Kristiansen, Michel Achard, René Dirven, and Francisco J. Ruiz de Mendoza Ibáñez (eds.), 81–115. Berlin/New York: Mouton de Gruyter. Chen, Liang 2004 Review of Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching, Michel Achard and Susanne Niemeier (eds.) (2004). LINGUIST List: Vol-15-1868. Sat Jun 19 2004. Chen, Rong 2003 Inversion: A Ground-before-Figure Construction. Berlin/New York: Mouton de Gruyter. Croft, William 2001 Radical Construction Grammar: Syntactic Theory in Typological Perspective. Oxford: Oxford University Press. Croft, William, and D. Alan Cruse 2004 Cognitive Linguistics. Cambridge: Cambridge University Press. Culicover, Peter 1999 Syntactic Nuts: Hard Cases in Syntax. Oxford: Oxford University Press.

Cognitive linguistic theories of grammar and grammar teaching Ellis, Rod 2002

89

The place of grammar instruction in the second/foreign language curriculum. In New Perspectives on Grammar Teaching in Second Language Classrooms, Eli Hinkel and Sandra Fotos (eds.), 17–34. Mahwah, New Jersey: Lawrence Erlbaum. Fauconnier, Gilles, and Mark Turner 2002 The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities. New York: Basic Books. Geuder, Wilhelm 2000 Oriented Adverbs: Issues in the Lexical Semantics of Event Adverbs. PhD dissertation, Universität Tübingen. Goldberg, Adele E. 1995 Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press. 2006 Constructions at Work: The Nature of Generalization in Language. Oxford: Oxford University Press. Goldberg, Adele E., and Alex Del Giudice 2005 Subject-auxiliary inversion: A natural category. The Linguistic Review 22: 411–428. Gries, Stefan, and Stefanie Wulff 2005 Do foreign language learners also have constructions? Evidence from priming, sorting, and corpora. Annual Review of Cognitive Linguistics 3: 182–200. Harmer, Jeremy 2001 The Practice of English Language Teaching. Harlow: Longman. Jackendoff, Ray, and Peter Culicover 2005 Simpler Syntax. Oxford: Oxford University Press. Langacker, Ronald 1987 Foundations of Cognitive Grammar. Vol. 1: Theoretical Prerequisites. Stanford, California: Stanford University Press. 1991 Foundations of Cognitive Grammar. Vol. 2: Descriptive Application. Stanford, California: Stanford University Press. 2005 Construction grammars: Cognitive, radical, and less so. In Cognitive Linguistics: Internal Dynamics and Interdisciplinary Interaction, Francisco J. Ruiz de Mendoza Ibáñez and M. Sandra Peña Cervel (eds.), 101–159. Berlin/New York: Mouton de Gruyter. Lewis, Michael 1993 The Lexical Approach. Hove, United Kingdom: Language Teaching Publications. 1997 Implementing the Lexical Approach. Hove, United Kingdom: Language Teaching Publications.

90

Cristiano Broccias

McCarthy, Michael, and Ronald Carter 1995 Spoken grammar: What is it and how can we teach it? ELT Journal 49: 207–218. Moon, Rosamund (ed.) 2002 Collins Cobuild Phrasal Verb Dictionary. Glasgow: Harper Collins Publishers. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik 1985 A Comprehensive Grammar of the English Language. London: Longman. Radford, Andrew 2004 Minimalist Syntax: Exploring the Structure of English. Cambridge: Cambridge University Press. Rowling, Joanne K. 1997 Harry Potter and the Philosopher’s Stone. London: Bloomsbury. Sandra, Dominiek, and Sally Rice 1995 Network analyses of prepositional meaning: Mirroring whose mind – the linguist’s or the language user’s? Cognitive Linguistics 6: 89– 130. Schmidt, Richard 1990 The role of consciousness in second language learning. Applied Linguistics 11: 129–158. Schmiedtová, Barbara 2004 At the Same Time … The Expression of Simultaneity in Learner Varieties. Berlin/New York: Mouton de Gruyter. Selinker, Larry 1972 Interlanguage. International Review of Applied Linguistics in Language Teaching 10: 209–231. Tomasello, Michael 2003 Constructing a Language: A Usage-Based Theory of Language Acquisition. Harvard: Harvard University Press. Tyler, Andrea, and Vyvyan Evans 2005 Applying cognitive linguistics to pedagogical grammar: The English prepositions of verticality. Revista Brasileira de Linguistica Aplicada 5: 11–42. Ungerer, Friedrich, and Hans-Jörg Schmid 2006 An Introduction to Cognitive Linguistics. London: Longman. First published 1996. Waters, Sarah 2006 The Night Watch. London: Virago. Willis, Dave 1990 The Lexical Syllabus. Glasgow: Collins Cobuild.

Corpora, cognition and pedagogical grammars: An account of convergences and divergences Fanny Meunier

Abstract The present paper addresses the convergences and divergences between cognitive and corpus linguistics with a view to determining if and how those two fields could meet not only in describing and analysing English grammar, but also in integrating the results of the analyses in pedagogical grammars. The presentation of the converging and diverging forces is followed by a discussion of the major obstacles to writing corpus- and cognitively-informed pedagogical grammars. The final section discusses future prospects and challenges in the writing of such grammars by using a question-and-answer approach. The notions of principled eclecticism and prioritization are presented, together with some methodological guidelines and illustrations. Keywords: corpus linguistics; cognitive linguistics; pedagogical grammar; principled eclecticism; prioritization

1. Introduction This article aims to outline the convergences and divergences between cognitive and corpus linguistics approaches to language within a pedagogical perspective. The applied aspect of the present volume being that of a pedagogical grammar, the core questions can therefore be stated as follows: First, under what conditions can cognitive linguistics and corpusbased studies meet in analysing English grammar? And secondly, how can their respective findings be merged to help learners of a second or foreign language get to grips with aspects of English grammar? To address these issues,1 Sections 2 and 3 sketch out the similarities and differences between the two fields. Section 4 consists in a critical account of the obstacles that may hinder the actual writing of corpus- and cognitively-informed pedagogical grammars. In the light of the findings presented, the concluding section aims to answer the research questions using

92

Fanny Meunier

a Socratic dialogue approach and to propose methodological guidelines for the writing of pedagogical grammars.

2. An account of converging forces To start with, it is worth noting that corpus linguistics and cognitive linguistics are relatively recent fields,2 which both burgeoned and came into bloom in the early 1980s, and which also happen to share the same abbreviation (i.e., CL). The convergences between the two fields go, however, far beyond these incidental facts. Stathi (2006: 27) states for instance that “both fields share basic assumptions in terms of adherence to a usagebased approach to language, that they demonstrate a lack of commitment to strict dichotomies …, and that they assign multi-word expressions a much more prominent role in linguistic theory than many other frameworks.” It could also be added that the adherence of the two fields to a usage-based approach to language somehow predisposes them to impact on domains such as language acquisition and language pedagogy. Other similarities include what I would call the central role of semantics (although each field has a different history in that respect, as will be shown below), the notion of a grammar of choice and, finally, a mutual curiosity about what each field can bring to the other. These converging forces are expanded, illustrated and commented on in the following paragraphs. On the cognitive side, the focus on usage-based approaches is evidenced, among others, by Langacker’s (1987) seminal work on cognitive grammar, where linguistic knowledge is seen as acquired through actual use (thereby rejecting mentalist approaches) and where the acquisition through use enables learners to generalize bits of knowledge thanks to recurring experiences of use. As for corpus linguistics, the importance of language use is not initially linked to acquisition but rather to language description. Corpus linguistics constitutes an increasingly popular method of analyzing language which exclusively uses authentic data as its raw material. Although the two fields have different aims (explanation vs. description, see Section 3), they can both be described as what Taylor (2002) calls bottom-up approaches to language, i.e., fundamentally empiricist3 in nature and basically starting from the analysis of samples of language use in order to describe and understand a more complex object or system. A key notion directly linked to what precedes is that of frequency. Again, while the interpretations of that notion vary from one field to another (see Section 3), recurrence of “language facts” nevertheless consti-

Corpora, cognition and pedagogical grammars

93

tutes a core issue for the two CLs. In cognitive linguistics “entrenchment” or “conventionalizations” are essential criteria, and frequency is both a possible impulse and consequence of those two criteria. Recurrence is often attributed a reinforcing effect, which leads to the entrenchment of linguistic expressions (established concepts are entrenched cognitive routines that are easily evoked), while lack of recurrence leads to decay (Langacker 1987: 100, 162). Here again the explanatory power of the cognitive approach is evident. In corpus linguistics, on the other hand, recurrence is rather seen as a key to typicality, and forms the basis of analyses of lexical and/or grammatical features of varieties of the language. Typical illustrations of that approach are Biber’s (1988) work on variation across speech and writing, or the corpus-based Longman Grammar of Spoken and Written English (Biber et al. 1999), which is a reference grammar of English aiming to describe the variety-dependent features of four core registers: conversation, fiction, newspaper language and academic prose. Coming back to the lack of commitment to strict dichotomies mentioned at the beginning of this section, it can be argued that this feature is probably linked to the inherently empiricist nature of the two CLs, both starting from the data first and reflecting only at a later stage on more theoretical issues. Both fields, for instance, reject the competence vs. performance dichotomy and postulate that also a usage-based approach will give partial access to competence as instantiated through performance data. As for the lexis-grammar dichotomy, the two CLs have greatly contributed to revising the lexis-grammar continuum. Corpus linguistics has promoted the lexicogrammatical analysis of language and has contributed to the discovery that a large part of the language can be said to be idiomatic, prefabricated or phraseological (to use only some of the most common terms found in the literature). Some of the most famous proponents of that trend in corpus linguistics include Sinclair (1991) and his open choice vs. idiom principle, Hoey (1998) and his work on colligation4 or Hunston and Francis’s (2000) work on pattern grammar.5 The analysis of mega-sized corpora, such as the 100-million-word British National Corpus,6 with the help of powerful text retrieval facilities has revealed the existence of patterns that introspection or intuition could not predict. The representativeness of recent corpora (i.e., the fact that they include written and spoken samples of numerous text types meant to be representative of a language as a whole7) has also revealed the close link of certain patterns to certain text types. Different types of communication make use of, among other things, different grammatical structures and can be characterized in terms of those grammar items (see for further illustrations Biber 1988; Biber, Conrad and Reppen

94

Fanny Meunier

1994; Conrad 1996a, 1996b; Grabe 1987), thereby leading to the “grammar of choice” idea (Larsen-Freeman 2002), which postulates that while the same idea can potentially be expressed by means of several different grammatical realizations, only some are preferred in specific contexts. The commitment of cognitive linguistics to the idea of communicative interaction patterns and to new perspectives on the lexis-grammar interface is also very strong. Langacker’s work (1987, 1999) posits that conceptual semantics, or content, is clearly primary to form. As for Paradis (2003), she states that the core idea in the cognitive school is that meanings of linguistic expressions are mental entities, and that semantics is the mapping of linguistic expressions to conceptual structure. The Hallidayan concept of choice or option is viewed in a more conceptual sense in cognitive linguistics, as illustrated by the notion of “construal” which means that each option conceptualizes, views, and expresses the given scene in a different way. Croft and Cruse’s (2004) notion of dynamic construal of meaning in situations of use refers to the fact that elements of meaning only receive a particular interpretation in a particular context and according to several constraints. Lexis and grammar, or semantics and syntax, can thus not be analysed in complete isolation. Similarly, Chen and Oller (this volume: 386) argue that “[a]ny particular event can be construed from many different perspectives, and may be expressed in a range of linguistic structures. The actual expression of the event depends on how the speaker conceptualizes the event and chooses to communicate it linguistically.” Again here, both CLs display convergences in stressing the importance of a linguistic analysis which goes, to quote Fitzpatrick (2007), “beyond the word.” As for the influence of the two CLs on research in second language acquisition (SLA) and pedagogy, I will first refer to corpus linguistics and the numerous links made between the results of corpus studies and their implications both for SLA studies and for the teaching of second or foreign languages. To name only but a couple of references addressing the corpusSLA link, I will mention Granger, Hung and Petch-Tyson (2002), who present a series of articles on computer learner corpora in second language acquisition and foreign language teaching (FLT), and MacWhinney’s volumes (2000), which focus on the CHILDES project (Child Language Data Exchange System) and address the potential of using CHILDES in SLA research. For more practical and pedagogically oriented applications in the field of corpus linguistics, I will refer the reader to Ketteman and Marko (2002), who offer a survey of issues on teaching and learning by doing corpus linguistics; to Burnard and McEnery (2000), Sinclair (2004) and Connor and Upton (2004), who provide comprehensive edited volumes on


95

the use of corpora in language teaching and learning, and finally to O’Keeffe, McCarthy, and Carter’s (2007) recent book, which addresses the steps needed to move “from corpus to classroom” (as the title of the book suggests) and illustrates the necessary mediation between research findings in corpus linguistics and classroom pedagogy. Coming back to cognitive linguistics, Pütz, Niemeier, and Dirven (2001) stress the relevance of its theoretical views for various fields of applied linguistics, primarily in the areas of language acquisition, learning and pedagogy. Ellis (2006) provides the reader with a detailed account of cognitive perspectives on SLA, and Achard and Niemeier (2004) present an edited volume on Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching. As for the links between cognitive linguistics and second and foreign language teaching and pedagogy, the main focus so far has been on semantically-oriented topics. Boers (2003) and Boers, Eyckmans, and Stengers (2007) focus for instance on lexical issues, such as metaphors and idioms. In terms of pedagogical grammar per se, despite Dirven (1989), Dirven and Taylor (1994) and Tyler and Evans (2004), far fewer publications are available. The potential of a cognitive approach to pedagogical grammar has, however, been clearly put forward by Evans and Tyler (2005), who argue that a cognitive linguistics approach to language analysis provides a more accurate and systematic account that, in turn, offers the basis for a more coherent, learnable presentation of the hitherto seemingly arbitrary aspect of English grammar under study. Similarly, De Knop and Dirven (this volume) state that FLT has lacked a strong theoretical linguistic framework, both in the past and today, to serve as a central pillar to support the teaching practice of all the various aspects of language … A serious candidate offering this all-embracing linguistic theory is cognitive linguistics, more specifically cognitive grammar. A last factor of convergence mentioned earlier is the mutual curiosity about what each field can bring to the other, and also perhaps the conviction that cross-fertilization is possible and useful. Recent publications pay tribute to this cross-fertilization: Gries, Hampe, and Schönefeld (2005) claim that usage-based approaches encompassing corpus-based and experimental perspectives result in a substantial increase in descriptive and explanatory reliability and validity; Taylor (this volume) argues that an account of the target language needs to be grounded in corpus data; Mukherjee (2004) shows the potential of corpus data in general in a usagebased cognitive grammar; and Gilquin (2004) carries out a corpus-based and cognitive study of the main English causative verbs. A last illustration

96

Fanny Meunier

is Stathi’s (2006) analysis of idioms entitled “Corpus Linguistics meets Cognitive Linguistics: A framework for the analysis of idioms.” Summarizing the converging forces that attract corpus linguistics and cognitive linguistics we can say that they both share a usage-based approach, a rejection of strict dichotomies, an interest in going beyond the word level, a belief in a grammar of choice motivated by contextual factors, an impact on SLA and language pedagogy, and a mutual interest in each other’s potential. It would be unfair, however, to deduce that those two fields have so much in common that they are hardly distinguishable. Section 3 aims to present some divergences between corpus linguistics and cognitive linguistics.

3. An account of diverging forces The common adherence to usage-based approaches is for instance interpreted differently in each field. As Ellis argues (2007: 24), all usage-based approaches “view the regularities of language as emergent phenomena: the rule-like regularities captured by linguists are mere descriptions, explananda [things to be explained] not explanans” or rather “explanantes” [explaining statements]. Cognitive linguistics, however, primarily exploits the heuristic power of language use for the sake of its explanatory value in terms of its conceptual potential (entrenchment of given conceptualizations and conventionalization of form/meaning pairings), which is an approach corpus linguistics does not really explore. In a lexically-oriented analysis of cognition vs. corpus, Gries (2006: 57) states that while many recent cognitive linguistic approaches to polysemy have concerned themselves with polysemous words as network-like categories with many interrelated senses (with varying degrees of commitment to mental representations), corpus linguistic approaches have remained rather agnostic as to how different word senses are related and have rather focused on distributional characteristics of different word senses.

This quote could easily be applied to grammatical aspects. Variationist approaches to grammar abound in corpus linguistics but the knowledge of how and why different grammatical structures are related and/or linked to specific varieties of the language is rarely addressed. Another difference underlying the reference to “usage” is that while cognitive linguistics uses the whole array of what Ellis (1994) labels “lan-


97

guage use data,” i.e., natural, elicited, clinical and experimental data, corpus linguistics8 works exclusively with natural language use data. As for the divergences linked to the notions of frequency and recurrence, they can be illustrated using a qualitative and quantitative metaphor. Generally, corpus linguistics is described as a quantitative and methodological approach to language, using a large amount of figures, percentages, frequency counts, measures of variance and co-variance, mutual information measures, and all sorts of statistical tests and validations (for an introduction to statistics in corpus linguistics, see Oakes 1998). Corpus linguistics aims at what McEnery, Xiao, and Tono (2006: 137) label a “theory of the typical,” in which “priority is given to describing the commonest uses of the commonest words,” together with the commonest lexical and/or grammatical patterns (see, for instance, Sinclair 1987a, 1987b or Francis, Hunston, and Manning 1998). Work in cognitive linguistics also invokes usage events and frequency issues, as amply demonstrated by Tomasello (2000), who stresses the key role of input frequency in the child language acquisition process. The notion of frequency is however used primarily here for its genetic aspect, i.e., the origin of given form/meaning pairings in language acquisition, a notion which is usually not addressed in corpus linguistics. Another difference between the two CLs is that corpus linguistics has what I would call a “stricter approach” to data description. It is common practice in corpus linguistics to refer precisely to the origin of the data, and to provide statistics such as the exact number of words examined or else the proportion of a certain feature within the total amount of data being analysed. Cognitive linguistics often seems to have a much looser approach to data (not in terms of the accuracy of description of the data but rather in terms of description of the sources). A phrase that I often came across in many cognitively-oriented publications was: “consider the following examples.” In most cases, it was left unclear, however, where those examples came from, how they had been collected, in what contexts they had been used by whom and how frequently they occurred. Going against this trend, cognitive linguistics is often reported as being a qualitative approach to language. Ellis (1999: 27), for one, states that cognitive linguistics “provides detailed qualitative analyses of the ways in which language is grounded in our experience and our embodiment, an embodiment which represents the world in a very particular way.” From a quite different viewpoint, i.e., that of the ethnography of communication, Hymes (1972) developed the notion of communicative competence as a correction to Chomsky’s narrow definition of linguistic competence.

98

Fanny Meunier

Communicative competence is not only about whether a particular utterance is formally possible, but also whether it is contextually appropriate and actually attested. One could relate corpus linguistics mainly to the attested occurrences and cognitive linguistics to the contextually appropriate ones. That leaves us with the formally possible sentences, traditionally the research limit of the generativist school and the Chomskyian framework. It would, however, be misleading and unfair to equate corpus studies to quantity and cognitive ones to quality. Widdowson (2000: 3) argues, for instance, that it is thanks to those quantitative measures, that corpus analysis has made it possible to discover a “reality about language hitherto not evident to its users.” Corpus studies have had an enormous impact on linguistic theory (as was apparent from the earlier discussion on the lexisgrammar interface for instance) and classroom methodology (see work on data-driven learning by Johns 2002). As for studies carried out in a cognitive linguistics framework, an increasing use of quantitative methods is being made in order to fine-tune the analysis (see for instance Stefanowitsch and Gries’s statistical measures for collostructional analysis, 2003). It would thus seem particularly well-suited, for fields which have a tendency to reject dichotomous approaches, to speak of a quantitativequalitative cline rather than a quantitative-qualitative dualism. Another issue I would like to mention is related to the attested prevalence of semantics over form (referred to in Section 2), which has a different history in each field. While cognitive linguistics had an almost innate interest in semantics, corpus linguistics somehow acquired it with time. To sum things up (at the risk of simplification), while the early stages of corpus linguistics focused essentially on lexis and single words (be they raw or part-of-speech tagged, see Meunier (1998) for an introduction to tools and methods in corpus linguistics), the focus has shifted to multi-words, to patterns and to their semantic motivation in a variationist perspective. As Teuberg (2001: 1) put it, “now that corpus linguistics has come of age, it is time for a broader programme.” This age of maturity of corpus linguistics has led to a greater awareness of its potential strengths (and also weaknesses) and corpus methods are now being used in many areas, such as SLA, psycholinguistics, sociolinguistics, pragmatics, cognitive linguistics, translation studies, literary studies, forensic linguistics, natural language processing, etc. (hence the “methodology” label often attributed to corpus linguistics). A last point of divergence worth commenting on is the notion of prototypicality vs. frequency. Gilquin (2006: 1), in her analysis of prototypical-


99

ity, argues that “when it is not used informally as a synonym of “typicality,” the term “prototypicality” refers to some intuitive assumption that a particular item is somehow more central to the category than other items.” She also states that most of the prototypes described by cognitive linguists are of this type while in corpus linguistics “it is not uncommon to equate prototypicality with frequency, the prototype being the most frequent item of a category in language (e.g., Stubbs 2004).” Stressing the fact that prototypicality is a psychological phenomenon, she then wonders whether frequency has any psychological reality when it comes to prototypes. After comparing different types of evidence, she concludes (2006: 2) that “contrary to common belief … frequency as attested in the corpus does not [always, my addition] coincide with salience as evidenced by the production and comprehension tests, a result which confirms some studies underlining the difference between corpus data and experimental data (cf. Roland and Jurafsky 2002 or Shortall, forthcoming).” This important issue will be further expanded in Section 5 where the combination of cognitive and corpus insights are applied to pedagogy. As has been discussed corpus linguistics and cognitive linguistics sometimes diverge in their descriptions and interpretations of linguistic facts (and perhaps also in their interpretation of what constitutes a linguistic fact) and in the methods used to reveal those facts. It appears, however, that such divergences may offer complementary views of the language. Before suggesting ways of integrating those two approaches – to the learner’s benefit – in a pedagogical grammar, the pedagogical objectives put forward in the corpus and cognitive linguistics literature will be examined in the light of their actual realizations.

4. Major obstacles to writing CL-based pedagogical grammars In this section, the pedagogical objectives that can be derived from the two CLs will be compared to the actual realizations in the domain of pedagogical grammars. While traditional pedagogical grammars of English abound, corpus-based or cognitive pedagogical grammars are few and far between. A number of factors can account for this lack of available materials. The two fields are still relatively young, and the way they view grammar does not correspond to the representations of grammar as typically found in traditional foreign language text books. Both CLs view grammar as going beyond the word level and, more importantly, also as going beyond the sentence level by taking contextual factors into account. While many aca-

100

Fanny Meunier

demic publications in corpus linguistics and in cognitive linguistics have addressed specific grammar topics, all-encompassing corpus-based or cognitive pedagogical grammars are not yet available. Three main obstacles to the integration of corpus and cognitive aspects can be listed here: first, while it may be relatively easy to include corpus-based research and/or corpus-based examples in grammars, it turns out to be more difficult to explain the emerging patterns or real-life data in cognitive terms; second, the never-ending debate on the value of authentic versus invented examples in pedagogical grammars may slow down the corpus/cognition enterprise; and third, learner’s expectations may diverge quite radically from what a corpus and/or cognitive pedagogical grammar may offer. These obstacles will be further illustrated in the coming paragraphs. To date, few grammar-oriented pedagogical applications have benefited from the results of corpus studies. Referring to the actual influence of corpus linguistics on pedagogical applications, Mair, in a panel discussion at the 2004 ICAME conference (reported in Aarts 2006: 397–398) mentions that a converse development to what happened in lexicography can be observed. He adds that: “Whilst almost all major dictionary makers have followed the COBUILD team in its pioneering effort to put lexicographical description on a solid corpus foundation …, comprehensive corpus-based works of grammatical reference are still rare.” Although Mair refers mainly to reference grammars such as The Longman Grammar of Spoken and Written English (Biber et al. 1999) in his argument, the situation is very similar for pedagogical grammars. Mair (in Aarts 2006: 397–398) goes on to say that while one usually turns to one’s reference grammar for clear and concise (if simplified) information on a point of linguistic usage, corpus-based work on English grammar tends to cater to linguists who wish to gain an idea of the complexity of grammar-in-context, of trade-offs between grammatical correctness and contextually motivated deviation, or the many other types of interdependence between form and meaning, or abstract pattern and concrete use.

Mair’s comments point towards one of the core challenges that both a corpus and/or a cognitive approach to pedagogical grammar face: viz. finding the trade-off between two antonymous goals, i.e., a desire to do justice to the richness of the data and a need for generalization. One such concrete attempt to combine a corpus and cognitive approach is Exploring Grammar in Context (Carter, Hughes, and McCarthy 2000).


101

The book is described9 as a unique reference and practice grammar, based entirely on real-life spoken and written language. These examples have been taken from the Cambridge International Corpus of English, “a stateof-the-art, multi-million word computerized corpus,” as the back cover puts it. There is also a good deal of spoken English with examples from the Cambridge and Nottingham Corpus of Discourse English (CANCODE). Despite its dual nature as both a reference grammar and a practice grammar, it contains the characteristics of a pedagogical grammar as defined by Greenbaum (1987). Exploring Grammar in Context is a course book grammar that can be used either in class under the control of a teacher or for self-study at home but it also contains shorter sections than a reference grammar and minimizes the use of metalanguage. This book also follows some of the recommendations made by Greenbaum in terms of the content and structure of an ideal pedagogical grammar: it considers the level of the user, is graded in various respects (i.e., the most frequent and simplest patterns should come first with additional and/or more complex patterns coming later), it anticipates the errors and it displays use in context. As for the cognitive aspects of Carter et al.’s book, it must be acknowledged that these are not explicitly addressed by the authors. Still, they partly appear from their analyses of why language users choose certain constructions in particular situations rather than others and also from the comments that they make on the effect of those choices on meaning. This is in line with what in cognitive linguistics is caught under the concept of construal. Carter et al.’s book tries to engage the reader in discovering patterns of grammar in use and in reflecting on the fact that elements of meaning receive a particular interpretation in a particular context and according to several constraints, as illustrated by the examples below. In the section devoted to discovering usage patterns of the passives and pseudopassives, the authors (Carter, Hughes, and McCarthy 2000: 97) indicate that “when there is an understanding of who carried out the action it is incorrect to add the agent,” such as in Rice is grown in India by Indian people. Although the use of the term “incorrect” could be seen as rather strong and that of “understanding” as rather vague, the comment indirectly refers to the fact that the choice between the different kinds of passive is a matter of event perspective (see, e.g., Chen and Oller, this volume). Another illustration of the importance of the context and the relationships between speaker and listener is found in the section on discourse markers (Carter, Hughes, and McCarthy 2000: 176), where the following observations are made: “Markers such as you know check that your listener understands you and that you both share the same viewpoint” or “Markers such

102

Fanny Meunier

as I don’t know and I think are sensitive to listeners and tend to soften opinions.” A last illustration of the authors’ desire to go beyond simple rule description comes from a section on tags, where they comment on the sentence ‘She is lovely, she is’ by saying that (Carter, Hughes, and McCarthy 2000: 184): … two positive clauses are found together. Such forms do not necessarily demand a reply; instead they often serve just to establish a shared mutual view of things. Tags are an essential feature of grammar in use in informal and intimate contexts of interactions and are particularly appropriate to contexts in which meanings are not simply stated but are negotiated and renegotiated.

While the cognitive character of the examples given here above may be said to be indirect, imprecise, or to pertain to just a few aspects of cognitive grammar, Carter et al.’s book nevertheless constitutes a first attempt to integrate the discovery, description and contextual motivation of patterns of grammar in use. The book also contains practice exercises to check understanding of these patterns. A second problematic issue in pedagogical grammars is the debate between authentic vs. invented or adapted language. Some insist on the fact that learners should be presented with authentic material exclusively (see Römer 2004, 2006) while some others plead for the use of easy-tounderstand prototypical, didactic and learner-centred examples (see Widdowson 2003). Römer’s view is in accordance with VanPatten (2002), who argues that samples of authentic language used among native speakers should be available from the beginning of instruction, and that this is, in fact, one of the fundamental tenets of communicative language teaching. A more balanced view can be found in Valdman (2003: 61). In his definition of pedagogical norms, he states that those norms should reflect the actual speech of target language speakers in authentic communicative situations, should conform to native speakers’ idealized view of their speech use, should conform to expectations of both [my emphasis] native speakers and foreign learners concerning the type of behaviour appropriate for foreign learners, and should take into account processing and learning factors. Reflecting actual speech does not necessarily mean copying the rough data as such, but adapting them by taking into account processing and learning factors. This means carrying out possible simplifications of the language if need be, thereby addressing what Long (1997) calls the distinction between potentially universal methodological principles (i.e., principles motivated


103

by research findings in SLA and cognitive science) and particular pedagogical procedures (i.e., procedures that realize the principles at the local level by taking into account learner- and teacher-centred factors). A third obstacle to the use of CL-based pedagogical grammars is the gap that may exist between the expectations of learners (in my experience10, many learners today still express a need for short, clear-cut, easyto-understand explanations and rules of grammar) and the desire of corpus or cognitive linguists to do justice to the way language works beyond the word level. I must admit that the classroom use of Exploring Grammar in Context (Carter, Hughes, and McCarthy 2000) did not prove too successful in my case.11 Learners were not at ease with what was considered by them to be a new type of grammar book which, to quote from their assessment forms, “lacked structure,” “was messy,” turned out to be too timeconsuming (expressed in comments such as “I would have preferred clearly presented rules instead of looking for the rules myself and having to reconstruct the theory”), and was overall too complex (“it is too difficult to remember when certain structures must be used in some contexts and not in others”). Apparently, learners have the impression that a variationist and contextually-oriented approach increases the number of things to be remembered. Comments like these also show that learners seem to expect decontextualized rules (possibly followed by a few exceptions) and that their expectations and views of grammar teaching are still very traditional and, probably, largely teaching-induced too. The challenge for both corpus and cognitive linguistics is to introduce learners’ to more variationist and cognitive approaches to grammar with the aim of facilitating both comprehension and production. Now that the converging and diverging forces of corpus and cognitive linguistics have been examined and that some of the problems related to the writing of a pedagogical grammar have been outlined, it is high time to come back to the two questions raised at the beginning of the present article, i.e., the conditions under which the two CLs can meet in analysing grammar, and the ways in which the findings of the two CLs can be integrated in pedagogical grammars.

5. Corpora, cognition and pedagogical grammars: Addressing methodological issues to promote collaboration The first question at the beginning of the paper was related to the conditions under which cognitive linguistics and corpus-based studies could

104

Fanny Meunier

meet in analysing English grammar. The fact that the two CLs display convergences (as shown in Section 2) and no irreconcilable divergences (see Section 3) constitutes the first two conditions for collaboration. A third condition could be described as the need to provide new perspectives on grammar, and here again the two CLs have greatly contributed to revealing new views on grammar. Other (and still to be met) conditions include the need to popularize corpus and cognitive findings, and to promote corpus literacy for teachers and learners of foreign languages. As for the second question, this was how the respective findings of the two CLs can be merged to help L2 learners get to grips with aspects of English grammar. To answer this, I will adopt a Socratic dialogue approach and posit what I believe to be two essential criteria, i.e., principled eclecticism and prioritization. Principled eclecticism usually refers to the use (and blending) of various teaching styles in a discriminating manner as required by learner needs and styles.12 As the present paper mainly deals with descriptive issues rather than teaching styles, principled eclecticism will also refer to the integration of different ways of representing grammatical aspects of the language and to providing the learners with access to those different representations. The second criterion, i.e., prioritization, should be understood as the concept of selecting, organizing, presenting or arranging items according to priority. As far as the operationalization of this concept of priority is concerned, it could encompass a number of quantitative and qualitative indicators. Quantitative indicators could include the frequency of a certain structure in the language and/or the range of registers or text types in which the structure is found. Qualitative indicators could include users’ needs (survival English, English for academic purposes, business English, English as a lingua franca, etc), levels of proficiency, etc. Those two key criteria will be illustrated in the questions and answers paragraphs. Question 1: Should a pedagogical grammar describe the formally possible, the contextually appropriate or the actually attested, as defined by Hymes (1972) (see Section 3)? Answer 1: A pedagogical grammar should ideally prioritize the attested and the contextually appropriate (which are essential both for productive and receptive skills), without necessarily ignoring the formally possible. It would make sense for a pedagogical grammar not to mention extremely rare uses, although some might argue that those can be useful, especially for receptive purposes. Corpus findings could be used to grade the various


105

options in terms of frequency of attested use. In the case of English for general purposes, one way of carrying out this almost impossible task (as the English language is not monolithic but consists of a variety of varieties) is to combine frequency and range, that is, to make sure that the examples selected are not only frequent but also appear in many different text types in order to ensure representativeness. In the case of English for specific purposes, the task is somehow easier as the work can be carried out on a very specific type of corpus (e.g., Airspeak). A concrete example of what corpus findings can bring to the description of a grammatical structure is Gabrielatos (2003), who compared the traditional English Language Teaching (ELT) typology of conditional clauses against corpus evidence. He showed that traditional ELT materials essentially use the typology in logic, usually presented as follows: Type 1 conditional (real conditional) e.g., If it rains, I will take my umbrella; Type 2 conditional (hypothetical conditional) e.g., If I won the lottery, I would buy a new house; and Type 3 conditional (counterfactual conditional) e.g., If you had told me the truth, I would never have criticized you), with the addition of two more types, zero and mixed conditionals, and give the status of exceptions to all other cases. One of the findings of Gabrielatos’s study is that 55% of the authentic corpus examples did not fit into the core ELT typology13 illustrated here above. Such results justify the claim that pedagogical grammars cannot blindly rely on descriptive grammars but must first of all evaluate them (see Dirven 1989). In that respect, corpus studies may help both refine the description of grammatical aspects and change the way the latter are usually presented. Cognitive insights also tend to be absent from most traditional ELT grammars, but it must be acknowledged that these never set out to be cognitively motivated. Coming back to the treatment of the conditional, many pedagogical grammars still refer to Type 1, Type 2, Type 3 conditionals without trying to link those types to semantic motivation or to why the expression of condition is organized the way it is. Another concrete example of the discrepancy between the content of pedagogical grammars and actual language use is Anderson’s study (2007), which focuses on word-order rules related to adjective position in French and on the (in)congruity of those rules with classroom input and texts written by native French speakers. Question 2a: Is the idea of an exhaustive pedagogical grammar a utopia? Answer 2a: Yes, even more than is the case for a descriptive grammar, a pedagogical grammar will never be exhaustive. A pedagogical grammar is

106

Fanny Meunier

inherently selective, as mentioned in Section 4, as it should contain shorter sections than a reference grammar, minimize the use of metalanguage, consider the level of the user, be graded in various respects, anticipate the errors and display use in context. The generalization and selectivity requirements of pedagogical grammars turn out to be even more of a challenge when one adopts a view of grammar which includes a detailed insight into the colligational (see Note 5 for illustration), collostructional,14 contextually and semantically motivated aspects of language. Question 2b: Given the previous answer, how can we strike a balance between the desire to do justice to the richness of the language data and the need for generalization and selectivity? Answer 2b: I am convinced that the pedagogical grammar of the future, if it aims to incorporate corpus and cognitive elements, will not be a paper grammar. A very practical and practicable way to accommodate as many aspects as possible is probably to go for a web-based, hyperlink type of grammar (and perhaps also to forget about the term “grammar book” altogether, but that discussion would lead us too far). This option would favor not only principled eclecticism in learning (as observational, descriptive or explanatory options could be provided, see Ruiz de Mendoza, this volume), but also descriptive eclecticism that might help cover various needs of the learners (rules, analytical presentations, functional approaches, cognitive types of explanations). If one takes the example of causation (such as the typical ‘make X do Y’ or ‘get something Ved’ patterns), a first general presentation of typical causation patterns could be provided (with short invented or adapted examples to illustrate the structures). Options should also be provided for more curious or advanced users who would like to access more refined collostructional analyses. Next, extra corpus examples could be added, together with exercises linking context, semantic motivations, etc. (see Gilquin 2004 for illustrations). In pedagogical grammar research, a collaborative and web-based approach could also be one way of collecting and taking stock of what has already been published and analyzed, and make it possible to gradually increment the grammar with further grammar points or with specific contrastive units (see, e.g., De Knop and Dirven or Cadierno in this volume for the importance of a contrastive approach to the conceptualization of motion events). In terms of prioritization, it must be acknowledged that not all aspects of grammar deserve corpus and/or cognitive treatment. It could, for instance, be argued that the fact that reference to a definitive time point in the past typically requires


107

the use of the past tense (e.g., She was born in 1980), or the fact that the past form of regular verbs is formed by adding –ed may not need to be explained at length. Some issues might benefit from a “one angle” treatment: it might seem reasonable to propose several options of presentation for the list of irregular verbs (one of them being a list of irregular verbs organized in decreasing order of frequency). Some other issues might be treated from both angles and good candidates could include phrasal verbs, expressions of futurity, causation, question tags, etc. As for the selection of what aspects of grammar should be covered from a double CL angle, the corpus/cognition joint venture is probably still much too recent to provide a clear list of items and a definite agenda. My intuition is that one important issue would be to assess the cost-benefit balance in terms of linguistic and pedagogical description vs. ease of learning, an issue addressed in psycholinguistic experimental studies. Question 3: Should we aim for authenticity at all costs? Answer 3: No. To plead for the use of authentic examples, I share Mukherjee’s view (quoted in Aarts 2006: 400), when he states that : “the grammar of the future is no doubt a corpus-based grammar.” In his view, corpusbased does not mean “restricted to corpus data,”, nor “should it ignore the fact that, in spite of all kinds of genre variation, much in grammar is language-specific and not genre-specific.” Corpus studies carried out on native corpora can inform teachers and writers of pedagogical materials on the typicality of such or such structures in an array of text types, making the most of the numerous corpora available (various text types, genres or sociolinguistic options). Here again it will be the task of the learning materials composer or writer to select, on the basis of a learners’ needs analysis, the most common options and the most authentic, representative examples which also foster learner motivation and interest. Question 4: Which should take precedence: prototypicality or frequency? Answer 4: I am not sure which of the two should take precedence in the case of high discrepancy between both. My guess is that it would probably depend on the user’s age and level of competence. In terms of grammar, it might be wise to go for prototypicality at an earlier stage and gradually shift towards frequency, whenever relevant. Sometimes prototypicality and frequency correlate (attributive adjectives typically precede the noun they modify) which somehow eases the pedagogical approach. Problems arise when prototypicality and frequency do not correlate. In the case of the

108

Fanny Meunier

passive, for instance, corpus studies have shown that by-passives were much less frequent than agentless passives (see also Chen and Oller, this volume). Biber (1988) reports on agentless and by-passives for a large number of spoken and written genres in British English. The highest mean score (per 1000 words) is for the written genre (official documents) with 18.6 agentless passives for 2.1 by-passives. In telephone conversations the mean score for agentless passives goes down to 3.4 and there were no bypassives at all. Inferring from such figures that a pedagogical explanation of the passive should not initially include the reference to an agent is probably too extreme. It should also be added that more information is needed on the psycholinguistic aspects of prototypicality and more particularly the universal vs. culture- and/or L1-dependent prototypes). Question 5: What can cognitive and corpus linguistics bring to the issue of learnability? Answer 5: If one views learnability as the input/output efficiency of some method or approach (e.g., rote learning of grammatical rules vs. a cognitive approach to grammatical aspects of the language), then more classroom experimental studies are definitely needed to validate the efficiency of different methods. If learnability is, however, equated with effectiveness, Taylor (this volume: 53) rightly points to the fact that the Contrastive Language Hypothesis (Lado 1957) has had only limited success in predicting learners’ errors. He adds that one can expect that target language structures will be difficult to learn to the extent that they symbolize “idiosyncratic” conceptualizations, i.e., conceptual categories which are not found in the learner’s mother tongue, or which are not completely isomorphic with those of the learner’s mother tongue. As for corpus linguistics, Granger (1996) also referred to the shortcomings of contrastive analysis and proposed the Contrastive Interlanguage Analysis model (CIA). The framework involves two major types of comparison: first, a comparison of native language and interlanguage (to promote insight into the overuse, underuse and/or misuse of specific aspects of the grammar) and secondly, a comparison of different interlanguages (to finetune the comparisons and address universal vs. L1 specific issues). The CIA model builds on the previous paradigms in SLA studies and introduces new comparative dimensions to research in second language acquisition. These complementary analyses make it possible to address, among other things, issues in language transfer and/or universals on a solid empirical basis. The combined insights of cognitive linguistics into the conceptual categories and of corpus linguistics


109

into contrastive interlanguage analysis will definitely prove invaluable in defining and classifying the priority issues mentioned in Question 2b, i.e., which aspects of grammar deserve corpus and/or cognitive treatment, for what type of learner population, and at what level.

6. Conclusion The present article has shown that a sufficient number of convergences between the two CLs can be found, and that, despite their divergences, the ground is fertile to start building pedagogical grammars that take into account the complementary findings of corpus and cognitive approaches. The builders of such grammars will probably have to subtly intertwine descriptions and interpretations, to become experts in principled eclecticism and prioritization, and to allow learners to opt for several paths in their learning quest by using the potential of new technologies. As for future cognitive and corpus accounts of learnability, it is likely that they will sometimes converge, sometimes diverge, but there is no doubt that they will provide exciting new avenues for a joint cognitive, corpus and pedagogical agenda. Finally, the double CL venture will only benefit learners if it finds its way to the classroom. Corpus findings and corpus literacy are only slowly, and not always unproblematically, making their way to the classroom thanks to corpus-based reference tools, data-driven learning methods and corpus consultation skills. Cognitive and corpus approaches to the teaching and learning of foreign languages may not be innate abilities but I am confident that they will become acquired tastes.

Acknowledgements I would like to thank the editors and reviewers of this article for their insightful comments and suggestions.

Notes 1. It should perhaps be noted, to use literary terms, that the narrator’s perspective will be that of a corpus linguist and English language teacher rather than that of a cognitive linguist.

110

Fanny Meunier

2. I will use the term “field” or “discipline” to refer to research into doing corpus linguistics itself as opposed to corpus linguistics as a “methodology” serving (applied) linguistics proper. The question as to whether or not corpus linguistics is an independent branch of linguistics or rather a methodology will not be addressed in detail here. For a few comments, see, however, Section 3. 3. The term “empiricist” refers to any type of empirical research that uses data drawn either from observation or experience. It thus encompasses experimental and corpus approaches. 4. Hoey (1998) defines the notion of colligation as the grammatical company a word keeps (or avoids keeping) either within its own group or at a higher rank, the grammatical functions that the word’s group prefers (or avoids), and the place in a sequence that a word prefers (or avoids). Hoey (2006) shows for instance that the expression ‘a word against’ has a semantic association with sending and receiving communication (e.g., hear a word against), that ‘send/receive a word against’ has a pragmatic association with denial (e.g., wouldn’t hear a word against), that it usually colligates with modal verbs (e.g., wouldn’t hear a word against) and with human subjects and human prepositional objects (e.g., The head wouldn’t hear a word against his staff). 5. A pattern is a phraseology frequently associated with (a sense of) a word, particularly in terms of the prepositions, groups and clauses that follow the word. Patterns and lexis are mutually dependent, in that each pattern occurs with a restricted set of lexical items, and each lexical item occurs with a restricted set of patterns, Hunston and Francis (2000: 3). 6. A pattern is a phraseology frequently associated with (a sense of) a word, particularly in terms of the prepositions, groups and clauses that follow the word. Patterns and lexis are mutually dependent, in that each pattern occurs with a restricted set of lexical items, and each lexical item occurs with a restricted set of patterns, Hunston and Francis (2000: 3). 7. For more information on the BNC, see http://www.natcorp.ox.ac.uk/. 8. The following list of text types comes from the BNC World edition (2001) and is illustrative of the variety of text types included in major general corpora: — spoken component: demographic (place of residence, age, etc.) vs. context-governed (speech recorded in particular types of setting); the contextgoverned component is further divided according to the nature of the setting (educational/informative; business; public/institutional; leisure), paralleled by a monologue/dialogue distinction — written component: categorized into domains (i.e., subject matter, viz. imaginative, arts, belief and thought, commerce, leisure, natural science, applied science, social science, world affairs) and medium (book, periodical, miscellaneous published, published, to-be-spoken). 9. For comprehensive introductions to corpus linguistics methodology, see Kennedy (1998) or McEnery and Wilson (2001).


111

10. The book description is available from http://www.amazon.fr/ExploringGrammar-Context-Upper-Intermediate-Advanced/dp/0521568447. 11. What is true in the Belgian environment that I am familiar with is even more true in different cultural and educational backgrounds. Chinese and Japanese visiting professors in my department at the Université Catholique de Louvain have repeatedly mentioned the need for highly structured grammatical explanations following the traditional PPP approach (i.e., presentation, pratice, production). 12. The grammar book was used with second year university students specializing in English. 13. See http://esl.about.com/library/weekly/aa012400b.htm for more information on principled eclecticism. 14. Tables, figures and example sentences of Gabrielatos’ study are available from http://eprints.lancs.ac.uk/140/01/Conditional_Sentences_ELT_typology_and_ corpus_evidence.pdf. 15. Collostructional analysis investigates the interaction of lexemes and the grammatical constructions associated with them (see Stefanowitsch and Gries 2003). For instance, the go-and-V structure displays strong collostructional strength with verbs like get or fetch and not with stative verbs like be or think. Stefanowitsch’s (2000: 261) thus states that the go-and-V construction displays a dynamic feature whose semantic contribution can be described as “motion along a path.”

References Aarts, Jan 2006

Corpus linguistics, grammar and theory. Report on a panel discussion at the 24th ICAME conference. In The Changing Face of Corpus Linguistics, Antoinette Renouf and Andrew Kehoe (eds.), 391–408. Amsterdam: Rodopi. Achard, Michel, and Susanne Niemeier (eds.) 2004 Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching. Berlin: Mouton de Gruyter. Anderson, Bruce 2007 Pedagogical rules and their relationship to frequency in the input: Observational and empirical data from L2 French. Applied Linguistics Advance Access, published on June 7, 2007, http://applij.oxfordjournals.org/papbyrecent.dtl reference doi 10.1093/applin/amm015.

112

Fanny Meunier

Biber, Douglas 1988 Variation across speech and writing. Cambridge: Cambridge University Press. Biber, Douglas, Susan Conrad, and Randi Reppen 1994 Corpus-based approaches to issues in applied linguistics. Applied Linguistics 15 (2): 169–189. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan 1999 The Longman Grammar of Spoken and Written English. Harlow: Longman. Boers, Frank 2003 Applied linguistics perspectives on cross-cultural variation in conceptual metaphor. Metaphor and Symbol 18: 231–238. Boers, Frank, June Eyckmans, and Hélène Stengers 2007 Presenting figurative idioms with a touch of etymology: More than mere mnemonics? Language Teaching Research 11 (1): 43–62. Burnard, Lou, and Anthony McEnery (eds.) 2000 Rethinking Language Pedagogy from a Corpus Perspective: Papers from the Third International Conference on Teaching and Language Corpora. Frankfurt/Main: Peter Lang. Carter, Ronald, Rebecca Hughes, and Michael McCarthy 2000 Exploring Grammar in Context: Upper-Intermediate and Advanced. Cambridge: Cambridge University Press. Connor, Ulla, and Thomas Upton (eds.) 2004 Applied Corpus Linguistics: A Multidimensional Perspective. Amsterdam: Rodopi. Conrad, Susan 1996a Academic discourse in two disciplines: Professional writing and student development in biology and history. PhD dissertation, Department of English, Northern Arizona University. 1996b Investigating academic texts with corpus-based techniques: An example from biology. Linguistics and Education 8: 299–326. Croft, William, and D. Alan Cruse 2004 Cognitive Linguistics. Cambridge: Cambridge University Press. Dirven, René 1989 Cognitive linguistics and pedagogic grammar. In Linguistic Theorizing and Grammar Writing, Gerhard Leitner and Gottfried Graustein (eds.), 56–75. Tübingen: Max Niemeyer. Dirven, René, and John Taylor 1994 English modality: A cognitive-didactic approach. In Perspectives on English: Studies in Honour of Professor Emma Vorlat, Keith Carlon, Kristin Davidse, and Brygida Rudzka-Ostyn (eds.), 542–556. Leuven; Paris: Peeters.

Corpora, cognition and pedagogical grammars Ellis, Nick 1999 2006 2007 Ellis, Rod 1994

113

Cognitive approaches to SLA. Annual Review of Applied Linguistics 19: 22–42. Cognitive perspectives on SLA: The associative cognitive CREED. AILA Review 19: 100–121. Dynamic systems and SLA: The wood and the trees. Bilingualism: Language and Cognition 10 (1): 23–25.

The Study of Second Language Acquisition. Oxford: Oxford University Press. Evans, Vyvyan, and Andrea Tyler 2005 Applying cognitive linguistics to pedagogical grammar: The English prepositions of verticality. Revista Brasileira de Linguistica Aplicada 5 (2): 11–42. Fitzpatrick, Eileen (ed.) 2007 Corpus Linguistics Beyond the Word: Corpus Research from Phrase to Discourse. Amsterdam: Rodopi. Francis, Gill, Susan Hunston, and Elizabeth Manning 1998 Collins COBUILD Grammar Patterns 2: Nouns and Adjectives. Harper Collins: London. Gabrielatos, Costas 2003 Conditional sentences: ELT typology and corpus evidence. Paper presented at the Annual Meeting of the British Association for Applied Linguistics, University of Leeds, 4–6 September 2003. Gilquin, Gaëtanelle 2004 A corpus-based cognitive study of the main English causative verbs: A syntactic, semantic, lexical and stylistic approach. PhD dissertation, Louvain-la-Neuve, Université catholique de Louvain. 2006 Towards an empirically grounded definition of prototypes. Paper presented at the International Conference on Linguistic Evidence, Tübingen, 2–4 February 2006. http://www.sfb441.uni-tuebingen.de/ LingEvid2006/abstracts/gilquin.pdf. Grabe, William 1987 Contrastive rhetoric and text-type research. In Writing Across Languages: Analysis of L2 Text, Ulla Connor and Robert Kaplan (eds.), 115–137. Reading, Massachusetts: Addison-Wesley Publishing Company. Granger, Sylviane 1996 From CA to CIA and back: An integrated approach to computerized bilingual and learner corpora. In Languages in Contrast: Text-based Cross-Linguistic Studies, Karin Aijmer, Bengt Altenberg, and Maria Johansson (eds.), 37–51. (Lund Studies in English 88) Lund: Lund University Press.

114

Fanny Meunier

Granger, Sylviane, Joseph Hung, and Stephanie Petch-Tyson (eds.) 2002 Computer Learner Corpora, Second Language Acquisition and Foreign Language Teaching. Amsterdam: Benjamins. Greenbaum, Sidney 1987 Reference grammars and pedagogical grammars. World Englishes 8 (3): 191–197. Gries, Stefan 2006 Corpus-based methods and cognitive semantics: The many senses of to run. In Corpora in Cognitive Linguistics: Corpus-Based Approaches to Syntax and Lexis, Stefan Gries and Anatol Stefanowitsch (eds.), 57–99. Berlin, Heidelberg, New York: Mouton de Gruyter. Gries, Stefan, Beate Hampe, and Doris Schönefeld 2005 Converging evidence: Bringing together experimental and corpus data on the association of verbs and constructions. Cognitive Linguistics 16 (4): 635–676. Hoey, Michael 1998 Introducing Applied Linguistics: 25 Years on. Plenary paper held at the 31st BAAL Annual Meeting: Language and Literacies, University of Manchester, September 1998. 2006 Corpus-driven approaches to grammar. Plenary paper held at The Search for Common Ground: Exploring the Lexis Grammar Interface (ELeGI) Conference, University of Hannover, 7 October 2006. Hunston, Susan, and Gill Francis 2000 Pattern Grammar: A Corpus-driven Approach to the Lexical Grammar of English. Amsterdam; Philadelphia: John Benjamins. Hymes, Dell 1972 On communicative competence. In Sociolinguistics, John B. Pride and Janet Holmes (eds.), 269–293. Harmondswortth: Penguin. Johns, Tim 2002 Data-driven learning: The perpetual challenge. Teaching and learning by doing corpus analysis. In Proceedings of the Fourth International Conference on Teaching and Language Corpora, Graz, 19–24 July 2000, Bernhard Ketteman and Georg Marko (eds.), 107–117. Amsterdam: Rodopi. Kennedy, Graeme 1998 An Introduction to Corpus Linguistics. London/New York: Longman. Ketteman, Bernhard, and Georg Marko (eds.) 2002 Teaching and Learning by Doing Corpus Analysis: Proceedings of the Fourth International Conference on Teaching and Language Corpora, Graz, 19–24 July 2000. Amsterdam: Rodopi.


115

Lado, Robert 1957 Linguistics Across Cultures: Applied Linguistics for Language Teachers. Ann Arbor: University of Michigan Press. Langacker, Ronald 1987 Foundations of Cognitive Grammar. Vol. 1. Stanford, California: Stanford University Press. 1999 Grammar and Conceptualization. Berlin: Mouton de Gruyter. Larsen-Freeman, Diane 2002 The grammar of choice. In New Perspectives on Grammar Teaching, Eli Hinkel and Sandra Fotos (eds.), 103–118. Mahwah, New Jersey: Lawrence Erlbaum. Long, Michael 1997 Authenticity and learning potential in L2 classroom discourse. In Language classrooms of tomorrow: Issues and responses, George M. Jacobs (ed.), 148–169. Singapore: SEAMEO Regional Language Centre. MacWhinney, Brian 2000 The CHILDES Project. Volumes I and II: Tools for Analyzing Talk & Transcription Format and Programs. 3rd ed. Mahwah, New Jersey: Lawrence Erlbaum. McEnery, Anthony, and Andrew Wilson 2001 Corpus Linguistics: An Introduction. 2nd ed. Edinburgh: Edinburgh University Press. McEnery, Anthony, Richard Xiao, and Yukio Tono 2006 Corpus-based Language Studies: An Advanced Resource Book. London: Routledge. Meunier, Fanny 1998 Computer tools for interlanguage analysis: A critical approach. In Learner English on Computer, Sylviane Granger (ed.), 19–37. London/New York: Longman. Mukherjee, Jobrayto 2004 Corpus data in a usage-based cognitive grammar. In The Theory and Use of Corpora: Papers from the 23rd ICAME Conference, Karin Aijmer and Bengt Altenberg (eds.) 85–100. Amsterdam: Rodopi. Oakes, Michael 1998 Statistics for Corpus Linguistics. Edinburgh: Edinburgh University Press. O’Keeffe, Anne, Michael McCarthy, and Ronald Carter 2007 From Corpus to Classroom: Language Use and Language Teaching. Cambridge: Cambridge University Press. Paradis, Caritas 2003 Is the notion of linguistic competence relevant in Cognitive Linguistics? Annual Review of Cognitive Linguistics 1 (1): 207–231.

116

Fanny Meunier

Pütz Martin, Susanne Niemeier, and René Dirven (eds.) 2001 Applied Cognitive Linguistics II: Language Pedagogy. Berlin: Mouton de Gruyter. Roland, Douglas and Daniel Jurafsky 2002 Verb sense and verb subcategorization probabilities. In The Lexical Basis of Sentence Processing: Formal, Computational, and Experimental Issues, Paola Merlo and Susanne Stevenson (eds.), 325–346. Amsterdam: Benjamins. Römer, Ute 2004 Comparing real and ideal language learner input: The use of an EFL textbook corpus in corpus linguistics and language teaching. In Corpora and Language Learners, Guy Aston, Silvia Bernardini, and Dominic Stewart (eds.), 151–168. Amsterdam: John Benjamins. 2006 Looking at “looking:” Functions and contexts of progressives in spoken English and “school” English. In The Changing Face of Corpus Linguistics: Papers from the 24th International Conference on English Language Research on Computerized Corpora (ICAME 24), Antoinette Renouf and Andrew Kehoe (eds.), 231–242. Amsterdam: Rodopi. Shortall, Terry forthc. Corpus, cognition, and curriculum. PhD dissertation, University of Birmingham. Sinclair, John 1987a Introduction. In Collins Cobuild English Language Dictionary, John Sinclair (ed.). London: Collins. 1987b Looking Up: An Account of the Collins Cobuild Project. London: Collins. 1991 Corpus, Concordance, Collocation. Oxford: Oxford University Press. 2004 How to Use Corpora in Language Teaching. Amsterdam/Philadelphia: Benjamins. Stathi, Katerina 2006 Corpus Linguistics meets Cognitive Linguistics: A framework for the analysis of idioms. In Cognitive-linguistic approaches: What can we gain by computational treatment of data?, Antonietta Alonge and Birte Lönneker-Rodman (eds.), 27–29. Hamburg: University of Hamburg. Stefanowitsch, Anatol 2000 The go-(PRT)-and-V construction in English. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 26: 259–270. Stefanowitsch, Anatol, and Stefan Gries 2003 Collostructions: Investigating the interaction of words and constructions. International Journal of Corpus Linguistics 8 (2): 209–243.


117

Stubbs, Michael 2004 On very frequent phrases in English: Distributions, functions and structures. Plenary lecture given at ICAME 25 (International Computer Archive of Modern and Medieval English), Verona, Italy, 19– 23 May 2004. (Available at http://www.unitrier.de/uni/fb2/anglistik/Projekte/stubbs/icame-2004.htm.) Taylor, John 2002 Cognitive Grammar. Oxford: Oxford University Press. Teuberg, Wolfgang 2001 Message to the Corpora List, Tuesday 20 March 2001, Archived and available from http://torvald.aksis.uib.no/corpora/2001-1/ 0378.html Tomasello, Michael 2000 First steps toward a usage-based theory of language acquisition. CognitiveLinguistics 11 (1/2): 61–82. Tyler, Andrea, and Vyvyan Evans 2004 Applying cognitive linguistics to pedagogical grammar: The case of over. In Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching, Michel Achard and Susanne Niemeier (eds.), 257–280. Berlin/New York: Mouton de Gruyter. Valdman, Albert 2003 The acquisition of sociostylistic and sociopragmatic variation by instructed second language learners: The elaboration of pedagogical norms. In Issues in Language Program Direction, Sally Sieloff Magnan (ed.), 57–78. Boston: Heinle. VanPatten, Bill 2002 Communicative classrooms, processing instruction, and pedagogical norms. In Pedagogical Norms for Second and Foreign Language Learning and Teaching, Susan M. Gass, Kathleen Bardovi-Harlig, Sally Sieloff Magnan, and Joel Walz (eds), 105–118. Amsterdam/Philadelphia: Benjamins. Widdowson, Henry 2000 The limitations of linguistics applied. Applied Linguistics 21 (1): 3– 25. 2003 Defining Issues in English Language Teaching. Oxford: Oxford University Press.

Part II Tools for conceptual teaching: Contrastive and error analysis

Cross-linguistic analysis, second language teaching and cognitive semantics: The case of Spanish diminutives and reflexive constructions1 Francisco José Ruiz de Mendoza Ibáñez

Abstract This chapter argues that Cognitive Linguistics offers both theoretical and applied linguists powerful analytical tools that allow them to capture systematically language-internal and language-external similarities and differences. In order to substantiate this assertion, some areas of grammar where English and Spanish differ are identified, especially the expression of diminutive values, and the Spanish reflex passive (which is compared with its English counterparts). The paper discusses the similarities and contrasts in these areas from the perspective of their cognitive motivation. This is done on the basis of relevant aspects of Cognitive Model Theory, especially the study of constraints on metaphor and metonymy, of conceptual interaction, and of the role of high-level metaphor and metonymy in grounding constructional relations. The chapter also discusses some of the implications for language teaching of explanatorily adequate contrastive analysis in systematically identifying areas of difficulty in terms of different conceptualization strategies. Keywords: conceptualization strategies; diminutives; middle construction; reflexive construction; inchoative construction; correlation principle; extended invariance principle; highlighting; idealized cognitive model; high-level metaphor/metonymy; metaphoric/metonymic chains; semantic network; pedagogical grammar; ICM of size

1. Introduction Many second language acquisition theorists have worked under the assumption that (at least some aspects of) second language (L2) acquisition can somehow be fashioned into the natural, innate, native-like first language (L1) acquisition process. In this connection, some proponents of Universal Grammar (e.g., Bley-Vroman, 1989; Hawkins, 2003) see L2 language learning as crucially affected by a “critical period” in life beyond which language is not developed innately, which makes learners resort to

122

Francisco José Ruiz de Mendoza Ibáñez

their general problem-solving skills. However, the evidence for such a period, even when discussing L1 acquisition, rests on very shaky grounds (cf. Bohn, 2005, Singleton, 2005). Thus, there is no agreement among competing accounts as to the number and kind of language acquisition capacities affected by the critical period (that go from just phonology, through some purportedly innate and/or non-innate aspects of language development, to all capacities), its onset and offset ages (which range from the sixth month of fetal life to the age of 16), and its origin (which is variously claimed to be neurobiological, cognitive-developmental, and affective-motivational). Nor is it clear how the purported critical period might affect L2 acquisition. Rather, the consensus seems to be that second-language proficiency is not necessarily subject to any critical period but is characterized by a steady decline in ability to learn with age (Hakuta, Bialystok, and Wiley 2002). This situation of theoretical impasse poses the practical problem of how to address second-language teaching in ways that are well grounded from the point of view of linguistic theory without neglecting the possibility of opening new pathways in terms of neurobiological and psychological plausibility. In relation to this problem, I will contend that Cognitive Model Theory (Lakoff 1987) offers scholars interested in optimizing L2 learning processes powerful analytical tools which, besides being sensitive to neurobiological and psychological validation (see Lakoff and Johnson 1999; Gibbs and Perlman 2006), are endowed with descriptive and explanatory adequacy from a linguistic perspective. I will specifically show how Cognitive Model Theory allows us to systematically capture similarities and differences between English and Spanish in two areas where the two languages differ considerably: Spanish diminutives, reflexive constructions and their English counterparts. I will also make some suggestions as to possible pedagogical implications of this kind of study for advanced English learners of Spanish as a second language and argue for the benefits of making pedagogical strategies comply with requirements of explanatory adequacy.

2. Pedagogical grammar and levels of adequacy A pedagogical grammar (PG) is understood as “the learning materials containing the best possible illustration, presentation and gradation of the learning problems in a given area of language learning” (Dirven 2001: 18). This conception of grammar rests on the contrastive analysis of L1 and L2,

The case of Spanish diminutives and reflexive constructions

123

especially in areas of L2 where learning difficulties are expected to arise. For example, many learners of English encounter difficulties with the vast amount of meaning extensions associated with the prepositional/adverbial element in phrasal verbs. In order to face this problem Dirven (2001: 19) proposes to include explicit semantic networks as part of a pedagogical grammar, as illustrated by the case of across (Table 1): Table 1. Semantic network for across a. b. c. d. e. f. g.

Basic sense: ‘from A to B;’ e.g., The children ran across the road without looking. Extension 1 from the basic sense: ‘at the other side;’ e.g., She was sitting across the table from me. Extension 2 from the basic sense: ‘make somebody understand;’ e.g., He doesn’t know how to get his ideas across to his pupils. Extension 3 from the basic sense: ‘form impressions about oncoming phenomena;’ e.g., She came across as a very intelligent person. Extension 1a from extension 1: ‘transfer;’ e.g., The teachers always see their pupils across the busy street. Extension 2a from extension 2: ‘make somebody identify something;’ e.g., Marketing is about putting across to the customers the qualities of a product. Extension 3a from extension 3: ‘find or meet by chance;’ e.g., I came across an old friend during my holiday.

According to Dirven (2001), the semantic network helps the designer of a PG to make sure that the major senses of a polysemous word are built into the learning materials. It is also a way of facilitating learning by setting up straightforward and explicit conceptual connections. Descriptive exhaustiveness and explicitness are two important benefits of a PG that can subsequently be enhanced by investigating differences and similarities with the learner’s L1. It may even be argued that this kind of PG can be made sensitive to problems of native-like selection of the most appropriate constructions for given situations. Queller (2001: 55) has argued, following Langacker’s (1988) usage-based model of language, that semantic polysemy networks are ultimately grounded in “clusters of lexically entrenched phrasal patterns that instantiate highly specific prototype schemas” that can be adapted for L2 teaching purposes. “Entrenchment” is a patterned process of schematization of semantic structures in connection with their formal realizations. Consequently, if we want learners to acquire native-like skills, we need to make explicit, in a didactic way, not only what conceptual connections there are, but also their nature and motivated

124


links of extension. Concentrating on get across, we would need to teach not only that the meaning “make somebody understand” is related to the basic sense of moving ‘from A to B,’ but also why this is so. Drawing on Cognitive Model Theory it is possible to develop an explanation of why these two senses of across are coherently related on the basis of imageschematic thinking and a metaphorical extension. An image schema is a schematic representation that arises from our interaction with the world in terms of spatial relations (i.e., concepts such as movement along a path, spatial orientations, part-whole configurations, and bounded threedimensional space or containers). Thus, across calls for the existence of an entity moving from one position to the opposite position as determined by its relationship with a surface. If an object is ‘caused to come across’ it is made accessible in spatial terms by moving it from one side to the other of a bounded area in such a way that we can interact with it and thus learn about it. In ‘getting an idea across’ this combination of image schemas is used as the source domain of a metaphor where the idea is seen as an object that, in being made accessible, becomes understandable. It would not be difficult to devise a pedagogical version of this kind of explanation that takes into account the learner’s L1 perspective. For example, Spanish learners of English would have no problem in understanding the metaphors mentioned above since they are not strange to Spanish at all. The problem lies in the constructional patterns in which they take part, which differ to some extent between Spanish and English. Thus, in Spanish it is not possible to say that an idea “comes across,” which suggests a surface. However, in Spanish and English ideas can “come into” and “go out of” your mind. This requires treating the mind as a bounded threedimensional region in space, as in Lo tengo en mente (‘I have it in my mind’). The picture is more complex since both languages can treat the mind as a surface too: No logro quitármelo de la mente de ninguna forma (‘There is nothing I can do to take it off my mind’). In English it is also possible to say: There is a problem on my mind, but in Spanish only the container schema works here: Tengo un problema en mi mente, which treats the mind as a three-dimensional entity. I suggest that in general an appropriate course of action would be to teach explicitly a general learner-friendly rule for the metaphorical use of across, which might be worded as follows: Sometimes we talk about ideas as if they were objects that we can see, touch, and handle. They can also move or we can make them


125

move. If an idea reaches me, I can handle it and create a picture in my mind of what it is like. That is why we say that ideas get across or that we get ideas across to someone, as if ideas could move (in contact with a surface) from where they are to where people can deal with them. Since entrenchment comes with exposure to a vast amount of usage data, it would be necessary to devise exercises based on this type of rule presentation (cf. Rudzka-Ostyn, 2003). Of course, other rules are needed to account for the use restrictions of the rest of the items in the descriptive network. If the rules are sufficiently explicit, non-native-like uses can be discarded straightforwardly. Thus, learners will know that ideas can be “brought across” or “carried across” to someone but that it would be rather strange to say that ideas are “pushed/shoved/flung/slung across” to a person, with verbs specifying manner of motion, since the sense of “across” that we are dealing with focuses on intentionally making the idea accessible to someone rather than on the way we do it.2 A grammar that supplies learners with fine-grained explanations of usage rules which allows them to exploit constructions in a native-like manner is an explanatorily adequate grammar. This level of adequacy combines what Lakoff (1990) has called the cognitive and generalization commitments in linguistics while emphasizing the usage-based dimension of the account, i.e., its ability to see linguistic “structure as arising from and interacting with actual language use” (Geeraerts 2006: 29). Promoting the conversion of explicit rules into implicit knowledge is not beyond the scope of an explanatorily adequate pedagogical grammar. It is possible to achieve this goal by using data samples of constructional exploitation that give language learners an opportunity to spark off their intuitive mechanisms. Of course, the way this process can be naturally enhanced in terms of age, motivation, and other factors will not be a part of this kind of grammar until these issues have been empirically settled (Robinson 2001).

3. Some explanatory tools The different strands of Cognitive Linguistics (Dirven 2005; Broccias, this volume) have been able to clarify many analytical categories and to develop an important number of powerful explanatory tools. Among the most important categories we have the notion of “construction” – a formmeaning pairing (Goldberg 1995), the equivalent of Langacker’s “symbolic

126


assembly” (cf. Langacker 2005) – and the different kinds of “idealized cognitive models” or ICMs (Lakoff, 1987). ICMs are rich conceptual structures that capture relevant aspects of reality on the basis of a number of structuring principles. We have already mentioned image schemas and metaphor. There are at least two other kinds of ICM: propositional models or frames (cf. Fillmore 1985) and metonymy. Frames are schematic representations of situation types (e.g., ‘buying and selling,’ ‘eating,’ ‘spying,’ etc.) describable in terms of participants and their roles. For example, the ingestion frame consists of an “ingestor” that consumes food and drink (“ingestibles”) often with the help of an “instrument.” Metonymy (Lakoff and Johnson 1980, 1999; Kövecses and Radden 1998) is a ‘stands for’ or ‘point-of-access’ relationship within a conceptual domain (which is constructed on the basis of an image schema or a frame). Metonymy contrasts with metaphor especially in that metaphor maps concepts across discrete conceptual domains in such a way that one of the domains allows us to reason about the other (e.g., we talk about time as if it were a moving entity, as in Time goes by). ICMs combine to give structure to the semantic part of grammatical constructions, as evidenced by our previous discussion of across. Ruiz de Mendoza (2007) has developed ICM theory by taking into account the following factors: (i) conceptual interaction patterns; (ii) different levels of representation; (iii) the principles that constrain metaphorical and metonymic production. The following subsections summarize the essential features of this development, which will then be applied to the case studies and pedagogical proposals in Sections 4 and 5.

3.1. Chaining and constraints Consider again the case of ‘get an idea across to someone,’ which involves the combination of the metaphorical mappings IDEAS ARE OBJECTS and UNDERSTANDING AN IDEA IS PERCEPTUALLY EXPLORING AN OBJECT, plus the figurative use of the complex image-schematic notion of motion from one side to another of an area while keeping contact with the surface as in (1). (1)

John got the idea across to the student > *The idea is at the student

Think of a literal counterpart of this expression, as in (2):


(2)

127

Jim got the child across to a safe place > The child is in/at a safe place.

What the two expressions have in common is the “caused-motion construction”, which is characterized by such roles as a causer of movement, the entity that is caused to move, a source of movement, a path, and a destination. However, each type of motion follows a different reasoning pattern. The reason for this is found in the way metaphorical targets govern the selection process of elements from the source domain. Thus, in (1) the student who comes to understand an idea is the destination of a moving object, but the object is not necessarily at his side. What we have instead is a combination of the ‘idea as a moving object’ metaphor with the metaphor that correlates understanding and spatial accessibility. Without this combination, there would be a target element without a clear source domain correspondence for the moving object metaphor. We may call a necessary combination of metaphors, as in getting an idea across, a “metaphoric chain.” Note that the chain is a requirement of the first metaphoric target, which demands a more refined elaboration of the basic correspondence between understanding and receiving an object, since just having the object does not involve understanding, i.e., full knowledge of its characteristics. Since receiving an object allows for full perception and manipulation, a new complementary metaphoric mapping is triggered whereby the actual correspondence is not between understanding and receiving the object, but also between understanding and interacting with the object. Ruiz de Mendoza and Santibáñez (2003) have argued that the target domain of metaphoric mappings places relevance constraints on the implicational structure of the source, which needs to parallel the target in every possible respect. This constraint is captured by the Correlation Principle (CP), according to which the greater the degree of correspondence that there is between the implicational structures of the source and target domains, the greater the degree of felicity of the resulting metaphor. By way of illustration, in the case of ARGUMENT IS WAR, an extremely heated debate between opposing political candidates is better described as an “all-out war” than just as a fight. The Correlation Principle has been explored in greater detail in Ruiz de Mendoza and Mairal (2007), where its activity is extended to metonymy and explored at various levels of genericity. It is complementary to the Invariance Principle, postulated by Lakoff (1990), which was developed by Ruiz de Mendoza (1998) into the Extended Invariance Principle (EIP). According to the EIP all the generic structure of the target domain of a metaphoric mapping has to be preserved

128


in a way that is consistent with the generic structure of the source. For example, in the metaphor PEOPLE ARE ANIMALS behavior maps onto behavior, and physical attributes onto physical attributes, but we do not map behavior onto physical attributes or the other way around. Suppose we want to describe the sportsman John Weatherford, a Land O’Lakes (Florida) quarterback, by means of a ‘horse’ metaphor as in John is a horse, a big strong kid. This metaphor suggests great physical strength. However, a different animal metaphor, as in He is very mulish, describes a headstrong person, one that behaves just as mules are thought to behave. In application of the EIP, agents map onto agents, objects onto objects, goals onto goals, and so on. The CP, in turn, regulates the selection of the most relevant agent, object, or goal type in terms of how it relates to the rest of the elements of structure within the same domain. Consider the range of implications associated with the use of the ‘caused-motion’ construction in the sentence My sister laughed me into despair. This is another case of figurative movement with a causer of motion (that maps onto the speaker’s sister), an affected entity (mapping onto the speaker) that is set in motion, and a destination (that maps onto the emotional state of despair). The correlations are neat in terms of the EIP: the causer of motion is the causer of an emotional state, the entity set in motion is the goal/experiencer of the induced emotional state, and the intended destination is the intended emotional state. Then, from the point of view of the CP, all the elements of the construction are adequate source elements to express an induced emotional state. But there is an important difference between caused-motion in this example and the case of figurative caused-motion with across discussed in (1) above, since unlike get across the verb laugh is not intrinsically a motion verb. Ruiz de Mendoza and Mairal (2007) have discussed this use of laugh as a case of high-level metaphor, i.e., a generic metaphor that has grammatical consequences, as is here evidenced by the impossibility of using the preposition “at” to introduce an object once laugh is used in the caused-motion construction. The adaptation of the verb laugh to the caused-motion construction – a coercion phenomenon of the kind discussed by Michaelis (2003) – demands understanding a verbal process where the activity of the agent has no direct physical impact on the object in terms of another kind of process where there is such an impact, as would be exemplified by a verb like kick. The meaning implications of this metaphor are not difficult to identify: the speaker feels that the emotional effect that being laughed at has had on him is comparable to the effect of directly applied physical force, as in push/drive/throw someone into despair.


129

Chaining also applies to metonymy and to combinations of metaphor and metonymy (Ruiz de Mendoza and Díez 2002). In Ruiz de Mendoza (2007) it is argued that necessary combinations of conceptual mappings, of whatever kind, result from the activity of the Mapping Enforcement Principle. This principle ensures that no item will be discarded from a mapping system if the item in question may be adapted to the meaning demands of the system in terms of the Extended Invariance and the Correlation Principles. One straightforward example arises from comparing the metonymic activity in Shakespeare is easy to read (AUTHOR FOR WORK) and Shakespeare is on the top shelf (AUTHOR FOR WORK FOR MEDIUM). In the ‘work’ metonymy the easy to read profiles the domain of an author’s creative writings, while in the ‘medium’ metonymy, on the top shelf additionally profiles the physical format in which the creative work is presented. This mapping arises as a requirement for the selection of the best possible source domain for the final target domain (the medium). ‘Shakespeare as an author’ is not an adequate source here since on the top shelf can hardly be profiled in the domain of authorship.

3.2. High-level conceptualization We will now make a distinction between low, primary, and high levels of conceptualization. We define the high level as the generic level of conceptual representation created by abstracting away structure common to multiple low-level or primary concepts. Cause-effect relationships and such notions as ‘action,’ ‘process,’ and ‘result’ are examples of high-level concepts. Low-level concepts (e.g., ‘chair,’ ‘mother,’ ‘buying,’ ‘hunting’) are non-generic representations created by making well-entrenched, coherent links between elements of our encyclopedic knowledge store. A primarylevel concept is any generic or non-generic concept that directly arises from bodily experience. Among primary-level concepts we find image schemas and notions designating temperature, smell, color, size, and emotions. We have already introduced high-level conceptualization in our discussion of the mapping between two different kinds of transitivity, a mapping that has been labeled EXPERIENTIAL ACTION IS EFFECTUAL ACTION by Ruiz de Mendoza and Mairal (2007). As we observed above, this mapping has consequences in terms of grammar, as evidenced by the recategorization of laugh from an experiential to an effectual verb type. There are other high-

130


level metaphors underlying various grammatical phenomena as summarized in Table 2: Table 2. High-level metaphor in grammar a.

Change of transitivity type: He talked me into business (COMMUNICATIVE ACTION IS EXPERIENTIAL ACTION).

b. c. d.

Nominalization: We couldn’t prevent the destruction of the town by the enemy (EVENTS ARE OBJECTS). Conversion of a verb into an idiomatic phrase: They gave the thug a big beating (ACTIONS ARE TRANSFERS). Use of the object construction to express states: She has a lot of fear (STATES ARE POSSESSIONS).

Metonymy can also be operational at the high level with effects on grammar. The productiveness of high-level metonymy and its consequences for grammar have been explored in some detail in Ruiz de Mendoza and Pérez (2001). Consider the following examples in Table 3. Table 3. High-level metonymy in grammar a.

Categorial conversion: He hammered the nail into the wall (INSTRUMENT FOR ACTION).

b.

c. d.

Subcategorial conversion: There is a lot of America in what she does (AN ENTITY FOR ONE OF ITS PROPERTIES); There were three Johns at the party (AN INDIVIDUAL ENTITY FOR A COLLECTION INCLUDING THAT ENTITY). Enriched composition (Jackendoff 1997: 61): She enjoyed/began the dance (AN OBJECT FOR AN ACTION IN WHICH THE OBJECT IS INVOLVED). Parametrization: This week, he’ll do the carpet and I’ll do the dishes (GENERIC FOR SPECIFIC).

A high-level metaphor or metonymy always requires the target domain to be a high-level propositional model, like ‘action,’ ‘process,’ ‘position,’ ‘state,’ ‘result,’ ‘agent,’ ‘object,’ or ‘instrument.’ Note that these notions are either image schemas or are grounded in image-schematic configurations. Thus, actions and processes (and associated concepts such as agent, instrument, and object) cannot be understood without making reference to motion (or dynamism), positions are locations in space, states are attributes of objects, and so on. This observation is consistent with the claim, formulated by Lakoff (1990), that much abstract reasoning is ultimately image-


131

schematic, and also consistent with the emphasis that cognitive linguists put on the embodied nature of thought, i.e., the grounding of many levels of thought and reasoning in bodily experience (cf. Lakoff and Johnson 1999: 36, 462). However, not all of our thinking is bodily-based, and certainly not spatially-based. There are other primary concepts such as those that designate color, temperature, smell, and the different kinds of emotions. These are the concepts that, together with image schemas, give rise to what Grady (1997) has termed “primary metaphors”, i.e., cross-domain mappings whose source domain is directly grounded in bodily experience. Table 4 offers some examples of primary metaphors (cf. Lakoff and Johnson 1999: Ch. 4, for a detailed list): Table 4. Primary metaphors a. b. c. d. e.

AFFECTION IS WARMTH (e.g., She is rather cold with people) HAPPY IS UP (e.g., I’m feeling up today) IMPORTANT IS BIG (e.g., He is a big wheel in the company) KNOWING IS SEEING (e.g., I see what you mean) ORGANIZATION IS PHYSICAL STRUCTURE (e.g., How will you

put all those ideas

together?)

High-level metaphors and metonymies motivate grammar directly. This is not the case with primary systems. Let us compare the primary metaphors listed in Table 4 with the high-level metaphors in Table 2 such as ACTIONS ARE TRANSFERS. This high-level metaphor is grounded in a cluster of primary concepts including a ‘causer of motion,’ an ‘object,’ ‘(physical) possession,’ ‘motion along a path,’ and a ‘destination of motion’ (e.g., He gave John a beating/a kick/a punch/a kiss/a shove, etc.). In spite of this primary-level component in the source, the metaphor ACTIONS ARE TRANSFERS is a high-level system, since the target domain is a high-level concept. In contrast, the primary metaphors in Table 4 do not have high-level target domains and do not motivate grammar. However, what primary concepts share with high-level concepts is their generic nature and that is how they differ from lower-level constructs. Thus, an image schema is a representation that draws away structure common to many cases of spatial and motor experience. This observation is necessary in order to understand why image schemas underlie some grammatical constructions and can be used to explain the value of so-called grammatical words like definite articles, demonstratives, adverbial particles, and the value of affixes having compa-

132


rable functions. This observation will be useful to understand the experientialist account of Spanish diminutives in the following section.

4. Spanish diminutives Our discussion of levels of conceptualization brings us to our first case study, diminutive values in Spanish. This is a very complex topic that has been treated in some detail from the perspective of Cognitive Model Theory in Ruiz de Mendoza (2000). Here we will make our discussion sensitive to the requirements of an explanatorily adequate pedagogical grammar.

4.1. Descriptive complexity The Spanish diminutive has traditionally been considered a rather arbitrary category (Alonso 1951; Montes 1972; Lázaro 1976). Thus, while -ito and illo seem to have affectionate and pejorative values respectively (e.g., un regalito ‘nice little gift,’ versus un regalillo ‘a modest or wretched little gift’), the reverse can also be the case (eg. eres un granujilla ‘you are a nice little rogue;’ menudo añito, ‘what a really horrible year’). There are more factors that make the study of Spanish diminutives an even more complex enterprise. One is local variation, which results in the use of other suffixes with different shades of meaning: -ico, -in usually have the same affective implications as -ito; -ejo, -uelo are related to -illo, and are basically pejorative; -ete is used in a humorous fashion. Another is the frequent avoidance of one suffix so as not to produce an ill-sounding effect; e.g., año (‘year’) gives rise to añito, añico, añete, but not *añín, *añillo. There are other relevant observations. Thus, Lázaro (1976) has argued that abstract nouns (e.g., gravedad, ‘gravity’) do not take the diminutive form, since it is difficult to talk about abstract entities in terms of size or of affective values. However, there are exceptions. Abstract nouns ending in ura (e.g., ternura, ‘tenderness’) can be found in the diminutive (e.g., ternurita in ¡Con qué ternurita la mira! ‘How very tenderly he looks at her!’). Lázaro (1976) speculates that diminutive forms of this kind may be used by analogy with concrete nouns with the same ending (e.g., cintura, ‘waist’). However, part of the reason may simply be a phonetic one, since it is difficult to combine derivational suffixes like –edad, and -ez, but not –ura, with the different diminutive endings. Note that concrete base words with the same endings are not generally found in the diminutive: antigüedad (‘an-


133

tique’) > *antigüedacita/-illa; tez (‘face’) > *tecita (however, it is possible to say tececita, thus avoiding the ill-sounding effect). This suggests that it is not their meaning which precludes abstract nouns from building diminutive forms. In fact, there are abstract nouns, other than those ending in – ura, that take the diminutive form: alegría (‘joy,’ ‘happiness’) > alegriita (cf. the concrete noun sandía ‘watermelon’ and its diminutive sandiita). In these cases the diminutive seems to be used for reasons of emphasis rather than to indicate affection. So we can conclude that abstract nouns are able to take diminutive forms for reasons of emphasis, unless the resulting form sounds bad. Lázaro (1976) has also observed that action nouns (e.g., alabanza, ‘praise’) are not found in the diminutive form unless they are recategorized as concrete nouns (licencia, ‘licence’ > licencita) or end in -ción/-zón/sión, or -ida (e.g., invasión, ‘invasion,’ acogida ‘welcome’). Again, Lázaro’s observation is only part of a more complex picture. It is true that action nouns are abstract in nature, which makes them less sensitive than concrete nouns to meaning modifications based on size or affective values. However, we still need to explain why action nouns ending in -ción/-zón/sión, or –ida can take the diminutive ending without affecting their abstract character. This time the reason is not phonetic, as evidenced by the possibility of affixing base words that are not action nouns: chanza (‘joke’) > chancita; panza (‘belly’) > pancita; pitanza (‘grub’) > pitancita. The real reason for the difference between action nouns ending in –anza and those ending in -ción/-zón/-sión, or -ida is that the latter tend to designate specific actions, which makes them sensitive to the diminutive form. Thus, invasión can refer to any act of invading or to one instance of invading. The diminutive can only be used in the latter case (e.g., Otra invasioncita más y estamos perdidos ‘Another damned invasion and we’ll meet our doom’).

4.2. The experientialist explanation Although diminutive categories in Spanish are difficult to describe, I will argue that the principles that underlie their use can be systematized for English learners of Spanish as a second language. Let us bear in mind that while diminutive suffixes are very frequent in Spanish, this is not the case in English, where they are restricted to a small unproductive set (e.g., -ette in statuette, -let in piglet, -ling in duckling, and y/ie in piggy). In English, the wide range of values attributed to the Spanish diminutive is usually

134


expressed by analytic means. For convenience, we will base our discussion on the very widespread –ito/-illo suffixes in Castellan. Since phonetic factors may either prevent diminutive affixation or compel to conflate values, we start by considering those cases where the contrast is possible: (3)

¡Qué perrito/*perrillo más mono! (‘What a cute little dog!’).

(4)

¡No era más que un perrillo callejero! (‘It was just a wretched little street dog!).

(5)

¡Menudo mesecito/*mesecillo que hemos tenido! (‘What a terrible month we’ve just had!’).

(6)

Unos mesecillos/mesecitos más y habrás terminado (‘Just a few more months and you will have finished’).

(7)

A lo lejos se oía una suave tonadilla/*tonadita (‘In the distance one could hear a sweet little ditty’).

(8)

¿Cuándo dejarás la dichosa cancioncita/*cancioncilla? (‘When will you stop that nasty little song?’).

(9)

Dame un besito/*besillo (‘Give me a nice little kiss’).

(10)

¡Qué granujilla/*granujita más simpatico! (‘What a nice little rogue!’).

(11)

Tengo un sueldecillo/*sueldecito de nada que apenas me permite vivir (‘I earn a modest little income, barely enough for a living’).

Traditionally, it has been postulated that there is a tendency for –ito to be more positive than –illo, which carries negative overtones. This position is supported by examples (3), (4), (9), and (11). However, as examples (4), (6), and (7) above show, -illo may be used in a positive way, while –ito may get a negative value, as in (8). So the issue is not only one of positive versus negative values. We need to explain why the axiological load of the two suffixes is so sensitive to variation and to find what principles regulate the range of meanings of diminutives. A plausible solution comes from considering our apparently inconsistent data from the perspective of Cogni-


135

tive Model Theory, more specifically from a description of the notion of size as a primary concept. We provisionally use the description in Table 5: Table 5. ICM of ‘size’ a. b. c. d. e.

Entities range in size from very small ones to very large ones. A small entity is often more manageable than a bigger one. A small entity is often less harmful than a bigger one. Small entities are likeable (since they can be controlled). Small entities are unimportant (since they will not do harm to us).

As a primary concept, the ‘size’ ICM includes more than just imageschematic information about a measurable property of objects (as given in (a)). It contains a description of how people interact with objects on the basis of such a property (i.e., (b) and (c)) plus the two corollaries that we can derive from them and which specify two types of emotional reaction. These corollaries provide us with an explanation of the basic values of the -ito/-illo suffixes. Thus, -ito is applied whenever there is a phonetically feasible -ito/-illo contrast to designate entities, properties, or events that are likeable, whether small size is involved or not; –illo, is used when the entity, property, or event is seen as unimportant. Note that the two corollaries are metonymic derivations: small entities either stand for the class of likeable entities or for the class of unimportant entities depending on whether the entity in question is manageable or harmless. This simple analysis accounts for all the values of examples (3)–(11) above. Thus, in (3) perrillo, but not perrito, is odd in the context of a sentence where the speaker is expressing admiration for the dog. In contrast, perrillo in (4) is more compatible than perrito with the context of worthlessness expressed by the rest of the sentence. A similar meaning is conveyed by (11), where sueldecito would not be possible since it would create an important meaning clash with de nada (‘absolutely unimportant’). Then, mesecito in (5), just like cancioncita in (8), is used in a context of speaker’s annoyance, which is conveyed through irony. Here, the diminutive, by strengthening the positive axiology of the words to which it applies, has the additional effect of enhancing the speaker’s irony. However, mesecillos in (6) is used to minimize the unwanted effects of hard labor in terms of time. Since hard labor is considered negative, the minimizing effect of –illo in this context has an overall positive effect. A similar situation arises from the use of –illo (7), where the intensity of music (which, if too high, could bother some people) is brought down to a level in which it

136


is perceived as smooth and gentle. The mitigating value of –illo is also evident in (10), where a rogue is seen as likeable. This is an interesting situation since we associate the notion of “likeable” with –ito. However, note that a rogue is intrinsically negative and the best way to convey something positive about a rogue is to minimize the importance of his misdeeds. The use of –ito in the same sentence would have conveyed irony, as was the case in (5) and (8), where the overall contexts were negative. This account is powerful enough to predict all possible uses of –ito/illo, including cases where the affixes combine with words that do not designate objects but properties or (aspects of) events: (12)

¿Qué hacéis? – Nada, mirandillo (‘What are you doing?’ – ‘Nothing, just looking around a little bit’).

(13)

Venga, vamos, corriendito (‘Come on, Let us go, up and running’).

(14)

Era como rojito, muy bonito (‘It was kind of a nice red, very nice’).

(15)

Es de un color como rojillo (‘It is kind of reddish in color’).

(16)

Ya eres grandecito para eso (‘You’re pretty grown-up for that’).

(17)

Es grandetoncillo, nada que temer (‘He is but a bit biggish, nothing to be afraid of’).

(18)

Lo he dejado apretadito (‘I have made it pretty tight’).

(19)

Venga, prontito (‘Come on, double quick’)

In (12) –illo minimizes the importance of the action. In (13) –ito is used to convey the positive idea of performing the action briskly. The same suffix in (14) suggests a likeable type of red color, while –illo in (15) softens the impact of the color. In (16) the possible bad connotations associated with behavior unbecoming of an adult are cancelled out by the positive -ito diminutive. In (17) the threatening aspect of a big person is mitigated by – illo, which suggests unimportance. Finally, the resultative aspect of the participle apretado in (18) is intensified by–ito, just as the meaning of the


137

adverb pronto in (19). In both cases, the underlying value is one of likeability: apretadito implies that the action of tying has been done so well that the hearer will like the result; in turn, prontito in (19) implies that the hearer is expected to act in a way that is desirable for the speaker. The two corollaries of the ICM of ‘size’ can help us to account for the apparent coincidence of meaning values in some examples of use of -ito/illo with the same word and in the same linguistic context: (20)

a.

b.

Venga, dame un pastelito … (Lit. ‘Come on, give me a delightful little cake;’ non-lit. ‘a nice little cake’) Venga, dame un pastelillo … (Lit. ‘Come on, give me a small little cake;’ non-lit. ‘a nice little cake’)

Examples like these have contributed to the erroneous claim – mentioned at the beginning of section 4.1 – that diminutives are unpredictable. But this is not the case. Pastelito in (20a) invokes a likeable object that will be beneficial for the speaker who is asking for it. Pastelillo in (20b) suggests a modest object that is not too valuable, in such a way that providing the speaker with it should represent little, if any, cost to the addressee. Each diminutive is part of a different exploitation strategy of the same underlying social convention (cf. Pérez and Ruiz de Mendoza 2002; Ruiz de Mendoza and Baicchi 2007), according to which we are not expected to act in ways that are costly to other people and, conversely, we are expected to do things which other people like. In (20a) the use of -ito implies that the speaker finds the referent desirable; in accordance with the first part of the convention, the hearer is then expected to do his best to cater for the speaker's desires. In (20b), on the other hand, the use of -illo suggests that the referent is really not very valuable, with the implication that providing the speaker with it should represent little, if any, cost to the addressee, which enhances the speaker’s expectations to obtain what he desires.

4.3. Pedagogical implications As has been noted above, it is possible to express in English the synthetic Spanish diminutive forms by means of analytical expressions generally making use of “little”, which provides a way of capturing in L1 the essential value of Spanish diminutives. Below are some correlations that may be

138


useful for an explicit teaching strategy based on asking learners to find the right suffix in L2 for the given L1 meaning: (i) NOUNS — With positive overtones: nice little hotel (‘hotelito’), delightful little cake (‘pastelito’), charming little village (‘pueblecito’), engaging little boy (‘niñito’), cute little dog (‘perrito’), dear little mouse (‘ratoncito’), gentle little kiss (‘besito’), dainty little dress (‘vestidito’), sweet little bird (‘pajarito’). — With negative overtones: nasty little film (‘peliculilla’), disturbing little creature (‘criaturilla’), wretched little car (‘cochecillo’), miserable little bird (‘pajarillo’), insignificant little man (‘hombrecillo’). There are other attitudinal adjectives that can translate diminutive values without appearing in combination with little. Consider: petty (little) details (‘detallitos’), a minor problem (‘un problemilla’), a slight taste (‘un saborcillo’), a fine young lad (‘un jovencito’). (ii) OTHER WORD CLASSES where the diminutive suffix carries an intensifying value: — Context-dependent equivalence: properly sitting, quietly sitting, delicately sitting, daintily sitting, nicely sitting, etc. (‘sentadito’); — One-to-one equivalence: Your children are exactly alike (‘Tus hijos son igualitos’); He's already quite grown-up (‘Ya está crecidito’); Let us see if you keep nice and quiet (‘A ver si te estás calladito’); They spoke very softly (‘Hablaron muy bajito’); She feels just a bit tired (‘Se nota algo cansadilla’); She's rather conceited (‘Es una creidilla’).

5. Reflexive constructions The theoretical status of Spanish reflexive constructions is another important source of controversy. Spanish reflexive constructions are characterized by the use of the pronoun se, which has traditionally been claimed to have a number of different values (see Maldonado, this volume), but which must be seen in the framework of and hence as part of larger constructions:


(21)

Se odian (‘They hate each other’) [reciprocal]

(22)

No se afeitó bien (‘He didn’t shave well’) [reflexive]

(23)

El trigo se siega en agosto (‘Wheat is reaped in August’) [reflex passive]

(24)

Se dice que se presentará en cualquier momento (‘It is said that he will show up any minute’) [impersonal]

(25)

Nunca se sabe (‘One never knows’) [indefinite]

(26)

Se lo dijeron demasiado tarde (‘They told her too late’) [dative]

139

One of the most interesting and controversial cases of se constructions is provided by the so-called reflex passive in (23), which is characterized by having a transitive verb and an implicit agent, which makes it very similar in meaning to the ser-passive construction (the equivalent of the English be-passive). In Cognitive Linguistics, Maldonado (1999; see also this volume) is the most exhaustive descriptive account of the different se constructions. Additionally, Ruiz de Mendoza and Peña (2006) have investigated the reflex passive from a contrastive Spanish-English perspective. This section will draw insights from the latter account with a view to improving the explanatory adequacy of this area of a pedagogical grammar.

5.1. Descriptive complexity The Spanish reflex passive construction shares features with the serpassive construction, as can be seen from the comparison between the aexamples and the c-examples in (27)–(29) below. However, unlike the serpassive, the reflex passive, can hardly take an agentive complement, as evidenced by the strong oddity (or even ungrammaticality) of the bexamples. (27)

a.

De repente, se abrió la puerta y apareció el mago (‘Suddenly, the door opened and the magician appeared’) [inchoative]

140


b.

c.

(28)

a. b.

c.

(29)

a. b.

c.

*De repente, se abrió la puerta por alguien y apareció el mago (‘Suddenly, the door opened by someone and the magician appeared’) [inchoative + agent] De repente, la puerta fue abierta por alguien y apareció el mago (‘Suddenly the door was opened by someone and the magician appeared’) [ser-passive] Este pan se corta con dificultad (‘This bread cuts with difficulty’) [middle] *?Este pan se corta con dificultad por cualquiera que lo intente (‘This bread cuts with difficulty by whoever tries’) [middle + agent] Este pan es cortado con dificultad por cualquiera que lo intente (‘This bread is cut with difficulty by whoever tries’) [ser-passive] La puerta de delante se abre con facilidad (‘The front door opens easily’) [middle] *?La puerta de delante se abre con facilidad por cualquiera que lo intente (‘The front door opens easily by whoever tries’) [middle + agent] La puerta de delante es abierta con facilidad por cualquiera que lo intente (‘The front door is opened easily by whoever tries’) [ser-passive]

Even though se reflex passives seem to share with passives the emphasis on the action rather than on the agent, there are differences that have been pointed out by Maldonado (1999; this volume). Thus, in se passives the endpoint of the action is given more prominence and the agent is schematized to a greater extent than in actual passives. In my view, this is the reason why se passives, unlike actual passives, are reluctant to take an agentive complement, as can be seen from the grammatical impossibility of (27b) and the strong oddity (or even the ungrammaticality) of sentences (28b) and (29b). There is a difference, however, between the sets of examples under (27) and those under (28) and (29). The latter two sets have an evaluative ingredient that applies to the way in which the process is carried out. This ingredient brings with it a slight shift of focus from the endpoint back to the process, which explains why some speakers would accept sentences like (28b) and (29b). Note that there is a variant of the evaluative se construction where the evaluative component acts on the result rather than


141

on the process. As evidenced by example (30b) below, this variant is as reluctant to take an agentive complement as the non-evaluative reflex passive illustrated by (27b): (30)

a.

b.

La falda roja sólo se lava bien en la lavadora nueva (‘The red skirt only washes well in the new washing machine’). *La falda roja sólo se lava bien en la lavadora nueva por cualquiera que lo intente (‘The red skirt only washes well in the new washing machine by whoever tries’).

In English the meaning conveyed by se reflex passives is usually conveyed through the intransitivization of the verb, thus resulting in the socalled inchoative (e.g., The door opened) and middle (e.g., The door opened easily) constructions. In this process the patient takes the syntactic subject function and the semantic subject (the agent) is made implicit. Thus, The door opened (easily) suggests that ‘someone (or some force) opened the door,’ as is evident from the implications of the following discourse development, where the anaphoric pronoun his links up with the implicit agent: (31)

After oiling the hinges the door opened easily, without any further effort on his part.

Radden and Dirven (2007) have pointed to the “enabling” function of the subject of middle constructions in English. In this function, the subject takes on a causal, agent-like value, as in Our new stadium seats 80,000, or These caves are dripping water, where such roles as the place and source can act as the subject of the construction. However, this feature of the construction does not preclude the existence of a true underlying agent. In the case of the stadium example, we know that someone has the possibility of seating 80,000 people in the stadium, as seen in (32a). The caves example is different, since drip is an intransitive verb without a prototypical agent (cf. 32b). So the alternative middle wording of the event cannot override the existence of an agent, simply because the whole process is construed as agentless whichever the constructional choice: (32)

a. b.

You know, we can seat 80,000 people in our new stadium. Water is dripping from these caves

142


Heyvaert (2003: 132) has observed that the middle construction “foregrounds the fact that the Subject-entity has properties which influence the occurrence of a particular process.” The accuracy of this observation is supported by the sentence The door opened easily, which makes explicit the properties the Subject-entity has had to acquire for the opening process to be facilitated. Heyvaert (2003: 132–137) also sets up a typology of middle constructions that goes beyond the traditional focus on “facilityoriented” middles, typically constructed with the adverb easily. Thus, middle constructions may indicate: (i) a quality judgment of the subject-entity, as in example (33a); (ii) the amount of time needed to carry out some process on the subject-entity, as illustrated in (33b); (iii) whether the process is possible, as in (33c); (iv) the existence of a telic or “destiny-oriented” focus, as seen in (33d); and (v) the relevance of a result, as in (33e). (33)

a. b. c. d. e.

This car handles like a sports sedan. This item usually ships within 2–3 days. This umbrella folds up. The travel pillow fixes to the headrest providing comfortable neck support. It washed well, with little shrinkage.

Spanish se reflex passives also foreground the Subject-entity properties that influence the process and are generally sensitive to the proposed typology, except for those cases where the construction specifies whether a process is possible, as evidenced by the Spanish renderings of the sentences in (33’): (33’)

a. b. c. d. e.

Este coche se lleva como un sedan deportivo. Normalmente, este artículo se expide en dos a tres días. *?Este paraguas se cierra (cf. Este paraguas cierra). La almohadilla se fija al reposacabezas proporcionando un cómodo apoyo para el cuello. Se lavó bien, encogiendo poco.

In spite of their strong similarity with the Spanish reflex passives the English inchoative and middle constructions are not end-prominence configurations. They focus on the process while making the agent implicit. Thus, the only inconsistency of (34b) below, which is an exact Spanish rendering of (34a), lies in the difficulty of using the Spanish equivalent of “in the process” to refer to the time span denoted by the verb of the first clause:


(34)

a. b.

143

The door opened with difficulty and in the process one of the hinges broke off the frame. ?La puerta se abrió con dificultad y mientras lo hacía uno de los goznes se soltó del marco.

So far, our account of se reflex passives and their contrast with the English inchoative and middle constructions has a descriptive status. The next subsection will account for the language-internal and cross-linguistic contrasts in terms of common underlying principles (cf. Section 3).

5.2. Experientialist explanation Ruiz de Mendoza and Pérez (2001) have dealt with the cognitive motivation of the English inchoative and middle constructions in terms of the high-level metonymies PROCESS FOR ACTION and PROCESS FOR ACTION FOR RESULT respectively. Ruiz de Mendoza and Peña (2006) further argue that the Spanish se and se evaluative constructions are also grounded in metonymy. The advantage of this kind of account is to be found in its explanatory elegance – both within and across the two languages in question – while being sensitive to the cognitive commitment. Let us start with the contrast between the English inchoative and the Spanish se reflex passive. One of the approaches taken in structural linguistics has been provided by Alarcos (1980), who argues that the reflex passive is not essentially different from other reflexive constructions with co-referential subject and object: (35)

a. b.

Juani sei afeitó (con mucha maña) (‘John shaved himself (very skillfully)’). La puertai sei abrió (con mucha facilidad) (‘The door opened (very easily)’).

If this proposal is correct, the strategy to avoid mentioning the agent in (35b) is to think of the true object as an actor acting upon itself, thus giving rise to the meaning implication that the actual nature of the agent is immaterial. This strategy is grounded in a combination of high-level metaphor and metonymy. First, we have the object-actor metaphor; then, we use the construction thus developed in a metonymic way by making it stand for the canonical actor-process-object conceptual configuration. However, the link

144


is not a direct one, but rather is mediated by a highlighting process of the resultative component of the figurative action. We thus have a figurative action (‘the door opening itself’) that stands for the result of the action (‘the door is open’) that in turn stands for the canonical configuration (‘someone actually opened the door’). We label this high-level metonymy grounded in metaphor (UNREAL) REFLEXIVE ACTION FOR RESULT FOR (ACTUAL) NON-REFLEXIVE ACTION. A comparable implication is obtained in English through intransitivization, which is a grammatical mechanism that allows us to see the object of an action as if it were the actor of a process. This operation is essentially metaphorical and results in the inchoative construction that is then used to stand for the original action against which it is profiled, thus giving rise to the metonymy (UNREAL) PROCESS FOR (ACTUAL) ACTION. Now, we turn our attention to the evaluative reflexive construction, as illustrated by example (28a), repeated here as (36). Consider the construction in the light of the ungrammaticality of (37): (36)

Este pan se corta con dificultad (‘This bread cuts with difficulty’).

(37)

*Este pan se corta (‘This bread cuts’).

Despite the similarity between this construction and the reflex passive, the obligatory status of the evaluative element points to a different underlying motivation. The same situation holds in English for the middle evaluative: (38)

This bread cuts with difficulty.

(39)

*This bread cuts.

We may postulate for the English middle evaluative construction the highlevel metonymy (UNREAL) PROCESS FOR (REAL) ACTION FOR RESULT. This metonymy is an extension of PROCESS FOR ACTION and motivates two related grammatical constructions. In the first one the scope of the evaluative element is the first domain of the metonymic chain (the unreal process), as revealed by the following paraphrase of (37): (40)

It is difficult to cut this bread.


145

In the second kind of middle evaluative, the scope of the evaluative element is the last domain of the metonymic chain, i.e., the result of the real action: (41)

These clothes wash well.

(42)

*It is well to wash these clothes.

In (41) it is not the process but the result of washing that is assessed by well. For this reason a paraphrase like the one in (40) is not possible. In Spanish the evaluative reflexive construction shares structure with the reflex passive but the metonymic chain is more complex: ‘bread cuts itself’ >

UNREAL ACTION >

‘bread gets cut’ > ‘bread is cut’ >

PROCESS >

RESULT >

‘someone cuts the bread’ REAL ACTION

Just as in English, either the process or the result can be within the scope of the evaluative element. In (36), the scope is the process, while in (30a) above we have an example of assessed result. Again, as with the English middle evaluative, we can have a paraphrase for (36), but not for (30a), based on the split structure test: (43)

a. b.

Es con dificultad que se corta este pan (‘It is difficult to cut this bread’). *Es bien que se lava la falda roja en la lavadora nueva (‘It is well to wash the red skirt in the new washing machine’).

All the metonymic operations described in this section are constrained by the Extended Invariance, Correlation, and Mapping Enforcement principles. These principles are instrumental in finding adequate source domains, with a plausible structural relationship with their targets, and with as many mappings as are necessary for full interpretation. For example, once we have determined that we want to evaluate the verbal process as in (38) or the result as in (41), it is necessary to shape the construction in such a way that these two elements of the ‘action’ schema acquire a degree of prominence. This is achieved through a process that Croft (1993) has called highlighting, which here has the function of making the relevant

146


element fall within the scope of the evaluative mechanism. In a complementary way, the CP and the EIP conspire to ensure that the resulting metonymic chain supplies the best possible source elements for each highlighted target. Thus, in the chain from the unreal action of the bread cutting itself to the real action of someone actually cutting the bread, the process element of being cut naturally precedes (and is therefore an adequate source for) the endpoint or result (i.e., the bread is cut). The Correlation Principle also determines that the evaluative adverbs cannot be dropped, as evidenced by (30b) and (39) above. Without the evaluative elements, we would tend to interpret these sentences as ordinary reflex passives. This interpretation would require us to conceive of non-assessed actions like cutting and washing as actions without an agent, which is obviously more difficult than with ‘opening’ or ‘closing,’ for which we may have an invisible agent. So, in both of our examples of the evaluative reflex passive the assessed element supplies the best possible source for the final (REAL) ACTION domain. Our description thus reveals the existence of a principled balance between the complexity of the underlying conceptual operation and the number of potential meaning implications where the principles guarantee the suitability of the constructions to their communicative purposes. Remember that what all the constructions under scrutiny have in common is their ability to present a controlled action as if it had happened by itself. The basic constructional strategy to achieve this aim is one of dispensing with the agent. But this strategy would be impossible without an adequately constrained metaphorical and metonymic motivation.

5.3. Pedagogical implications Our discussion has allowed us to determine with accuracy the extent of the equivalence between the inchoative and the reflex passive constructions, on the one hand, and between the se evaluative and the middle evaluative constructions, on the other hand. English learners of Spanish will find the equivalences useful but they will have to be made aware of the subtle differences that have been brought up by our analysis in terms of ICMs: (i)

The four constructions are ways of talking about actions without mentioning the actual agents. In Spanish this is achieved by pretending that the object is both actor and object at the same time. In


(ii)

(iii)

(iv)

147

English we have a different strategy, i.e., pretending that the object is the subject and that there is no object. The English constructions give prominence to the process. The Spanish constructions give greater prominence to the endpoint of the process. The evaluative construction in English is formally based on the inchoative construction but with a compulsory evaluative element. In a parallel way, the evaluative construction in Spanish is formally built on the standard reflex passive also with a compulsory evaluative element. The scope of the evaluative element, both in the case of the English and the Spanish constructions, is either the process or the result of the process.

A reasonable teaching strategy would start with the elements that the English and Spanish have in common and then depart from them into the usage areas where English and Spanish differ. For example, learners may be required to first consider sentence (46) in the light of sentences (43)–(45) and then to indicate why the asterisked case is dispreferred in Spanish: (43)

La puerta del garaje no abre (‘The garage door won’t open’).

(44)

La puerta del garaje no se abre (‘The garage door doesn’t open itself’).

(45)

Si la puerta del garaje no abre bien, insistid (‘If the garage door won’t open well, try again’).

(46)

?*Si la puerta del garaje no se abre bien, insistid (‘If the garage door doesn’t open itself well, try again’).

The learner of Spanish is expected to understand why “no abre” in (43) – which is an exact translation equivalent of the English inchoative – is preferred over the reflexive “no se abre” in (44) to express the idea that the speaker feels unable to open the door. “No se abre” simply indicates the impossibility for the door to open, while “no abre” suggests that the speaker has somehow been involved in unsuccessfully trying to open the door. The conceptual strategy in (44) is not useful to convey the idea of “refusal” of the door to be opened since we have the following underlying chain:

148


‘the door does not open itself’

>

‘the door is not open’

>

‘no one opens the door’

However, in (43) we have a chain that is consistent with the intended meaning through a (now fairly conventionalized) pragmatic implication: ‘the door is not getting opened’ > ‘no one is opening the door’ > ‘no one is capable of opening the door’ In (44) we are interested in the result of the action, while in (43) we focus on the fact that the process has not taken place, so we presume that someone must have tried. Finally, for the same reasons, example (45) is preferred over (46) to convey the idea that the addressee has made some attempt to open the door, but with an evaluative element. In (45) the evaluative adverb applies to the process, suggesting difficulty in its development. In (46) the adverb has the unsuccessful result within its scope. Since it is strange to assess a state of affairs that has not taken place, the resulting sentence is odd.

6. Conclusion In the context of the goal to create explanatorily adequate pedagogical grammars based on explicit teaching, this chapter has argued that Cognitive Linguistics offers both theoretical and applied linguists explanatorily adequate analytical tools that allow them to capture systematically language-internal and language-external similarities and differences. In order to substantiate this assertion, we have discussed the similarities and contrasts between English and Spanish in the expression of diminutive and inchoative/middle values in the two languages from a cognitive semantics perspective. We have done so by studying some constraints on metaphor and metonymy production, the way these two cognitive models interact (metaphor-metonymy and metonymic chains) and the role of high-level metaphor and metonymy in grounding constructional relations. Finally, the paper has discussed some specific implications for language teaching of theory-driven contrastive analysis in systematically identifying areas of difficulty in terms of different conceptualization strategies.


149

Notes 1.

2.

This research has received financial support from the Spanish Ministry of Education and Science, grant no. HUM2004-05947-C02-01/FILO, which is co-financed through FEDER funds. I wish to express my gratitude to René Dirven and Sabine De Knop for their preliminary comments on this chapter. Any remaining flaws are my own responsibility. BNC and Google searches show that while “push/shove/fling/sling” combine with “across” in non-metaphorical uses, “push” only occurs metaphorically with “ideas,” as in Ed envisioned a bus that could push the idea across the nation. This observation is consistent with the fact that the metaphor IDEAS ARE OBJECTS often combines with UNDERSTANDING IS GRASPING (AN OBJECT). The source in IDEAS ARE OBJECTS can only have objects intended for manipulation.

References Alarcos Llorach, Emilio 1980 Valores de /se/. Estudios de Gramática Funcional del Español. Madrid: Gredos. Alonso, Amado 1951 Noción, emoción, acción y fantasía en los diminutivos. Estudios Lingüísticos: Temas Españoles, 195–229. Madrid: Gredos. Bley-Vroman, Robert 1989 What is the logical problem of foreign language learning? In Linguistic Perspectives on Second Language Acquisition, Susan M. Gass and Jacquelyn Schachter (eds.), 41–68. Cambridge: Cambridge University Press. Bohn, Ocke-Schwen 2005 A fond farewell to the critical period hypothesis for non-primary language acquisition. In In Search of a Language for the MindBrain: Can the Multiple Perspectives be Unified? Anjum P. Saleemi, Ocke-Schwen Bohn, and Albert Gjedde (eds.), 285–310. (The Dolphin 33) Aarhus: Aarhus University Press. Croft, William 1993 The role of domains in the interpretation of metaphors and metonymies. Cognitive Linguistics 4 (4): 335–370. Dirven, René 2001 English phrasal verbs: Theory and didactic application. In Applied Cognitive Linguistics II: Language Pedagogy, Martin Pütz, Susanne Niemeier, and René Dirven (eds.), 3–27. Berlin/New York: Mouton de Gruyter.

150


2005

Major strands in Cognitive Linguistics. In Cognitive Linguistics: Internal Dynamics and Interdisciplinary Interaction, Francisco J. Ruiz de Mendoza Ibáñez and M. Sandra Peña Cervel (eds.), 17–68. Berlin/New York: Mouton de Gruyter. Fillmore, Charles 1985 Frames and the semantics of understanding. Quaderni di Semantica 6 (2): 222–254. Geeraerts, Dirk 2006 Methodology in Cognitive Linguistics. In Cognitive Linguistics: Current Applications and Future Perspectives, Gitte Kristiansen, Michel Achard, René Dirven, and Francisco J. Ruiz de Mendoza Ibáñez (eds.), 21–49. Berlin/New York: Mouton de Gruyter. Gibbs, Raymond W., and Marcus Perlman 2006 The contested impact of cognitive linguistic research on the psycholinguistics of metaphor understanding. In Cognitive Linguistics: Current Applications and Future Perspectives, Gitte Kristiansen, Michel Achard, René Dirven, and Francisco Ruiz de Mendoza Ibáñez (eds.), 211–228. Berlin/New York: Mouton de Gruyter. Goldberg, Adele E. 1995 Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press. Grady, Joseph 1997 THEORIES ARE BULDINGS revisited. Cognitive Linguistics 8 (4): 267– 290. Hakuta, Kenji, Elen Bialystok, and Edward Wiley 2002 Critical evidence: A test of the Critical Period Hypothesis for Second Language Acquisition. American Psychological Society 14: 31–38. Hawkins, Roger 2003 “Representational deficit” theories of (adult) SLA: Evidence, counterevidence and implications. Paper presented at the 13th Annual Conference of the European Second Language Association, Edinburgh. Heyvaert, Liesbet 2003 A Cognitive-Functional Approach to Nominalization in English. Berlin/New York: Mouton de Gruyter. Jackendoff, Ray 1997 The Architecture of the Language Faculty. Cambridge, Massachusetts: Massachusetts Institute of Technology Press. Kövecses, Zoltan, and Günter Radden 1998 Metonymy: Developing a cognitive linguistic view. Cognitive Linguistics 9: 37–77.


151

Lakoff, George 1987 Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press. 1990 The Invariance Hypothesis: Is abstract reason based on image schemas? Cognitive Linguistics 1 (1): 39–74. Lakoff, George, and Mark Johnson 1980 Metaphors We Live By. Chicago: University of Chicago Press. 1999 Philosophy in the Flesh. New York: Basic Books. Langacker, Ronald W. 1988 A usage-based model. In Topics in Cognitive Linguistics, Brygida Rudzka-Ostyn (ed.), 127–161. Amsterdam/Philadelphia: John Benjamins. 2005 Construction Grammars: cognitive, radical, and less so. In Cognitive Linguistics. Internal Dynamics and Interdisciplinary Interaction, Francisco Ruiz de Mendoza Ibáñez and Sandra Peña (eds.), 101– 159. Berlin/New York: Mouton de Gruyter. Lázaro Mora, Fernando A. 1976 Compatibilidad entre lexemas nominales y sufijos diminutivos, Boletín Instituto Caro el Cuervo, XXXI: 41–57. Maldonado, Ricardo 1999 A Media Voz: Problemas Conceptuales del Clítico Se. México D. F.: Universidad Nacional de México. Michaelis, Laura 2003 Word meaning, sentence meaning, and syntactic meaning. In Cognitive Approaches to Lexical Semantics, Hubert Cuykens, René Dirven, and John Taylor (eds.), 93–122. Berlin/New York: Mouton de Gruyter. Montes Giraldo, José Joaquín 1972 Funciones del diminutivo en español: ensayo de clasificación. Boletín Instituto Caro el Cuervo, XXVII: 71–88. Pérez, Lorena, and Francisco Ruiz de Mendoza 2002 Grounding, semantic motivation, and conceptual interaction in indirect directive speech acts. Journal of Pragmatics 34 (3): 259–284. Queller, Kurt 2001 A usage-based approach to modeling and teaching. In Applied Cognitive Linguistics II: Language Pedagogy, Martin Pütz, Susanne Niemeier, and René Dirven (eds.), 55–83. Berlin/New York: Mouton de Gruyter. Radden, Günter, and René Dirven 2007 Cognitive English Grammar: The Simple Sentence. Amsterdam /Philadelphia: John Benjamins.

152


Robinson, Peter (ed.) 2001 Cognition and Second Language Instruction. (Cambridge Applied Linguistics). Cambridge: Cambridge University Press. Rudzka-Ostyn, Brygida 2003 Word Power: Phrasal Verbs and Compounds, A Cognitive Approach. Berlin/New York: Mouton de Gruyter. Ruiz de Mendoza, Francisco 1998 On the nature of blending as a cognitive phenomenon. Journal of Pragmatics 30 (3): 259–274. 2000 El modelo cognitivo idealizado de ‘tamaño’ y la formación de aumentativos y diminutivos en español. In Estudios Cognoscitivos del español, Ricardo Maldonado (ed.), 335–374. Asociación Española de Lingüística Aplicada, Spain, and Universidad Autónoma de Querétaro, Mexico. 2007 High level cognitive models: in search of a unified framework for inferential and grammatical behavior. In Perspectives on Metonymy, Krzysztof Kosecki (ed.), 11–30. Frankfurt/Main: Peter Lang. Ruiz de Mendoza, Francisco, and Annalisa Baicchi 2007 Illocutionary Constructions: Cognitive motivation and linguistic realization. In Explorations in Pragmatics: Linguistic, Cognitive, and Intercultural Aspects, Istvan Kecskes and Laurence Horn (eds.), 95–128. Berlin/New York: Mouton de Gruyter. Ruiz de Mendoza, Francisco, and Olga Díez 2002 Patterns of conceptual interaction. In Metaphor and Metonymy in Comparison and Contrast, René Dirven and Ralph Pörings (eds.), 489–532. Berlin/New York: Mouton de Gruyter. Ruiz de Mendoza, Francisco and Sandra Peña in press Grammatical metonymy within the “action” frame in English and Spanish. In Current Trends in Contrastive Linguistics: Functional and Cognitive Perspectives, María de los Ángeles Gómez-González, Lachlan Mackenzie, and Elsa González-Álvarez (eds.). (Studies in Functional and Structural Linguistics) Amsterdam/Philadelphia: John Benjamins. Ruiz de Mendoza, Francisco, and Lorena Pérez Hernández 2001 Metonymy and the grammar: Motivation, constraints, and interaction. Language and Communication 21: 321–357. Ruiz de Mendoza, Francisco, and Ricardo Mairal 2007 High-level metaphor and metonymy in meaning construction. In Aspects of Meaning Construction in Lexicon and Grammar, Günter Radden, Klaus-Michael Köpcke, Thomas Berg, and Peter Siemund (eds.), 33–49. Amsterdam/Philadelphia: John Benjamins.


153

Ruiz de Mendoza, Francisco, and Francisco Santibáñez 2003 Content and formal cognitive operations in construing meaning. Italian Journal of Linguistics 15 (2): 293–320. Singleton, David 2005 The Critical Period Hypothesis: A coat of many colours, IRAL 43, 269–285. Berlin: Walter de Gruyter.

Spanish middle syntax: A usage-based proposal for grammar teaching Ricardo Maldonado

Abstract This paper argues that the analysis of the clitic se in Spanish as a reflexive pronoun misrepresents the overall functions that the clitic displays. Instead it is proposed that while having a reduced number of reflexive uses the clitic se is a middle voice marker. Middle voice is defined as a system of constructions whose main property is to portray events remaining in the subject’s dominion. The subject is in most cases an experiencer. It is proposed that the crucial function of the middle voice marker is to highlight the affectedness undergone by the experiencer and this focusing function leads the way for two general subschemas. The first subschema involves self-directed actions and the second subschema depicts the pivotal moment of change-of-state. While the first subschema accounts for grooming actions and self-beneficial events, the second accounts for focal change-of-state events depicting inchoative and spontaneous events as well as emotional changes and changes of location. Further developments of the general focusing strategy are explained in terms of the absolute/energetic contrast (Langacker 1991) as it accounts naturally for the existence of sudden, abrupt and even unexpected construals. Finally, the paper also identifies the most problematic areas for the English speaker learning Spanish and suggests that the middle constitutes a natural, motivated and coherent system which can be exploited in the learning process in a productive and efficient way. Keywords: reflexive; middle; reflexive vs. middle; middle construction; clitic se; grooming action; self-beneficial event; emotional state; internal emotion; mental image; inchoative; motion; experiencer subject; experiential dominion; second language acquisition; gustar verbs; emotional reaction; body part; alienable vs. inalienable possession; dominion; focus construction; unexpected event; counter-expectation; participant involvement; speaker involvement; subjectivity; subjectification; grammaticalization; absolute event; energetic event; Spanish; English

156

Ricardo Maldonado

1. Introduction It would be hard to remain blind to the fact that “small rules” in grammar constitute a considerable problem both for linguistic analysis and for grammar instruction. Big rules are elegantly presented by almost any theoretical model in terms of important generalizations that are supposed to represent the most salient organization of a language. Smaller rules tend to be seen as deviations from the cannon or as forms with aberrant behavior that deserve little attention to the extent that they can be obscured in a footnote, listed as exceptions or simply hidden under the carpet. However, leaving aside smaller rules may be an obstacle both for linguistic in-depth research and for grammar teaching. If the student is left with a big rule and a list of exceptions with no motivation or no internal coherence, he is simply left alone with a skeletal representation of a language with no history, no culture and no creativity. In this paper I will argue that the traditional approach to the clitic se in Spanish has been particularly inefficient precisely because in most analyses the clitic is treated as a reflexive marker with a general rule of subject-object coreference. This strategy has left behind a considerable number of so-called “exceptions” with no internal coherence. The clitic is indeed coreferential, se marks third person singular and second and third person plural. It represents the whole class of coreferential clitics (me 1SG, te 2SG, nos 1PL, and for Spain os 2PL).1 Traditional analyses (Aid 1973; Alonso and Henríquez Ureña 1953; Gili Gaya1955; Goldin 1968; González 1985; Grimshaw 1982; Sells, Zaenen, and Zec 1986; Butt and Benjamín 2004; and many others) as well as instructional textbooks (Terrell, Andrade, and Egasse 2006; Canteli Dominicis and Reynolds 1994; King and Suñer 1999; Alonso, Castañeda, Martínez, Miquel, Ortega, and Ruiz 2005 to name a few) analyze examples like (1) as a clear reflexive construction while those like (2) are exceptions treated separately from the general coding pattern of the language: (1)

a.

b.

Valeria se vio en el espejo. Valeria RFLX.3SG saw in the mirror ‘Valeria saw herself in the mirror’ Estando en México Valeria se vio actuando en Londres. Being in Mexico Valeria RFLX saw 3SG dancing in London ‘Being in Mexico Valeria saw herself dancing in London’

Spanish middle syntax

(2)

a.

b.

157

Don Nico se murió. Don Nico MID.3SG died ‘Don Nico died (unexpectedly)’ Tachita ya se fue. Tachita already MID.3SG went ‘Tachita has already left’

Reflexives involve transitive verbs where subject and object are coreferential, as in (1). They allow a split representation of the only participant in the event in such a way that the subject and the coreferential object may be in separate mental spaces, as is Valeria in (1b). Under traditional views, the cases of (2) are “deviant reflexives.” Given that in (2) the verbs morir ‘die’ and ir ‘go’ are intransitive there is no point in arguing for a reflexive analysis. One must question whether the reflexive is the general construction from which other uses can be derived. Among the wide variety of “deviant reflexives” only self care “grooming verbs” (lavarse la cara ‘wash one’s face,’ peinarse ‘comb one’s hair’) show a reflexive feature, namely coreferentiality. All other uses share fewer reflexive properties with reflexives and need an alternative analysis. In this paper I suggest that “deviant reflexives” are in fact middles, a grammatical category that is well known in classical studies of Greek and other languages. Middles, however, have not received enough attention in current linguistic analyses until recent cognitive studies of syntax (Kemmer 1993, 1994; Maldonado 1988, 1992, 1993, 1999; Manney 2001, Nava and Maldonado 2005) have pointed out its relevance for grammatical analysis. As I will try to show, the middle construction accounts for the set of non-reflexive meanings depicted by the clitic se. In this paper I will illustrate the kinds of errors commonly made by native English speakers learning Spanish in order to identify the conceptual problems of a variety of non-reflexive se constructions. The examples used in this paper are extracted from a database of 1,382 examples of native speakers of English learning Spanish. These are errors I have collected over the years as a Spanish instructor at the University of California at Santa Barbara. Moreover, I have compared my sample with 176 errors generously provided by three prominent Spanish instructors in three different parts of the world: Martha Jurado (Mexico City), Marta Montemayor (San Antonio, Texas), and Alejandro Castañeda (Granada, Spain). I have selected in the four locations the most representative examples that showed up. Since my only goal in using these examples is to illustrate the kind of errors commonly found in the production of second language learners, I

158

Ricardo Maldonado

have not done any statistical analysis and I make no claim about the second language acquisition process. There are important reasons to focus on the non-reflexive uses of se. Foreign language learners of Spanish hardly ever produce reflexive errors like *Yo veo en el espejo ‘I see in the mirror’ with the meaning of (1a). In contrast under-use and over-use of middle markers show up frequently in everyday use. In cases like (3) a middle marker is missing, while in (4) the use of middle marker me is ungrammatical: (3)

a.

b.

(4)

Quiero informar que a partir de hoy dejo la coordinación. ‘I want to inform (you all) that as of today I am no longer the coordinator’ … pero no vas, ¿verdad? [pero no te vas, ¿verdad?] but not go, truth? but not MID.2SG go, truth? Intended reading: ‘but you are not leaving, right?

*No me puedo lavar el coche. not MID.1SG can wash the car Intended reading: ‘I cannot wash my car’

[no puedo lavar] cannot wash

In contrast with previous approaches, in this paper I argue for a grammar instruction strategy that recognizes the middle as a well-organized and coherent system. I propose that middles must be introduced as opposed to reflexives and that using the reflexive as the base form to derive middle constructions is misleading. The middle marker develops meanings totally unrelated to reflexives. Observe again example (2a), the middle se marker imposes a reading of unexpectedness and in (2b) it develops an inchoative reading: the trajectory followed along a path characteristic of go is reduced to simply signaling the moment in which the subject leaves some place. Notice that the ungrammaticality of (3b) corresponds to the absence of a middle marker depicting the inchoative reading. In (4) the improper use of se by a learner of Spanish may be an overgeneralization of self-directed actions as in Yo me cuido ‘I take care of myself.’ I will attempt to identify the types of situation where middles are most common in Spanish and I will show how they are related to each other in a coherent system. This should provide the language instructor with an organized set of situations that may facilitate the learning process in foreign language learners. The paper is structured as follows: Section 2 is devoted to distinguishing reflexive from middle construals and to show the ways in which the


159

two event types tend to be marked in the languages of the world with special reference to Spanish. Section 3 offers a grammar strategy for teaching middle constructions. This section is divided into four subsections showing the behavior of the most problematic middle constructions for the second language learner. More particularly, 3.1 shows the behavior of self-directed action in transitive verbs. Section 3.2 further analyzes middle transitive constructions depicting full exploitation of the object as well as increased participation of the subject. The behavior of middles with verbs of emotional reactions is dealt with in 3.3. A brief note on the gustar ‘like’ type of verbs is made in 3.4. Verbs of motion are treated in section 3.5. Section 4 offers a basic didactic application for middle constructions while Section 5 contains some concluding remarks on the middle construction and its acquisition, as well as its relevance for pedagogical grammars.

2. Reflexive and middle se Reflexives are defined as constructions in which the subject and the object of a transitive verb share the same referent, as in (1). Consequently the reflexive construction normally exhibits intransitive behavior. Still it can be shown that reflexive constructions are not totally intransitive. They exhibit an intermediate degree of transitivity as the subject acts on the self. We can see from the Yucatec Mayan example (taken from Martínez and Maldonado 2006) that although the reflexive construction is intransitive, it preserves the notion of subject control characteristic of a transitive clause. The intransitive verb lúub as used in (a) designates the mere act of falling. In contrast the reflexive construction of (5b) depicts an act of falling that the subject controls: (5)

a.

b.

(j)-lúub-Ø-Ø le wakax-o´. PERF-fall-COMP-B3SG DEM cow-DEM ‘The cow fell down’ Waan-e´ t-u-lúub-s-(i)k u-ba. P3SG-RFLX Juan-DEM DUR-A3SG-all-CAUS-IMPERF ‘Juan lets himself fall (so as not to fall down from the tree)’

Spanish reflexives not only preserve subject control but also involve a complex representation of the event in which there is some concrete or abstract split. Kemmer (1993) suggests that in reflexives there are two distinguishable entities put in correspondence via coreferentiality. In ex-

160

Ricardo Maldonado

ample (6) there is a split representation of Vale: one facing the other. It is quite common for split reflexives to show up in separate mental spaces (Fauconnier 1985) as in (7): (6)

Vale se enfrentó consigo misma. Vale RFLX.3SG faced with-she same, i.e., with herself ‘Vale faced herself’

(7)

Me imaginé bailando con Tongolele. RFLX imagined dancing with Tongolele ‘I imagined myself dancing with Tongolele’

This split representation has grammatical consequences. For emphatic or contrastive purposes se can be expanded by the prepositional phrase a mí mismo ‘to myself,’ a ti mismo ‘to yourself,’ a sí mismo ‘to himself/herself,’ in exactly the same manner that object pronouns can be expanded by their corresponding prepositional phrases: OBJECT EXPANSION

REFLEXIVE EXPANSION

lo ‘3rd’ Æ a él ‘to him’ te ‘2nd’ Æ a ti ‘to you’ me ‘1st’ Æ a mí ‘to me’

se Æ a sí mismo ‘to himself te Æ a ti mismo ‘to yourself’ me Æ a mí mismo ‘to myself

(8)

Es cierto, lo respetas a él, pero no te respetas a ti mismo. ‘It is true, you respect him, but you don't respect yourself’

In contrast, middle constructions, as defined below, do not profile the way an agent acts on himself, instead they focus on the change of state undergone by the experiencer. This is reflected by the fact that middles do not have a split representation and cannot be expanded by a mismo phrase. (9)

Me enfermé al salir de la fiesta. ‘I got sick as I left the party’

(10) *Me enfermé a mi mísmo al salir de la fiesta. Intended reading: ‘I sickened myself as I left the party’ At this point I must clarify that the term ‘middle’ can have two meanings corresponding to two different linguistic schools. In the English grammar tradition it refers to constructions like This car sells well, where the proper-


161

ties of the thematic subject car are seen as being responsible for the outcome of the event. I will call this the English Middle construction (EMC). The EMC is well attested in a variety of languages and in Spanish it is marked by the clitic se. See Maldonado (1992, 1999) and Ruiz de Mendoza (this volume) for two alternative analyses within the cognitive linguistics tradition. In a less restrictive use of the term “middle” which is largely characteristic of typological studies, the EMC is but one of a variety of middle constructions that depict actions, events or changes of state pertaining to the subject’s own dominion (Maldonado 1992, 1999). The notion of dominion is defined as a virtual area to which some participant has mental or physical access to manipulate, control or have mental contact with a set of objects located within it (Langacker 1991). Now, the “middle voice shows that the action is performed with special reference to the subject” (Smyth 1956: 390). In Benveniste’s (1950: 149) words the subject “is indeed inside the process of which he is the agent.” Current crosslinguistic data suggest that more than focusing on the agent, middle constructions depict the change of state undergone by the experiencer in the event (see examples (15)–(18) below). While transitive active constructions correspond to situations where two participants (most commonly agent and patient) interact, middle voice marking corresponds to situation types involving only the subject, typically an experiencer. Middles differ from reflexives in that the possibility of distinguishing two separate images of the same participant is either low or non-existent (Kemmer 1993, 1994). In (11) the experiencer undergoes some change-of-state where she is the only participant. It is not the case that Ceci does something to put herself down, she simply undergoes an emotional change. Likewise in (12) the grandfather simply became happy because of the weather conditions: (11) Ceci se deprimió. Ceci MID.3SG depressed ‘Ceci got depressed’ (12) Con un día así de bello el abuelo se alegra. With such a beautiful day the grandpa MID.3SG glad ‘With such a beautiful day grandpa becomes glad’ The contrast of these middle examples with reflexives is dramatic. The emphatic situation in which Ceci willfully depresses herself would be en-

162

Ricardo Maldonado

coded by a reflexive construction expanded with the mismo ‘self’ phrase as in (13). The same is true for the grandfather’s action in (14): (13) De tanto decirse que era insoportable, Ceci acabó deprimiéndose a sí misma. ‘After so much telling herself that she was unbearable, Ceci ended up depressing herself’ (14) El abuelo se alegra a sí mismo recordando su juventud. ‘Grandpa gladdens himself remembering his youth’ The Spanish contrast between reflexives and middles parallels what is found in two-form languages (Kemmer 1994). In these languages (e.g., Hungarian or Russian), middles and reflexives get a different marker and the reflexive is normally longer than the middle. Haiman (1983) has defined this type of contrast as “iconic” in the sense that the degree of complexity of the event is reflected by the degree of elaboration and complexity of the morphosyntactic marking. Long forms are consistently used to express reflexive construals that involve a volitional act imposed on the self. In contrast, short forms never mark reflexive split-representations of the same participant; instead they mark middles with the same type of meaning attested for the Spanish middle clitic se. the short forms construe the event as spontaneous, short, sudden, unplanned or even unexpected. Well-known cases of such a contrast are the Hungarian reflexive pronoun magat versus the verbal suffix -kod- or -koz- and the Russian reflexive pronoun sebja versus the verbal suffix -sja, which Haiman (1983: 797) gave as examples of an iconic contrast: Hungarian (15) Meg-üt-ött-e mag-á-t PERF-hit-PAST-3SG SELF-his-ACC ‘He hit himself' (valami-be) (16) Bele-üt-koz-öttPERF-hit-self-PAST-3SG.INDEF (something-ILLATIVE) ‘He bumped into something’


163

Russian (17) On utixomiril sebja. He pacified RFLX (‘His better nature prevailed over his enraged self’ = ‘He controlled himself’) (18) On utixomiri+sja. He pacified+MID (‘He settled down after sowing his wild oats’ = ‘He calmed down’) Langacker and Munro (1975) suggest that for Uto-Aztecan languages the reflexive object is distinct from the subject when it is a separate word (the long form), but non-distinct when it is expressed as an affix on the verb. Haiman (1983), on the other hand, claims that the contrast between long and short forms is determined by iconic motivation where the linguistic separateness of an expression is said to correspond to the conceptual independence of the object or the event which it represents. The long forms of Russian and Hungarian depict reflexives like Spanish se (a sí mismo), while the short markers cover the same type of phenomena encountered for the se form without a mismo phrase. Russian, Hungarian and Spanish are not exceptional in encoding the reflexive/middle contrast. It is also found in Greek (Manney 2001), Dutch (van der Leek 1991), Turkish (Haiman 1983), Tarascan (Nava and Maldonado 2005), Yucatec Mayan (Martínez and Maldonado 2006), and many other languages reviewed by Kemmer (1993, 1994). In all these languages it is never the case that the short form encodes a reflexive construction. Though Spanish cannot be identified as a two-form language, it follows – just like the other Romance languages – the pattern of two-form languages in allowing the reflexive object to be expanded into an oblique phrase like sí mismo. These expansions, normally used for emphatic purposes, cannot apply to middle constructions with a short form. (19)

Tachita se paró (*a sí misma). ‘Tachita stood (*herself) up’

The use of a sí misma in (19) would derive a reflexive reading where Tachita would have to use her arms or other means as if her legs would be injured or paralyzed. To the extent that she acts on herself as she would act

164

Ricardo Maldonado

on another participant the construal with a sí misma in (19) is reflexive not middle. Of particular interest is the fact that middle constructions tend to appear in about the same type of situations across the languages of the world. As pointed out by Kemmer (1993, 1994), there are situations that lend themselves to be coded in middle terms. One can expect internal emotions and mental images to be coded as middle constructions as they happen within the experiencer subject. This is in fact the case: verbs of emotion and emotional expression generally constitute the prototype, as in Classical Greek olophyre-sthai ‘lament.’ There are other situations involving no other participant than the subject that render themselves to be coded by middle marking. Table 1, taken from Maldonado (in press), contains the group of situations where the middle marker se is used. These situations match those identified by Kemmer (1993) in a wide variety of languages: Table 1. Spanish basic middle voice situations

Interaction limited to body part or inalienable possession ~ grooming or body care

Lavarse ‘wash,’ peinarse ‘comb’

Self-benefit actions ~ benefactive middle Conseguirse ‘get,’ allegarse ‘obtain’ Non-translational motion ~ change in Pararse ‘stand up,’ sentarse ‘sit body posture down,’ voletarse ‘turn,’ estirarse ‘stretch out’ Change in location ~ translational motion Irse ‘leave,’ subirse ‘get on top of something,’ meterse ‘go into’ Internal change (emotional) ~ emotional Alegrarse ‘gladden,’ entristecerse reaction middle ‘sadden,’ enojarse ‘become angry’ Verbal actions manifesting emotions ~ emotive speech actions

Quejarse ‘complain,’ lamentarse ‘lament’

Internal change (mental) ~ cognition middle

Acordarse ‘remember,’ imaginarse ‘imagine’

Changes of state whose energetic source is not identified ~ spontaneous events

Romperse ‘break,’ quebrarse ‘crack,’ cerrarse ‘close,’ abrirse ‘open,’ etc.


165

For all the cases in Table 1, there is consistency in the way middles are restricted to depicting events remaining in the subject’s dominion (Maldonado 1999, in press). One can expect internal emotions and mental images to be coded as middle constructions since they both designate events happening within the experiencer subject. This is in fact the case, the prototype for the middle are verbs of emotion alegrarse ‘gladden’ and verbal expression of emotions quejarse ‘complain.’ Emotive speech and cognition verbs also take place within the subject’s experiential dominion. Middle marking is thus predicted for all these cases. More problematic are motion and spontaneous events for it is less evident how they correspond to the sole experience of the subject. In previous works (Maldonado 1988, 1993, 1999), I have shown that the main function of the Spanish middle marker is to focus on the pivotal moment of change undergone by the experiencer. Thus, neither the circumstances nor the forces inducing the event are profiled. This property is shared not only by the prototypical cases algrarese ‘gladden,’ enfermarse ‘become sick,’ quejarse ‘complain,’ but also by physical changes like irse ‘leave,’ and pararse ‘stand up.’ Since the focus is on the result rather than on the process (as also observed by Ruiz de Mendoza in this volume), inchoative, inceptive and spontaneous events are part of the middle system in languages that have such a system. In other languages they are commonly encoded by aspectual markers or by intransitive verbs. In languages like English they may be expressed by intransitive verbs sadden, stand up or by the aspectual verb get as in he got sick, he got happy. In languages across the world inchoative meanings are commonly encoded by middle markers and Spanish follows that strategy. I suggest that motion and spontaneous middle events develop from the focusing properties of the middle construction as only the experiencer’s change-of-state is being focused upon. In the same way that middles focus on the emotional change of the subject, they also focus on the crucial point in which the physical change takes place. The energy used to produce a change is not evident either because it is applied internally (pararse ‘stand up,’ sentarse ‘sit down’) or because it is not profiled, as in spontaneous events (romperse ‘break’). What matters is the change itself. Middles share the property of focusing on the change of state undergone by an experiencer subject. They may focus on a change of position, a change of location or a change of state. Thus, the event will be compressed in different ways according to the meaning of verbs to which the middle marker applies. Verbs of emotion will depict sudden or abrupt changes, verbs of motion and location will reduce the path. I would suggest that a grammar showing such internal coherence pattern should provide the

166

Ricardo Maldonado

basis for language instruction. In what follows I propose to motivate the acquisition of middles by presenting them in groups where the focus applies to different conceptual categories. I will show that the type of errors most commonly found in foreign language learners evidences the fact that this basic focusing function has not been made clear to them. I will suggest the type of observations the instructor could make to facilitate the acquisition of a cumbersome area that, over the years, has been made even more difficult by assuming an inadequate reflexive baseline.

3. An instructional path for middle constructions It would be pointless to impose all the previous categories on the second language learner in one shot. It makes sense to introduce different parts of the system independently to allow learners to build a grammar based on internal coherent groups that have related schemas (see Langacker, this volume). I suggest presenting two big groups of middles, which must be subdivided into smaller constructions. The first group corresponds to all the constructions that have some resemblance to the reflexive construction. These involve self-care and self-benefaction events. The second group corresponds to change-of-state focusing, a category quite distant from reflexives. This group involves emotional, physical and locative change. I suggest that instead of having a list of random uses the two groups of constructions may facilitate the acquisition of the whole spectrum of intricate constructions. Although it is true that the learner will have to deal with smaller details, the internal coherence of the two groups should facilitate the learning process. This is a task that has been made particularly difficult due to the lack of transparent mental correspondences among a variety of constructions. As I introduce each group I will show typical errors made by foreign speakers. Since the variety of phenomena to be found depends a great deal on the native language of the learner and his level of Spanish proficiency, I will focus on advanced English speakers learning Spanish.

3.1. Self-directed actions: Transitive verbs The behavior of transitive verbs with self-directed actions may constitute the most obvious case where middles and reflexives overlap. Although the theoretical problem of distinguishing middles and reflexives in this area is


167

not an easy job, the instructional task is not so complicated. The clitic se corefers with the subject and is consistently the experiencer indirect object. In this construction the direct object is normally an inherent object, an inalienable possession, as in (9a) and (19b), or some (concrete or abstract) meaningful object, as in (19c), that is changed as a result of the experiencer’s affectedness. Unquestionable examples of inherent objects are body parts. As analyzed in Maldonado (1992, 1999) the body part constitutes the active zone (AZ) of the event (Langacker 1991) which may be profiled as the direct object. The clitic se is thus equated with the indirect object as it refers to the whole body. Less specific verbs like bañarse ‘bathe’ (Adrián se bañó ‘Adrian bathed’) do not have an active zone thus se encodes the direct object. With more specific grooming verbs the interaction with body parts is constant. It is part of everyday routines that do not require particular consciousness. As expected the middle construction cannot take the sí mismo expansion: (20)

a. b. c.

Se lavó la cara (*a sí mismo). ‘S/he MID washed his face (*to her/himself)’ Me corté las uñas (*a mí mismo). ‘I MID cut my nails (to myself)’ Me despejé la mente (*a mí mismo) con una buena caminata. ‘I MID cleared my mind (*to myself) with a good walk’

It is important to observe that since the whole body operates as the dominion where the body part is located, the whole, encoded by se, operates as the possessor of the part. Consequently, the se construction precludes the use of a possessive marker, which would be redundant (*Le corté sus uñas a Adrián Lit.: ‘I cut to Adrian his nails’).2 Using the possessive wrongly implies that the face and the nails are independent of the participant. This is particularly problematic for English speakers since they tend to transfer the English possessive construction into Spanish, as in (21a) and (21b), instead of using the middle marker, as in (20a) and (20b): (21)

a. b.

*Lavó su cara. *Corté mis uñas.

The interaction with body parts requires further specification. The middle marker is used only when we interact with a body part using our hands,

168

Ricardo Maldonado

otherwise the construction is simply intransitive (Maldonado 1995). Observe the following contrast: (22)

Tenía tanto sueño que no podía abrir (*se) los ojos. ‘He was so sleepy that he couldn’t open his eyes’

(23)

Amaneció con una infección en los ojos y se (*Ø) los tuvo que abrir con los dedos. ‘He woke up with an infection in his eyes and he had to open them with his fingers’

This pattern is highly productive. We move our eyebrows, our eyes or our shoulders without se as our energy comes from within. The clitic would only be used if we were to use our hands as an external instrument: (24)

Se levantó la pierna porque la tenía dormida. ‘He lifted his leg because it was numb’

Clothing items behave like inalienable objects in a whole-part relationship. They operate as the active zone being profiled and the middle marker represents the whole. Anything happening to the part affects the whole in a positive or negative way. In the following example the affectedness is negative: (25) Me manché la camisa. I MID stained my shirt (26) Te rasgaste el pantalón. You MID ripped your pants off From the basic middle construal with inalienable objects the pattern extends to interactions with alienable objects. Since these objects are not already part of the subject, the clitic se encodes that such object is brought into the subject’s dominion as in (28). Without se the action is done for someone other than the subject as in (27):


(27)

a. b. c.

(28)

a.

b.

169

Sirvió una copa. ‘He served a drink’ Preparó un café. ‘He prepared a coffee’ Compró un pantalón. ‘He bought some pants’ Se

sirvió una copa. RFLX.3SG serve-PST a glass ‘He served a drink for himself [Lit.: ‘He served himself a drink’] Te preparaste un café. RFLX.2SG prepared a coffee ‘You made (a cup) of coffee for yourself’ [Lit.: ‘You made yourself a cup of coffee’]

From the use of the benefactive marker for in the English gloss it can be seen that bringing the object into the subject’s dominion implies that the subject benefits from such an action. Now, if the object ends up being outside the subject’s dominion he will be affected in a negative way. Although there are dialectal differences on which specific verbs can enter the construction3, a transitive construction with negative implications takes the middle clitic se to emphasize affectedness. In what follows I will simplify the gloss by only marking the presence of a middle marker in the clause with the abbreviation MID: (29)

Me dejé la bolsa en la tienda. [Example from Spain] I MID left/forgot the bag at the store’

(30)

Me olvidé las llaves. [Example from Argentina] I MID forgot the keys

(31)

Te perdiste el discurso del director. [Example from Mexico] You MID missed the director’s speech

While in (31) te is already lexicalized in the verb, in (29) and (30) me can be omitted. It is introduced to highlight not only affectedness but also the fact that the action is not intentional (see Section 3.5 for an account of unexpectedness). The middle with alienable objects portrays positive or negative affectedness on the subject as the object either enters or leaves his

170

Ricardo Maldonado

dominion. This crucial property is not evident for non-native speakers of Spanish. A common error made by foreign learners is using the middle marker with plain transitive change-of-state verbs like lavar ‘wash,’ cerrar ‘close,’ and the like, which do not portray the object’s entering or leaving the subject’s dominion. Example (4), repeated here for convenience as (32), is only one of many instances of this phenomenon, as evidenced by (33) and (34): (32)

*No me puedo lavar el coche. Intended reading: ‘I cannot wash my car’

[no puedo lavar]

(33)

*Me cerré la puerta. Intended reading: ‘I closed the door on me’

(34)

*A él le gusta que ella se cambie su personalidad. [cambie] Intended reading: ‘He likes that she changes her personality’

[cerré]

Obviously, learners realize that se is used to highlight self-affectedness and attempt to exploit this strategy as much as they can. However, they are not usually aware that with transitive verbs there is a clear conceptual restriction: middles and reflexives designating positive or negative affectedness only apply to changes imposed on objects located in the subject’s dominion. One such possible change is for the object to leave such dominion. It is this restriction that has to be made clear to learners. A subtle but crucial contrast can finally be observed between alienable and inalienable objects. The interaction with inalienable objects is part of everyday routines; it requires a lower degree of control and volition; and the notion of affectedness is not particularly prominent. This is the expected reading for a typical middle construal where things happen within the subject. In contrast, bringing objects into the subject’s dominion, as is the case for alienable nouns, requires a higher degree of volition and control. Consequently, the notion of positive or negative affectedness is more prominent and the construal falls into the category of reflexive constructions. Surely, the difference is relative rather than absolute. Still, the two polar ends can be easily distinguished. The two schemas can be represented in the following diagrams where the subject/indirect object (S/IO) interacts with the direct object (DO) which, in the case of inalienable object may be a profiled active zone (AZ):


171

S/IO S/IO DO/AZ

DO

DOMINION Figure 1. Middle with inalienable object

Figure 2. Reflexive with alienable object

If the clitic se is imposed on verbs that already designate the content of the schema represented in Figure 2, the beneficial implications already present in the base of the verb will be emphasized. This is the case of conseguir ‘obtain’ and comprar ‘buy,’ which imply – without any further specification – that what is obtained is for the subject. In (35a) getting a job may have good implications, in (35b) and (36) the use of se profiles those beneficial implications, which may be elaborated by adverbs like maravilloso ‘marvelous’ and precioso ‘precious:’ (35)

a. b.

(36)

Adrián consiguió un empleo. ‘Adrian got himself a job’ Adrián se consiguió un empleo (maravilloso).4 ‘Adrian got himself a marvelous job’

Me

compré un pantalón (precioso). pant (beautiful) ‘I bought me a pair of (beautiful) pants’ [i.e., ‘I bought a pair of (beautiful) pants for myself] RFLX.1SG buy a

3.2. More on transitive middles The schema represented in Figure 2 extends to verbs of concrete or abstract consumption to derive a more specific construal: the full exploitation middle, a construction where the subject maximally exploits the object (Maldonado 2000). In verbs of consumption comer ‘eat,’ fumar ‘smoke,’ beber ‘drink,’ tragar ‘swallow,’ etc. and its abstract manifestations saber ‘know,’ aprender ‘learn’ the subject not only brings the object into her/his dominion but also exploits it in different ways. Crucially, the use of the clitic se highlights the fact that the whole object is consumed. I call this the maximal exploitation middle. Thus, in the (37b) and (38b) examples be-

172

Ricardo Maldonado

low, the respective utterances entail that Victor ate the whole portion of meat and that Adrian has read the whole paper. It is also implied that they both accomplished it in a short span of time. The transitive construction without se remains neutral about both the total affectedness of the object and the time efficiency of the middle construction: (37)

a. b.

Victor sólo comió un poco de carne. ‘Victor only ate some meat’ Victor se comió la carne (en tres minutos). FULL EXPLOITATION

c.

(38)

a. b.

‘Victor ate the (whole) meat (in three minutes)’ *Victor se comió la carne y quedó un poquito. Intended reading: “Victor ate up the meat and there is some of it left’ Adrián leyó el periódico con cuidado. ‘Adrian read the paper with care’ Adrián se leía el periódico de una hora. FULL EXPLOITATION

‘Adrian would read the (whole) paper in one hour’ The contrast is parallel to the lexical difference in English between drink and drink up, where the particle entails full exploitation. One could claim that the clitic se is nothing but an aspectual marker that changes activities into accomplishments. This conclusion would be wrong, however, since the meaning imposed by se is quite more specific. Notice that the clitic itself is not responsible for the aspectual change. From the examples in (39) it can be seen that there are accomplishments without se: (39)

a. b.

Sacó de la bolsa la última torta y la comió despacio. ‘He got the last cake from his bag and ate it slowly’ El viejo bebió un trago a pico de botella y le nacieron unas llamitas en las pupilas. ‘The old man drank a sip from the bottle and little flames came out of his pupils’

Full exploitation middles apply only to accomplishments; thus most of the properties of accomplishments must be met in the full exploitation construction. While the restrictions on the object noun are stringent, aspect is


173

more flexible. The event tends to be perfective as in (37b) or (38b) but imperfective events are also possible, as in (40): (40)

a. b.

Se estaba comiendo la carne cuando oyó un disparo. ‘He was eating his meat when he heard a shot’ Don Nico se bebía su tequila antes de comer. ‘Don Nico would drink his tequila before supper’

As for the object it must meet every property expected for accomplishments. For a whole thing to be totally affected it must be clearly identified, isolatable and easy to manipulate. The object must be bounded and individuated although it needs not be specific, thus mass nouns and generics are out (* Se tomó café ‘He drank up coffee,’ Se comió tortillas ‘He ate up tortillas’). This distinction is again not evident for learners of Spanish. They tend to make generic construals with se leaving out the object as in (41), using generic objects (types) as carne in (42) or generic plural nouns as periódicos ‘newspapers’ in (43) which are incompatible with what the transitive full exploitation se middle construction encodes:5 (41)

*En el norte la gente se come mucho. Intended reading: In the north people eat a lot

[come mucho]

(42)

*Tachita no se come carne. Intended reading: ‘Tachita does not eat meat’

[come carne]

(43)

[leyó periódicos] *José Angel se leyó periódicos. Intended reading: ‘José Angel read newspapers’

An important property of the construction is that full exploitation also entails full subject involvement. In some dialects of Latin American Spanish full involvement has extended to verbs of action: (44)

a. b.

Se echó una cena deliciosa. ‘He made a delicious dinner’ Tongolele se bailó una rumba inolvidable. ‘Tongolele danced an unforgettable rumba (with all her might)’

174

Ricardo Maldonado

The middle marker not only indicates that the object is of good quality but also that the subject is highly involved as in (44a) and highly expressive as in (44b). The increase in involvement is clearly attested in the Spanish middle system not only for transitive verbs of consumption but for motion and emotion (see Sections 3.3 and 3.4). The clitic se consistently designates a higher degree of subject involvement as in (45b). Without se the feeling is neutral as in (45a). Notice that adverbial phrases reducing the subject’s involvement would be incompatible with the use of the middle marker, as in (45c): (45)

a. b. c.

Juan compadeció a los muchachos. ‘Juan felt pity for the young men’ Juan se compadeció de los muchachos. ‘Juan felt pity for the young men’ ??Juan se compadeció de los muchachos sin mayor compromiso. ‘Juan felt pity for the young men without major compromise’

In contrast, the use of deliberadamente ‘deliberately,’ which underlines volitional subject control, is acceptable with the se middle construction in (46a) but is questionable in the plain transitive form (46b): (46)

a. b.

Juan se aprovechó de tu experiencia deliberadamente. ‘Juan took advantage of your experience deliberately’ ??Juan aprovechó tu experiencia deliberadamente. ‘Juan took advantage of your experience deliberately’

Thus, aprovechar and aprovecharse differ in that the middle marker emphasizes the increase of participation of the subject in bringing the object into his or her sphere of action (and this to the detriment of some other participant). The following example shows that this is a contrast not commonly acquired by foreign learners of Spanish: (47)

*Me dijeron que un señor aprovechó la niña. [se aprovechó de] Intended reading: ‘I was told that a man took advantage of the girl (i.e., he abused her)’


175

The increase in affectedness and involvement is a defining property of middles. This is also found in emotion and motion middles (see Sections 3.3 and 3.4). I have proposed elsewhere (see Maldonado 1992, 1999) that since all the event energy is “compacted” into one participant the event becomes more ENERGETIC (see also Section 3.5). The higher degree of involvement of the experiencer is one manifestation of this phenomenon; speed and abruptness of the action are two others. Let us first observe the behavior of middle se in verbs of emotion.

3.3. Emotion middles Dynamic emotion middles are easily acquired if we distinguish verbs of emotion such as odiar ‘to hate’ and amar ‘love,’ querer ‘want/love’ from verbs of emotional reaction such as demprimirse ‘get depressed.’ Notice that the use of se with verbs of emotion leads to reflexive construals: (48)

Me

odio

cuando me

RFLX.1SG hate-1SG when

(49)

pongo deprimido.

MID.1SG put-1SG depressed

‘I hate myself when I get depressed’ Pobre hombre, nunca llegó a quererse. poor man, never arrived to love-RFLX.3SG ‘Poor man, he never got to love himself’

Interestingly enough, verbs of emotion behave like verbs of cognition. With the exception of the two verbs recordar ‘remember’ and olvidar ‘forget,’ which denote an involuntary cognitive process, all other verbs of cognition and perception denote a conscious state or a voluntary act and can take a reflexive reading (Maldonado 1999): (50)

Yo me entiendo/comprendo/observo. I RFLX.1SG understand/comprehend/observe ‘I understand/comprehend/observe myself’

The Spanish verbs of emotional reaction include alegrarse ‘get happy,’ enojarse ‘get mad,’ divertirse ‘become amused/enjoy,’ frustrarse ‘get frustrated,’ encabritarse ´get furious,’ perderse ‘get lost,’ irritarse ‘get irritated,’ alarmarse ‘be alarmed,’ asustarse ‘get scared,’ emocionarse ‘be moved,’ etc. Middle se is used with these verbs to stress the experiencer’s emotional involvement. In fact, it has been claimed (Maldonado 1999) that

176

Ricardo Maldonado

such a construction contrasts with a dative experiencer construction precisely in that middle se makes the experiencer responsible for his feelings. Example (51) encodes Gabriela’s tendency to be bothered by children whereas in (52) Gabriela actively hates children: (51) A Gabriela los niños le enojan. to Gabriela the children DAT.3SG annoy-3PL ‘Children annoy Gabriela’ (52) Gabriela se enoja con los niños. Gabriela MID.3SG annoy-3SG with the children ‘Gabriela gets mad with children’ Yet one of the most common errors of Spanish learners is to leave out the obligatory marker se with verbs of emotional reaction as in (53): (53)

[preocuparte] *No tienes que preocupar por tu dinero . not have that worry for your money worry-MID Intended reading: ‘You don’t have to worry about your money’

The pattern where the experiencer’s involvement increases is so productive that it extends to a whole set of deponent verbs (i.e., verbs that are passive in form, but active in meaning) that also involve a high degree of participation/expression. Languages with a middle system tend to have a class of deponent verbs without a transitive or intransitive counterpart. Latin (oblivisco-r ‘forget,’ vereo-r ‘tear’) and Turkish (hastal-án ‘get sick’) are only a couple of many languages having deponents (Kemmer 1993). Spanish also has a group of deponent middles that cannot be used in any other voice form (*fue arrepentido ‘was repented,’ *fue jactado, ’was bragged’) but which students can easily remember if they keep in mind that these verbs involve the subject’s higher participation. Notice in (54a), (54b) and (54c) that the subject actively participates in the emotional act: (54)

a. Juan se (*Ø) arrepintió de sus tonterías. ‘Juan regretted his foolish acts’ b. Juan se (*Ø) jactó de sus buenos resultados. ‘Juan bragged of his good results’ c. Juan se (*Ø) quejó de la política económica. ‘Juan complained about the economic policy’


177

These verbs were all intransitive in early Spanish until the 17th century, when the once optional use of se became obligatory and the high degree of the subject’s involvement became grammaticalized or lexicalized as a middle verb.6 Since these verbs come from intransitives, no reflexive derivation applies here. Spanish deponent forms are synchronically construed as basic middles.7 The non-derivative interpretation of middle constructions has already been put forward in current cognitive analyses (Manney 2001 for Modern Greek, Messineo 2002 for Toba, Palancar 2004 for Otomí, Nava and Maldonado 2005 for Tarascan). As for Spanish, deponent verbs designate what I will call energetic readings (see Section 3.5). Emotion middles (preocuparse ‘to worry’) are easily distinguished from emotion verbs (odiar ‘to hate’) used as reflexives (odiarse ‘hate oneself’). Emotion verbs have active subjects, just like verbs of cognition and perception. In contrast, emotion middles always involve verbs of emotional reaction and the clitic se is consistently used to increase the degree of participation of the experiencer in the event, regardless of whether the middle has a transitive or an intransitive counterpart. For grammar teaching, focusing on the higher degree of participation of the experiencer may be more useful than deriving the middle construction from an alleged verbal (in)transitive root and it is certainly better than making the student memorize these verbs as exceptions.

3.4. A note on gustar One phenomenon that is particularly problematic for second language learners is benefaction. Learners tend to overemphasize the experiencer’s involvement in so-called “inverse” verbs (Delbecque and Lamiroy 1996; Vázquez 1995), such as gustar ‘like’ in (55). Here the human experiencer is not the subject. These verbs are particularly problematic for English speakers for several reasons. First, unlike English, the experiencer is a dative indirect object and the stimulus or impulse triggering the human reaction is the nominative subject. Second, these verbs are emotional states – not emotional changes-of-state – and depict human dispositions towards things. The participant’s involvement is already lexicalized in this verb class. Thus, no involvement needs to be emphasized via the clitic se. The most common error is to use both the dative clitic le and the middle marker se, as in (55b), instead of only employing the dative clitic as in (55a):

178 (55)

Ricardo Maldonado

a. A él le gustan las ciudades grandes. big to him DAT.3SG like the cities ‘He likes big cities’ Lit. Big cities appeal to him’ b. *A él se gustan las ciudades grandes. big to him MID.3SG like the cities

The other common error is to use the middle marker instead of the dative as in the following examples: (56) *Nos gustamos bailar. MID.1PL like-1PL dance-INF Intended reading: ‘We like dancing’

[Nos gusta] DAT.3PL like-3SG

(57) *Se encantan con la música. MID.3SG enchant-3PL with the music Intended reading: ‘They love music’

[Les encanta] DAT.3PL enchant-3SG

(58) *A Teresa se fascina con las cámaras [A Teresa le fascinan] DAT.3SG fascinates-3SG digitales to Teresa MID.3SG fascinates with the cameras digital Intended reading: ‘Teresa is fascinated by digital cameras’ The non-native speaker of Spanish needs to learn that this class of verbs has two properties: (i) these verbs denote emotional states, not changes of state; thus no change-of-state can be focused by se; (ii) the experiencer of these verbs being marked for dative8 is already active and does not need to be activated by the middle marker. Thus, the experiencer participant of gustar ‘like,’ encantar ‘enchant,’ and fascinar ‘fascinate’ is incompatible with the middle marker. As we move away from cases that somehow still resemble reflexive constructions to emotional changes, the typically focusing properties of middles gain prominence. In cases involving motion, these properties are even more salient and determine several construals where the event is compressed to highlight the critical moment of change. These will be introduced as the second cluster of middle construals in the following sections.


179

3.5. Motion middles Particularly problematic for the learner of Spanish are events of physical motion. There is a general tendency amongst learners to leave out the middle marker in changes of position, e.g., levantarse ‘stand up,’ changes of location, e.g., subirse ‘get on’ and translational motion, e.g., irse ‘leave.’ The following errors by second language learners illustrate this problem: CHANGE OF BODY POSTURE

(59) *Yo no acuesto temprano como mi madre. [me acuesto] I NEG lay down early as my mother Intended reading: ‘I don’t go to bed early like my mother’ (60) *esa noche levanté asustado. that night raised scared Intended reading: ‘That night I woke up all scared’

[me levanté]

TRANSLATIONAL MOTION

(61) *pero no vas, ¿verdad? but not go, right Intended reading: ‘But you are not leaving, right?’

[te vas]

[me bajé del] (62) *El camión no paró y yo bajé el camión saltando. the bus not stopped and I went off the bus jumping Intended reading: ‘The bus didn’t stop and I jumped off the bus’ As I have already suggested, the main function of the Spanish middle marker is to focus on the core of the event. This schematic representation emerges naturally from the prototype where the change-of-state is focused. To the extent that the event remains in the subject’s dominion, attention is centered on the change-of-state undergone by the subject. The forces driving the event are of secondary importance. When the middle applies to a transitive construction, the attention is focused on the change-of-state such that the initiating force is downplayed. This quite productive pattern leads to spontaneous constructions of the middle in (63b) and (64b) (also called pseudopassives, reflexive passives, inchoatives and so on) corresponding to the transitive examples in (63a) and (64a) respectively. The reduction of the degree of transitivity in (63b) and (64b) is evident:

180

(63)

Ricardo Maldonado

a. b.

(64)

a. b.

El niño rompió la taza. ‘The boy broke the cup’ La taza se rompió. ‘The cup broke’ El niño perdió la pluma. ‘The boy lost the pen’ La pluma se perdió. ‘The pen got lost’

These examples have been analyzed, in formal traditional approaches, as the result of a subject deletion rule (Aid 1973; Aissen 1987; Goldin 1968; González 1985; Grimshaw 1982; Sells, Zaenen and Zec 1986; and many others). Indeed, se focuses on the pivotal moment of change as the agent is eliminated. An obvious consequence is that agent responsibility is eliminated, too. While this holds for cases deriving from transitive verbs, the phenomenon covers a much wider area. What is interesting is the fact that the middle also derives into energetic events from intransitive verbs to focus on the crucial moment of change. Langacker (1991: 389–393) has proposed a basic contrast between ABSOLUTE events such as She was sleeping and She went home depicting processes that do not profile any sort of energy and ENERGETIC ones such as She woke up and She dashed home, where some type of energy is profiled. An energetic view of middles may facilitate the student’s attempt to find a common denominator in a variety of unexplained uses. Events being focused are energetic in that they only show the pivotal moment of change (Maldonado 1988, 1992). Instead of doing the long scanning of the event as going along a path to some place or picking up something from one place and putting it in another location, only the crucial moment of change is profiled. Thus, change-of-position (pararse ‘stand up’ and sentarse ‘sit down’) are typical middle uses in the motion domain. This is in fact the kind of meaning the second language learner should have access to without having to refer to any kind of reflexive construal. A schematic representation may be helpful at this point. Figure 3 represents the unmarked long scanning construction as indicated by the arrow; Figure 4 is the representation of the basic import of se focalizing (the small rectangle) the pivotal moment of change:


S

Figure 3. Absolute

181

S

Figure 4. Energetic Middle

A parallel representation in the abstract domain pertains to verbs of emotional reaction (enojarse ‘get mad,’ asustarse ‘get frightened’). The energetic middle construal in Figure 4 has further extensions; abruptness and unexpectedness are the most notable ones. Let us look at abruptness first. Since in the middle event we focus on the result we lack information about evolutionary development of the action. Thus, middle energetic construals tend to be conceptualized as rapid or abrupt, as the following contrastive examples illustrate: (65)

a.

Valeria se/Ø despierta diario a las seis de la mañana. ‘Valeria MID.3SG/Ø wakes up everyday at six in the morning’ b. Valeria se (*Ø) despertó por un segundo pero se (*Ø) volvió a dormer. ‘Valeria MID.3SG woke up for a second and she went MID.3SG back to sleep’ c. Juan Carlos se (*Ø) despertó abruptamente gritando de terror. Juan Carlos MID.3SG woke up suddenly shouting in terror’

In (65a) we see that both forms alternate: either the moment of waking up or its resulting state of being up is focused here. But the examples in (65b) and (65c) show that se chooses the pivotal moment of change of state. Now, also sleeping must somehow have had a transition from not sleeping to sleeping – a distinction that can corroborate the absolute/energetic contrast. Whereas (66a) encodes the long-lasting event, (66b) depicts the crucial moment when Daniel falls asleep.9 If Daniel slept through the whole night the clitic se could not be used. This contrast is valid through the whole second subschema which comprises a variety of change-of-state focus constructions:

182 (66)

Ricardo Maldonado

a. b.

Daniel durmió toda la noche. ‘Daniel slept all through the night’ Daniel se durmió en clase. ‘Daniel fell asleep in class’

The cases of ir ‘go’ (67a) and subir ‘go up’ (68b) illustrate the same kind of contrast albeit in the domain of translational motion. Without se they constitute continuous, on-going, absolute actions with a long traceable trajectory. On the other hand, the use of se removes the image of ongoingness and portrays a punctual event irse ‘leave’ as in (67b) or depicts a sudden and abrupt change subirse as in (68b), both of which are energetic: (67)

a. b.

(68)

a. b.

Valeria fue al bar. ‘Valeria went to the bar’ Valeria ya se fue (al bar). ‘Valeria left (for the bar)’ Juan subió el Popocatepetl. ‘Juan went up the P. mountain’ Juan se subió a la silla. ‘Juan got on the chair (jumping)’

The aspectual contrast is attested by the fact that subir can be qualified by adverbs like poco a poco ‘bit by bit’ (69a); in contrast, in the se middle construction the event takes place in one shot as in (69b): (69)

a. b.

(*Se) subió la montaña poco a poco. [not benefactive]10 ‘He went up the mountain bit by bit’ Apareció un ratón y Valeria se subió a la silla de un salto. ‘There appeared a mouse and Valeria got on the chair in one jump’

One should not be surprised to find that this pattern also has metaphorical abstract representations as in the following example: (70)

La caída de la bolsa hizo que el gobierno se desviara de la política económica actual. ‘The stocks drop made the government depart from (i.e., make a sudden change from) the current economic program’


183

We can see now that the middle dynamic construal accounts naturally for the energetic behavior of the se construction. From (68) to (70) we showed a variety of dynamic examples depicting either punctual energetic actions or abrupt and sudden changes. Figures 5 and 6 represent the absolute/energetic contrast in the domain of translational motion. They represent the specific case of subir/subirse:

Figure 5. Subir Absolute

Figure 6. Subirse Energetic

A further extension of this pattern pertains to events that are seen not only as rapid or abrupt but even unexpected. The absolute/energetic contrast is clearly manifested in (71) where a falling event can be seen in an absolute construal, as is the case of the rain simply falling in a neutral manner, versus the energetic view with Adrian falling suddenly, accidentally and unexpectedly: (71)

a. b.

La lluvia (*se) cae. ‘The rain is falling’ Adrián se cayó. ‘Adrian fell down’

Notice that the use of se in the absolute construal (71a) is ungrammatical. It is important to realize that unexpectedness has an extra layer of energy. In this type of construal the speaker’s expectations are put forth. Since the falling event is not seen as common or neutral, Adrian’s fall runs against the speaker’s expectations as the default representation for humans is in the vertical position. We have a force dynamic construal of the event via subjectification. The event is energetic not only in that it happens suddenly but the force-dynamic construal (Talmy 1985) makes it even more dynamic. The encounter of forces has a physical base. For instance in caerse we may fall without resisting gravity (without se) or we may fall despite our resis-

184

Ricardo Maldonado

tance to gravity, as encoded by se. In the latter case the result takes place against our expectations. From that base, more abstract encounters take place. The unexpectedness of falling down with se is corroborated by example (72), where the diver falls in the water volitionally and the clitic se cannot be used: (72)

El clavadista (*se) cayó al agua con toda elegancia. ‘The diver fell in the water elegantly’

Counter-expectation explains the case of morir ‘die,’ which can be seen either as a natural biological event – an absolute construal without se (73) – or as a happening that runs against our expectations, in which case we obtain a force-dynamic energetic event, marked by se, as in (74): (73)

a.

b.

(74)

a. b.

Don Nico murió suavemente, se quedó dormido y ya no despertó. ‘Don Nico died softly, he fell asleep and he didn’t wake up’ Cuando don Nico murió, su hijo ya tenía treinta años. ‘When Don Nico died, his son was already 30 years old’ Don Nico se murió sin que su hijo pudiera hablar con él. ‘Don Nico died before his son could talk to him’ A Juan se le (*Ø) murió su papa. ‘As for Juan his father died on him’

What these examples show is that unexpectedness comes from the way the event is observed. This contrasts with (75), a sentence which does not take se because it is a news report of an accident and thus supposed to be objective, distant and free of the speaker’s expectations: (75)

Un autobús choca en la carretera de Toluca. Mueren 28 personas. ‘A bus crashes on the Toluca highway. Twenty-eight people die’

There is nothing exceptional about this contrast. The clitic se marks two things: (i) focusing on the pivotal moment of change, and (ii) seeing the event as contradicting the speaker’s expectations. The change from coding an event in the world to marking a situation in the speaker’s belief is one of the grammaticalization paths identified by Traugott (1982, 1986, 1988) and


185

Spanish energetic se follows precisely that process of historical change. The pattern is quite productive as can be seen from further cases of forcedynamic events: (76)

En el parto, la cabeza del bebé fue lo primero que (*se) apareció. ‘In the childbirth the head of the baby was the first thing that appeared’

(77)

Juan se (*Ø) apareció en la fiesta sin haber sido invitado. ‘Juan showed up at the party without having been invited’

Since we expect the baby’s head to show up first in the delivery, the use of se in (76) is ruled out. Now for Juan to appear at a party where he has not been invited as in (77) is an unexpected and, most probably, an unpleasant event. The variety of se focusing constructions examined so far suggests the existence of a general schema, represented in Figure 4, from which more specific subschemas emerge. This subsystem corresponds to the gradual organization of focusing functions represented in (78): (78) PIVOTAL MOMENT OF CHANGE-OF-STATE Spontaneous change (romperse ‘break’) > punctual change (curarse ‘recover,’ dormirse ‘fall asleep,’ entirstecerse ‘sadden’) > sudden or abrupt change (subirse ‘get on’) > unexpected change (morirse ‘die’) The general concept corresponds to focusing on the pivotal moment of change but more specific meanings obtain in the various constructions that have been discussed so far. With transitive verbs we may have spontaneous events where only the change-of-state is in focus: romperse ‘break, quebrarse ‘crack’ and so on. As for intransitive constructions we also have spontaneous energetic events depicting punctual changes: enfermarse ‘become ill,’ curarse ‘recover from illness.’ We may also have punctual physical or emotional changes-of-state: dormirse ‘fall asleep,’ despertarse ‘wake up,’ cansarse ‘become tired,’ aburrirse ‘become bored,’ sorprenderse ‘be surprised.’ There may also be sudden changes of location: salirse ‘go out,’ meterse ‘go in,’ subirse ‘go up,’ bajarse ‘go down,’ desviarse ‘deviate,’ soltarse ‘let go,’ irse ‘leave,’ regresarse ‘return,’ etc.

186

Ricardo Maldonado

Most of these verbs have an absolute intransitive counterpart. The most “aberrant” case is the one involving unexpectedness as in caerse ‘fall down,’ morirse ‘die,’ aparecerse ‘show up,’ etc. Particularly interesting is that these last examples match spontaneous constructions where agent responsibility is eliminated via subject deletion (romperse ‘break’ and quebrarse ‘crack’) and which also have a counter-to-expectation reading. This happens as all the inductive forces driving the event are downplayed to simply focus on the change itself. It follows from all this that the second language learner will have to acquire a simple focusing rule that extends to energetic construals of the event in the form of suddenness, speed, energy input and unexpectedness. In fact, these are precisely the kinds of inferences that develop from seeing the change-of-state in one shot without having the relevant information as to how this change-of-state came about.

4. A didactic application While this chapter does not systematically elaborate on methodological issues, a minor suggestion may help find a way to teach these most “aberrant” energetic cases. One may think of a variety of exercises for the learner to attain a good grasp of the absolute/energetic contrast. In all of them the learner must choose between employing the clitic and leaving it out. This can be done in a gap-fill way, using either isolatedcontrastive sentences or short stories providing naturalistic contexts. I have created a story that offers a variety of contexts where the contrast should be evident: La final de Wimbledon. Introduzca en la líneas en blanco la forma correcta del verbo poniendo especial atención a la necesidad de emplear o no el clítico se. Era la final que todo mundo esperaba. León prometía destruir a Daniel e impedir que éste volviera a ganar la copa. Después dos sets ambos jugadores todavía (correr) ____________ de un lado al otro de la cancha con total energía. Empezaron el tercer set bastante angustiados. León sacó con toda velocidad. La pelota (ir) ____________ hasta la pared. En un esfuerzo inigualable, Daniel (ir) ____________ tras ella. Al verlo venir, uno de los jueces abruptamente (subió) ____________ a la silla para que el atleta no lo golpeara con la raqueta. Cuando Daniel estaba a punto de contestar, (estrellar) ____________ contra el muro, (tropezar) ____________ con la silla del juez de línea y (caer) ____________ estrepitosamente sin siquiera


187

meter las manos. La sangre le (correr) ____________ por toda la cara y la temperatura del cuerpo poco a poco le (subir) ____________. Inesperadamente, la madre de Daniel (meter) ____________ a la cancha. León (asustar) _________ y cobardemente prefirió (ir) ____________ de ahí antes de que la madre de Daniel (morir) ____________ de un ataque cardiaco. La madre corrió hacia su hijo, lo abrazó y llorando le gritaba: “No (morir) ____________ hijito de mis amores”, pero el hijito (dormir) ____________ plácidamente. Todo parecía indicar que ya no había esperanza de que regresara pero de pronto Daniel (despertar) ____________. Su madre (sorprender) ________ tanto que de pronto (paralizar) ____________ y, de manera totalmente inesperada, (desmayar) ____________. Daniel estaba nerviosísimo. Le dieron una copa de brandy para que se recuperara del susto y él (tomarla) ____________ de un trago … Su madre llevaba acostada ya cinco minutos. Daniel empezó a sospechar que quizá su madre estuviera fingiendo un desmayo como lo había hecho tantas veces. Daniel (aprender) ____________ de memoria la receta que una vez le dio el doctor para curarla: había que gritar en voz alta que le iban a poner una inyección. Tan pronto como Daniel empezó a decir las palabras mágicas su madre inesperadamente (despertar)________________. Wimbledon final. Fill in the blank with the correct form of the verb. Make sure to use the clitic se where needed. It was the final everyone had been expecting to see. Leon promises to destroy Daniel and stop him from winning the cup again. After two sets the two players were still running (corrían) from one side of the court to the other full of energy. They started the third set anxiously. Leon served. The ball went (fue) all the way to the wall. Daniel ran (corrió) after it. As the referee saw him coming he got on (se subió) the chair so that Daniel would not hit him with his racquet. When Daniel was about to hit the ball back he crashed (se estrelló) against the wall, he stumbled (se tropezó) against the referee’s chair and he fell down (se cayó) loudly without even using his hands. Blood was running (corría) all over his face and his body temperature was going up (subía) bit by bit. Unexpectedly, Daniel’s mother entered (se metió) the court. Leon got scared (se asustó) and in a cowardly manner he chose to leave (irse) before Daniel’s mother would die (muriera) of a heart attack. However, Daniel’s mother spotted her son, ran up to him, and embraced him crying as she shouted “don’t die (no te mueras), my dear son.” But her son had fainted (se había desmayado), hardly breathed and

188

Ricardo Maldonado

seemed to be sleeping (parecía dormir) placidly. It all indicated that there was no hope of regaining consciousness, but all of a sudden Daniel woke up (se despertó). His mother was so utterly surprised (se sorprendio) that she became paralyzed (se paralizó) and in a totally unexpected manner passed out (se desmayó). Daniel was hypernervous. They gave him a glass of brandy to recover from the impact and he drank it up (se bebió) in one shot … His mother had been on the floor for about five minutes and Daniel started to suspect that maybe she was pretending to have passed out as she had done so many times before. Daniel had learned by heart (se aprendió) the prescription the doctor had given him once. He yelled in a loud voice that he was going to give her a shot. As soon as he started saying the magical words his mother woke up (se despertó). This exercise provides enough cases to help learners distinguish between (i) long scanned events and short punctual ones, and (ii) acceptable, though somewhat extreme, events and those that contradict our expectations. All of these instances together represent an area of Spanish language use for which no internal coherence has yet been provided in either current analyses or in instructional grammars. It is hoped that a twin strategy of teaching general conceptual principles and intensive practice in specific construals embedded in a narrative context may help learners to discover and exploit that coherence for themselves.

5. Conclusions The traditional grammatical account of the clitic se is based solely on a reflexive interpretation. In this paper I have tried to show that this approach may not be the best way for second language learners to capture the wide variety of uses that the clitic se covers. I have proposed an alternative view in which the internal coherence of the middle system may help the teacher to get a clearer picture of a cumbersome area in Spanish grammar. This new picture may facilitate ways to introduce different parts of the middle system such that the learner may comprehend the nuclear conceptualization of “focus on the change-of-state, characteristic of simplified events remaining in the subject’s dominion.” The learner of Spanish will also be able to see that this basic pattern spreads in very specific directions and that it does so in a motivated manner. I have proposed the existence of two main subschemas that subsume the entire range of constructions analyzed in this paper: (i) self-directed actions, which include events where


189

the object is brought into the subject’s dominion, and (ii) change-of-state focus. All constructions grow from one of these schemas. The following semantic network shows the basic organization of the middle system in Spanish:

Change-of-state

self-directed actions: lavarse la cara ‘wash one’s face’ Objects brought into the subject’s dominion: comprarse un coche ‘buy a car for oneself’

Pivotal moment of change

Spontaneous: romperse Punctual change of state: curarse ‘recover’ dormirse ‘fall asleep’ pararse ‘stand up’

Full exploitation: comerse el pastel ‘eat up the cake’

Full involvement: bailarse un tango ‘dance away a tango’

Increased participation: quejarse ‘complain’, enojarse ‘get mad’, jactarse ‘brag’

Sudden change: subirse ‘get on’ Unexpected change: morirse ‘die’

The first subschema evolves from self-interactions with the subject’s body parts and other inalienable objects to depict routine actions such as washing and combing one’s hair. As we deal with alienable objects, some beneficial implications are drawn from the fact that the object is brought into the subject’s dominion (comprarse una camisa ‘buy a shirt for oneself’). This kind of interaction leads to a construction with a high degree of transitivity in which the subject maximally exploits the object for his or her own benefit (fumarse un cigarro ‘smoke a cigar’). This kind of action also implies a high degree of subject participation, which sets the basis for constructions of full involvement (bailarse un tango ‘dance away a tango’). One of the most prominent properties of middle constructions is the high degree of subject involvement in the event as internal emotions and emotional reactions become manifested. We may obtain this kind of construal from full exploitation events but the most common source for increased participation comes from the fact that the whole event is condensed into

190

Ricardo Maldonado

the subject and that his or her feelings, needs and thoughts are brought into the open from within. Here we go back to the central properties of the middle construction where the change-of-state is put into focus. This is the most productive pattern and it is instantiated in a variety of realms. The change-of-state focus accounts for an unrestricted number of transitive verbs where the agent is deleted to produce a spontaneous event. It also accounts for all the intransitive verbs of motion that have been listed as exceptional in traditional analyses: change of body posture as well as dynamic and abrupt change of location. Together with events running counter to normal expectations, all of these focus, however, on the actual moment of change and contrast with plain intransitive counterparts that scan the process along a temporal or a physical full-fledged path. I have stressed the fact that via the middle marker the event becomes energetic. Thus, the event is not only abrupt or rapid but also requires a higher level of the subject’s involvement in the event. When that notion is applied to verbs of emotional reaction, the predicted effect takes place. The subject experiencer actively participates in the emotional change. While the reflexive cases are not particularly problematic for the second language learners, the main middle constructions have remained, until now, obscured by the inadequate imposition of a reflexive view that is unable to motivate the existence of a wide variety of uses that are everything but exceptional. I have tried to offer sufficient explanations for the kind of notions a language teacher could use to make the complexities of the middle system a coherent and digestible group of related notions. Once this learning objective has been attained, the possibility of addressing further issues such as reflexive passives and impersonal constructions may be addressed. The assumption that bigger rules at a relatively high schematic level represent the structure of a language has been overprized. Constructional views of language (Croft 2001; Goldberg 1995; Langacker 1987, 1991) have emphasized the fact that language is a structured inventory of constructions which the speaker masters according to specific contexts. General rules are both schematizations and abstractions of smaller rules that group together around related notions. I have offered a general schema that crystallizes first into two groups and then into different specific but interrelated construals that motivate each other. While the schema is preserved for all constructions, the lower-level internal coherence among different uses is such that one construction motivates the acquisition of another. It is this phenomenon that seems to determine the formation or schemas of intermediate generality. This view has been tested in abundant studies in language acquisition by Tomasello and collaborators (1992, 2000, 2003)


191

and in many other publications, where a “piece meal” process of acquisition determines the emergence of more general schemas. Goldberg’s (2006) experimental studies on the acquisition of dative, caused motion and resultative constructions make the same basic point. It all amounts to top-down learning vs. bottom-up learning. I have highlighted that top-down learning using a general subject-object coreference reflexive rule and a list of unrelated exceptional uses is far too distant from mirroring the acquisition process of the clitic se. Learning lists of exceptions does not activate dynamic memory. In contrast, relating smaller schemas not only activates dynamism in memory but it also brings in the possibility of relating those schemas to cultural patterns. Once rules are put into proper contexts in accordance with cultural associations, the learner will be able to integrate language patterns into coherent groups (Tomasello 2003). Admittedly, this paper has not attempted to give any clues as to the ways in which smaller rules may be linked to cultural patterns. I have only provided the appropriate grammar notions for the teacher to take the crucial step of motivating and linking constructional sets. I assume that teaching in context is the best way to facilitate the learning experience. In a usage-based model of grammar, everything comes in context (Langacker 1987, 1991). I have not offered an elaborated suggestion as to the methodological ways to introduce these notions in the classroom. The small exercise that I provided only attempts to offer a sample of the linguistic contexts where energetic construals would contrast with absolute ones. I should hope, however, that this kind of approach will shed some light on a particularly difficult area in the second language acquisition of Spanish grammar.

Notes 1. In this paper the following abbreviations will be used: 1 ‘first person,’ 2 ‘second person,’ 3 ‘third person, CAUS ‘causative,’ COMP ‘completive,’ DEM ‘demonstrative,’ DUR ‘durative,’ IMPERF ‘imperfect,’ MID ‘middle,’ PERF ‘perfective,’ PL ‘plural,’ POSS ‘possessive,’ PST ‘past,’ RFLX ‘reflexive,’ SG ‘singular.’ 2. In informal and Mexican Spanish middles and possessives may co-occur for emphatic purposes – see Maldonado (2002) for further analysis. 3. While dejar ‘leave’ and olvidar ‘forget’ cannot be used with that construction in Mexico, they are perfectly normal in Spain and Argentina. 4. In Spain the verb hacer ‘make’ is used instead of conseguir but the construal is the same: Adrián se hizo con un empleo maravilloso Lit.: ‘Adrian made himself with a great job.’

192

Ricardo Maldonado

5. See Maldonado (2000) for further specifications of the object. 6. Other intransitive verbs show the same behavior. This is the case of enfermarse ‘get sick.’ The verb can still be used intransitively as an inchoative verb as in (1) below. In (2) it is used as a transitive causative verb with the cause in subject position and the passive experiencer in direct object position. However, the middle in (3) is now the norm. By using the middle marker, the passive object me ‘me’ of (2) now becomes an active experiencer in the middle construction of (3), which is now coreferential with the subject yo ‘I’: (1) Fue en ese invierno que mi padre enfermó. ‘It was in that winter that my father got sick’ (2) El exceso de trabajo me enferma. ‘The excess of work makes me sick’ (3) (Yo) me enfermé con el exceso de trabajo. ‘I MID got sick with the excess of work’ 7. Several constructions having a se clitic and an oblique phrase have been analyzed as antipassives by Constenla (1997) and Bogard (1999). In Maldonado (2005) I have argued against that approach. The most outstanding features of antipassives have to do with referentiality, control and aspect as found in ergative languages but all these have exactly the opposite properties in middles. 8. Langacker (1991) defines the dative experiencer as an active participant in the target domain. As opposed to other objects, datives are consistently more active that accusatives, instrumentals and locatives. Now, until the 17th century the subject of these verbs was nominative. There are some traces of the old form in constructions like Antonio gusta de visitar a sus padres ‘Antonio likes visiting his parents.’ Then the construction aligned with other experiencer dative constructions. And there is still a lot of variation in popular dialects Yo gusto de María Lit.: ‘I like of Mary’ or María no se gusta de mí Lit.: ‘Mary does not like of me.’ Given so much variation the need for a clear rule for the second language learner is vital. 9. Ruiz de Mendoza (this volume) suggests seeing parallel phenomena as a contrast between process and result. Although the proposal is in the right direction, the notion of result is not precise enough as it fails to distinguish the pivotal moment of change from the result of an action. The crucial point is that Spanish has independent constructions to depict results, namely the resultative construction with estar ‘be’ or quedar ‘remain’: Está dormido ‘He is asleep’ and Se quedó dormido ‘He fell and remained asleep.’ 10. For some dialects it is possible to say Se subió la montaña ‘He made it all the way to the top of the mountain’ to mean that such an act is a big achievement. This construal belongs to the transitive use of the clitic se as a full exploitation construal. I will give the basic properties of the construal in this paper. For further properties of the construction, see Maldonado (2000).


193

References Aid, Frances 1973 Semantic Structure in Spanish: A Proposal for Instructional Materials. Washington: Georgetown University Press. Aissen, Judith 1987 The Tzotzil Clause Structure. Dordrecht/Boston: Reidel. Alonso, Amado, and P. Henríquez Ureña 1953 Gramática castellana. Buenos Aires: Losada. Alonso, Rosario, Alejandro Castañeda, Pablo Martínez, Lourdes Miquel, Jenaro Ortega, and José Placido Ruiz Campillo 2005 Gramática Básica del Estudiante de Español. Barcelona: Difusión, Centro de investigación y publicaciones de idiomas. Benveniste, Emile 1950 Active and middle voice in the verb. Reprinted in 1971. Problems in General Linguistics, 153–161. Coral Gables: University of Miami Press. Bogard, Sergio 1999 Construcciones antipasivas en español. Nueva Revista de Filología Hispánica XLVII (2): 305–327. Butt, Jon, and Carmen Benjamín 2004 A New Reference Grammar of Modern Spanish. New York: McGraw-Hill. Canteli Dominicis, María, and John J. Reynolds 1994 Repase y Escriba. Curso Avanzado de Gramática y Composición. New York: Wiley. Constenla, Adolfo 1997 Construcciones reflexivas de carácter antipasivo en guatuso y castellano. In Memoria del VI Congreso de Filología, Lingüística y Literatura "Víctor Manuel Arroyo,” Adolfo Constenla, Margarita Rojas Gonzáles, and Carlos Francisco Monge (eds.), 205–213. Universidad Nacional. Escuela de Literatura y Ciencias del Lenguaje. Croft, William 2001 Radical Construction Grammar: Syntactic Theory in Typological Perspective. Oxford: Oxford University Press. Delbecque, Nicole, and Beatrice Lamiroy 1996 Towards a typology of the Spanish dative. In The Dative 1: Descriptive Studies, William Van Belle and Willy Van Langendonck (eds.), 73–117. Amsterdam/Philadelphia: John Benjamins. Fauconnier, Gilles 1985 Mental Spaces: Aspects of Meaning Construction in Natural Language. Cambridge, Massachusetts: Massachusetts Institute of Technology Press.

194

Ricardo Maldonado

Gili Gaya, Samuel 1955 Curso Superior de Sintaxis Española. Barcelona: Editorial Spes. Goldberg, Adele E. 1995 Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press. 2006 Constructions at Work: The Nature of Generalization in Language. Oxford: Oxford University Press. Goldin, Mark 1968 Spanish Case and Function. Washington: Georgetown University Press. González, Nora 1985 Object and Raising in Spanish. New York: Garland. Grimshaw, Jane 1982 On the lexical representation of Romance reflexive clitics. In Mental Representations of Grammatical Relations, Joan Bresnan (ed.), 87– 148. Cambridge, Massachusetts: Massachusetts Institute of Technology Press. Haiman, John 1983 Iconic and economic motivation. Language 59: 781–819. Kemmer, Suzanne 1993 Middle Voice. Amsterdam: John Benjamins. 1994 Middle voice, transitivity and events. In Voice: Form and Function, Barbara Fox and Paul Hopper (eds.), 179–230. Amsterdam: John Benjamins. King, Larry, and Margarita Suñer 1999 Gramática española: Análisis y práctica. New York: McGraw-Hill. Langacker, Ronald 1987 Foundations of Cognitive Grammar. Vol. 1: Theoretical Prerequisites. Stanford, California: Stanford University Press. 1991 Foundations of Cognitive Grammar. Vol. 2: Descriptive Application. Stanford, California: Stanford University Press. Langacker, Ronald, and Pamela Munro 1975 Passives and their meaning. Language 51: 789–830. Maldonado, Ricardo 1988 Energetic reflexives in Spanish. Berkeley Linguistics Society. Berkeley. 1992 Middle voice: The case of Spanish se. PhD dissertation, University of California at San Diego. 1993 Dynamic construals in Spanish. Studi italiani di linguistica teorica e applicata XXII (3). 1995 Reflexividad y niveles de activida. Memorias del II Congres Nacional de Lingüística de la Asociación Mexicana de Lingüística Aplicada, México, 43–65.

Spanish middle syntax 1999

195

A media voz: problemas conceptuales del clítico se en español. México: Instituto de Investigaciones Filológicas, UNAM. 2000 Conceptual distance and transitivity increase in Spanish reflexives. In Reflexives Form and Function, Zygmunt Frajzyngier (ed.). Amsterdam: Benjamins. 2002 Objective and subjective datives. Cognitive Linguistics 13 (1):1– 65. 2005 Surface syntax, systemic imagery. In Imagery in Language: Festschrift in Honour of Professor Ronald W. Langacker, Barbara Lewandowska-Tomasczyk and Alina Kwiatkowska (eds), 187–215. Franfurt/Main: Peter Lang. in press Syntactic voice in cognitive grammar. In Handbook of Cognitive Linguistics, Dirk Geeraerts and Hubert Cuyckens (eds). Oxford: Oxford University Press. Manney, Linda 2001 Middle Voice in Modern Greek: Meaning and Function of a Morphosyntactic Category. Amsterdam: Benjamins. Martínez, Israel, and Ricardo Maldonado 2006 Middles and reflexives in Yucatec Maya. Paper presented at the Conceptual Structure Discourse and Language Conference, 6 November 2007, University of California, San Diego. Messineo, Cristina 2002 La marcación verbal activa/inactiva en toba (guaycurú). LIAMES (Revista de Lingüística Indígena Americana), Vol. 2, Instituto de Estudios da Linguagem, UNICAMP, Brasil, 38–50. Nava, Fernando, and Ricardo Maldonado 2005 Basic voice patterns in Tarascan. In Conceptual Structure and Language, 461–478. Stanford, California: Center for the Study of Language and Information. Palancar, Enrique 2004 Middle voice in Otomi. International Journal of American Linguistics 70 (1): 52–85. Sells, Peter, Annie Zaenen, and Draga Zec 1986 Reflexivization variation: Relations between syntax, semantics and lexical structure. In Studies in Grammatical Theory and Discourse Structure. Vol. 1: Interactions of Morphology, Syntax and Discourse, Lida Masayo, Stephen Weschlerand, and Draga Zec (eds.). Stanford, California: Center for the Study of Language and Information Publications. Smyth, Herbert 1956 Greek Grammar. Cambridge, Massachusetts: Harvard University Press.

196

Ricardo Maldonado

Talmy, Leonard 1985 Force dynamics in language and thought. Papers from the Regional Meeting of the Chicago Linguistic Society 21: 293–337. Reprinted in 2000. Toward a Cognitive Semantics, Cambridge Massachusetts: Massachusetts Institute of Technology Press. Terrell, Tracy, Magdalena Andrade, and Jeanne Egasse 2006 Dos Mundos. Comunidad y Comunicación. 6th ed. New York: McGraw-Hill. Tomasello, Michael 1992 First Verbs: A Case Study of Early Grammatical Development. Cambridge: Cambridge University Press. 2000 Do young children have adult syntactic competence? Cognition 74: 209–253. 2003 Constructing a Language: A Usage-Based Theory of Language Acquisition. Cambridge: Cambridge University Press. Traugott, Elizabeth 1982 From propositional to textual and expressive meanings: Some semantic-pragmatic aspects of grammaticalization. In Perspectives on Historical Linguistics, Winfred P. Lehmann and Yakov Malkiel (eds.), 245–271. Amsterdam/Philadelphia: John Benjamins. 1986 From polysemy to internal reconstruction. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 12: 539–550. 1988 Pragmatic strengthening and grammaticalization. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 14: 406–416. van der Leek, Frederike 1991 Iconicity and two-form reflexive systems. Chicago Linguistics Society 27: 447–463. Vázquez Rozas, Victoria 1995 El Complemento Indirecto en Español. Santiago de Compostela: Universidad de Santiago de Compostela.

What can language learners tell us about constructions? Javier Valenzuela Manzanares and Ana María Rojo López

Abstract This paper has a double goal; on the one hand, it attempts to contribute to the discussion on the ontological status of constructions by reviewing the evidence provided in the relevant literature for the existence of constructions in L2 learners of English. On the other hand, it also attempts to contribute to such research by presenting a number of studies focused on Spanish learners of English, which yield interesting results for the pedagogical applications of construction research. To this purpose, we have designed three studies which employ some of the methods used in the existing literature to investigate constructions. In order to evaluate the existence of constructions in the learners’ linguistic system, we have started by replicating a sorting study by Bencini and Goldberg (2000) using Spanish learners of English as subjects. Secondly, we have investigated the Spanish learners’ use of constructions by examining their performance in the International Corpus of Learner English (ICLE). We focus on one specific construction, i.e., the ditransitive construction, and compare the Spanish learners’ preferences for either the dative or the prepositional versions of the construction with those of English native speakers and two other L2 speakers, i.e., Polish and German. And thirdly, we have designed an acceptability judgement task which aims at investigating whether learners would accept the versions of the ditransitive construction most frequently found in the ICLE as somehow ‘more grammatical’ than other forms of the construction which they used less frequently. Subjects were also requested to provide acceptability ratings for examples which replicated the learners’ misuse of the construction found in the qualitative analysis of the ICLE sentences. The paper provides further evidence for the existence of constructions in L2 learners, and yields results which are coherent with an exemplar-based view of language acquisition and which suggest that Tomasello’s (2003) verb-island hypothesis also applies to foreign language learning. It also shows how constructional knowledge can have an effect on L2 performance, offering thus new insights on the role of constructions in foreign language teaching. Keywords: constructions; foreign language learning; sorting; acceptability judgement tasks; pedagogical implications; corpus study

198

Javier Valenzuela Manzanares and Ana María Rojo López

1. Introduction One of the cognitive traits of human beings is our need for patterns. We need to resort to some sort of ‘cognitive model’ which allows us to impose some order on a complex and ever-changing world. From this perspective, the process of learning a language can also be seen as the process of acquiring the relevant cognitive patterns which codify the conventions of use of that language. In cognitive linguistics, these patterns are called ‘constructions,’ which are regarded as the basic units of linguistic organization. The notion of construction is defined by Goldberg (1996: 68) in the following way: A construction is … a pairing of form with meaning/use such that some aspect of the form or some aspect of the meaning/use is not strictly predictable from the component parts or from other constructions already established to exist in the language.1

According to construction grammarians, our knowledge of a language consists in an inventory of all the constructions in a language (i.e., a “constructicon”). From morphemes to lexemes, to idiomatic units and schematic abstract combinations, all our linguistic information takes the form of constructions. Most constructional approaches2 adopt a usage-based view of grammar to explain how all this constructional information is acquired. In this view, grammar is regarded as “the cognitive organization of one’s experience with language” (Bybee 2005): through the repeated co-occurrence of different linguistic configurations, constructions emerge from usageevents of language speakers. This stance has important consequences not only for how we should conceptualize the acquisition of constructional information but also for other issues such as their mental representation and their ontological status. There is now extensive research on constructions and their role in different aspects of language use. There are plenty of studies dealing with numerous constructions from diverse languages (far too many to cite here; see Barcelona and Valenzuela 2005 for a brief overview). The theoretical advantages of construction grammars over other grammatical approaches have been thoroughly discussed in Goldberg (1995) and Hilferty (2003); Construction Grammar has also been studied in relation to language processing (Bencini and Goldberg 2000; Goldberg and Bencini 2005; Ahrens 2003; Kaschak and Glenberg 2000), and language production (Chang, Bock and Goldberg 2003). There are also quite detailed accounts of first

What can language learners tell us about constructions?

199

language acquisition from a constructional perspective (Abbot-Smith and Behrens 2006; Tomasello 2003). Most of these studies have focused on demonstrating the role that constructions play for native speakers, both in adult usage and in child language acquisition. Recently, however, evidence has also been gathered about the role that constructions could play in the linguistic systems of foreign language learners (e.g., Liang 2002; Waara 2004; Gries and Wulff 2005). This type of evidence not only provides a more solid foundation for the argument in favour of the psychological reality of constructions, but also opens a rather interesting avenue to the possible pedagogical contributions of the notion of constructions. If constructions can be shown to play a role in (foreign) language learning, this research will have clear implications for second language acquisition and pedagogy. Many aspects of foreign language learning, ranging from the role of input frequency3 to how to conduct error analysis, can now be reassessed in the light of this new conception of usage and exemplar-based grammatical systems. The present article attempts to contribute both to the discussion on the ontological status of constructions and to the pedagogical applications of construction research. We review the evidence for the existence of constructions in L2 learners, and at the same time contribute to it by presenting a number of studies focused on Spanish learners of English.

2. A sentence-sorting experiment 2.1. Introduction Bencini and Goldberg (2000) carried out a sentence-sorting experiment which was used to support the existence of constructions in native speakers of English.4 This study has been later adapted and replicated with L2 speakers. For example, Liang (2002) analyzed the behaviour of Chinese speakers of English, and Gries and Wulff (2005) studied German language learners of English. The procedure followed in these three studies was basically the same: subjects were given sixteen sentences, which were built by crossing four different verbs (throw, slice, get and take)5 with four different types of argument structure constructions: transitive (e.g., Pat threw the hammer), ditransitive (e.g., Chris threw Linda the pencil), caused motion (e.g., John threw the key onto the roof) and resultative (e.g., Lyn threw the box apart). The L2 learners were instructed to sort the sixteen sen-

200


tences into four different piles according to the ‘overall meaning of the sentence,’ so that sentences more similar in meaning would be placed together in the same pile. The goal here was to check which element was the main contributor to sentence meaning, the verb or the construction. The sentences could therefore be sorted by grouping together all the sentences that contained the same verb, or alternatively by using a less obvious strategy, based on the type of construction that the verb appeared in. In Bencini and Goldberg’s (2000) first study, 7 out of 17 participants were found to sort entirely by construction; no participant sorted entirely by verb and the other 10 carried out mixed sorts. The way in which the ‘mixed’ sortings were measured was by counting how many cards should be changed in a given pile so that the classification would be purely by construction (Cdev, for Construction Deviation) or purely by verb (Vdev, for Verb Deviation). Thus, a purely verb-based sort would have a score of 0 Vdev (since no card has to be changed from its pile) and a score of 12 Cdev (since three cards have to be changed from each four-card pile); alternatively, a sort done purely by constructions would have a 0 Cdev and a 12 Vdev. In Bencini and Goldberg’s first experiment, the average number of changes required for the classification to be made by the verb only was significantly6 higher (Vdev 9.8) than the average number of changes necessary to arrive at a classification based totally on the type of construction (Cdev 3.2). However, these results might have been affected by the fact that subjects were initially given examples of how sentences with more or less the same words could in fact have different meanings (e.g., kick the bucket vs. kick the dog). To avoid this kind of influence, the authors repeated the experiment without providing any examples. The second time round, however, the differences were minimized, with the contrast between verb-sort and construction-sort classifications being statistically non-significant (Vdev 5.5 and Cdev 5.7). As mentioned above, Liang (2002) replicated this study with Chinese learners of English. The L2 learners were tested at three different levels of language proficiency: beginners with only two years of instruction in English, intermediate learners who had passed the national entrance examination to university and advanced learners who had passed the Chinese national test for non-English majors. Liang found a significant correlation between the subjects’ level of English and their construction-based sorts. For beginners (N = 46) the mean deviation from a sort entirely based on the verb (Vdev) was 5.8 and the mean deviation from a sort entirely based on the construction (Cdev) was 6.2; intermediate learners (N = 31) showed


201

a mean Vdev of 6.2 and a mean Cdev of 5.3 and advanced learners (N = 33) obtained a mean Vdev of 8.2 and a Cdev of 4.9. These results demonstrated that the higher the subjects’ level of English, the more constructionbased sorts they produced. Gries and Wulff (2005) provided another study with L2 learners, i.e., German learners of English, replicating Bencini and Goldberg’s second experiment (that is, the one that did not provide the subjects with any examples). As mentioned earlier, they used the same set of verbs, though one of them, slice, was replaced by cut, similar in meaning but less infrequent and, probably, better known by foreign speakers. Surprisingly, their results were closer to Bencini and Goldberg’s first study, with foreign language learners focusing predominantly on a construction-based sorting (Vdev 8.50 vs. Cdev 3.45). Gries and Wulff took the analysis a step further and determined the preference of each sentence towards a constructional sorting. They used hierarchical cluster analysis to establish how often each sentence was put into a group with each of the other stimuli. Their cluster analysis revealed no significant preference of any of the verbs for grouping together, although cut displayed a different behaviour from the other verbs, since its mean co-occurrence frequency was more than 40% higher than that of the other verbs. We have replicated this sorting experiment using Spanish learners of English as subjects. Our study not only provides additional results to these previous studies but also investigates a language in which three of the four argument structure constructions studied do not exist. Spanish does have a transitive construction, but lacks any ditransitive, resultative or caused motion constructions which the subjects in the previous experiments might have transferred from their L1 when performing the sorting task.

2.2. Method 2.2.1. Participants Fifty second-year undergraduate students of Translation and Interpreting from the University of Murcia in Spain agreed to participate in the experiment (mean age 19.6). All of them were native speakers of Spanish with the exception of one subject who reported herself as Spanish-Arab bilingual, and two subjects who reported themselves as Spanish-German bilinguals. All of them were fluent in English (mean years of English: 11.02; mean of their last English-language exam score: 8.14 out of 10).

202


2.2.2. Materials and stimuli A set of sixteen cards was prepared for each subject, each one with a different English sentence printed in the center. Each set of cards was accompanied by a questionnaire that students had to fill in with information about their mother tongue, their command of other languages, the number of years of English language instruction and the mark obtained in the last English exam they had taken. Each set of cards was shuffled randomly (using the weave method) and clipped together with its corresponding questionnaire. The different sets were put inside individual envelopes containing three more clips. The sentences were based on those used by Bencini and Goldberg but following Gries and Wulff’s change of the verb slice for the more frequent cut. The total number of sentences, 16, was obtained by crossing the four different verbs (cut, throw, take and get) with four argument structure constructions (transitive, ditransitive, caused motion and resultative). As in the previous experiments, special care was taken not to repeat any other content word apart from the verb across the set of stimuli. Table 1. List of stimuli in the sorting experiment Transitive 1. 2. 3. 4.

Barbara cut the bread Pat threw the hammer Audrey took the watch Michele got the book

5. 6. 7. 8.

Jennifer cut Terry an apple Chris threw Linda the pencil Paula took Sue a message Beth got Liz an invitation

Resultative 9. 10. 11. 12.

Nancy cut the watermelon open Lyn threw the box apart Rachel took the wall down Dana got the balloon inflated

13. 14. 15. 16.

Meg cut the ham onto the plate John threw the key onto the roof Kim took the rose into the house Laura got the ball into the net

Ditransitive

Caused motion

2.2.3. Procedure The participants were tested as a group. Each subject was given one of the envelopes with a randomly shuffled set of cards and the language questionnaire. They were first asked to fill in the questionnaire and then to sort the sixteen cards into four piles of four cards each, based on the overall


203

meaning of the sentence, so that sentences closer in meaning would go into the same pile. They were asked to clip each pile and put them back inside the envelope. They were also told that there were no right or wrong answers, the aim of the experiment being only to investigate how people sort sentences according to their overall meaning.

2.3. Results We analyzed the results following the procedure carried out by Bencini and Goldberg (2000), calculating how many changes would be necessary to achieve either a fully construction-based sort or a fully verb-based sort. In our case, the average number of changes required for a verb-based sorting was 8.94 while the average number of changes for a construction-based sorting was 3.52. These results reveal that Spanish L2 learners also have constructions, since the difference between the two measures was found to be significant when performing a t-test for dependent samples (t = 4.44; df = 49; p < 0.001). Such results are especially interesting in the case of Spanish, a language in which the constructions tested either do not exist at all or at least have a different form. This means that it is less likely that our results were influenced by interference from the students’ native language. So, arguably, when Spanish learners of English grouped the sentences according to the type of construction, no transfer from L1 took place. On the contrary, their sorting can be regarded prima facie as having been exclusively based on their knowledge of L2.7 In Spanish there is no motion construction similar to He cut the ham onto the plate (Spanish lit.*‘Cortó el jamón hacia encima del plato’). Being a verb-framed language, Spanish needs to resort to different means to express the action conveyed in the English preposition onto. The most common strategy employs a second verb (e.g., Cortó el jamón y lo puso en el plato; English lit. ‘He cut the ham and put it on the plate’) but there are also ways to convey the same information by using other lexical means (e.g., Cortó un plato de jamón; English lit.: ‘He cut a plate of ham’).8 Neither has Spanish a resultative construction similar to the one exemplified by the sentence He cut the watermelon open (Spanish lit.: *‘Cortó la sandía abierta’). Again, in order to convey this meaning, Spanish speakers need to resort to an altogether different verb. This is usually a verb which explicitly denotes the result, while the action itself is expressed by means of a manner adverb or a gerund (e.g., Abrió la sandía de un

204


corte/cortándola; English lit.: ‘He opened the watermelon with one cut/by cutting it’). Another possibility is to use an adverbial which expresses the result by specifying the number of pieces that were cut (e.g., Cortó la sandía en dos; English lit.: ‘He cut the watermelon in two’). The only two constructions which exist in Spanish are the transitive and the ditransitive. The English and Spanish transitive constructions are similar in structure (e.g., English He cut the watermelon and Spanish Cortó la sandía). But the Spanish ditransitive has a different form, since it requires the use of a dative clitic doubling construction9 (e.g., English He showed Mary his flat vs. Spanish Le enseñó el piso a Mary), in which the goal argument (i.e., a Mary) is doubled by the clitic le. Moreover, the compulsory use of the preposition a in Spanish to express the goal argument makes the construction more similar to the English prepositional dative construction (e.g., He showed his flat to Mary) than to the double object one.10 These grammatical differences between English and Spanish make it less probable that the Spanish learners who sorted the sentences by type of construction did so on the basis of their L1 knowledge. By analogy with Gries and Wulff’s (2005) study, we also performed a hierarchical cluster analysis (using Euclidean distance as a measure) in order to calculate how often each of the experimental sentences was grouped together with each of the other fifteen sentences. The analysis identifies four clear constructional clusters which illustrate the subjects’ tendency towards a construction-based sorting. In the dendrogram in Figure 1 below, sentences that were most frequently classified together appear closer to each other; the higher the subjects’ agreement, the more to the left they show up in the tree. For example, in the first cluster, we see that subjects tended to group together sentences 14, 15 and 16 quite frequently; sentence 13 also appeared commonly classified within this group, although slightly less frequently than the rest. The fact that the dendrogram does not reveal any significant difference for the only construction which has a direct counterpart in Spanish (i.e., the transitive construction) seems to support our claim that the results from the sorting experiment are unlikely to be due to L1 transfer. Interestingly enough, our dendrogram also shows results similar to those of Gries and Wulff (2005), namely, that sentences with the verb cut (i.e., sentences 1, 5, 9 and 13) revealed a lower tendency towards a constructional sorting than the rest of the sentences of the same construction type. What is more, just as in their study, the only instance of cut to show a greater tendency towards a constructional sorting than a verb-based sorting was the resultative cut (sentence 9). As Gries and Wulff (2005: 194) ob-


205

serve, more empirical evidence would be needed to determine whether or not these differences are interesting from a theoretical point of view. It is worth mentioning, however, that Bencini and Goldberg (2000) report a similar case in their study, in which the verb which is more biased to a verb-based sort is slice.11 0 5 10 15 20 +---------+---------+---------+---------+---------+

25

14

«±

15

««±

16

«° ²«««««««««««««««««««««««««««««««««««««««««««««±

13

«««°

¬

11

«««´«««««««±

¬

12

«««°

¬

²«««««««««««««««««««««««««««««««««±

«««««´«««««°

¬

¬

10

«««««°

¬

¬

3

«««««±

²«««°

4

««««««««««±

¬

2

«««««°

¬

1

«««««««««««°

¬

6

«««««´«±

²«««««««««««°

7

«««««° ²«««««±

¬

8

«««««««°

5

«««««««««««««°

9

²«««««««««««««««««««««±

¬

²«««««««««««««««««««°

1. cut transitive

9. cut resultative

2. throw transitive

10. throw resultative

3. take transitive

11. take resultative

4. get transitive

12. get resultative

5. cut ditransitive

13. cut caused-motion

6. throw ditransitive

14. throw caused-motion

7. take ditransitive

15. take caused-motion

8. get ditransitive

16. get caused-motion

Figure 1. Dendrogram of the sorting experiment

206


2.4. Summary of sorting study The results of this experiment clearly show how Spanish learners of English, when looking for semantic similarity among sentences, tend to rely on the information supplied by the constructional configuration rather than on the meaning of the verb. The strong bias towards a constructional sorting was found to be statistically significant both when comparing the verbbased vs. construction-based reclassification measures and carrying out a hierarchical cluster analysis. As for the latter, the resulting dendrogram splits into four clusters that unmistakably correspond to the four argument structure constructions. Somewhat surprisingly, this was the case for all four constructions, even though Spanish does not have a direct counterpart for three of them (resultative, caused-motion and ditransitive). Our results thus seem to support Gries and Wulff’s (2005) conclusion that constructions do have a psychological status even in the mind of foreign language learners.

3. Analysis of the International Corpus of Learner English 3.1. Research on native-speaker corpora In our second study, we wanted to focus on one specific English construction, namely, the ditransitive construction12. As has been mentioned in the previous section, this is an interesting construction to look at partly because the nearest counterpart in Spanish (the Indirect Object clitic doubling construction) has a different form, making L1 interference effects less likely. There exists a wealth of research on the use of this construction in several languages (e.g., inter alia, Anagnostopoulou 2002; Bleam 2003; Chung and Gordon 1998; Croft 2003; Demonte 1995; Goldberg 1995; Gropen et al. 1989; Haspelmath 2005; Jackendoff 1990; Levin 1993; Pinker 1989; Thompson 1995). Some authors have focused on the construction preferences of individual verbs, investigating the interactions between verbs and the grammatical constructions associated with them, i.e., verb categorization patterns (e.g., Levin 1993; Gropen et al. 1989; Goldberg 1995). A particularly fruitful method for research on verb-specific construction preferences is the corpus-based collostructional analysis (CA) developed by Stefanowitsch and Gries (2003). Traditional corpus linguistic approaches perform a colloca-


207

tional analysis based on raw frequency counts, that is, they simply count how many times a word occurs in a linguistic structure disregarding the overall frequency of the word in the corpus. By contrast, CA measures the association strength of a given verb with a given construction, making it a far more powerful tool. More specifically, to determine collostructional strength, the following four frequencies will have to be calculated: (i) the frequency of the lexeme in a particular construction; (ii) the frequency of the lexeme in all other constructions; (iii) the frequency of the construction with other lexemes; and (iv) the frequencies of all other constructions with other lexemes. Table 2. Collexemes most strongly attracted to the ditransitive construction13 Collexeme Collostruction strength give (461) 0

assign (3) 2.61E-04

allocate (4) 2.91E-06

allow (18) 1.12E-10

tell (128) 1.6E-127

charge (4) 3.02E-04

wish (9) 3.11E-06

lend (7) 2.85E-09

send (64) 7.26E-68

cause (8) 5.56E-04

accord (3) 8.15E-06

deny (8) 4.5E-09

offer (43) 3.31E-49

ask (12) 6.28E-04

pay (13) 2.34E-05

owe (6) 2.67E-08

show (49) 2.23E-33

afford (4) 1.08E-03

hand (5) 3.01E-05

promise (7) 3.23E-08

cost (20) 1.12E-22

cook (3) 3.34E-03

guarantee (4) 4.72E-05

earn (7) 2.13E-07

teach (15) 4.32E-16

spare (2) 3.5E-03

buy (9) 6.35E-05

grant (5) 1.33E-0.6

award (7) 1.36E-11

drop (3) 2.16E-02

Gries and Stefanowitsch (2004) included a study on native speakers’ use of the ditransitive construction using information extracted from the British component of the International Corpus of English (ICE-GB). They provided a ranking of the association between the ditransitive construction and a number of specific verbs (see Table 2 above), which supports Goldberg’s (1995: 38) analysis of the basic sense and extensions of the ditransi-

208


tive. This ranking helps to explain why some verbs are regarded as inherently more ditransitive than others. Out of the thirty verbs listed in Table 2, the first twenty would most probably be felt to be ditransitive by native speakers, whereas intuitions are likely to differ for the last ten verbs. Gries and Stefanowitsch’s (2004) collostructional analysis reveals that give is in fact the verb most strongly associated with the form and meaning of the ditransitive construction. This result is completely compatible with the claim that give is most similar in meaning to the basic ‘transfer’ sense of the ditransitive, according to which ‘an agent causes a recipient to receive a theme.’ Another interesting finding of their analysis is that this basic ‘transfer’ sense is less dominant for the next strongest collocates after give. Instead, these other verbs seem to instantiate different extended senses. For example, tell instantiates the meaning of communication as transfer, whereas offer involves the agent’s willingness to create the conditions for the recipient to receive a theme and show instantiates the meaning of perceiving as receiving.

3.2. Applications of corpus research to foreign language learning One of the applications of this kind of corpus-based research is to use the data obtained from native corpora as a point of reference against which to compare native speakers’ performance on different production tasks with that of foreign language learners. One example of this would be Gries and Wulff (2005: 187–191), who use the analysis by Gries and Stefanowitch (2004) to explain differences of verb priming effects14 in a sentencecompletion task carried out with German learners of English. They found that those verbs which are more strongly associated with the ditransitive construction were more sensitive to priming to the ditransitive, whereas those verbs which were more clearly associated with the prepositional dative were more sensitive to priming to the prepositional construction. Gries and Wulff (2005: 188) reported a strong positive significant correlation between the verb-specific constructional frequencies in the corpus and the verb’s likelihood to be primed to a certain construction. Gries and Wulff (2005: 190) extended the collostructional method even further by applying it to the German equivalents of the verbs used in their priming study. With such analyses they attempted to demonstrate that the verb priming effects were in fact due to the verb’s preferences in L2 (English) and not to the verb’s translation equivalents in the subjects’ L1 (German). Their results


209

revealed a small correlation between the priming effects and the corpusbased indices for German (r2 = 0.05; df =6; p = 0.577). In fact, this correlation was eight times as small as the one between the verb priming effects and the corpus-preferences in English. A different method to apply corpus-based research to foreign language learning consists in comparing data from native speaker corpora to those from a foreign language learner corpus. Unfortunately, since not many of these are available in the market, most researchers are forced to compile their own corpora to meet their own specific needs. One of the few exceptions is the International Corpus of Learner English (ICLE), which we propose to use in the present study. This corpus has already been used in several studies15. For example, Callies and Szczesniak (2006) focused on fifteen highly frequent dative alternation verbs identified in the literature and compared how they were used in both alternating constructions (i.e., ditransitive vs. prepositional) by native speakers (as evidenced in the British National Corpus) and German and Polish learners (as evidenced in the ICLE). Moreover, they also analyzed whether the learners’ constructions were governed by similar restrictions to those which govern the native speakers’ use of the ditransitive and the prepositional dative construction respectively.

3.3. Aim of our corpus study We have taken the work by Callies and Szczesniak (2006) as a starting point for our corpus research. Our study aims to compare native speakers’ preferences for using either the ditransitive or the prepositional dative construction with certain verbs and the preferences of advanced Spanish learners of English. We concentrate on the 12 most frequent dative alternating verbs identified in Callies and Szczesniak (2006) and compare their results for the use of each verb by native speakers (in the BNC) and by German and Polish learners (in the ICLE) with the use by Spanish learners in the ICLE. We also carry out a qualitative analysis of the learners’ mistakes to identify the types of problems that they have when using the ditransitive or the prepositional dative construction.

210


3.4. General description of the ICLE and its Spanish subcorpus The ICLE corpus is the result of a project launched by Sylviane Granger at the University of Louvain in October 1990. It contains 2.5 million words of English written by language learners from 11 different mother tongue backgrounds: Bulgarian, Czech, Dutch, Finnish, French, German, Italian, Polish, Russian, Spanish and Swedish. All the learners are university undergraduates and, thus, are usually young adults in their late teens or early twenties. They all have learned English in a non-English-speaking country, so they are actually learners of English as a Foreign Language rather than learners of English as a Second Language. The corpus includes 3,640 essays with a total number of 2,500,353 words. All the texts are academic essays, mainly of the argumentative type, and have an average length of 705 words. As indicated in the description of the ICLE provided by Granger et al. (2002), this is an important variable to take into account, since this text type has no exact equivalent in professional writing. The ICLE is made up of eleven national subcorpora, each containing an average of 200,000 words. In the general description of the corpus, potential users are warned that the subcorpora are not very big in size, which makes the ICLE only appropriate for researching highfrequency linguistic phenomena. The Spanish subcorpus comprises 251 essays with a total number of 200,376 words. The university centres involved in the Spanish subcorpus are the Universidad Complutense de Madrid, the Centro de Estudios Superiores de Madrid, the Universidad de Alcalá and the Escuela Oficial de Idiomas de Madrid.

3.5. Procedure All the Spanish essays were put together in a subcorpus of 200,376 words. To start with, we focused on the 74 verbs (see Table 3 below) that are most frequently found in the literature on the ditransitive construction (e.g., Gropen et al. 1989; Goldberg 1995; Levin 1993). We included both the verbs typically regarded as ditransitive and those that Levin defines as ‘benefactive,’ i.e., verbs which take the preposition for in the prepositional dative construction (e.g., I baked Mary a cake vs. I baked a cake for Mary): We used the Monoconc program to check for the different tense forms of each of these verbs in the corpus: simple present, third person present


211

singular, simple past, past participle and the –ing form. From the outset, all those verbs which appeared with a frequency lower than 10 were disregarded from the analysis. After discarding the least frequent verbs (most of them with 0 frequency), we were left with 15 verbs. Out of these 15 verbs, we finally focused on the 12 most frequent ones, which coincided with the ones analyzed by Callies and Szczesniak (2006): bring, give, offer, pay, read, sell, send, show, take, teach, tell and write.16 Seven of these verbs also overlap with those listed by Gries and Stefanowitsch (2004) as some of the collexemes most strongly attracted to the ditransitive construction (i.e., give, offer, pay, send, show, teach, tell); the other five (i.e., bring, read, sell, take, write) were not in Gries and Stefanowitsch’s table, but are frequently found in most lists of ditransitive verbs reported in the existing literature. Table 3. List of the ditransitive verbs most frequently found in the literature advance, allocate, allot, allow, ask, assign, award, bake, bequeath, blast, bring, build, cite, cook, cost, deny, earn, e-mail, fax, feed, fling, flip, forward, get, give, grab, grant, guarantee, hand, kick, knit, leave, lend, loan, lob, mail, make, netmail, offer, owe, pass, pay, permit, poke, pose, promise, quote, radio, read, refer, refuse, rent, reserve, sell, send, serve, set back, sew, ship, shoot, show, slap, spin, take, teach, telegraph, telephone, tell, throw, toss, trade, win, wire, write

Every instance in which any of the verb forms selected showed up was then copied into either the ditransitive or prepositional dative construction or both. Every example of either the ditransitive or prepositional dative construction was written down irrespective of whether or not it contained any grammatical mistakes. Later, the ungrammatical sentences were individually analyzed to determine the type of mistake and establish what the errors might reveal about the learners’ acquisition of constructions.

3.6. Results Our results for the 15 ditransitive verbs most frequently found in the ICLE are listed in Table 4 below. This table shows both the total number of times each verb appeared in the corpus and the number of times it was used in a ditransitive double object construction:

212


Table 4. Total frequency vs. frequency of ditransitive double object use for each ICLE verb Total frequency bring

Double object frequency

66

6

get

443

0

give

322

96

leave

86

0

make

556

5

offer

71

25

pay

139

21

read

54

0

sell

16

2

send

19

0

show

228

34

take

303

0

teach

43

14

108

16

94

1

tell write

Next, we compared these results to the native speakers’ use of the verbs that are more strongly associated with the ditransitive construction as reported by Gries and Stefanowitsch (2004) in Table 2 above. We can report the following interesting findings. There are obvious divergences in choice of the ditransitive verbs found in ICE-GB (the corpus used by Gries and Stefanowitsch) and in ICLE, as well as in the frequencies with which these verbs occur. On the one hand, the fact that the sets of verbs to appear in ditransitive constructions differ from each other could be due to the fact that ICLE is a more limited corpus, with fewer words than ICE-GB. So, we have some verbs that are used by native speakers in BNC but not by language learners in the ICLE. On the other hand, the restrictions regarding the subject matter of the ICLE essays could explain the fact that some of the verbs used in ICE-GB do feature in ICLE, but sometimes in different uses from those requiring a ditransitive construction. Despite these differences, some similarities could be found in the relative frequencies of many of the verbs, with items like give, tell, offer, show or teach being among the


213

most frequent verbs found in ditransitive constructions in both corpora. Such similarities indicate that there are some central uses of the ditransitive construction which are shared by both L1 and L2 speakers. In order to determine whether Spanish learners associate some verbs more frequently than others with either a ditransitive (ditr.) or a prepositional dative (prep.) construction, we started by noting down the number of times each verb was used in ICLE with each of the two constructions (see column L1 Spanish in Table 5 below). Table 5. BNC and ICLE frequencies of ditransitive and prepositional constructions L1 English

L1 Spanish

ditr.

prep.

prep.

3

31

4

2

9

19

4

2

give

152

73

47

49

132

68

115

39

offer

9

7

18

7

10

11

17

4

pay

7

6

4

17

5

19

2

12

read

0

0

0

0

0

0

0

0

sell

0

6

1

1

0

3

1

3

send

1

14

0

0

0

9

0

21

show

13

13

28

6

23

5

43

9

take

0

5

0

0

0

9

0

6

teach

32

2

11

3

48

2

11

3

tell

55

0

13

3

49

1

145

1

0

2

0

1

0

0

0

7

write

prep.

ditr.

L1 German ditr.

bring

ditr.

L1 Polish prep.

To be able to compare our results with Callies and Szczesniak’s (2006) data, we eliminated from our list of 15 verbs in Table 4 the three forms get, make and leave. We were then left with the 12 verbs listed in Table 5 below. Finally, we compared our results for these 12 verbs with the data Callies and Szczesniak (2006) reported for native speakers in BNC (see column L1 English) and for German and Polish learners of English in ICLE (see columns L1 Polish and L1 German): When comparing the Spanish learners’ data with those of English native speakers and Polish and German learners, some intriguing patterns in the use of the different verbs were discovered:

214


— Two verbs (i.e., teach and tell) behaved similarly regardless of the learner’s mother tongue, showing a higher number of ditransitive uses than of prepositional dative ones. — Two other verbs (i.e., pay and show) were used differently by native speakers and language learners. While English L1 speakers showed a similar proportion of ditransitive and prepositional dative uses of both verbs, all learners of English, irrespective of their L1, used pay more frequently in prepositional constructions and show in ditransitive ones. — In two other cases (i.e., bring and offer), Spanish learners behaved like German learners but differently from Polish and native speakers of English. The figures for bring in prepositional and ditransitive constructions were similar for both L1 Spanish and L1 German speakers (prep. 2 vs. ditr. 4), the frequency in the ditransitive construction being slightly higher than in the prepositional one. By contrast, L1 Polish learners and native speakers of English used bring more frequently in the prepositional dative construction (L1 Polish prep. 19 vs. ditr. 9 and L1 English prep. 31 vs. ditr. 3). The use of the verb offer was also similar for Spanish and German learners, who, nevertheless, differed from Polish and English speakers. Thus, whereas Spanish and German learners tended to use offer more frequently in ditransitive constructions, Polish learners and English speakers showed similar proportions of the verb for each of the two constructions (L1 Polish prep. 11 vs. ditr. 10 and L1 English prep. 7 vs. ditr. 9). — Finally, Spanish learners differed from all the other learners and native speakers in their use of the verb give. While Polish, German and English speakers used give more frequently in ditransitive constructions, Spanish learners produced about the same number of instances for each construction, with the prepositional one being slightly more frequent (prep. 49 vs. ditr. 47). Certainly, these patterns point to interesting differences and similarities in the constructional systems of second language learners with different L1s. However, due to the small size of the corpus used (less than 200.000 words), they should be regarded as thought-provoking hints in need of further verification. In order to investigate these data in further detail and determine some of the factors which may govern the learners’ use of the ditransitive and the prepositional constructions, we carried out a qualitative analysis of the contexts where each verb appeared in ICLE. The most striking feature of Spanish learners’ ditransitive constructions was the fact that


215

the great majority of these structures used a pronoun as recipient. This could explain, for instance, the difference found in the Spanish learners’ use of the verb offer. Out of the 18 cases in which offer appeared in a ditransitive construction, 16 had a pronoun and 2 a proper noun as recipient. The fact that the ditransitive construction is favoured with pronouns supports Callies and Szczesniak’s (2006) hypothesis of the role syntactic weight plays in constructions with dative alternation verbs (see also Wasow 2002). What is more, such a finding is also interesting from the point of view of second language acquisition. The data seem to suggest that Tomasello’s (2003) notion of ‘constructional islands’17 may be even more specific: learners do not only acquire a construction associated with a specific verb, but rather a specific form of this construction; that is, the data suggest an exemplar-based view of acquisition centred on specific items. Thus, although further research would be needed, it may be possible to hypothesize that language learners master first a specific form of the construction with pronouns, which later on gets extended to proper nouns and other types of phrases. A close examination of the mistakes made in the ICLE utterances also supports the idea that Spanish learners favour the use of the ditransitive construction when there is a pronominal recipient. We found several cases of ungrammatical ditransitive constructions in which learners introduced the recipient using the preposition to (e.g., *These careers offer to the students the opportunity of practicing them). However, in all these cases the recipient that appeared with the preposition was a nominal phrase (e.g., *…offered to the young soldiers a hard life), with the exception of one sentence in which the preposition introduced two coordinated proper nouns (e.g., *She wants to show to Leontes and Perdita a statue). The fact that no mistakes were made where pronouns were involved seems to support the hypothesis that Spanish learners master the ditransitive construction with pronominal recipients earlier than those with other types of recipients. Finally, the qualitative analysis of the ICLE sentences also served to account for some of the differences reported between native speakers and our foreign language learners. As mentioned above, English speakers used the verb pay, for example, in comparable proportions in ditransitive and prepositional constructions whereas Spanish learners favoured the use of pay in prepositional dative constructions. The reason for this tendency among the Spanish learners of English became apparent when we analyzed the particular contexts in which the verb pay appeared in the ICLE corpus. Out of

216


the 17 cases in which pay was used in a prepositional construction, 15 were examples of the construction pay attention to.

3.7. Summary of ICLE analysis Clearly, the results reported above have to be taken with some caution due to the limited size and specific nature of the Spanish ICLE subcorpus that we examined. Still, we can safely conclude that Spanish learners of English employ the ditransitive construction in free production tasks such as essay writing. We found interesting similarities in the kinds of verbs that appeared to be more frequently associated with this construction and those showing up in Gries and Stefanowitsch’s (2004) analysis of the International Corpus of English, a native speaker corpus. It seems that some of the central uses of the ditransitive construction (those associated with verbs such as give, tell, offer, show or teach) are shared by both L1 and L2 speakers. More interesting and detailed results were found when comparing the proportion of using the ditransitive vs. the prepositional version of these verbs in our ICLE subcorpus with, first, the preferences of other L2 speakers, namely German and Polish, and secondly, those of the native speakers as reported by Callies and Szczesniak (2006). Here, we found that some verbs like teach or tell occur, in about the same proportion, with the ditransitive and prepositional uses among both Spanish L2 learners and native speakers. However, this does not hold true for most of the other verbs involved. Such results cannot be easily accommodated by a wordsand-rules approach to language acquisition (Pinker 1999), and seem to point instead to an item-based process similar to Tomasello’s theory of constructional islands. What is more, we found that even within the ditransitive uses of some verbs, there were particular constructional configurations that appeared as considerably more entrenched than others. Thus, the ditransitive option was quite often chosen when the construction involved a personal pronoun (16 out of 18 cases in the ditransitive use of offer); personal pronouns also appeared less frequently in incorrect uses of the construction. This again seems to be an indication that acquisition in foreign language learning also proceeds in an exemplar-based fashion, and that specific configurations of linguistic material are entrenched to different degrees even within the same verb.


217

4. Evidence from an acceptability rating task 4.1. Introduction Finally, an acceptability rating task was designed in order to test whether some of the trends found in the ICLE subcorpus would be confirmed in a task which focused on different linguistic skills. The analysis of the ICLE data had revealed that the ditransitive construction is not a ‘ready-made’ category. The variations found in the use of the construction across different verbs support the argument for the usage-based nature of constructions and for the existence of constructional islands. Moreover, we even found data showing that within one specific verb, the ditransitive construction may appear more frequently with some configurations than with others. As previously mentioned, we observed that, in the case of verbs that allow for the dative-ditransitive alternation, the ditransitive version was favoured with some special configurations, for example, with pronouns. The ICLE Spanish L2 speakers used ‘GIVE PRON THEME’ far more often than any other ditransitive possibility, e.g., ‘GIVE PROPER NOUN THEME’ or ‘GIVE FULL NOUN THEME,’ this last configuration being markedly underrepresented. Therefore, in our last study, we wanted to examine whether this differential use of the construction in L2 learners’ production would also correspond to a difference in their acceptability judgements of different ditransitive configurations involving the presence of a pronoun, a full noun or a proper noun.

4.2. Method 4.2.1. Participants Sixty-eight students of English Philology at the University of Murcia agreed to participate in the experiment. All of them were native speakers of Spanish; they were all in the final years of their degree: 17 in the third year, 19 in the fourth year and 32 in the fifth year. Their level of proficiency in English ranged from intermediate to advanced.

218


4.2.2. Stimuli A list with 36 sentences was compiled using four different verbs, i.e., give, offer, show and tell, all allowing the ditransitive alternation. We also tested the alternation involving different NPs for the recipient: a noun-phrase corresponding to a person (e.g., mother, employee, friend, and son), a proper noun (e.g., Peter, Paul, David, and George) or a personal pronoun (him and her). For an overview, see Table 6. Table 6. Different NP combinations for the prepositional and ditransitive alternations Prepositional version

Ditransitive version

[Verb – NP – PP (to – Full Noun)]

[Verb – Full Noun – NP]

[Verb – NP – PP (to – Proper Noun)]

[Verb – Proper Noun – NP]

[Verb – NP – PP (to –Pronoun)]

[Verb – Pronoun – NP]

Additionally, the subjects’ acceptability judgements were also tested for three different types of ungrammatical sentences, as shown in Table 7. Table 7. Ungrammatical sentences included in the grammatical rating task Construction

Example

[Verb – PP (to – Full Noun) – NP]

*Peter gave to his father a present.

[Verb – PP (to – Proper Noun) – NP]

*Peter gave to David a present.

[Verb – PP (to – Pronoun) – NP]

*Peter gave to him a present.

All of them were anomalous because the Prepositional Phrase corresponding to the Recipient was placed immediately after the verb. They followed the order typical of the ditransitive construction [i.e., Verb-RecipientTheme/Argument] but a preposition was added to introduce the recipient, flouting, thus, the grammatical rule which requires the preposition to be used only in the prepositional dative construction after the theme. These sentences were based on the ungrammatical utterances that subjects had been found to produce in ICLE. Just like the grammatical sentences, the ungrammatical ones were built crossing the four verbs with the three different types of recipient: a noun phrase, a proper noun and a pronoun.


219

In total, there were 9 types of sentences per verb; since there were four verbs, the total number of critical stimuli was 36.

4.2.3. Procedure The participants were tested as a group. Each participant was given a questionnaire with the following instructions: Rate the following list of sentences in terms of acceptability 1. 2. 3. 4. 5.

Horrible; completely impossible Bad, but not completely unacceptable Doubtful, but perhaps acceptable Acceptable, although a bit unnatural Completely acceptable and natural

No time limit was imposed.

4.3. Results The results were analyzed by collapsing all data for the four different verbs. In Table 8, information from the general categories can be seen. The numbers correspond to the mean ratings of all students.18 Table 8. Acceptability rating results for prepositional and ditransitive alternations (collapsing all verbs) Prepositional

Ditransitive

Full noun

4.13

Full noun

4.04

Proper N

3.91

Proper N

4.22

Pronoun

3.68

Pronoun

4.43

*Full noun

2.72

*Proper N

2.31

*Pronoun

2.52

220


To start with, the preference for the ditransitive construction over the prepositional one was found to be much greater in sentences containing a personal pronoun. Ditransitive sentences with a pronoun have a mean of 4.43 compared to a mean of 3.68 for their prepositional counterpart (for both constructions N = 263). A Wilcoxon paired signed rank test (p < 0.001) revealed that such difference in rating was statistically significant. The same preference of a ditransitive construction over a prepositional one can be observed for each of the verbs individually; the preferential difference between the ditransitive and the prepositional option is significant in all of them; this is the case even for the two verbs where the difference is smaller (offer, p = 0.001 and show, p = 0.002). However, this tendency does not hold true for sentences containing an NP with a full noun (which seem to be biased towards the prepositional option) and as for sentences involving a Proper Noun, the difference is not so large as in the case of the ditransitives with a pronoun (ditransitive = 4.22, prepositional = 3.91; again a significant difference, p < 0.001). Besides, the preference for the ditransitive version using Proper Nouns only holds for two of the verbs, i.e., offer and tell (p = 0.008 and p = 0.003, respectively). In the second place, when comparing the ratings of the three types of ditransitives (with full noun, with proper noun and with pronoun), again the Pronoun ditransitive receives higher ratings than the other two versions (and these differences with the other two choices, Full Noun-ditransitive and Proper Noun-ditransitive, are significant, with p < 0.001 and p = 0.007, respectively). Small differences were, nevertheless, found for the different verbs. Whereas the ditransitive with a pronoun was the highest rated option for give, the Pronoun ditransitive for tell and show was rated higher than the Full Noun-ditransitive but there was no significant difference with the Proper Noun-ditransitive. In sentences with offer, no significant differences were found between the three different types. This could be taken as another indication that Tomasello’s (2003) constructional islands are also found in L2 acquisition; different verbs have different degrees of entrenchment with a given construction, suggesting that acquisition proceeds in a usage- and item-based incremental fashion. Thirdly, regarding ungrammatical sentences with a prepositional recipient right after the verb, the version which was considered most acceptable was that involving a full noun NP (e.g., *He gave to his mother a present); the difference with the ratings of ungrammatical sentences with pronouns was significant: p = 0.015. The fact that subjects seem to accept more easily the wrong version of the ‘prepositional phrase’ is not that strange, since the prepositional phrase versions also received higher ratings when pre-


221

sented in a grammatical sentence. The wrong versions with pronouns were accepted more readily than those with proper nouns, again something that is coherent with the fact that pronoun ditransitives were the most generally accepted ones in the grammatically correct sentences. Finally, as we have just pointed out, the ungrammatical sentences which receive the highest acceptability ratings are those with a full noun object. The only exception to this is the group of sentences with the verb tell, in which the version that is best accepted is the one with the pronoun. This could be due to the fact that the ditransitive construction with a pronoun is still the most entrenched pattern for the verb tell, so that subjects are ready to accept it better even when the grammatical make-up of the construction is anomalous.

4.4. Summary of acceptability rating task The results of this study confirm the findings of the analysis of ICLE (see Section 3). The version of the ditransitive construction which appears to be preferentially entrenched in the minds of Spanish learners of English is that containing a personal pronoun. This version was significantly perceived as being more ‘grammatical’ than its prepositional counterpart, and also, ditransitive sentences with a pronoun were regarded as more acceptable than other ditransitive configurations, i.e., those including proper nouns or full nouns. These results are again compatible with approaches to acquisition which adopt an exemplar-based view of grammatical organization, and more difficult to explain from other rule-based approaches. The exemplarbased view may also explain the ratings obtained for the ungrammatical sentences, i.e., those containing a prepositional phrase with to immediately after the verb, instead of its more natural position after the theme argument. The versions that were rated as more acceptable included a full noun in the ungrammatical prepositional phrase, which correlated with the higher ratings also received by full nouns in the (grammatical) prepositional versions. Such was the case for three of the verbs tested, but not for the fourth one, tell, which exhibited a different behaviour. This could once more be taken as a sign of the intricacies of constructional acquisition, in which verb-based island-constraints and exemplar-based specific configurations interact and criss-cross in complex ways.

222


5. Conclusion Constructions are more than just a purely grammatical notion. They reflect a basic cognitive trait of human beings: our tendency to establish patterns which help us impose some order on an apparently chaotic and unpredictable reality. In the case of language, the cognitive structures that arise from our linguistic experience are what we have called here ‘constructions.’ Accordingly, constructions should be viewed and analyzed as any other mental structure: conforming to the known principles of general cognitive organization. As Bybee (2005: 22–23) suggests: A theory based on usage … which takes grammar to be the cognitive organization of language experience can reference general cognitive abilities: the importance of repetition in the entrenchment of neuro-motor patterns, the use of similarity in categorization and the construction of generalizations across similar patterns.

Given such a view of grammar, constructions emerge from the repeated cooccurrence of linguistic configurations during usage-events of language speakers, both in native speakers (e.g., Bybee 2005) and in the linguistic systems of foreign language learners (e.g., Ellis 2003). This confers on constructions a great potential to account for linguistic phenomena in a cognitively natural way. As previously mentioned, most studies on constructions have focused on their role in native speakers of English. By contrast, the present paper joins works such as Liang (2002) or Gries and Wulff (2005) by providing further evidence for their role in the grammatical systems of L2 learners. For a start, the data from our sorting task have clearly revealed that Spanish learners of English rely on constructional information when they have to decide on the overall similarity of a group of different sentences. Secondly, it was found that constructions also play a role in explaining both the frequency of use and idiosyncratic composition of some syntactic configurations present in the ICLE. For example, unlike native speakers of English, Spanish learners used double object constructions much more frequently with a pronoun as recipient than with any other possibility, i.e., with a proper noun or a full noun. Finally, exemplar-specific constructional information has also been shown to influence learners when performing an acceptability rating task: again, the pronoun versions of the ditransitive construction consistently received higher ratings than those including a proper noun or a full noun.


223

Our three empirical studies yield results which can be useful at two levels. At a theoretical level, they can be seen as a contribution to the debate on the psychological reality of constructions; the data obtained in our three studies provide systematic evidence in favour of the psychological reality not only of constructions but of the notion of ‘constructional islands’ postulated by Tomasello (2003) for L1 acquisition. Our studies reveal that the process of learning a foreign language also takes place by acquiring specific constructions which get extrapolated into more general structures later on. In the particular case of ditransitive constructions, the topic of our second investigation, our data indicate that L2 learners seem to have mastered this construction to a greater degree with some verbs than with others. A very specific form of the construction, namely, that involving a pronoun as recipient, also seems to have a special status. Perhaps this construction serves as a ‘landmark’ which could in subsequent stages of linguistic development be extended to other forms of the construction, with different non-pronominal types of recipients. If so, this would be coherent with exemplar-based approaches to grammar. At a more practical level, we can see these experiments as a new set of powerful tools that allow us to analyze the emerging linguistic system of foreign language learners and, accordingly, suggest useful connections between cognitive linguistics and the practice of foreign language teaching, joining other works like Boers and Lindstromberg (2006). Investigating empirically the L2 learners’ use of a given construction provides real data which can help L2 researchers and teachers to determine the stage of the students’ learning process, locate their main problems and establish their needs. If constructional patterns actually play a significant role in the language learning process, then such a role should be acknowledged and reflected not only when setting the relevant objectives of foreign language teaching but also creating useful and effective exercises. Therefore, the results from the present research paper should not be overlooked by second language acquisition experts, foreign language professionals and materials writers, among others. The fact that learners organize their syntactic knowledge around constructions connected to specific verbs is a finding that they should be interested in. Conceiving the processes of foreign language acquisition from a usageand item-based perspective can be reasonably expected to provide a solid theoretical foundation to current teaching approaches that complement the explicit teaching of grammatical rules with more contextualized practice. These kinds of exercises help students acquire the relevant patterns and thus master the conventions of use of a particular language (cf. Broccias,

224


this volume). Furthermore, it is hoped that this focus on patterns and constructions opens up a new avenue of applications of cognitive linguistics and establishes another bridge to its further pedagogical exploitation.

Notes 1. Recently, however, Goldberg (see, e.g., Goldberg 2006) has opted for a more comprehensive definition of construction: constructions should be posited in grammar even if they are fully predictable. 2. Within the cognitive linguistics tradition there are slightly different versions of construction grammar, e.g., Fillmore and Kay (cf. Fillmore et al., in progress), Lakoff (1987), Langacker (1987, 1991), Croft (2001) or Goldberg (1995, 2006), inter alia. For interested readers, the differences among them are discussed in Broccias (this volume), Croft and Cruse (2004), Evans and Green (2006), Goldberg and Bencini (2005) or Schönefeld (2006). 3. Issue 24 of the journal Studies in Second Language Acquisition is a monograph on the role of frequency in second language learning in which psycholinguist Nick Ellis writes a deep and insightful review of frequency effects in second language acquisition and responds to some of the arguments other authors make for and against such effects (Ellis 2002). 4. For the psychological validity of this sorting paradigm, see Bencini and Goldberg (2000: 644). 5. Gries and Wulff (2005) used a slightly different set of verbs, replacing slice for cut; see below. 6. Throughout the rest of the chapter, we will use the term ‘significant’ in its statistical sense, that is, referring to p-values lower than 0.05. 7. Admittedly, the transfer from L1 to L2 is far from easy to determine; Cabrera and Zubizarreta (2005) discuss the role of both lexical and constructional transfers in early and advanced L2 learners; for a slightly different view, see Perpiñan and Montrul (2006). 8. Note that neither of these translation alternatives, however, conveys exactly the same meaning as example 13 above. 9. Also called the Indirect Object Clitic Doubling Construction (Bleam, 2003) 10. For more detailed information on the properties of Spanish double object constructions and their similarities with English ones, see Bleam (2003). 11. “Among the participants who performed mixed sorts, there was only one instance of grouping all of the four sentences which contained the same verb. The verb was ‘slice’.” (Bencini and Goldberg 2000: 648) 12. It is worth mentioning at this point that there is some disagreement concerning the terminology used. Some authors use the term ‘ditransitive’ in a wider sense, encompassing both double object and prepositional object alternations,


13.

14.

15. 16.

17.

18.

225

whereas others use the term in a more restrictive way as a synonym of the double-object case (e.g., Levin 1993 or Goldberg 1995). Thus, in the literature the sentence I sent a package to my parents can be found labelled as a Dative Construction (Perpiñan and Montrul 2006), a Prepositional Dative Construction (Bleam 2003), a Prepositional Object Construction, (Gries and Stefanowitsch 2004) or an Indirect Object Construction (Haspelmath 2005). On the other hand, the sentence I sent my parents a package is found as a Ditransitive Construction (Goldberg 1995; Gries and Stefanowitsch 2004; Levin 1993). The numbers in brackets correspond to the frequency of the verb in the corpus, and the next figure is the “collostructional strength”, that is, the measure of specific association between a given verb and the construction studied (specified by means of the Fisher-Yates Exact Test); for a more detailed explanation of Gries and Stefanowitsch’s Collostructional Method, see Gries and Stefanowitsch (2004). In psychology, when the presence of a given stimulus facilitates the recognition or production of a subsequent one, the first one is said to ‘prime’ the second. For a review of ICLE-based research, see Granger (2003, 2004). Callies and Szczesniak (2006) included three other verbs in their study: carry, hand and pass. However, no examples of these verbs were found in our analysis of the Spanish ICLE subcorpus. Tomasello’s verb-island hypothesis (e.g., Tomasello 2003) postulates that children’s grammars initially consist of inventories of verb-specific predicate structures. In this view, children’s linguistic systems are at first built around verbs which are used with arguments specific to those verbs and not around any other more syntactically-abstract or schematic unit. Therefore, in their speech some constructions may appear used with some particular verbs but not with others. These mean ratings cannot be used in statistical computations and should be regarded as purely indicative since we are dealing with an ordinal scale; hence the use of the Wilcoxon paired signed rank test below.

References Abbot-Smith, Kirsten F., and Heike Behrens 2006 How known constructions influence the acquisition of other constructions: the German passive and future constructions. Cognitive Science 30 (6), 995–1026.

226


Ahrens, Kathleen 2003 Verbal integration: The interaction of participant roles and sentential argument structure. Journal of Psycholinguistic Research 22: 497– 516. Anagnostopoulou, Elena 2002 The Syntax of Ditransitives: Evidence from Clitics. Berlin: Mouton de Gruyter. Barcelona, Antonio, and Javier Valenzuela 2005 An overview of cognitive linguistics. In Nuevas Tendencias en Lingüística Aplicada, Imelda Katherine Brady, Marta Navarro Coy and José Carlos Periñán Pascual (eds.), 197–230. Murcia: Quaderna Editorial. Bencini, Giulia, and Adele E. Goldberg 2000 The contribution of argument structure constructions to sentence meaning. Journal of Memory and Language 43: 640–651. Bleam, Tonia 2003 Properties of the double object construction in Spanish. In A Romance Perspective on Language Knowledge and Use, Rafael A. Núñez Cedeño, Luis López, and Richard Cameron (eds.), 233–252. Amsterdam: Benjamins. Boers Frank, and Seth Lindstromberg 2006 Cognitive linguistic approaches to second or foreign language instruction: rationale, proposals and evaluation. In Cognitive Linguistics: Current Applications and Future Perspectives, Gitte Kristiansen, Michel Achard, René Dirven, and Francisco J. Ruiz de Mendoza Ibáñez (eds.), 305–358. Berlin: Mouton de Gruyter. Bybee, Joan 2005 From usage to grammar: The mind’s response to repetition. LSA Presidential Address, January 8, 2005. Available at http://www.unm. edu/~jbybee/Bybee%20plenary.pdf . Cabrera, Mónica, and Maria Luisa Zubizarreta 2005 Overgeneralization of causatives and transfer in L2 Spanish and L2 English. In Selected Proceedings of the 6th Conference on the Acquisition of Spanish and Portuguese as First and Second Languages, David Eddington (ed.), 15–30. Somerville, Massachusetts: Cascadilla Proceedings Project. Callies, Marcus, and Konrad Szczesniak 2006 Argument realization, information status and syntactic weight: A learner-corpus study of the dative alternation. In Fortgeschrittene Lernervarietäten, Patrick Grommes and Maik Walter (eds.). Tübingen: Niemeyer.


227

Chang, Franklin, Kathryn Bock, and Adele E. Goldberg 2003 Do thematic roles leave traces in their places? Cognition 90 (1): 29– 49. Chung, Ting Ting Rachel, and Peter Gordon 1998 The acquisition of Chinese dative constructions. BUCLD 22: 109– 120. Croft, William 2001 Radical Construction Grammar: Syntactic Theory in Typological Perspective. Oxford: Oxford University Press. Croft, William 2003 Lexical rules vs. constructions: A false dichotomy. In Motivation in Language: Studies in Honor of Günter Radden, Hubert Cuyckens, Thomas Berg, René Dirven, and Klaus-Uwe Panther (eds.), 49–68. Amsterdam: Benjamins. Croft, William, and D. Alan Cruse 2004 Cognitive Linguistics. Cambridge: Cambridge University Press. Demonte, Violeta 1995 Dative alternation in Spanish. Probus 7: 5–30. Ellis, Nick C. 2002 Frequency effects in language processing: A review with implications for theories of implicit and explicit language acquisition. Studies in Second Language Acquisition 24 (2): 143–188. 2003 Constructions, chunking, and connectionism: The emergence of second language structure. In The Handbook of Second Language Acquisition, Catherine J. Doughty and Michael H. Long (eds.), 63–103. Malden, Massachusetts: Blackwell. Evans, Vyvyan, and Melanie Green 2006 Cognitive Linguistics: An Introduction. Edinburgh: University of Edinburgh Press. Fillmore, Charles J., Paul Kay, Laura Michaelis, and Ivan Sag forthc. Construction Grammar. Stanford, California: Center for the Study of Language and Information Publications. Goldberg, Adele E. 1995 Constructions: A Construction-Grammar Approach to Argument Structure. Chicago/London: University of Chicago Press. 1996 Construction grammar. In Concise Encyclopedia of Syntactic Theories, Keith Brown and Jim Miller (eds.), 68–71. Oxford: Pergamon. 2006 Constructions at Work: The Nature of Generalization in Language. Oxford/New York: Oxford University Press. Goldberg, Adele E., and Giulia Bencini 2005 Support from Language Processing for a Constructional Approach to Grammar. In Language in Use: Cognitive and Discourse Perspectives on Language and Language Learning, Andrea Tyler, Mari Ta-

228


kada, Yiyoung Kim and Diana Marinova (eds.), 3–18. Georgetown University Round Table on Languages and Linguistics. Washington, DC: Georgetown University Press. Granger, Sylviane 2003 The International Corpus of Learner English: A new resource for foreign language learning and teaching and second language acquisition research. TESOL Quarterly 37 (3): 538–546. 2004 Computer learner corpus research: Current status and future prospects. In Applied Corpus Linguistics: A Multidimensional Perspective, Ulla Connor and Thomas Upton (eds.), 123–145. Amsterdam; Atlanta: Rodopi. Granger, Sylviane, Estelle Dagneaux, and Fanny Meunier 2002 The International Corpus of Learner English (ICLE): Handbook and CD-ROM. Louvain-la-Neuve: Presses Universitaires de Louvain. Gries, Stefan, and Anatol Stefanowitsch 2004 Extending collostructional analysis: A corpus-based perspective on alternations. International Journal of Corpus Linguistics 9 (1): 97– 129. Gries, Stefan, and Stefanie Wulff 2005 Do foreign language learners also have constructions? Evidence from priming, sorting and corpora. Annual Review of Cognitive Linguistics 3: 182–200. Gropen, Jess, Steven Pinker, Michelle Hollander, Robert Goldberg, and Ronald Wilson 1989 The learnability and acquisition of the dative alternation in English. Language 65: 203–257. Haspelmath, Martin 2005 Ditransitive constructions: The verb “give.” In The World Atlas of Language Structures, Martin Haspelmath, Matthew S. Dryer, David Gil, and Bernard Comrie (eds.), 426–429. Oxford: Oxford University Press. Hilferty, Joseph 2003 In defense of grammatical constructions. PhD dissertation, University of Barcelona. Jackendoff, Ray S. 1990 Semantic Structures. Cambridge, Massachusetts: Massachusetts Institute of Technology Press. Kaschak, Michael P., and Arthur M. Glenberg 2000 Constructing meaning: The role of affordances and grammatical constructions in sentence comprehension. Journal of Memory and Language 43: 508–529.


229

Lakoff, George 1987 Women, Fire and Dangerous Things. Chicago: University of Chicago Press. Langacker, Ronald 1987/91 Foundations of Cognitive Grammar. Vols. 1 and 2. Stanford, California.: Stanford University Press. Levin, Beth 1993 English Verb Classes and Alternations. Chicago; London: University of Chicago Press. Liang, J. 2002 How do Chinese EFL learners construct sentence meaning: Verbcentered or construction-based? MA thesis, Guangdong University of Foreign Studies. Perpiñán, Silvia, and Silvina Montrul 2006 On binding asymmetries in dative alternation constructions in L2 Spanish. Selected Papers from the 7th Hispanic Linguistics Symposium, 135–148. Somerville, Massachusetts: Cascadilla Press. Pinker, Steven 1989 Learnability and Cognition. Cambridge, Massachusetts: Massachusetts Institute of Technology Press. 1999 Words and Rules: The Ingredients of Language. London: Weidenfeld and Nicolson. Schönefeld, Doris (ed.) 2006 Constructions all over: Case studies and theoretical implications. Special Volume 1 – Constructions SV1-1/2006. Stefanowitsch, Anatol, and Stefan Gries 2003 Collostructions: Investigating the interaction of words and constructions. International Journal of Corpus Linguistics 8 (2): 209–243. Thompson, Sandra A. 1995 The iconicity of ‘dative shift’ in English. In Syntactic Iconicity and Linguistic Freezes, Marge E. Landsberg (ed.), 155–175. Berlin: Mouton de Gruyter. Tomasello, Michael 2003 Constructing a Language: A Usage-based Theory of Language Acquisition. Harvard: Harvard University Press. Waara, Rene 2004 Construal, convention, and constructions in L2 speech. In Cognitive Linguistics, Second Language Acquisition and Foreign Language Teaching, Michel Achard and Susanne Niemeier (eds.), 51–75. Berlin/New York: Mouton de Gruyter.

230


Wasow, Thomas 2002 Post-Verbal Behaviour. Stanford, California: Center for the Study of Language and Information Publications.

Conceptual errors in second-language learning Marcel Danesi

Abstract Research on creative writing and overall discourse in classroom second-language (SL) learning suggests that the errors that are most disruptive of comprehension are conceptual, rather than strictly linguistic (phonological, syntactic, etc.) or communicative (interactive and strategic) (Russo 1997; Danesi 2003). Conceptual errors result from the tendency of SL learners to assume that conceptual structures in the native and target languages are encoded in grammatically and lexically parallel ways. For example, students of Italian as an SL assume that “falling in love” can be expressed in a similar way in the target language, typically producing erroneous phrases such as cadere in amore. The study of such errors suggests that an expanded Contrastive Analysis (CA) identifying how grammar, vocabulary etc. reflect differential conceptual structures (or more precisely image schemas) will go a long way towards addressing this problematic aspect of classroom SL learning. Essentially, a revamped CA would have to take into account current work in cognitive linguistics and conceptual metaphor theory into SL instructional methodology and syllabus design. The purpose of this paper is to discuss what a conceptually-based CA would entail, reporting on a writing experiment conducted with students and SL teachers of Italian at Middlebury College, Vermont, USA, in 2003 and 2005. The objective of the study was to compare students exposed to conceptual training versus those trained in other ways (control groups) in terms of the types and tokens of conceptual errors that emerged in assigned writing tasks. The study produced data that has made it possible to classify conceptual errors in various ways and to separate them from other kinds. These will be analyzed, computed, and explained in terms of conceptual metaphor theory. Implications that the results have for pedagogical grammar and syllabus design will then be discussed. Ultimately, the objective of a conceptually-based CA is to ensure that learners have access to the conceptual structures inherent in the target language and culture in a systematic, sequential, and integrated fashion with other areas of language learning. Keywords: error analysis; conceptual errors; SLA; second language acquisition; conceptual metaphor theory; conceptual fluency theory; conceptual training; pedagogical grammar; syllabus design

232

Marcel Danesi

1. Introduction Error-based research on classroom second-language (SL) discourse and writing suggests that the errors that are most disruptive of comprehension are conceptual (meaning-based), rather than strictly “form-based” (phonological, syntactic, etc.) or communicative (interactive and strategic) (Danesi 1993, Russo 1997). Conceptual errors, as they may be called, result from the tendency of SL learners to assume that meanings in the native and target languages are encoded with identical or parallel structures (phrases, idioms, etc.). For example, as Russo (1997) found in an extensive study of such errors, English-speaking students of Italian as SL assume instinctively that expressions such as “falling in love” or “born with a silver spoon in one’s mouth” are delivered in the target language in structurally-parallel ways. As a result, they typically produce erroneous parallel expressions such as cadere in amore and nato con un cucchiaino d’argento in bocca. In previous work, I have suggested that identifying how such concepts are encoded differentially will go a long way to addressing the persistent problem of conceptual blunders such as these and towards imparting true proficiency in classroom SL learning (Danesi 2003). This would involve, first and foremost, taking into pedagogical account the insights into language coming out of the ongoing work in cognitive linguistics and, more particularly, in conceptual metaphor theory. This paper has a twofold purpose: (i) to revisit the notion of conceptual error introduced in previous work (for example, Danesi 1986, 1995; Russo 1997; Kecskes 1998, 2000), so as to expand it and discuss what it reveals about SL learning generally in classroom contexts, reporting on various studies conducted on conceptual errors and on the associated notion of conceptual fluency; (ii) to discuss relevant implications for SL teaching and syllabus design.

2. Conceptual Fluency Theory The inability of students to speak and write in ways that go beyond the “literalness” of textbook materials and to use idiomatic and culturallysensitive phrasal structures of the SL has been the motivation behind several very interesting pedagogical proposals, starting in the 1970s, including the so-called Notional-Functional Syllabus (Van Ek 1975; Wilkins 1976) and Contrastive Rhetoric pedagogy (Kaplan 1978; Piper 1985; Leki 1991; Grabe and Kaplan 1996; Connor 1996). Since the introduction and rapid

Conceptual errors in second language-learning

233

growth of conceptual metaphor theory and cognitive linguistics generally within general linguistics, starting in the 1980s (Lakoff and Johnson 1980, 1999; Lakoff 1987; Johnson 1987), several interesting discussions and frameworks for researching access to SL concepts in SL learning started surfacing already in the 1980s and spreading through the 1990s and 2000s (Alexander 1983; Dirven 1985; Danesi 1986; Low 1988; MacLennan 1994; Kövecses and Szabo 1996; Cameron and Low 1999; Boers 2000, 2004; Littlemore 2001a, 2001b, 2001c, 2003a, 2003b, 2004; CharterisBlack 2002; Cameron 2003; Holme 2004; Achard and Niemeier 2004; Cameron and Stelma 2005; Richardt 2005; Reagan 2006; Littlemore and Low 2006). Of these, the one that I have previously labeled conceptual fluency theory (CFT), starting in the late 1980s, is the target of concern in this paper, since CFT has since come under critical scrutiny and constituted the source of various critiques and research projects (Valeva 1996; Russo 1997; Kecskes 2000; Kecskes and Papp 2000a, 2000b; Velasco Sacristán 2005). The objective here is to revisit its main theoretical principles and discuss its applications to error analysis. The basic premise in CFT is that knowledge of conceptual systems plays a much larger role in guiding the acquisition of true SL proficiency than has ever been thought by the mainstream models of SL learning and pedagogy. Conceptual fluency is defined as the ability to give appropriate structural form to the all kinds of meanings, literal and non-literal that constitute the semantic system of the SL. CFT was motivated originally by the common observations of SL teachers that, although classroom-based student discourse often manifests a high degree of form-based or formal fluency (FF)—that is, the ability to manipulate the formal grammatical units and formulaic communicative structures of the SL—it invariably seems to lack the expressive (conceptually-based) fluency that characterizes the discourse of native speakers. Students typically use SL structures as “carriers” of their own native language (NL) concepts. When the NL and SL conceptual systems coincide in some area of discourse, then the student discourse is serendipitously native-like; when they do not, it manifests an asymmetry between language form and conceptual content. What student discourse often lacks, in other words, is conceptual fluency. From the extensive research on concepts in cognitive linguistics over three decades, it has become obvious that figurative language, metonymy, and other modes of abstraction are all interactive in conceptual fluency. In fact, the massive amount of data collected in cognitive linguistics suggests rather strongly that many abstract concepts, if not most, as Lakoff believes, are encoded and knowable primarily as “metaphorized ideas,” “meto-

234

Marcel Danesi

nymized ideas,” and so on (detailed summaries and diverse perspectives of the relevant work in this domain can be found in Gibbs 1994; Goatley 1997; Dirven and Verspoor 1998; Fauconnier and Turner 2002; Kövecses 2002; Danesi 2004; Geeraerts 2006). If the findings are in fact correct, then the pedagogical implications they suggest can no longer be ignored. Indeed, it can be claimed that the incorporation of CFT into a viable approach will go a long way towards making true proficiency a realizable outcome of SL pedagogy. Needless to say, the distinction between literal and non-literal expressions of concepts is (and always has been) a conveniently reductive one. In actual fact, there are many degrees of literalness and non-literalness in conceptualization that are influenced by various kinds of psychological and social factors. However, for the purposes of the present discussion, it is a sufficient one. The question of how to teach literal concepts in SL classroom contexts has always had a fairly straightforward answer. It is assumed that they are best taught in line with their referential nature – namely, by “demonstrating” them through explanatory and ancillary devices that allow their meanings to be shown in some concrete way (through pictures, through simple dialogues, etc.). By and large, this type of pedagogy has always produced, and will continue to produce, good results in all kinds of classroom situations. It can be called concretization for the sake of argument and defined as the technique of demonstrating an SL referent in concrete terms. The dilemma in SL pedagogy, it would seem, often lies in how to teach non-literal concepts. How does one teach students to express the concept of “love” appropriately when discussing it in the SL, given its abstract nature and many non-literal uses? It is precisely in attempting to answer this kind of pedagogical question that CFT presents itself as a potentially useful framework, because it suggests that many abstract concepts can be “concretized” in exactly the same way as literal-concrete ones. Why is this so? The answer is suggested by the research findings within cognitive linguistics itself, which show, essentially, that common abstract concepts are delivered in terms of concrete ones through the medium of metaphor, metonymy, and other generalized processes. Conceptual fluency can be be defined more specifically as the ability to access the appropriate source domains in the programming of SL discourse that allow for a non-literal allocation of meaning to language forms. Conceptual errors emerge when students utilize an NL source domain that is different from the SL one to deliver an abstract concept. So, when Englishspeaking students of Italian produce a phrase such as cadere in amore for “falling in love” they are using their NL image schema of a TRAP and


235

linking it erroneously to the concept of amore, which is based on a different (even if affined) image schema. The claim here is that this type of “conceptual interference,” as it can be called, occurs more frequently and hampers the flow of meaning more seriously than has been thought in SL pedagogy in the past.

3. Conceptual fluency The above discussion lays the basis for sifting out conceptual errors from other kinds of typical interlanguage errors (phonological, morphological, etc.). The question that arises is whether or not a conceptual error is really no more than a fancy term for the misuse of an idiomatic expression, as some critics have suggested (for example Valeva 1996), and thus really a type of lexical error. The answer to this critique does not come from the domain of pedagogy but from the more general research domain in cognitive linguistics, which has amply documented that the use of idioms and metaphors in discourse is not a simple lexical option. Such use reveals the very nature of abstract conceptualization. The process of concretization – the linkage of concrete source domains with abstract target domains – occurs at all levels of language, not just the lexical one. As a concrete example, consider the selection of certain verbs in particular types of sentences in Italian. The verb fare (“to make”) is used in reference to weather – Fa caldo (“It is hot,” literally “It makes hot;” Fa freddo “It is cold,” literally “It makes cold”). The verb essere (“to be”) is used instead in reference to objects – Il caffè è caldo “The coffee is hot;” Il caffè è freddo “The coffee is cold”). And the verb avere (“to have”) is used in reference to people – Lui ha caldo “He is hot, literally “He has heat;” Lei ha freddo “She is cold” literally “She has cold”). The use of one verb or the other – fare, essere, or avere – is not a matter of simple idiomatic choice. There are several ways to concretize these concepts. One of these may be via the image schema of a CONTAINER, in which HEAT or COLD is located, and this might be the trigger that determines the selection of the verb. The concretization process might then be explained as follows. If the CONTAINER is the environment, the HEAT or COLD is “made” by Nature (Fa caldo; Fa freddo); if it is a human body, then the body is conceptualized as “having” the HEAT or COLD within itself (Lui ha caldo; Lei ha freddo); and if it is an object, then the object “is” the CONTAINER (Il caffè è caldo; Il caffè è freddo). Metaphorical competence (as it has been called in previous

236

Marcel Danesi

work) is, clearly, the ability to access appropriate image schemata and source domains in the concretization of abstract concepts. Another plausible conceptual explanation is suggested as well by Radden and Dirven (2007). In English the so-called dummy subject it is used in likely reference to the atmospheric setting (It is hot = Fa caldo). It functions, in semiotic terms, as an index pointing conceptually to atmospheric phenomena (cold, heat, darkness, etc.), as can be seen in an array of expressions – It is raining, It is freezing, etc. – which the index projects against a general atmospheric ground alluding to states or changes of state. In the case of Lui ha caldo (“He is hot”), the conceptual schema involves experiencing something. The same schema is expressed as a dative in some languages, such as German: Mir ist kalt ‘To me is cold.’ The difference between English be and Italian avere (“to have”), or the dative in German reveals the kind of reaction that “experiencers” have to states such as heat and cold. And in the case of Il caffè è caldo (“The coffee is cold”), one could say that, rather than the container having the heat property, it is the contained itself that possesses that quality. Whatever conceptual system we adopt to explain the above expressions, the point to be made here is that one can no longer explain differences of this kind as simple matters of idiomaticity, but rather of unconscious patterns of thought relating concepts and language forms. It is, in other words, a phenomenon that can be accounted for differently from the way it has been traditionally explained by SL pedagogy, namely as an idiomatic or communicative aspect of language learning. If the proposed explanation is valid then, it is part of a systematic type of “conceptual competence,” so to speak, that is, arguably, as teachable as purely form-based and communicative competencies. Access to this competence and control of its categories can be called conceptual fluency (Danesi 2003), because it involves various kinds of concretization strategies in translating concepts into forms. Conceptual fluency involves knowledge of how to access source domains such as the ones described above (Dirven 1993). Thus, to be “conceptually fluent” is to know, in large part, how a language encodes abstract conceptual schemas (CONTAINER, EXPERIENCER, etc.) on the basis of various concretization processes—that is, how to turn concepts into language forms. This kind of knowledge, like form-based and communicative (pragmatic) knowledge, is by and large unconscious in native speakers. Conceptual fluency is highly intertwined with linguistic (purely formbased) and communicative competence in SL learning, as a host of recent studies have suggested (see Littlemore and Low 2006 for a good survey).


237

Kecskes (1998, 2000), for instance, has also shown that conceptual fluency plays a considerable role in guiding learners to successfully formulating Situation-Bound Utterances (SBUs), which are highly conventionalized, prefabricated units of discourse whose occurrence is sensitive to standardized communicative situations – “Hi, how’s it going?” “Not bad, and you?” etc. Similarly, Danesi (2000) and Sebeok and Danesi (2001) found that the “meaning flow” which characterizes discourse situations is guided by conceptual reasoning. The following brief stretch of recorded conversation between two students on the University of Toronto campus in 1999 shows, for example, how the PEOPLE ARE ANIMALS conceptual metaphor shaped the discourse sequence of their conversation (Danesi 2000): Student 1: Student 2: Student 1: Student 2:

You know, that prof is a real snake. Yeah, I know, he’s a real slippery guy. He somehow always knows how to slide around a tough situation. Yeah, tell me about it! Keep away from his courses; he bites!

Such sequences have been found to characterize large stretches of common conversations (Danesi 1999, 2000; Armour 2001). It would seem that conceptual fluency guides the unconscious “navigation” through the source domains that make up discourse. Conceptual fluency can, thus, be defined more specifically as the ability to navigate source domains in the SL during discourse programming. SL learners will have difficulties in taking part in real discourse situations until they have acquired the ability to recognize metaphors, to access their source domains and to produce further metaphorization themselves. One technique for fleshing out the unconscious conceptual fluency of native-speakers is to ask them to draw up meaning associations of specific concepts. This was the object of a University of Lugano study in 1998, conducted on 149 students. The study consisted of a very simple procedure. The students were asked to write sentences for given color terms that would show how they can be used in non-literal ways. The sentences were then collected and categorized statistically according to frequency. The research team counted the number of times a color was used in sentences to demonstrate a specific emotion. For example, bianco (“white”) was used by 144 students in a sentence that related the concept to “light” such as Ho visto tutto al bianco della luce (“I saw everything in the white of light”). The result for bianco can be found in Table 1. Percentages stand for the number of students listing the given concept.

238

Marcel Danesi

Table 1. NL speakers’ non-literal meaning associations of bianco (in percentages) Non-Literal Meaning Uses

Frequency (rounded off)

luce “light”

97%

purezza “purity”

91%

pulito “clean”

87%

neve “snow”

85%

candore “candor”

85%

colore “color”

84%

sposa “bride”

80%

latte “milk”

79%

anima “soul”

77%

verginità “virginity”

76%

freddo “cold”

70%

Similar results were tabulated for the other colors given (10 in total) – bianco (“white”), nero (“black”), rosso (“red”), verde (“green”), azzurro (“blue”), marrone (“brown”), giallo (“yellow”), rosa (“pink”), viola (“purple”), arancione (“orange”). The same procedure was repeated several months later with 89 students at the University of Perugia students (Danesi 2000). From the data collected several interesting findings emerged. One was that the non-literal showed a remarkable consistency. The same task of using colors in non-literal ways was given to 84 English-speaking students of Italian at the University of Toronto and 56 at Middlebury College in Vermont, in 2005, in order to assess the degree to which they had access to potential non-literal uses for color terms. The meanings the 140 non-native subjects provided for bianco are charted below in Table 2. The overall results show two interesting differences between the SL and NL subjects: (i) the former came up with some of the same meanings, but the frequencies of these varied considerably; (ii) various meanings not present in Italian were introduced (such as those marked with an asterisk for bianco above). The first finding is probably due to the fact that there is an overlap in the meanings of English and Italian color terms due to the similar cultural histories of their respective speakers. The second finding


239

indicates that the conceptual systems possessed by NL speakers of Italian and SL students are different. Overall, the non-literal meanings given by the NL and SL groups for the ten color terms used in the study varied by 58.2%, that is, about 58% of the meanings given by the SL group differed from those given by the NL group. Table 2. SL speakers’ non-literal meaning associations of bianco (in percentages) Non-Literal Meaning Uses

Frequency (rounded off)

neve “snow”

88%

latte “milk”

81%

*salute “health”

80%

verginità “virginity”

75%

sposa “bride”

70%

*ghiaccio “ice”

60%

*amicizia “friendship”

58%

anima “soul”

52%

*eroismo “heroism”

46%

In a follow-up study (in January of 2006), a group of intermediate level SL learners of Italian at the University of Toronto (26 in total) were given specific training in using the non-literal meanings of color terms identified by the NL groups, as a part of pedagogical units dealing with abstract concepts. For example, the information provided for the target domain of EMOTIONS consisted of such listings as given in Table 3 below. Similar information was provided for such concepts as SOCIAL STATES ARE COLORS, POLITICAL IDEOLOGIES ARE COLORS, etc. The class was then taught with straightforward techniques. For example, students were given exercises that were designed to allow them to apply their conceptual knowledge in systematic (and even mechanical) ways. An example has been given below the table (answers are provided here in parentheses for the sake of convenience).

240

Marcel Danesi

Table 3. Color terms and their non-literal uses in NL and SL Color terms

NL equivalents Non-literal uses

NL equivalents

arancione

orange

–

–

azzurro

blue

il Principe azzurro

Prince Charming

bianco

white

notte bianca

sleepless night

giallo

yellow

giallo dalla rabbia

extremely angry

grigio

gray

vita grigia

dull life

marrone

brown

fare un marrone

make a big mistake

nero

black

umore nero

dark mood

rosa

pink

acqua di rosa

superficiality

rosso

red

diventare rosso/a

become embarrassed

verde

green

al verde

(financially) broke

Sample exercise

Fill in the blanks with appropriate color terms. Anche tu sei di umore _____ (nero)? (“Are you in a dour (black) mood too”?) Sono quasi sempre al _____ (verde) e faccio una vita _____ (grigia). (“I’m always broke and I live a dull life.”) Ieri la mia amica mi ha fatto passare una notte _____(bianca)! (“Yesterday my friend made me go through a sleepless night!”) Lui mi fa diventare ______ (giallo) dalla rabbia! (“He makes me very angry!”)

Other techniques used included pattern practice, translation, role-playing, and brief writing tasks. At the same time a control group – a comparable intermediate class of 24 SL learners taking the same course in a different classroom during the same academic term – were not given any specific training in the metaphorical uses of color terms. The two groups were then


241

given a series of 8 writing tasks at the end of term: “Write a brief paragraph describing anger in Italian using descriptive language, such as colors, as best you can;” “Write a brief paragraph describing envy in Italian using descriptive language, such as colors, as best you can;” etc. The student texts were then analyzed in terms of (i) how many metaphorical meanings were used (as instructed to do) and (ii) how many conceptual interferences surfaced (whenever the students opted for metaphorical language). The first finding was that of the 8 writing tasks assigned, the experimental group, i.e., the group of students that received training in using the nonliteral meanings of color terms, used metaphorical descriptions in all of them, whereas the control group used them sporadically. Ratings of the appropriateness of the metaphorical uses in the texts were given by native speakers of Italian (two professors of the language who were on the campus as visiting professors). The breakdown is as follows – each column gives the number of students who used one or more color terms with metaphorical meanings per task: Table 4. Writing task: Use of color terms with metaphorical meanings Writing task

Experimental group (N = 26)

Control group (N = 24)

Number 1

22

–

Number 2

19

2

Number 3

19

2

Number 4

25

8

Number 5

20

4

Number 6

12

–

Number 7

24

3

Number 8

23

6

20.5

3.125

Average

Of the 164 times that the experimental group used a metaphorical meaning for a color term, 135 were judged as appropriate by the two native speakers, with an 82.31% accuracy rating; of the 25 times that the control group used a color term, only 4 were judged as appropriate by the native speakers, with a 16.00% accuracy rating. When examined by the research team,

242

Marcel Danesi

it was found, moreover, that in all cases of inappropriateness, the source was NL conceptual interference. For example, one student of the control group wrote: Anch’io sono sempre verde di gelosia (Lit.: “I too am also green with envy always”), showing the use of verde with an English metaphorical meaning. In sum, the results showed two relevant things. First, the use of nonliteral concepts by the control group was negligible, whereas the conceptually-trained group used it often. Second, the experimental group showed a remarkable ability to apply the appropriate metaphorical meanings to the given writing tasks, whereas the control group did not. And when students from the latter group opted to use color terms, they invariably applied NL meanings to them. The study thus showed that conceptual fluency can be taught explicitly and that, when it is not, the tendency of students is either to avoid using metaphor altogether or to apply NL meanings in its use. Needless to say, such classroom-based research is only relatively reliable. It does not allow generalisations nor prove anything in any empirical way. It only suggests a pattern that may exist. As Boers and Lindstromberg point out (2006), such small-scale classroom experiments are valuable, but they must be supplemented by larger scale surveys. Also, it should be mentioned that such research really reveals that we live in diverse “cultural worlds,” so to speak. Without going into the whole issue of the relation between language and culture, suffice it to say here that in CFT the two would be considered reflexes of each other (see Danesi 2004 for the relevant discussions, theories, and findings on this hypothesis).

4. Conceptual errors As recent research has shown, the inability to convert concepts into language (formal) structures through concretization is the source of many of the most disruptive interlanguage errors made typically by students (Kövecses and Szabo 1996; Kecskes 2000; Phillip 2003). The learners in the study mentioned above were asked what errors they considered to be the most serious ones. Virtually all in the control group (23 of 24) identified their errors in vocabulary usage, especially with regards to color terms, as the most serious ones. They saw the other errors as easily “correctable” and thus not as serious. A conceptual error (CE) can be defined as a sort of “symptom” in the SL learner’s output, that is, as an erroneous attempt to frame an SL concept that results in an inappropriate SL form. A CE is to be distinguished from a form-based, or formal, error (FE) which can be de-


243

fined as a symptom of a conceptually-correct output, but which is formally (morphologically, syntactically, etc.) incorrect. The above set of findings raises several key issues with regard to classroom SL teaching. The main one is that differences between the NL and SL conceptual systems will have to be taken directly into account in the formulation of any effective pedagogical system for imparting true native-like proficiency. Conceptual systems are culturally transmitted – through language, through representations (movies, TV programs, novels, newspapers, etc.), and through discourse situations. Therefore, when students learn a new language, not only are two “language systems” in contact – but also two “conceptual systems.” The relevant systems can be symbolized as follows: L1 CS1 L2 CS2

= = = =

the learner’s native language the learner’s native conceptual system the target language the conceptual system of the target culture

The task of gaining proficiency in the SL can be formulated as the ability to express oneself in the L2 while utilizing the CS2, rather than using the L2, but relying primarily on the CS1. During the earliest phases of SL learning, students unconsciously rely on their CS1 to decipher novel input. They also express themselves largely through a process of “conceptual translation,” relying on the CS1 to direct the choice of L2 structures or, vice versa, using new L2 structures with CS2 meanings, without being consciously aware of this. Gradually, and with greater mastery of the L2, the students should gain higher and higher levels of conceptual fluency. But as previous research has shown, this will occur only if they have first of all absorbed the CS2 and next been taught how to access it directly. It is, of course, possible that even during the earliest stages learners will produce native-like utterances, even though they are relying on the CS1. The reason for this is positive conceptual transfer, whereby L2 expressions result as native-like because the CS1 and CS2 overlap in a specific source domain. Clearly, the degree of concepts shared by two language systems in contact will vary, depending on the particular languages and cultures in question. It is safe to assume, in effect, that there will be overlap, and that this is a source of positive conceptual transfer for learners. Transfer and interference are tied to what can be called “conceptual productivity.” Indirect evidence from the writing tasks described above suggests, in fact, that concepts with a high number of potential source do-

244

Marcel Danesi

mains tend to generate fewer conceptual errors than those with a lower number simply because there are bound to be more instances of overlap between the CS1 and the CS2 (Danesi and Di Pietro 1991). Is the concept of love more or less productive than, say, justice in a specific culture? If so, does it produce fewer conceptual errors? Is there a way of determining, or even “measuring,” the relative productivity of these two concepts? Interest in answering such questions has been minimal, until very recently. Only one movement, called the Contrastive Rhetoric Movement, which actually took root four decades ago with Kaplan (1966), can be seen to have always had some of the same pedagogical interests as CFT. The key finding of this movement has consistently been that a large portion of negative transfer occurs on a rhetorical conceptual level rather than on purely a lexical or syntactic level. Typically, unless they have learned the appropriate domains of the CS2, the students’ attempts at SL discourse will “violate the expectations of the native reader” (Kaplan 1966: 4). The claim of the movement is that conceptual fluency – as it has been called here – is as teachable as linguistic or communicative competence. This claim has, in fact, been borne out not only by the studies mentioned above, but also by similar studies (Lazar 1996; Deignan, Gabrys, and Solska 1997; Velasco Sacristán 2005). The findings of the movement suggest that the most common errors in writing occur when the individual concepts, or by generalization, the CS1 and the CS2 have what I have called here different productivity levels. For example, writing tasks involving notions of JUSTICE are more likely to demonstrate negative conceptual transfer than writing tasks involving notions of LOVE. With a small group of research assistants at the University of Toronto, a pilot project was initiated in early 2000 (Danesi 2003) and repeated in 2006 with another team (in order to assess its replicability) to determine if the above hypothesis was indeed a valid one. First, the two teams gathered data derived from a variety of English and Italian written and oral texts (newspapers, magazines, popular books, etc., recordings of radio and television programs). From the data, a list of conceptual metaphors related to two abstractions – IDEAS and LOVE – was compiled. This was achieved simply by collecting statements related to the two concepts. From the list, the relevant source domains could be easily distinguished and categorized. Thus, for example, it was found that in English IDEAS (and their incorporation in such activities and events as laws, performances, lectures, etc.) was rendered by such source domains as MOVING THINGS, LIGHT AND DARKNESS, BUILDINGS, PLANTS, COMMODITIES, VISION, GEOMETRY, FOOD, PEOPLE, FASHION, among others.


245

The research team found, altogether (on the basis of nearly 200 written and oral texts), a total number of 89 source domains used to deliver the concept of IDEAS in English. This can be called its Productivity Index (PI = 89). Given the number of texts used, by regular statistical inference, it can be considered to be a reliable metric for this specific concept: that is, there is no reason to believe that a compilation of similar data would yield significantly different results (within regular margins of statistical error). Some source domains were enlisted more frequently than others. This could, of course, be a matter of style (some types of texts tend to manifest the utilization of certain source domains more than others), or a true index of productivity within the target domain. The pilot projects did not control for this aspect of conceptual productivity (henceforward CP). It emerged as a consequence of the two studies and, thus, is an issue that will have to be investigated in future work on CP. For the present purposes, the PI can be defined, simply, as a relative quantitative indication of the productivity of a concept, given a specific sample of data. The PI of IDEAS was then compared to the PI of LOVE (using the same database). The latter manifested itself through the medium of such source domains as PHYSICS (“There were sparks between the two actors”), HEALTH DISEASE (“The roles focused on the sick relationship between them”), INSANITY (“He’s gone mad over her”), MAGIC (“She has bewitched her lover”), among others. The PI for LOVE was found to be 36. This suggests that LOVE is a less productive concept than IDEAS in English. To compare the two, the notion of Relative Productivity Index (RPI) was used. This is defined, simply, as the quotient that emerges when the lower PI is divided into the higher one. In the data collected, this turns out to be 2.47 (RPI = 2.47). This means, in effect, that there are 2.47 times more source domains used for delivering IDEAS than there are for delivering LOVE in English. What does this imply? It suggests, arguably, that in English culture, IDEAS is a concept that is given much more representational salience, so to speak, than is LOVE. This does not mean that the latter is not important, but, simply that there are fewer ways to conceptualize it in everyday cultural interaction. It is left for future work to investigate whether differential RPIs lead to differences in cultural forms of expression. The notion of CP can be used to compare conceptual systems cross-culturally in a specific way. Knowledge of some of the source domains above – VISION and HEALTH – is relatively independent of culture. However, there are some source domains that are – or appear to be – dependent upon cultural

246

Marcel Danesi

knowledge. A comparative analysis will not only identify these, but also allow us to compare the RPI of certain concepts cross-culturally. In the Italian data, it was found that most of the source domains used to deliver the concept of IDEAS were identical (even though their morphological and syntactic manifestations varied predictably). For example, GEOMETRICAL FIGURES, VISION, FOOD, etc. were frequent in the Italian data. More than completely identical source domains, however, the teams found that the nature of the source domain, in most cases, was slightly different. For example, rather than PLANTS (“That idea has many ramifications”), which also exists in Italian, the notion of FERTILITY emerged as more productive in delivering the target domain (“Quella politica è veramente arida” = “That policy is truly arid”), as did METEOROLOGICAL EVENTS (“Le sue sono idee tempestose” = “His ideas are tempestuous;” “Piovono nuove idee in quell’università” = “New ideas are raining on that university”) and WEIGHT (“Una delle sue idee è, comunque, un’idea assai leggera” = “One of his ideas is rather light”). Overall, the Italian PI came to 123. The RPI between Italian and English in this domain was, therefore, 1.38 (which means that the Italian data showed 1.38 more source domains than the English data in the delivery of IDEAS). The PI for LOVE in Italian came to 99. Within Italian, this showed that the RPI (= 1.24) between the two target concepts of IDEAS and LOVE is probably statistically irrelevant. However, it is significant when the PIs for LOVE in English and Italian are compared (RPI = 2.75). Thus, there were almost three times as many source domains used to verbalize and represent LOVE in the Italian data than there were in the English data. Related to the notion of PI is what can be called Source Domain Productivity (SDP). This provides, simply, a comparative measure of the vehicles utilized within source domains: for example, the SEEING source domain is highly productive in both English and Italian for the delivery of IDEAS; however, the MOVING THINGS one is rather limited in both languages, since very few vehicles in the domain are selected to deliver the concept. In effect, the SDP of the former is higher than that of the latter in both languages: in English it is 25 and 5, in Italian 29 and 12 for SEEING and MOVING THINGS respectively. Once again, this suggests that some source domains are more frequent in discourse, representation, etc. than are others. In effect, within each source domain, there are subdomains that provide the concept-user with an array of specific vehicles that can be utilized to provide subtle detail to some concept. The higher the SDP, the more likely the utilization of that source domain.


247

Two writing tasks were then given to a class of 35 intermediate students of Italian at the University of Toronto in 2005. The first task was to write in Italian about “The importance of reading philosophical works today” and the second one was to write about “How love is expressed at Saint Valentine’s Day.” The conceptual errors found in each of the writing tasks were compiled together and classified as: (i) E1, or errors generated by the CS1 in the domain of IDEAS and (ii) E2, or errors generated by the CS1 in the domain of LOVE. The total numbers of such errors detected were as follows: E1 = 24; E2 = 65. As the relevant RPIs had predicted, the number of conceptual errors was almost three times larger for the LOVE concept. As a follow-up study, in the summer of 2006, students of two intermediate Italian classes at Middlebury College in Vermont were asked to write on the same topics. One of the two classes – the “experimental class” – was trained to perceive the conceptual differences between the CS1 and the CS2 (English and Italian) whereas the control group was not. The nonliteral density (NLD) of the texts was determined by dividing the number of non-literal sentences by the total number of sentences. A sentence was tagged as non-literal if it was a token or instantiation of the underlying culturally appropriate concept. The repetition of a non-literal form was not counted more than once as it was considered to be an elaboration. The most relevant finding was that the NLD for experimental group with regards to LOVE was 2.75% more than the control group’s NLD, indicating a consistency with the PI of LOVE in Italian. Little difference was noted in the NLDs for IDEAS between the two groups. Leaving aside conceptual interferences in the data, this finding suggests above all else that the ability to use concepts productively is a definite factor in the development of conceptual fluency. The CEs were subsequently compared to FEs in the texts. Table 5 lists the actual numbers of error tokens produced by the two groups. Table 5. Writing task: Formal vs. conceptual errors Writing task

Experimental group (N = 18)

Control group (N = 16)

FEs

292

277

CEs

34

398

The result showed (i) that CEs far outnumbered other kinds of errors in the control group and (ii) that the two groups were comparable with respect to FEs. Thus, a second major finding of the study shows that conceptual er-

248

Marcel Danesi

rors far outnumber formal ones in writing tasks. Nevertheless, the findings need to be tested in future work and with students at different stages of learning. The studies seem to suggest, at this juncture, that simple exposure to and training in conceptual fluency leads to concrete results.

5. Expanding CFT A common finding in a follow-up analysis of the writing texts produced by the metaphorically-trained groups of students is that although they showed greater access to the CS2 source domains than did the other students, they were rarely able to access what can be called the connotative meanings or “modalities” associated with the concepts. Take, for example the metaphorical statement “The professor is a snake,” which is an instantiation of PEOPLE ARE ANIMALS. It is rare that students are able to manage, let alone, use the kinds of non-literal concepts typical of native language use. Learners are less likely to provide further descriptive modality to the above evaluation of the professor’s personality, if such a need should arise: “He’s a cobra;” “He’s a viper; “He’s a boa constrictor;” “He’s a rattlesnake;” etc. In effect, within each source domain, there are subdomains that provide the native speaker with an array of connotative modalities that can be utilized to project subtle detail on a target domain. Recent studies of recent work on concepts indicate a taxonomic structure comprising three levels of granularity (summarized in Dirven and Verspoor 1998); these allow us to take the above problem into account in SL learning. Some concepts have a highly general referential function. These are called superordinate. The formula PEOPLE ARE ANIMALS itself is a superordinate concept, because it refers to the general phenomenon of personality. Other concepts have a typological function. These are called basic. The choice of specific metaphorical vehicles from the ANIMALS source domain – snake, rat, etc. – produces, in effect, basic concepts because the vehicular choices allow for reference to types of personalities. Finally, some concepts have a detailing function. These are called subordinate. The selection of sub-types of snake – cobra, viper, etc. – are all subordinate concepts that might be needed for specialized purposes. The problem detected in the writing texts can now be reformulated as a problem in “access to the subordinate level” of source domains. This has obvious implications for CFT. No data have as yet been collected on this phenomenon. It is left as a target of future research.


249

6. Implications and concluding remarks As the research on CFT has suggested for over two decades, the time has come to seriously consider developing methodological frameworks, materials, etc. that take conceptual fluency and its subcategory of metaphorical fluency directly into account. The techniques used to impart conceptual fluency in the studies described briefly above were all based, in fact, on the utilization of rather traditional forms of teaching, including such things as: — preparing dialogical material that exemplifies conceptual systems at work; — exposing the students to cultural realia that illustrate how they are employed by native speakers; — preparing exercise and activity materials, whereby the students must identify the source domains and explain them; — requiring the students to write their own explanations of the domains. So, for instance, when teaching an English-speaking student about the weather in Italian (the topic), the instructors made sure: (i) to inform him or her about the conceptualization of HEAT and COLD as substances that are contained in specific contexts (the vehicles) or as reflexes of the EXPERIENCER mode of conceptualization; (ii) to teach him or her how to use the verbs fare, avere, and essere as reflexes of the vehicles (including relevant morphosyntactic information); (iii) to develop appropriate textual and practice materials based on this explanatory framework. Showing teachers how to utilize the CS2 directly in SL pedagogy was the objective of a graduate summer course I gave to student-teachers of Italian at Middlebury College in the summers of 2004 and 2005. The subjects were taught the basics of CFT and then asked to prepare appropriate teaching units. The first thing that the student-teachers did was to write dialogues in conceptually-appropriate ways. For instance, a dialogue on birthdays in Italian would have to include the AGE IS A DENUMERABLE QUANTITY concept, which produces linguistic metaphors such as Li porti bene gli anni (Lit. “You carry your years well”) and Gli anni incominciano a pesare sulle sue spalle (Lit. “The years are beginning to weigh on her shoulders”). Neither of these expressions was found in typical first-year or intermediate textbooks of the language. At the end of the course, the dialogues were evaluated by three native speakers of the language (professors from Italy visiting Middlebury College that summer) who found them to be

250

Marcel Danesi

“authentic” when compared to those found typically in textbooks. The explanatory, grammatical, and activity units that the student-teachers constructed, following up on the dialogue-writing exercises, also demonstrated the facility with which conceptually appropriate practical material could be devised by teachers. The subjects explained the verbs (portare), nouns (anni), and particles (ne) as structures connected to the basic conceptual modes. Then, expressions such as Ho due anni più di te (Lit.: “I have two years more than you”) were explained within this conceptual framework, thus allowing them to expand upon the purely structural components of the unit. Typical exercises that showed the relation between forms and concepts were also written by the student-teachers: Quanti anni hanno i tuoi amici? (“How old are your friends?”), Quanti ne hanno? (“How old are they?”), etc. Fill-ins, completions, multiple choices, etc. were also suggested, followed by typical role-playing and textual analysis activities. Without going into details, suffice it to say that the Middlebury study showed that the notion of CFT is as teachable and usable in the creation of units as is any other pedagogical notion. By simply structuring designated units of study around the CS2 and then by presenting the appropriate grammar and communication patterns, the result seems to be a pedagogical product that is as usable as is any other kind of pedagogical artifact. One student-teacher wrote the following e-mail to me at the end of the course: “I am now aware of the conceptual differences between the two languages and glad to be able to teach them.” There are many ways in which CFT can be incorporated into classroom practice. In my view, the most significant one is in the area of syllabus design. The main idea would be to identify and catalogue the source domains that deliver specific concepts in discourse, together with a correlative analysis of the grammatical/communicative structures that encode them. As Savignon (1992) has suggested, it is perhaps more appropriate, and certainly more useful, to think of different kinds of syllabi as cooperative and complementary contributors to SL acquisition in the classroom, not as antagonistic or mutually exclusive competitors. Most current syllabus designs are constructed to impart mainly what I have called formal fluency. I believe that a conceptual syllabus will come to have an increasingly larger role to play in the future, if the findings on conceptual errors discussed here are in any way indicative of how SL learning proceeds in classroom situations. To conclude, the notions of conceptual fluency and of conceptual error raise several critical questions that will have to be addressed in the future –


251

answers to which I believe will go a long way towards imparting true native-like proficiency in SL classrooms. Here are the two most pivotal ones: — To what extent and in what ways, if any, does conceptual fluency relate to, or is embedded in, the native speaker’s world knowledge? It must always be kept in mind that metaphorically-shaped knowledge, assuming that it is the most difficult one for SL learners to access in the CS2, is probably just one possible form in which knowledge of the world is encoded and decoded by humans. As Levin (1988: 10) has aptly remarked, there appear to be many modes of knowledge: “innate knowledge, personal knowledge, tacit knowledge, spiritual knowledge, declarative and procedural knowledge, knowing that and knowing how, certitude (as well as certainty), and many other varieties.” The more appropriate goal for the SL teacher should be, therefore, to determine to what extent language use is based on conceptual knowledge and to what extent it is based on other forms of knowledge. — If concepts are to be placed at the core of language courses and syllabi, on what basis should they be selected and sequenced? In my view, the conceptual syllabus should be integrated with grammatical and communicative syllabi, as mentioned. Units could be planned around concepts such as IDEAS, LOVE, AGE, BIRTHDAYS, etc. Not all aspects of language and SL learning are tied to conceptual fluency, as it has been defined here. Interlanguage studies have amply documented error phenomena that are purely grammatical, communicative, etc. We should, of course, continue to assess the role played by formal fluency in the overall process of classroom language learning. However, in my view the notion of conceptual fluency can no longer be ignored. The work of Lakoff and Johnson and others has shown that there is systematicity to non-literal concepts. The process of learning the CS2 is, arguably, identical to the one enlisted for learning the L2. This implies, as a corollary, that teachers will have to become more conversant with current theories of language, in the same way that they have had to consider various other theories of learning in the past. As illustrated in this paper, this should not constitute a problematic area given outcomes such as those produced by the Middlebury project described above. Equipped with more precise notions of what constitutes a concept, what productivity is, etc., teachers can anticipate learning difficulties and

252

Marcel Danesi

conceptual errors, and thus be able to take appropriate pedagogical steps to help the learners overcome them. It must be emphasized, however, that CFT is not the “magic key” that will guarantee successful SL learning in all situations. Like good piano teachers, experienced language teachers know all too well that each student is different, that each learning task presents its own kinds of problems and, thus, that patience is the operative word in all situations and contexts of learning. But frameworks such as CFT seem to have broad applicability, indicating that all kinds of learning tasks invariably involve culturallybased meaning processes.

References Achard, Michel, and Susanne Niemeier (eds.) 2004 Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching. Berlin: Mouton de Gruyter. Alexander, Richard G. 1983 Metaphors, connotations, allusions: Thoughts on the language culture connexion in learning English as a foreign language. LAUT Working Papers, Series B, 91. Trier: LAUT. Armour, William S. 2001 “This guy is Japanese stuck in a white man’s body:” A discussion of meaning making, identity slippage, and cross-cultural adaptation. Journal of Multilingual and Multicultural Development 22: 1–18. Boers, Frank 2000 Metaphor awareness and vocabulary retention. Applied Linguistics 21: 553–571. 2004 Expanding learners’ vocabulary through metaphor awareness: What expansion, what learners, what vocabulary? In Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching, Michel Achard and Susanne Niemeier (eds.), 211–232. Berlin: Mouton de Gruyter. Boers, Frank, and Seth Lindstromberg 2006 Cognitive linguistic applications in second or foreign language instruction: Rationale, proposals, and evaluation. In Cognitive Linguistics: Current Applications and Future Orientations, Gitte Kristiansen, Michel Achard, René Dirven, and Francisco J. Ruiz de Mendoza Ibáñez (eds.), 305–355. Berlin; New York: Mouton de Gruyter. Cameron, Lynn 2003 Metaphor in Educational Discourse. London: Continuum.


253

Cameron, Lynn, and Graham Low (eds.) 1999 Researching and Applying Metaphor. Cambridge: Cambridge University Press. Cameron, Lynn, and Juurd H. Stelma 2005 Metaphor clusters in discourse: Methodological issues. Journal of Applied Linguistics 1: 107–136. Charteris-Black, Jonathan 2002 Second language figurative proficiency: A comparative study of Malay and English. Applied Linguistics 23: 104–133. Connor, Ulla 1996 Contrastive Rhetoric: Cross-Cultural Aspects of Second-Language Writing. Cambridge: Cambridge University Press. Danesi, Marcel 1986 The role of metaphor in second language pedagogy. Rassegna Italiana di Linguistica Applicata 18: 1–10. 1993 Metaphorical competence in second language acquisition and second language teaching: The neglected dimension. In Language, Communication and Social Meaning, James E. Alatis, (ed.), 489–500. Washington, D.C.: Georgetown University Press. 1995 Learning and teaching languages: The role of conceptual fluency. International Journal of Applied Linguistics 5: 3–20. 1999 The interconnectedness principle and the semiotic analysis of discourse. Applied Semiotics 6/7: 394–401. 2000 Semiotics in Language Education. Berlin: Mouton de Gruyter. 2003 Second Language Teaching: A View from the Right Side of the Brain. Dordrecht: Kluwer Academic Publishers. 2004 Poetic Logic: The Role of Metaphor in Thought, Language, and Culture. Madison: Atwood Publishing. Danesi, Marcel, and Robert J. Di Pietro 1991 Contrastive Analysis for the Contemporary Second Language Classroom. Toronto: OISE Press. Deignan, Alice, Danuta Gabrys, and Agnieszka Solska 1997 Teaching English metaphors using cross-linguistic awareness-raising activities. ELT Journal 51: 352–360. Dirven, René 1985 Metaphor as a basic means for extending the lexicon. In The Ubiquity of Metaphor, Wolf Paprotté and René Dirven (eds.), 85–119. Amsterdam: John Benjamins. 1993 Metonymy and metaphor: Different mental strategies of conceptualization. Leuvense Bijdragen 82: 1–25. Dirven, René, and Marjolijn Verspoor 1998 Cognitive Exploration of Language and Linguistics. Amsterdam: John Benjamins.

254

Marcel Danesi

Fauconnier, Gilles, and Mark Turner 2002 The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities. New York: Basic. Geeraerts, Dirk (ed.) 2006 Cognitive Linguistics: Basic Readings. Berlin: Mouton de Gruyter. Gibbs, Raymond W. 1994 The Poetics of the Mind: Figurative Thought, Language, and Understanding. Cambridge: Cambridge University Press. Goatley, Andrew 1997 The Language of Metaphors. London: Routledge. Grabe, Wallace, and Robert Kaplan 1996 Theory and Practice of Writing. New York: Longman. Holme, Randal 2004 Mind, Metaphor and Language Teaching. Basingstoke: Palgrave. Johnson, Mark 1987 The Body in the Mind: The Bodily Basis of Meaning, Imagination and Reason. Chicago: University of Chicago Press. Kaplan, Robert D. 1966 Cultural thought patterns in inter-cultural education. Language Learning 16: 1–20. 1978 Contrastive rhetoric: Some hypotheses. ITL: Review of Applied Linguistics 39/40: 61–72. Kecskes, Istvan 1998 The state of L1 knowledge in foreign language learners. Word 49: 321–341. 2000 Conceptual fluency and the use of situation-bound utterances in L2. Links & Letters 7: 143–158. Kecskes, Istvan and Tunde Papp 2000a Metaphorical competence in trilingual language production. In Acquisition of English as a Third Language, Jasone Cenoz and Ulrike Jessner (eds.), 99–120. Clevedon: Multilingual Matters. 2000b Foreign Language and Mother Tongue. Hillsdale, New Jersey: Lawrence Erlbaum. Kövecses, Zoltan 2002 Metaphor: A Practical Introduction. Oxford: Oxford University Press. Kövecses, Zoltan, and Paul Szabo 1996 Idioms: A view from cognitive semantics. Applied Linguistics 17: 326–355. Lakoff, George 1987 Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press.


255

Lakoff, George, and Mark Johnson 1980 Metaphors We Live By. Chicago: Chicago University Press. 1999 Philosophy in the Flesh. New York: Basic. Lazar, Gillian 1996 Using figurative language to expand students’ vocabulary. ELT Journal 50: 43–51. Leki, Ilona 1991 Twenty-five years of contrastive rhetoric: Text analysis and writing pedagogies. TESOL Quarterly 25: 123–143. Levin, Samuel R. 1988 Metaphoric worlds. New Haven: Yale University Press. Littlemore, Jeannette 2001a Metaphor as a source of misunderstanding for overseas students in academic lectures. Teaching in Higher Education 6: 333–351. 2001b Metaphoric competence: A possible language learning strength of students with a holistic cognitive style? TESOL Quarterly 35: 459– 491. 2001c An empirical study of the relationship between the holistic/analytic cognitive style dimension and second language learners’ communication strategy preferences. Applied Linguistics 22: 241–265. 2003a The effect of cultural background on metaphor interpretation. Metaphor and Symbol 18: 273–288. 2003b The communicative effectiveness of different types of communication strategy. System 31: 331–347. 2004 Interpreting metaphors in the EFL classroom. Les Cahiers de l’APLIUT 23: 57–70. Littlemore, Jeannette, and Graham Low 2006 Metaphoric competence, second language learning, and communicative language ability. Applied Linguistics 27: 268–294. Low, Graham D. 1988 On teaching metaphor. Applied Linguistics 9: 125–47. MacLennan, Carol H.G. 1994 Metaphors and prototypes in the learning and teaching of grammar and vocabulary. International Review of Applied Linguistics 32: 97– 110. Phillip, Gillian S. 2003 Connotation and collocation: A corpus-based investigation of colour words in English and Italian. PhD dissertation, University of Birmingham. Piper, David 1985 Contrastive rhetoric and reading in second language: Theoretical perspectives on classroom practice. Canadian Modern Language Review 42: 34–43.

256

Marcel Danesi

Radden, Günter, and René Dirven 2007 Cognitive English Grammar. Amsterdam/Philadelphia: John Benjamins. Reagan, Timothy 2006 Learning theories and metaphorical discourse: Reflections on second language learning and constructivist epistemology. Semiotica 161: 291–308. Richardt, Susanne 2005 Metaphor in Languages for Special Purposes. New York: Peter Lang. Russo, Gerard A. A. 1997 A conceptual fluency framework for the teaching of Italian as a second language. PhD dissertation. Toronto: University of Toronto. Savignon, Sandra J. 1992 Problem solving and the negotiation of meaning. In Problem Solving in Second Language Teaching, Caterina Cicogna, Marcel Danesi, and Anthony Mollica (eds.), 11–25. Welland: Soleil. Sebeok, Thomas A., and Danesi, Marcel 2001 The Forms of Meaning: Modeling Systems Theory and Semiotic Analysis. Berlin: Mouton de Gruyter. Valeva, Gloria 1996 On the notion of conceptual fluency in a second language. In Papers in Second Language Acquisition and Bilingualism, Aneta Pavlenko and Rafael Salaberry (eds.), 22–38. Ithaca: Cornell University Press. Van Ek, Jan A. 1975 The Threshold Level in a European Unit/Credit System for Modern Language Teaching by Adults. Strasbourg: Council of Europe. Velasco Sacristán, Maria S. 2005 Metaphor and ESP: Metaphor as a useful device for teaching L2 business English learners. Ibérica 10: 115–131. Wilkins, David A. 1976 Notional Syllabuses. Oxford: Oxford University Press.

Part III Conceptual learning: Construal of motion, temporal structure, and dynamic action

Motion events in Danish and Spanish: A focus-on-form pedagogical approach Teresa Cadierno

Abstract This article discusses the possible contribution of Talmy’s (1985, 1991, 2000) typological framework of motion events to focus on form, a pedagogical approach to grammar teaching that involves drawing L2 learners’ attention to linguistic elements in the context of communication (Long 1991). The article, which focuses on the expression of motion events in Danish and Spanish, shows how Talmy’s typological approach and the insights from empirical research into the acquisition of motion expressions by L1 and L2 speakers can constitute the basis for a focus-onform pedagogical intervention that focuses on both comprehension and production processes on the part of the L2 learner. Keywords: contrastive study; language typology; motion events; focus on form; pedagogical approach; comprehension; production; Danish; Spanish

1. Introduction The main aim of this article is to discuss the contribution of cognitive linguistics to focus on form, a pedagogical approach to grammar instruction which has received a great deal of attention over the last decade from both second language acquisition (SLA) researchers and foreign language teachers.1 This approach, originally proposed by Long (1991), involves drawing L2 learners’ attention to linguistic elements in the context of communication, i.e., in lessons whose overriding focus is on meaning and communication. It is argued that the pedagogical option of focus on form, with its key emphasis on the integration of grammar and communication in foreign language teaching, can greatly benefit from the theoretical perspective of cognitive linguistics, given its view of language as consisting of conventionalized form-meaning mappings used for communicative purposes (Langacker 1987). The article focuses on the conceptual domain of motion. It will be shown how Talmy’s (1985, 1991, 2000) typological framework for the

260

Teresa Cadierno

expression of motion events as well as the insights from empirical research into the expression of these events by L1 and L2 speakers (e.g., Berman and Slobin 1994; Slobin 1996b, 1997, 1998, 2000, 2003, 2004; Cadierno 2004; Cadierno and Ruiz 2006; Kellerman and Van Hoof 2003; Navarro and Nicoladis 2005; Negueruela et al. 2004) can constitute the basis for a focus-on-form pedagogy that focuses on both comprehension and production processes on the part of the L2 learner. The article is organized as follows. Section 2 provides a diachronic overview of the different pedagogical options in relation to grammar teaching, with a particular emphasis on the focus-on-form approach. Section 3 presents Talmy’s (1985, 1991, 2000) typological framework for the expression of motion events, and briefly summarizes empirical work on first language acquisition inspired by this framework. Section 4 addresses the implications of this framework for second language acquisition, and reviews existing empirical work on the acquisition of motion constructions by adult L2 learners of Spanish. Section 5 presents examples of how motion constructions can be the object of a focus-on-form pedagogical intervention. Finally, Section 6 summarizes the main points discussed in the article.

2. Grammar teaching in foreign language classrooms The role of grammar instruction in foreign language curricula has been an object of debate for the past 30 years. According to Long (1991) and Long and Robinson (1998), three main pedagogical options can be identified with respect to grammar teaching, namely focus on formS, focus on meaning and focus on form. The focus-on-formS approach characterizes synthetic syllabi – the most common one being the structural syllabus – where linguistic elements, e.g., phonemes, morphemes, words, sentence patterns – are selected and presented in isolation in a step by step fashion according to given criteria (e.g., frequency and difficulty). Language acquisition in this approach is seen as “... a process of gradual accumulation of parts until the whole structure of language has been built up” (Wilkins 1976: 2). Traditional teaching methods such as Grammar Translation, and the Audiolingual Method, with their associated classroom practices of explicit presentation of grammar rules, repetition of models, and explicit error correction are typical examples of this approach where communicative L2 use has little, if any, role to play. The language acquisition view behind this pedagogical option has not been supported by existing empirical research on the

Motion events in Danish and Spanish

261

acquisition of grammar by adult L2 learners (Long and Robinson 1998). Rather than mastering linguistic structures in one single step, learners have been shown to acquire specific linguistic structures in a gradual fashion, typically passing through various stages of non-target like uses of target L2 forms (e.g., Andersen 1984; Meisel 1987). With the advent of communicative language teaching at the end of the 1970s, the focus on meaning approach started to dominate the language teaching profession. In this pedagogical option, the role of grammar instruction in second language acquisition was seriously downplayed, and sometimes even suggested to be detrimental for the language learner. A frequent claim made at the time was that “... learning an L2 incidentally (i.e., without intention, while doing something else) or implicitly (i.e., without awareness) from exposure to comprehensible target language samples (was) sufficient for successful second or foreign language acquisition ... just as it appears to be for first language acquisition ...” (Long and Robinson 1998: 18). Foreign language classroom practices such as immersion programs and content-based instruction reflect this pedagogical option in which learners experience the language as a medium of communication and not as an object of study in itself. Within SLA circles, this noninterventionist position was epitomized by Krashen’s influential Monitor Model (1981) which claimed that knowledge of consciously learned language was distinct from unconsciously acquired language knowledge. As learned knowledge could not be turned into acquired knowledge, grammar instruction was assigned a very limited role in language teaching, whose main purpose was to afford opportunities for real communication in the L2. As empirical research into the role of grammar instruction in SLA flourished in the last couple of decades, a reconsideration of the role of grammar teaching in foreign language classrooms began to emerge. Research on learning outcomes in the Canadian French immersion programs showed that learners exposed to long-term communicative language teaching with no explicit grammar teaching failed to achieve high levels of accuracy with certain grammatical forms (e.g., Harley and Swain 1994; Swain 1985). Additionally, a large number of classroom-based studies provided evidence for the positive effects of grammar instruction, both in terms of accuracy and rate of acquisition. For recent reviews, see Ellis (2002) and Norris and Ortega (2000). In addition to this empirical evidence, the 1980s theoretical position that language acquisition was largely an unconscious process was challenged by Schmidt’s (1990, 1993) influential claim that conscious attention to form, or noticing, was a necessary condition for language learning.

262

Teresa Cadierno

The empirical and theoretical work just mentioned led to the suggestion that communicative foreign language teaching that was meaning-focused could benefit from some degree of form-focused instruction. In agreement with this line of thought Long (1991: 56) proposed a third option in language teaching, namely, focus on form, which consisted of “... overtly draw(ing) students’ attention to linguistic elements as they arise incidentally in lessons whose overriding focus is on meaning, or communication.” Later Long and Robinson (1998: 23) offered a more operational definition of this approach: “... during an otherwise meaning-focused classroom lesson, focus on form often consists of an occasional shift of attention to linguistic code features – by the teacher and/or one or more students – triggered by perceived problems with comprehension or production.” Focus on form thus refers to how focal attentional resources are allocated during language teaching. Critically, however, the allocation of attention to form is not only limited to L2 teaching contexts (Long and Robinson 1998). The same kind of process would take place when a native speaker, in the middle of a conversation, pauses to reflect on a particular aspect of the language s/he is using (e.g., whether to use this or that morphological verb form; whether to refer to one’s interlocutor with the Spanish forms tú or usted (you informal or you formal). The native speaker’s fundamental orientation is to meaning and communication, but factors may arise that lead him/her to temporarily pay attention to the language per se. Since Long’s original definition of focus on form, there has been considerable variation in how this pedagogical option has been understood and operationalized in empirical research (Doughty and Williams 1998). One parameter that has often been discussed in the literature is whether to take a reactive approach to focus on form as Long (1991) originally suggested, i.e., to provide feedback to learners’ problem areas as they naturally occur in communicative activities, or to take a proactive approach, i.e., to select in advance aspects of the target language that have been observed to be problematic for the learners. The two approaches involve different emphasis in curricular planning: whereas the reactive approach emphasizes the teacher’s ability to notice pervasive errors on the part of the learners, and to have at the ready techniques for drawing their attention to them, the proactive approach emphasizes the a priori design of tasks that can ensure the use of previously observed problematic areas by the learners when they are engaged in communication (Doughty and Williams 1998). As there is yet no definite research to support the effectiveness of one approach over the other, both are suggested to be effective, depending on the classroom context.


263

Another important parameter is the degree of explicitness of focus on form, that is, the extent to which an implicit or more explicit approach to grammar instruction is adopted. An implicit focus on form has the aim of attracting learners’ attention to form as unobtrusively as possible, and thus assumes an implicit learning mode, i.e., “a nonconscious and automatic abstraction of the structural nature of the material arrived at from experience of instances” (N. Ellis 1994: 1). An explicit focus on form, on the other hand, has the aim of directing learners’ attention to the problem area involved, and assumes that providing learners with explicit knowledge about a specific linguistic feature will facilitate its eventual acquisition by aiding (i) the process of noticing the form in the input and (ii) the process of noticing-the-gap, i.e., helping the learners realize the difference between their own production and that of native speakers (Ellis 1995, 2003). The degree of explicitness of focus on form is best conceived of as a continuum from more implicit to more explicit techniques and types of intervention. At the implicit end we find techniques such as input flood, which entails the provision of numerous instances of the same linguistic form in the input. At the explicit end we find the use of tasks where learners have to reflect upon, discuss or process specific linguistic items in specific ways. Examples of such tasks would be the metatalk that takes place during dictogloss tasks, consciousness-raising tasks where a given grammar problem is solved interactively in groups as the task outcome, and processing instruction, an instructional technique that will be discussed in more detail in Section 5, and whose aim is to help learners establish appropriate form-meaning connections when processing L2 input. Finally, along the continuum we have techniques such as input enhancement, whose aim is to enhance the perceptual saliency of linguistic structures in the input and thus increase their chance of being noticed. Enhancement techniques can be applied to both written input (e.g., highlighting, color-coding or any other kind of font manipulation of specific linguistic forms) and oral input (e.g., intonational focus on learner errors). Classroom-based studies that have examined the relative effectiveness of more or less implicit-explicit instructional techniques generally show that combinations of focus-onform techniques are likely to be more effective than the use of individual techniques. Combinations that have resulted in positive outcomes are those of input flood plus input enhancement (e.g., White 1998), and intonational focus plus corrective recasting (e.g., Doughty and Varela 1998). In sum, as indicated by Nassaji and Fotos (2004), research into the effects of grammar instruction in SLA clearly shows that some grammar teaching is necessary if language learners are to achieve high levels of pro-

264

Teresa Cadierno

ficiency in the foreign language. Grammar instruction, however, is best implemented as a focus-on-form approach where linguistic expressions are presented in communicative contexts designed to promote learners’ awareness of their form-meaning relationships. The focus-on-form approach to grammar teaching overcomes the limitations of the two previous approaches (i.e., focus on formS and focus on meaning). On the one hand, it recognizes that even though implicit learning can certainly take place in foreign language classrooms, adult learners can benefit from having their attention explicitly drawn to linguistic aspects of the L2. On the other hand, it goes beyond the traditional and formal linguistics-based conception of language assumed in the focus-on-formS approach, in which grammar was primarily seen as a collection of meaningless forms governed by a set of rules. The type of grammar teaching embodied in the focus-on-form approach, with its key focus on form and meaning relations, and the use of language in communicative contexts, is thus more in consonance with the language view advocated by cognitive linguistics, and is expected to benefit from its insights. Terminologically, however, the terms “focus on meaning” and “focus on form” are unfortunate in that they might suggest a categorical dissociation between semantics and (grammatical) form, a dichotomy well established in formal approaches to language. However, even though language teaching theorizing is still very much influenced by such a language view, the concepts of “focus on meaning” and “focus on form” are meant to refer to whether or not the learner’s attention is more or less explicitly drawn to aspects of the L2 during language teaching activities. Focus on form does not then imply a focus on “form” at the cost of “meaning,” but instead emphasizes the need to direct learners’ attention to certain aspects of the language when they are engaged in meaningful and communicative L2 language use (e.g., as in, for example, task-based language teaching), and thus focuses on the establishment of appropriate form-meaning mappings on the part of the learner. Even though there are other branches of linguistics that embrace the inseparability of meaning and form, and specifically focus on their relations such as Halliday’s (1985) Systemic Functional Linguistics or other functional theories of language, it is argued that the view of meaning adopted by cognitive linguistics and the central role that meaning has in linguistic description makes it especially suited for a pedagogical option that explicitly focuses on form-meaning relations and the communicative functions of language. In the next section we develop this point in greater detail.


265

3. Motion events: A typological approach on first language acquisition Cognitive linguistics is a school of linguistics that comprises different approaches and lines of research, but they all share a common concern with the symbolic and communicative functions of language (Evans and Green 2006). Specifically, language is viewed as a symbolic system, i.e., “an open-ended set of linguistic signs or expressions, each of which associates a semantic representation of some kind with a phonological representation” (Langacker 1987: 11).2 Linguistic expressions, which include lexical, morphological and syntactic structures, are considered to be symbolic units consisting of conventionalized form-meaning mappings used for communicative purposes. This view of language places semantics at the center of linguistic description, i.e., linguistic meaning is considered to be an essential part of grammar. In cognitive linguistics, meaning is considered to be embodied, i.e., “the structures used to put together our conceptual systems grew out of bodily experience and make sense in terms of it” (Lakoff 1987: xiv–xv). Furthermore, meaning is equated with conceptualization, which, in turn, must be interpreted broadly, and thus includes all facets of sensorimotor and emotive experience, and the apprehension of the social, linguistic and cultural context (Langacker 1996). Linguistic meaning is then viewed as encyclopedic in nature, that is, as integrating both semantic and pragmatic aspects of language. Thus, the meaning representation we may have of a given entity is comprised by everything we know about that entity, and thus draws upon complex bodies of knowledge. These complex bodies of knowledge are referred to as “cognitive domains” in Langacker’s (1987) terms, “frames” in Fillmore’s (1975), and “idealized cognitive models” in Lakoff’s (1987). The cognitive linguistics view of meaning contrasts with that of formal theories of language where semantic structure is seen as a truth-conditional relationship between an utterance and objective reality. As argued in Cadierno (to appear) this objectivist view of meaning has been the predominant one in both second language acquisition research and foreign language teaching where the emphasis has tended to be on the referential aspects of meaning. It is thus argued that this “new” conception of meaning, with its emphasis on the conceptual, pragmatics and cultural aspects of language, can make an important contribution to both the field of second language acquisition research and the pedagogical option of focus on form. See Niemeier (2004) for a discussion of the contribution of cognitive linguistics to the teaching of cultural aspects in language pedagogy.

266

Teresa Cadierno

Within cognitive linguistics, Talmy’s typological framework on motion events (Talmy 1985, 1991, 2000) is a promising line of inquiry for the investigation of second language acquisition and teaching, as it addresses the systematic relationship in language between form and meaning in different languages. Talmy’s work shows that even though our understanding of motion seems to be based on a universal SOURCE-PATH-GOAL image schema (Johnson 1987; Lakoff 1987), languages differ as to how the conceptual components of a motion event are packaged into linguistic forms. These six components are: (i) motion: the presence per se of translational motion, i.e., motion where “an object’s basic location shifts from one point to another in space” (Talmy 2000: 35); (ii) figure: the moving object; (iii) ground: the object with respect to which the figure moves; (iv) path: the path followed by the figure with respect to the ground; (v) manner: the manner in which the motion takes place; (vi) cause: the cause of its occurrence. On the basis of the typical form-meaning patterns found in the depiction of motion events in different languages, two broad types of languages have been identified.3 In satellite-framed languages (S-languages), such as Chinese and all branches of the Indo-European family except Romance Languages, the verb typically conflates motion and manner or cause, and path is encoded separately by a satellite. A satellite has been defined by Talmy (2000: 102) as “the grammatical category of any constituent other than a noun-phrase or prepositional-phrase that is in a sister relation to the verb root.” The following example in English and its Danish counterpart illustrate this typological pattern where manner of motion is expressed by the verbs ran and løb, and path of motion is expressed by the satellites out and ud: (1)

a. b.

The man ran out of the house. Manden løb ud af huset. Lit.: man-Det. ran out of house-Det.

In verb-framed languages (V-languages), such as Romance, Semitic and Polynesian languages, on the other hand, the verb typically conflates motion and path, while manner and cause are expressed separately in an adverbial or gerund. The following Spanish example illustrates this second typological pattern. In the example, manner of motion is expressed by the gerund corriendo, whereas path of motion is coded by the verb salió: (2)

El hombre salió de la casa corriendo. ‘The man went out of the house running.’


267

These lexicalization patterns, however, reflect general tendencies and not absolute differences across languages. As indicated by Talmy (2000), occasional forms of alternative conflation patterns can occur in a given language, such as the presence in English of Latinate verbs (e.g., enter, descend, ascend) which conflate motion and path. Additionally, V-languages such as Spanish evidence a split system of conflation where two different lexicalization patterns are observed for two different types of motion events, namely, one for bounded motion events, that is, motion events that involve the crossing of a spatial boundary (e.g., motion out of/over/into a bounded region), and one for unbounded motion events, i.e., motion events that do not involve such a spatial boundary-crossing (Aske 1989; Slobin and Hoiting 1994). In non-boundary-crossing events (e.g., running towards a house, without entering), Spanish speakers can express manner of motion in the verb. In boundary-crossing events, on the other hand, manner of motion must be expressed in a separate constituent. The two following examples illustrate this contrast: (3)

a. b.

El hombre corrió hasta la casa. ‘The man ran towards the house.’ El hombre entró en la casa corriendo. ‘The man entered the house running.’

The reason for this difference in the two types of languages is that in Vlanguages “… crossing a spatial boundary is conceived of as a change of state, and that state changes require an independent predicate …” (Slobin 1997: 441). As V-languages mark changes of state by means of motion verbs that incorporate a path element (e.g., enter, exit), the option of expressing manner of motion must be carried out by means of separate constituents. In sum, S- and V-languages critically differ with respect to how manner information is expressed in boundary and non-boundary-crossing events. In S-languages, manner of motion can be conflated with motion in the main verb in both boundary and non-boundary situations – as in (4a) and (4b) below. (4)

a. b.

The man ran into the house. The man ran towards the house.

In V-languages, on the other hand, such conflation is only possible in nonboundary situations – see (5a), (5b) and (5c) below.

268 (5)

Teresa Cadierno

a. b. c.

El hombre corrió hasta la casa. ‘The man ran towards the house.’ El hombre corrió en la casa. ‘The man ran inside the house.’ El hombre entró en la casa corriendo. ‘The man entered the house running.’

In addition to the inter-typological differences mentioned above, intratypological variation has also been reported within a given typological pattern. Both lexico-morpho-syntactic and cultural facts have been suggested to have an effect on the differences found between languages belonging to the same type of language (Slobin 2004). With respect to the former, it has been argued by Cadierno (2004) and Cadierno and Lund (2004) that Danish can be considered to be a more prototypical example of an S-language than, for example, English, as it lacks Latinate verbs which conflate motion and path of motion (e.g., enter, ascend, descend). In addition, Danish possesses an elaborate system of satellites, i.e., verb particles, which systematically indicate an opposition between translocative and nontranslocative motion, e.g., ud-ude ‘out (translocative)-outside,’ ind-inde ‘in(to) (translocative)-inside,’ op-oppe ‘up (translocative)-upstairs,’ nednede ‘down (translocative)-downstairs,’ and hjem-hjemme ‘home (translocative)-at home.’ Following Zlatev, David and Blomberg (in press: 7), translocative motion is defined as “the continuous change of an object’s average position according to a spatial frame of reference.” Compare: (6)

a. b.

The child runs out of the house. The child crawls up the stairs.

(7)

a. b.

The child runs outside in the garden. The child crawls up on top of the stairs.

Sentences (6a) and (6b) are both examples of translocative motion as they specify the change of the figure’s location with respect to a goal – in the case of the first sentence, and with respect to a system of geo-centric coordinates – in the case of the second (Zlatev, David and Blomberg in press). The two sentences differ, though, in that the first one pictures a bounded motion event whereas the second one pictures a noun-bounded one. In contrast, sentences (7a) and (7b) are examples of non-translocative motion. The following examples in Danish illustrate the contrastive use of the Danish satellite system to express translocative and non-translocative motion:


(8)

a. b.

(9)

a. b.

269

Barnet løber ud af huset. ‘The child runs out of the house.’ Barnet kravler op af trappen. ‘The child crawls up the stairs.’ Barnet løber ude i haven. ‘The child runs outside in the garden.’ Barnet kravler oppe på trappen. ‘The child crawls up on the stairs.’

Examples (8a) and (8b) illustrate translocative motion events, whereas examples (9a) and (9b) illustrate non-translocative motion events. In addition, (8a) pictures a bounded motion event, whereas the other examples picture unbounded motion events. The importance of lexico-morpho-syntactic factors has also been pointed out in the literature when discussing intra-typological variation within S-languages with respect to manner of motion. For example, the differences reported in Slobin (2004) with respect to the amount of manner information included in narratives produced by speakers of different Slanguages (e.g., Slavic languages such as Russian vs. Germanic languages such as Dutch, German and English) have led this author to suggest placing languages on a cline of manner salience rather than on a strict bipartite typology. Similarly, researchers such as Özçaliúkan and Slobin (1998), Ibarretxe-AntuĖano (2004) and Engberg-Pedersen and Trondhjem (2004) have observed that the presence of more complex morphological resources for spatial expression in three V-languages (i.e., Turkish, Basque, and West-Greenlandic) seems to lead to more elaborated path descriptions than those observed in more prototypical V-languages such as Spanish. With respect to cultural factors, researchers such as Wilkins (2004) and Bavin (2004) have argued that the relatively greater attention to path details observed in V-languages such as Arrernte and Walpiri can be possibly related to the culture-specific concern for motion and orientation – journeys – among nomads in Central Australia. Talmy´s typological framework has been empirically investigated within child language acquisition by Slobin and his colleagues (e.g., Berman and Slobin 1994; Slobin 1996b, 1997, 2000).4 A key piece of work is Berman and Slobin (1994) which constitutes an extensive cross-linguistic and developmental investigation of native speakers – both children and adults – from both S-languages (English and German) and V-languages

270

Teresa Cadierno

(Hebrew, Spanish, and Turkish). The results of these investigations which have been discussed and elaborated on in Slobin (1996b, 1997, 1998, 2000, 2003, 2004) showed clear differences in the rhetorical style followed by the native speakers of the two types of languages. Speakers of S-languages exhibited a higher degree of elaboration in their description of path of motion than speakers of V-languages, i.e., they tended to provide richer and more detailed descriptions of paths (trajectories). This was evidenced by a more frequent use of ground-adjuncts, i.e., prepositional phrases referring to the source, medium or goal of movement (e.g., the English description fell into the pond vs. the Spanish description se cayó ‘fell down’), as well as by a more frequent use of so-called event conflation, i.e., the incorporation of the different composites of trajectories (path and ground, including source, medium and goal) within a single clause (e.g., The deer threw the boy over a cliff into the pond). Speakers of S-languages also exhibited a higher elaboration of manner of motion than speakers of V-languages. This was evidenced by a more frequent use of manner of motion verbs (token analysis) and a higher variety of these types of verbs (type analysis). As suggested by Slobin (1997, 2000), languages seem to have a “two-tiered” lexicon of manner of motion verbs: neutral, everyday verbs such as walk, run and fly, and more expressive or exceptional verbs such as dash, swoop, and scramble. Despite the intra-typological differences mentioned above with respect to the attention given to manner of motion by speakers of different S-languages, S-languages generally have a more extensive and elaborated second-tier than V-languages. Manner of motion is therefore more salient in S-languages than in V-languages, and consequently, their speakers are used to make finer manner distinctions than speakers of Vlanguages. In Slobin’s own words (2004: 251), speakers of high-mannersalient languages [S-languages] “… regularly and easily provide information about manner when describing motion events, whereas (speakers of low-manner-salient languages provide) manner information … only when manner is foregrounded for some reason” (Slobin 2004: 251). Speakers of S-languages, finally, exhibited a relative higher attention to the dynamics of movement along paths, in comparison to speakers of V-languages who exhibited a relative higher attention to scene setting and static descriptions. Thus, speakers of S-languages tended to specify the details of trajectories and leave settings to be inferred – see example (10a) below. In contrast, speakers of V-languages tended to describe aspects of the static scene in which the movement took place, leaving trajectories to be inferred – as in example (10b).


(10)

a. b.

271

The deer threw them off over a cliff into the water. Lo tiró. Por suerte, abajo estaba el río. El niño cayó en el agua. ‘(The deer) threw him. Luckily, below, was the river. The boy fell into the water.’ (Slobin 1996b: 204).

The systematic differences found in children learning typologically different languages have been explained by Slobin (1996a) as reflecting different patterns of thinking for speaking. Thinking for speaking involves picking those characteristics of objects and events that fit some conceptualization of the event, and are readily encodable in the language (Slobin 1996a). When acquiring a native language, the child learns particular ways of thinking for speaking, i.e., s/he “learns to attend to particular aspects of experience and to relate them verbally in ways that are characteristic of that language” (Berman and Slobin 1994: 611). Each language thus trains its native speakers to pay different kinds of attention to particular details when talking about them. As Slobin (1996a) himself has suggested, this training takes place in our early childhood and might be resistant to restructuring in adult second language acquisition.

4. Motion events in second language acquisition The results on the acquisition of the expression of motion events by native speakers of S- and V-languages pose the following question for SLA: How do adult L2 learners come to express motion events in an L2 that is typologically different from their L1? Following Slobin’s thinking for speaking hypothesis, learning an L2 would entail learning another way of thinking for speaking, that is, learning which particular details of a motion event must be attended to in the input and expressed in the foreign language (i.e., attention to location-static descriptions vs. movement-trajectories, and relatively more or less attention to manner of motion), and learning the conventional ways in which the semantic components of a motion event are mapped onto L2 surface forms in connected discourse (Cadierno 2004; Cadierno and Lund 2004). The task of the L2 learner is thus not just to learn individual motion verbs in isolation, but to learn how to structure the whole semantic domain of motion in the L2, i.e., learn the characteristic mappings of the conceptual components of a motion event onto the surface features of the L2. A general hypothesis consistent with the thinking for speaking hypothesis would be that the learners’ L1 typological patterns will, at least ini-

272

Teresa Cadierno

tially, constitute the point of departure for the form-meaning mappings established in the L2. This hypothesis is in agreement with claims made in the literature (e.g., VanPatten et al. 2004) that learners, especially in the early and intermediate stages of language acquisition, tend to make partial and non-target-like form-meaning mappings that are often influenced by their L1. In line with this general hypothesis, Cadierno and Lund (2004) posited a set of more specific hypotheses concerning the initial mappings involved in the interpretation and production of manner of motion by learners with typologically different L1s and L2s, namely, Danish – an S-language – and Spanish – a V-language. For example, given that Danish has a more finegrained lexicon of manner of motion verbs than Spanish and, consequently, Danish speakers have been trained to categorically distinguish between more different manners of motion than Spanish speakers, it was predicted that Spanish learners of Danish would initially tend (i) not to process the fine-grained manner distinctions present in especially L2 second-tier manner verbs and (ii) not to use these manner verbs in production, overgeneralizing, instead, a single or few first-tier manner verbs to all communicative contexts (e.g., gå ‘to walk’ in all walking contexts independent of the manner of walking involved, which in Danish could be expressed by a larger variety of manner of verbs than in Spanish, such as slentre ‘stroll,’ stavre ‘trot,’ traske ‘trudge,’ or vakle ‘stagger’). In contrast, it was predicted that Danish learners of Spanish would initially tend to add manner information when interpreting and producing Spanish motion constructions. For example, in order to express the idea expressed in (11a) below, they might say something like the utterance in (11b): (11)

a. b.

Folk strømmede gennem gaderne. ‘People streamed through the streets.’ La gente pasaba por las calles como un río. ‘People passed through the streets like a river.’5

Likewise, as Danish allows for the mapping of manner and motion onto the main verb in both boundary and non-boundary-crossing situations, whereas Spanish only allows this mapping in the latter case, it was predicted that Spanish learners of Danish would initially tend to interpret Danish boundary-crossing expressions as non-boundary-crossing. Thus, the sentence expressed in (12a) would be wrongly interpreted as in (12b):


(12)

a. b.

273

Han løber ind i huset. ‘He runs into the house.’ *El corre dentro de la casa. ‘He runs inside the house.’

In addition, Spanish learners of Danish would also be expected to avoid the conflation of manner and motion in the verb in boundary-crossing situations. Conversely, it was expected that Danish learners of Spanish would initially tend to interpret Spanish non-boundary-crossing expressions such as the example in (13) to refer to both/either a non-boundary (e.g., He ran inside the house) and a boundary-crossing (e.g., He ran into the house) situation. (13)

Corrió en la casa. ‘(S/He) ran inside the house.’

Finally, Danish learners of Spanish were expected to produce non-targetlike L2 Spanish expressions conflating manner and motion in boundarycrossing situations, as (14a) to mean (14b): (14)

a. b.

*Ella corrió en la casa. ‘She ran inside the house.’ She ran into the house.

So far only a few studies have attempted to empirically investigate how adult learners acquire motion constructions in an L2 that is typologically different from their L1. Most of the studies (e.g., Cadierno 2004; Navarro and Nicoladis 2005; Cadierno and Ruiz 2006) have focused on the acquisition of a V-language – Spanish – by intermediate and advanced adult learners with L1 S-languages – Danish and English, and have compared the performance of these L2 learners to control groups of native speakers, i.e., native speakers of Danish and English.6 Whereas the first two studies examined the expression of path and ground of motion, the latter study also investigated the semantic component of manner of motion. With respect to data collection procedures, Cadierno (2004) and Cadierno and Ruiz (2006) elicited written narrative data by means of the “frog stories,” whereas Navarro and Nicoladis (2005) elicited narrative oral data by means of two video films. A full description of the results of these studies cannot be provided here due to space limitations – the interested reader can consult Cadierno (to

274

Teresa Cadierno

appear) – but the available research so far reveals a mixed picture with respect to the role of the learners’ L1 in the expression of motion events in the L2. Some traces of the L1 patterns were, for example, found in Cadierno (2004) and Cadierno and Ruiz (2006) where intermediate Danish learners of Spanish tended to exhibit a relatively higher degree of complexity and elaboration of the semantic component of path of motion than the one exhibited by the Spanish speaker control group. This was evidenced by a “satellization” of the Spanish motion constructions, e.g., the use of inaccurate constructions incorporating anomalous path particles not found in the Spanish native speakers data as the examples in (15) shows: (15)

a. b.

El niño fue arriba de una roca. Lit.: The boy went on top of a rock. ‘The boy climbed onto a rock.’ El ciervo mueve al niño y a su perro abajo en un precipicio. Lit.: The deer moves the boy and his dog down in an abyss. ‘The deer threw the boy and his dog down into the abyss.’

In these examples the Danish learners used verbs that indicate the fact-ofmovement with no specification of direction (e.g., ir ‘go,’ mover ‘move’) followed by a satellite encoding vertical path (e.g., arriba ‘up,’ abajo ‘down’). As argued in Cadierno (2004), these usages seem to reflect instances of what has been called communication transfer (R. Ellis 1994), which refers to the role of the learners’ L1 as a strategy to compensate for deficiencies in their interlanguage. In the case of the learners at hand, it seems that when confronted with communicative situations in which they lack the knowledge of L2 appropriate verbs of motion, they tend to rely on an L1-based strategy (i.e., use a motion verb plus a directional adverb) to fill in the gaps in their knowledge of the L2. In other cases, the inaccurate expressions produced by these Danish learners of Spanish involved boundary-crossing motion events, as the examples in (16) show: (16)

a. b.

*El perro saltó fuera de la ventana. ‘The dog jumped out of the window.’ *El perro se cayó fuera de la ventana. ‘The dog fell out of the window.’

Even though V-languages seem to allow certain manner verbs in boundarycrossing situations, i.e., manner verbs denoting high-energy motor patterns that are more like punctual or instantaneous acts than activities (e.g., jump,


275

throw oneself, plunge) (Slobin 2004), the use of the directional adverb fuera ‘out’ in this type of construction by the learners contrasts with the construction favored by the Spanish native speakers, who, instead, produced constructions such as El perro saltó/se cayó de/por la ventana ‘The dog jumped/fell from/through the window’ where the particle expressing path of motion fuera ‘out’ was omitted. In this type of constructions the Danish learners thus seem to add some extra semantic information, namely information on the path of motion by means of a satellite. What is interesting is that this semantic information must be obligatory expressed in the learners’ L1 – Danish – by a satellite, as the following examples in (17) illustrate: (17)

a. b.

Så hoppede hunden ud af vinduet. ‘So jumped the dog out of the window.’ Hunden faldt ud af vinduet. ‘The dog fell out of the window.’

As argued in Cadierno (2004), the use of this type of construction by the Danish learners seems to reflect a learning type transfer (R. Ellis 1994), which refers to the influence of the learners’ L1 in interlanguage hypothesis construction. Specifically, these learners seem to be inadvertently transferring the lexicalization pattern of the L1 into the L2, thus explicitly encoding semantic information that is obligatory encoded in their L1 but omitted in the L2. In contrast to the findings mentioned above, the Danish learners of Spanish investigated in Cadierno (2004) and Cadierno and Ruiz (2006) did not transfer the characteristic typological pattern of the L1 into the L2 in some other aspects analyzed. For example, these learners did not use event conflation in Spanish, a construction not easily allowed in this language but common in Danish. In Cadierno (2004) it was argued that this could be due to the learners’ psychotypology, i.e., the perceived distance between the characteristic constructions of the two languages (Kellerman 1979, 1986). Additionally, Cadierno and Ruiz (2006) found no significant differences between the Danish learners of Spanish, the Italian learners of Spanish and the Spanish native speakers with respect to the expression of manner of motion. In other words, Danish learners of Spanish, who were expected to highly attend to this semantic component of a motion event in the L2, given its saliency in their L1, did not differ with respect to the other two subject groups in their amount of use of manner of motion verbs or

276

Teresa Cadierno

other alternative means of expressing manner (e.g., subordinated manner clauses and adverbial expressions). In sum, the research reported above indicates that learners at the lateintermediate and advanced stages of language acquisition seem to have been able to acquire the characteristic L2 form-meaning mappings, and the general rhetorical style of the L2 when talking about motion. Existing research into the simultaneous use of speech and gesture (Kellerman and Van Hoof 2003; Negueruela et al. 2004) shows, however, that L2 learners’ gestures may reveal L1-based thinking patterns that are not detectable in otherwise fluent and target-like L2 speech. One possible explanation for the results mentioned above is that, in agreement with Cadierno and Lund’s (2004) hypotheses, the influence of the L1 patterns on the expression of motion events may be stronger at the initial stages of language acquisition. The idea behind this explanation is that with increased exposure to motion constructions in the L2 input, adult L2 learners gradually learn to pay attention to the L2 relevant aspects of experience, and to establish the appropriate form-meaning mappings. It addition, it could be hypothesized that learners with an L1 V-language and an L2 S-language (e.g., Spanish learners of English/Danish) would experience greater acquisitional difficulties than learners with an L1 S-language and an L2 V-language (e.g., English/Danish learners of Spanish). This hypothesis is based on the fact that even though for both types of learners the form-meaning mappings involved in the expression of motion events require complex associations not restricted to a one-to-one form-meaning correspondence, their degree of saliency is not uniform. More semantic information is presented in a backgrounded fashion in S-languages as compared to V-languages, i.e., manner of motion is generally less salient in Slanguages than in V-languages given its characteristic lexicalization in the verb root (Talmy 2000). Learners of S-languages are, therefore, expected to have greater difficulties in the acquisition of motion constructions than learners of V-languages, especially in the domain of manner of motion. In other words, it is hypothesized that it will be harder to learn to discriminate among different manners of motion for which learners have not been trained by their native language than to learn to make fewer discriminations than they are used to. To exemplify, given that Spanish learners of Danish are not used to discriminate among different fine-grained manners of motion, it is hypothesized that they will not attend to these fine-grained manner distinctions in the L2 input and will thus tend to use a less varied lexicon of manner of motion verbs than Danish native speakers. This hypothesis has not been subjected to an empirical investigation yet, but in-


277

formal observations of Spanish learners of Danish suggest that after having learned Danish and lived in Denmark for more than ten years, they have not managed to discriminate among the many fine-grained manner of motion distinctions encoded in Danish motion verbs.7

5. Teaching motion constructions in foreign language classrooms The motion domain has been a rather neglected domain in foreign language teaching. An analysis of some common textbooks used in Denmark for the teaching of Spanish as a foreign language at high-school level (e.g., En vivo by Laursen, Tegnberg-Hansen and Wiese 1993; A qué sí by Liébana and Tegnberg 2005) as well as for the teaching of Danish as a second language (e.g., Dansk for hele verden by Holm and Pinholt 1990; Mellem linjerne by Jeppesen and Maribo 2001) evidence the lack of any explicit reference to motion constructions. This absence is especially surprising in the case of the Spanish manuals, given their contrastive focus, and thus their tendency to focus on linguistic aspects where the two languages – Danish and Spanish – differ with respect to each other (e.g., Spanish aspectual system, use of ser vs. estar, use of indicative vs. subjunctive mood). The insights obtained from Talmy’s typological approach as well as from Slobin and his colleagues’ empirical work on L1 acquisition tell us, however, that there are good reasons why motion constructions should be taught in foreign language classrooms. First, Talmy’s typological approach shows that the conceptual domain of motion exhibits systematic crosslinguistic variation, i.e., languages vary with respect to how the conceptual components of a motion event are lexicalized in different languages. Second, Slobin’s work reveals the pervasive effects of the typology on native speakers’ acquisition and use of motion constructions in connected discourse. Thirdly, even though empirical research into the acquisition of these constructions by beginning L2 learners is still lacking, learners at the early stages of language acquisition have been hypothesized to establish inappropriate form-meaning mappings both with respect to the interpretation and the production of motion constructions. Finally, even though highintermediate and advanced learners with an L1 S-language and an L2 Vlanguage (e.g., Danish/English learners of Spanish) have been shown to be able to learn, for the most part, the new thinking-for-speaking of the foreign language, a greater level of acquisitional difficulty has been hypothesized for learners with an L1 V-language and an L2 S-language (e.g., Spanish learners of Danish/English), a hypothesis that is consistent with

278

Teresa Cadierno

informal observations made by the present researcher as well as by Caballero (2005). In this section we present a proposal of pedagogical intervention on motion constructions that is consistent with the focus-on-form approach to language teaching. Specifically, we focus on the teaching of the semantic component of manner of motion for adult learners with an L1 V-language – Spanish – and an L2 S-language – Danish.8 As this has been identified as a potential problematic area for this type of learners, the pedagogical intervention is based on a proactive approach to language teaching (see Section 2). In consonance with the hypotheses stated in Cadierno and Lund (2004), both comprehension and production activities are described.

5.1. Instruction on comprehension One focus-on-form approach to grammar teaching that particularly addresses the involvement of L2 learners’ comprehension processes is the socalled processing instruction, which aims at “alter(ing) the processing strategies that learners take to the task of comprehension and to encourage them to make better form-meaning connections than they would if left to their own devices” (VanPatten 1996: 60). In other words, the goal is to help learners process form-meaning relations that they have not previously processed – because they may not have been salient to them, for example, or that they have processed incorrectly – due to, for example, the influence of their L1. As an example, given that research on L2 sentence processing suggests that beginner and intermediate L2 learners tend to assign the role of agent to the first noun phrase in NP-V-NP sequences (e.g., VanPatten 1984; LoCoco 1987), processing instruction can be targeted to help learners process O-V-S sentences in the appropriate way. As evidenced by empirical research (e.g., VanPatten and Cadierno 1993a, 1993b; Cadierno 1995; VanPatten and Sanz 1995), processing instruction is a valuable pedagogical tool to help learners make appropriate connections between L2 language forms and their meanings.9 Processing instruction has three basic components: (i) explanation of the relation between a given form/construction and the meaning it conveys; (ii) information about processing strategies that may lead the learners to process the input incorrectly; for example, if research (e.g., VanPatten 1984) has shown that English learners of Spanish in the early and intermediate stages of language acquisition tend to assign the semantic role of agent and the grammatical role of subject to the first noun phrase of se-


279

quences such as Lo llama por teléfono la chica (Lit. Him-Direct Object calls by phone the girl-Subject), learners’ attention can be drawn towards this inaccurate processing strategy and they can be informed about the appropriate ways to process the construction; (iii) structured input activities specially designed to focus learners’ attention on a given linguistic structure in the input, and enable them to identify and comprehend its meaning. These activities should include both oral and written input, and move from sentence to connected discourse. No production of targeted linguistic structures is present in processing instruction as its focus is on the establishment of form-meaning connections during comprehension. An instructional sequence based on processing instruction for learners with an L1 V-language – Spanish – and an L2 S-language – Danish – could proceed as follows. With respect to the explanation of the relation between a given form/construction and the meaning it conveys, learners would be presented with a characterization of the typical form-meaning patterns found in the depiction of motion events in their L1 and L2. This description, which would be based on Talmy’s typological approach, would especially emphasize the different lexicalization patterns with respect to the semantic components of path and manner of motion. That is, it would focus on the fact that, in Spanish, path of motion is characteristically mapped onto the verb whereas manner of motion is typically expressed by a separate constituent. In Danish, on the other hand, it is manner of motion that is characteristically mapped on the verb, whereas path is coded by an elaborate set of satellites. This explanation would naturally be illustrated with different examples, such as the Spanish and Danish examples presented in (18a) and (18b): (18)

a. b.

La botella entró a la cueva flotando. Flasken flød ind i grotten. ‘The bottle entered the cave floating.’

After this general introduction to the characteristic conflation patterns of the two types of languages (i.e., their L1 Spanish, a V-language, and their L2 Danish, an S-language), the learners would be made aware of the consequences of the typology for the expression of manner of motion: (i) the fact that this semantic component is more elaborate in their L2 than in their L1 both in terms of the variety of manner of motion verbs existent in the lexicon and of the frequency with which they are used and (ii) the fact that manner of motion can be mapped onto the main verb in both boundary and non-boundary situations in their L2, but only in non-boundary situations in

280

Teresa Cadierno

their L1. It is suggested that the presentation of different (second-tier) manner of motion verbs is done according to different subdomains (e.g., manners of walking, jumping, gliding), through the aid of Total Physical Response (TPR),10 another comprehension-based approach to language teaching based on the coordination of speech and action (Asher 1977). The language instructor could show the students the meanings of the different manner of motion verbs by “acting out” the relevant manners of motion. Alternatively, or in addition to, this “acting out” could be presented by the instructor by means of video clips or short films. Next, learners would be provided with information about the processing strategies that they might be inclined to follow given the characteristic conflation patterns of the two languages: (i) given that they have not been “trained” to attend to different manners of motion in a categorical way, the semantically fine-grained manner distinctions present in especially L2 second-tier manner verbs might be initially ignored, the verbs being decoded instead only for motion and (ii) given the different conflation patterns in boundary vs. non-boundary-crossing situations, they might tend to interpret L2 boundary-crossing situations as non-boundary crossing. Learners should be made aware that this L1-based interpretation would furthermore be reinforced by the fact that, in Danish, path of motion is expressed by means of an elaborate satellite system where subtle phonetic differences (e.g., +/– glottal stop, +/– stress and vowel length) differentiate between a translocative function – used in boundary crossing contexts – and a locative function – used in non-boundary-crossing contexts, e.g., ud/ude ‘out (translocative)/outside,’ ind/inde ‘into (translocative)/inside,’ op/oppe ‘up (translocative)/upwards’). Some examples of this contrast are presented in (19) and (20) below: (19)

a. b.

(20)

a. b.

Han løb ind i huset. ‘He ran into the house.’ Han løb inde in huset. ‘He ran inside the house.’ Hun kravlede ud i haven. ‘She crawled out into the garden.’ Hun kravlede ude i haven. ‘She crawled outside in the garden.’

Needless to say, all the information provided to the students about processing strategies should be accompanied by relevant examples.


281

Finally learners would engage in structure input activities, both at the sentence and discourse levels, where their attention would be focused on motion constructions in the input and more specifically for our purposes, on the semantic component of manner of motion. In order to maximize intake efficiency, input activities should first focus on motion events involving non-boundary-crossing situations. Later activities should focus on the contrast between motion events involving both boundary and nonboundary-crossing situations. For each type of activity, different subdomains of manners of motion (e.g., walking, jumping) could be treated separately. With respect to activities focusing on non-boundary-crossing situations, learners could be asked to match oral/written stimuli with visual ones, i.e., listen to/read isolated sentences incorporating fine manner verb distinctions (e.g., Peter vakler hen ad gaden ‘Peter is staggering along the street’) and select the picture/video clip (out of two) that matches it (e.g., in one picture/video clip, a man is staggering along the street; in the other, he is strolling along the street). Additionally, and following a typical procedure of TPR, learners could act out the manner of motion encoded by commands provided by the language instructor. Moving into connected discourse, learners could be shown a series of pictures/video clips comprising a short story. The visual stimuli could, for example, show several events in one person’s day (many of these need obviously involve motion events where manner of motion is depicted). After seeing the pictures/video clips a couple of times, students are given a sheet that contains a list of sentences that refer to the person’s motion events. These sentences are presented in a scrambled order, and the students are asked to put them in the correct order. Afterwards, they are allowed to see the pictures/video clips again, and check whether their responses have been correct. In a third type of activity, learners could be asked to read short texts (these could be of different genres) where a great deal of physical motion is described. Learners would be first asked to read the texts for general meaning, then, in pairs, find given manner of motion verbs, and indicate what their meanings are. In a following text, learners themselves could find all the instances of manner of motion verbs and provide their meanings. Finally, learners could be presented with some general statements about the manner in which other persons perform certain actions, and be asked to tick whether or not this also applies to them (e.g., when I am in a hurry in the morning, I have to run to the bus; when I get too much to drink, I tend to stagger). With respect to activities focusing on the contrast between boundary and non-boundary-crossing situations, these would be of comparable na-

282

Teresa Cadierno

ture as to the ones described above. For example, learners could be asked to listen/read sentences describing motion events that describe both boundary and non-boundary-crossing situations (e.g., Hun løber ind i huset ‘She runs into the house’) and be asked to select pictures/video clips that match the sentences (e.g., choose between a picture that shows a girl running into a house, and another one where the same girl is running around inside the house). A more discourse-based type of activity could be one in which learners are asked to read a given text and try to understand its general meaning, and then, in pairs, identify motion constructions and classify them into those referring to boundary and non-boundary situations. Finally, and similarly to the activity proposed above, learners could be required to agree or disagree with respect to several statements about certain typical actions. These statements would naturally contain manner motion constructions referring to both boundary and non-boundary-crossing contexts (e.g., when children get back from school, they usually rush into the house; when babies start crawling, they love to crawl up the stairs). In addition to the activities mentioned above, and on the basis of Slobin’s (1996b, 1997) work on translations, language instructors could also engage learners in comparing original and translated novel passages rich in descriptions of motion events. The original and the translation should naturally represent the learners’ L1 and L2. Learners could be directed to analyze in groups how given motion events are presented in the original language and how they have been translated to their L2. This type of activity, like the ones described above, focuses learners’ attention on the targeted L2 construction, but is perhaps more in line with the consciousness-raising tasks proposed by Ellis (2003), as it involves a higher degree of reflection and meta-discussion about the targeted construction. The comprehension-based approach to grammar teaching incorporated in processing instruction should be followed by instruction on production, which is described in the following section.

5.2. Instruction on production With respect to production activities focusing on fine manner distinctions, some activities could involve information transfer, i.e., activities where the input to the task takes the form of some non-verbal device such as pictures, maps or diagrams which then must be communicated verbally to the hearer (Widdowson 1978). Some of these activities would constitute one-way tasks, i.e., tasks where information is passed from one person to another,


283

whereas others would be two-way tasks, i.e., tasks where the information is equally distributed. Examples of activities, all to be carried out in pairs, would be: (i) one-way picture drawing tasks, where the two learners would take turns in describing a series of pictures depicting different manners of motion to one another. One learner would describe the pictures and the other would draw the actions being described; before completing the task, the learners could be provided with a possible list of manner of motion verbs to be used; (ii) one-way map direction tasks, where, again the two learners would take turns in giving directions from one point to another on a map to one another. One student would give directions, and the other would draw the route being described;11 (iii) two-way spot-the-difference tasks, where two learners are given two similar pictures and are asked to identify the differences between the two. Some of the differences between the two pictures should involve different manner of motion (e.g., two people walking, jumping, or creeping in different ways); and (iv) two-way picture sequencing tasks, where two learners are required to reconstruct a story involving a great deal of motion on the basis of, for example, eight pictures. Each learner is given four pictures which are hidden from the other learner; each student describes a picture in turn and together they try to find the right sequence of the story. Additionally, learners could perform a decision-making task in pairs or small groups, where they would listen to several weather forecasts for the following days, and decide, on the basis of that, the kinds of free time activities they could perform each day. Learners could, for example, base their decisions on pictures provided to them where different motion activities are depicted. The decisions reached by the different students could then be compared in a plenum. With respect to activities focusing on the contrast between boundary and non-boundary-crossing situations, these could be similar to the ones described above. For example, in a one-way picture drawing task, one learner could be asked to describe a series of pictures depicting boundary and non-boundary-crossing situations to another learner, who, in turn, is required to draw the corresponding pictures. Similarly, in a two-way spotthe-difference task, the two learners could be given pictures that differ with respect to whether or not they depict boundary-crossing situations. In a related type of activity, learners could be asked to construct itineraries in a specific country from travel brochures. These should naturally include landmarks involving both boundary-crossing and non-boundary-crossing situations (e.g., entering key buildings – cathedrals, museums; walking on the beach, climbing some pyramid).

284

Teresa Cadierno

6. Concluding remarks The aim of this article was to show how the insights from cognitive linguistics, and more specifically from Talmy’s typological approach on the expression of motion events, can not only be fruitful for the investigation of how motion constructions are acquired by adult L2 learners, but also constitute the basis for pedagogical intervention on a construction, which, otherwise, tends to be neglected in foreign language teaching. It has further been argued that Talmy’s typological analysis of the form-meaning mappings in the motion domain fits nicely with the main assumptions of focus on form, one of the pedagogical options in language teaching frequently used nowadays. Moreover, we would argue that an approach to language teaching whose main aim is to integrate grammar and communication and focus on form-meaning relations should turn its eyes towards a linguistic enterprise which crucially views language as “… a system for the expression of meaning and for carrying out its symbolic and interactive functions” (Evans and Green 2006: 11–12). The in-depth analyses provided by cognitive linguists can provide applied linguists with valuable descriptions of the types of form-meaning relations encoded in the learners’ mother tongue and the foreign language they are attempting to learn. As it has occurred in the conceptual domain of motion, these descriptions could then constitute the basis for theoretically motivated studies of second language acquisition, and, as argued in this paper, for the development of contrastive pedagogical interventions. The proposal outlined in this article constitutes a first attempt at developing a pedagogical tool for the teaching of motion constructions that is based on the insights from cognitive linguistics, and is consistent with a proactive focus-on-form approach to grammar teaching. Needless to say, this proposal would still need further elaboration before it is implemented in the L2 classroom. For example, if one were to follow a task-based language teaching curriculum, the activities proposed above could be organized around given tasks with common themes (e.g., free time activities; traveling in a given country). Furthermore, the applied linguist/teacher should clearly specify: (i) the overall goal of the task(s) (i.e., to provide an opportunity for the interpretation and use of motion constructions); (ii) the input that constitutes its point of departure (e.g., pictures, maps, brochures, videos); (iii) the types of activities to be carried out, as well as their sequencing;12 (iv) the teacher’s and the students’ roles in the task; and (v) the settings for the task, i.e., types of classroom arrangements as well as a


285

specification of whether the task is to be implemented only inside the classroom or also outside of it (Nunan 1989; Ellis 2003). Lastly, but not the least, the pedagogical proposal outlined above could easily be subjected to empirical investigation. For example, classroombased experimental or quasi-experimental studies could be designed to investigate whether the cognitive linguistics-inspired grammar teaching proposed in this article has the expected positive effects on the acquisition of motion constructions by L2 learners.

Notes 1. I’d like to thank Søren W. Eskildsen for his useful comments on this paper. 2. Another common assumption is that language is viewed as intrinsically linked to human cognition and general cognitive processes, and not, therefore, as an autonomous cognitive faculty. This has been referred to as the “cognitive commitment,” i.e., the commitment to provide linguistic descriptions and explanations that accord with what is known about human mental processing in general (Lakoff 1990; Langacker 1987). As expressed by Gibbs (1996: 27), “linguistic structures are seen as being related to and motivated by human conceptual knowledge, bodily experience, and the communicative functions of discourse.” 3. A third typological pattern identified by Talmy is the one that characterizes American Indian languages where the main verb conflates motion and figure. This pattern, however, is also present in other languages such as English as the following example illustrates, It rained in through the bedroom window. Recent research into serial-verb languages (e.g., Thai and Mandarin Chinese), however, has led to the proposal of a further typological pattern, namely, equipollently-framed languages which are characterized by the expression of both manner and path by elements that are equal in both formal linguistic terms and in their force or significance (Zlatev and Yangklang 2004; Slobin 2004). 4. Talmy’s typology has also constituted the basis for contrastive analyses of novels and novel translations (Slobin 1996b, 2000), studies on gesture and language (e.g., McNeill 1997, 2000; McNeill and Duncan 2000; Özyürek and Kita 1999; Kita and Özyürek 2003), and studies on memory and categorization processes (e.g., Gennari, Sloman, Malt and Fitch 2002; Naigles and Terrazas 1998; Slobin 2000, 2003). 5. This example is, in fact, found in a commonly used Danish-Spanish dictionary (Hansen and Gawinski 1996), written by Danish native speakers.

286

Teresa Cadierno

6. In addition, Cadierno and Ruiz (2006) included a control group of learners whose L1 and L2 belonged to the same typological pattern, i.e., Italian learners of Spanish. 7. In fact, this researcher was not aware of such distinctions until she started to do research in this area, and began to notice the frequent use of these verbs by Danish native speakers. 8. The activities proposed in the next section are designed for adult L2 learners who are at the intermediate stages of language acquisition. It is important to stress that teachers do not need to carry out all the activities described, as they are meant as examples of possible activities that could be carried out when teaching motion constructions in foreign language classrooms. Finally, when implementing these activities the teacher needs to consider possible different learning styles among his/her students. 9. Some studies (e.g., Salaberry 1997; Allen 2000) have not found processing instruction to be superior to traditional output-based instruction. However, the instructional treatment administered to the processing groups was not altogether in accordance with the principles that characterize processing instruction (see Wong 2004 for a detailed critique of these studies). 10. I’d like to thank Michel Achard for this suggestion. 11. As indicated by Robinson (2003) within the framework of his Cognition Hypothesis, this activity can be made more or less complex by, for example, increasing or decreasing the number of landmarks present on the map, by having or not having the route previously marked on the map, by using maps of areas that are known or unknown by the students, and by allowing or not allowing the students to have planning time before the activity. 12. The interested reader could consult Robinson’s (2001a, 2001b, 2003) framework for task sequencing on the basis of task complexity.

References Allen, Linda Q. 2000 Form-meaning connections and the French causative. Studies in Second Language Acquisition 22: 69–84. Andersen, Roger 1984 What’s gender good for, anyway? In Second Language: A CrossLinguistic Perspective, Roger Andersen (ed.), 77–99. Rowley, Massachusetts: Newbury House. Asher, James J. 1977 Learning Another Language through Actions: The Complete Teacher’s Guidebook. Los Gatos, California: Sky Oaks.

Motion events in Danish and Spanish Aske, John 1989

287

Path predicates in English and Spanish: A closer look. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 15: 1–14. Bavin, Edith L. 2004 Focusing on where: An analysis of Warlpiri frog stories. In Relating Events in Narrative. Typological and Contextual Perspectives, Sven Strömqvist and Ludo Verhoeven (eds.), 17–35. Mahwah, New Jersey: Lawrence Erlbaum. Berman, Ruth A., and Dan I. Slobin (eds.) 1994 Relating Events in Narrative: A Crosslinguistic Developmental Study. Hillsdale, New Jersey: Lawrence Erlbaum. Caballero, Rosario 2005 Genre constraints in the use of manner of motion verbs in narrative. Paper presented at the XXII National Conference of the Spanish Association of Applied Linguistics (AESLA). University of Baleares Islands, March 2005. Cadierno, Teresa 1995 Formal instruction from a processing perspective: An investigation into the Spanish past tense. Modern Language Journal 79 (2): 179– 193. 2004 Expressing motion events in a second language: A cognitive typological perspective. In Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching, Michel Achard and Susanne Niemeier (eds.), 13–49. Berlin: Mouton de Gruyter. to appear Learning to talk about motion in a foreign language. In Handbook of Cognitive Linguistics and Second Language Acquisition, Peter Robinson and Nick Ellis (eds.). Mahwah, New Jersey: Lawrence Erlbaum. Cadierno, Teresa, and Karen Lund 2004 Cognitive linguistics and second language acquisition: Motion events in a typological framework. In Form-Meaning Connections in Second Language Acquisition, Bill VanPatten, Jessica Williams, Susanne Rott, and Mark Overstreet (eds.), 139–154. Mahwah, New Jersey: Lawrence Erlbaum. Cadierno, Teresa, and Lucas Ruiz 2006 Motion events in Spanish L2 acquisition. Annual Review of Cognitive Linguistics 4: 183–216. Doughty, Catherine, and Elizabeth Varela 1998 Communicative focus on form. In Focus on Form in Classroom Second Language Acquisition, Catherine Doughty and Jessica Williams (eds.), 114–138. Cambridge: Cambridge University Press.

288

Teresa Cadierno

Doughty, Catherine, and Jessica Williams 1998 Focus on Form in Classroom Second Language Acquisition. Cambridge: Cambridge University Press. Ellis, Nick (ed.) 1994 Implicit and Explicit Learning of Languages. London: Academic Press. Ellis, Rod 1994 The Study of Second Language Acquisition. Oxford: Oxford University Press. 1995 Interpretation tasks for grammar instruction. TESOL Quarterly 29 (1): 87–105. 2002 The place of grammar instruction in the second/foreign language curriculum. In New Perspectives on Grammar Teaching in Second Language Classrooms, Eli Hinkel and Sandra Fotos (eds.), 17–34. Mahwah, New Jersey: Lawrence Erlbaum. 2003 Task-Based Language Learning and Teaching. Oxford: Oxford University Press. Engberg-Pedersen, Elisabeth, and Frederikke B. Trondhjem 2004 Focus on action in motion descriptions: The case of WestGreenlandic. In Relating Events in Narrative: Typological and Contextual Perspectives, Svend Strömqvist and Ludo Verhoeven (eds.), 59–88. Mahwah, New Jersey: Lawrence Erlbaum. Evans, Vyvyan, and Melanie Green 2006 Cognitive Linguistics: An Introduction. Edinburg: Edinburg University Press. Fillmore Charles J. 1975 An alternative to checklist theories of meaning. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 1: 123–131. Gennari, Silvia P., Steven A. Sloman, Barbara C. Malt, and W. Tecumseh Fitch 2002 Motion events in language and cognition. Cognition 83: 49–79. Gibbs, Raymond W. 1996 What is cognitive about cognitive linguistics? In Linguistics in the Redwoods: The Expansion of a New Paradigm in Linguistics, Eugene H. Casad (ed.), 27–53. Berlin: Mouton de Gruyter. Halliday, Michael A. 1985 An Introduction to Functional Grammar. London: Edward Arnold. Hansen, Johan W., and Birthe Gawinski 1996 Dansk-Spansk Ordbog. Norbok: Munksgaard Ordbøger. Harley, Birgit, and Merrill Swain 1984 The interlanguage of immersion students and its implications for second language teaching. In Interlanguage, Alan Davies, Clive Criper, and Antony Howatt (eds.), 291–311. Edinburg: Edinburg University Press.


289

Holm, Lars, and Per Pinholt 1990 Dansk for Hele Verden. Copenhagen: Gyldendahl. Ibarretxe-Antuñano, Iraide 2004 Motion events in Basque narratives. In Relating Events in Narrative. Typological and Contextual Perspectives, Sven Strömqvist and Ludo Verhoeven (eds.), 89–111. Mahwah, New Jersey: Lawrence Erlbaum. Jeppesen, Bodil, and Grethe Maribo 2001 Mellem Linjerne. Copenhagen: Alinea. Johnson, Mark 1987 The Body in the Mind: The Bodily Basis of Meaning, Imagination, and Reason. Chicago: University of Chicago Press. Kellerman, Eric 1979 Transer and non-transfer: Where are we now? Studies in Second Language Acquisition 2: 37–57. 1986 An eye for an eye: Cross-linguistic constraints on the development of the L2 lexicon. In Cross-linguistic Influence in Second Language Acquisition, Eric Kellerman and Michael Sharwood Smith (eds.), 37–57. Oxford: Pergamon. Kellerman, Eric, and Anne-Marie Van Hoof 2003 Manual accents. International Review of Applied Linguistics 41: 251–269. Kita, Sotaro, and Asli Özyürek 2003 What does cross-linguistic variation in semantic coordination of speech and gesture reveal? Evidence for an interface representation of spatial thinking and speaking. Journal of Memory and Language 48 (1):16–32. Krashen, Stephen 1981 Second Language Acquisition and Second Language Learning. New York: Pergamon Press. Lakoff, George 1987 Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press. 1990 The invariance hypothesis: Is abstract reason based on image schemas? Cognitive Linguistics 1: 39–74. Langacker, Ronald 1987 Foundations of Cognitive Grammar, Vol. 1: Theoretical Perspectives. Stanford, California: Stanford University Press. 1996 Cognitive grammar. In Concise Encyclopedia of Syntactic Theories, Keith Brown and Jim Miller (eds.), 51–54. Oxford: Pergamon. Laursen, Elisabeth, Birgit Tegnberg-Hansen, and Jørn Wiese 1993 En Vivo. Copenhagen: GEC GADs Forlag.

290

Teresa Cadierno

Liébana, Eva, and Birgit Tegnberg 2005 A Qué sí– Et Begyndersystem i Spansk. Copenhagen: Systime. LoCoco, Veronica 1987 Learner comprehension of oral and written sentences in German and Spanish: The importance of word order. In Foreign Language Learning: A Research Perspective, Bill VanPatten, Trisha R. Dvorak, and James F. Lee (eds.), 119–129. Cambridge, Massachusetts: Newbury House. Long, Michael H. 1991 Focus on form: A design feature in language teaching methodology. In Foreign Language Research in a Cross-Cultural Perspective, Kees de Bot, Ralph B. Ginsberg, and Claire Kramsch (eds.), 39–52. Amsterdam: John Benjamins. Long, Michael H., and Peter Robinson 1998 Focus on form: Theory, research and practice. In Focus on Form in Classroom Second Language Acquisition, Catherine Doughty and Jessica Williams (eds.), 15–41. Cambridge: Cambridge University Press. McNeill, David 1997 Imagery in motion event descriptions: Gestures as part of thinkingfor-speaking in three languages. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 23: 255–267. 2000 Analogic/analytic representations and cross-linguistic differences in thinking for speaking. Cognitive Linguistics 11 (1/2): 43–60. McNeill, David, and Susan D. Duncan 2000 Growth points in thinking-for-speaking. In Language and Gesture: Window into Thought and Action, David McNeill (ed.), 141–161. Cambridge: Cambridge University Press. Meisel, Jürgen 1987 Reference to past events and actions in the development of natural second language acquisition. In First and Second Language Acquisition Processes, Carol Pfaff (ed.), 206–224. Cambridge, Massachusetts: Newbury House. Naigles, Letitia R., and Paula Terrazas 1998 Motion-verb generalizations in English and Spanish: Influences of language and syntax. Psychological Science 9 (5), 363–369. Navarro, Samuel, and Elena Nicoladis 2005 Describing motion events in adult L2 Spanish narratives. In Selected Proceedings of the 6th Conference on the Acquisition of Spanish and Portuguese as First and Second Languages, David Eddington (ed.), 102–107. Somerville, Massachusetts: Cascadilla Press.


291

Nassaji, Hossein, and Sandra Fotos 2004 Current developments in research on the teaching of grammar. Annual Review of Applied Linguistics 24: 126–145. Negueruela, Eduardo, James P. Lantolf, Stefanie R. Jordan, and Jaime Gelabert 2004 The “private function” of gesture in second language speaking activity: A study of motion verbs and gesturing in English and Spanish. International Journal of Applied Linguistics 14 (1): 113–147. Niemeier, Susanne 2004 Linguistic and cultural relativity – Reconsidered for the foreign language classroom. In Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching, Michel Achard and Susanne Niemeier (eds.), 95–118. Berlin/New York: Mouton de Gruyter. Norris, John, and Lourdes Ortega 2000 Effectiveness of L2 instruction: A research synthesis and quantitative meta-analysis. Language Learning 50: 417–528. Nunan, David 1989 Designing Tasks for the Communicative Classroom. Cambridge: Cambridge University Press Özçaliúkan, ùeyda, and Dan I. Slobin 1998 Learning how to search for the frog: Expression of manner of motion in English, Spanish, and Turkish. In Proceedings of the 23rd Annual Boston University Conference on Language Development, Annabel Greenhill, Heather Littlefield, and Cheryl Tano (eds.), 541–552. Somerville, Massachusetts: Cascadilla Press. Özyürek, Asli, and Sotaro Kita 1999 Expressing manner and path in English and Turkish: Differences in speech, gesture, and conceptualizations. In Proceedings of the Twenty First Annual Conference of the Cognitive Science Society, Martin Hahn and Scott C. Stoness (eds.), 507–512. Mahwah, New Jersey: Lawrence Erlbaum. Robinson, Peter 2001a Task complexity, cognitive resources and syllabus design: A triadic framework for examining task influences on SLA. In Cognition and Second Language Instruction, Peter Robinson (ed.), 185–316. New York: Cambridge University Press. 2001b Task complexity, task difficulty and task production: Exploring interactions in a componential framework. Applied Linguistics 22: 27– 57. 2003 The cognition hypothesis, task design, and adult task-based language learning. Second Language Studies 21 (2): 45–105. Salaberry, Rafael 1997 The role of input and output practice in second language acquisition. The Canadian Modern Language Review 53: 422–451.

292

Teresa Cadierno

Schmidt, Richard 1990 The role of consciousness in second language learning. Applied Linguistics 11 (2): 17–46. 1993 Awareness and second language acquisition. Annual Review of Applied Linguistics 13: 206–226. Slobin, Dan I. 1996a From “thought and language” to “thinking for speaking.” In Rethinking Linguistic Relativity: Studies in the Social and Cultural Foundations of Language, Vol. 17, John Gumperz and Stephen Levinson (eds.), 70–96. Cambridge: Cambridge University Press. 1996b Two ways to travel: Verbs of motion in English and Spanish. In Grammatical Constructions: Their Form and Meaning, Masayoshi Shibatani and Sandra A. Thompson (eds.), 195–220. Oxford: Clarendon Press. 1997 Mind, code, and text. In Essays on Language Function and Language Type: Dedicated to T. Givón, Joan Bybee, John Haiman, and Sandra A. Thompson (eds.), 437–467. Amsterdam/Philadelphia: John Benjamins. 1998 A typological perspective on learning to talk about space. In Learning to Talk about Time and Space. Proceedings of the 3rd NELAS Conference, Hanna Ragnarsdóttir and Sven Strömqvist (eds.), 1–30. Reykjavík and Göteborg: University Colleague of Education and Department of Linguistics, University of Göteborg. 2000 Verbalized events: A dynamic approach to linguistic relativity and determinism. In Evidence for Linguistic Relativity, Susanne Niemeier and Rene Dirven (eds.), 107–138. Amsterdam: Benjamins. 2003 Language and thought online: Cognitive consequences of linguistic relativity. In Language in Mind: Advances in the Study of Language and Thought, Dedre Gentner and Susan Goldin-Meadow (eds.), 157– 192. Cambridge, Massachusetts: Massachusetts Institute of Technology Press. 2004 The many ways to search for a frog: Linguistic typology and the expression of motion events. In Relating Events in Narrative: Typological and Contextual Perspectives, Sven Strömqvist and Ludo Verhoeven (eds.), 219–257. Mahwah, New Jersey: Lawrence Erlbaum. Slobin, Dan I., and Nini Hoiting 1994 Reference to movement in spoken and signed languages: Typological considerations. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 20: 487–503. Swain, Merrill 1985 Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In Input in Second


293

Language Acquisition, Susan M. Gass and Carolyn Madden (eds.), 235–253. Rowley, Massachusetts: Newbury House. Talmy, Leonard 1985 Lexicalization patterns: Semantic structure in lexical forms. In Language Typology and Syntactic Description, Vol. 3: Grammatical Categories and the Lexicon, Timothy Shopen (ed.), 36–149. Cambridge: Cambridge University Press. 1991 Path to realization: A typology of event conflation. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 17: 480–519. 2000 Toward a Cognitive Semantics: Typology and Process in Concept Structuring. Cambridge, Massachusetts: Massachusetts Institute of Technology Press. VanPatten, Bill 1984 Learners’ comprehension of clitic pronouns: More evidence for a word order strategy. Hispanic Linguistics 1: 57–67. 1996 Input Processing and Grammar Instruction: Theory and Research. Norwood, New Jersey: Ablex. VanPatten, Bill, and Teresa Cadierno 1993a Input processing and second language acquisition: A role for instruction. The Modern Language Journal 77: 45–57. 1993b Explicit instruction and input processing. Studies in Second Language Acquisition 15 (2): 225–243. VanPatten, Bill, and Cristina Sanz 1995 From input to output: Processing instruction and communicative tasks. In Second Language Acquisition Theory and Pedagogy, Fred R. Eckman, Diane Highland, Peter W. Lee, Jean Mileham, and Rita R. Weber (eds.), 169–185. Mahwah, New Jersey: Lawrence Elrbaum. VanPatten, Bill, Jessica Williams, Susanne Rott, and Mark Overstreet (eds.) 2004 Form-Meaning Connections in Second Language Acquisition. Mahwah, New Jersey: Lawrence Erlbaum. White, Joanna 1998 Getting the learners’ attention: A typographical input enhancement study. In Focus on Form in Classroom Second Language Acquisition, Catherine Doughty and Jessica Williams (eds.), 85–113. Cambridge: Cambridge University Press. Widdowson, Henry 1978 Teaching Language as Communication. Oxford: Oxford University Press. Wilkins, David A. 1976 Notional Syllabuses. Oxford: Oxford University Press.

294

Teresa Cadierno

Wilkins, David P. 2004 The verbalization of motion events in Arrernte. In Relating Events in Narrative: Typological and Contextual Perspectives, Sven Strömqvist and Ludo Verhoeven (eds.), 143–157. Mahwah, New Jersey: Lawrence Erlbaum. Wong, Wynne 2004 The nature of processing instruction. In Processing Instruction: Theory, Research, and Commentary, Bill VanPatten (ed.), 33–63. Mahwah, New Jersey: Lawrence Erlbaum. Zlatev, Jordan, and Peerapat Yangklang 2004 A third way to travel: The place of Thai in motion-event typology. In Relating Events in Narrative: Typological and Contextual Perspectives, Sven Strömqvist and Ludo Verhoeven (eds.), 159–190. Mahwah, New Jersey: Lawrence Erlbaum. Zlatev, Jordan, Caroline David, and Johan Blomberg in press Translocation, language and the categorization of experience. In Space, Language, and Experience, Vyvyan Evans and Paul Chilton (eds.). London: Equinox.

Motion and location events in German, French and English: A typological, contrastive and pedagogical approach Sabine De Knop and René Dirven

Abstract A pedagogical grammar – in the cognitive sense – considers language as a component of overall cognition and attempts to apply this insight to the processes of foreign language learning, especially in the preparation of instruction materials and their exploitation in the classroom. The starting point for any type of pedagogical grammar is a fairly exhaustive repertory of all the patterns or constructions of a language, either previously recognized or else recently dis-covered or still to be discovered. For the stock-taking of different patterns, the role of language typology and contrastive linguistics is primordial. This paper first concentrates on typological distinctions concerning the manner and path of motion between Germanic and Romance languages in the conceptual fields of motion and location. The study further covers the extensions of the prototypical concept of spatial motion into reduced motion of instrumental body parts, fictive motion and metaphorical motion. The examples are meant to concretize the challenges for the learning tasks of foreign language learners in their mastering of the variation in the conceptualization of motion and location in various types of languages, especially those in German as a target language for speakers of French or English. Keywords: language typology; contrastive study; motion schema; location schema; trajectory; path of motion; manner of motion; reduced motion; fictive motion; metaphorical motion; Germanic languages; Romance languages

1. Different languages – different categorizations 1.1. Foreign language learning as conceptual learning Building on a broad cognitive foundation, cognitive linguistics approaches language as an integrated system of lexical and grammatical concepts and of communicative interaction patterns. As Danesi (this volume) shows, it is

296

Sabine De Knop and René Dirven

common experience in foreign language teaching that grammatical patterns which reflect foreign conceptual categories and foreign schemas of categorization will cause greater learning difficulties to foreign language learners. This paper therefore addresses the central question of pedagogical grammar: how can foreign language learners better manage to acquire these more abstract foreign language categories. To put it in Lantolf’s terms, the real issue is this: “to what extent is it possible for people to become cognitively like members of other cultures; that is, can adults learn to construct and see the world through culturally different eyes?” (Lantolf 1999: 29– 30). In this paper we will restrict ourselves to the domain of spatial conceptualization, more particularly to the conceptualization of motion and location in both the physical and non-physical, e.g., metaphorical dimensions. We will do so against the background of the typological and contrastive studies by Talmy (1985), Slobin (1996, 2000, 2006), Cadierno (2004; this volume), Cadierno and Lund (2004), Levinson (2003), Özçaliskan (2003, 2004, 2005a, 2005b), Croft et al. (in preparation), and still others.

1.2. Components of motion events The conceptual unit in which motion and location are framed is an EVENT SCHEMA (see Heine 1993, Dirven and Radden 2004) which is the abstract representation of the participants in an event, here a motion or location event. According to Talmy (1978, 1985) a motion event contains the five basic components of figure (the “thing” moving), the verb action of motion, the trajectory or ground of the motion, itself consisting of a source or point of departure, a path followed by the moving object and the goal, as in She ran out of the kitchen down the stairs into the garden. Let us present this event schema in the following diagram. We will spell out “ground” in all its elements separately, as shown below in (1). (1) MOTION SCHEMA FIGURE She

MOTION ran

SOURCE out of the kitchen

GROUND PATH down the stairs

MANNER GOAL into the garden

(ran)

Roughly speaking, the figure is constituted by the subject of the sentence, and the ground by all the further nominal phrases (direct object, comple-

Motion and location events in German, French and English

297

ment, adjunct, etc.). Moreover, this sentence does not only contain the trajectory but also the manner of motion encapsulated in the lexical item ran (see also Talmy 1985: 61). As stated already, source, path and goal form the conceptual trajectory of any motion event, to which further also the direction of the motion is to be assigned. Leonard Talmy (1985) discovered strongly different linguistic patterns in the conceptualization of motion events across typologically different, or even related, languages, e.g., between Romance and Germanic languages. These differ especially in the expression of the path and the manner of motion. In Romance languages the manner of motion is far less – or not at all – focused upon, but in Germanic languages it is almost obligatorily differentiated: (2)

a. b. c.

Elle alla dans le jardin. ‘She went into the garden.’ She went/walked/ran/rushed into the garden. Sie ging/lief/rannte/stürmte in den Garten.

Here we witness a differentiation in the expression of auto-motion. But the difference is still more fundamental. While French uses the verb aller (‘to go’) in a schematic or abstract sense for any form of motion or change to another position, German and English must further differentiate between the means causing the motion; such that different verbs must be used: gehen (on foot), fahren (using a means of transport: car, lorry, ship, train), fliegen (in an aeroplane), etc. Of course in French there are similar verbs such as marcher ‘to walk,’ rouler ‘to drive,’ voler ‘to fly,’ but they are not obligatorily used like in German, where the speaker has no choice when he wants to express a motion event; s/he will have to conceive the manner of and/or the means of motion before speaking and choose the appropriate verb. As for the path of motion, Romance languages preferably express this by full verbs such as French entrer ‘to go/come in,’ sortir ‘to go/come out,’ aller ‘to go,’ traverser ‘to cross,’ whereas Germanic languages denote the ground elements by grammatical forms. These elements are called by Talmy “satellites:” a satellite is “the grammatical category of any constituent other than a nominal complement that is in a sister relation to the verb root” (Talmy 1991: 486). Satellites are either particles (come out) or prepositions introducing nominal groups and accompanying full verbs (e.g., dans, into, in) in (2)). Talmy therefore calls Romance languages “verb-framed” languages and Germanic languages “satellite-framed” languages.

298


Taking Talmy’s observation as our starting point, we will compare examples from French (a Romance language) with corresponding ones from German and English (Germanic languages), whereby English, due to its historical French influence, is somewhere in between the two others, though basically much closer to German. This may also entail that, more generally, we cannot simply speak of two clear-cut types of verb-framed and satellite-framed languages, but that these are rather two extreme points on a continuum with many intermediate stages and mixed cases such as Korean (see Bowerman 1996). We will now see in more detail how all this applies first to the manner of motion and next to the path of motion.

2. Motion events: Manner of motion and path of motion 2.1. Manner of motion Psycholinguistic evidence for the different conceptualization of the manner of motion in Romance and Germanic languages is provided in various research experiments by Slobin and by Özçaliskan. Starting from “The frog stories,” (Berman and Slobin 1994), in which English and Spanish speakers have to tell the picture stories verbally, Slobin shows that the “diversity of English verb + satellite constructions is impressive” (1996: 199) and that Spanish narrators devote less explicit attention to movement and ground than do English narrators: “Speakers of S-languages [i.e., satelliteframed languages] have been trained, by their language, to make more distinctions of motor pattern, rate, effect, and evaluation of movement, in comparison with speakers of V-languages [i.e., verb-framed languages] … V-languages seem to have far fewer expressive manner verbs than Slanguages” (Slobin 2000: 113). In her comparison of English and Turkish, Özçaliskan (2003: 218) comes to a similar conclusion: “English pays more attention to the manner dimension of motion events than Turkish, using a greater amount and variety of motion event types that encode manner.” Looking for motion verbs in novels and newspapers in the two languages she describes, she finds 138 types of English motion verbs expressing manner against only 39 types in Turkish (Özçaliskan 2003: 219). Also (Pourcel in press: 3) confirms this tendency for French, which “tends to background the element of manner to the extent that it is often left unsaid altogether.” She uses the example L’oiseau est sorti de sa cage, which in a literal translation would yield the un-English sentence The bird exited its


299

cage. Here English requires a verb expressing the manner of motion like fly, jump, etc. Comparing the three languages, French, German and English, we see that French uses, just like aller in (2), the very abstract motion verb sortir, and that German and English use the same manner of motion verbs as in (2). (3) MOTION SCHEMA FIGURE Il Er He

MOTION sortit ging/lief/ rannte went/walked/ ran

SOURCE de la cuisine aus der Küche out of the kitchen

MANNER O (ging/lief/ rannte) (walked/ran)

While French speakers abstract away from manner, German and English speakers, for each specific motion event, transmit the information of neutral, slower or quicker pace. If French speakers feel the wish or urgent need to express the manner of motion, they will rather use a construction with the present participle or with an adverb. Therefore a literal translation of one of the German and English examples in (3) could be: Il sortit de la cuisine en courant, lit. ‘He exited the kitchen, running,’ containing a general verb plus a present participle construction. But such constructions, though possible and grammatically correct, are not commonly used in French. Other possibilities are the use of an adverb ending in –ment, adverbial expressions like those with pas (‘pace’) or still others.1 Using a long list of English verbs expressing manner of motion, Pourcel (2005) shows that their translation into French is often only possible when using a construction with marcher + adverbial phrase; some of her examples are: to drudge – marcher péniblement; to march – marcher au pas; to plod – marcher d’un pas lent; to stomp – marcher d’un pas lourd, bruyant; to tiptoe – marcher sur la pointe des pieds. Her conclusion is clear: “the concept of manner … is highly codable in satellite-framing languages only” (Pourcel 2005: 3).2

300


2.2. Path of motion Also for the different coding of the path of motion in typologically different languages, Slobin and Özçaliskan provide evidence. In his comparative study of English and Spanish, Slobin (1996) stresses that English speakers pay more attention to path details than Spanish speakers (Slobin 1996: 201). In Özçaliskan’s cross-linguistic study of Turkish and English (2003) it becomes clear that Turkish “typically encodes direction of motion in the main verb of a clause (e.g., He enters, exits, ascends, descends), whereas English prefers to encode direction of motion by using particles or prepositions, making the main verb slot available for a manner verb (e.g., He walks, runs, crawls in/out/across)” (Özçaliskan 2003: 221). Like these two verb-framed languages, Spanish and Turkish, also French uses full verbs like traverser (‘cross’), sortir (‘go out’), entrer (‘come in’), etc. German and English prefer, though with remarkable differences, to express the path of motion at the level of the satellite. Typical German satellites are separable particles in so-called particle verbs: Wir gehen hinaus ‘we go out,’ Wir kommen herein ‘we come in,’ Gehen wir rüber? ‘shall we go across (the road)?,’ as may be seen in the following example: (4)

a. b. c.

Les athlètes traversent le fleuve (à la nage). The athletes cross the river/swim across (to the other bank). Die Athleten schwimmen ans andere Ufer.

Again the French full verb3 traverser ‘to cross’ is much more general in meaning than the German or English equivalents. English in (4b) has both patterns at its disposal here, a fact which Talmy (1985: 72) was fully aware of. Although German has the equivalent verb überqueren ’to cross,’ it is also a fact that this grammatically correct form is not actually used. The justification can be found in the preference of German to use satellites to express the path of motion. Thus in German only the expression in (5) is really authentic and idiomatic. (5)

Lass uns rübergehen. ‘Let’s go across (to the other side of the road).’

In such idiomatic language use in a concrete context of situation, the ground need not even be evoked, because it is visibly given in the speech


301

situation. In French it is totally absent, in English only the deictic location (here in Let’s cross here) is given and in German the path is expressed by the particle rüber ‘over/across.’ The conceptual differences between verb-framed and satellite-framed languages are even bigger if we have several ground elements of the trajectory SOURCE-PATH-GOAL. It does not surprise that French as a verb-framed language cannot take several ground satellites, but expresses part of this trajectory by a different verb each time, which can take one satellite only as in (6a), whereas English and German can specify the complex ground trajectory with one verb and several satellites: (6)

a. b. c.

Il sortit de la salle de séjour, traversa la cuisine pour aller au jardin. Er ging/lief/rannte aus dem Wohnzimmer durch die Küche in den Garten. He went/walked/ran out of the living-room through the kitchen into the garden.

In French the path of motion is expressed by the verb and consequently each phase in this motion trajectory must be expressed by a different verb. In German and English the path is expressed in the satellites so that the various phases of the trajectory or path can be concatenated as prepositional phrases. Within the ground trajectory, a further distinction is made between the source element on the one hand, and the path and goal elements, on the other. In German the former, fixed point of departure takes the dative declension, the latter, more mobile parts take the accusative. This aspect of the opposition between location and motion will be gone into in Section 3.

3. Motion and location: Lexical, morphosyntactic and figurative challenges As discussed so far, the different conceptualizations of motion events in Romance and Germanic languages are characterized by their preferential realizations of elements in the event schemas for motion, and in the use of general vs. differentiated motion verbs. In addition, the conceptual opposition between motion and location is marked in German by related verb pairs and the accusative vs. dative case system, and in English by the choice between some prepositions. We will successively look at the com-

302


bined choices of lexical and case expressions in the motion schema vs. the location schema (3.1), at motion to a goal vs. motion at a location (3.2), at reduced motion schemas for body parts (3.3), at fictive motion (3.4) and finally at abstract or metaphorical motion (3.5).

3.1. Motion schema and location schema The use of generalized verbs in French holds in many domains. In English, the domain of motion requires differentiated verbs, but the domain of location does not. In contrast to this, German obliges its speakers to use differentiated verbs both for motion and for location. A location schema consists of the relatively simple combination of FIGURE + LOCATION (verb) + GROUND. Thus French and English use the generalized verbs mettre/to put for the manual motion of objects, and être/to be for location, whereas German speakers must obligatorily express the orientational position of the object being moved or of the located object. On top of this, German also exploits its case system, using the accusative for dynamic motion to a goal and the dative for static motion at a given point and for location. These differences are obligatorily combined in every construal of a motion event as in the (c) examples of (7), (9), (11) and (13) or a location event as in (8), (10), (12) and (14) – see below. Note that the examples in (7) and (8) denote horizontal positions, those in (9) and (10) vertical positions. Because of the complexity of the German conceptualization in these event schemas, we will first go into the lexical specifications, and after that, into the case system. While French and English, though the latter with more variation, express the location of both animate beings or inanimate objects by means of generalized verbs, German almost always specifies the orientation of the object’s position, i.e., it spells out whether the object is in a horizontal, i.e., lying position (legen/liegen) or in a vertical, i.e., upright (stellen/stehen) position4 (see also Serra-Borneto 1996: 377). The same applies to sitting and hanging entities by means of the verb pairs setzen/sitzen and hängen/hängen respectively:


303

(7) HORIZONTAL MOTION

(8) HORIZONTAL LOCATION

a. Papa met le journal sur la table. b. Dad puts the paper on the table. c. Vati legt die Zeitung auf den Tisch. [ACC]

a. Le journal est sur la table. b. The paper is on the table. c. Die Zeitung liegt auf dem Tisch. [DAT]

(9) VERTICAL MOTION

(10) VERTICAL LOCATION

a. Papa met le vase sur la table. b. Dad puts the vase on the table. c. Vati stellt die Vase auf den Tisch. [ACC]

a. Le vase est sur la table. b. The vase is on the table. c. Die Vase steht auf dem Tisch. [DAT]

(11) MOTION EVENT

(12) LOCATION EVENT

SITTING UP

SITTING

a. Papa met le bébé dans la chaise. a. Le bébé est dans la chaise. b. Dad puts the baby in the high b. The baby is in the high chair. chair. c. Vati setzt das Baby in den c. Das Baby sitzt in dem Kinderstuhl. [ACC] Kinderstuhl. [DAT] (13) MOTION EVENT HANGING UP

a. Papa met le cadre au mur. b. Dad puts/hangs the picture on the wall. c. Vati hängt das Bild an die Wand. [ACC]

(14) LOCATION EVENT HANGING

a. Le cadre est au mur. b. The picture is on the wall. c. Das Bild hängt an der Wand. [DAT]

The difficulty for French and English speakers learning German is precisely and primarily this differentiation of the location of objects or persons. Serra-Borneto (1995: 462) defines the problem as follows:

304


… locative verbs like stehen and liegen are simply devices the speaker uses to convey information about the orientation of the object(s) he is referring to in discourse. Thus, in order to decide how to characterize the position of an object, the speaker first has to set his attention on the most relevant dimension of the object and then must match it with one of the abstract spatial axes. These cognitive operations of course imply a certain amount of “schematization,” which is ‘a process that involves the systematic selection of certain aspects of the referent scene to represent the whole, while disregarding the remaining aspects.’ (Talmy 1983: 225), because one particular feature of the overall shape of the object is isolated in order to determine its orientation.

In colloquial German one sometimes hears sentences with the more general verb tun ‘do’ corresponding to French mettre or English put as in (15): (15)

a. b. c.

Vati hat die Zeitung auf den Stuhl getan. ‘Dad has put the newspaper on the chair.’ Ich tue die Vase auf den großen Tisch. ‘I put the vase on the big table.’ Er hat meine Tasse in den Schrank getan. ‘He has put my cup into the cupboard.’

Sentence (15a) is acceptable because the speaker does not know exactly “in what position” the newspaper is (In German restaurants, newspapers are even fixed to a stick and can be hung up). Sentences (15b) and (15c), on the other hand, are very informal and would probably be rejected by purists as “bad German.” So, the use of tun is not much comfort to the learner, since this is below the general standard-variety level of the foreign language to be aimed for. The learning problem is not a cultural one, since in most cultures, artefacts and people have canonical positions.5 Canonically, a baby is normally placed in its high chair in a sitting position, and a picture in a hanging position. Still, the learning problems remain immense, because learners must grow into the habit of focusing on these canonical orientational positions when talking about them instead of abstracting away from these physical aspects as they are accustomed to in their native language. Thus learning a foreign grammar is also learning to re-orientate one’s attention to different aspects in the visual scenery. In spite of the complexity of the German way of seeing things and expressing them, the set of oppositions above may offer a very concrete, visual and even tactile way of experiencing the learn-


305

ing problem and associating the linguistic expressions with these bodily experiences. Here training in the sense of Lantolf and Thorne’s (2006) activity theory can be applied in concrete activities creating the bodily experience for the learning process. As all teachers of German as a foreign language are well aware, the learning of spatial relationships as conceptualized in German and realized through different verbs cause major learning difficulties to French speakers. But, the most severe learning difficulties are caused by the subtle interaction of verb, preposition and case system. This will be investigated in the following section.

3.2. Motion to goal vs. motion at location Though the previous presentation is correct for as far as it goes, the situation is in fact far more refined than that. There is not just a simple contrast between motion and location, but also, and equally frequently, a contrast between motion to a goal and motion at a fixed location. Here the problem is to learn to see the way the German language realizes the concept of “motion at a fixed location” as different from “motion to a goal.” This different way of seeing is simultaneously to be associated with the alternation between dative and accusative as it appears with the so-called “two-way prepositions” (Smith 1995: 293): an, auf, hinter, in, neben, über, unter, vor and zwischen (‘at,’ ‘on,’ ‘behind,’ ‘in,’ ‘next to,’ ‘over,’ ‘under,’ ‘in front of/before’ and ‘between.’ The problem for French-speaking learners can perhaps be better understood if one takes into account the preferential conceptualizations of French, which privileges more abstract motion and location verbs and tends to incorporate the path in the verb itself, whereas German – and to some extent also English – tends towards a more concrete specification of motion and location events. A traditional approach, thought to be helpful, has been the exploitation of the conceptual distinction between a “newly arising” relation of figure and ground and a “pre-existing” spatial relation between them, whereby it is necessary to distinguish between prototypical motion events as in (16) and (17) below, and the less prototypical cases in the later subsections, such as the more peripheral cases of reduced motion (Section 3.3), fictive motion (Section 3.4), and abstract or metaphorized motion (Section 3.5). (16)

a.

Er geht auf die Straße. ‘He walks into the street.’ [He was not in the street yet.]

306


b.

Er geht auf der Straße. ‘He is walking in the street.’ [He was already in the street and starts walking there.]

Still, the use of the accusative and the dative is not merely one between a relationship coming into existence (16a) and one already existing before (16b) as suggested by the comments added between brackets in (16). As Taylor (this volume) points out, the use of “rules” in FLT is indispensable, but it all depends on the quality of the rule’s formulation. Thus rules of thumb – for the opposition in (16), one could suggest “new relationship = accusative; already existing relationship = dative” – are extremely helpful, hence useful, as long as one realizes their limited value in explanatory power, as we shall see in more detail below in Section 3.4. In actual fact, the choice of the particular case depends most of all on the difference between the two event schemas under discussion, i.e., the motion schema and the location schema. The motion schema has as its ground the whole trajectory or any part(s) of it, i.e., SOURCE-PATH-GOAL. Thus we can see, as already discussed before in (6c), why a sentence such as Er geht aus dem Laden auf die [ACC] Straße, ‘He walks out of the shop into the street,’ is fully acceptable: aus dem Laden expresses the source, auf the path and die Straße the goal. But the location schema has no such trajectory and can merely express location, not any part of a trajectory. Hence German cannot say *Er geht aus dem Laden auf der [DAT] Straße because this would entail a contradictory conflation of two entirely different event schemas. The specification of location can be part of two different constructions: it can be part of the core or nucleus of an event schema as in (17a) or else it can function as a non-nuclear element with action verbs as in (17b) or (17c). (17)

a. b. c.

Der Hausmeister wohnt in der Uni. ‘The caretaker lives in the university.’ Seine Frau arbeitet in der Uni. ‘His wife works in the university.’ Ihr Kind spielt herum in der Uni. ‘Their child plays around in the university.’

In a location event as in (17a) the specification of the location in the university is an obligatory complement of the location verb wohnen ‘to live somewhere.’ But in (17b) and (17c) the action verbs arbeiten ‘to work’ and


307

spielen ‘to play’ constitute an action schema and although the specification of the place of action is always possible, it is not a necessary element of the core of an action schema. This appears from the well-known test using the paraphrase “and is doing so in X.” This works for the action schemas in (17b) and (17c), She works/plays and does so in the university, but not for the location schema in (17a): *He lives and does so in the university. It appears thus clearly that a motion schema, a location schema, and an action schema must be strictly discerned, because they determine the core elements of event schemas (for further discussion, see Chapter 11 in Radden and Dirven 2007). Although the difference in the use of the accusative and dative in (16) has already been described in much detail in a number of other cognitive approaches (e.g., Leys 1989, 1995; Di Meola 1998; Draye 1996; Meex 2002; Meex and Mortelmans 2002; Serra-Borneto 1997; Smith 1987, 1993, 1995), it has – with the exception of Meex (2002) – almost never been explored in a systematic way along lines of event schemas. Only in SerraBorneto’s (1997) study about prepositions and in Di Meola’s (1998) description of the use of the preposition entlang we can find some vague hints at the schemas we are describing in the following sections.

3.3. Reduced motion (of body parts) In addition to these prototypical event types of motion, action or location there are a number of less prototypical or more peripheral types of motion events, particularly those involving parts of the body such as the hands, the fingers, the teeth, etc., which are more difficult cases and hence require an approach in terms of figure and trajectory as the ground of the motion event. Their less prototypical character is due to the fact that these motion events describe movements of inalienable body parts and therefore cannot have a source nor a path, but only a goal as their ground. A first group consists of cases where body parts, e.g., the teeth, the eyes’ looks or the fingers, penetrate into an object or other body part seen as a container. Here a pedagogical grammar could visualize the situation by presenting the idea of a goal in the shape of a container, which also holds for French, but far less systematically for English: (18)

a. b.

in einen Apfel beißen ‘to bite into an apple’/‘mordre dans une pomme’ tief in die Augen schauen

308


c.

‘to look deep in the eyes’/‘regarder profondément dans les yeux’ sich ins Gesicht kratzen ‘to scratch one’s face’/‘se gratter au visage’

The following examples have a body part as a goal, but it is an object and not a body part which penetrates this goal: (18)

d. e.

sich in den Finger schneiden ‘to cut one’s finger’/‘se couper au doigt’ Schampoo in die Haare einmassieren ‘to shampoo one’s hair’/‘appliquer du shampoing dans ses cheveux’

Several of these expressions are also used metaphorically, e.g., sich ins Fleisch schneiden lit. ‘to cut into one’s flesh,’ which is also used in the sense of ‘to do harm to oneself,’ or ins Gras beißen, ‘to bite the dust’ (lit. into grass), meaning ‘to die.’ But the motion itself is not, or not any more, metaphorized than other parts of the expression. In this respect, these metaphorizations are completely different from cases such as sich an die Hoffnung klammern, ‘to cling to hope’ (see Section 3.5 below), where precisely the motion verb is metaphorized and linked to an abstract ground. A second group contains expressions conceptualizing a contact situation between a body part, especially the hand, and an object conceived as a vertical surface (19a) and (19b) or a horizontal surface (19c) and (19d): (19)

a.

b.

c.

d.

etwas an die Tafel schreiben ‘to write something on the blackboard’ ‘écrire quelque chose au tableau’ an die Tür klopfen ‘to knock on the door’ ‘frapper à la porte’ auf ein Blatt Papier schreiben ‘to write on a sheet of paper’ ‘écrire sur une feuille de papier’ Sonnenlotion auf die Haut gießen ‘to pour sun lotion on the skin’ ‘verser de la lotion solaire sur la peau’


309

In these examples German, and to some extent English as well, focuses on the slight dynamic aspect of the motion event, which motivates the accusative. Although the source cannot be expressed, it is part of the conceptual trajectory, of course, and can to some degree also be visualized, e.g., a motion from the teeth to the apple, from the hand to the board, from the hand to the door. To facilitate the learning of this German way of viewing things, these motions can be visualized by the graphic representations of these verbs. Such graphs may be highly suggestive to help associate such partial motion events. The arrow in the figures below represents the concrete movement of objects and/or parts of the body, which motivates the use of the accusative:

Figure 1. The goal is a container

Figure 2. The goal is a vertical or horizontal surface

The learner of German may have to be taken to a point where s/he will take in such examples quite consciously, since the motion is not necessarily explicit, although it can usually be reconstructed on the basis of our experience of the world around us: you bite into an apple with your teeth, you look at someone’s eyes with your own eyes, you write on the blackboard with the chalk in your hand, you knock on the door with the knuckles of your fingers etc. An underlying metonymic relationship connects the activity expressed by the verb to the implicitly involved part of the body (teeth or hand) or object (e.g., knife). If the teacher or learning materials create the conditions for a bodily activity so as to make the learners aware of the

310


implicit metonymic relationship that motivates the use of the accusative, the learner may come to perceive such aspects of the experiential world in a more differentiated way and learn to “see” the formerly unspecified event as a motion event.

3.4. Fictive motion The notion of fictive motion stems from Talmy (1991) and is also – under the name of ‘subjective’ or ‘mental motion’ – discussed by Langacker (1987: 171), Matsumoto (1996: 185) and Matlock (2004a, 2004b). It relates to the scanning motion we make in the mind when speaking about trajectories such as roads, rivers, mountain ranges, railways, etc. in the topological environment. While the phenomenon of fictive motion is found in many languages and does not, in this respect, constitute a specific feature of German, the interesting thing is that German uses the accusative for the path and goal elements. As shown in (20) and (21): it is only the path that is preferentially selected and the source or goal are less apt to become prominent, although it is always possible as (21) shows: (20)

Der Weg läuft durch ein Naturschutzgebiet. ‘The road goes through a natural reserve.’

(21)

Der Rhein fließt von der Schweiz durch drei Bundesländer bis in die Niederlande. ‘The Rhine flows from Switzerland through three federal states up to the Netherlands.’

These cases of fictive motion start, like most cases discussed in the literature, from real motion verbs and relate to motion that is scanned in the mind only, not in reality and therefore is fictive, subjective and abstract. But in fact fictive motion covers a much wider conceptual area and can also take place in cases where no motion verb and no motion event seems to be involved, but where the fictive motion is realized as a satellite as in (22): (22)

Meine Freundin wohnt um die Ecke. ‘My friend lives round the corner.’


311

Obviously, this utterance informs the hearer about the location of the friend’s dwelling place, but it does so by tracing a path from where the speaker is at speech time or from his or her own dwelling place to the friend’s place. This motion is not stated explicitly, but implied in the use of the motion event presupposed by the use of the path expression um die Ecke ‘round the corner’ in the location slot of the location schema, evoked by wohnen ‘to live somewhere’ (compare with (17a) discussed above). The meaning of sentence (22) could be spelled out in full as follows: “If you want to know where my friend lives, you have to go from here (or from my place) round the corner.” What is involved here is a conceptual blending or integration of two event schemas: the location schema (expressed by the verb) and the motion schema (expressed by the satellite um). Here it may also become clear that an explanation in terms of the well-known rule of thumb for the dative/accusative alternation (an already existing relation = dative; a newly arising situation = accusative) is of little or no avail in such conceptually highly complex, but pragmatically fully justified and very frequently used blends of the location schema and the motion schema. Especially in the following figurative uses of this blend of a fictive motion with a location construction, it may at first sight look as if only a location schema is involved, but on closer inspection the sentences in (23) cannot but mean that the dynamic path situations in the ground of the location schema are conceptualized as the final stage of an imagined or fictive trajectory belonging to a motion event. This also appears from the possibility of adding some motion verb, as exemplified in (23). It must be clear, however, that the verb form between square brackets is not part of the linguistic expression, but only of their underlying blended location and motion schemas. (23)

a.

b.

c.

Wir sind endlich über den Berg [gelangt]. We are over the mountain [arrived] = ‘We have overcome all obstacles.’ Er ist über alle Berge [gelangt]. He is over all mountains [gone] = ‘He is far away.’ Wir sind über dass Schlimmste hinweg [gekommen]. We are over the worst [come] = ‘We have left the worst behind us.’

312


Clearly, it is again the whole motion event that has served as the input of the metaphorical expression, not the verb or any other separate element of the motion event. The first cognitive linguist to realize the contrastive relevance of the fictive motion schema for foreign language learning is Serra-Borneto (1997: 192). But he does not make any explicit distinction between the presence of a verb of motion or measure (24a) and its total absence (24b). (24)

a. b.

Das Wasser reicht ihm bis über die Schenkel. ‘The water is above his thighs.’ Er ist über seine besten Jahre hinweg. [He is over his best years] ‘He has left the best years of his life behind him.’ (Serra-Borneto 1997: 192)

Although the situation in (24a) looks static, there is much more to it according to Serra-Borneto: “you can imagine the eyes of the speaker following a trajectory from the ground up to the thighs and beyond them.” (SerraBorneto 1997: 192). He goes one step further when he claims that “the situation (and the verb) is static but dynamism is subjectively added to the configuration” (1997: 192). At first sight this explanation may look a bit subjective, but it is certainly welcome and acceptable in language instruction, though it might require much more imagination for the metaphorical use in (24b). There is very strong psycholinguistic evidence based on empirical experiments for the visual processing of fictive motion. Matlock (2004b: 227) speaks of the undeniable necessity of “visualization:” while reading examples like The table goes from the kitchen wall to the sliding glass door, “we automatically visualize a table that is long and narrow, perhaps a long oval table or a rectangular table.” She comes to the conclusion that “Human experience with motion goes beyond actual movement and perceived movement – it also includes mentally simulated or imagined motion.” (Matlock 2004b: 235). In our opinion, this psycholinguistic fact is a very strong hint to incorporate the notion of fictive motion into language instruction and to provide for explicit visualizations in learning materials. However, we need to assume a further cognitive process in order to be able to understand the interpretation of a special type of abstract motion, i.e., metaphorical motion, which will be discussed in the next section.


313

3.5. Metaphorical motion Whereas fictive motion as in The road runs along the coast requires mental scanning, this is not the case with metaphorical motion as in The years run by. What is expressed in this metaphor is that the years change quickly. As Lakoff and Johnson (1980) found out, the more abstract domains of time and of change are mapped onto the concrete domain of space. This is in the first place a conceptual process and linguistic metaphors like years running by are based on underlying conceptual metaphors like TIME IS MOTION and CHANGE IS MOTION (The use of upper case marks the conceptual status). Clearly, a metaphorical source domain like MOTION is not reserved for one target domain, but may serve several other and far more abstract target domains, e.g., that of mental activity as illustrated in the conceptual metaphor ABSTRACT MENTAL ACTIVITY IS MOTION. The concept of “motion” itself can be linked to that of “location” since MOTION IS CHANGE OF LOCATION. As the input for the metaphorization process we may have, not just one single concept like “motion”, but the whole of the motion schema. This schema can therefore be metaphorically extended in its CHANGE IS MOTION component itself or in its (changed) location component, or in both. The final location is of course the goal component. If the goal is mapped onto abstract domains, the verb signaling the change may or may not be metaphorized. For speakers of French and English learning German it cannot be surprising to find that such general tendencies are also operative in German, just like in their native languages. Thus spatial distinctions such as those between containers and surfaces will also be extended to more abstract areas of experience, especially in the context of situations describing abstract changes. Here one of the main learning problems for the learner is to find out whether the abstract goal is conceptualized as a container, as a surface or as still some other basic spatial relation. In the abstract motion event, German uses expressions with the preposition in and exploits the CONTAINER metaphor6 for a great variety of contexts such as becoming member of a group, splitting a group into subgroups, concentrating on one’s work, falling in love, translating, etc. : (25)

a.

b. c.

Er fügt sich in die Gruppe Lit.: ‘He joins (himself in) the group,’ i.e., ‘He becomes a member of the group.’ Die Lehrerin hat die Klasse in drei Gruppen eingeteilt ‘The teacher has split the class into three groups.’ Der Schüler vertieft sich in die Aufgaben

314


d. e.

‘The pupil goes deeper into, i.e., concentrates on, the study questions.’ Peter hat sich in seine Mitarbeiterin verliebt ‘Peter has fallen in love with his colleague.’ in eine andere Sprache übersetzen ‘to translate into another language.’

Here again, the path is not explicitly expressed but it can be imagined in our minds starting from our knowledge of physical and fictive motion events (See the arrow between the figure and the goal in Figure 1). Moreover, these examples differ from the examples of fictive motion because now the goal does not contain any features of a physical location, but only abstract locations such as a group of persons (25a) and (25b), mental entities like study questions (25c), an individual personality (25d), or some conceptual content laid down in words (25e). Smith (1995: 313) offers an explanation for the verb übersetzen ‘to translate’ by appealing to a version of the rule of thumb for the accusative/dative alternation, i.e., that the choice of case marking depends on whether a change takes place (requiring the accusative) or not (requiring the dative). He illustrates this by Hans hat den Brief ins Deutsche übersetzt ‘Hans translated the letter into German.’ But this concept of “change” is far too general in that it is not clear what exactly gets changed here. In actual fact, no internal change or change of location takes place in translations, but at best there is a kind of abstract motion that might possibly be involved: the conceptual content of a text in one language is mentally recreated in a different language, thereby creating a new text. Of course, Smith’s explanation is not wrong, but it does not pertain to the psychological reality of the translation process, but only to its folk representation via the implicitly invoked conduit metaphor for communication. According to this folk metaphor of verbal communication, contents get packed into word containers in one language, carried across to a receiver and there unpacked or decoded, or – as in the case of translation – repacked by the translator into the word containers of a different language. With the German preposition an ‘to/at/on,’ the abstract goal of the metaphorized verb is construed as a vertical surface: (26)

a. b.

sich an die Hoffnung klammern ‘to cling to hope’ an einen Vertrag gebunden sein ‘to be tied, i.e., committed, to a contract’


c.

315

sich an eine Entscheidung halten ‘to hold to = to stick to a decision’

In expressions with auf, the goal is shaped as a horizontal surface to land on: (27)

a. b.

c.

auf ein altes Problem eingehen ‘to enter into, i.e., to deal with, an old problem’ auf andere Gedanken bringen ‘to take somebody to other thoughts, i.e., to get him to think of different things‘ Er geht mir auf die Nerven ‘He goes, i.e., works, on my nerves = he upsets me’

Again, these examples can be represented in a picture which as a didactic representation visualizes the motion from the figure to the vertical or horizontal goal (see Figure 2). But language is not always such a clear-cut, neat system where all instances of usage are either in or outside a category. There are also many socalled fully idiomatic cases in every pattern of language. It is even the great merit of hundreds of cognitive linguists, e.g., Gibbs and O’Brien (1990) to have integrated the study of idiom into the core business of the science of language. Therefore, even if the following cases are more peripheral and used to be called (unexplainable) idiomatic uses, they belong to the real essence of language. On closer inspection, there is even a lot of systematicity in idioms, and there is no way of stopping the drive of abstract motion, nor its expression by means of the German accusative. (28)

a. b. c. d. e.

jemandem etwas ins Gesicht sagen ‘to say something (straight) to someone’s face’ Er gehört ins Gefängnis ‘He belongs, i.e., ought to be, in prison’ ins Gerede kommen ‘to get talked about’ Komm mir nicht in die Quere ‘Don’t cross my path’ = ‘Don’t contradict me’ Hämmere dir das in den Schädel ‘Hammer this into your skull’ = ‘Memorize this’

316


f. g.

jemandem ins Wort fallen ‘to fall into someone’s word’ = ‘to interrupt someone’ die Beine in die Hand nehmen ‘to take the legs in the hand’ = ‘to run very quickly and disappear.’

Other idiomatic expressions conceptualize the goal as a surface and express the satellite by means of auf. (29)

a. b.

auf einen grünen Zweig kommen ‘to come onto a green branch’ = ‘to have no money problems’ etwas auf die lange Bank schieben ‘to shove (something) on the long bench’ = ‘to put something off.’

Other idiomatic expressions as in (30) exploit the conceptual metaphors and INACCESSIBILITY IS A (BLOCKING) CONTAINER (see Morgan 1997). In order to “see” the truth or the real meaning of things, one must, in English, look “through” the blocking container or find things “out” of it by unblocking the container and thus making things accessible to sight and knowing. In German there is rather the image of a BLOCKING VERTICAL SURFACE, behind which one must look. To some important extent, the learning of foreign languages is learning which images a given language has chosen to express more universal experiences. In other words, learning foreign languages is learning to see how they see the world: KNOWING IS SEEING

(30)

a. b.

hinter die Fassade schauen Lit. ‘to look behind the façade’ = ‘to see through something’ Er ist hinter das Geheimnis gekommen Lit. ‘He has come behind the secret’ = ‘He has found out about it.’

Since no psycholinguistic confirmation of abstract motion in the case of figurative expressions has been found thus far, it may be speculative, though not unreasonable, to hypothesize that a kind of mental motion is the metaphorical interpretation of the examples above. If students of German are asked to describe the motion in the examples above they will often make a gesture representing this visualization. As it is a preferred learning


317

aid, teachers can exploit this possibility when teaching such constructions of abstract motion. 3.6. Consequences for the learning of German The difficulties encountered in case marking with abstract verbs can be considerably reduced if FL teachers attempt a straightforward explanation by trying to visualize the abstract motion, taking into account the relevant conceptual metaphors. They might perhaps proceed as follows: Imagine the act of klammern ‘to cling to’ in example (26a). How do you perform the act of klammern ‘cling to?’ Which body part do you need in order to be able to klammern? Show how you can klammern. The learners will then probably suggest hands. Take another example, e.g., (30a) hinter die Fassade schauen as shown in Figure 3. X

Y

V Figure 3. The goal is behind a blocking surface

If something (X) is hidden behind something else (Y) from the viewer’s viewpoint (V), then it is hidden from sight and knowledge. The question the teacher can ask to start with is why it is necessary to look “behind” something. The students will probably answer that it is because some object or person is hidden behind something else. After that he can ask what is happening when you look behind something. Which body part do you

318


use to do so? The students will again visualize the abstract motion by means of a gesture and often try to represent an implicit trajectory or path from a viewer towards an object which is hidden behind something else. Alternatively the teacher can ask the students to draw a little picture. The next step in the teacher-student dialogue could be to show that the underlying conceptual metaphor in all our examples, i.e., ABSTRACT MENTAL ACTIVITY IS MOTION, is usually realized by means of the accusative in German. 4. Conclusions At the descriptive and explanatory level, the present event-schema-based analysis of the German conceptualization of motion and location events in contrast to those in French and English has revealed the rich potential of a more comprehensive conceptual approach to the various frames and event schemas underlying the lexicon and grammar in a given conceptual domain. Indeed, it looks as if the whole German spatial prepositional system and its case marking system may find a natural explanation in terms of the event schemas for motion and location. Both schemas can nicely show the motivation for the two-way prepositions taking either the accusative or dative, as shown in the above sections. But it looks to us that there is more to it. The location schema could possibly also account for the exclusive dative use with the prepositions aus, bei, mit, seit and von,7 which all express the SOURCE in the motion schema as a fixed location in space. Similarly, the motivation for the exclusive accusative use with the prepositions bis, durch, für, gegen, um, and wider is likely to reside in the element PATH (durch, für, um) or GOAL (bis, gegen) of the trajectory in the motion schema or the integrated location and fictive-motion schema, either in a concrete or in a more abstract sense. Evidently such hypotheses are to be tested in empirical research and consequently can be confirmed or rejected. It will also have become clear from the conceptual fields of motion and location that the numerous differences between French, English and German can be systematically explained – and incorporated into teaching materials on this basis – only by taking into account the way in which these languages have realized their conceptualizations. These insights may offer important chances for FLT. Language teachers often react to sentences produced by learners by simply remarking: “You don’t say that in German,” without any further explanation why this might be so. On the basis of the explanations about the differences between Romance and Germanic


319

languages as offered in various cognitive-linguistic studies, learners can become aware of the deeper conceptual differences by means of simple contrastive examples. This means not only that learners must acquire the lexical items, e.g., the specific verbs expressing the type of motion and location, but also, and most importantly, that they must learn to approach and experience lexical and grammatical concepts in the source and target languages differently. In this learning process they will also have to consider the satellite constructions of a verb, which cannot be used freely but are an obligatory component of the German, and partly also of the English, way of seeing and expressing things and events. Finally, students will grasp the abstract regularities through the use of schemas or pictures. Up to now in foreign language instruction, learners have at best been advised very vaguely to “learn to think in the foreign language.” As we have seen, this can only be achieved with a teaching method which aims for an understanding of the foreign language and culture based on bodily experiences in the form of activities, visual representations and cognitive insight. Thus we are now in a better position to understand the underlying meaning of this motto. Notes 1.

2.

3.

Pourcel and Kopecka (2005) show that there is much more variability in the expression of motion events in French than generally accepted from Talmy’s typology. They suggest that the above pattern is not the only one available in French. For more details, see Pourcel and Kopecka 2005. If manner, though, is expressed in the main verb in French, it is the result of being the expression of a “motion activity” and not of a “motion event.” To quote Pourcel and Kopecka (2005: 5), “[the] semantic emphasis in an activity consists of the Manner of motion as an end in itself, further specifying a motion in progress. We suggest that it is relevant, therefore, to make a distinction between ‘motion activity’ and ‘motion event.’ Unlike motion events, activities are essentially concerned with conveying information relating to the Manner of motion.” So, consequently, in their example “Marc (F) court (M) dans la rue (G),” we have a motion activity and not an event. In her investigation of French motion verbs Kopecka (2006) shows that French can also express the path of motion in a prefix revealing a satelliteframed pattern attributed to Germanic and Slavic languages, e.g., “L’oiseau s’est envolé (P + M) du nid” (Pourcel and Kopecka 2005: 11). It seems that the expression of path in a prefix is widely spread in French. For more examples see Kopecka (2006).

320 4.

5.

6.

7.

Sabine De Knop and René Dirven Newman (2004) postulates that “sitting, standing, and lying would appear to be the three main positions which humans assume, though admittedly with some cultural variations. We typically experience all three in any 24 hour cycle, …” (205). Interestingly enough, this three-way distinction between sitting, standing and lying must also be expressed in a number of African languages like in Kxoé or Mbay, in which “existential constructions and predicative adjective constructions require one of the three verbs ndì ‘sit,’ dà ‘stand’ and tò ‘lie’” (Newman 2004: 206). Some more examples of languages from all over the world (like Ameridian, Caucasian or Oceanic languages) which make this difference in postures can be found in Newman (2004). Serra-Borneto (1997) uses the CONTAINER schema to rather justify the use of the dative, and indeed he brings examples where the dative is the correct form, e.g., “Die Mutter konnte ihr Kind vor diesen Gefahren nicht behüten.” (1997: 196). The CONTAINER schema is composed of “entailments,” like protection or resistance, limitation or restriction of forces, the relative fixity of location, … (1997: 196–202) This further subdivision is not relevant for our examples. For more details see Serra-Borneto (1997). Here the two prepositions nach ‘to’ and zu ‘to’ seem to contradict the generalization proposed for German dative prepositions. However, only detailed corpus research can falsify or confirm the proposed generalization.

References Berman, Ruth A., and Dan I. Slobin 1994 Development of linguistic forms: English. In Relating Events in Narrative: A Crosslinguistic Developmental Study, Ruth A. Berman and Dan I. Slobin (eds.), 127–188. Hillsdale, New Jersey: Lawrence Erlbaum. Bowerman, Melissa 1996 The origin of children’s spatial semantic categories: Cognitive versus linguistic determinism. In Rethinking Linguistic Relativity, John J. Gumperz and Stephen C. Levinson (eds.), 145–176. Cambridge: Cambridge University Press. Cadierno, Teresa 2004 Expressing motion events in a second language: A cognitive typological perspective. In Cognitive linguistics, Second Language Acquisition, and Foreign Language Teaching, Michel Achard and Susanne Niemeier (eds.), 13–49. Berlin: Mouton de Gruyter.


321

Cadierno, Teresa, and Karen Lund 2004 Cognitive linguistics and second language acquisition: Motion events in a typological framework. In Form-Meaning Connections in Second Language Acquisition, Bill VanPatten, Jessica Williams, Susanne Rott, and Mark Overstreet (eds.), 139–154. Mahwah, New Jersey: Lawrence Erlbaum. Croft, William, Johanna Barddal, Willem Hollmann, Violeta Sotirova, and Chiaki Taoka in prep. Revising Talmy’s typological classification of complex events. Ms. Manchester, United Kingdom. Di Meola, Claudio 1998 Semantisch relevante und semantisch irrelevante Kasusalternation am Beispiel von ‘entlang.’ Zeitschrift für Sprachwissenschaft 26: 204– 235. Dirven, René, and Günter Radden 2004 Putting things together: Syntax. In Cognitive Exploration of Language and Linguistics: Cognitive Linguistics in Practice 1, René Dirven and Marjolijn Verspoor (eds.), 79–105. Amsterdam/ Philadelphia: John Benjamins. Draye, Luk 1996 The German Dative. In The Dative: Descriptive Studies. Vol. 1, William Van Belle and Willy Van Langendonck (eds.), 155–215. Amsterdam/Philadelphia: John Benjamins. Gibbs, Raymond W., and Jennifer Ellen O’Brien 1990 Idioms and mental imagery: The metaphorical motivation for idiomatic meaning. Cognition 36: 35–68. Heine, Bernd 1993 Auxiliaries, Cognitive Forces and Grammaticalization. Oxford: Oxford University Press. Kopecka, Anetta 2006 The semantic structure of motion verbs in French: Typological perspectives. In Space in Languages: Linguistic Systems and Cognitive Categories, Maya Hickmann and Stéphane Robert (eds.), 83–101. Amsterdam/Philadelphia: John Benjamins. Lakoff, George, and Mark Johnson 1980 Metaphors We Live By. Chicago/London: University of Chicago Press. Langacker, Ronald W. 1987 Foundations of Cognitive Grammar. Vol. 1. Stanford, California: Stanford University Press.

322


Lantolf, James P. 1999 Second culture acquisition: Cognitive considerations. In Culture in Second Language Teaching and Learning, Eli Hinkel (ed.), 28–46. Cambridge: Cambridge University Press. Lantolf, James P., and Steve Thorne 2006 Sociocultural Theory and the Genesis of Second Language Development. Oxford: Oxford University Press. Levinson, Stephen C. 2003 Space in Language and Cognition: Explorations in Cognitive Diversity. Cambridge: Cambridge University Press. Leys, Odo 1989 Aspekt und Rektion räumlicher Präpositionen. Deutsche Sprache 17: 97–113. 1995 Dativ und Akkusativ in der deutschen Sprache der Gegenwart. Leuvense Bijdragen 84: 39–62. Matlock, Teenie 2004a Fictive motion as cognitive simulation. Memory and Cognition 32 (8): 1389–1400. 2004b The conceptual motivation of fictive motion. In Studies in Linguistic Motivation, Günter Radden and Klaus-Uwe Panther (eds.), 221–248. Berlin: Mouton de Gruyter. Matsumoto, Yo 1996 Subjective motion and English and Japanese verbs. Cognitive Linguistics 7: 183–226. Meex, Birgitta 2002 Die Wegpräposition “über.” In Perspectives on Prepositions, Hubert Cuyckens and Günter Radden (eds.), 157–176. Tübingen: Max Niemeyer. Meex, Birgitta, and Tanja Mortelmans 2002 Grammatik und Kognition. Germanistische Mitteilungen 56: 48–66. Morgan, Pamela S. 1997 Figuring out figure out: Metaphor and the semantics of the English verb-particle construction. Cognitive Linguistics 8 (4): 327–357. Newman, John 2002 The Linguistics of Sitting, Standing, and Lying. Amsterdam/Philadelphia: John Benjamins. 2004 Motivating the uses of basic verbs: Linguistic and extralinguistic considerations. In Studies in Linguistic Motivation, Günther Radden and Klaus-Uwe Panther (eds.), 193–218. Berlin: Mouton de Gruyter. Özçaliskan, Seyda 2003 Metaphorical motion in crosslinguistic perspective: A comparison of English and Turkish. Metaphor and Symbol 18 (3): 189–228.

Motion and location events in German, French and English 2004

323

Typological variation in encoding the manner, path, and ground components of a metaphorical motion event. Annual Review of Cognitive Linguistics 2: 73–102. 2005a Metaphor meets typology: Ways of moving metaphorically in English and Turkish. Cognitive Linguistics 16 (1): 207–246. 2005b On learning to draw the distinction between physical and metaphorical motion: Is metaphor an early emerging cognitive and linguistic capacity? Journal of Child Language 32 (2): 291–308. Pourcel, Stéphanie 2005 Linguistic relativity in cognitive processes. Manuscript presented at 1st UK Cognitive Linguistics Conference: New Directions in Cognitive Linguistics, University of Sussex, October 2005. in press What makes path of motion salient? Proceedings of the Annual Meeting of the Berkeley Linguistics Society 13. Pourcel, Stéphanie, and Anetta Kopecka 2005 Motion expression in French: Typological diversity. Durham and Newcastle Working Papers in Linguistics 11: 139–153. Radden, Günter, and René Dirven 2007 Cognitive English Grammar. Amsterdam/Philadelphia: John Benjamins. Serra-Borneto, Carlo 1995 Liegen and stehen in German: A study in horizontality and verticality. In Cognitive Linguistics in the Redwoods: The Expansion of a New Paradigm in Linguistic, Eugene H. Casad (ed.), 459–505. Berlin/New York: Mouton de Gruyter. 1996 Polarity and metaphor in German. In The Construal of Space in Language and Thought, Martin Pütz and René Dirven, 373–394. Berlin/New York: Mouton de Gruyter. 1997 Two-way prepositions in German: Image and constraints. In Lexical and Syntactical Constructions and the Construction of Meaning, Marjolijn Verspoor, Kee Dong Lee, and Eve Sweetser (eds.), 187– 204. Amsterdam/Philadelphia: John Benjamins. Slobin, Dan I. 1996 Two ways to travel: Verbs of motion in English and Spanish. In Grammatical Constructions: Their Form and Meaning, Masayoshi Shibatani and Sandra A. Thompson (eds.), 195–220. Oxford: Clarendon Press. 2000 Verbalized events: A dynamic approach to linguistic relativity and determinism. In Evidence for Linguistic Relativity, Susanne Niemeier and René Dirven (eds.), 107–138. Amsterdam/Philadelphia: John Benjamins. 2006 What makes manner of motion salient? Explorations in linguistic typology, discourse, and cognition. In Space in Languages: Linguis-

324


tic Systems and Cognitive Categories, Maya Hickmann and Stéphane Robert (eds.), 59–81. Amsterdam/Philadelphia: John Benjamins. Smith, Michael B. 1987 The semantics of dative and accusative in German: An investigation in cognitive grammar. PhD dissertation, Department of Linguistics, University of California at San Diego. 1993 Cases as conceptual categories: Evidence from German. In Conceptualizations and Mental Processing in Language: A Selection of Papers from the First International Cognitive Linguistics Conference in Duisburg, 1989, Richard A. Geiger and Brygida Rudzka-Ostyn (eds.), 531–565. Berlin/New York: Mouton de Gruyter. 1995 Semantic motivation vs. arbitrariness in grammar: Toward a more general account of the DAT/ACC contrast with two-way prepositions. In Insights in Germanic Linguistics: Methodology and Transition, Irmengard Rauch and Gerald F. Carr (eds.), 293–323. Berlin/New York: Mouton de Gruyter. Talmy, Leonard 1978 Figure and ground in complex sentences. In Universals of Human Language. Vol. 4: Syntax, Joseph H. Greenberg, Charles A. Ferguson, and Edith A. Moravcsik (eds.), 625–649. Stanford, California: Stanford University Press. 1985 Lexicalization patterns: Semantic structure in lexical forms. In Language Typology and Syntactic Description. Vol. 3: Grammatical Categories and the Lexicon, Timothy Shopen (ed.), 57–149. New York: Cambridge University Press. 1991 Path to realization: A typology of event conflation. Proceedings of the Annual Meeting of the Berkeley Linguistics Society 17: 480–519.

Making progress simpler? Applying cognitive grammar to tense-aspect teaching in the German EFL classroom Susanne Niemeier and Monika Reif

Abstract This paper centres around the idea that an integration of cognitive grammar principles into tense-aspect teaching may help EFL learners to develop a more meaningful understanding of the concepts underlying these grammatical constructions. In the first part of the paper, a short review of recent studies on tense-aspect acquisition and learning is provided, as well as some background information on tenseaspect teaching in the German EFL context. The second part is concerned with theoretical linguistic considerations and outlines a synthesis of several cognitive approaches to the English tense-aspect system. On the basis of this theoretical framework, suggestions are put forward in the final part as to the implementation of concrete strategies and materials in tense-aspect instruction. Keywords: cognitive grammar; EFL teaching; grammatical aspect; lexical aspect; mental spaces; reality status; relevance time; tense; (un)boundedness; Germany

1. Introduction Imagine the following scenario: A student attending a German secondary school learns in her first year of English that the present progressive expresses “… dass jemand gerade etwas tut oder dass ein Vorgang noch im Gange ist” [… that someone is doing something at speech time or that a process is still ongoing] (Schwarz 2001: 133), as stated in the textbook English G 2000. In her next contribution in class, the student utters the sentence I’m loving my horse, but the teacher corrects her and explains that the verb love is an exception which never takes the progressive. A few months later, this very student comes across the slogan of an advertisement of a well-known fast food chain, “I’m lovin’ it.” At this stage, confusion is great and the student might well give up on trying to detect regularities

326

Susanne Niemeier and Monika Reif

behind certain grammatical phenomena such as the English tense-aspect system. As we will outline in more detail in Section 2, textbook grammars1 used in the German EFL context have so far placed strong emphasis on form and use, while the meaningfulness of grammatical constructions (in the sense of a conceptual characterization of grammar) has been widely neglected. This might be one reason why German students tend to experience considerable difficulties with grammatical units whose underlying concepts differ from those of their L1 or are non-existent in their mother tongue (cf. Dürich 2005). Thus, the predominance of structural approaches in the EFL teaching materials of the last few decades seems to have left learners with a need for plausible conceptual-semantic concepts (cf. also Bardovi-Harlig 1995). Furthermore, a study by Bardovi-Harlig and Bofman (1989) has shown that especially the context-appropriate usage of certain grammatical constructions in English is considered a major source of error for EFL students, a finding which highlights a need for more contextualized and usage-based input. Although a certain discrepancy may exist between what the textbooks suggest and what is actually taught in EFL classrooms – depending on the individual teacher’s view of the nature of grammar –, at least "the first few years of English language instruction in German schools are very much determined by a particular coursebook series" (Römer 2005: 171). From our own experience, even language courses at universities frequently consist of large parts of textbook work. Zydatiß’ (1976b: 352) statement on the progressive aspect, which he considers “one of the elements within the English language whose syntax and semantics have remained rather elusive concepts for most learners,” is still valid today. Observations like these disclose a general desideratum for a slightly different approach to grammar. Grammar teaching should a) allow learners to develop their own meaningful understanding of the concepts underlying grammatical constructions, and b) be suited for integration into communicative language learning environments. As Achard (2004) has argued, cognitive aproaches to grammar may offer possibilities to combine grammar instruction and communicative methodology. We will therefore focus on the introduction of cognitive linguistic concepts into tense-aspect instruction and suggest teaching strategies and materials that, in addition to accommodating form and contextualized use, pay particular regard to the notions of meaning and construal, thus reflecting the basic principles of cognitive grammar. Our paper is organized as follows: Section 2 presents some background information on tense-aspect teaching in German schools. Section 3 ad-

Applying cognitive grammar to tense-aspect teaching

327

dresses three major points of criticism which must be leveled at “traditional” tense-aspect teaching/learning materials from a cognitive linguistic perspective, while Section 4 states reasons for the pedagogical potential of cognitive grammar frameworks. In Section 5, a synthesis of several cognitive approaches to the English tense-aspect system is outlined, providing the linguistic background for our pedagogical considerations and teaching suggestions in Section 6. This last part is primarily concerned with grammatical constructions that pose a problem to German learners of English, partly due to divergences between the L1 and the target language system, partly due to inherent “properties” of the phenomena themselves such as conceptual and grammatic complexity, which may exacerbate cognitive processing. Concrete suggestions for strategies and teaching/learning materials in tense-aspect instruction are put forward. 2. Tense and aspect in German EFL pedagogy: Leaving the past behind According to the German curriculum for English as a foreign language at secondary level, one of the major objectives of EFL instruction is communicative competence: “Grundlegendes Ziel des Englischunterrichts ist die Vermittlung der Fähigkeit, sich mündlich und schriftlich mit anderen zu verständigen, die Englisch als Muttersprache oder als ‘lingua franca’ benutzen” [The basic aim of English instruction is enabling the learners to communicate orally and in writing with people speaking English as native speakers or as a lingua franca]. The comprehension and native-like use of different grammatical phenomena such as tense and aspect are also explicitly stated as (more concrete) learning targets in the curriculum, which indicates that the term ‘communicative competence’ is meant to include grammatical competence as well. This view largely conforms to more recent insights by EFL researchers, methodologist and practicioners, who “have commented that grammatical competence is essential for communication … but cannot be attained solely through exposure to meaningful input” (Hinkel and Fotos 2002: 5). The effectiveness of both predominantly structural syllabuses and solely communicative approaches has been questioned in the linguistic and pedagogical literature of the last decades, though obviously for different reasons. Advocates of communicative methodologies have criticized traditional formal grammar instruction because many learners came to know the grammar rules “but could not use the target language communicatively”

328


(Hinkel and Fotos 2002: 4). For instance, there appeared to be a substantial discrepancy between the number of errors instructed learners committed in the area of tense-aspect morphology and the number of contextually inappropriate instances of tense-aspect use they produced. In a study of compositions written by advanced EFL learners, Bardovi-Harlig and Bofman (1989: 26–28) found that semantically inconsistent use of tenses ("misuse") occurred much more frequently than errors in verbal morphology ("ill-formedness"). Learners were able to produce morphosyntactically well-formed verb strings such as has been developed, but these forms frequently occurred “in environments not predicted by the target language” (Bardovi-Harlig 1995: 154). On the other hand, it has become obvious that the purely communicative approaches, many of which emerged in the U.S.A and Canada during the 1980s as a response to Krashen‘s Monitor Model, can not be called flawless either. Especially the non-existence of grammar instruction in the communicative language classroom has been a highly debated issue, since learners in immersion programmes would not intuitively achieve grammatical accuracy. Pedagogues as well as applied linguists have therefore expressed their fear of a fossilization of learner errors (cf. Hinkel and Fotos 2002: 5). We argue in favor of an approach to EFL teaching that incorporates both communicative principles and explicit grammar instruction (cf. also Achard 2004). While structuralist and generativist views of grammar appear to be largely incompatible with communication-focused teaching strategies, cognitive grammar may be a new option to bridge the gap between form- and meaning-related issues in the EFL classroom – though this hypothesis does, of course, need empirical confirmation. Within the vast area of grammar instruction, we decided to focus on tense and aspect teaching because the English tense-aspect system occupies a prominent place in the grammar sections of German school curricula for the subject of English, as well as in higher education language courses and in most other EFL programmes. It is standard practice for German universities to include basic knowledge of English tense and aspect in their criteria for student placement in language courses. Also, most textbook grammars (as well as pedagogical grammars for self-study purposes) tend to devote a great deal of attention to tense-aspect morphology and use. Nevertheless, it seems that the English tense-aspect system constitutes a major source of error even for advanced learners of English – and certainly not only for German students (cf. also Dürich 2005; Housen 2002; Kennedy 2003). As Kennedy (2003: 4) notes, “experienced teachers of English


329

have found that students from many different language backgrounds … tend to have certain problems in common when they learn English,” one of the major problem areas being verb use including tense, aspect and modality. Some of the “hurdles” for students that have been identified in the learning process of the English tense-aspect system are the following. Both Lock (1996) and Riddle (1986) have found that the use of tense morphology to refer to non-temporal concepts frequently lead to misconceptions and errors on the part of the learners. Tenses are often used inconsistently when it comes to speech events in which the “speaker’s point of view and purpose in performing [the] speech act condition the choice between the present and past tenses in actual discourse” (Riddle 1986: 267). An example would be (1a), where the present tense is used as a foregrounding device to indicate salience and to make the narrative livelier. Similarly, Tyler and Evans (2001: 63–64) notice that the use of past tense morphemes to symbolize the speaker’s view of the (non-)actuality of a situation is sometimes met with a high degree of confusion by EFL learners (see the examples of hypothetical and counterfactual situations in (1b) and (1c). (1)

a. b. c.

Erm, I’m just sitting in front of the car last night and erm … (BNC: KC2 3048) Let’s suppose your visitors were Special Branch. (BNC H8M 2665) If he had told her the truth, she would not have believed him. (BNC: HR8 858)

With regard to aspect, several studies have shown that especially EFL learners whose L1 does not have grammatical aspect at all (like German) or possesses a different aspectual system (like the Slavic languages; see Schmiedtová and Flecken this volume) often experience difficulty acquiring English aspect. In his corpus analysis of spoken and written learner English of intermediate and advanced German students, Zydatiß (1976a) found many instances of inappropriate usage of progressive and nonprogressive (or “simple”) aspect. Examples like the following demonstrate that errors in the area of aspect may still be prevalent in the interlanguage of German EFL learners at an advanced learning stage. (2)

a.

*The kitchen is the working place of the wife. Some men are cooking in the kitchen too when it is their hobby. (Zydatiß 1976a: 30)

330


b.

*Most people are thinking that the tower’s name is Big Ben but Big Ben is only the name of the bell. (Römer 2005: 172)

A study conducted by Bardovi-Harlig and Reynolds (1995) has revealed that tutored learners of English initially hesitate to apply the nonprogressive aspect to verbs which are typically used to express durational activities, such as walk or play. The non-progressive (e.g., walked, played) is thus underrepresented in the interlanguage of EFL beginners, while the progressive aspect (e.g., was walking, was playing) tends to be overused in the environment of activity verbs at first (cf. Bardovi-Harlig 1995: 158– 159). Conversely, students start with applying the progressive aspect predominantly to activities which are characterized by a dynamic and atelic nature and only later come to extend its use to more marginal activities and other imperfectivized situations (cf. Housen 2002: 166). The use of the progressive aspect in the slogan “I’m lovin’ it,” for instance, constitutes one of the more difficult cases for students, since the verb love is typically used in its stative sense (rather than with the dynamic meaning of ‘enjoy’). It is certainly dangerous to draw conclusions as to specific reasons for the learners’ manifold errors in the area of English tense and aspect on the basis of the above results. Learner errors need not necessarily be attributed to the materials used in the EFL classroom or to linguistic explanations given by the teachers. Nevertheless, an analysis of concrete textbook materials might help us find some link between potentially misleading explanations and recurrent learner errors, which would be a first step towards an “effectivization” of tense-aspect teaching. In the following section, we will therefore address a selection of issues which we deem as being improvable from a cognitive linguistic perspective. 3. Current textbooks: Relics of the past? In our analysis of materials for EFL learners we have concentrated on two textbook grammars that are widely distributed in German secondary schools, English G 2000 and Learning English – Orange Line. While more extensive pedagogical grammar books, such as the Longman and Collins grammars (cf. Alexander 2005; Sinclair 1994), usually contain some introductory information on verbs and tenses, pointing out what they consider to be the difference between time and tense, the two textbook grammars do not mention any general function of the English tense system. The tense


331

and aspect section in the textbook Orange Line for year 1 starts off with the morphosyntactic formation of the present progressive and states a rule for its use: “Wenn du ausdrücken willst, was jemand gerade macht, verwendest du die Verlaufsform des Präsens (Gegenwart)” [If you intend to express what someone is doing at speech time, you use the progressive form of the present] (Reisener and Steinbach 2003: 159). Further rules for the formation and use of the present non-progressive (termed "simple present" in the textbook) follow in a later unit. None of the grammars differentiate explicitly between tense and aspect but gradually introduce combined tense-aspect components such as the present progressive, the present non-progressive and so forth by describing their formation and providing rules for their use. At this point it already becomes clear that the focus of these grammars is on form and use, while the concepts underlying the grammatical structures are left aside. For EFL learners, however, it is crucial that semantic concepts be established before (or simultaneously with) the introduction of morphosyntactic forms to help them develop a meaningful understanding of the English tense and aspect system (cf. Bardovi-Harlig 1995: 166). Thus, an integration of the semantic-conceptual level into the pedagogical materials, alongside with morphosyntactic rules and examples of contextualized use, might be useful and even necessary for the learners’ cognitive construction of the English tense-aspect system. A second point of criticism that must be raised against the approaches taken by the two textbook grammars is the extensive memory load students are burdened with through the learning by heart of rules and exceptions. Let us consider, for example, the way in which the different tenses and aspects are introduced in the textbooks. Instead of providing separate rules for the use of the present progressive, the present non-progressive, the past progressive, the past non-progressive and so forth, it might facilitate the learning process to explain to students the conceptual difference between constructions such as in (3a) and (3b). Once the learners have understood that the progressive and non-progressive aspects are simply different ways of construing a situation, they might be able to apply this knowledge to the construal of other situations as well, largely independently of the tense that is being used. (3)

a. Eric built a snowman. b. Eric was building a snowman.

332


Although we generally appreciate the presentation of a grammatical unit in different contexts as well as the information on the various purposes for which it can be used, it is still not necessary from a CG point of view to treat the various usages of a grammatical unit as prescriptive rules that need to be memorized. CG does not regard rules as constructive statements, i.e., as prescriptive norms for the formation and use of a certain grammatical unit, but as templatic schemas which are immanent in actual language use and can therefore be derived from concrete examples by means of abstraction (cf. Langacker 2000a: 59). Provided that learners are given the opportunity to develop their own conceptual frame of reference for the target-language tense and aspect system, they might be able to analyse genuine language examples against this frame of reference themselves. Thus, they may eventually extend their interlanguage system and their use of the respective grammatical constructions accordingly. Usage-based sample texts for the different usages of a grammatical construction and concomitant conceptual explanations may therefore render rule learning negligible and help to avoid unnecessary overgeneralizations that often go along with simplified rules. One example for such an overgeneralization is the way in which the English G 2000 textbook for the third year of English at secondary school tries to illustrate the differences between the present perfect and the past non-progressive. They set up the following rule: “In Sätzen mit dem present perfect wird nie ein genauer Zeitpunkt genannt” [In sentences containing the present perfect an exact point in time is never given] (Schwarz 2002: 115). According to this formula, an utterance like (4) would be considered ungrammatical, since a definite reference time (the starting point of the time period up to the present) is included. Thus, students would presumably be confused if they encountered such a sentence. Rules like the one above would, however, become superfluous if students were familiarized with the time concepts involved in tense and aspect marking. (4)

The Dalai Lama … has been living in exile at Dharamsala, India, since 1959. (BNC: K5M 10402)

A last point that applies to pedagogical grammars in general is Tyler and Evans’ (2001: 64) critique of the discontiguous distribution of "rules" for different uses of a certain linguistic construction. They have found, for instance, that the presentation of the non-temporal uses of tense morphemes are “often scattered throughout a grammar with no attempt to tie the non-temporal use back to the basic temporal sense.” That is, the links


333

between phenomena that are conceptually related to one another are not made explicit and learners may not be provided with an explanation for the motivation behind the use of tense morphemes with non-temporal meanings. Assuming, however, that the more basic form-meaning pairings serve “as a starting point [for learners] for reconstructing the target TA [tenseaspect] system” (Housen 2002: 189) if the form-meaning relations of the L2 system have no apparent counterpart in the L1, it must be considered crucial for the learning process that less typical uses be treated as motivated extensions from the basic ones. Several linguists (e.g., Langacker 2001; Tyler and Evans 2001) have therefore argued in favour of a kind of grammar instruction whose later stages involve showing in what way the different uses are related to one another (cf. also Boers and Lindstromberg 2006: 330). In summary, we claim that an adaption of current textbook materials in the German EFL context to principles of cognitive grammar might offer a promising way of combining form- and meaning-related issues in grammar teaching. We would like to stress that we do not advocate a complete replacement of what we have termed a little loosely “traditional grammar teaching,” but want to argue for an integration of a third dimension into tense-aspect teaching, i.e., the conceptual-semantic dimension. Furthermore, it might be necessary to replace the constructed sample sentences, which are still dominant in the majority of today’s grammar- and textbooks, by more authentic, corpus-based examples. Learners should be provided with real linguistic instances which they can take as “models” (by means of abstraction) for their own language production (cf. Sinclair 1994: x). 4. Why cognitive grammar? In our view, all grammar teaching, including the pedagogical mediation of English tense and aspect, should be founded in both theoretical and descriptive linguistics, i.e., a descriptive model utilized in language programmes and pedagogical materials should be compatible with a certain linguistic theory and adapt the principles underlying the respective linguistic approach. Given the fairly limited amount of time available for language and grammar instruction at schools and other EFL institutions, it goes without saying that only partial areas of complex grammatical phenomena such as the English tense-aspect system can be dealt with. This makes it even more crucial for the linguistic framework in question to be

334


able to account for grammatical phenomena in their full complexity, not allowing for explanatory gaps or ungrammatical overgeneralizations at any stage during the EFL learning process. Furthermore, grammar-focused teaching sessions should ideally go hand in hand with the communication- and meaning-centred phases advocated in the recent literature on foreign language instruction. The approach to grammar taken by language pedagogues may not only have an impact on the learners’ interlanguage and, concomitantly, on the learners’ awareness of target language conceptualisations, but also on the ability and willingness of the learners to use grammatical constructions effectively and creatively in communicative situations. For three reasons, cognitive approaches to grammar are likely to allow for a kind of grammar instruction that is compatible with communicative teaching and learning. Firstly, cognitive grammar is based on the assumption that grammatical units, just like lexical ones, are meaningful in the sense that they possess a phonological and a conceptual pole (cf. Langacker 1991). In contrast to lexical units, however, they are used to express a rather abstract meaning, as, for instance, “being relevant for a time before the communicative present” in the case of the past tense marker. For EFL learners, this view of grammar implies that they may be better able to remember tense-aspect “rules” once they have recognized the relevant formmeaning connections (cf. Verspoor, forthcoming). Secondly, the notion of polysemy enables EFL instructors to point out the motivation behind certain meaning extensions, e.g., from the temporal use of the past tense morpheme to its non-temporal uses. And thirdly, cognitive grammar encourages inductive grammar learning due to its usagebased nature (cf. Achard 2004: 180). As already mentioned above, grammatical phenomena are seen as constructional schemas that are abstracted from real language use (cf. Achard 2004: 181; Langacker, this volume). Thus, if students are presented with different usage-based examples for a particular grammatical unit, they may be able to work out regularities in form-meaning connections themselves. 5. Theoretical prerequisites for teaching English tense and aspect 5.1. “Today was tomorrow yesterday” – On time and tense In the title of this song, the Staple Singers catch in a nutshell three essential components of a tense system. The base of any tense system is the present


335

moment of speech, or “speech time” (ST). Speech time is always the moment in time at which a communicative instance is produced. In the song title, the present speech moment is part of the time region indicated by the lexical item today. In the same vein, yesterday and tomorrow do not include the present moment of speech time, but refer to time intervals before or beyond speech time. The second component is the time span which the proposition of the utterance is set relevant for by the speaker. Let us simply call this relevant time span “relevance time” (RT). The relevant time span for the proposition today was tomorrow is past time, more precisely yesterday, as indicated by the grammatical past tense marker was and the temporal adverb yesterday. While the adverb yesterday is responsible for conveying the temporal setting of the clause, today and tomorrow are nouns in subject and predicative functions. The third component in the relation between tense and time is the time at which a situation, i.e., a process or a state, is instantiated, for the sake of simplicity called “situation time” (SiT). Relevance time can correspond exactly to situation time, as in the song title above; it can comprise only a temporal section of situation time, as in (5); or it can be completely different from situation time, as in the newspaper headline Snowstorm kills 20 people. Although the situation took place in the past, the newspaper headline wants to present it as being relevant to the present; in the further text of such news stories, the past tense is used. We find constructions like these quite frequently, not only in journalist articles, but also in lively narratives. Clearly then, tense does not, in the first place, locate the processes or states themselves on the time axis, but it rather allows the speaker to select a time span that is relevant for what he/she says. Let us take a closer look at (5) to make this clear: (5)

Emily and Charlotte came home to see her, but she was dead … (BNC: FNY 448)

As Klein (1994: 22) argues, such sentences do not simply situate the fact of someone being dead in the past, because the situation still applies at speech time. Except in science fiction, the victim is still dead at the moment of reporting this past state. So the past tense in was dead does not primarily fulfil the function of locating the state of being dead in past time, but it rather expresses that the relevant time span for what the speaker is saying lies somewhere in past time. In contrast, situation time, i.e., the state

336


of the victim’s being dead, comprises both past and present time. Thus, relevance time only includes a part of situation time: the relevant time stretch, which lies in the past. As Klein remarks, we could not use the present tense in a context like (5) above, i.e., she is dead, even though the validity of the utterance would be given. From a cognitive linguistic point of view, this understanding of tense can be optimally combined with insights both from Fauconnier (1994, 1997) as laid down in his “mental space theory” and from Langacker (1987, 1991) in his “model of evolving reality”, which is a synthesis of a central part of his cognitive grammar model. Let us take Fauconnier first. The basic idea of Fauconnier’s mental space model is that when discourse participants communicate with each other, they mentally construct small conceptual packets, called “mental spaces” (cf. Fauconnier 1998: 252). These mental spaces contain elements, i.e., conceptual information on the things or persons we talk about, as well as information on the reality status of the situation and its relevance time. The status of things or persons in the discourse is expressed by the determiner system and is known as reference; the status of situations and their relevance time is expressed by the tense, aspect and modality system. A new mental space can be set up or opened and returned back to at any moment in the discourse. Most mental spaces are implicitly opened, but it can also be done in an explicit manner. Explicit space builders can be time adverbials, such as the adverb yesterday in the above song title Today was tomorrow yesterday or the expression in 1999, or the first clause in the sequence of two or more in a narrative context such as (5). They all denote the temporal setting serving as the background for a situation. Explicit space builders for the reality status of a situation are expressions such as Possibly or Gavin thinks that … Other major space builders are grammatical means, such as the marking of tense and aspect, to which we can add modality (cf. Fauconnier 1994: 17). Grammatical tenses and aspects, and their combinations, serve to indicate relative relations between spaces and crucially, to keep track of the discourse ‘position’ of the participants – which space is in focus …, which one serves as base and what shifts are taking place (Fauconnier 1994: xi, emphasis in original).

A fundamental distinction is made between base space and all other mental spaces. Base space is the situation at speech time, the here and now, in which the speaker is the deictic centre. It serves as an anchor for ex-


337

pressing both the reality status and the temporal relevance of any situation that is being communicated. The reality status of a situation can be indicated either by means of tense forms or by modal verbs. Situations set relevant for base space are referred to by the use of the present tense. From base space, mental spaces can be opened for situations whose relevance time is anterior to speech time and expressed by the past tense, or else posterior to speech time and expressed by the will form. However, the socalled past tense not only expresses temporal relevance, it may equally express epistemic relevance, as we shall see below. Let us first focus on the reality status and the expression of some instances. In an assertion, the reality of the state of affairs is taken for granted and therefore not explicitly marked. But any other reality status such as hypotheticality or counterfactuality needs to be marked explicitly. A hypothetical space, for instance, can be created by means of the space builder if in combination with a set of tense forms that indicate epistemic as well as temporal relevance. In terms of their reality status, the situations are located either within the reality of base space (6a), or close to, but outside it (6b), or at a great epistemic distance from the reality of base space as in (6c) and (6d). Simultaneously, different relevance times are involved as the adverbs right now, straight away, then in (6a), (6b) and (6c) respectively, illustrate. (6)

a. b. c. d.

If he is here, as you say, let’s ask him right now. If he comes, let’s ask him straight away. If he came, we would ask him then. If he had come, we could have asked him.

The distinction between base space in (6a), which allows the present time adverb right now, and hypothetical space in (6b) and (6c), which do not allow that adverb, clearly confirms that temporal spaces and others such as hypothetical spaces are expressed by the same set of tenses in English. Both uses can best be understood in their relation to the double function of base space, i.e., its reality status function and its relevance time function. Example (6d) illustrates how the interaction between temporal and epistemic relevance is expressed grammatically: Since the situation is meant to be located in temporal as well as epistemic distance from the base, what is traditionally called a “double backshift” takes place grammatically, as can be seen by the use of the past perfect in had come. A speaker thus has many grammatical options at hand to communicate his/her thoughts, while from the hearer’s perspective, tense and aspect

338


marking “allows us to reconstruct the reality and time spaces set up from base space” (Fauconnier 1997: 78) and the relations between them. The choice of tense form or modal verb, or any conventionalized combination of them, shows us that reality status and relevance time go hand in hand. Summarizing, we can say that tense in English and most other tensed languages is a matter of reality status and temporal relevance vis à vis the “external" reality and temporal “location” of speech time, information on both of which is kept available in mental spaces. In contrast to this, the function of aspect in English is rather a matter of the internal temporal constitution of situations (cf. Comrie 1976: 3). Thus, aspect is a question of alternative construal, whereby the speaker often, though not always, has the choice between two contrasting construals, which are grammatically realized by different aspectual forms, i.e., the progressive and the nonprogressive form. Although English only has one clearly marked aspectual form, the progressive construction be + Vinf + ing, as in He is having an affair (BNC: CE2 2671), we adopt the view held by Radden and Dirven (2007: 177–196) that there is a cross-wise aspectual contrast in English, but related to either processes or states (see Section 5.2). Let us now contrast some of the English tense forms, first the present non-progressive and the past non-progressive with states. What is the relation with respect to speech time in (7a) and (7b)? (7)

a. b.

Many Aborigines live in the outback. Most Aborigines lived in the outback.

The predication of the two utterances is the same, namely . In (7a), however, the predication is asserted to hold in a time span including speech time, whereas the relevance time for (7b) is past time. That is, the second situation is portrayed as having relevance for a time span before speech time, the past tense marker symbolizing temporal distance from speech time. But this does not imply whether at speech time the situation still obtains or is finished, as may appear from the possibility to combine both sentences, e.g., Many Aborigines lived in the outback before the European settlers came and still live there today. So what the speaker in (7b) says is that the predication is a lasting state that obtained in the past, no more and no less. Like states, processes can equally be construed as indefinitely lasting. If they are repeated over a longer period of time, including speech time, processes tend to be construed as “habitual states,” as in (9a) below. Still, the


339

present tense is used because the proposition of the utterance is set relevant to present time. (8)

a. b. c. d.

Verena works out in the gym three times a week. Right now she is working out in the gym. During the holidays she is working out in the gym. Workout is good for your health.

While in (8a), the situation encoded by the present tense is a habitual state, the situation in (8b) is construed as being in progress at speech time. Due to the temporal adverb right now as an unambiguous space builder for speech time, it becomes clear that the working out in the gym in (8b) is temporally coincident with the base, whereas the process referred to in sentence (8c) need not be, although both processes are encoded by the present progressive. (8c) covers a habitual state which is construed as not lasting indefinitely (as habits usually tend to), but as temporary, i.e., as having implicit boundaries. The person in (8c) is only working out in the gym for a certain limited period of time, i.e., during her holidays, but she might still be a regular visitor of the gym during this particular time span, e.g., go there three times a week. Moreover, she is not necessarily lifting weigths at the moment of speaking, like the person in (8b). Neither in (8a) nor in (8c) situation time and relevance time need to coincide exactly. Though different, something similar applies to (8d), where the predication of the sentence is presented as an ever-lasting truth that is relevant for the present, although its validity is assumed to hold over a much longer time span, including the present moment of speech. As already silently assumed so far, we want to claim explicitly, keeping in line with Langacker (1991: 249), that English has only two, and not three, tenses, i.e., present tense and past tense.2 On a conceptual level, it can be argued that only the present and past tenses are used to signal relevance time for situations that are construed as having reality status, i.e., situations located in factual reality (comprising immediate reality and known past reality; see Figure 1 below). Situations with (potential) future situation time are variously construed by the utterances in (9). (9)

a. b.

The next train to Garvie leaves at midnight. (BNC: CEH 1317) This year, there are going to be hundreds of lamb births on his farm.

340


c. d.

The sad thing is that all the crippled lambs will be discarded like waste. Some lambs may very well be born dead.

These examples instantiate the different types of reality status, sketched in Figure 1. irreality (counterfactual space) potential past reality (potentiality space) known past reality

potential present reality

potential future reality

immediate reality ST

projected future reality

tense forms

going to form epistemic will

modal verbs

modal verbs

tense forms

Figure 1. Model of evolving reality (adapted from Langacker 1991)

We find factual reality in (9a), projected reality in (9b) and (9c) and potentiality in (9d). The difference between projected reality and potentiality/modality is substantial, since in each case the reality status depends on different sources of knowledge or evidence. Projected reality results from the evolving nature of the world and can therefore be construed by the prospective form going to (9b) or by means of predictive will (9c). The will form places relevance time in the future and is based on the evolving nature of things; the going to form has present time as relevance time and is based on present evidence interpreted on the basis of past experience of the speaker. Hence, the going to form expresses greater certainty than the will form. Modality, as in (9d), is fully different from projected reality in that it depends entirely on the speaker’s subjective assessment. The speaker can only make a reasonable guess about the future occurrence of dead born


341

lambs, though the probability of the instantiation of the situation is stressed by the adverb very well. The difference between future situations expressed by the present tense as in (10a) and by going to or will as in (9b) and (9c) sometimes appears to be exiguous. Nevertheless, the economic nature of grammar lets us expect a conceptual and functional difference between the two forms, which – in our opinion – can be attested. While the situation in (9a) is construed as already being part of reality, on the basis of its being an instatiation of a mental timetable or virtual schedule at speech time (cf. Langacker 2001), the future reality of the situations in (9b) and (9c) is predicted on the basis of present evidence. The reality status of the situations in (9b) and (9c) is not yet given, although it is almost certain to be achieved. So far we have dealt with the temporal use of tense morphemes. It was stressed that temporal use does not primarily mean locating situations in time, but rather indicating what time span is relevant for what we are saying. So the temporal use of tense and other morphemes is already much more subtle than what is usually understood by “temporal meaning.” This may also make it easier to see that the same morphological forms can also be employed to encode non-temporal meaning. Just as we can open, starting from base space, further reality spaces with past, present or future relevance time, we can also open up other mental spaces, as, for instance, potentiality space. One kind of potentiality space is modality space; others are hypotheticality space or counterfactuality space, as in (10a) and (10b). There can also be interactive spaces such as narrative space (10c) or politeness space (10d). (10)

a. b. c. d.

If only he knew! (BNC: AEB 3109) If he had told her the truth, she would not have believed him. (BNC: HR8 858) Erm, I’m just sitting in front of the car last night and erm … (BNC: KC2 3048) I wanted to ask you something. (BNC: HTN 2787)

The two spaces in (10a) and (10b) are counterfactual ones, i.e., spaces whose reality status is in contradiction with base space: In (10a) the person that is meant does not know, and in (10b) the male person has not told the truth. Although situation time in (10a) is present time, the past tense is used to express epistemic distance vis-à-vis the base. This example shows that past tense morphology cannot only be used to indicate relevance time that is located anterior to (and thus at a distance from) the base, but also to

342


express epistemic distance from the base. The same applies to (10b): Since situation time is past and, at the same time, the situation is construed as a counterfactual one, the past perfect is used, indicating temporal as well as epistemic distance from the base. In fact, it is because base space is the anchor for two functions, i.e., speech time and reality status, that it can serve two different roles: temporal and/or epistemic ones. In (10c) the narrative space allows a conflict between situation time, which is past time, and relevance time, which is presented as if it is present time. In (10d) we see the reverse: Situation time is present time (the speaker asks the question now), but for reasons of politeness the speaker presents its relevance time as non-present time, which has the effect that the request seems to be less urgent such that the hearer may feel less facethreatened. In summary, tense morphology cannot only be used to symbolize temporal relevance, but it can equally be used to express epistemic relevance, salience and attenuation in certain contexts. Both the temporal and nontemporal uses of tense forms can therefore be ascribed to a common conceptual basis, the proximal/distal (or immediacy/non-immediacy) schema (cf. Langacker 1991: 249). That is, the present tense is always used to express proximity/immediacy – be it temporal, epistemic or “narrative” proximity – while the past tense always indicates distance/non-immediacy, either with respect to relevance time, reality status or social commitment. 5.2. Progress isn’t simple: progressive vs. non-progressive aspect While so far, we have taken an external view on situations, this section is concerned with the internal temporal constitution of situations. We suggest a theory of aspect that takes into account both lexical aspect (Aktionsart) and grammatical aspect, assuming that lexis and grammar interact with each other on a conceptual level during communication. What distinguishes English from non-aspectual languages, such as German or the Scandinavian languages, is the fact that it does not only insist “on marking every finite verb group for ... tense, whether or not the time orientation would be clear without it” (Lock 1996: 163), but additionally requires the marking of grammatical aspect. Every time we produce a finite clause, we must decide whether to use the present tense or the past tense, the non-progressive aspect or the progressive aspect, the non-perfect form or the perfect form.3 These various grammatical units are always combined and interact with each other.


343

However, in contrast to other European languages with formal aspect, especially the Slavic languages with their markings for the imperfective and perfective construals of situations (see Schmiedtová and Flecken, this volume), English only has a clearly marked progressive aspect, but no special marker for the contrasting element in the pair. There is no agreement about the existence of a non-progressive aspect in English. Still, we claim, in keeping with Radden and Dirven (2007: 177–196) that there is a crosswise aspectual contrast in English, related to either processes or states. In Section 5.1 we introduced the notion of situation, which is meant to encompass both states that things (i.e., persons, objects etc.) are in and processes that things undertake or undergo. As Klein (1995: 146) says, it is useful to distinguish between situations which are seen as limited in time and others which are not. Thus, situations can be classified into two broad categories according to their inherent temporal structure: inherently unbounded situations, such as , and potentially bounded situations, which can have implicit boundaries, such as <WANDER ABOUT THE MUSEUM>, or explicit boundaries, such as . As these examples demonstrate, verbs do frequently not stand alone in a communicative instance but interact with their complements, which have an impact on the type of situation that is evoked. Thus, it is crucial to consider the entire verbal predicate for the purpose of categorisation. Moreover, we would like to stress that we are not referring to real world situations per se, but to the linguistic expression of situations, which is at least to some part dependent on linguistic convention (cf. Xiao and McEnery 2004: 328). Inherently unbounded situations are internally homogeneous and insusceptible to change. By contrast, potentially bounded situations are internally heterogeneous and susceptible to change, and they are expected to come to an end at some point. Grammatical aspect, i.e., the use of either the non-progressive or the progressive form, interacts with lexical aspect in that it offers the speaker a means to construe an idealized situation in different ways. Depending on the type of situation, i.e., whether the situation is inherently bounded or unbounded, and depending on whether the situation is construed as a single situation or as a repeated situation, grammatical aspect can have different conceptual effects. Let us have a look at the following sentence pair: (11)

a. Eric built a snowman. b. Eric was building a snowman.

344


The situation , which is realized differently in the contexts of (11a) and (11b) as far as grammatical aspect is concerned, is per se bounded because it involves change through time and because its beginning (initial boundary) and endpoint (final boundary) are inherent. At some point in time, Eric starts to roll the snowballs that are then stacked up and decorated with a hat, a scarf, a broom and so on to become the finished snowman in the end. While the non-progressive aspect in (11a) expresses that the situation is viewed in its entirety and that both its beginning and endpoint fall within the scope of predication, the progressive aspect in (11b) has the effect of unbounding the situation. Schmiedtová and Flecken (this volume) call this process “defocusing of boundaries,” i.e., both the initial and the final boundary of the situation are excluded. It is important to note, though, that the higher order schematic conception of a bounded situation is still evoked by the progressive aspect: it functions as a base which provides the possibility of defocusing the boundaries (cf. Langacker 2000b: 227; Schmiedtová and Flecken, this volume). While the situation in (11a) is perceived as being complete in itself and is thus not susceptible to change anymore, the situation in (11b) is construed as being in progress and is thus (at least potentially) susceptible to change (cf. Williams 2002: 88). The same applies to potentially bounded situations with implicit boundaries, such as in (12a) and (12b). The only difference between the snowman example and the museum example is that the boundaries of the situation are explicit due to the internal structure of the situation, while the boundaries of <WANDER ABOUT THE MUSEUM> stay implicit. A museum visit starts at some point in time and ends at some point, but there is no “change in state” involved, like in the snowman example where we have an initial state (no snowman) and a final state (the finished snowman) (cf. also Klein 1994, 1995). (12)

a. We wandered about the museum at night. b. We were wandering about the museum at night.

Let us now have a look at a different situation type, i.e., inherently unbounded situations. When used with the non-progressive aspect, situations such as are construed as lasting states that are not susceptible to change. What effect, then, does the use of the progressive form have with this type of situation? In contrast to (11b) and (12b), where the progressive has the effect of unbounding a situation, the use of the progressive in (13b) expresses that the situation is not seen as a lasting


345

one, but as a temporary state. Thus, it is construed as having implicit boundaries and as being susceptible to change. We can therefore speak of a cross-wise aspectual contrast (cf. Radden and Dirven 2007), which fundamentally differs from Langacker’s views. (13)

a. Many Aborigines live in the outback. b. Many Aborigines are living in the outback.

To sum up, we can say that non-progressive and progressive aspect have an effect on the construal of a situation, i.e, they indicate in what way the internal constitution of a situation is viewed. With inherently bounded situations, the non-progressive aspect expresses that the situation is seen as complete in itself and thus insusceptible to change, as in (11a) and (12a) (cf. Williams 2002: 89). The progressive aspect, by contrast, has the effect of defocusing the boundaries of the situation. That is, it “zooms in” on the situation (cf. Langacker 2000b: 228), viewing the situation as ongoing and potentially susceptible to internal change, as in (11b) and (12b). When used with inherently unbounded situations, the non-progressive construes the situation as continuing indefinitely without there being any expectation of change, as in (13a) (cf. Williams 2002: 89). The progressive, by contrast, imposes implicit temporal boundaries on the situation, focusing on its temporariness. The situation is seen as potentially susceptible to external change since it is construed as a temporary state that might sooner or later come to an end; thus, the focus lies on the possible transition to adjacent states. Both the notions of (un)boundedness and susceptibility to change have great explanatory value and may lend themselves for application in pedagogical contexts, as we will see in the last sections of this paper. 6. Applying cognitive grammar in the EFL classroom 6.1. Getting started: Preliminary considerations Let us finally consider the crucial question of how grammar instruction based on conceptual-semantic and usage-centered approaches to grammar could proceed, and how it may assist learners in making the form-meaning connections relevant for their interlanguage development. Since our reflections are still at an early stage, the following suggestions for strategies and teaching/learning materials have not yet been em-

346


pirically tested. They are aimed at intermediate level students who are enrolled in the subject of English studies, i.e., who study English as a major subject within different degree programmes. We expect that an empirical study by one of the authors, scheduled for the summer of 2008, will provide us with results as to the efficiency of cognitive approaches in tenseaspect teaching. It is especially keeping track of the speaker’s or hearer’s discourse position towards the reality status and the relevance time of the situations evoked that may cause difficulty for EFL learners because they can only rely on an L2 interim language growing towards an approximative native language system. As soon as the focus shifts to longer discourses that do not contain notable amounts of explicit space builders, students tend to lose track of the time reference shifts and the epistemic shifts within the unfolding discourse. Our analysis of essays written by German EFL students has revealed that even advanced learners (i) tend to produce unnatural switches of tenses in longer texts and (ii) have immense problems noticing and expressing the temporal relations between the different spaces. Thus, a mental “space building plan” may help them recognize the relevance time path of the space configurations (which is reflected by the grammatical tenses) as well as the temporal connections between the different spaces (which are indicated by the perfect). A particular area of difficulty for German students is made up of hypothetical and counterfactual situations, which are expressed differently in English and in German. While we have seen that the notions of hypotheticality and counterfactuality are signalled by means of tense morphemes in English, German uses the subjunctive to symbolize these epistemic concepts.4 Therefore, a potential risk of negative “transfer” from the students’ L1 exists, which may result in grammatically incorrect constructions such as the following: (14)

*If I would tell my parents that I have quit college they would not be too pleased about it. [from an intermediate learner’s essay]

Second language acquisition (SLA) research has shown that interference of the L1 at least partly accounts for certain developmental patterns in the learners’ interlanguages. Within the recent cognitive approaches to SLA, there has been widespread acknowledgement that “learners draw on their L1 in forming interlanguage hypotheses … [and] do not construct rules in a vacuum” (Ellis 2002: 52). Therefore, a contrastive analysis of the


347

English and German tense-aspect systems may provide us with a rough idea of the conceptual frameworks German EFL learners have to deal with. From a contrastive point of view, not only hypotheticality and counterfactuality, but also the English progressive/non-progressive distinction is very likely to result in learning difficulty, since the German language possesses no equivalent grammaticalized forms for expressing the internal temporal constitution of a situation. To stress the ongoing progression of a situation, German solely adheres to lexical means. Particles and adverbials such as gerade ‘right now’ or momentan ‘at the moment,’ as well as the idiomatic expression dabei sein, etwas zu tun ‘be in the middle of doing something’ are common ways of indicating the defocusing of boundaries of an inherently bounded situation. In some dialectal contexts, it is also acceptable to use the construction be + Prep + Vinf, as in Ich war am Schreiben ‘I was in the process of writing,’ commonly known as the Rhenish progressive. Thus, students need not only become aware of the fact that a certain notion such as aspect might manifest differently in languages. They also need to realize that in English, the speaker has to make a choice between the progressive and non-progressive aspect every time he/she wants to refer to a situation – unlike in German, where even the lexical marking of aspect is not necessarily required, provided that the pragmatic context allows the recipient to infer from it the internal structure of the situation (cf. also Niemeier 2005). However, contrasts between the L1 and the target language are not the only factors influencing the learners’ interlanguage development (cf. Ellis 2002: 51–54). Although the non-temporal concepts of foregrounding (salience) and politeness (attenuation) are also expressed by means of tense morphemes in German, learners still tend to encounter difficulty with these usages. SLA research has come up with two possible explanations for such observations. According to Kellerman (1983: 114), learners have perceptions about the transferability of certain L1 structures. “If a feature is perceived as infrequent, […] semantically or structurally opaque, or in any other way exceptional, what we could in other words call ‘psycholinguistically marked,’ then its transferability will be inversely proportional to its degree of markedness” (Kellerman 1983: 114). Thus, students might hesitate to make the relevant “cross linguistic tie-ups” (Kellerman 1983: 114) in the case of the concepts of salience and attenuation because they may perceive these functions of tense morphology as rather secondary (compared to the time-reference function), and therefore as less likely to be transferable.

348


The second explanation concerns the highly polysemous nature and the concomitant degree of “ambiguity” of tense morphemes in English, which might further interfere with the learners’ interlanguage by complicating the emergence of the necessary form-meaning associations. Since learners are usually taught the temporal interpretation of tense morphemes first, these primary form-meaning pairings may affect the learners’ readiness for further form-meaning associations (cf. Ellis 2006: 102–106). We would therefore strongly argue in favour of an approach to tense and aspect that highlights the conceptual commonalities underlying the different form-meaning mappings. 6.2. Keeping track A construct that, in adapted form, could have considerable pedagogical value to the teaching of the temporal and non-temporal concepts underlying tense is, as already discussed in Section 5.1, Fauconnier’s (1994, 1997) “mental space model.” This model not only possesses explanatory merit for the sequencing of tenses in discourses, but it also has the potential to provide learners with a better insight into the interaction between reality status and relevance time functions. A further advantage of this construct is that it lends itself quite well for visualization. A simplified time model could be drawn on the blackboard as a visual basis for illustrating the relevance time of and temporal relations between different spaces. Students would then be presented with a sample discourse and with different cards displaying images of the situations contained in the discourse. It would be the students’ task to locate the situations in temporal “space.” In a second step, they would be asked to indicate the temporal path of the space configurations as well as the temporal connections between the spaces by means of arrows. In a last step, the connection between form and (conceptual) meaning should be made more explicit by discussing which morphological forms are responsible for indicating the respective relevance time path. At a later stage, the epistemic dimension of reality status could be added to the model. Students could also be invited to copy a sample model into their exercise books, which they would then be allowed to use in essays, exercises and tests as a “mental pillar.” Our hypothesis is that students who have previously encountered serious problems with tense distribution in longer discourses may be able to develop a mental “space building plan.” This “plan” may enable them to keep track of the discourse position of the


349

speaker/hearer and to follow temporal and epistemic shifts within the discourse. Thus, the likelihood for them to “lose track,” i.e., to produce unnecessary and non-native like tense shifts in essays or elicitation exercises, may be reduced. 6.3. Making progress As for grammatical aspect, we have seen that it has proven to be a major source of error and confusion for German students. Learners are often not aware of the concepts underlying the progressive and non-progressive forms, but tend to believe that omitting the progressive marker is tantamount to using a neutral form. Thus, they need to notice that also the nonprogressive aspect, often misleadingly called the “simple” aspect, implies a certain mode of construal (cf. Niemeier 2005: 19). To give students a better grasp of the rather abstract concept of grammatical aspect, one may introduce the notions of (un)boundedness and susceptibility to change, which were elucidated in Section 5.2. Learners need to become aware of the differences between inherently bounded situations and inherently unbounded situations. The concept of (un)boundedness could be visualized by means of a box that is either closed (with bounded situations) or opened to the left and the right (with unbounded situations). Bounded situations should be depicted as having a starting point and an endpoint and as consisting of several developmental segments, as illustrated in Figure 2a for the situation Eric built a snowman. By contrast, situations construed as unbounded should be characterized by internal homogeneity and by an exclusion of boundaries.

Figure 2a. Inherently bounded situation

350


Figure 2b. Defocusing of boundaries

Once the learners have understood the difference between (inherent) boundedness and unboundedness, a further step would be to move on to grammatical aspect. As pointed out in Section 5.2, the effect of the progressive and non-progressive forms can differ, depending on the inherent properties of a situation. In the following, we would like to demonstrate how EFL teachers could illustrate to students the effect the progressive has with inherently bounded situations, an effect which can be best described as defocusing boundaries or zooming in on a situation. One possibility of introducing the progressive aspect with inherently bounded situations would be to visualize the construct of zooming in on a situation by use of animated graphics: On the first picture, a bounded situation is displayed, for instance the process of Eric building a snowman from its beginning to its end (see Figure 2a). By and by, the “lens” zooms in on the situation and comes to rest with the middle segment of the process, where Eric is rolling the snowballs to grow them thicker (see Figure 2b). Thus, only a segment of the situation is still viewable and both boundaries have moved out of focus. Another suggestion has been made by Niemeier (2005), which she calls the “keyhole method.” A situation viewed in its entirety, such as , is illustrated sketchily on a transparency. Then a jig in a form of a keyhole is laid upon the transparency, focusing the view on a segment of the process and leaving its boundaries aside. The students could then be asked to describe further “keyhole pictures,” i.e., pictures depicting processes that are viewed through a keyhole, in order to practise the progressive form. A card showing a keyhole could also be used for negative feedback. Every time a learner forgets to differentiate between progressive and non-progressive aspect, the teacher or a peer could raise the card to remind the student (and the class) of the boundedness/unboundedness distinction, thus giving the learners an opportunity for self-repair (cf. Niemeier 2005: 23–26). Both methods may help students to realize that situations do not only have inherent aspect, but that a situation can be construed differently by means of grammatical forms. Thus, they may be able to establish their own form-meaning connections with the aid of visual and explanatory materials based on cognitive approaches to grammar.

7. Conclusion In this paper we have put forward the hypothesis that an integration of cognitive principles into tense-aspect teaching/learning materials will help students to develop a more meaningful understanding of these grammatical constructions. We have seen that tense and aspect are two major sources of error for German EFL learners, partly due to interference of their L1, partly due to the complexity of the English tense-aspect system. Sections 5 and 6 have shown that a synthesis of various cognitive approaches to tense and aspect a) offers a coherent account of the concepts underlying these grammatical constructions, and b) lends itself for application in EFL contexts due to its explanatory and illustrative merit. We have made some concrete suggestions as to how EFL teachers could proceed in a tense-aspect grammar session and which visual aids may be useful in class. An empiricalvalidation of our hypothesis still remains a desideratum for further research.

352


Acknowledgement We wish to thank René Dirven for his support and his valuable comments on a first draft of this paper. All remaining flaws are of course entirely our own responsibility.

Notes 1.

2. 3.

4.

In Germany, textbook grammars are relatively short grammar sections contained in the back of an EFL textbook or, alternatively, small booklets complementing the textbook. Other Cognitive linguists categorize the epistemic will as a tense form, i.e., future tense (cf. Radden and Dirven 2007). For reasons of space limitations, the perfect/non-perfect distinction could not be dealt with in detail in this paper, although, of course, this construction interacts with tense and aspect as well. The subjunctive is also used to indicate indirect speech in German, whereas English again applies tense morphology for this purpose. For a purely temporal interpretation of the tense distribution in indirect speech in English see Langacker (1991: 253–256).

References Achard, Michel 2004 Grammatical instruction in the Natural Approach: A Cognitive Grammar view. In Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching, Michel Achard and Susanne Niemeier (eds.), 165–194. Berlin: Mouton de Gruyter. Alexander, Louis G. 2005 Longman English Grammar. London: Longman. Bardovi-Harlig, Kathleen 1995 The interaction of pedagogy and natural sequences in the acquisition of tense and aspect. In Second Language Acquisition Theory and Pedagogy, Fred R. Eckman, Diane Highland, Peter W. Lee, Jean Mileham, and Rita R. Weber (eds.), 151–168. Mahwah, New Jersey: Lawrence Erlbaum. Bardovi-Harlig, Kathleen, and Theodora Bofman 1989 Attainment of syntactic and morphological accuracy by advanced language learners. Studies in Second Language Acquisition 11: 17– 34.


353

Bardovi-Harlig, Kathleen, and Dudley W. Reynolds 1995 The role of lexical aspect in the acquisition of tense and aspect. TESOL Quarterly 29: 107–131. Boers, Frank, and Seth Lindstromberg 2006 Cognitive linguistic applications in second or foreign language instruction: Rationale, proposals, and evaluation. In Cognitive Linguistics: Current Applications and Future Perspectives, Gitte Kristiansen, Michel Achard, René Dirven, and Francisco J. Ruiz de Mendoza Ibáñez (eds.), 305–358. Berlin/New York: Mouton de Gruyter. Comrie, Bernard 1976 Aspect: An Introduction to the Study of Verbal Aspect and Related Problems. Cambridge: Cambridge University Press. Dürich, Kristiane 2005 The acquisition of the English tense and aspect system by German adult learners. Magisterarbeit [MA thesis], Technische Universität Chemnitz. Ellis, Nick C. 2006 Cognitive perspectives on SLA: The Associative-Cognitive CREED. AILA Review 19 (1): 100–121. Ellis, Rod 2002 Second Language Acquisition. Oxford: Oxford University Press. Fauconnier, Gilles 1994 Mental Spaces: Aspects of Meaning Construction in Natural Language. Cambridge: Cambridge University Press. 1997 Mappings in Thought and Language. Cambridge: Cambridge University Press. 1998 Mental spaces, language modalities, and conceptual integration. In The New Psychology of Language: Cognitive and Functional Approaches to Language Structure, Michael Tomasello (ed.), 251–280. Mahwah, New Jersey/London: Lawrence Erlbaum. Hinkel, Eli, and Sandra Fotos 2002 From theory to practice: A teacher’s view. In New Perspectives on Grammar Teaching in Second Language Classrooms, Eli Hinkel and Sandra Fotos (eds.), 1–13. Mahwah, New Jersey/London: Lawrence Erlbaum. Housen, Alex 2002 The development of tense-aspect in English as a second language and the variable influence of inherent aspect. In The L2 Acquisition of Tense-Aspect Morphology, Rafael Salaberry and Yasuhiro Shirai (eds.), 155–198. Amsterdam/Philadelphia: Benjamins.

354


Kellerman, Eric 1983 Now you see it, now you don’t. In Language Transfer in Language Learning, Susan M. Gass and Larry Selinker (eds.), 112–134. Rowley, Mass.: Newbury House Publishers. Kennedy, Graeme 2003 Structure and Meaning in English: A Guide for Teachers. Harlow: Pearson Longman. Klein, Wolfgang 1994 Time in Language. London and New York: Routledge. 1995 A simplest analysis of the English tense-aspect system. In Anglistentag 1994 Graz: Proceedings of the Conference of the German Association of University Teachers of English, Wolfgang Riehle and Hugo Keiper (eds.), 139–152. Tübingen: Niemeyer. Langacker, Ronald W. 1987 Foundations of Cognitive Grammar, Vol. I: Theoretical Prerequisites. Stanford, California: Stanford University Press. 1991 Foundations of Cognitive Grammar, Vol. II: Descriptive Application. Stanford, California: Stanford University Press. 2000a A dynamic usage-based model. In Usage-Based Models of Language, Michael Barlow and Suzanne Kemmer (eds.), 1–64. Stanford, California: Center for the Study of Language and Information Publications. 2000b Grammar and Conceptualization. Berlin/New York: Mouton de Gruyter. 2001 Cognitive linguistics, language pedagogy, and the English present tense. In Applied Cognitive Linguistics I: Theory and Language Acquisition, Martin Pütz, Susanne Niemeier, and René Dirven (eds.), 3– 40. Berlin/New York: Mouton de Gruyter. Lock, Graham 1996 Functional English Grammar: An Introduction for Second Language Teachers. Cambridge: Cambridge University Press. Niemeier, Susanne 2005 Boundedness/Unboundedness: Blick durch das Schlüsselloch. Angewandte Kognitive Linguistik für den Englischunterricht [Boundedness/Unboundedness: The keyhole method. Applied Cognitive Linguistics for TEFL]. Zeitschrift für Angewandte Linguistik (ZfAL) 43: 3–32. Radden, Günther, and René Dirven 2007 Cognitive English Grammar: The Simple Sentence. (Cognitive Linguistics in Practice 2) Amsterdam/Philadelphia: Benjamins. Reisener, Helmut, and Gudrun Steinbach (eds.) 2003 Learning English – Orange Line New 1. Stuttgart: Klett.


355

Riddle, Elizabeth 1986 The meaning and discourse function of the past tense in English. TESOL Quarterly 20: 267–286. Römer, Ute 2005 Progressives, Patterns, Pedagogy: A Corpus-Driven Approach to English Progressive Forms, Functions, Contexts and Didactics. Amsterdam/Philadelphia: John Benjamins. Schwarz, Hellmut (ed.) 2001 English G 2000. Band B1, für das 5. Schuljahr an Realschulen [Vol. B1, for the 5th grade at lower secondary schools]. Berlin: Cornelsen. 2002 English G 2000. Band B3, für das 7. Schuljahr an Realschulen [Vol. B3, for the 7th grade at lower secondary schools]. Berlin: Cornelsen. Sinclair, John (ed.) 1994 Collins Cobuild English Grammar. London: Harper Collins. Tyler, Andrea, and Vyvyan Evans 2001 The relation between experience, conceptual structure and meaning: Non-temporal uses of tense and language teaching. In Applied Cognitive Linguistics I: Theory and Language Acquisition, Martin Pütz, Susanne Niemeier, and René Dirven (eds.), 63–105. Berlin/New York: Mouton de Gruyter. Verspoor, Marjolijn subm. Cognitive linguistics and its applications to second language teaching. In Encyclopedia of Language and Education. Vol. 6: Knowledge About Language. Berlin/New York: Springer. Williams, Christopher 2002 Non-Progressive and Progressive Aspect in English. Fasano: Schena editore. Xiao, Zhonghua, and Anthony McEnery 2004 A corpus-based two-level model of situation aspect. Journal of Linguistics 40: 325–363. Zydatiß, Wolfgang 1976a Tempus und Aspekt im Englischunterricht [Tense and Aspect in English Teaching]. Kronberg: Scriptor. 1976b Learning problem expanded form: A performance analysis. IRAL 14: 351–371.

Aspectual concepts across languages: Some considerations for second language learning Barbara Schmiedtová and Monique Flecken

Abstract In this paper, we focus on some terminological issues concerning the notion of aspect. We address the notions of grammatical aspect vs. Aktionsart, perfectivity vs. telicity, and imperfectivity vs. progressivity. We observe that these terms are often mixed up in the literature, which leads to some fundamental misconceptions in the theoretical description of different aspectual systems as well as in L1 and L2 acquisition studies. The descriptive approach we follow is strictly empirical and based on spoken production data. For our cross-linguistic comparisons, we draw upon data from native speakers of Czech, English, Dutch, German, and Russian. The theoretical framework of the paper is based on the idea that aspectual markers are not merely grammatical categories with a particular function, but more importantly they denote underlying cognitive concepts. These grammaticalized concepts determine native speakers’ preferences in event construal, are language-specific (L1-based), and play a decisive role in second language learning. In order to deal with the difficulties arising in L2 learning, it is crucial to attempt to avoid terminological confusion. We think that this can be achieved by adopting a more conceptual and empirical approach to the analysis of aspect. Keywords: grammatical aspect; Aktionsart; telicity; aspect terminology; perfectivity; imperfectivity; progressivity; psycholinguistic reality; empirical research; grammaticalization; conceptualization; language typology; second language learning; false friends; Czech; Russian; Dutch; English; German

1. Introduction In this paper we will attempt to show and discuss some of the complexities in terminology that regularly come up in theoretical analyses of aspect in cross-linguistic research. Examples of terms that are often confused and that we focus on are grammatical aspect vs. Aktionsart, telic vs. perfective, and imperfective vs. progressive. In our view, this terminological confu-

358

Barbara Schmiedtová and Monique Flecken

sion often leads to crucial misconceptions with regard to the functional description of aspectual systems, the way in which L2 acquisition of aspect is viewed, and also how it is taught in schools and language courses. Obviously, our research is especially relevant for the first part of pedagogical grammar, which is that of descriptive adequacy (see Ruiz de Mendoza, this volume), but not for its final part, which is that of providing improved teaching methods. We can merely present a number of relevant linguistic issues and descriptions that we believe should be taken into account by applied linguists writing pedagogical grammars. Disregarding the discussion on the Critical Period hypothesis, one can state that from a learning point of view it seems nearly impossible for advanced learners to have full command of the aspectual distinctions in the target language (e.g., Schmiedtová 2004; Slabakova 2005; von Stutterheim and Carroll 2006). Equally challenging appears to be the task of learning to express temporal relations in non-aspect languages (for example German) by native speakers of aspect-dominant languages (such as Czech or Russian). This is particularly evident in learners’ ways of structuring information in narratives (e.g., Schmiedtová and Sahonenko, in press; Carroll et al., in press). The difficulties that second language learners of all proficiency levels face when dealing with aspectual relations in the L2 arise partly because of the high complexity and prominence of the aspectual systems as such and the differences between the L1 and the L2 systems. But perhaps they also occur because traditional analyses (e.g., Comrie 1976; Dahl 1985; Smith 1997) of the aspectual categories do not provide the necessary guidelines for teachers to formulate instructions that would make the acquisition of aspect more systematic and thus successful. Learners have to gain competence not only in connecting the form and the corresponding meaning(s), but also in making that connection on the basis of usage principles in discourse. These kinds of competence have to be coherently integrated within the learning process, which is a difficult task for both teachers and learners. Our approach to investigating aspectual systems and their use in discourse is entirely empirical. We base our claims and conclusions on spoken data produced by native speakers and learners. Our framework reflects actual native speaker preferences1 for using aspectual markers in a particular language and, in addition, it describes the internal organization of the respective aspectual system. In line with the current trend in cognitive linguistics, we assume that aspectual categories do not merely depict grammatical features, but that they also mirror conceptual structures and hence

Aspectual concepts across languages

359

have psycholinguistic reality. The aim of this paper is to sketch some of the existing problems, increase awareness of them, and stimulate a discussion. We will address several terminological issues by introducing an empirically based approach to the classification of grammatical aspect providing examples from L1 as well as L2 data. Our material includes data from Czech, Dutch, German and Russian native speakers as well as from Russian and Czech learners of German. The structure of the present paper is as follows: in the next section we will discuss a number of terminological fallacies, then show some empirical data in order to support our claims concerning grammatical aspect, and finally draw our conclusions with a couple of remarks regarding second language learning. 2. Aspect terminology 2.1. Grammatical aspect and lexical aspect (Aktionsart) One of the frequently occurring problems in the literature on aspect is the lack of uniformity concerning the theoretical notion of aspect. We distinguish between two categories: grammatical aspect and Aktionsart2. The former aspect is a purely grammatical category marked by inflectional morphology (e.g., affixes in Slavic languages, the be V–ing form in English). In our approach, grammatical aspect (i.e., in general the morphosyntactic marking of aspectual categories) denotes grammatically encoded concepts. We agree with Klein (1994: 30) when he says that: “... the convential ways of characterizing [grammatical] aspect, whilst intuitively often appealing, are [apparently] not very satisfactory: they have much more the status of metaphorical descriptions than of precise and clear definitions.” These conventional ways include the terms ‘viewpoints,’ ‘viewing a situation from the outside or the inside,’ ‘situation is seen as completed/non-completed.’ We do not adhere to the conventional view that grammatical aspect is a way of seeing situations, which involves lexicosemantic as well as grammatical elements, because this view does not provide a suitable theoretical framework for analyzing empirical data in crosslinguistic context. Aktionsart, by contrast, is a semantic category that expresses temporal characteristics of verb meanings and meanings of verbal predicates (lexical content). Several Aktionsart classifications have been proposed (e.g., Vendler 1967; Smith 1997) but none of them are unproblematic. In our

360


framework, we apply Klein’s classification from 1994. It is based on the Topic Time (TT) notion: Topic Time (TT) is the time for which a particular assertion is made. For example, in She was ill, the TT precedes the time of utterance (TT vs. ); — two-states (two TT-contrasts, e.g., She opened the window – this assertion holds true for two different contrasts vs. , vs. ). Despite the fact that Klein’s framework does not make an explicit difference between Vendler’s accomplishment and achievement, we prefer to use Klein’s classification because Topic Time is a well-defined and transparent notion, which is applicable to numerous languages other than English. In any case, the notion of Aktionsart is less important for our research since our main focus is on aspect. Usually, the notion of aspect comprises grammatical aspect as well as Aktionsart (an exception to this trend is Bertinetto and Delfitto 2000). In line with this misconception, some researchers assume that the acquisition of grammatical aspect is always guided by learners’ knowledge of semantic features encoded on the verb (Aktionsart). In other words, it is believed that grammatical aspect is not acquired independently, but must be accompanied or even preceded by knowledge of Aktionsart. Hence, both categories are usually described and analyzed as a whole. The original proposal goes back to Andersen and Shirai’s Aspect Hypothesis (also called Primacy of Aspect Hypothesis, Prototype Hypothesis, Aspect before Tense Hypothesis) from 1994, on the basis of which they accounted for different types of acquisitional data in Pidgin and Creole


361

languages. Interestingly, this hypothesis seems to be tenable for numerous L1 and L2 varieties; it has stimulated a lot of research related to aspect (for example, Li and Shirai 2000; Stoll 2005; Johnson and Fey 2006) and has been useful for stating initial research hypotheses in the study of child language. It is important to realize, however, that the correlation between lexical and grammatical features, be it aspect or tense, as can be observed in first language acquisition, does not necessarily play a role in the description of the underlying linguistic system. Researchers following the Aspect Hypothesis have failed to adequately differentiate between lexical and grammatical elements, which has led to a mix-up between semantics and the grammatical categories of aspect and tense. For example early occurrence of accomplishment/achievement verbs with past tense marker –ed in L1 English is considered to represent the child’s knowledge of perfectivity. This influential hypothesis does not address the core issue of how to keep the notion of grammatical aspect and Aktionsart apart, nor does it provide a systematic description of these categories. We think that a suitable description of an underlying aspectual system is a necessary prerequisite not only for our general understanding of aspect and its acquisition, but also for developing appropriate teaching methods. Another shortcoming of the Aspect Hypothesis is that it formulates acquisitional patterns for aspectual notions in contrast (i.e., the order of acquisition of perfective vs. imperfective markers). When concentrating on English, which has only one grammaticalized aspectual marker (the suffix –ing), it may indeed make sense to set up an opposition between a verb marked for ongoingness (i.e., progressivity – I am sleeping) and a verb inflected for past tense (He slept all day yesterday), and label the latter as perfective. This seems to work because the “perfective meaning” (completion) arises here through the past tense morphology. However, note that simple past in English is an aspectually unmarked form that is open to +/– perfective interpretation. Thus, the verbal form in He slept is not inherently perfective, but receives its “perfective meaning” (completion) merely by pragmatic knowledge. It could very well be the case that the person who slept yesterday is in fact at the moment of speech still sleeping. This information is simply not part of the temporal semantics of this utterance, and it is also not grammatically encoded (this misconception is present e.g., in Slabakova and Montrul 2002). In some other cases, the pragmatic information is accompanied by lexical features of the verb as in She broke my arm or of the verbal predicate as in He ate up his sandwich. In these examples, the verbs including their arguments inherently express a change of state, which makes the “perfective” reading possible (aided by the past

362


tense morphology). But again, the verbs are not marked for perfectivity by means of aspectual inflectional morphology. Simple forms in English, whether in past or present tense, are with regard to grammatical aspect open (neutral or unspecified). Since these forms lack the ongoing marker altogether they can be called “non-progressive,” but they do not express any aspectual meaning that is contrastive to progressivity. Simple forms in the present tense in English have many different meanings, but most of these meanings arise through the linguistic context (e.g., through the addition of adverbial phrases) or a specific speech act (e.g., an informative act). These meanings are conveyed by lexical and not grammatical features, and therefore the several different meanings that English simple forms can have – habitual, scientific present, etc. – do not represent an aspectual opposition to the progressive aspect. As will be explained below (see Section 2.2), the most prominent meaning – habituality – arises only in specific contexts and is, as we hypothesize, the result of the grammaticalization process of the progressive marker. By grammaticalization we mean the process in which grammatical morphemes gradually develop out of lexical constructions and become more and more used as fully-fledged constructions in an ever-expanding range of contexts. These grammatical constructions are becoming part of the core grammar of a language (cf. Bybee et al. 1994). For English it is true that in certain contexts the simple form can convey “a holistic viewpoint,” for example in He reaches the finish. Note, however, that this meaning, in contrast to the meaning of the –ing form, are not grammaticalized and belong to the lexico-semantic and not the grammatical area. All this is very different from languages that use two grammaticalized aspectual markers whose meanings are truly contrastive. All Slavic languages, for example, can express both meanings – perfective and imperfective – grammatically on the verb. Although these systems do not apply to all verbs and there are some exceptions to the rule, we see a fundamental difference between the Czech/Russian and the English systems. There is an opposition between two different aspectual categories – perfective vs. imperfective – in Slavic languages, neither of which is expressed by past tense marking, whereas no such grammatical opposition exists in English (only the progressive is grammaticalized in English). We believe that this mix-up has been dominating and partially misguiding the overall discussion about aspect typology and acquisition (see for example, general aspect analysis: Verkuyl 1993; Smith 1997; acquisition: Stoll 1998; Wagner 2006). An exception to this trend is Slabakova’s review of recent research on the acquisition of aspect (2002). Slabakova


363

(2002: 176) points out that many studies have blended three different temporal contrasts, that is past vs. present tense, perfective vs. imperfective grammatical aspect, and Aktionsart distinctions. In fact, we would say that fully grammaticalized grammatical notions (e.g., grammatical aspect) only interact (but do not merge) with other temporal categories, such as tense, adverbials, or Aktionsart. To a large extent the English progressive marker –ing is the ideal example of such a fully grammaticalized and independent grammatical category. Adopting our view makes it possible to tear apart grammatical aspect, Aktionsart, and tense and it would possibly improve teaching methodologies because teachers would be able to explain these categories in a more systematic and independent way. 2.2. Telicity vs. perfectivity Another problem we would like to tackle is the confusion between the terms telic and perfective. Similar to the issues discussed above, this problem too is related to an inaccurate differentiation between Aktionsart and grammatical aspect. In our view, the notion of telicity belongs to the domain of lexical features inherent in the verb/verbal predicate while perfectivity is a grammatical category. We define telic verbs or telic verbal predicates as expressing an inherent endpoint, which must not necessarily be realized in a situation (e.g., to fall, to write a paper). It is in principle plausible to assume that all languages have verbs expressing +/– telicity. However, only a number of aspect-prominent languages can convey +/– perfectivity grammatically. In other words, although the two terms are closely related in meaning and can interact with each other at the level of expression, they involve two different layers of linguistic analysis and are hence not synonymous. To illustrate this difference let us consider the following examples from English and Czech. (1)

He ate an apple.

(2)

(On) S-nČd-l jablko. He-NOM PERF-eat-PAST-3SG apple-ACC ‘He ate (an) apple’

In example (1) the verb to eat is a one-state verb denoting only one change of state (Klein 1994) and for the sake of argument we assume that together

364


with the indefinite object an apple it forms a telic predicate. The same utterance in Czech is presented in example (2). In Czech the verb jíst ‘to eat’ is also a one-state verb, but unlike in English, in the example above it occurs as a perfective, marked grammatically by the prefix s-. As in English, we are dealing here with a telic predicate – to eat an apple, but the verb is overtly marked as perfective. So the Czech utterance involves two different features: telicity on the predicate plus the perfective aspect inflected on the verb. Only the first feature is present in the English example. As discussed briefly above, we can see that telicity and perfectivity involve two different operations, yet, in these examples they result in a comparable semantic structure: having reached the right boundary of the situation, i.e., the endpoint of the situation. In spite of this parallelism if we change the tense of the English verb from past to present we observe a shift from (–) aspect to (+) aspect. More precisely, from (–) progressive to (+) progressive as in example (3). (3)

He is eating an apple.

If the aspect is not changed from (–) to (+) progressive, as in (4), the meaning of the utterance becomes problematic. (4)

?He eats an apple.

In (4) the tense change makes the utterance ill formed in contexts of ongoing situations because of the conflict between the presence of an object and the simple present. Normally this combination renders a habitual reading, but then further temporal specification (e.g., He eats an apple every day) or a particular context conveying the habituality (e.g., What does your diabetic friend do when he suffers a hypo?) is required. It is true that English simple forms often denote habituality, but we strongly believe that this is merely a consequence of the grammaticalization of the -ing form.3 In itself, the simple form does not convey habitual meaning grammatically. This can be seen in example (4), where habitual meaning only arises when specific habitual contexts are provided, i.e., either lexical devices (temporal adverbials) or context. Another option for making (4) grammatical is to change the simple verb form into the progressive: He is eating an apple, as in (3). That means that in English a change in tense goes hand in hand with a change in aspectual value: The addition of the –ing suffix (or a temporal adverbial) is obligatory in a context of ongoingness in the present tense.


365

What is relevant here is that despite the change in aspect, the telicity of the English predicate remains unaffected. The Czech example in (5) demonstrates that a shift in tense does not influence the aspectual value, nor the telicity of the utterance. (5)

(On) S-ní jablko. He-NOM PERF-eat-PRESENT-3SG apple-ACC ‘He eats (an) apple (up)’

The interpretation of example (5) is that the situation to eat an apple in Czech is presented as inevitably reaching its endpoint in a very near future. This is very unlike the English predicate, which is telic (a semantic category), but by no means perfective (a grammatical category). In other words, by using a perfective prefix a Czech language user conceptualizes and presents the situation depicted in (5) as perfective. In principle, the Czech aspectual system allows the expression of perfectivity in the present tense,4 which is not possible for English. As shown in (3) and (4), it is compulsory in English to use the progressive in here-and-now contexts. This shows that only the progressive aspect has been grammaticalized in English. In Czech, on the other hand, verbs must be marked either for perfectivity or imperfectivity in all tenses. This is because both aspects have been grammaticalized. The English aspectual system, by contrast, does not contain a systematic opposition between two different grammatical aspects: the “perfective” interpretation of verbal predicates such as to eat up is not brought about grammatically (perfectivity) but it is conveyed lexically by the particle up (telicity). In this sense, perfectivity does not equal telicity. It leads to fundamental problems when the unspecified simple form in the context of telic verbs/verbal predicates is put in opposition to the aspectually marked progressive form. As we will show in more detail below (Section 2.4), Slavic languages have both poles of this aspectual contrast at their disposal and thus represent a completely different system with not only different forms, but also with different underlying concepts. With respect to learning, the difficulty arises when teachers draw parallels between a marked perfective and the English simple form: these are basically false friends.

366


2.3. Imperfectivity The last terms we would like to attend to in Section 2 are the notions of imperfectivity and perfectivity. We will first address the former category. Slavic languages use simplex forms to express imperfectivity (e.g., in Czech psát – ‘to write’) to express imperfectivity apart from the marked imperfective, the so-called secondary imperfective. The secondary imperfective is marked by inflectional morphology, that is, in Czech the suffix (o)va-, and in Russian the suffixes -iva-/-yva-, -va-, -a-/-ja- (e.g., in Czech/Russian vypis-ova-t/ vy-pis-yva-t' – ‘to be in the process of writing out’). There is also a small group of frequently used simplex verbs denoting perfectivity without an explicit morphological marker (e.g., Czech dát – ‘to give’). Because these verbal forms lack any overt grammatical marking of their aspectual value, a question arises. Does their aspectual meaning come from the inherent verbal semantics (Aktionsart) or is it rooted in the grammar (grammatical aspect)? Despite this serious terminological problem, which has not yet been thoroughly investigated,5 we hypothesize that Slavic simplex forms differ from those in English, German, or Dutch. A possible justification for this line of thinking is the following. Usually, adding a prefix6 to a Czech simplex imperfective verb results in changing the aspectual features into the perfective, as in example (6). (6) Prefixation of the simplex imperfective form Czech

pít

Æ

IMPF-simplex

English

‘to drink’

VY- pít PERFdrink ‘to PERFdrink’ § ‘to drink up’

Æ

The situation is different when dealing with simplex perfective verbs (7). (7) Prefixation of the simplex perfective form Czech

dát

Æ

PERF-simplex

English

‘to give’

Æ

U-dat PREF-PERFgive ‘to report’

In example (7), the prefix u- only changes the meaning of the verb, but not the aspectual value. That is, the verb remains perfective and a new lexical


367

entry is derived. Another relevant point to be mentioned here is that simplex perfective forms, such as dát ‘to give,’ can only be used in perfective contexts. For generic and imperfective contexts the marked imperfectivized form – dá-VA-t – must be employed. This shows that the simplex form has an aspectual value – the perfectivity – on its own. Because of these observations we theorize that the perfective value is already encoded in the stem of the verb regardless of the lack of overt marker(s). There are no comparable cases in English, German, or Dutch. Therefore, we argue that (i) in contrast to English, simplex forms in Slavic languages have a default grammatical aspect (in addition to their inherent Aktionsart) and (ii) that simplex forms in English, German, or Dutch only make use of Aktionsart and are underspecified with respect to grammatical aspect. Turning back to L2 learners of Slavic languages, the dichotomy in the domain of simplex forms must pose a learning challenge since simplex forms are unmarked by default, nevertheless they carry an unambiguous aspectual meaning. Because of this, we are again dealing with a kind of false friend when translating (and teaching) the Czech dát as English ‘to give.’ The next section will focus on some difficulties in characterizing and defining the notions of imperfectivity and perfectivity across languages. 2.4. Perfectivity vs. imperfectivity: Conceptual differences This section focuses on the comparison between two binary aspectual systems: the Czech and the Russian systems. Although these two Slavic languages show many typological similarities, our research (e.g., Schmiedtová and Sahonenko in press) shows that in the aspectual domain there are crucial differences in native speakers’ preferences, as well as in the distribution of the forms within the system. These differences may pose a real challenge to L2 learners. As stated above, both languages encode two contrasting grammatical aspectual categories: the perfective and the imperfective. Both languages also use a number of simplex verbs, but in what follows, we will only focus on grammatically marked aspects. In principle, there are two operators that can change the aspectual value of a verb. The first operation is adding a prefix to the verbal stem. These prefixes do not only change the grammatical aspect, but they can also affect the semantics of the verb, i.e., derive a new lexical item. Moreover, with some verbs it is only the lexical meaning that changes. So, the challenge here is that the lexical and the grammatical

368


modification can hardly be separated from one another (Comrie 1976; Schmiedtová 2004). The other operation is adding a suffix. Suffixation leads to secondary imperfectivization of the verb (regardless of the type of verb stem) and the change is mainly grammatical (from perfective to imperfective aspect). These claims hold true for Russian as well as Czech. Let us consider a couple of examples. (8) Prefixation of the simplex imperfective form Czech: psát

Russian: pisat’

IMPF-simplex

Czech: VY-psa-(-t) Russian: VY -pisa(-t’) PREF-writePERF English: ‘to write out (all keywords)’/’to announce (a job)’ (9) Suffixation of the simplex perfective form Czech: dát

Russian: dat’

PERF-simplex

Czech: dá-VA(-t) Russian: da- VA(-t’) PERF-give-2ndaryIMPERF English: ‘to be giving’ (10) Suffixation of a prefixed perfective form Czech: VY-psat Russian: VY-pisat’ PREF-writePERF Czech: VY-pis-OVA(-t) Russian: VY –pis-YVA(-t’) PREF -2ndaryIMPERF English: ‘to be writing out (all keywords)’/‘to be announcing (a job)’ In (8) a simplex imperfective is turned into a perfective by the prefix vy-, which changes the meaning. Note that one and the same operator affects two linguistic areas: lexicon and grammar.The point of example (9) is to


369

illustrate the change of a simplex perfective to a marked imperfective verb (i.e., change in grammatical aspect only). The same suffix (-(o)va/-(y)va) can be attached to a prefixed verb denoting perfectivity. As in (9), the suffix in (10) also changes the grammatical aspect. The question to ask here is: what are the conceptual consequences of these operations? We do not completely adhere to how perfectivity and imperfectivity are usually described in the literature, e.g., Langacker (1987) or Bybee (1992: 144): “... perfective, which indicates that the situation is to be viewed as a bounded whole, and imperfective, which in one way or another looks inside the temporal boundaries of the situation ...” We want to be more specific and claim that the crucial difference between the perfective and imperfective is the degree of focus on the right boundary of a situation (e.g., in the situation, in which a person is drinking up a glass of water, the right boundary is reached when the glass is empty and the person is in the post state of having finished a glass of water). The following two figures may serve to clarify this point. Right Boundary

Post State PERFECTIVE Figure 1. Scope of the perfective aspect in Slavic languages Defocused Right Boundary

IMPERFECTIVE Figure 2. Scope of the secondary imperfective aspect in Slavic languages

As illustrated in Figure 1, the function of the perfective in Czech and Russian is to encode that a situation has reached its right boundary and also that an assertion is made about the possible post state of this situation

370


(speaker's focus is on the right boundary). In contrast, the secondary imperfective accesses the time interval prior to the right boundary, but (!) does not ignore the right boundary of the situation altogether, rather the secondary imperfective defocuses it, as in Figure 2. So, in both instances, the perfective as well as the imperfective aspect, the attention centers around the right boundary. This view of the imperfective aspect puts the frequently assumed similarity between the progressive (e.g., in English or Dutch) and the imperfective into question. Even though such a comparison might be linguistically interesting, our analyses show that the two aspectual operations are very different (for more details, see Section 3). In the next section we will provide empirical evidence for the conceptual differences between the Czech and the Russian aspectual systems, as described above. 3. Underlying concepts in cross-linguistic comparison: Empirical data In this section, we will first explain, using production data from Czech, Dutch, and Russian native speakers as well as advanced L2 learners of these languages, that grammatical aspect is not only a matter of grammatical form, but also of conceptualization. This conceptual structure is reflected in the language-specific preferences of native speakers when using different aspectual forms in their L1, as well as reflected in the overall degree of grammaticalization within each system. This new take on analyzing aspectual distinctions is pursued by our research group at the University of Heidelberg and finds its origin in research conducted by Christiane von Stutterheim and Mary Carroll. We base our analyses and description of aspectual systems on production data elicited from large samples of native speakers as well as learners. The experimental approach consists of an online production task, in which speakers (N = 30) are asked to retell short everyday situations in answer to the question What is happening? (translated into all relevant languages), i.e., video clips depicting somebody drinking a glass of water, a dog running into a house, etc. In order to test our hypotheses, we make use of several sets of stimuli that are grouped according to situation type (e.g., causative actions, locomotions with +/– endpoints, etc.). This approach forces speakers to choose a particular aspectual form, which is appropriate or obligatory for a specific situation type. To strengthen our arguments we also use other methods, such as speech onset times and eye-tracking meas-


371

urements. By adopting this line of empirical research, combining linguistic analyses of production data with psycholinguistic methodology, we believe that we are able to tap into speakers’ conceptualization patterns7. The focus of the previous studies (e.g., Carroll and von Stutterheim 2003; von Stutterheim and Carroll 2006; Klabunde and von Stutterheim 1999) was on Semitic, Germanic, and Romance languages. It has been shown that the way events are depicted is highly dependent on the feature +/– grammatical aspect. It has also been found that the underlying principles for event construal are perspective driven and strongly linked to patterns of grammaticalization. Additionally, recent L2 studies have provided evidence that even very advanced learners fall back on conceptualization strategies from their L1 when construing temporal events in a L2 (cf. von Stutterheim and Nüse 2003; Schmiedtová and Sahonenko in press). These findings also hold true for learners describing situations that are more complex than single events such as narratives. Carroll and Lambert (2003) have shown that the use of aspectual categories influences the overall information structure in more complex tasks, such as composing written or oral narrative texts. The next sections will deal with conceptual representations that underlie the grammaticalized aspectual categories of Czech, Russian, and Dutch. 3.1. Differences between the Czech and the Russian aspectual systems In the previous section we have discussed the concepts that are encoded by the two aspectual categories: the perfective and the imperfective. In what follows, we will present production data of native speakers of Czech and Russian (for both N = 30). We will see that, despite very similar underlying aspectual systems, Czech and Russian native speakers have different aspectual preferences when construing events in their mother tongue. These preferences, as we believe, reflect the way in which these speakers view and conceptualize a situation. First of all, in both the Czech and Russian native speaker data, there is a pronounced tendency to relate events to the right boundary (for Czech: 87% of all speakers; for Russian 77% of all speakers). This means that speakers mark an evident right boundary or endpoint of a situation as depicted in the stimulus. The stimuli set consisted of two types of scenes: in type one, the right boundary of a situation was visible in the clip and actually reached; in the other type, only a potential right boundary could be inferred but it was not depicted as being reached in the clip. The difference

372


between the ways in which native speakers of Czech and Russian verbalized stimuli of the second type lies in the fact that Czech native speakers mention the endpoint more frequently (for Czech 65% of all speakers; for Russian 25% of all speakers). In addition, Czech native speakers use the perfective form, independent of the scene type. Russians, on the other hand, showed a clear preference for using the secondary imperfective in all scenes. When they used the perfective form it was exclusively for scenes showing the right boundary being reached. In other words, speakers of different languages follow different preferential patterns when they encode events. We believe that these preferences which so far have been described from a linguistic point of view (i.e., surface structure) are rooted in differences in conceptualization of events. In one and the same stimulus, Czech native speakers concentrate on the time interval at and after the right boundary whereas Russian native speakers are sensitive to the time interval preceding the right boundary. At the same time, the data show that the distribution of the aspectual forms within each system differs, too. That is, in Russian the imperfectivizing suffix -(y)va is productive and can be applied to many verbs. In Czech, by contrast, this suffix only combines with a small group of verbs. Additionally, as pointed out by Schmiedtová (2004), the perfective form, when used in the present tense, can have a here-and-now meaning in Czech. This is completely impossible in Russian where the present perfective always refers to the future. We hypothesize that in Czech the increased use of the perfective form goes hand in hand with the prominence of the underlying conceptualization (as depicted in Figure 1). In other words, the extensive use of the perfective aspect in Czech results in a perspective focusing on the right boundary of a situation and/or its post state. The same logic applies to Russian, where the frequent use of the secondary imperfective goes hand in hand with the imperfective perspective (i.e., focus on the time interval preceding the right boundary without excluding it completely). It remains an open question, however, in what direction this influence takes place. The relevant point here is that despite big similarities between the two aspectual systems, Czech and Russian native speakers differ considerably as far as their aspectual preferences are concerned. In summary, our experimental data show that there is interplay between grammatical categories and conceptual structures. Furthermore, we see that even speakers of typologically related languages display different conceptually driven perspectives (preferential patterns) when selecting information for event construal.


373

With respect to L2 learning, in Schmiedtová and Sahonenko (in press) we showed that advanced Czech and Russian learners of German adhere to their respective L1 preference. For example, Czech learners use the concept of perfectivity in L2 German although German does not have grammatical aspect at all. The adherence to this concept becomes apparent in more frequent mentioning of endpoints in the form of local adjuncts (e.g., into the house) when retelling video clips depicting locomotions with ± endpoints. Even though German native speakers are also inclined to mention endpoints frequently (as pointed out in, e.g., von Stutterheim and Lambert 2005), the number of endpoints verbalized by L1 Czech speakers of German exceeds the average for German native speakers. This is a relevant finding because it illustrates that patterns found for native speakers for event depiction in their native language still drive the perspectivization in L2 production. This important issue presents a considerable challenge to language teachers, since, for learners, being aware of the meanings of various aspectual categories is a good starting point for achieving native-like competence in a second language. 3.2. Progressive in English and Dutch: grammaticalization and conceptual structure This part of the paper is devoted to the Dutch language. This is because in Dutch the progressive marker aan het + V-INF zijn is currently being grammaticalized (Flecken 2006). We are aware that a truly grammaticalized aspectual marker is morphological in nature and that the Dutch marker is still a periphrastic construction. However, we speculate that in the course of the grammaticalization process it will be reduced to a verbal morpheme. This seems to be already noticeable when considering native speakers’ shortened pronunciation of this construction. Because the Dutch grammatical system is in the middle of this process, we envisage that learners are confronted with the hard task of figuring out how the system operates. We will first present some empirical data illustrating the range of applications of this marker. Furthermore, we will show that the range is expanding, following the grammaticalization process described in Bybee et al. (1994), which motivates our focus on verb type. We will briefly discuss some differences between the Dutch construction aan het + V-INF zijn and the German construction am + V-INF sein and we will draw parallels between the Dutch and the English progressive marker.

374


Finally, we will demonstrate that progressivity and imperfectivity denote two different temporal concepts. First of all, it is necessary to define our notion of grammaticalization. As mentioned above, grammaticalization means expansion of the range of contexts in which a particular construction is applied. The starting point for grammaticalization is the use of a particular construction in its prototypical lexical environment. This use is inherently linked to a specific meaning of the grammatical feature, which slowly spreads out to less prototypical uses/contexts (Comrie 1976; Bybee et al. 1994). Regarding the meaning of the Dutch progressive marker, in our data we observe that modifying a Dutch verb with the aan het-construction depicts situations as ongoing, as in example (11). (11) Ik ben aan het lezen. ‘I am reading’ The aspectual marker in (11) defocuses both the initial and the final boundary of the situation and hence the temporal reference applies only to the here-and-now. The meaning of the Dutch aan het-construction is, therefore, identical with the meaning of the English –ing, which has the same function. Let us take a closer look at the similarities between Dutch and English. At first sight, the Dutch marker looks like a locative construction because of the locative aan (like the English prepositions at/on) (Boogaart 1999). Interestingly, the English progressive marker might have evolved out of a locative construction as well. This original construction looks similar to the contemporary Dutch periphrastic construction (12) (example taken from Bybee et al. 1994: 132): (12) He is on hunting. ‘He is hunting’ Comparing (11) and (12), we can see that the original meaning of both constructions could have been ‘to be in the place of doing something.’ This originally locative meaning evokes a very deictic here-and-now context, and we assume that, in a way, this condition was the starting point for the grammaticalization of the –ing form (also in Jespersen 1949; Comrie 1976). We claim that it is also the starting point in the grammaticalization process of the aan het-construction in current Dutch. In English, we see


375

that this precondition is no longer necessary for application of –ing, as is apparent when looking at examples (13) and (14). (13) Katja is having an affair with Christopher. (14) Doro is practicing law. The meaning of the -ing form in (13) and (14) is not necessarily restricted to the deictic (locative) here-and-now, but it is extended over a longer period of time (as in (13)), and it can even describe a habitual feature (as in (14)). In Dutch, this type of application of the aan het-construction is not (yet?) possible. The meaning of this construction mostly refers, at this point in time, to agentive subjects who are in the midst of an activity at reference time or in the very deictic past as in (15a) and (15b). (15)

a. Ik ben aan het werken. ‘I am working’ b. Gisteren was ik aan het studeren. ‘Yesterday, I was studying’

We presume that in contrast to English –ing, the use of the Dutch construction in true habitual contexts is more constrained.8 The traditional view of the aan het-construction in Dutch literature is that it is merely “a locative construction with a ‘progressive-like’ meaning” (e.g., Boogaart 1999: 167). This view, however, does not take into account that the Dutch grammatical system is evolving and hence the progressive marker is becoming part of the core grammar. In our view, we take the above observations to mean that the Dutch progressive construction is at the onset of a similar grammaticalization process but that, the English progressive marker is in a far more advanced stage within this process. This has been shown empirically: in our data English native speakers, when construing events, use the –ing in all cases whereas the simple form is completely absent. That is, all native speakers of English (N=60) in our sample resort to the progressive marker when asked to tell what is happening or even what happens. Going back to the Dutch language, in order to sketch a more accurate development of the aan het-construction, we focus on the types of verbs (Aktionsart, in line with Klein 1994) that take the marker aan het (in line with Bybee’s approach to grammaticalization).

376


The first step of grammaticalization, thus the prototypical context for using progressive markers, is to use it in situations denoting an activity, e.g., wandelen (‘to take a walk’), zwemmen (‘to swim’), but also een boek lezen (‘to read a book’), de tafel poetsen (‘to clean the table’). In the prototypical phase, the prerequisite for using the aan het-construction is the possibility of defocusing boundaries. All predicates that inherently refer to one of the boundaries (such as to fall) do not combine with the aan het marker at this stage of grammaticalization. The verb type which meets all these conditions is the one-state verb, such as zwemmen ‘to swim.’ In the next grammaticalization phase, the two-state verb referring to a rather long time span is included (e.g., veranderen ‘to change’) followed by the twostate verb denoting a short time interval (e.g., breken ‘to break’). The last step is the expansion to zero-state verbs, such as houden van ‘to love.’ Interestingly, in English the grammaticalization process of the –ing suffix has reached this last phase: It is grammatical to say I am loving it (in the sense of ‘I am enjoying it’) or She is having a baby (although they have two different temporal meanings). To illustrate this process for Dutch, we present some preliminary results of an acceptability judgment task using a five degree scale ranging from completely acceptable (5) to completely unacceptable (1). We asked 30 Dutch native speakers to make a choice between a simple verb form and a verb marked by an aan het-construction in here-and-now contexts. We differentiated between the four types of verbs described above: one-state verbs, two-state verbs with long and short duration, zero-state verbs. It turned out that one-state verbs (e.g., lezen ‘to read,’ tekenen ‘to draw,’ schilderen ‘to paint,’ knutselen ‘to tinker,’ pianospelen ‘to play the piano’) triggered the most frequent use of the aan het construction. The second best attractor for aan het was the two-state verb with a long duration9 (as in afmaken ‘to finish,’ afwassen ‘to do the dishes,’ veranderen ‘to change’), followed by the two-state verb with a short duration, e.g., vallen ‘to fall,’ exploderen ‘to explode,’ breken ‘to break.’ The zero-state verbs did not elicit any choices for the aan het-construction in the here-and-now-context. As far as acceptability is concerned, this task has allowed us to interpret the values that the participants attached to the form they did not choose. They always had to grade the other form in terms of its acceptability in a given context. The most important finding was that participants rated the simple form as unacceptable in here-and-now contexts for the verbs expressing a game-like activity, examples of which are zwemmen ‘to swim,’ tafeltennissen ‘to play table tennis,’ schilderen ‘to paint.’ Moreover, they rated the aan het form as unacceptable in clauses with motion verbs plus a


377

depicted endpoint (as in *Ik ben in het water aan het springen ‘I am jumping into the water’). These results make sense: The latter verb type expresses the shortest possible duration, namely the time interval right before reaching the final boundary, which makes defocusing of boundaries quite impossible. A further interpretation of these results is that in a number of cases the aan het-construction was considered compulsory by the participants. As pointed out above, this is the case for situations expressing activities taking place in the here-and-now. The simple form in these cases was rated unacceptable because using the simple form renders a habitual meaning in these contexts. For example, following the question Wat ben je aan het doen? ‘What are you doing?’ all Dutch native speakers in our sample choose the aan het form in combination with one-state verbs, e.g., Ik ben aan het werken ‘I am working.’ The simple form, Ik werk ‘I work,’ is rated as completely unacceptable in such contexts. In summary, when activity verbs and verbal predicates are used in a here-and-now context the aan het marker is obligatory. Again, this is comparable to English, because the difference between I am dancing and I dance is that the former implies an activity that is taking place at the time of utterance; whereas the latter refers to a habitual activity (a hobby or perhaps even a job). Bybee et al. (1994) label this phenomenon as grammaticalization of zero (i.e., the unmarked form receives a different meaning in certain contexts). Of course we realize that the depiction of the grammaticalization process is rather different from the question of what the actual attractors are for using the aan het-construction. It cannot solely depend on the verb type, but will rather be a matter of the entire predicate.10 An interesting comparison to draw at this point is between Dutch and German. Though both languages are typologically similar, one important difference is that Dutch is grammaticalizing a marker for ongoingness, whereas in German ongoingness is mainly expressed by lexical means. German has a construction, which is form-wise very similar to the Dutch one. Consider example (16). (16)

German: Dutch: English:

Rieke ist (gerade) am/beim Kochen. Rieke is aan het koken. ‘Rieke is cooking’

The German periphrastic construction is merely a regional and stylistic variant of Standard German while in Dutch it is an obligatory marker in

378


such a context, compared to the unmarked simple verb form. Furthermore, the progressive markers in English as well as Dutch are systematically used by native speakers for the expression of other temporal concepts, such as the expression of simultaneity between two events in present tense. The German construction is never produced in such contexts (see Schmiedtová 2004; Flecken 2006). Furthermore, unlike in German, the aan hetconstruction in Dutch can also be combined with a direct object (e.g., Ik ben een boek aan het lezen ‘I am reading a book’). Looking at these similarities from a learner’s point of view, we have another occurrence of false friends. Learners have to deal with two very similar forms that do not show a similar distribution across verbs and, in addition, are employed by speakers for different purposes. The last point to be addressed in this section is the difference between progressive and imperfective aspect. As we have shown in Section 3.1, speakers of Slavic languages do not ignore the right boundary of the depicted situation when using the marked imperfective, but rather include it in their conceptualization and verbalization of situations. In other words, by using this form speakers refer to the time interval anchored in the hereand-now and to the linkage of this time interval to the right boundary. The Dutch and the English progressive, by contrast, are used to link situations to the deictic here-and-now without any explicit temporal information about the right (or left) boundary. The progressive marker merely expresses ongoingness. This is especially true in Dutch where the grammaticalization process of the aan het marker has started out exactly from this context. To relate this observation to the conceptualization of temporal events, we know from eye-tracking studies that Dutch and English speakers concentrate only on the ongoing process of situations regardless of whether they depict a right boundary (von Stutterheim and Carroll 2006; Carroll et al. in press). We speculate that Slavic speakers, when using the secondary imperfective to describe ongoing situations of the same type as above, will also pay attention to the right boundary. To conclude this section, it is important for researchers, teachers and learners to take into consideration the conceptual differences between the imperfective and the progressive aspects.


379

4. Conclusions The present paper centers around the idea that the analysis of grammatical aspect contains at least the following two different areas: the form and the meaning. Another idea is the usage and applicability of aspectual forms in context that are determined by the preferences of native speakers. When investigating aspectual forms cross-linguistically many similarities can be observed. The tricky issue is, however, that the mere existence of a form in a sytem or similar forms across systems does not necessarily entail an equally frequent production. To this end, we have demonstrated on the basis of a comparison between German and Dutch that similar forms with comparable meanings do not show the same distribution in native speakers’ production. The same holds true for the language pair Czech and Russian. Despite the similarities between the two aspectual systems, Czech and Russian native speakers show different preferences for applying aspectual forms. These preferential patterns are closely linked to differences in conceptualization, which only become evident when examining empirical material collected by means of experimental methods. The second area of analyzing aspectual systems is meaning. We have claimed that categories such as progressive and imperfective aspect, albeit applicable in comparable contexts, encode different temporal concepts. Again the same statement holds for the terms telic and perfective. They too are not interchangeable and, in addition, belong to two different kinds of aspect: lexical (telic) vs. grammatical (perfective). Note that even when two forms and their temporal meanings are very similar there can still be a difference with respect to the conditions under which these forms can be employed. This is directly connected to the degree of grammaticalization of the respective aspectual form. This has been presented on the basis of the progressive markers in English and Dutch. Another point to be mentioned here is that many divergences pointed out in this paper do not only occur between typologically distinct languages (such as Russian and German), but also between languages that are typologically closely related (e.g., languages within the Slavic or Germanic group). To summarize the differences between the different languages that we addressed, consider Figure 3 below. These observations are highly relevant for teaching and learning. It is reasonable to assume that to focus on form is the least complex approach to teaching aspect, although we have illustrated that even in this area false friends can be identified. As far as meaning is concerned the issues are equally serious. Several aspectual categories

380


that we dealt with are used synonymously in the literature, even though they denote semantically and conceptually different entities. Now, what about L2 learning? It is true that at the onset of acquisition false friends can aid and support the learning process. Looking at advanced learners, on the other hand, provides a considerable piece of evidence that false friends hinder learners in the possibility of achieving nativeness (e.g., English learners of Czech in Schmiedtová 2004). Note that advanced learners are in perfect command of the aspectual forms and even their meaning (i.e., they do not make any grammatical errors), but they do not successfully (not in a native speaker-like fashion, that is) use the principles that govern the application of the forms. In other words, they do not follow native-like preferences, but rather rely on patterns of use from their respective L1s (research conducted by our group in Heidelberg). Language

Czech/Russian

English

Dutch

German

Form

sufixes/ prefixes

-ing

aan het + V (inf) zijn

am/bei + V (inf) sein

Temporal

± reaching of the right boundary

defocusing boundaries



Term

imperfective/ perfective

progressive

progressive

?

Degree of grammatica lization

both aspects fully grammaticalized

fully grammaticalized

in the process of being grammaticalized

not in the process of being grammaticalized

function

Figure 3. Overview aspectual devices in different languages

We are not sure whether these preferences can be learned at all (for a discussion of the feasibility of ultimate attainment, see e.g., van Boxtel 2005). Nevertheless, it is essential to attempt to encourage the learning of aspectual distinctions as a whole. That means that linguists and language teachers have to realize that the debate on aspect is not only a matter of terminology, but that aspect is a conceptual category that requires empirical research. We believe that the approach to view aspect as a conceptual category and to adhere to empirical research when investigating this linguistic domain would be beneficial to teachers as well as learners.


381

Notes 1. The term preference refers in our framework to highly automated processes that speakers activate during speech production. The structures involved in these processes consist of concepts that are expressed through grammaticalized (linguistic) means. 2. To avoid confusion, in our terminology we label lexical aspect Aktionsart. 3. Bybee et al. (1994) label this phenomenon grammaticalization of zero (of the unmarked form). 4. The present perfective in Slavic languages is considered to have a future interpretation (for Czech: Petr 1987; for Russian: Isaþenko 1982). Recent research, however, has shown that this is not necessarily the case for Czech, where perfectively marked verbs in the present tense can have a present tense (here-andnow) interpretation (Schmiedtová 2004, 2005). 5. A possible way of testing whether the aspectual value of the unmarked simplex forms is part of the grammar (grammatical aspect) or the lexicon (Aktionsart) is to conduct a priming experiment. This research question will be addressed in our lab at the University of Heidelberg in the near future. 6. There are about 20 different prefixes available in Czech that are used to make a verb perfective. Each of them is associated with a cluster of meanings, most of them exhibit polysemy and homonymy, and the realization of a given meaning of a prefix is highly dependent on the context in which the prefix occurs. The same holds for Russian. 7. Preliminary results clearly indicate that grammatical features guide speakers’ attention patterns: To be more precise, the focus on the right boundary as predicted by our linguistic analyses of Czech and Russian is visible in speakers’ eye movements (significant difference in amount of fixations in the critical region) and speech onset times (a significantly later speech onset times for speakers who are right boundary-minded). The patterns that were found in the production data thus have a psycholinguistic reality. 8. We are currently testing this hypothesis with Dutch native speakers by means of an acceptability judgement task. 9. The duration was brought about through the description of the situation. The verb itself does not reveal the duration of the situation. For example, in the case of veranderen, the situation was described as “changing the interior of one’s apartment,” elongated with several adverbials expressing that you have been working on this for a very long time so far and you will not finish this in the near future. 10. This approach to the aan het-construction is being pursued by Marianne B. Starren’s research group in the Business Communication Department at the Radboud University in Nijmegen, The Netherlands.

382


References Andersen, Roger, and Yasuhiro Shirai 1994 Discourse motivations for some cognitive acquisition principles. Studies in Second Language Acquisition 16: 133–156. Bertinetto, Pier Marco, and Denis Delfitto 2000 Aspect vs. actionality: Why they should be kept apart. In Tense and Aspect in the Languages of Europe, Östen Dahl (ed.), 189–225. Berlin: Mouton de Gruyter. Boogaart, Ronny 1999 Aspect and temporal ordering: A contrastive analysis of Dutch and English. PhD dissertation, Amsterdam/The Hague: Holland Academic Graphics. Boxtel, Sonja van 2005 Can the late bird catch the worm? Ultimate attainment in L2 syntax. PhD dissertation, Nijmegen/Utrecht: LOT. Bybee, Joan 1992 Entry on tense and aspect. In International Encyclopedia of Linguistics, William Bright (ed.), 144. Oxford: Oxford University Press. Bybee, Joan, Revere Perkins and William Pagliuca 1994 The Evolution of Grammar: Tense, Aspect and Modality in the Languages of the World. Chicago: University of Chicago Press. Carroll, Mary, and Monique Lambert 2003 Information structure in narratives and the role of grammaticised knowledge: A study of adult French and German learners of English. In Information Structure and the Dynamics of Language Acquisition, Christine Dimroth and Marianne Starren (eds.), 267–287. Amsterdam: John Benjamins. Carroll, Mary, and Christiane von Stutterheim 2003 Typology and information organization: perspective taking and language-specific effects in the construal of events. In Typology and Second Language Acquisition, Anna Ramat (ed.), 365–402. Berlin: Mouton de Gruyter. Carroll, Mary, Monique Lambert, Silvia Natale, Marianne Starren, and Christiane von Stutterheim in press Being specific: The role of aspect in event construal when grounding events in context. A cross-linguistic comparison of advanced second language learners (L1 French-L2 English, L1 German-L2 English, L1 Dutch-L2 French, L1 Italian-L2 German). In Dynamics of Learner Varieties, Stefanie Haberzettl (ed.). Berlin: Mouton de Gruyter. Comrie, Bernard 1976 Aspect. Cambridge: Cambridge University Press.


383

Dahl, Östen 1985 Tense and Aspect Systems. Cambridge, Oxford: Basil Blackwell. Flecken, Monique 2006 The expression of simultaneity in L1 Dutch. MA thesis, Radboud University, Nijmegen, The Netherlands. Isaþenko, Aleksandr 1982 Die russische Sprache der Gegenwart. München: Max Hueber. Jespersen, Otto 1949 A Modern English Grammar. Part IV: Morphology. Heidelberg: Carl Winter. Johnson, Bonnie, and Marc Fey 2006 Interaction of lexical and grammatical aspect in toddlers’ language. Journal of Child Language 33: 419–435. Klabunde, Ralf, and Christiane von Stutterheim (eds.) 1999 Representations and Processes in Language Production. Wiesbaden: Deutscher Universitäts Verlag (Studien zur Kognitionswissenschaft). Klein, Wolfgang 1994 Time in Language. London: Routledge. Langacker, Roger W. 1987 Foundations of Cognitive Grammar, Vol. I: Theoretical Prerequisites. Stanford, California: University Press. Li, Ping, and Yasuhiro Shirai 2000 The Acquisition of Lexical and Grammatical Aspect. Berlin: Mouton de Gruyter. Petr, Jan 1987 Mluvnice þeãtiny: Skladba. Praha: Academia. Schmiedtová, Barbara 2004 At the Same Time ... The Expression of Simultaneity in Learner Varieties. Berlin: Mouton de Gruyter. 2005 The notion of aspect in Czech and Russian. Paper presented at Dynamics of Learner Varieties, Berder, France, March 2005. Schmiedtová, Barbara, and Natalya Sahonenko in press Die Rolle des grammatischen Aspekts in Ereignis-Enkodierung: Ein Vergleich zwischen Tschechischen und Russischen Lernern des Deutschen. In Fortgeschrittene Lernervarietäten: Korpuslinguistik und Zweitspracherwerbsforschung, Patrick Grommes and Maik Walter (eds.). Tübingen: Max Niemeyer Verlag. Slabakova, Roumyana 2002 Recent research on the acquisition of aspect: An embarrassment of riches? Second Language Research 18 (2): 172–188. 2005 What is so difficult about telicity marking in L2 Russian? Bilingualism: Language and Cognition 8 (1): 63–77.

384


Slabakova, Roumyana, and Silvina Montrul 2002 On viewpoint aspect interpretation and its L2 acquisition: A UG perspective. In Tense-Aspect Morphology in L2 Acquisition, Yasuhiro Shirai and Rafael Salaberry (eds.), 363–396. Amsterdam: John Benjamins. Smith, Carlota 1997 The Parameter of Aspect. Dordrecht: Kluwer Academic Press. Stoll, Sabine 1998 The role of Aktionsart in the acquisition of Russian aspect. First Language 18: 351–378. 2005 Beginning and end in the acquisition of the perfective aspect in Russian. Journal of Child Language 32: 805–825. Stutterheim, Christiane von, and Mary Carroll 2006 The impact of grammaticalised temporal categories on ultimate attainment in advanced L2-acquisition. In Educating for Advanced Foreign Language Capacities: Constructs, Curriculum, Instruction, Assessment, Heidi Byrnes (ed.), 40–53. Georgetown: University Press. Stutterheim, Christiane von, and Monique Lambert 2005 Crosslinguistic analysis of temporal perspectives in text production. In The Structure of Learner Varieties, Henriette Hendricks (ed.), 203–230. Berlin: Mouton de Gruyter. Stutterheim, Christiane von, and Ralf Nüse 2003 Processes of conceptualization in language production. Linguistics (Special Issue: Perspectives in Language Production) 851–881. Vendler, Zeno 1967 Verbs and Times. In Vendler Zeno (ed.) Linguistics in Philosophy, 97–121. Ithaca, New York: Cornell University Press. Verkuyl, Henk 1993 A Theory of Aspectuality. Cambridge: Cambridge University Press. Wagner, Laura 2006 Aspectual bootstrapping in language acquisition: telicity and transitivity. Language Learning and Development 2 (1): 51–76.

The use of passives and alternatives in English by Chinese speakers Liang Chen and John W. Oller, Jr.

Abstract Cognitive linguistics introduces speaker construal as a dynamic element into the study of grammar (Langacker, 2002). The relationship between a speaker and an event that he conceptualizes and portrays involves focal adjustments and imagery (Langacker 1987). Moreover, each native language has trained its speakers how to choose to construe and to attribute differentiated cognitive significance to an event and its parts for expressive purposes. Since this training is carried out in childhood, it is exceptionally resistant to restructuring in adult L2 acquisition (Slobin, 1996). This suggests that even advanced L2 learners may still have difficulty in choosing one grammatical construction over alternative forms due to nativelanguage transfer. To test this idea, we examine English passive constructions in written texts produced by 25 advanced learners of English in China. Results show that they frequently have difficulty with the choice between the transitive active structure and the passive even though they are capable of producing both forms. We suggest that a Chinese speaker can learn to use the English passive construction appropriately and productively by discovering how native speakers do it in much the same way a child learns English (cf. Tomasello, 2003). In the foreign language context, this can be accomplished by helping L2 learners discover the pragmatic connection between speaker construal and usage-patterns. Natural conversations in well-acted high quality sound motion pictures can be used to facilitate the discovery of motivated active and passive structures and the pragmatic mapping relations between their distinct forms and functions in English (Chen and Oller 2005). Keywords: passive; alternatives; speaker construal; perspective; degree of agency; transitive event; flexibility; construction; packaging; conceptual transfer; naturalistic setting; structural device; motion pictures; motivation; frog stories; advanced learners of English; Chinese

1. Introduction Theories of cognitive grammar (CG) and other approaches to cognitive linguistics have advocated speaker construal as a dynamic element in lan-

386

Liang Chen and John W. Oller, Jr.

guage acquisition. CG gives special attention to the ways in which speakers parse up and construe the parts of a particular event or situation for expressive purposes (Langacker 2002). The notion of “construal,” as Langacker (1987: 487–488) defines it, is “the relationship between a speaker (or hearer) and a situation that he conceptualizes and portrays.” This notion implies that the language user has an active role in organizing and structuring his or her world. Any particular event can be construed from many different perspectives, and may be expressed in a range of linguistic structures. The actual expression of the event depends on how the speaker conceptualizes the event and chooses to communicate it linguistically. Because language always provides more than one way to encode events for discourse purposes, it can be argued that a speaker must often choose between two or more possible routes of construction during the online process of communication. In our own theoretical and empirical studies of discourse processes and language acquisition (see Oller and Giardetti 1999; Oller et al. 2005; Chen and Oller 2005), we have emphasized the dynamic interactions of deeply layered and ranked systems of sense, movement, and language (Oller, Oller, and Badon 2006). Language processing involves a series of mostly unconscious processes: coordinating lexical items, phonological forms, syntactic patterns, and semantic/pragmatic interactions. The situational use of a particular sequence of expressive forms – syllables, morphemes, lexical items, phrases, clauses, and higher constructions – involves perspective taking. The choice of one expressive form over another may often indicate a switch in perspective or the emergence of a particular conception of the conceived events and situations (Langacker 2002: 12). It is true that many scholars have thought of this process as one of selecting from a list of alternatives, but we prefer to think of it in terms of a dynamic process of construction and construal in which the selective aspect of processing is implicit. In constructing discourse it is almost certainly the case that the many possible options exist merely as hypothetically potential possibilities. They do not, and could not, exist as a discrete prefabricated list of alternatives to be pored over as one might go over a menu at a restaurant. Rather, these dynamic possibilities change in size and nature with the ongoing construction of the discourse. For this reason we prefer the term Systems Grammar (SG) as contrasted with any term that either implies static prefabricated structures or that fails to emphasize the dynamic aspect of diverse interacting systems (see also Oller and Giardetti 1999; Oller, Oller, and Badon 2006). To describe a transitive event involving two participants, the speaker needs to decide on the relationship between the two participants, i.e., who

The use of passives and alternatives in English

387

is the agent and who is the patient. If the action is embedded within a larger and more complex scene, the role of the participants may vary between agent and patient across distinct events. In building up a structure, the speaker must also implicitly decide on which participant to focus and must “select” a linguistic structure among a range of structural options to encode that perspective choice. Consider a simple event consisting of a swarm of bees chasing a dog. A speaker can describe the action from the perspective of the agent (1a) or from the perspective of the patient (1b). Alternatively, the speaker can choose to neutralize the roles of the participants, and describe the event as if each participant engages in an action without necessarily causing the action of the other (1c). (1)

a. b. c.

The bees were chasing the dog. The dog was being chased by the bees. The dog was running and the swarm of bees was flying.

The choice of any one of these expressive options and many others1 depends on several factors including the speaker’s parsing and construal of the sequence of events, the communicative intentions of the speaker, the discourse context in which the communication occurs, and the accessibility of available structural options “during the evanescent time frame of constructing utterances in discourse” (Slobin 1991: 12). These constructive processes are defined (or determined, etc.) by the fact that they do not come out prefabricated or all at once and by the limited time available for the construction of discourse. Recognizing the dynamics of construal in language acquisition and language use has important implications for the acquisition of the passive construction in a non-primary language (L2). Consistent with the fundamental claim, per Langacker (2002) and other researchers working within the CG framework, that linguistic expression is inherently “perspectivized,” the passive structure is an independent construction that is not derived from the active structure via a transformation. From the CG perspective, positing a transformational derivation of a passive structure from an active adds unnecessary complexity and fails to correctly represent the fact that the two structures involve distinct conceptual perspectives on the sequence of events depicted. CG argues that the “choice” of passive structure over active structure is meaning-based and is crucially subject to the dynamic construal of the participant roles involved in a typical transitive sequence of events. Although we may construe some sequences of events as intrinsically transitive from a grammatical point of view, this construal is

388


not forced on us by the sequences themselves. It is something we do with event sequences as we express them in language. According to the CG view, a passive sentence such as The dog was being chased by bees and an active sentence such as The bees were chasing the dog are not exactly synonymous. Likewise, from the dynamic perspective of SG, the conceptualizations are developed differently and have different interpretive and expressive dynamics. It is true that in both sentences, the participant encoded by the noun phrase, the bees, plays the agent role, while the participant encoded by the noun phrase, the dog, plays the patient role. However, the two sentences differ with respect to which participant is the main focal point (i.e., what the sentence is about). In the active sentence, what is construed as the focal element is the agent (i.e., the bees) and the phrase referring to them occupies the subject position. In the passive sentence, by contrast, the patient (i.e., the dog) is construed as the focal element and appears in subject position. The relative prominence accorded to the participants of the event (i.e., the bees versus the dog) accounts in part for the semantic contrast between the use of an active structure or a passive one. Once we realize that passive sentences and active sentences “represent alternative construals of the profiled event” (Langacker 1990: 13), we can see that the passive structure emphasizes, or brings to the foreground, the patient of a transitive event and at the same time de-emphasizes, and pushes to the background, the role of the agent (Shibitani 1985; Dixon 1991). We could say that in the passive, the patient is profiled more prominently; therefore, the use of the passive in narratives may help to maintain a smooth discourse flow by consistently retaining the main character as discourse topic (Berman and Slobin 1994). However, there are a range of alternative structures available in a language for the meanings and discourse functions expressed by the passive. While the passive is a prototypical structure for describing an event from the perspective of the patient, it is not the only one. In fact, the English passive comprises a family of related constructions including the full passive construction (e.g., The dog was chased by the bees), the so-called truncated passive construction (e.g., The dog was chased away), and the get-passive (e.g, The dog got chased by the bees). Describing an event from the perspective of the patient does not commit the speaker to a passive construal. The speaker may legitimately use an active sentence to describe the event, and still maintain the perspective of the patient. For example, the speaker may choose to say something like (2):


(2)

a. b. c.

389

The dog runs away because the bees are chasing him. The dog runs away while the bees are chasing after him. The dog runs howling by with this swarm of bees chasing him.

And, of course, many other possibilities exist. Learning a language, primary or non-primary, involves the mastery of an open-ended and dynamic inventory of possible meaningful linguistic constructions. A passive construction is merely one among many possible dynamic constructions. Language learning also involves learning how to use these constructions to manipulate perspective and construct coherent and cohesive connected discourse (Slobin 1994). The examination of the acquisition of passives and their equivalents in narrative and expository texts has provided insight into the process of first language acquisition (Berman 1993; Slobin 1994; Jisa and Kern 1995; AkÕncÕ 2001; Jisa et al. 2002; van Hell et al. 2005). Discovering how alternative constructions are built up within meaningful discourse contexts may help us understand how second language acquisition normally works and how we can make it work better. Understanding that active and passive sentences often stand in a paraphrastic relation to one another may be useful for learning to use passive constructions in a second language, but it cannot provide the basis for knowing which route to prefer on any given occasion. To understand how speakers of a second language use the passive, it is important to consider not only whether L2 learners can produce grammatically well-formed passive sentences, but also how they use active and passive structures in the dynamic processes of building up conceptual content in a range of discourse contexts. Slobin (1994: 341) put it this way: [A]s soon as one turns from isolated sentences to connected discourse, it is evident that passives serve well-described functions of information packaging and information flow, and that they exist as alternatives among collections of options provided by any particular language.

We agree with the gist of this statement and would only note that, of course, languages that do not have passive structures do not provide that dynamic option. We would also stress, as we have above, that the list of options does not appear anywhere in consciousness as the discourse is being produced, even in the case of the seemingly binary distinction between active and passive. The choices, as we are confident Slobin and CG theorists would agree, are implicit ones that occur in the dynamic flow of discourse production and interpretation.

390


2. Passives in second language acquisition For all of the foregoing reasons, and because they can also be challenging to L2 learners, passive constructions have attracted considerable interest among researchers and L2 teachers. According to Hinkel (2002b: 233– 234), “teaching the meanings, uses, and functions of the passive voice represents one of the thorniest problems in second language (L2) grammar instruction” and “many learners even at advanced levels often do not form passive constructions correctly and do not use passive voice in appropriate contexts.” Second language learners of English, for example, often have difficulty with the constraints on predicates that can appear in passive sentences, and L2 learners frequently produce ungrammatical sentences such as “I square danced yesterday …First we were decided partner and corner” (Watabe, Brown, and Ueta 1991: 126), where the passive form of the verb decide is the source of ungrammaticality. Second language learners also frequently have difficulty in telling when a passive is more appropriate than an active structure, in spite of the fact that they understand the almost universal function of the passive in profiling the patient of a transitive event (Langacker 1982). Seliger (1989), for example, asked 6 native English speakers (undergraduate students and secretaries) and 6 Hebrew-English bilinguals (all college students enrolled in regular courses as well as advanced composition courses for foreign students and all with a TOEFL score of 500 or above) to respond orally to four topics that were intended to serve as cues for eliciting English passive sentences. These cues included: — — — —

Describe how an omelet is made. Describe how a baby is diapered. Describe how oranges are picked and sent to market. Describe how mail is sent and delivered.

Results on all four topics showed that English speakers produced far more passives than native Hebrew speakers using English as their L2. In addition, English speakers were found to produce more passives for the last two topics than for the first two, suggesting the possibility of “topic-dependent” patterns in English passives. In contrast, Hebrew-English bilinguals were found to show no preference for any of the topics in their responses. Seliger (1989) observed that L2 learners often have difficulty in putting the L2 lexicon and grammar into appropriate use in discourse despite their


391

seeming proficiency with them in some isolated instances. This was certainly true for Korean learners of English studied by Jung (1996). Jung studied the pragmatic error patterns in English passives in argumentative essays written by Korean learners of English as a foreign language. Subjects were 200 college students divided equally between sophomores and juniors at two universities in Korea. All of them had opted to be English majors or minors and the course they were in was an elective for them. The use of passives was analyzed for discourse functional errors (e.g., violation of role prominence or violation of defocusing) or emotional (affective) function errors. Passives in Korean and English are relatively similar in discourse function. However, they differ significantly in emotional function, i.e., the use of passives to reflect the attitude of the speaker toward the described events. Passives can express emotions or subjective feelings in both English and Korean. Take English for example. The use of John got promoted suggests a sense of favorable affectedness as contrasted to the neutral John was promoted. Similarly, the use of John got killed suggests an adversative connotation. While in English the emotional function is limited to the get-passive, the emotional function is reflected in all types of passives in Korean. Results showed that Korean learners produced more emotional function errors than discourse function errors. Jung attributed this contrast to the effect of negative pragmatic transfer, i.e., the prevalence of the emotional function of the Korean passive led Korean speakers to extend the use of English passive to both favorable and adversative contexts. We may interpret this pragmatic transfer as an inappropriate use of Korean construal dynamics in English. In addition, Korean learners made a substantial number of discourse function errors, despite the similarity of the discourse functions in Korean and English passives. In their study of native language interference in the use of passives by Japanese learners of English (ESL) and English learners of Japanese (JSL), Watabe, Brown, and Ueta (1991) found that both groups of second language learners produced passive sentences that were either grammatically ill-formed or contextually inappropriate. Although the Japanese ESL learners in their samples had studied English for an average of 9.8 years and had even spent an average of 31.1 months in the United States, they still had a definite tendency “to transfer the functions of the native passive forms to the target passive forms” (Watabe, Brown, and Ueta 1991: 132). The same tendency held for JSL learners who had been studying Japanese for an average of 4 years and who had lived in Japan for 18 to 22 months. There is more than one source of L2 learner difficulty with the passive construction. For one, languages differ in the potential options they provide

392


for perspective taking on an event sequence (Fillmore 1977). In his crosslinguistic study of passives and alternatives in children’s narratives, Slobin (1994), for example, observed that English get + participle construction, Turkish agentless passives, English and German inchoative/middle intransitives, and Spanish reflective mediopassives perform similar functions as event construal devices. Another factor is that each native language conditions its speakers concerning the construal options, shaping to some extent how cognitive significance is assigned within an event sequence and how components are singled out for expressive attention. Because the conditioning for this kind of shaping occurs in childhood, it is exceptionally resistant to restructuring in adult L2 acquisition (Slobin 1996). However, a third factor that may perhaps be even more influential in shaping outcomes is that the pedagogical grammars which underlie much modern foreign language teaching commonly assume (explicitly or implicitly) that the passive is derived from the active sentence (Odlin 1994). For this reason “much of the L2 instruction associated with the passive voice includes the derivation of passive structures from active” (Hinkel 2002b: 233). The discussion of passive is often limited to the notion of voice and topicalization, without taking into account the degree of agency that gives rise to a whole family of perspective-taking devices. Learners and their teachers are offered step-by-step instruction on changing an active into a passive sentence (Master 1996; Steer and Carlisi 1998). The notion of a set of rules is considered essential for pedagogical grammar (Odlin 1994), but as Hinkel (2002a: 196) points out, “pedagogical grammar rules are frequently simplistic and do not account for the large number of cases or examples that learners come across in real life.” Often, the important element of systematicity that exists in the motivated use of the passive or active voice in connected discourse has not been appropriately understood or presented to learners in a comprehensible way. In view of these facts, perhaps it should not surprise us that L2 learners are not able to master and flexibly use the passive structure along with the full range of other perspective-taking devices in the target language. 3. The present study 3.1. Objectives The current study examined the range of linguistic devices used by Chinese EFL learners and native speakers of English to encode and express the


393

specific perspectives they choose to communicate in the process of constructing a coherent story on the basis of a wordless picture story book (Berman and Slobin 1994). The wordless picture book, Frog, Where Are You? (Mayer 1969) consists of 24 pictures portraying a series of complicated events involving the two main protagonists (a boy and his dog). They are on their way to search for the boy’s runaway pet frog. There are also four secondary characters (a ground squirrel, an owl, some bees, and a deer). The interactions between the two main characters and the four secondary characters provide a rich context for the study of passives and alternative perspective-taking devices (see Table 1 below). We focus on five episodes of the story where there is some interaction between one of the main characters and one of the secondary characters. The key episodes are described below in (3). (3) The five episodes selected for analysis Episode I

A ground squirrel comes out of a hole and bites the boy’s nose.

Episode II

The boy falls out of a tree when an owl comes out of a hole in the tree.

Episode III

The dog is running away from a swarm of bees.

Episode IV

The boy mistakes the antlers of a deer for a bush and gets carried off.

Episode V

The deer pitches the boy off a cliff and down into a pond below.

Each of the episodes invites the narrator to construe it as involving a transitive event in which the main characters (i.e., the boy or the dog) are likely to be taken as affected patients, or to be subject to the actions of the secondary characters (i.e., the ground squirrel, the owl, the bees, and the deer). But the eventual linguistic encoding of each event, according to Slobin (1994, see also Berman and Slobin 1994), is determined by four communicative decisions involving the choice of (i) topic; (ii) locus of control and effect; (iii) cognitive perspectives on the event sequence; and (iv) the degree of agency involved. The topic is always considered to be the focal point and the profiled element. The locus of control/effect consists of

394


whatever entity or entities undergo the action of the agent and experience whatever effect may be caused by that action. When a speaker chooses the expression The bees are chasing the dog, the speaker is selecting the bees as the topic and the dog as the locus of control/effect. The event perspectives refer to the three dynamic components of a transitive event, i.e., CAUSE, BECOME, and STATE. For example, The deer threw the boy into a pond = AGENT [the deer] CAUSE PATIENT [the boy] BECOME STATE [into the pond]. Slobin (1994: 345) defines the three event perspectives as follows: A Cause-View represents an event as having an actor that, in some way, causes a change of state in an undergoer [The bees are chasing the dog.]; a Become-View orients to a change of state without attribution of external causality [The dog was running.]; and a State-View orients to a state in itself [The dog was in the pond.]

The degree of agency is more difficult to define. Slobin (1994) distinguishes three levels: high, mid, and minimal degrees of agency, with a cline of others in between. Examples are given in (4). (4) Degrees of agency HIGH MID MINIMAL

The bees are chasing the dog. The dog was being chased by the bees. The dog was running away (from the bees).

Together, these event construal dimensions determine “the number of participants mentioned, syntactic roles of participants (in one or more clauses), word order and construction type, and lexical choices” (Slobin 1994: 343). The main objective of the present study, then, is to compare native speakers’ descriptions of these episodes with those of Chinese EFL learners along the dimensions just described. The comparison makes it possible to study the range of structural options used by each group to encode the varying degrees of agency and causation inherent in the construal of these episodes.


395

3.2. Method Participants The subjects included 12 advanced Chinese learners of English. These learners, aged 28–35, all started to learn English in mainland China when they entered middle school at approximately the age of twelve. They majored in English in college in China, and were enrolled in a doctoral program in an American university at the time of data collection. They had been in the U.S. for an average of 30.2 months, with a range from 2 years and 5 months to 5 years and 1 month. Twelve native speakers of English were used as a comparison group (MacWhinney 2000). Language elicitation and transcription Each participant was shown the picture book Frog, Where Are You? (Mayer 1969) page by page from the beginning to the end. Once all the pictures were shown, the researcher returned to the first page and asked each participant to tell a story based on the entire book. In an attempt to minimize interviewer control over participant narrations, only minimal instructions, such as “This is a story about a boy and a dog,” or verbal prompting, such as “What’s next?” or “What about the boy?” were given (Berman and Slobin 1994: 22–25). Each oral narrative was audiorecorded, then transcribed and coded according to the conventions of the Child Language Data Exchange System (CHILDES; MacWhinney 2000). The CHILDES system consists of a transcription protocol (CHAT) and a series of language analysis programs (CLAN). The recorded narrative texts were transcribed verbatim in clauses following the guidelines given by Berman and Slobin (1994: 655–664). A research assistant who is a native speaker of English first transcribed the recording, and, for interrater agreement, the first author reviewed all the audiotaped samples for correspondence to the transcript. Word-by-word agreement was determined to be 100%.

396


Coding Descriptions of the five episodes were coded along the four event construal dimensions suggested by Slobin (1994). The coding scheme, which is a modified version of AkÕncÕ’s (2001) scheme, is illustrated below in Table 1 with an example from the native or non-native narratives in the data. Table 1. The coding of passives and alternatives by event perspective (VIEW) and topic and locus of control-and-effect2 TOPIC AND LOCUS OF CONTROL-AND-EFFECT

EXAMPLE

CAUSE-VIEW

1

The secondary character is AGENT + TOPIC and the main character is PATIENT within a clause containing a transitive verb.

The bees were chasing the dog.

2

The main character is PATIENT + TOPIC and the secondary character is AGENT in a full get-passive construction.

The boy got bitten by a gopher.

3

The main character is PATIENT + TOPIC and the secondary character is AGENT in a full be-passive construction.

The dog was being chased by the bees.

4

The main character is PATIENT + TOPIC in a truncated passive construction, i.e., one in which the agent is not overtly mentioned.

The boy was lifted into the air.

5

The main character is ACTOR + TOPIC and the secondary character is AGENT in a subordinate clause.

The dog runs away as bees follow him.

6

The main character is ACTOR + TOPIC and the secondary character is mentioned in a peripheral phrase.

The dog is running away from the bees.


397

BECOME-VIEW

7

The main character is ACTOR and the secondary character is ACTOR in two successive clauses connected with because.

The boy tumbles down from the branch because an owl came out from the hole.

8

The main character is ACTOR and the secondary character is ACTOR in two coordinate clauses.

This owl comes out and the boy falls.

9

The secondary character is ACTOR.

An owl comes out.

10

The main character is ACTOR.

The boy fell off the tree.

STATE-VIEW

11

The main character is PATIENT + TOPIC.

The boy is stuck on the head of the deer.

NO MENTION OF THE EPISODES

12

The narrator skips the episode.

The procedures used for transcription were followed in the coding of the data by different raters, and interrater agreement was determined to be 98.5%. As Table 1 shows, there is a family of constructions in English to encode the construal of prototypical transitive events with varying degrees of salience or importance of the two participants (AGENT and PATIENT), and the three dynamic event perspectives (CAUSE, BECOME, and STATE). Where there is a decrease in the salience or importance of AGENT and CAUSE, there is a corresponding increase in the salience or importance of PATIENT, BECOME, and STATE (Slobin 1994). 3.3. Results A quantitative analysis Table 2 below shows the families of constructions and the number of times a particular construction occurred in the data. The Chinese EFL learners, like the native speakers of English in the comparison group, used transitive active constructions, passive constructions (full or agentless), and intransi-

398


tive active constructions to express various kinds of construals of the events in question.The two groups of speakers did not differ significantly in their choice of event perspectives. Neither group of speakers made much use of the STATE-VIEW. The dynamic nature of the episodes in the story line seems to exclude the STATE-VIEW. Table 2. Passives and alternatives by native speakers of English and Chinese EFL learners TOPIC AND LOCUS OF CONTROL-AND-EFFECT

NATIVE SPEAKERS OF ENGLISH

CHINESE EFL LEARNERS

CAUSE-VIEW

1

The secondary character is AGENT + TOPIC and the main character is PATIENT within a clause containing a transitive verb.

19

25

2

The main character is PATIENT + TOPIC and the secondary character is AGENT in a full get-passive construction.

5

0

3

The main character is PATIENT + TOPIC and the secondary character is AGENT in a full be-passive construction.

5

6

4

The main character is PATIENT + TOPIC in a truncated passive construction, i.e., one in which the agent is not overtly mentioned.

3

2

5

The main character is ACTOR + TOPIC and the secondary character is AGENT in a subordinate clause.

3

0

6

The main character is ACTOR + TOPIC and the secondary character is mentioned in a peripheral phrase.

3

1


399

BECOME-VIEW

7

The main character is ACTOR and the secondary character is ACTOR in two successive clauses connected with because.

4

0

8

The main character is ACTOR and the secondary character is ACTOR in two coordinate clauses.

7

5

9

The secondary character is ACTOR.

6

11

10

The main character is ACTOR.

1

4

2

0

4

8

STATE-VIEW

11

The main character is PATIENT + TOPIC.

NO MENTION OF THE EPISODES

12

The narrator skips the episode.

Both language groups also tended to take a CAUSE-VIEW more frequently than a BECOME-VIEW (38 versus 18 for native speakers of English, and 34 versus 20 for Chinese EFL students). They did not differ in frequency with respect to the CAUSE-VIEW or the BECOME-VIEW. The two groups of speakers did not differ in the choice of topic and locus of control/effect either. Both the main characters and the secondary characters were selected as topic, and typically the main characters (i.e., the boy and the dog) were seen as locus of control/effect. The two groups of speakers differed, however, in their flexibility in expressing the varying degrees of agency. We can see this more clearly in Table 3 (see below). In their descriptions of the sixty (5 episodes × 12 speakers) potential episodes, Chinese EFL learners either used transitive active sentences to express a high degree of agency (with respect to secondary characters), attributed a non-agentive-perspective towards these episodes via the use of intransitive active sentences, or made no mention of the events at all. These descriptions, which express either a high or zero degree of agency, account for 86.7% of the total number of descriptions of the five episodes. By contrast, native speakers made more frequent and

400


diverse use of a range of constructions to express various degrees of agency. Table 3. Degrees of agency expressed by native speakers of English and Chinese EFL learners NATIVE SPEAKERS OF ENGLISH

CHINESE EFL LEARNERS

DEGREE OF AGENCY

FAMILIES OF CONSTRUCTION

High

Transitive active sentences

19

25

Mid & minimal

Fall Passive, Agentless passive, lower agency in adjoining phrases or clause

20

8

Zero

Intransitive or no mention at all

21

27

A qualitative analysis Because of our small samples, it may be useful to do some more intensive qualitative analysis of the information contained in Tables 2 and 3. For instance, it is interesting to compare how native speakers of English and Chinese EFL learners chose to describe each of the five episodes in terms of degree of agency. For Episode I, which involved the boy and the ground squirrel, seven of the twelve native speakers of English mentioned the biting event. Three used the active transitive structure to do so, and four used a full (be or get) passive sentence. The latter represents a lower degree of agency than the active form. In addition, the get-passive suggests some kind of volitional control or responsibility on the part of the affected entity. By contrast, only three of the twelve Chinese EFL learners mentioned the biting event, and all of them used the active sentence structure. The rest of the speakers in both groups either only mentioned the appearance of the ground squirrel or chose not to mention the episode at all. For Episode II, which involved the boy and the owl, half of the twelve native speakers of English mentioned the boy’s fall from the tree. A variety of structures were used, and they ranged from high degree of agency to minimal degree of agency. At the high end, one speaker said There’s an


401

owl in there who bumps him [the boy] down to the ground and another speaker said An owl flew out of the hole in the tree and knocked him down out of the tree. The diminished agency of the owl in the boy’s fall is marked by mentioning the owl in an adjoined clause introduced by because. Examples of this type are illustrated in (5). (5) Lower agency in adjoining phrases or clauses in Episode II a. b. c. d.

The boy tumbles down from the branch because of an owl who’s popped up from the hole [20f].3 The boy falls out of the tree because an owl came out of the hole [20g]. The little boy falls off the tree frightened by an owl [20h]. And the boy falls off the tree because the owl came out of the hollow part [20l].

One native speaker used a coordinate structure to describe the boy’s fall, thus diminishing the degree of agency even further: This owl comes out and the boy falls. While the listener may infer that the boy’s fall from the tree is due to the (sudden) appearance of the owl, the consequence-cause relation is not made explicit. By contrast, only four of the twelve EFL learners mentioned the boy’s fall at all. They used either a transitive active clause, An owl came out and pushed boy down and the boy fell to the ground, a full be-passive, The boy was frightened off the tree by the owl, or a coordinate structure, The owl flew out and Sam fell down. None of the EFL learners used adjoined phrases or clauses to express diminished degree of agency. Again, we see evidence that our Chinese EFL learners, even with years of formal instruction as well as years of exposure in the target language environment, are still struggling with the use of linguistic devices to express varying degrees of agency. Next, let us consider Episode III, which involved the dog and the swarm of bees. The native speakers of English used a variety of structures to describe the chasing event with varying degrees of agency. Half of the native speakers used active clauses, such as The bees start chasing the dog, and two of them used the full passive form, such as The dog gets/is chased by the bees. Or, sometimes, they embedded the transitive event in a relative clause as when one speaker said It looks like he got bit by a couple of bees that were chasing them. This structure allows these speakers to keep the dog as the topic and locus of control-and-effect. As in the previous two

402


episodes, a salient feature in the productions by native speakers of English is that they frequently mentioned the bees in an oblique phrase, indicating that the bees were contributing some degree of agency to the event, as in (6). (6) Lower agency in adjoining phrases or clauses in Episode III a. b. c. d. e.

And the dog runs howling by with this swarm of bees chasing him [20f]. The dog runs away as bees follow him [20h]. The dog is running away from all the bees [20c]. The dog was running away from the bees [20d]. So both of them are in kinda chase scenes running away from these other animals [20e].

The Chinese EFL learners, however, used either the active structure with the bees as topic and the dog as locus of effect in the form The bees began to chase the dog, or the by-passive construction with the dog as the topic and locus of effect in the form The dog was chased by the bees. The active structure was used by seven of the EFL learners, while the passive structure by the remaining five EFL learners. Again, we see evidence that our Chinese EFL learners are less flexible in exploiting the range of perspective-taking devices than are the native speakers of English. The native speakers also used a wide variety of structures to represent the event sequence in Episode IV, where the boy ended up in the deer’s antlers. These structures included: active sentences, The deer picks him up; past participle structures, The deer takes off with the boy strewed across his antlers; the full be-passive, Now the boy has been picked up by some antlered beast; a mediopassive, The boy gets caught/stuck/snagged on a deer; a stative, The little boy is stuck on his head; and one or more inchoative motion verbs, The boy falls on the stag. One speaker also used the adverb unintentionally evidently to express a diminished degree of agency, The deer pops out of the rock after being disturbed starts giving him a ride unintentionally. While all the native speakers mentioned the event in one way or another, three Chinese EFL learners did not mention it at all. The structures used by the nine EFL learners who did mention it were again less variable and less complex than those used by the native speakers. Four learners described the event from the perspective of the deer as topic and source of agency, using active sentences such as, The deer carried Tom on his head.


403

Three learners saw the event from the perspective of the boy as the subject of a verb of motion, suggesting either high agency with respect to the boy, e.g., The boy climbed unto the head of the deer, or lessened agency, marking the boy’s lack of volition, e.g., The boy fell on the horns of the deer. In either case, however, these three learners did not attribute any agency to the deer. The same was true for the use of an agentless passive by one EFL learner, Suddenly he was lift[ed] up. Another structure used involved a prepositional phrase indicating the state of the boy, The deer starts running with the boy in between his horns. Finally, evidence for the disparity between the native speakers of English and the EFL learners was also found in the descriptions of Episode V, in which the deer throws the boy off a bluff into a pool or stream of water below. Five native speakers chose to describe this sequence of events with a single active clause, e.g., The deer threw the boy off the cliff into the pond. Two native speakers used full get-passives, and one used full bepassive. Another native speaker used the complex verb cause to in a relative clause to attribute external causation to the event: The deer stops abruptly which causes the boy to lose his balance and fall with the dog down into the stream or a little puddle. The other three native speakers chose to describe the same episode with a coordinate structure and to leave it to the reader to infer the relationship between the action of the deer versus that of the dog: The deer runs with the boy on his head and stops at a cliff and the boy and the dog fall into a lake. The majority of our EFL learners (11 of 12) described the episode in a transitive active clause, e.g., The deer threw the boy into the river. The remaining EFL learner used a coordinate structure: The deer stops at the cliff and Sam and Bobby fall down into the pond. In summary, the Chinese EFL learners in the present study made less flexible use of the range of perspective-taking devices that are available in the target language than did the native speakers of English. In particular, they exhibited difficulty in expressing varying degrees of agency in the descriptions of the potentially transitive causative event sequences that we focused on in the study. 4. Conclusions and discussion It is frequently observed in studies of L2 acquisition that learners may be able to acquire the formal aspects of a language and achieve a high level of proficiency in the phonological and syntactic components of the target

404


system, but still struggle with appropriate use of the full range of structural options in discourse. Knowing a language does not guarantee accurate, appropriate, and fluent use of the language (Chen 2004). Yoshida (1990: 20 [cited in Kecsceks 2002: 183]) has made this very clear when he says although I might have knowledge of what to say with whom in what circumstances, that does not necessarily mean that I am able to perform accordingly. Moreover, even if I could perform in an ‘American’ way if I consciously strived to do so, that does not mean that I feel comfortable doing so.

Seliger (1989: 33) observes that “the inability to acquire the semantically constrained distributional rules concerned with the selection of a particular form from among options may be one of the characteristics of foreignism that remain.” This observation was also supported in the present study of the linguistic encoding of perspective taking through passives and alternatives in a story generation context. It seems clear that the frog stories actually provided rich opportunities for our Chinese EFL participants and the native comparison group to demonstrate a wide variety of distinct construals of a series of events. Each of the event sequences involved two animate entities. Each sequence, as the native speakers (and many of the Chinese EFL subjects as well) demonstrated, was susceptible to coding in a variety of structural devices. Of course, the results we have reported are constrained by our relatively small sample size, and also by the particular story-line as constructed by Mayer (1969). These factors suggest caution in making generalizations, but we believe that our findings give insight into the way passives and other structural alternatives are used in encoding sequences of events. The ability to flexibly construe events in narrative requires both the cognitive ability to consider distinct perspectives on an event sequence and the linguistic flexibility to express those distinct perspectives through appropriate structures from a wide range of options that may be afforded by any given language. The examination of topic, locus of control-and-effect, and event perspectives (i.e., CAUSE-VIEW, BECOME-VIEW, and STATEVIEW) show that our Chinese EFL learners, like our comparison group of native speakers of English, all clearly do have the cognitive flexibility to be able to take different perspectives on events. Both groups of speakers are capable of attending to agent and change of state as separate components of an event, for instance, and we know that these are fundamental to the selection of event perspective.


405

While the cognitive capacity for selection of event perspective emerges in very young children (Clark 2004), the choice of event perspective in narrative is more demanding (Slobin 1994). Even nine-year-old children experience difficulty in arriving at selection of event perspectives, and the flexible use of expressive options to express perspective in connected discourse is part of later language development (van Hell et al. 2005). Of course, our Chinese EFL learners have developed the cognitive capacity and skills in perspective choice as part of their acquisition of their mother tongue. Therefore, the cognitive demand on selection of event perspective in narrative construction should, by itself, pose no greater difficulty for our Chinese EFL learners than for the native speakers of English. What is more demanding for our EFL learners, however, is to represent the different perspectives appropriately via the rich range of linguistic devices that are provided, at least in potential, by the target language, in this case by English. The reasons are manifold. First, each native language has trained its speakers how to choose to construe and to attribute differentiated cognitive significance to an event and its parts for expressive purposes. Since this training is carried out in childhood, it is exceptionally resistant to restructuring in adult second language acquisition (Slobin 1996). The Chinese saying shì zài rén wéi (thing-exist-man-do, meaning “It all depends on human effort”) is sometimes used to characterize the Chinese way of thinking. This cultural way may lead Chinese speakers to emphasize the agent of an action, and the Chinese language to favor active sentences over passive sentences (Lian 1993). Furthermore, it has been proposed that sentence patterns are structured under completely different sets of conceptual systems in English and Chinese (Tai 2003). One may wonder whether the difficulty with the flexible use of passives and alternatives experienced by our Chinese EFL learners is due to a conceptual difference or to a constructional difference between the native language and the target language4. To address this concern, we need to compare the use of passives and alternatives in English and Chinese in the same narrative context. Preliminary results suggest the effect of conceptual transfer. Chen and Guo (2006), in their developmental study of passive and alternatives in Chinese narratives (i.e., frog stories) observe that the twelve adult Chinese speakers used the following five types of constructions in their descriptions of the five episodes examined in the present study.

406


(7) Passives and alternatives in Chinese frog narratives a.

b.

c.

d.

e.

Simple active transitive sentence YƯ dà qún mìfƝng zhuƯ-zhƝ g΅u. one big group bees chase-ZHE dog ‘A swarm of bees were chasing the dog.’ BA-Construction Lù b΁ zhè xi΁o nánhái cóng xuányá shàng rƝng-dào. above throw-arrive deer BA the little boy from cliff he l΃ qù le. river inside go LE ‘The deer threw the little boy down the cliff into the river.’ BEI-construction (the so-called passive sentence) Tǌrán tƗ bèi (lù) j·-le-q΃-lái. suddenly he BEI deer lift-LE-rise-come ‘Suddenly he was lifted up (by the deer).’ Coordinate structure with cause-effect relation implied MƗotóuyƯng fƝi-le-chǌ-lái. owl fly- LE-exit-come Xi΁ohái xià-de cóng shù shàng diào-le-xià-lái. boy scare- DE from tree above fall- LE-descend-come ‘The owl flew out. The boy was so scared to fall down the tree.’ Simple intransitive sentence Zhè xi΁o nánhái cóng shù shàng diào-le-xià-lái. The little boy from tree above fall- LE-ascend-come ‘The little boy fell down from the tree.’

Speakers who use structures such as (7a) and (7b) express a high degree of agency while those who use structures like (7e) or make no mention of this event at all express a low degree of agency. The structures such as (7c) and (7d) sit in the middle. Out of the sixty possible descriptions of the five episodes in the Chinese narratives, twenty-five expressed a high degree of agency, and twenty-four expressed a low degree of agency leaving the remaining eleven to express a mid-degree of agency. This pattern is very similar to the one we found for our Chinese EFL learners in their expression of varying degrees of agency in English. The effect of conceptual transfer in event construals is intrinsically related to the linguistic packaging of event construals. For native speakers the rich variety of linguistic forms required for flexible event construal is available from early on (Berman and Slobin 1994). Native speakers de-


407

velop habitual patterns of event construal and are at least theoretically capable of making “conventionalized choices of grammatical devices” to encode their conceptualizations in discourse (see, e.g., Achard 2004: 185). By contrast, EFL learners, or L2 learners of any language, presumably can be expected to have difficulty in differentiating and using the full potential range of linguistic options for the expression of perspectives on event sequences in narrative. The pressure of online production of narrative discourse was presumably higher for the EFL learners in this study, and for L2 learners in general, and may prevent them from accessing and integrating the linguistic constructions online.5 5. Pedagogical implications A close look at the way the two groups expressed (or did not express) degree of agency reveals that our Chinese-speaking EFL learners, in spite of their years of both classroom-based English learning in China and more natural exposure to the English language and culture in the United States, still come up short of the full potential of the English language in construal of subtle aspects and relations within event sequences. In their connected discourse we see room for growth in linguistic flexibility. We also see evidence in favor of our starting premise that even skilled adult L2 learners may have difficulty in using subtly distinct grammatical constructions. We believe that the results of the present study have some pedagogical implications. English learners in the present study had not only formally studied EFL in classrooms but had also been immersed in an Englishspeaking environment for substantial periods of time. However, the subtleties of passives and other alternatives to express varying degrees of agency still appear to be just beyond reach. This provides further evidence for the observation that mere exposure to the target language in a naturalistic setting may not be enough. Rather, there is a need to “get the learners to (re)discover the motivated structures and principles that govern a foreign language” (Pütz, Niemeier and Dirven 2001: xv). We cannot rely exclusively on unconscious learning. As Pütz, Niemeier, and Dirven (2001: xv) put it, “both processes – unconscious acquisition and awareness in learning – go hand in hand and are always present in language instruction scenarios, albeit in widely varying degrees.” Optimal language development, in both L1 and L2, requires both significant immersion in the target language (Kecskes 2001) and interactive exposure to large quantities of natural speech in context (Langacker 2001).

408


The present study also provides empirical support for the idea that grammar “can no longer be viewed as a central, autonomous system to be taught and learned independent of meaning, social function, and discourse structure” (Celce-Murcia 1991: 466). Given cognitive linguistics’ emphasis on the physical, social, cultural, and linguistic contexts of language and language development, it is only natural to situate the teaching of grammar within cultural, conceptual, functional, social, and discourse contexts. If the acquisition and use of passive constructions involved only the formulation of a rule of passive transformation, we would expect that the native speakers of Chinese who were L2 English learners in the present study should have little or no problem achieving the full range of possible structures. None of them had any difficulty producing well-formed passive constructions. However, the capacity to produce such well-formed structures, or to comprehend even more complex ones, does not guarantee the full range of flexibility that is attained by native speakers of English. It seems that we do not need to teach students how to formulate a rule of passive formation, nor do we need to prescribe steps for them to follow in order to change an active sentence to a passive sentence (Steer and Carlisi 1998; Master 1996). What we need to do instead is to make rich event sequences accessible and to enable learners to discover how native speakers of the target language conceptualize the various ways that participants in an event sequence can be profiled, and, more particularly, how degrees of agency can be differentiated. Instead of focusing merely on producing well-formed structures, we believe that foreign language teaching should “allow the learner to focus on the conditions that motivate specific structural choices” (Achard 2004: 180). By “conditions” we have in mind the kinds of pragmatic considerations, mostly subconscious, that evidently influence the choices and decisions made by native speakers in the processes of event construal. That is, learners should pay attention to who or what is being emphasized, focused on, or foregrounded, and what the attitude of the producer of the discourse is to the various elements (phonological, lexical, syntactic, semantic, and pragmatic) that are being coordinated in the production of the surface forms of the discourse. These aspects of language are deeply dependent on the inferential understanding of discourse relations. Although these conditions are largely determined subconsciously some, such as agency, can evidently be finely distinguished within the context of distinct construals. Teachers should provide richly developed contexts of experience that will allow L2 learners to discover these fine distinctions on their own as they experience the pragmatic relations between surface forms and deeper meaningful interpretations and intentions. We


409

have made some specific suggestions about how such subtle language acquisition problems can be solved in our paper on the use of well-selected sound-motion pictures (see Chen and Oller 2005). This present study has demonstrated empirically that the linguistic packaging and the dynamic construal of event sequences is certainly challenging even to very advanced and highly experienced L2 learners. We believe it is desirable, even necessary, therefore, to show how structural devices discussed here and elsewhere (see Chen and Oller 2005) present different degrees of agency and how subtle differences in form and meaning are deployed and exploited in different contexts. While exposure to the contexts of interaction (that is, whatever is talked about in discourse production) greatly enhances L2 acquisition across the board, the teacher in the foreign language classroom and the conscious effort exerted by the learner also play essential roles in enabling second language learners to decipher and fully unpack meaning relations and their linguistic construal in target language packages (Chen and Oller 2005). With that in mind, we believe that improvement in the acquisition of the deeper subtleties of the constructions investigated in the present study will follow from greater exposure to and increased attention to the differences in speaker construal of passives and alternative structures in L2 teaching. We have elaborated elsewhere how high-quality sound-motion pictures (as contrasted with still picture vignettes) can be used to illustrate and elicit appropriate linguistic forms identifying every essential argument in an event and how it can be done “from multiple perspectives and with differing packaging” (Chen and Oller 2005: 275). Language teachers in the classroom can in fact take advantage of Frog, Where Are You?, which is available on DVD or VHS from Scholatics Weston Woods to familiarize students with varying degrees of agency and the variety of expressive alternatives. Although there are no words in the frog story DVD, native-speaker narration can easily be added. After viewing the video, students can be encouraged to tell the story in their own words in the target language. The teacher may even use the transcripts of the narration of the wordless picture book Frog, where are you? (Mayer 1969), which are available in the CD included in MacWhinney (2000), and encourage students to compare their own narration with the narrations produced by native speakers. During this process, the teacher can orient the students toward the possible event construals and their linguistic packaging. As a pedagogical practice, making the systematicity, non-arbitrariness, and motivations of linguistic structures explicit to L2 learners in the classroom,6 and embedding L2 teaching and learning within its natural cultural contexts will help motivate learners and provide an au-

410


thentic way to develop ordinary intuitions concerning the target language. Of course, these suggestions on the teaching of construals and their linguistic packaging, as well as our suggestion regarding the pedagogical significance of consciousness-raising, call for carefully designed and wellcontrolled empirical studies. Otherwise, they may fare no better than the long criticized “armchair” fantasies of the past. Cognitive linguistics is a rapidly growing area of study. Taylor (1995: 19) conservatively predicted that cognitive linguistics is an approach that “is like to exert an increasing influence on the direction of linguistic research for some years to come.” Ten years later, Taylor (2004: 17) has said that cognitive linguistics “has established itself as a viable alternative to the Chomskyan paradigm, and indeed, at the beginning of the twenty-first century, is on the cusp of entering the mainstream.” In fact, it is quite likely that “more linguists around the world do cognitive linguistics than do generative grammar” (Newmeyer 2003: 683). Yet studies that focus on the application of cognitive linguistic theory to SLA research and pedagogy are scarce (Niemeier 2004). This is unfortunate, given that the effectiveness of pedagogical application is considered an important empirical test for cognitive linguistic theory (Langacker 2001: 3). It is our hope that the present study will stimulate SLA researchers and language pedagogues to rethink theories of language acquisition view and also to investigate the pedagogical applications of cognitive linguistics. Acknowledgements We are deeply indebted to Dr. Dan Slobin for his pioneering cross-linguistic developmental studies on alternative means of event packaging in narratives, especially the use of passives and related constructions. We also wish to thank the editors and anonymous reviewers for their insightful comments on prefinal versions of this paper. The help from Dr. Harriet Jisa and Ms. Katie Coleman is greatly appreciated as well.

Notes 1.

For example, as one reviewer points out, we may represent a different construal of the event with a simple addition of the word behind to (1c). According to the reviewer, the sentence The dog was running and the swarm of bees


2.

3. 4. 5.

6.

411

was flying behind carries the conventional implicature of a causal link, which we think (1c) does not necessarily have. Jisa et al. (2002: 170) point out that “[T]he number of alternative options provided by the different languages can either increase or decrease the functional load attributed to passive constructions.” We agree. In principle, other constructions such as cleft sentences or left dislocation can assume some of the functions of English passives and alternatives. In fact, most of the structures here can be rephrased as clefts or dislocations. However, no cleft or left dislocation structures are found in the data. Here and throughout the paper, the number and letter combination in brackets following each example is the subject identification number. We thank one of the reviewers for bringing this point to our attention. The greater pressure experienced by our Chinese EFL learners during their online production of narrative discourse in English can be seen from their slower and more deliberate speech style as compared with the more conversational style of the native speakers. We thank one of the reviewers for pointing this out. Happily, as the research shows, teachers do not necessarily need to resort to explicit instruction concerning motivations to make the systematicity and nonarbitrariness of linguistic structures explicit to L2 learners. As Chen and Oller (2005) point out, there are ways of dealing with the subtleties and dynamics of pragmatic language use quite indirectly through well-acted narratives, real conversations, and the like. In addition, the fast-growing Internet and multimedia technology make it possible for foreign language learners to gain interactive exposure to large quantities of natural speech in context.

References AkÕncÕ, Mehmet-Ali 2001 Development of perspective in narrative texts of Turkish-French bilingual children in France. In Research on Child Language Acquisition: Proceedings of the 8th conference of the International Association for the Study of Child Language, Vol. 1, Margareta Almgren, Andoni Barreña, María-José Ezeizabarrena, Itziar Idiazabal, and Brain MacWhinney (eds.), 57–77. Somerville: Cascadilla Press. Achard, Michel 2004 Grammatical instruction in the natural approach: A cognitive grammar view. In Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching, Michel Achard and Susanne Niemeier (eds.), 165–194. New York: Mouton de Gruyter.

412


Berman, Ruth A., and Dan I. Slobin 1994 Relating Events in Narrative: A Cross-Linguistic Development Study. Hillsdale, New Jersey: Lawrence Erlbaum. Celce-Murcia, Marianne 1991 Grammar pedagogy in second and foreign language teaching. TESOL Quarterly 25 (3): 459–480. Chen, Liang 2004 Review of Applied Cognitive Linguistics. Vols. 1 and 2 by Martin Pütz, Susanne Niemeier, and René Dirven (eds.) (2001). Applied Linguistics 25 (3): 440–443. Chen, Liang, and Jiansheng Guo 2006 A developmental study of passives and alternatives in Mandarin Chinese narratives. Manuscript. University of Georgia and California State University at East Bay. Chen, Liang, and John W. Oller, Jr. 2005 Using well-chosen motion pictures in L2 classrooms. Canadian Modern Language Review 62 (2): 263–284. Clark, Eve V. 2004 How language acquisition builds on cognitive development. Trends in Cognitive Sciences 8: 472–478. Dixon, Robert M.W. 1991 A New Approach to English Grammar, on Semantic Principles. Oxford: Clarendon Press. Fillmore, Charles 1977 Scenes-and-frames semantics: Linguistic structures processing. Fundamental Studies in Computer Sciences 5: 55–81. Hinkel, Eli 2002a Teaching grammar in writing classes: tenses and cohesion. In New Perspectives on Grammar Teaching in Second Language Classrooms, Eli Hinkel and Sandra Fotos (eds.), 181–198. Mahwah, New Jersey: Lawrence Erlbaum. 2002b Why English passive is difficult to teach (and learn). In New Perspectives on Grammar Teaching in Second Language Classrooms, Eli Hinkel and Sandra Fotos (eds.), 233–260. Mahwah, New Jersey: Lawrence Erlbaum. Jisa, Harriet, and Sophie Kern 1995 Discourse organization in French children’s narratives. In Proceedings of the 26th Annual Child Language Research Forum, Eve V. Clark (ed.), 177–188. Stanford, California: Center for the Study of Language and Information Publications. Jisa, Harriet, Judy Reilly, Ludo Verhoeven, Elisheva Baruch, and Elisa Rosado 2002 Passive voice constructions in written texts. Written Language and Literacy 5: 163–182.


413

Jung, Woo-Hyun 1996 A pragmatic analysis of English passives in second language acquisition. Paper presented at the 10th Annual Meeting of the International Conference on Pragmatics and Language Learning. March 22–23, Urbana, Illinois: University of Illinois. Kecskes, Istvan 2001 The “graded salience hypothesis” in second language acquisition. In Applied Cognitive Linguistics, Vol. 1: Theory and Language Acquisition, Martin Pütz, Susanne Niemeier, and René Dirven (eds.), 249– 269. Berlin/New York: Mouton de Gruyter. 2002 Situation-Bound Utterances in L1 and L2. Berlin/New York: Mouton de Gruyter. Langacker, Ronald W. 1982 Space grammar, analyzability and the English passive. Language 58: 22–80. 1987 Foundations of Cognitive Grammar: Theoretical Prerequisites. Stanford, California: Stanford University Press. 1990 Concept, Image, and Symbol: The Cognitive Basis of Grammar. Berlin: Mouton de Gruyter. 2001 Cognitive linguistics, language pedagogy, and the English present tense. In Applied Cognitive Linguistics, Vol. 1: Theory and Language Acquisition, Martin Pütz, Susanne Niemeier, and René Dirven (eds.), 3–39. Berlin/New York: Mouton de Gruyter. 2002 Concept, Image, and Symbol: The Cognitive Basis of Grammar. 2nd ed. New York: Mouton de Gruyter. Lian, Shuneng 1993 Ying Han Duibi Yanjiu (Contrastive Studies of English and Chinese). Beijing: Chinese Youth Press. MacWhinney, Brian 2000 The CHILDES Projects. 3rd ed. Mahwah, New Jersey: Lawrence Erlbaum. Master, Peter 1996 Systems in English Grammar: An Introduction for Language Teachers. Englewood Cliffs, New Jersey: Prentice Hall. Mayer, Mercer 1969 Frog, Where Are You? New York: Dial Press. Newmeyer, Frederick J. 2003 Grammar is grammar and usage is usage. Language 79 (4): 682–707. Niemeier, Susanne 2004 Applied cognitive linguistics and newer trends in foreign language teaching methodology. In Language in Use: Cognitive and Discourse Perspectives on Language and Language Learning, Andrea Tyler, Mari Takada, Yiyoung Kim, and Diana Marinova (eds.), 100–

414


111. Washington, District of Columbia: Georgetown University Press. Odlin, Terence (ed.) 1994 Perspectives on Pedagogical Grammar. New York: Cambridge University Press. Oller, John W., Jr., and J. Roland Giardetti 1999 Images that Work: Creating Successful Messages in Marketing and High Stakes Communication. Westport, Connecticut: Quorum Books. Oller, John W., Jr., Liang Chen, Stephen D. Oller, and Ning Pan 2005 Empirical predictions from a general theory of signs. Discourse Processes 40 (2): 115–144. Oller, John W., Jr., Stephen D. Oller, and Linda C. Badon 2006 Milestones: Normal Speech and Language Development across the Life Span. San Diego, California: Plural Publishing, Inc. Pütz, Martin, Susanne Niemeier, and René Dirven (eds.) 2001 Applied Cognitive Linguistics. Vol. 1: Theory and Language Acquisition. Berlin/New York: Mouton de Gruyter. Seliger, Herbert W. 1989 Semantic transfer constraints on the production of English passives by Hebrew-English bilinguals. In Transfer in Language Production, Hans W. Dechert and Manfred Raupach (eds.), 21–33. Norwood, New Jersey: Ablex Publishing Corporation. Shibatani, Masayoshi 1985 Passives and related constructions: A prototype analysis. Language 61 (4): 821–848. Slobin, Dan I. 1991 Learning to think for speaking: Native language, cognition, and rhetorical style. Pragmatics 1: 7–26. 1994 Passives and alternatives in children’s narratives in English, Spanish, German, and Turkish. In Voice: Form and function, Barbara A. Fox and Paul J. Hopper (eds.), 341–364. Amsterdam/Philadelphia: John Benjamins. 1996 From “thought and language” to “thinking to speaking.” In Rethinking Linguistic Relativity, John J. Gumperz and Stephen C. Levinson (eds.), 70–96. Cambridge: Cambridge University Press. Steer, Jocelyn M., and Karen A. Carlisi 1998 The Advanced Grammar Book. Boston, Massachusetts: Heinle & Heinle. Tai, James 2003 Cognitive relativism: Resultative construction in Chinese. Language and Linguistics 4: 301–316.


415

Taylor, John R. 1995 Linguistic Categorization: Prototypes in Linguistic Theory. 2nd ed. Oxford: Clarendon Press. 2004 Linguistic Categorization: Prototypes in Linguistic Theory. 3rd ed. Oxford: Oxford University Press. Tomasello, Michael 2003 Constructing a Language: A Usage-Based Theory of Language Acquisition. Cambridge, Massachusetts: Harvard University Press. Watabe, Masakazu, Cheryl Brown, and Yumiko Ueta 1991 Transfer of discourse function: Passives in the writing of ESL and EJL learners. International Journal of Applied Linguistics 29: 115– 134. van Hell, Janet G., Ludo Verhoeven, Marjan Tak, and Moniek van Oosterhout 2005 To take a stance: A developmental study of the use of pronouns and passives in spoken and written narrative and expository texts in Dutch. Journal of Pragmatics 37: 239–273. Yoshida, Kensaku 1990 Knowing vs. behaving vs. feeling: Studies on Japanese bilinguals. In Language proficiency: Defining, Teaching, and Testing, Louis A. Arena (ed.), 19–40. New York: Plenum Press.

Author index A Aarts, Jan 100, 107, 111 Abbot-Smith, Kirsten F. 199, 225 Achard, Michel 25, 29, 30, 33, 88, 95, 111, 117, 150, 226, 229, 233, 252, 286, 287, 291, 320, 326, 328, 334, 352, 353, 407, 408, 411 Ahrens, Kathleen 198, 226 Aid, Frances 156, 180, 193 Aijmer, Karen 113, 115 AkÕncÕ, Mehmet-Ali 389, 396, 411 Alarcos Llorach, Emilio 149 Alatis, James E. 253 Alexander, Louis G. 330, 352 Alexander, Richard G. 233, 252 Allen, Linda Q. 286 Allwood, Jens 9, 30 Allwright, Richard 81, 82, 88 Almgren, Margareta 411 Alonge, Antonietta 116 Alonso, Amado 132, 149, 156, 193 Alonso, Rosario 156, 193 Altenberg, Bengt 113, 115 Anagnostopoulou, Elena 206, 226 Andersen, Roger 261, 286, 360, 382 Anderson, Bruce 105, 111 Andrade, Magdalena 156, 196 Andronis, Mary 64 Arena, Louis A. 415 Armour, William S. 237, 252 Asher, James J. 280, 286 Aske, John 267, 287 Aston, Guy 116 B Badon, Linda C. 386, 414 Baicchi, Annalisa 137, 152 Ball, Christopher 64

Barcelona, Antonio 198, 226 Barddal, Johanna 321 Bardovi-Harlig, Kathleen 117, 326, 328, 330, 331, 352, 353 Barlow, Michael 25, 30, 33, 354 Barreña, Andoni 411 Baruch, Elisheva 412 Batstone, Rob 82, 88 Bavin, Edith L. 269, 287 Behrens, Heike 199, 225 Bencini, Giulia 197, 198, 199, 201– 203, 205, 224, 226, 227 Benjamín, Carmen 156, 193 Benveniste, Emile 161, 193 Berg, Thomas 152, 227 Berlin, Brent 61 Berman, Ruth A. 260, 269, 271, 287, 298, 320, 388, 389, 393, 395, 406, 412 Bernardini, Silvia 116 Bertinetto, Pier Marco 360, 382 Bialystok, Elen 122, 150 Biber, Douglas 58, 61, 93, 100, 108, 112 Bleam, Tonia 206, 224–226 Bley-Vroman, Robert 121, 149 Blomberg, Johan 268, 294 Bloom, Paul 34 Boas, Hans C. 33 Bock, Kathryn 198, 227 Boers, Frank 15, 30, 95, 112, 223, 226, 233, 242, 252, 333, 353 Bofman, Theodora 326, 328, 352 Bogard, Sergio 192, 193 Bohn, Ocke-Schwen 122, 149 Bolinger, Dwight 42, 61 Boogaart, Ronny 374, 375, 382 Bowerman, Melissa 298, 320 Boxtel, Sonja van 380, 382

418

Author index

Brady, Imelda Katherine 226 Bresnan, Joan 194 Bright, William 382 Broccias, Cristiano 2, 29, 67, 68, 73, 83, 86, 88, 125, 223, 224 Brown, Cheryl 390, 391, 415 Brown, Keith 227, 289 Brugman, Claudia 9, 30, 51, 61 Brumfit, Christopher 88 Burnard, Lou 94, 112 Burt, Marina 39, 62 Butt, Jon 156, 193 Bybee, Joan 198, 222, 226, 292, 362, 369, 373–375, 377, 381, 382 Byrnes, Heidi 384 C Caballero, Rosario 278, 287 Cabrera, Mónica 224, 226 Cadierno, Teresa 3, 106, 259, 260, 265, 268, 271–276, 278, 286, 287, 293, 296, 320, 321 Callies, Marcus 209, 211, 213, 215, 216, 225, 226 Cameron, Lynn 233, 252, 253 Cameron, Richard 226 Canteli Dominicis, María 156, 193 Carlisi, Karen A. 392, 408, 414 Carlon, Keith 112 Carr, Gerald F. 324 Carr, Thomas 40, 63 Carroll, Mary 358, 370, 371, 378, 382, 384 Carter, Ronald 82, 90, 95, 100–103, 112, 115 Casad, Eugene H. 288, 323 Castañeda, Alejandro 156, 157, 193 Celce-Murcia, Marianne 408, 412 Cenoz, Jasone 254 Chafe, Wallace 17, 30 Chang, Franklin 198, 227 Charteris-Black, Jonathan 233, 253

Chen, Liang 3, 81, 88, 94, 101, 108, 385, 386, 404, 405, 409, 411, 412, 414 Chen, Rong 85, 88 Chilton, Paul 294 Chomsky, Noam 3, 38, 42–44, 53, 61, 82, 97, 411 Chung, Ting Ting Rachel 206, 227 Cicogna, Caterina 256 Clark, Eve V. 405, 412 Coleman, Katie 410 Comrie, Bernard 228, 338, 353, 358, 368, 374, 382 Connor, Ulla 94, 112, 113, 228, 232, 253 Conrad, Susan 61, 93, 112 Constenla, Adolfo 192, 193 Cook, Vivian J. 43, 53, 61 Criper, Clive 288 Croft, William 2, 4, 29, 60, 61, 67, 70, 75–77, 88, 94, 112, 145, 149, 190, 193, 206, 224, 227, 296, 321 Cruse, D. Alan 76, 88, 94, 112, 224, 227 Cuenca, Isabel 10, 31 Culicover, Peter 81, 88, 89 Cuyckens, Hubert 30, 35, 64, 195, 227, 322 D Dagneaux, Estelle 228 Dahl, Östen 358, 382, 383 Danesi, Marcel 2, 3, 10, 30, 231– 234, 236–238, 242, 244, 253, 256, 295 Dapremont, Elena M. 12, 30 David, Caroline 294 Davidse, Kristin 112 Davies, Alan 288 Davis, Philip W. 35 de Bot, Kees 290 De Knop, Sabine 1, 3, 95, 106, 149, 295

Author index de los Ángeles Gómez-González, 152 De Rycker, Teun 1, 258 de Saussure, Ferdinand 64 Dechert, Hans W. 414 Deignan, Alice 244, 253 Del Giudice, Alex 82, 89 Delbecque, Nicole 177, 193 Delfitto, Denis 360, 382 Demonte, Violeta 206, 227 Di Meola, Claudio 307, 321 Di Pietro, Robert J. 244, 253 Díez, Olga 129, 152 Dimroth, Christine 382 Dirven, René 3, 30–33, 35, 38, 42, 48, 51, 52, 54, 55, 58, 61–64, 88, 95, 105, 106, 112, 116, 122, 123, 125, 141, 149, 150–152, 226, 227, 233, 234, 236, 248, 252, 253, 256, 292, 295, 296, 307, 321, 323, 338, 343, 345, 352– 355, 407, 412–414 Dixon, Robert M.W. 42, 57, 62, 388, 412 Doughty, Catherine J. 227, 262, 263, 287, 288, 290, 293 Draye, Luk 307, 321 Dryer, Matthew S. 228 Duffley, Patrick J. 59, 62 Dulay, Heidi 39, 62 Duncan, Susan D. 285, 290 Dürich, Kristiane 326, 328, 353 Dvorak, Trisha R. 290 E Eckman, Fred R. 293, 352 Eddington, David 226, 290 Egan, Thomas 59, 62 Egasse, Jeanne 156, 196 Ellis, Nick C. 29, 30, 95–97, 113, 222, 224, 227, 263, 274, 287, 288, 348, 353

419

Ellis, Rod 39, 40, 62, 81, 89, 96, 113, 261, 274, 275, 282, 285, 288, 346, 347, 353 Elston, Heidi 64 Endresen, Rolf T. 33 Engberg-Pedersen, Elisabeth 269, 288 Escobedo, Joyce 59, 64 Evans, Vyvyan 9, 35, 82, 90, 95, 113, 117, 224, 227, 265, 284, 288, 294, 329, 332, 355 Eyckmans, June 95, 112 Ezeizabarrena, María-José 411 F Fauconnier, Gilles 13, 30, 31, 34, 67, 79, 80, 89, 160, 193, 234, 254, 336, 338, 348, 353 Ferguson, Charles A. 324 Fey, Marc 361, 383 Fillmore, Charles 26, 31, 75, 126, 150, 224, 227, 265, 288, 392, 412 Finegan, Edward 112 Fitch, W. Tecumseh 285, 288 Fitzpatrick, Eileen 94, 113 Flecken, Monique 3, 329, 343, 344, 357, 373, 378, 383 Fotos, Sandra 89, 115, 263, 288, 291, 327, 328, 353, 412 Fox, Barbara 194, 414 Frajzyngier, Zygmunt 195 Francis, Gill 93, 97, 110, 113, 114 Fried, Mirjam 33 G Gabrielatos, Costas 105, 111, 113 Gabrys, Danuta 244, 253 Garrett, Merril F. 34 Gass, Susan M. 117, 149, 293, 354 Gawinski, Birthe 285, 288 Geeraerts, Dirk 64, 125, 150, 195, 234, 254 Geiger, Richard A. 64, 324

420

Author index

Gelabert, Jaime 291 Gennari, Silvia P. 285, 288 Gentner, Dedre 292 Geuder, Wilhelm 86, 89 Giardetti, J. Roland 386, 414 Gibbs, Raymond W. 122, 150, 234, 254, 285, 288, 315, 321 Gil, David 228 Gili Gaya, Samuel 194 Gilquin, Gaëtanelle 95, 98, 106, 113 Ginsberg, Ralph B. 290 Givón, Talmy 21, 31, 292 Gjedde, Albert 149 Glenberg, Arthur M. 198, 228 Goatley, Andrew 234, 254 Goldberg, Adele E. 2, 4, 26, 27, 31, 67, 72–75, 80, 82, 83, 86, 87, 89, 125, 150, 190, 194, 197–199, 201–203, 205–207, 210, 224– 227 Goldberg, Robert 206, 210, 228 Goldin, Mark 156, 180, 194 Goldin-Meadow, Susan 292 Goldsmith, John 13, 31 González, Nora 156, 180, 194 González-Álvarez, Elsa 152 Gordon, Peter 206, 227 Grabe, Wallace 232, 254 Grabe, William 94, 113 Grady, Joseph 131, 150 Granger, Sylviane 94, 108, 113– 115, 210, 225, 228 Graustein, Gottfried 39, 62, 112 Green, Melanie 224, 227, 265, 284, 288 Greenbaum, Sidney 39, 62, 90, 101, 114 Greenberg, Joseph H. 324 Greenhill, Annabel 291 Gregg, Kevin 40, 41, 62 Gries, Stefan 74, 89, 95, 96, 98, 111, 114, 116, 199, 201, 202, 204, 206–208, 211, 212, 216, 222, 224, 225, 228, 229

Grimshaw, Jane 156, 180, 194 Grommes, Patrick 226, 383 Gropen, Jess 206, 210, 228 Gumperz, John J. 292, 320, 414 Guo, Jiansheng 405, 412 H Haberzettl, Stefanie 382 Hahn, Martin 291 Haiman, John 9, 21, 31, 42, 62, 162, 163, 194, 292 Hakuta, Kenji 122, 150 Halliday, Michael A. 94, 264, 288 Hamawand, Zeki 59, 62 Hampe, Beate 95, 114 Hansen, Johan W. 285, 288 Harley, Birgit 261, 288 Harmer, Jeremy 81, 82, 89 Haspelmath, Martin 206, 225, 228 Hawkins, Roger 121, 150 Heine, Bernd 296, 321 Hendricks, Henriette 384 Heyvaert, Liesbet 142, 150 Hickmann, Maya 321, 324 Highland, Diane 293, 352 Hilferty, Joseph 198, 228 Hinkel, Eli 89, 115, 288, 322, 327, 328, 353, 390, 392, 412 Hoey, Michael 93, 110, 114 Hoiting, Nini 267, 292 Hollander, Michelle 228 Hollmann, Willem 321 Holm, Lars 277, 289 Holme, Randal 233, 254 Holmes, Janet 114 Hopper, Paul 194, 414 Horn, Laurence 152 Housen, Alex 328, 330, 333, 353 Howatt, Antony 288 Hudson, Richard A. 2, 4 Hughes, Rebecca 100, 101, 103, 112 Hung, Joseph 94, 114

Author index Hunston, Susan 93, 97, 110, 113, 114 Hymes, Dell 97, 104, 114 I Ibarretxe-Antuñano, Iraide 289 Idiazabal, Itziar 411 Isaþenko, Aleksandr 381, 383 J Jackendoff, Ray S. 59, 62, 81, 89, 130, 150, 206, 228 Jacobs, Georg M. 115 Janssen, Theo 32 Jaspers, Danny 61 Jeppesen, Bodil 277, 289 Jespersen, Otto 374, 383 Jessner, Ulrike 254 Jisa, Harriet 389, 410, 411, 412 Johansson, Maria 113 Johansson, Stig 112 Johns, Tim 98, 114, 130 Johnson, Bonnie 361, 383 Johnson, Keith 88 Johnson, Mark 13, 32, 50, 63, 122, 126, 131, 151, 233, 251, 254, 255, 266, 289, 313, 321 Jordan, Stefanie R. 291 Jung, Woo-Hyun 391, 413 Jurafsky, Daniel 99, 116 K Kaplan, Robert D. 113, 232, 244, 254 Kaschak, Michael P. 198, 228 Kay, Paul 26, 31, 53, 61, 224, 227 Kecskes, Istvan 10, 31, 152, 232, 233, 237, 242, 254, 407, 413 Kehoe, Andrew 111, 116 Keiper, Hugo 354 Kellerman, Eric 260, 275, 276, 289, 347, 354 Kemmer, Suzanne 25, 30, 33, 157, 159, 161–164, 176, 194, 354

421

Kennedy, Graeme 110, 114, 328, 354 Kern, Sophie 389, 412 Ketteman, Bernhard 94, 114 Kim, Yiyoung 228, 413 King, Larry 156, 194 Kita, Sotaro 285, 289, 291 Klabunde, Ralf 371, 383 Klein, Wolfgang 335, 343, 344, 354, 359, 360, 363, 375, 383 Klooster, Wim 61 Köpcke, Klaus-Michael 152 Kopecka, Anetta 319, 321, 323 Kosecki, Krzysztof 152 Kövecses, Zoltán 13–15, 31, 126, 150, 233, 234, 242, 254 Kramsch, Claire 290 Krashen, Stephen 39–41, 59, 62, 63, 261, 289, 328 Kristiansen, Gitte 30, 88, 150, 226, 252, 353 Kurtyka, Andrzej 15, 32 Kwiatkowska, Alina 195 L Lado, Robert 53, 63, 108, 115 Lakoff, George 9, 13, 32, 37, 42, 47, 50, 63, 122, 125–127, 130, 151, 224, 229, 233, 251, 254, 255, 265, 266, 285, 289, 313, 321 Lambert, Monique 371, 373, 382, 384 Lamiroy, Beatrice 177, 193 Landsberg, Marge E. 229 Langacker, Ronald W. 2, 7–14, 17, 18, 21, 22, 24–26, 28, 32, 37, 42, 45–47, 49, 50, 52, 56, 58, 59, 63, 67, 69, 70–72, 82, 85, 87, 89, 92–94, 115, 123, 125, 151, 155, 161, 163, 166, 167, 180, 190, 192, 194, 195, 224, 229, 259, 265, 285, 289, 310, 321, 332– 334, 336, 339–342, 344, 345,

422

Author index

352, 354, 369, 383, 385–388, 390, 407, 410, 413 Lantolf, James P. 291, 296, 305, 322 Larsen-Freeman, Diane 94, 115 Laursen, Elisabeth 277, 289 Lazar, Gillian 244, 255 Lázaro Mora, Fernando A. 151 Lee, James F. 290 Lee, Kee Dong 323 Lee, Peter W. 293, 352 Leech, Geoffrey 61, 90, 112 Lehmann, Winfred P. 196 Leitner, Gerhard 39, 62, 112 Leki, Ilona 232, 255 Levin, Beth 206, 210, 225, 229 Levin, Samuel R. 251, 255 Levinson, Stephen C. 292, 296, 320, 322, 414 Lewandowska-Tomasczyk, Barbara 195 Lewis, Michael 81, 89 Leys, Odo 307, 322 Li, Ping 249, 361, 383 Lian, Shuneng 405, 413 Liang, J. 199, 200, 222, 229 Liébana, Eva 277, 290 Lindner, Susan 9, 15, 33 Lindstromberg, Seth 15, 30, 223, 226, 242, 252, 333, 353 Littlefield, Heather 291 Littlemore, Jeannette 233, 236, 255 Lloyd, Barbara B. 34 Lock, Graham 329, 342, 354 LoCoco, Veronica 278, 290 Long, Michael H. 102, 115, 162, 227, 259–262, 290 Lönneker-Rodman, Birte 116 López, Luis 226 Low, Graham D. 233, 236, 253, 255 Lund, Karen 268, 271, 272, 276, 278, 287, 296, 321

M Mackenzie, Lachlan 152 MacLaury, Robert E. 32 MacLennan, Carol H.G. 233, 255 MacWhinney, Brian 94, 115, 395, 409, 411, 413 Madden, Carolyn 293 Mairal, Ricardo 127–129, 152 Maldonado, Ricardo 3, 18, 33, 63, 138–140, 151, 152, 155, 157, 159, 161, 163–165, 167, 168, 171, 175, 177, 180, 191, 192, 194, 195 Malkiel, Yakov 196 Malt, Barbara C. 285, 288 Manney, Linda 157, 163, 177, 195 Manning, Elizabeth 97, 113 Maribo, Grethe 277, 289 Marinova, Diana 228, 413 Marko, Georg 94, 114 Martínez, Israel 156, 193 Martínez, Pablo 159, 163, 195 Masayo, Lida 195 Master, Peter 392, 408, 413 Matlock, Teenie 12, 34, 310, 312, 322 Matsumoto, Yo 12, 34, 310, 322 Mayer, Mercer 393, 395, 404, 409, 413 McCarthy, Michael 82, 90, 95, 100, 101, 103, 112, 115 McEnery, Anthony 94, 97, 110, 112, 115, 343, 355 McNeill, David 285, 290 Meex, Birgitta 307, 322 Meisel, Jürgen 261, 290 Merlo, Paola 116 Messineo, Cristina 177, 195 Meunier, Fanny 3, 91, 98, 115, 228 Michaelis, Laura 128, 151, 227 Mileham, Jean 293, 352 Miller, Jim 227, 289 Miquel, Lourdes 156, 193 Mollica, Anthony 256

Author index Monge, Carlos Francisco 193 Montes Giraldo, José Joaquín 151 Montrul, Silvina 224, 225, 229, 361, 384 Moon, Rosamund 86, 90 Moravcsik, Edith A. 324 Morgan, Pamela S. 316, 322 Mortelmans, Tanja 307, 322 Mukherjee, Jobrayto 95, 107, 115 Munro, Pamela 163, 194 Munsell, Paul 40, 63 N Nadel, Lynn 34 Naigles, Letitia R. 285, 290 Nassaji, Hossein 263, 291 Natale, Silvia 382 Nava, Fernando 157, 163, 177, 195 Navarro, Samuel 260, 273, 290 Navarro Coy, Marta 226 Negueruela, Eduardo 260, 276, 291 Neuvel, Sylvain 64 Newman, John 320, 322 Newmeyer, Frederick J. 410, 413 Nicoladis, Elena 260, 273, 290 Niemeier, Susanne 3, 30–33, 88, 95, 111, 116, 117, 149, 151, 229, 233, 252, 265, 287, 291, 292, 320, 323, 325, 347, 349, 350, 352, 354, 355, 407, 410–414 Norris, John 261, 291 Nunan, David 285, 291 Núñez, Rafael E. 13, 32 Núñez Cedeño, Rafael A. 226 Nüse, Ralf 371, 384 Nuyts, Jan 34 O O’Brien, Jennifer Ellen 315, 321 O’Connor, Mary C. 26, 31 O’Keeffe, Anne 95, 115 Oakes, Michael 97, 115 Odlin, Terence 392, 414

423

Oller, John W., Jr. 3, 94, 101, 108, 385, 386, 409, 411, 412, 414 Oller, Stephen D. 386, 414 Ortega, Lourdes 261, 291 Ortega Olivares, Jenaro 156, 193 Ortony, Andrew 34, 63 Overstreet, Mark 287, 293, 321 Özçaliúkan, ùeyda 269, 291 Özyürek, Asli 285, 289, 291 P Pagliuca, William 382 Paivio, Allan 60, 63 Palancar, Enrique 177, 195 Pan, Ning 414 Panther, Klaus-Uwe 14, 34, 64, 88, 227, 322 Papp, Tunde 233, 254 Paprotté, Wolf 253 Paradis, Caritas 94, 115 Pavlenko, Aneta 256 Pederson, Eric 34 Peña Cervel, M. Sandra 33, 61, 89, 150 Pérez Hernández, Lorena 130, 137, 143, 151, 152 Periñán Pascual, José Carlos 226 Perkins, Revere 382 Perlman, Marcus 122, 150 Perpiñan, Silvia 224, 225 Petch-Tyson, Stephanie 94, 114 Peterson, Mary A. 34 Petr, Jan 381, 383 Pfaff, Carol 290 Phillip, Gillian S. 242, 255 Pinholt, Per 277, 289 Pinker, Steven 206, 216, 228, 229 Piper, David 232, 255 Pörings, Ralph 152 Pourcel, Stéphanie 298, 299, 319, 323 Pride, John B. 114 Putseys, Yvan 61

424

Author index

Pütz, Martin 31–33, 95, 116, 149, 151, 323, 354, 355, 407, 412– 414 Q Queller, Kurt 123, 151 Quirk, Randolph 68, 90 R Radden, Günter 14, 31, 34, 58, 63, 64, 88, 126, 141, 150–152, 227, 236, 256, 296, 307, 321–323, 338, 343, 345, 352, 354 Radford, Andrew 68, 90 Ragnarsdóttir, Hanna 292 Ramat, Anna 382 Rauch, Irmengard 324 Raupach, Manfred 414 Reagan, Timothy 233, 256 Reddy, Michael J. 14, 34, 54, 63 Redeker, Gisela 32 Reif, Monika 3, 325 Reilly, Judy 412 Reisener, Helmut 331, 354 Renouf, Antoinette 111, 116 Reppen, Randi 61, 93, 112 Reynolds, Dudley W. 330, 353 Reynolds, John J. 156, 193 Rice, Sally 9, 34, 83, 90, 101 Richards, Jack 53, 63 Richardson, Daniel C. 12, 34 Richardt, Susanne 233, 256 Riddle, Elizabeth 329, 355 Riehle, Wolfgang 354 Robert, Stéphane 321, 324 Robinson, Peter 29, 30, 125, 152, 260–262, 286, 287, 290, 291 Rojas Gonzáles, Margarita 193 Rojo López, Ana Maria 3, 74, 197 Roland, Douglas 99, 116 Römer, Ute 102, 116, 326, 330, 355 Rosado, Elisa 412 Rosch, Eleonor 10, 34 Rott, Susanne 287, 293, 321

Rowling, Joanne K. 86, 90 Rudzka-Ostyn, Brygida 15, 34, 42, 62–64, 112, 125, 151, 152, 324 Ruiz, Lucas 260, 273–275 Ruiz de Mendoza Ibáñez, Francisco J. 3, 30, 33, 61, 88, 89, 106, 121, 126–130, 132, 137, 139, 143, 150–153, 161, 165, 192, 226, 252, 353, 358 Ruiz Campillo, José Placido 193 Russo, Gerard A.A. 231–233, 256 Rutherford, William E. 39, 53, 64 S Sag, Ivan 227 Sahonenko, Natalya 358, 367, 371, 373, 383 Salaberry, Rafael 256, 286, 291, 353, 384 Saleemi, Anjum P. 149 Sandra, Dominiek 9, 34, 83, 90 Santibáñez, Francisco 127, 153 Sanz, Cristina 278, 293 Savignon, Sandra J. 250, 256 Schachter, Jacquelyn 149 Schmid, Hans-Jörg 71, 90 Schmidt, Richard 82, 90, 261, 292 Schmiedtová, Barbara 3, 86, 90, 329, 343, 344, 357, 358, 367, 368, 371–373, 378, 380, 381, 383 Schönefeld, Doris 95, 114, 224, 229 Schwarz, Hellmut 325, 332, 355 Sebeok, Thomas A. 237, 256 Seliger, Herbert W. 390, 404, 414 Selinker, Larry 49, 64, 81, 90, 354 Sells, Peter 156, 180, 195 Serra-Borneto, Carlo 302, 303, 307, 312, 320, 323 Seuren, Pieter 61 Sharwood Smith, Michael 39, 64, 289 Shibatani, Masayoshi 292, 323, 414 Shirai, Yasuhiro 353, 360, 382–384

Author index Shopen, Timothy 4, 293, 324 Shortall, Terry 99, 116 Sieloff Magnan, Sally 117 Siemund, Peter 152 Simonsen, Hanne G. 33 Sinclair, John 93, 94, 97, 116, 330, 333, 355 Singleton, David 122, 153 Slabakova, Roumyana 358, 361, 362, 383, 384 Slobin, Dan I. 260, 267–269, 271, 275, 277, 282, 285, 287, 291, 292, 296, 298, 300, 320, 323, 385, 387–389, 392–397, 405, 406, 410, 412, 414 Sloman, Steven A. 285, 288 Smith, Carlota 358, 359, 362, 384 Smith, Michael B. 59, 64, 305, 307, 314, 324 Smyth, Herbert 161, 195 Solska, Agnieszka 244, 253 Sotirova, Violeta 321 Starren, Marianne 381, 382 Stathi, Katerina 92, 96, 116 Steer, Jocelyn M. 392, 408, 414 Stefanowitsch, Anatol 98, 111, 114, 116, 206–208, 211, 212, 216, 225, 228, 229 Steinbach, Gudrun 331, 354 Stelma, Juurd H. 233, 253 Stengers, Hélène 95, 112 Stevenson, Susanne 116 Stewart, Dominic 116 Stoll, Sabine 361, 362, 384 Stoness, Scott C. 291 Strömqvist, Sven 287–289, 292, 294 Stubbs, Michael 99, 117 Stutterheim, Christiane von 358, 370, 371, 373, 378, 382–384 Suñer, Margarita 156, 194 Svartvik, Jan 90 Swain, Merrill 261, 288, 292 Sweetser, Eve 12, 13, 31, 34, 323 Szabo, Paul 233, 242, 254

425

Szczesniak, Konrad 209, 211, 213, 215, 216, 225, 226 T Tai, James 405, 414 Tak, Marjan 415 Takada, Mari 228, 413 Talmy, Leonard 3, 4, 12, 34, 37, 183, 196, 259, 260, 266, 267, 269, 276, 277, 279, 284, 285, 293, 296–298, 300, 304, 310, 319, 321, 324 Tano, Cheryl 291 Taoka, Chiaki 321 Taylor, John R. 2, 9, 22, 30, 32, 35, 37, 42, 47–49, 52, 54, 58–60, 62, 64, 92, 95, 108, 112, 117, 151, 306, 410, 415 Tegnberg, Birgit 277, 289, 290 Terrazas, Paula 285, 290 Terrell, Tracy 39, 41, 63, 156, 196 Teuberg, Wolfgang 98, 117 Thompson, Sandra A. 206, 229, 292, 323 Thorne, Steve 305, 322 Tomasello, Michael 3, 4, 58, 59, 65, 74, 90, 97, 117, 190, 196, 197, 199, 215, 216, 220, 223, 225, 229, 353, 385, 415 Tono, Yukio 97, 115 Traugott, Elizabeth 184, 196 Tremblay, Rachel 59, 62 Trondhjem, Frederikke B. 269, 288 Tsohatzidis, Savas L. 64 Tuggy, David 9, 35 Turner, Mark 13, 31, 35, 67, 79, 80, 89, 234, 254 Tyler, Andrea 9, 35, 82, 90, 95, 113, 117, 228, 329, 332, 355, 413 U Ueta, Yumiko 390, 391, 415 Ungerer, Friedrich 71, 90 Upton, Thomas 94, 112, 228

426

Author index

Ureña, P. Henríquez 156, 193 V Valdman, Albert 102, 117 Valenzuela Manzanares, Javier 3, 74, 197, 198, 226 Valeva, Gloria 233, 235, 256 Van Belle, William 193, 321 van der Leek, Frederike 163, 196 Van Ek, Jan A. 232, 256 van Hell, Janet G. 389, 405, 415 Van Hoof, Anne-Marie 260, 276, 289 Van Langendonck, Willy 193, 321 van Oosterhout, Moniek 415 van Wolde, Ellen 33 VanPatten, Bill 65, 102, 117, 272, 278, 287, 290, 293, 294, 321 Varela, Elizabeth 263, 287 Vázquez Rozas, Victoria 177, 196 Velasco Sacristán, Maria S. 233, 244, 256 Vendler, Zeno 359, 360, 384 Verhoeven, Ludo 287–289, 292, 294, 412, 415 Verkuyl, Henk 362, 384 Verspoor, Marjolijn 234, 248, 253, 321, 323, 334, 355 W Waara, Rene 199, 229 Wagner, Laura 362, 384 Walter, Maik 226, 383 Walz, Joel 117 Wasow, Thomas 215, 230 Watabe, Masakazu 390, 391, 415 Waters, Sarah 85, 90 Webb, David 913 Weber, Rita R. 293, 352 Weschlerand, Stephen 195 White, Joanna 263, 293 Widdowson, Henry 98, 102, 117, 282, 293

Wierzbicka, Anna 9, 35, 41, 50, 57, 65 Wiese, Jørn 277, 289 Wiley, Edward 122, 150 Wilkins, David A. 232, 256, 260, 293 Wilkins, David P. 269, 294 Williams, Christopher 344, 345, 355 Williams, Jessica 262, 287, 288, 290, 293, 321 Willis, Dave 81, 90 Wilson, Andrew 110, 115 Wilson, Ronald 228 Woisetschlaeger, Erich 13, 31 Wong, Wynne 286, 294 Wulff, Stefanie 74, 89, 199, 201, 202, 204, 206, 208, 222, 224, 228 X Xiao, Richard 97, 115 Xiao, Zhongua 343, 355 Y Yangklang, Peerapat 285, 294 Yoshida, Kensaku 404, 415 Z Zaenen, Annie 156, 180, 195 Zec, Draga 156, 180, 195 Zlatev, Jordan 9, 35, 268, 285, 294 Zubizarreta, Maria Luisa 224, 226 Zydatiß, Wolfgang 326, 329, 355

Subject index A aan het-construction (see construction) absolute event (see event) ABSTRACT MENTAL ACTIVITY IS MOTION 313, 318

abstract motion (see motion) abstract noun 49, 132 acceptability 56, 197, 217–219, 221, 222, 376, 381 acceptability judgement task 197, 381 acceptability rating task 217, 221, 222 accusative 192, 301, 302, 305–307, 309–311, 314, 315, 318 accusative/dative alternation (see alternation) acquisition 1, 3, 4, 38– 41, 74, 81, 83, 92, 94, 95, 97, 121, 159, 166, 190, 197–199, 211, 215, 216, 220, 221, 223, 233, 250, 259– 261, 263, 265, 269, 271–273, 276–278, 285, 286, 325, 346, 357, 358, 360–362, 380, 387, 389, 392, 403, 405, 407–410 acquisition and learning (see learning) unconscious acquisition 407 action 3, 27, 78, 86, 101, 124, 129– 131, 133, 136, 140, 143–146, 148, 155, 158, 159, 161, 162, 164, 166, 168, 169, 173–175, 181–183, 188, 189, 192, 203, 257, 280–283, 296, 306, 307, 370, 387, 393, 403, 405 dynamic action 3, 257 grooming action 155 self-benefit action 164

self-directed action 155, 158, 159, 166, 188 action schema 307 action verb (see verb) ACTIONS ARE TRANSFERS 130, 131 adequacy 122, 125, 139, 358 explanatory adequacy 122, 139 adjective-noun combination 79 advanced learner 74, 200, 277, 328, 346, 358, 371, 380, 385 AFFECTION IS WARMTH 131 agency 385, 392–394, 399, 400– 403, 406–409 degree of agency 392–394, 399– 402, 406, 407 Aktionsart (also lexical aspect) 342, 357, 359–361, 363, 366, 367, 375, 381 alienable possession (also alienable objects) 155, 168–170, 189 alternation 209, 215, 217–219, 224, 305, 311, 314 accusative/dative alternation 303, 305, 311, 314 bounded/unbounded alternation 343, 345, 349 dative/prepositional group alternation 209, 215, 217–219, 224 infinitive construction/gerundial complement alternation 43, 47–49, 58 infinitive construction/that-clause alternation 43, 47, 55, 56 mass/count noun alternation 37, 45–47, 50, 54, 56, 59, 60, 173 ambiguity 348 applied linguistics 53, 95

428

Subject index

approach 1, 3, 8, 21, 25, 26, 28, 37, 43, 44, 45, 52, 53, 56, 67, 68, 73, 76, 80–82, 91–93, 95–100, 103, 104, 106–111, 143, 156, 158, 180, 188, 191, 192, 198, 206, 216, 221, 223, 234, 259–265, 277–280, 282, 284, 295, 305, 307, 318, 319, 325–328, 331, 333, 334, 345, 346, 348, 351, 357–370, 375, 379–381, 385, 410 bottom-up approach 67, 76, 80, 92 communicative approach 327, 328 corpus-based approach 58, 91, 94, 98, 100, 103, 105, 108, 197, 209 didactic approach 1 exemplar-based approach 197, 199, 215, 216, 221, 223 lexical approach 81, 87 maximalist approach 57, 67, 76, 80 top-down approach 67, 76, 81 typological approach 259, 265, 277, 279, 284 usage-based approach 25, 26, 28, 92, 93, 95, 96 ARGUMENT IS WAR 127 aspect (also grammatical aspect) 3, 9, 17, 25, 26, 37–40, 42, 51, 56, 58, 59, 68, 72, 87, 91, 95–97, 100–102, 104–106, 108, 121, 126, 136, 172, 192, 196, 198, 199, 231, 236, 245, 251, 262, 264, 265, 270, 271, 275–277, 301, 304, 309, 310, 325–334, 336–338, 342–352, 357–373, 378–384, 386, 403, 407, 408 lexical aspect (see Aktionsart) progressive aspect (also progressivity) 3, 56, 326, 330, 342,

343, 344, 345, 350, 357, 361, 362, 365, 374, 378 non-progressive aspect 329, 330, 331, 332, 338, 342–345, 347, 349–351, 362 aspect hypothesis 360, 361 atelic(ity) 330 audiolingualism 260 authentic(ity) 92, 100, 102, 105, 107, 250, 300, 333, 410 AUTHOR FOR WORK 129 AUTHOR FOR WORK FOR MEDIUM 129 avoidance strategies 132, 143, 242, 273, 332 B bilingualism 201, 390 blending (also conceptual blending) 13, 79, 80, 85, 87, 104, 311 blending theory 67, 68, 79, 80 BNC (see British National Corpus) body part 155, 164, 167, 189, 295, 302, 307, 308, 317 body part motion (see motion) bottom-up approach (see approach) boundary-crossing 267, 272–274, 280, 281, 283 bounded/unbounded alternation (see alternation) boundedness 325, 345, 349–351 British National Corpus (also BNC) 60, 93, 110, 149, 209, 212, 213, 329, 332, 335, 338, 339, 341 C case marking 314, 317, 318 cause of motion (see motion event) caused motion 72, 73, 80, 86, 191, 199, 201, 202 caused-motion construction (see construction) centrality of meaning 7, 8 CG (see Cognitive Grammar)

Subject index change construction (see construction) CHANGE IS MOTION 313 change of state (see state) change-of-state verb (see verb) Chinese 3, 111, 199, 200, 266, 285, 385, 392, 394, 395, 397–408, 411 Chinese learner of English 3, 199, 200, 385, 392, 394, 395, 397– 407, 411 clause 16, 17, 33, 43, 46, 47, 55, 56, 62, 69, 71, 78, 86, 102, 105, 110, 142, 159, 169, 270, 276, 300, 335, 336, 342, 360, 376, 386, 394–403 clitic se in Spanish 155, 156 coercion 128 cognition 50, 70, 91, 96, 100, 103, 107, 164, 165, 175, 177, 285, 295 Cognitive Grammar (also CG) 7, 8, 15–17, 20, 22, 25, 26, 29, 42, 57–60, 67–73, 76, 82, 332, 386– 389 cognitive grammar model (see model) cognitive semantics 9, 14, 121, 148 colligation 93, 110 collostruction(al) 98, 106, 111, 206–208, 225 collostructional analysis 98, 111, 206, 208 colour (also color) 3, 18, 54, 79, 129, 131, 136, 237–242, 255, 263 communication 21, 54, 60, 74, 81, 93, 97, 110, 208, 250, 255, 259, 261, 262, 274, 284, 314, 327, 328, 334, 342, 386, 387 COMMUNICATIVE ACTION IS EXPERIENTIAL ACTION 130

communicative activities 262

429

communicative approach (see approach) communicative competence 97, 236, 244, 327 communicative language teaching (see teaching) communicative-based (see teaching) competence/performance dichotomy 93 complement(ation) 26, 37, 38, 42– 45, 47–49, 55, 56, 58–60, 78, 139, 140, 223, 297, 306, 343 completion exercises (see teaching) component structure 19 composite structure 19 composition 7, 19, 130, 222, 328, 390 comprehension 99, 103, 228, 231, 232, 259, 260, 262, 278–280, 290, 293, 327 comprehension of L2 (see teaching) discourse comprehension (also text comprehension) 231, 232 conceptual blending (see blending) conceptual difference 247, 250, 301, 319, 331, 370, 378, 405 conceptual domain (see domain) conceptual error (see error) conceptual fluency 2, 3, 10, 231– 233, 236, 237, 242–244, 247– 251 conceptual interference 235, 241, 242, 247 conceptual metaphor (see metaphor) conceptual metaphor theory 231– 233 conceptual metonymy (see metonymy) conceptual productivity 243, 245 conceptual prominence 10, 17, 20, 23 conceptual relation (see relation)

430

Subject index

conceptual structure 94, 126, 231, 358, 370, 372, 373 conceptual substrate 7, 14, 23 conceptual teaching (see teaching) conceptual training 231 conceptual transfer 243, 244, 385, 405, 406 conceptualization 37, 50–54, 56, 58, 60, 96, 106, 108, 121, 129, 132, 148, 188, 234, 235, 249, 265, 271, 295–298, 301, 302, 305, 318, 357, 370, 371, 372, 378, 379, 388, 407 conceptualization strategies 121, 148, 371 CONDUIT metaphor (see metaphor) conflation 267, 270, 273, 275, 279, 280, 306, 324 conscious awareness (also consciousness raising) 39 conscious learning (see learning) constraints on metaphor (see metaphor) constraints on metonymy (see metonymy) construal 3, 7, 9–11, 15, 17, 19–21, 51, 68, 70, 73, 86–87, 94, 101, 155, 158, 162, 164, 168, 170, 171, 173, 175, 178, 180, 181, 183, 184, 186, 188–192, 302, 326, 331, 338, 343, 345, 349, 357, 371, 372, 385, 387, 388, 391, 392, 394, 396–398, 404, 406–410 construal of motion (see motion) construction (also pattern) 2–3, 7, 11, 13, 14, 18–20, 22–28, 37, 43–47, 58, 67–70, 72–77, 79–88, 93, 94, 97, 98, 100–102, 106, 110, 111, 121–128, 130, 131, 138–147, 155–160, 162–174, 176–183, 185, 188–192, 197– 204, 206–218, 220–225, 236, 240, 242, 250, 260, 266–268,

271–282, 284–286, 295, 297– 300, 306, 311, 315, 317, 319, 320, 325–327, 331, 332, 334, 335, 338, 346, 347, 351, 352, 361, 362, 371–377, 379–381, 385–391, 394, 396–398, 400, 402, 405–411 aan het construction 376 caused-motion construction 68, 73, 127, 128 change construction 67, 68 ditransitive construction 3, 27, 31, 69, 197, 206–208, 210– 212, 214–218, 220–223 focus construction 155, 181 gerundial complement construction 43, 44, 47–49 grammatical construction 19, 22, 25, 31, 37, 46, 79, 111, 126, 131, 144, 206, 228, 325–327, 332, 334, 351, 362, 385, 407 inchoative construction 121, 144, 147 infinitive construction 43, 44, 47–49, 56, 58, 59, 78 mental construction 7, 11, 13, 14, 23, 24 middle construction 3, 121, 141– 143, 155, 157–161, 163–167, 172–174, 177, 182, 189, 190, 192 middle evaluative construction 144, 146 passive construction 3, 385, 387–391, 396–398, 402, 408, 411 psychological reality of constructions 199, 223 reflexive construction 121, 122, 138, 143–145, 156, 159, 162, 163, 166, 170, 178 ser-passive construction 139 simultaneity construction 67, 68

Subject index that-clause construction 43, 46, 47, 55 transitive construction 69, 70, 159, 169, 172, 179, 201, 204 Construction Grammar 2, 26, 29, 67, 68, 70, 72, 73, 75–77, 82, 198 constructional island 215–217, 220, 223 constructional meaning 19, 72 constructional schema 7, 18, 19, 25, 334 CONTAINER metaphor (see metaphor) contrastive analysis (also contrastive linguistics, contrastive study) 1, 2, 52, 60, 108, 121, 122, 148, 259, 285, 295, 296, 346 conventional expression 7, 28 convergence 91, 92, 94, 95, 104, 109 corpus linguistics 3, 91–94, 96–100, 108, 110, 116 corpus study (also corpus-based study, see approach) corpus-based approach (see approach) correlation principle 121 count noun 1, 37, 45, 46, 49, 50, 54, 56, 59 count/mass noun alternation (see alternation) counter-expectation 155 counterfactual(ity) 105, 329, 337, 341, 346, 347 cross-linguistic 39, 143, 269, 277, 300, 357, 359, 370, 379, 392, 410 Czech 210, 357–359, 362–373, 379–381 Czech as L2 (see L2 Czech) Czech learner of German 359, 373

431

D Danish learner of Spanish 272–276 data 2, 3, 41, 43, 58, 60, 67, 92, 93, 95, 97–100, 102, 106, 107, 109, 110, 125, 134, 161, 208, 209, 213, 214, 217, 219, 222, 223, 231, 233, 238, 244–248, 273, 274, 357–360, 370–375, 381, 395–397, 411 dative 70, 139, 176–178, 191–192, 197, 204, 208–215, 217, 218, 236, 301, 302, 305–307, 311, 314, 318, 320 dative/accusative alternation (see alternation) dative/prepositional group alternation (see alternation) decision-making task (see teaching) defocused right boundary 369 defocusing of boundaries 344, 347, 377, 380 degree of agency (see agency) deictic (also deixis) 301, 336, 374, 375, 378 descriptive adequacy 358 didactic approach (see approach) didactic grammar (also learner’s grammar, school grammar, teacher’s grammar, textbook grammar) 1, 39, 326, 328, 330, 331 diminutive 3, 121, 122, 132–138, 148 discourse comprehension (see comprehension) discourse production (see production) distal (see proximal/distal) ditransitive 3, 26–28, 69, 72, 197, 199, 201, 202, 204–224 ditransitive construction (see construction) divergence 2, 91, 96–99, 104, 109, 212, 327, 379

432

Subject index

domain 49, 51, 54, 59, 69, 92, 99, 110, 124, 126–131, 144, 145, 180–183, 192, 234–237, 239, 243–250, 259, 265, 271, 276, 277, 284, 296, 302, 313, 318, 363, 367, 380 conceptual domain 54, 126, 259, 277, 284, 318 source domain 124, 127, 129, 131, 145, 234–237, 243–246, 248–250, 313 spatial domain 51, 69 target domain 127, 129–131, 192, 235, 239, 245, 246, 248, 313 dominion (see experiential dominion) drill(ing) (see teaching) Dutch 3, 163, 210, 269, 357, 359, 366, 367, 370, 371, 373–381 dynamic action (see action) dynamic relation (see relation) dynamic situation 3, 175, 183–185, 190, 270, 302, 309, 311, 330, 394, 398, 409 dynamic/static opposition 24, 71, 270, 271, 302, 312, 394, 397– 399, 404 E eclecticism 91, 104, 106, 109, 111 EFL (see English as a foreign language) ELT (see English language teaching) emotion(al) 9, 50, 86, 128, 129, 131, 135, 155, 159, 161, 164– 166, 174–178, 181, 185, 189, 190, 237, 239, 391 emotion middle 175, 177 emotion verb (see verb) emotional reaction 135, 155, 159, 164, 175–177, 181, 189, 190

emotional state 50, 128, 155, 177, 178 empirical research (also empirical study) 110, 223, 259–262, 277, 278, 318, 346, 357, 371, 380, 386, 410 empiricist 92, 93, 110 endpoint 140, 146, 147, 344, 349, 363–365, 370, 371, 373, 377 end-prominence configuration 142 energetic event (see event) English 3, 13, 24, 26, 37, 38, 41, 45, 47, 49, 50, 52, 54, 56, 58, 60, 69, 72–75, 78, 80, 82, 84–87, 93, 95, 99–101, 104, 105, 108, 109, 121–124, 133, 137, 139, 141– 148, 155, 157, 160, 165–167, 169, 172, 177, 197, 199–204, 206–210, 213–217, 221, 222, 224, 232, 234, 236, 238, 242, 244–247, 249, 266–269, 273, 276–278, 285, 295, 297–305, 307, 309, 313, 316, 318, 325– 334, 337–339, 342, 343, 346– 348, 351–352, 357, 359–368, 370, 373–380, 382, 383, 385, 388, 390–392, 395, 397–408, 411 English as a foreign language (also EFL) 325–331, 333–347, 350– 352, 392, 394, 397–407, 411 English as L2 (see L2 English) English language teaching (also ELT) 86, 105, 327, 334 English learner of German 303, 313 ENTITY FOR ONE OF ITS PROPERTIES

130 entrenchment 58, 85, 87, 93, 96, 125, 220, 222 error 2, 3, 40, 49, 53, 55, 101, 106, 108, 119, 157, 158, 166, 170, 176–179, 199, 211, 231–235, 242, 244, 245, 247, 250–252,

Subject index 260, 262, 263, 326, 328–330, 349, 351, 380, 391 conceptual error 3, 231, 232, 235, 242, 244, 247, 248, 250, 252 formal error (also error of form) 235, 380 error analysis 2, 3, 119, 199, 231, 233 event 3, 7, 16, 20, 22–25, 48–50, 54, 59, 69, 71, 77, 80, 86, 94, 97, 101, 106, 130, 135, 136, 141, 155, 157, 159, 161–167, 173, 175, 177–185, 188–190, 198, 222, 244, 246, 259, 260, 265– 271, 274–277, 279, 281, 282, 284, 293–299, 301–303, 305– 307, 309–314, 318–321, 323, 329, 357, 371–373, 375, 378, 385–394, 396–410 absolute event 155, 180 energetic event 155, 180, 184, 185 self-beneficial event 155 transitive event 385, 386, 388, 390, 393, 394, 397, 401 unexpected event 155 event schema 296, 301, 302, 306, 307, 311, 318 EVENTS ARE OBJECTS 130 exemplar-based approach (see approach) exercises (see teaching) completion exercises (see teaching) grammatical exercises (see teaching) experiencer subject 155, 164, 165 EXPERIENTIAL ACTION IS EFFECTUAL ACTION 129

experiential dominion 155, 161, 165, 167–171, 179, 188, 189

433

experimental research 87, 95, 97, 99, 107, 108, 110, 191, 204, 241, 242, 247, 285, 370, 372, 379 explanation in grammar 41–45 explanation in language teaching (see teaching) explanatory adequacy (see adequacy) explicit focus on form (see focus on form) explicit grammar teaching (see teaching) explicit instruction (also explicit teaching, see teaching) extension of meaning (see meaning extension) F facilitator (see teacher as facilitator) false friends 357, 365, 367, 378– 380 fictive change 12 fictive motion (see motion) figurative language 233, 255 figure and ground 3, 8, 51, 56, 71, 109, 122, 236, 266, 270, 273, 296–298, 300, 301, 305–308, 311, 312, 323, 393, 400, 401 first language acquisition 81, 199, 260, 261, 265, 361, 389 flexibility 14, 385, 399, 404, 407, 408 FLT (see foreign language teaching) focus construction (see construction) focus on form 39–41, 259, 260, 262–265, 278, 284, 379 focus on formS 260, 264 focus on meaning 260, 261, 264 explicit focus on form 263 implicit focus on form 263 proactive focus on form 262, 278, 284 reactive focus on form 262 foreign language teaching (also FLT) 31, 94, 95, 197, 223, 259, 262,

434

Subject index

265, 277, 284, 296, 306, 318, 392, 408 formal error (see error) form-focused activities (see teaching) form-meaning connection 263, 278, 279, 334, 345, 351 form-meaning mapping 259, 264, 265, 272, 276, 277, 284, 348 fossilization 55, 81, 328 frame 34, 43, 79, 80, 126, 143, 152, 242, 265, 268, 318, 332, 387, 412 French as L2 (see L2 French) French learner of German 3, 303, 305, 309, 313 frequency 58, 74, 79, 83, 92, 97, 98, 104, 105, 107, 199, 201, 207, 208, 210–214, 222, 224, 225, 237, 238, 260, 279, 399 “frog stories” 273, 298, 385, 393, 395, 404–406, 409 G generalization 13, 26, 27, 44, 53, 57, 74, 80, 84, 100, 106, 125, 156, 222, 244, 290, 320, 404 GENERIC FOR SPECIFIC 130 German as L2 (see L2 German) German learner of English 3, 201, 208, 327 Germanic language 3, 269, 295, 297, 298, 301, 319 gerundial complement construction (see construction) gerundial complement/infinitive construction alternation (see alternation) gesture 276, 285, 291, 316, 318 goal 13, 49, 70, 73, 100, 128, 148, 204, 266, 268, 270, 296, 297, 301, 302, 305–310, 313–317 gradation 15, 122

graded input 84 grammar 1–5, 7, 8, 15, 20, 21, 23, 25, 26, 28, 29, 37–41, 45, 47, 49, 51–53, 55–58, 61, 62, 67–69, 72–74, 76, 79–83, 85–87, 98– 109, 111, 121, 122, 125, 129– 132, 139, 148, 155, 156, 158– 160, 165, 166, 177, 188, 191, 198, 222–227, 231, 250, 259, 260, 261, 263, 265, 278, 282, 284, 285, 295, 296, 304, 307, 318, 325–328, 330–334, 336, 341, 342, 345, 351, 352, 358, 362, 366, 368, 375, 381, 385, 390, 392, 408, 410 grammar instruction (also grammar teaching, see teaching) grammar model (see model) grammatical aspect (see aspect) grammatical construction (see construction) grammatical exercises (see teaching) grammatical metonymy (see metonymy) grammatical relation 71 grammaticality 43, 56 grammaticalization 155, 184, 357, 362, 364, 370, 371, 373–379, 381 grooming action (see action) ground (see figure and ground) grounding 121, 131, 148 gustar verbs 155, 159, 177, 178 H HAPPY IS UP 131 high-level metaphor (see metaphor) high-level metonymy (see metonymy) highlighting 39, 121, 144, 145, 263 horizontal 302, 303, 308, 309, 315

Subject index

435

I ICLE (see International Corpus of Learner English) ICM (see idealized cognitive model) ICM of size 121 iconic(ity) 21, 22, 162, 163 idealized cognitive model (also ICM, see model) IDEAS ARE OBJECTS 126, 149 identification 45, 52, 411 idiom (also idiomatic expression) 15, 28, 31, 46, 69, 70, 81, 93, 95, 96, 232, 235, 315, 316, 347 image schema 124, 126, 129–131, 231, 234, 235, 266 imagery 37, 51, 385 imperfective (also imperfectivity) 3, 173, 343, 357, 361–363, 365– 372, 374, 378–380 implicit focus on form (see focus on form) IMPORTANT IS BIG 131 INACCESSIBILITY IS A (BLOCKING) CONTAINER 316 inalienable possession 155, 164, 167 inchoative 121, 139–144, 146–148, 155, 158, 165, 179, 192, 392, 402 inchoative construction (see construction) inchoative verb (see verb)

278, 281, 282, 284, 312, 313, 326, 327 input activities (see teaching) input enhancement 263 input flood 263 instruction (see teaching) explicit instruction (see teaching) processing instruction (see teaching) task-based instruction (see teaching) INSTRUMENT FOR ACTION 130 intake 40, 281 intake vs. input 40, 41 interaction pattern 94, 126, 295 interference 203, 206, 235, 242, 243, 346, 351, 391 International Corpus of Learner English (also ICLE) 197, 206, 209–218, 221, 222, 225, 228 intransitive 10, 11, 18, 73, 76, 86, 141, 157, 159, 165, 168, 176, 177, 180, 185, 190, 192, 392, 398, 399, 406 intransitive verb (see verb) introspection 93 intuition 93, 107, 208, 410 invariance principle 121, 127, 129 inventory 1, 25, 28, 45, 46, 69, 87, 190, 198, 389

INDIVIDUAL ENTITY FOR A COLLECTION INCLUDING THAT ENTITY

K

130 inference 186, 245 infinitive construction (see construction) infinitive construction/gerundial complement alternation (see alternation) infinitive construction/that-clause alternation (see alternation) input 8, 40, 79–81, 84, 97, 105, 108, 186, 199, 243, 263, 271, 276,

KNOWING IS SEEING

131

L LAD (see language acquisition device) landmark (see trajector and landmark) language acquisition 4, 39, 74, 81, 83, 92, 94, 95, 97, 108, 121, 155, 158, 190, 197, 199, 215, 216, 223, 224, 231, 253, 259–261, 265, 266, 269, 271, 272, 276–

436

Subject index

278, 284, 286, 346, 361, 386, 387, 389, 390, 405, 409, 410 language acquisition device (also LAD) 40, 41, 59 language teaching (also language pedagogy, see teaching) language typology 259, 295, 357 language-specific 10, 29, 39, 52, 60, 76, 77, 107, 357, 370 learnability 22, 108, 109, 228 learner’s grammar (see didactic grammar) learning 1, 2, 3, 7, 10, 15, 21, 25– 29, 38–41, 52, 53, 55, 60, 74, 81, 82, 84, 87, 94, 95, 98, 102, 106– 109, 121–123, 155, 157, 158, 166, 190, 191, 197–199, 208, 209, 216, 223, 224, 231–233, 236, 243, 248, 250–252, 257, 261, 263, 264, 271, 275, 286, 295, 296, 303, 304, 309, 312, 313, 316, 317, 319, 325–327, 329, 331–345, 347, 351, 357– 359, 365, 367, 373, 379, 380, 389, 407, 409 conscious learning 40 learning and acquisition 41, 325 learning difficulty 3, 53, 123, 251, 296, 305, 347 learning materials 107, 122, 123, 309, 312, 327, 345, 351 top-down/bottom-up learning 191 lexical approach (see approach) lexical aspect (see Aktionsart) lexical meaning 8, 9, 19, 367 lexicalization 267, 275, 276, 279 lexicalization pattern 267, 275, 279 lexicon and grammar 8, 28, 81, 318, 368, 390 literal meaning 238, 239, 241 location 11, 20–22, 73, 130, 155, 157, 164, 165, 179, 180, 185, 190, 266, 268, 271, 295, 296,

301–303, 305–307, 311, 313, 314, 318, 320, 338 location event 295, 296, 302, 303, 305, 306, 318 location schema 295, 302, 306, 307, 311, 318 low-level schema 26, 84 L2 Czech (also Czech as L2) 370, 380 L2 English (also English as L2) 226, 408 L2 French (also French as L2) 111 L2 German (also German as L2) 373 L2 instruction (also L2 teaching, see teaching) L2 Spanish (also Spanish as L2) 273 M manner 9, 46, 51, 59, 71, 86, 104, 125, 160, 183, 187, 188, 203, 266, 267, 269–276, 278–282, 285, 295, 297–300, 319, 336 manner of motion (see motion event) mapping 94, 126–129, 131, 145, 259, 264, 265, 271, 272, 276, 277, 284, 348, 385 mapping enforcement principle 129, 145 mass noun 1, 14, 37, 45–47, 49, 50, 54, 56, 59, 60, 173 mass/count noun alternation (see alternation) maximalist approach (see approach) meaning extension (also extension of meaning) 123, 334 meaning-focused activities (see teaching) meaning-form connection (see formmeaning connection) meaningfulness of grammar 7, 8, 15 mental construction (see construction)

Subject index mental image 155, 164, 165 mental scanning 313 mental space 7, 13, 79, 157, 160, 325, 336, 338, 341, 348 mental space model (see model) mental space theory 336 metalanguage 101, 106 metaphor(ical) 3, 7, 13, 14, 15, 51, 54, 69, 73, 95, 97, 121, 124, 126–131, 143, 144, 146, 148, 182, 231–235, 237, 240–242, 244, 248, 249, 252, 295, 296, 302, 312–314, 316, 318, 359 conceptual metaphor 231, 232, 233, 237, 244, 313, 316–318 CONDUIT metaphor 14, 54, 314 constraints on metaphor 121, 148 CONTAINER metaphor 313 high-level metaphor 121, 128, 130, 131, 143, 148 primary metaphor 131 SURFACE metaphor 316 metaphorical transfer (see transfer) metaphor understanding 150 metaphorized motion (see motion) metonymy (also metonymic) 14, 51, 121, 126, 127, 129–131, 135, 143, 144–146, 148, 233, 234, 309 constraints on metonymy 121, 148 grammatical metonymy 143 high-level metonymy 130, 143, 144 middle 3, 18, 121, 140–146, 148, 155, 157–183, 188–192, 392 middle/reflexive (see reflexive/middle) middle construction (see construction) middle evaluative construction (see construction)

437

modality 112, 248, 329, 336, 340, 341 model 2, 5, 42, 67–69, 75, 80, 83, 85, 87, 108, 121, 126, 130, 148, 156, 198, 233, 260, 265, 333, 336, 348 cognitive grammar model 336 grammar model 69, 336 idealized cognitive model 121, 126, 135, 137, 146, 265 mental space model 336, 348 network model 67, 68, 80, 83 usage-based model 2, 33, 67, 68, 71, 76, 80, 81, 123, 191, 222 morphosyntactic 45, 46, 55, 162, 301, 331, 359 motion 3, 7, 11, 12, 24, 32, 34, 49, 68, 70–73, 80, 106, 111, 126– 128, 130, 131, 155, 159, 164, 165, 174, 175, 178–180, 182, 183, 190, 191, 199, 201–203, 205, 206, 257, 259, 260, 266– 286, 295–303, 305–319, 376, 385, 402, 403, 409 abstract motion 299, 305, 312– 318 body part motion 295, 302, 307, 308, 317 construal of motion 3 fictive motion 7, 34, 295, 302, 305, 310–314, 322 metaphorized motion 305 non-translational motion 164 reduced motion 295, 302, 305 translational motion 164, 179, 182, 183, 266 motion at location (see motion event) motion event 3, 106, 259, 260, 266– 271, 274–277, 279, 281, 282, 284, 296–303, 305, 307, 309– 314, 319 cause of motion 266, 394, 396, 398, 399, 404

438

Subject index

manner of motion 125, 266, 267, 269–273, 275, 276, 278–281, 283, 295, 297–299 motion at location 305 motion to goal 305 path of motion 266, 268, 270, 274, 275, 279, 280, 295, 297, 298, 300, 301, 319 MOTION IS CHANGE OF LOCATION

313 motion picture 71, 385, 409 motion schema 295, 302, 306, 307, 311–313, 318 motion to goal (see motion event) motion verb (see verb) motivation 15, 22, 27, 37, 51, 58, 68, 70, 82, 84, 87, 98, 102, 105– 107, 121, 125, 143, 144, 146, 156, 163, 232, 318, 333, 334, 385, 409, 411 N narrative 188, 269, 273, 329, 335, 336, 341, 342, 358, 371, 388, 389, 392, 395, 396, 404–407, 410, 411 native speaker 23, 27, 40, 55, 67, 102, 157, 170, 178, 197, 199, 201, 207–209, 212–217, 222, 233, 236, 241, 248, 249, 251, 262, 263, 269, 271, 273–277, 285, 286, 327, 357, 358, 367, 370–373, 375–381, 385, 392, 394, 395, 397–406, 408, 409, 411 naturalistic setting 385, 407 network 9, 27, 28, 67–69, 73, 80, 82, 83, 87, 88, 96, 123, 125 network model (see model) non-arbitrariness of grammar 41, 83 non-boundary crossing 280 non-progressive (see aspect) non-reflexive (see also reflexive/middle) 18, 144, 157, 158

non-translational motion (see motion) noticing 82, 84, 85, 261, 263, 346 noun phrase 46, 67, 68, 70, 79, 218, 278, 388 noun-noun combination 79 O OBJECT FOR AN ACTION IN WHICH THE OBJECT IS INVOLVED 130

one-way map direction task (see teaching) one-way picture drawing task (see teaching) ongoingness 182, 361, 364, 377, 378 orientation 11, 58, 124, 262, 269, 302, 304, 342 ORGANIZATION IS PHYSICAL STRUCTURE 131

P packaging of information 389 part/whole relation (see relation) participant involvement 155 particle 15, 131, 172, 250, 268, 274, 275, 297, 300, 301, 365 particle verb (see verb) parts of speech 71, 77, 78, 88 passive 3, 16, 17, 75, 101, 108, 139, 140, 143, 146, 176, 179, 190, 192, 385, 387–391, 393, 396– 411 passive construction (see construction) passivization 75 path of motion (see motion event) pattern (see construction) pedagogical grammar 1–4, 37–41, 47, 49, 52, 53, 55–58, 68, 74, 80, 82, 83, 85–87, 91, 95, 99–106, 109, 121, 122, 125, 132, 139, 148, 159, 231, 295, 296, 307, 328, 330, 332, 358, 392

Subject index pedagogical implication 7, 15, 25, 28, 37, 122, 197, 234, 407 PEOPLE ARE ANIMALS 128, 237, 248 perfective (also perfectivity) 3, 173, 191, 343, 357, 361–369, 371– 373, 379–381 perspective 7, 11, 20, 51, 94, 101, 109, 121, 371, 372, 385–389, 392–394, 396–399, 402–405 perspectivization 373 phrasal verb (see verb) pivotal moment of change 155, 165, 180, 181, 184, 185, 192 polysemy (also polysemous) 9, 16, 18, 23, 44, 51, 68, 72, 73, 96, 123, 334, 348, 381 potentiality 340, 341 pragmatic 42, 77, 81, 98, 110, 148, 236, 265, 347, 361, 385, 386, 391, 408, 411 predicate 34, 44, 48, 56, 78, 83, 225, 267, 343, 359–361, 363– 365, 376, 377, 390 predictability 27 preferential 220, 301, 305, 372, 379 preposition 13, 17, 19, 22, 49, 51, 67, 70, 75, 82, 83, 84, 110, 128, 203, 204, 210, 215, 218, 297, 300, 301, 305, 307, 313, 314, 318, 320, 374 prepositional group/dative alternation (see alternation) primary metaphor (see metaphor) prioritization 91, 104, 106, 109 proactive focus on form (see focus on form) problem-solving skill (see teaching) PROCESS FOR ACTION 143, 144 PROCESS FOR ACTION FOR RESULT

143 processing instruction (see teaching) processing strategies 278, 280 production 5, 99, 103, 111, 126, 148, 157, 198, 208, 216, 217,

439

225, 259, 260, 262, 263, 272, 277–279, 282, 333, 357, 370, 371, 373, 379, 381, 389, 407– 409, 411 discourse production 389, 409 production activities (see teaching) production of L2 (see teaching) text production 384 productivity 26, 46, 243–245, 251 profiling 10, 19, 51, 390 progressive aspect (also progressivity, see aspect) prominence (also salience) 10, 17, 20, 23, 99, 140, 142, 145, 147, 178, 245, 269, 329, 342, 347, 358, 372, 388, 391, 397 prototype 17, 18, 27, 37, 47, 50, 67, 68, 77, 82–84, 87, 99, 108, 123, 164, 165, 179 prototypicality 22, 47, 98, 107 proximal/distal 342 proximity 342 psycholinguistic(s) 4, 67, 98, 107, 108, 312, 316, 357, 359, 371, 381 psychological reality of constructions (see construction) Q qualitative method 97, 98, 104, 197, 209, 214, 215, 400–403 quantitative method 97, 98, 104, 245, 397–400 R Radical Construction Grammar 2, 4, 29, 67, 68, 70, 72, 75–77 reactive focus on form (see focus on form) reality 13, 32, 86, 98, 99, 126, 222, 223, 265, 310, 314, 325, 336– 342, 346, 348, 357, 359, 381 recurrence 92, 97

440

Subject index

reduced motion (see motion) reference 11, 12, 17, 23, 39, 50, 77, 78, 93, 94, 96, 100, 101, 106, 108, 109, 130, 159, 161, 206, 208, 222, 235, 236, 248, 268, 277, 332, 336, 346, 347, 374, 375 reference point 17, 23 reference time 332, 375 reflex passive 121, 139, 140–147 reflexive 3, 33, 121, 122, 138, 139, 143–145, 147, 155–163, 166, 170, 175, 177–180, 188, 190, 191 reflexive/middle 163 reflexive construction (see construction) reflexive verb (see verb) register 93, 104 regularity 26, 27, 96, 319, 325, 334 relation 17–20, 23, 51, 67, 68, 70, 71, 73, 76, 83, 121, 122, 148, 198, 242, 250, 260, 264, 266, 278, 279, 284, 297, 305, 311, 333, 335–338, 358, 385, 389, 401, 406–409 part/whole relation 18, 20, 22, 23, 76, 124, 168 spatial relation 16, 51, 124, 305, 313 temporal relation 71, 346, 348, 358 relevance time (see time) right boundary 364, 369, 371, 372, 378, 380, 381 Romance language 163, 295, 297, 298, 371 rule(s) 1, 2, 5, 28, 37, 39–42, 53, 55, 56, 68, 72, 96, 102, 103, 105, 106, 108, 124, 125, 156, 180, 186, 190, 192, 216, 218, 221, 223, 260, 264, 306, 311, 314, 327, 331, 332, 334, 346, 362, 392, 404, 408

Russian 162, 163, 210, 269, 357– 360, 362, 366–373, 379–381 Russian learner of German 359, 373 S salience (see prominence) satellite 266, 268, 274, 275, 279, 280, 297–301, 310, 311, 316, 319 satellite-framed language (also Slanguage, satellite-oriented language) 266–269, 272, 273, 276– 279, 298, 301 satellization 274 schema 1, 2, 7, 18, 19, 25, 26, 28, 37, 46, 47, 49, 68, 69, 76, 80, 83, 123, 124, 126, 129, 130, 131, 145, 166, 170, 171, 185, 189, 190, 231, 234–236, 266, 289, 296, 301, 302, 306, 307, 311– 313, 318–320, 332, 334, 342 schematic(ity) 10, 15, 17, 19, 22, 23, 25–28, 37, 43–47, 49–51, 59, 69, 70, 71, 74, 80, 81, 124, 126, 130, 135, 179, 180, 190, 198, 225, 297, 344 school grammar (see didactic grammar) second language acquisition (also SLA) 81, 83, 94–96, 98, 103, 108, 113, 121, 150, 155, 158, 191, 199, 215, 223, 224, 231, 253, 259, 260, 261, 263, 265, 266, 271, 284, 346, 347, 389, 390, 405, 410 self-beneficial event (see event) self-benefit action (see action) self-directed action (see action) semantic composition 19 semantic network 82, 121, 123, 189 sentence-sorting experiment 199– 206 ser-passive construction (see construction)

Subject index simultaneity construction (see construction) situation time (see time) skill (see teaching) S-language (see satellite-framed language) SLA (see second language acquisition) Slavic language 269, 319, 329, 343, 359, 362, 365–367, 369, 378, 381 source domain (see domain) SOURCE-PATH-GOAL schema 266, 301, 306 space 13, 52, 54, 60, 77–80, 124, 130, 266, 273, 313, 318, 336– 339, 341, 342, 346, 348 Spanish as L2 (see L2 Spanish) Spanish learner of Danish 272, 273, 276, 277 Spanish learner of English 124, 197, 199, 201, 203, 206, 209, 215, 216, 221, 222, 276 spatial configuration 21, 22, 71 spatial domain (see domain) spatial image 131, 295, 316 spatial relation (see relation) speaker involvement 155 specificity 9, 46, 47, 51, 69 speech time (see time) state 12, 69, 73, 80, 86, 107, 128, 130, 148, 155, 160, 161, 164– 166, 170, 175, 177–179, 181, 185, 188, 190, 236, 239, 267, 335, 337–339, 343–345, 360, 361, 363, 369, 372, 376, 377, 394, 397–399, 403, 404 change of state 80, 160, 161, 165, 178, 181, 236, 267, 361, 363, 394, 404 emotional state (see emotion) state verb (see verb) temporary habitual state 339 types of state 360

441

unbounded state 343–345 130 static situation 24 static/dynamic opposition (see dynamic/static opposition) structural device 404, 409 subject/auxiliary inversion 82, 84, 85 subjectification 155, 183 subjectivity 155 surface 11, 15, 42, 124, 125, 126, 271, 308, 309, 313–317, 372, 408 SURFACE metaphor (see metaphor) susceptibility to change 345, 349 syllabus design 231, 232, 250, 291 symbolic structure 8, 18, 26, 69 symbolic unit 37, 45–47, 52, 69, 265 syntax-lexicon continuum 68, 70 STATES ARE POSSESSIONS

T target domain (see domain) task (see teaching) task-based instruction (see teaching) teacher as facilitator 87 teacher’s grammar (see didactic grammar) teaching (also instruction, teaching method, teaching strategies) 1–3, 7, 8, 14, 15, 21, 25, 29, 37, 39, 67, 68, 80–84, 86, 87, 94, 95, 102–104, 109, 119, 121–123, 138, 147, 148, 155, 156, 158, 159, 166, 177, 188, 191, 200, 202, 219, 223, 232, 243, 249, 259–266, 277–280, 282, 284– 286, 290, 295, 312, 317–319, 325, 326, 328, 330, 333–345, 348, 351–351, 358, 361, 363, 367, 379, 390, 395, 401, 407– 409 communicative language teaching 102, 261, 262

442

Subject index

communicative-based 87 completion exercises 186–188, 240 comprehension of L2 180, 188, 190, 373 conceptual teaching 3, 119 decision-making task 283 drill(ing) 67, 81, 87 exercises 41, 102, 106, 125, 186, 188, 191, 223, 239, 240, 249, 250, 348 explanation in language teaching 47–49, 56–58, 106, 108, 111, 124, 125, 133–135, 143, 190, 236, 249, 278, 279, 311, 312, 314, 317, 330, 332, 333 explicit grammar teaching 261 explicit instruction (also explicit teaching) 81, 138, 148, 233, 263, 411 focus on form (see focus on form) foreign language teaching (see foreign language teaching) form-focused activities 41, 262 grammar instruction (also grammar teaching) 39, 67, 82, 103, 155, 156, 158, 177, 259, 260, 261, 263, 278, 282, 284, 285, 326–328, 333–345, 390 grammatical exercises 41, 102, 106, 186, 223, 250, 348, 349 input activities 279, 281 language teaching (also language pedagogy) 3, 4, 7, 8, 14, 25, 37–39, 52, 67, 80–83, 87, 92, 95, 96, 102, 116, 121, 122, 148, 197, 223, 253, 259, 261, 262, 264, 265, 277, 278, 280, 284, 296, 392, 408 L2 instruction (also L2 teaching) 123, 262, 291, 392, 409 meaning-focused activities 41, 262

one-way map direction task 283 one-way picture drawing task 283 problem-solving skill 122 processing instruction 263, 278, 279, 282, 286 production activities 278, 282 production of L2 157, 166–186, 260, 271–277, 358, 389–392 skill 42, 104, 109, 123, 217, 405 task 87, 105, 107, 167, 197, 201, 208, 216–218, 221, 222, 238, 241, 243, 244, 247, 252, 262–264, 271, 278, 282, 284, 286, 295, 348, 358, 370, 371, 373, 376, 381 task-based instruction 264, 284 test 8, 43, 97, 99, 145, 200, 203, 217, 220, 225, 307, 348, 370, 385, 410 traditional teaching 188, 260, 333 two-way picture sequencing task 283 two-way spot-the-difference task 283 visualization 60, 312, 316, 348 writing task 231, 240, 242–244, 247, 248 telic(ity) 142, 357, 363–365, 379 temporal meaning 333, 341, 376, 379 temporal relation (see relation) temporal structure 3, 257, 343 temporary habitual state (see state) tense 3, 11, 24, 26, 33, 46, 107, 210, 325–339, 341, 342, 346–348, 351, 352, 361–365, 372, 378, 381 test (see teaching) text comprehension (see discourse comprehension) text production (see production)

Subject index textbook grammar (see didactic grammar) that-clause construction (see construction) that-clause/infinitive construction alternation (see alternation) thinking for speaking time 5, 11, 16, 17, 24, 40, 50, 56, 72, 86, 98, 103, 106, 126, 133, 135, 142, 146, 172, 199, 200, 219, 240, 249, 261, 283, 284, 286, 301, 313, 330, 332–344, 346–348, 351, 360, 370, 372, 375–378, 381, 387, 388, 395, 407 relevance time 325, 335–342, 346, 348 situation time 335, 339, 341, 342 speech time 311, 325, 331, 335, 336, 338, 339, 341, 342 tool 98, 109, 115, 121, 122, 125, 148, 207, 223, 278, 284 top-down approach (see approach) trajector and landmark 11, 15, 17, 19, 20, 23, 71, 223, 283, 286 trajectory 158, 182, 270, 271, 295, 296, 297, 301, 306, 307, 309– 312, 318 transfer 27, 70, 108, 123, 130, 131, 167, 203, 204, 208, 224, 243, 244, 274, 275, 282, 346, 385, 391, 405, 406 transitive construction (see construction) transitive event (see event) transitive verb (see verb) transitivity 129, 130, 159, 179, 189, 384 translation 49, 54, 98, 147, 208, 224, 240, 243, 282, 285, 298, 299, 314 translational motion (see motion) Turkish 163, 176, 269, 270, 291, 298, 300, 392

443

two-way picture sequencing task (see teaching) two-way spot-the-difference task (see teaching) types of state (see state) typological approach (see approach) typological distinction 295 typological framework 259, 260, 266, 269, 287 typological variation 268, 269 typology 105, 142, 259, 269, 277, 279, 285, 295, 319, 323, 324, 357, 362 U unbounded/bounded alternation (see alternation) unbounded state (see state) unboundedness 350, 351 unconscious acquisition (see acquisition) understanding 14, 15, 24, 25, 41, 101, 102, 124, 126–128, 149, 266, 319, 325, 326, 331, 336, 351, 361, 408 UNDERSTANDING AN IDEA IS PERCEPTUALLY EXPLORING AN OBJECT

126 unexpected event (see event) unit of language 2, 25, 69, 71, 87 usage-based 2, 7, 8, 25, 26, 28, 58, 67, 68, 71, 76, 80–82, 92, 93, 95, 96, 123, 125, 155, 191, 198, 217, 326, 334 usage-based approach (see approach) usage-based grammar 222 usage-based model (see model) usage-based perspective 7, 25 usage event 7, 25, 69, 71, 80, 97 V vantage point 11

444

Subject index

verb 1, 11, 15–17, 24, 26, 27, 37, 38, 41–44, 46, 49, 55, 58, 60, 70–76, 78, 85, 86, 95, 107, 110, 111, 125, 128–130, 141, 142, 155, 157, 159, 163–167, 169, 171, 173, 175–178, 180, 181, 186, 187, 190–192, 197, 199– 204, 206–221, 223–225, 235, 249, 250, 262, 266–268, 270, 272–274, 276, 279–281, 285– 286, 296–302, 304–306, 309– 314, 317, 319, 320, 325, 328– 330, 337, 338, 342, 343, 359– 369, 372–378, 381, 390, 403 action verb 306 change-of-state verb 170 emotion verb 177 inchoative verb 192 intransitive verb 86, 141, 159, 165, 180, 190, 192 motion verb 24, 128, 267, 270– 272, 274–276, 279, 281, 283, 291, 298, 299, 301, 308, 310, 311, 319, 376, 402 particle verb 300 phrasal verb 15, 32, 107, 123 reflexive verb 18, 138, 139, 156 state verb 170, 363, 376, 377 transitive verb 11, 139, 157, 159, 166, 170, 174, 180, 185, 190, 396, 398 verb-framed language (also Vlanguage, verb-oriented language) 203, 266, 267, 269, 271– 274, 276–279, 298, 300, 301 vertical 52, 54, 58, 183, 274, 302, 303, 308, 309, 314, 315 visualization (see teaching) visualize 307, 312, 317, 318, 350 V-language (see verb-framed language) vocabulary 79, 81, 231, 242

W word class 2, 70, 71, 75, 138 word meaning 72 writing 91, 93, 99, 103, 210, 216, 231, 232, 240, 242–244, 247, 248, 250, 327, 347, 358, 366, 368 writing task (see teaching)

Cognitive Approaches to Pedagogical Grammar: A Volume in Honour of RenA© Dirven (Applications of Cognitive Linguistics)

Cognitive Approaches to Pedagogical Grammar: A Volume in Honour of René Dirven (Applications of Cognitive Linguistics)

Cognitive English Grammar (Cognitive Linguistics in Practice)

Phonology: A Cognitive Grammar Introduction (Cognitive Linguistics in Practice)

World Englishes: A Cognitive Sociolinguistic Approach (Applications of Cognitive Linguistics)

Cognitive Exploration of Language and Linguistics (Cognitive Linguistics in Practice)

Cognitive Exploration of Language and Linguistics (Cognitive Linguistics in Practice)

Cognitive Foundations of Grammar

Cognitive Linguistic Approaches to Teaching Vocabulary and Phraseology (Applications of Cognitive Linguistics)

Cognitive Poetics: Goals, Gains and Gaps (Applications of Cognitive Linguistics)

Cognitive Linguistics in Action: From Theory to Application and Back (Applications of Cognitive Linguistics)

Investigations in Cognitive Grammar

Methods in Cognitive Linguistics (Human Cognitive Processing)

Methods in Cognitive Linguistics (Human Cognitive Processing)

Investigations in Cognitive Grammar (Cognitive Linguistic Research, Volume 42)

Title Cognitive Foundations of Grammar

The Cognitive Linguistics Reader (Advances in Cognitive Linguistics)

Phonology: A Cognitive Grammar Introduction

New directions in cognitive linguistics

Cognitive Linguistics: An Introduction

Cognitive Linguistics: An Introduction

Cognitive Linguistics – Basic Readings

Cognitive Linguistics: An Introduction

Cognitive Linguistics. An Introduction

Cognitive Linguistics (Cambridge Textbooks in Linguistics)

The cognitive linguistics reader

Cognitive Foundations of Linguistic Usage Patterns: Empirical Studies (Applications of Cognitive Linguistics)

New Directions in Cognitive Linguistics (Human Cognitive Processing)

The Oxford Handbook of Cognitive Linguistics

Analogy as Structure and Process: Approaches in Linguistics, Cognitive Psychology and Philosophy of Science (Human Cognitive Processing, Volume 14)

New Directions in Cognitive Linguistics (Human Cognitive Processing)

Cognitive Approaches to Pedagogical Grammar: A Volume in Honour of RenA© Dirven (Applications of Cognitive Linguistics)

Cognitive Approaches to Pedagogical Grammar: A Volume in Honour of René Dirven (Applications of Cognitive Linguistics)

Cognitive English Grammar (Cognitive Linguistics in Practice)

Phonology: A Cognitive Grammar Introduction (Cognitive Linguistics in Practice)

World Englishes: A Cognitive Sociolinguistic Approach (Applications of Cognitive Linguistics)

Cognitive Exploration of Language and Linguistics (Cognitive Linguistics in Practice)

Cognitive Exploration of Language and Linguistics (Cognitive Linguistics in Practice)

Cognitive Foundations of Grammar

Cognitive Linguistic Approaches to Teaching Vocabulary and Phraseology (Applications of Cognitive Linguistics)

Cognitive Poetics: Goals, Gains and Gaps (Applications of Cognitive Linguistics)

Cognitive Linguistics in Action: From Theory to Application and Back (Applications of Cognitive Linguistics)

Investigations in Cognitive Grammar

Methods in Cognitive Linguistics (Human Cognitive Processing)

Methods in Cognitive Linguistics (Human Cognitive Processing)

Investigations in Cognitive Grammar (Cognitive Linguistic Research, Volume 42)

Title Cognitive Foundations of Grammar

The Cognitive Linguistics Reader (Advances in Cognitive Linguistics)

Phonology: A Cognitive Grammar Introduction

New directions in cognitive linguistics

Cognitive Linguistics: An Introduction

Cognitive Linguistics: An Introduction

Cognitive Linguistics – Basic Readings

Cognitive Linguistics: An Introduction

Cognitive Linguistics. An Introduction

Cognitive Linguistics (Cambridge Textbooks in Linguistics)

The cognitive linguistics reader

Cognitive Foundations of Linguistic Usage Patterns: Empirical Studies (Applications of Cognitive Linguistics)

New Directions in Cognitive Linguistics (Human Cognitive Processing)

The Oxford Handbook of Cognitive Linguistics

Analogy as Structure and Process: Approaches in Linguistics, Cognitive Psychology and Philosophy of Science (Human Cognitive Processing, Volume 14)

New Directions in Cognitive Linguistics (Human Cognitive Processing)

Recommend Documents