Up and down the Cline - The Nature of Grammaticalization

Up and down the Cline - The Nature of Grammaticalization

Typological Studies in Language (TSL) A companion series to the journal Studies in Language

General Editor

Michael Noonan

Assistant Editors

Spike Gildea, Suzanne Kemmer

Editorial Board Wallace Chafe (Santa Barbara) Bernard Comrie (Leipzig) R. M. W. Dixon (Melbourne) Matthew Dryer (Buffalo) John Haiman (St Paul) Bernd Heine (Köln) Paul Hopper (Pittsburgh) Andrej Kibrik (Moscow) Ronald Langacker (San Diego)

Charles Li (Santa Barbara) Edith Moravcsik (Milwaukee) Andrew Pawley (Canberra) Doris Payne (Eugene, OR) Frans Plank (Konstanz) Jerrold Sadock (Chicago) Dan Slobin (Berkeley) Sandra Thompson (Santa Barbara)

Volumes in this series will be functionally and typologically oriented, covering specific topics in language by collecting together data from a wide variety of languages and language typologies. The orientation of the volumes will be substantive rather than formal, with the aim of investigating universals of human language via as broadly defined a data base as possible, leaning toward cross-linguistic, diachronic, developmental and live-discourse data.

Volume 59 Up and down the Cline - The Nature of Grammaticalization Edited by Olga Fischer, Muriel Norde and Harry Perridon

Up and down the Cline The Nature of Grammaticalization Edited by

Olga Fischer Muriel Norde Harry Perridon University of Amsterdam

John Benjamins Publishing Company Amsterdam
/
Philadelphia

8

TM

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences – Permanence of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data New Reflections on Grammaticalization II Conference (2002) : University of Amsterdam) Up and down the Cline - The nature of grammaticalization / edited by Olga Fischer, Muriel Norde and Harry Perridon. p. cm. (Typological Studies in Language, issn 0167–7373 ; v. 59) Papers presented at the New Reflections on Grammaticalization II Conference held at the University of Amsterdam on April 4-6, 2002. Includes bibliographical references and indexes. 1. Grammar, Comparative and general--Grammaticalization-Congresses. I. Fischer, Olga. II. Norde, Muriel, 1968- III. Perridon, Harry. IV. Title. V. Series. P299. G73N484 2004 415-dc22 isbn 90 272 2968 6 (Eur.) / 1 58811 504 6 (US) (Hb; alk. paper) isbn 90 272 2969 4 (Eur.) / 1 58811 505 4 (US) (Pb; alk. paper)

2004041137

© 2004 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

TSL[v.20020404] Prn:7/05/2004; 13:22

F: TSL59CO.tex / p.1 (v)

Table of contents

Preface Introduction: In search of grammaticalization Olga Fischer, Muriel Norde, and Harry Perridon

vii 1

On directionality in language change with particular reference to grammaticalization Martin Haspelmath

17

Rescuing traditional (historical) linguistics from grammaticalization theory Brian D. Joseph

45

The English s-genitive: A case of degrammaticalization? Anette Rosenbach An investigation into the marginal modals dare and need in British present-day English: A corpus-based approach Martine Taeymans Redefining unidirectionality: Is there life after modality? Debra Ziegeler

73

97 115

From pronominalizer to pragmatic marker: Implications for unidirectionality from a crosslinguistic perspective Foong Ha Yap, Stephen Matthews, and Kaoru Horie

137

Conditionals and subjectification: Implications for a theory of semantic change Jacqueline Visconti

169

Unidirectionality in the grammaticalization of modality in Greek Anastasios Tsangalidis How cognitive is grammaticalization? The history of the Catalan perfet perifràstic Ulrich Detges

193

211

TSL[v.20020404] Prn:7/05/2004; 13:22



F: TSL59CO.tex / p.2 (vi)

Table of contents

Perfect and resultative constructions in spoken and non-standard English 229 Jim Miller Grammaticalization and standardization Lea Laitinen

247

External factors behind cross-linguistic similarities Ilona Herlin and Lari Kotilainen

263

What constitutes a case of grammaticalization? Evidence from the development of copulas from demonstratives in Passamaquoddy Eve Ng

281

Multi-categorial items as underspecified lexical entries: The case of Kambera wàngu Marian Klamer

299

The acquisition of polysemous forms: The case of bei2 (“give”) in Cantonese Kwok-shing Wong

325

Phonetic absence as syntactic prominence: Grammaticalization in isolating tonal languages Umberto Ansaldo and Lisa Lim

345

Grammaticalization of word order: Evidence from Lithuanian Sergey Say

363

Language index

385

Name index

387

Subject index

393

TSL[v.20020404] Prn:13/04/2004; 15:27

F: TSL59PR.tex / p.1 (vii)

Preface

The 1980s and 1990s witnessed a rapidly increasing interest in grammaticalization phenomena. Such was the number of people engaged in this branch of historical linguistics that an entire conference was dedicated to it – the New Reflections on Grammaticalization Conference, held at the University of Potsdam in 1999. The proceedings of this conference, edited by Ilse Wischer and Gabriela Diewald, appeared in the Typological Studies in Language series in 2002 with John Benjamins. Due to its success, a follow-up to this conference was organized by the editors of this volume at the University of Amsterdam, April 4–6, 2002, which was attended by over a hundred participants. Judging by the vast range of theoretical issues and languages covered, as well as the liveliness of the discussions, the popularity of grammaticalization studies and its significance for the field of historical linguistics looks unabated. This volume contains a selection of the papers presented at the New Reflections on Grammaticalization II Conference. Rather than presenting a random compilation, we have chosen to make this a thematically coherent volume, and invited papers that specifically addressed topics relating to a modification of the ‘theory’ of grammaticalization, such as the principle of unidirectionality, alternative grammaticalization pathways (or chains), the synchronic (internal or external) factors that may interfere with what is seen as ‘usual’ in grammaticalization processes, and internal or universal factors which make grammaticalization a truly separate mechanism in language change. We have selected seventeen papers, written by leading theoreticians and specialists covering a number of different language families. A large number of people have assisted and supported us in the organization of the conference and in editing this volume. We gratefully acknowledge the financial support of the Royal Netherlands Academy of Arts and Sciences (KNAW). We would also like to thank all the contributors and participants, who by their expertise and enthusiasm made the conference such a success. Special thanks are due to the students who helped us during the conference: Margot van den Berg, Annerieke Boland, Corrien Blom, Pieter Claeys, Marion Elenbaas, Tessa Liebman and especially Ernie Ramaker, who also took care of the conference website. We would also like to thank Anke de Looper and Kees Vaes of Benjamins Publishers and the editor of the series for their help towards the publication of this volume, and Jetty Peterse for her help with the index. Finally, we are particu-

TSL[v.20020404] Prn:13/04/2004; 15:27

F: TSL59PR.tex / p.2 (viii)

 Preface

larly indebted to all the referees who generously commented on the first round of papers and made innumerable suggestions for improvement: Johan van der Auwera, Adrie Barentsen, Lyle Campbell, Holger Diessel, Marcel Erdal, Bernd Heine, Kees Hengeveld, Brian Joseph, Ans van Kemenade, Marian Klamer, Roger Lass, Jouko Lindstedt, Bettelou Los, Helena Raumolin-Brunberg, Eva SchultzeBerndt, Elizabeth Traugott, Seongha Rhee, Sali Tagliamonte, Jeroen Wiedenhof and Debra Ziegeler.

Amsterdam, November 2003 Olga Fischer Muriel Norde Harry Perridon

TSL[v.20020404] Prn:13/04/2004; 15:31

F: TSL5901.tex / p.1 (1)

Introduction In search of grammaticalization Olga Fischer, Muriel Norde, and Harry Perridon University of Amsterdam

.

Introduction

During the past two decades, the revived interest in grammaticalization phenomena has firmly established itself as one of the most flourishing branches of historical linguistics. It boasts some of the most widely used textbooks in the field (Heine et al. 1991; Hopper & Traugott 2003 [1993]), as well as a veritable flood of case-studies and now even its own lexicon (Heine & Kuteva 2002). Judging by the sheer volume of pages devoted to it, it would seem that grammaticalization ‘theory’ provides the principal framework to account for the origin of and changes within grammar and hence a forceful tool for reconstruction. And indeed, in the early days of grammaticalization studies it was generally assumed that it had generated (virtually) exceptionless laws of grammatical change (e.g. Lehmann 1995 [1982]; Traugott & Heine 1991; Haspelmath 1999). In present-day historical linguistics there are no indications of a fading interest in grammaticalization – our notion of grammaticalization is continuously being expanded and refined (see e.g. Giacalone Ramat & Hopper 1998 and Wischer & Diewald 2002). But perhaps more importantly, its very nature as a ‘theory’, and some of its seemingly unchallengeable principles, including the principle of unidirectionality, have increasingly become the object of critical examination. In 2001 the journal Language Sciences devoted a special issue to grammaticalization in order to provide “a critical assessment”. The editor of this issue (Lyle Campbell) along with the contributors (Richard Janda, Brian Joseph, Frederic Newmeyer and Muriel Norde) pointed to a number of problems that the theory raises. These pertain to some widely accepted principles of (the theory of) grammaticalization, such as the principle of unidirectionality and the idea that grammaticalization is primarily driven by semantic or pragmatic factors, but the

TSL[v.20020404] Prn:13/04/2004; 15:31



F: TSL5901.tex / p.2 (2)

Olga Fischer, Muriel Norde, and Harry Perridon

contributors to this issue also took stock of the phenomenon of grammaticalization itself. They critically evaluated the claim that grammaticalization is a unitary, continuous process and that it has explanatory value. They also considered the mechanisms that are said to play a role in grammaticalization (such as semantic bleaching, phonetic reduction, analogy and reanalysis), to see in how far they are connected or can be seen to work independently. The issues raised were not settled in this volume, which threw up quite a bit of dust in the grammaticalization field. It seemed appropriate therefore to take up some of these issues again by means of case studies, paying special attention to both the philological details involved in any development that could be seen as a case of grammaticalization as well as the theoretical implications of each case. This has resulted in the present collection of papers, some of them pursuing the points raised in the Language Sciences volume, some of them, on the other hand, criticizing these points and the examples that had been brought up to support them. What all the papers have in common, however, is that they present truly new reflections on grammaticalization, enriching our notion of what (de)grammaticalization entails, and what it does not entail.

. The principle of unidirectionality One of the most pressing themes in recent research on grammaticalization is whether or not it is possible for a grammatical item to become less grammatical, in other words, to move ‘up’ on Hopper and Traugott’s (2003 [1993]) ‘cline of grammaticality’ (content item > grammatical word > clitic > inflectional affix). In earlier work it was generally believed that such degrammaticalization is not possible, see e.g. Lehmann (1995 [1982]: 19): “[. . . ] no cogent examples of degrammaticalization have been found. This result is important because it allows us to recognize grammaticalization at the synchronic level. Given two variants which are related by the parameters of grammaticalization [. . . ] we can always tell which way the grammaticalization goes, or must have gone. The significance of this for the purposes of internal reconstruction is obvious [. . . ].” In recent years, however, the unidirectionality hypothesis has been criticized both on theoretical (e.g. by Newmeyer 1998) and methodological grounds (e.g. Lass 2000) as well as with reference to a number of well-described changes in which the directionality of grammatical change appears to be reversed. Martin Haspelmath’s paper in the present volume acknowledges that some of this criticism is justified and that some counterexamples do exist, which implies that an absolute universal has to be weakened to a statistical universal, but he emphasizes at the same time that the number of exceptions is very small. Referring to ‘true’ exceptions to unidirectionality as cases of ‘antigrammaticalization’, he lists eight such

TSL[v.20020404] Prn:13/04/2004; 15:31

F: TSL5901.tex / p.3 (3)

Introduction

cases, including the English and Continental Scandinavian s-genitive (from affix to clitic) and Irish muid (from 1st person plural subject suffix to independent pronoun). For all other cases that have been put forward as exceptions Haspelmath identifies seven types of change that according to him are not instances of antigrammaticalization, such as ‘delocutive’ word-formation (i.e. a metalinguistic use of grammatical words, as in ifs and buts, isms), back-formations (teens, burger) or adverb-to-verb/noun conversions (to up, to off ). Another type of change which is not antigrammaticalization in Haspelmath’s view is a process he terms ‘retraction’, which means that less grammaticalized items regain ground at the expense of their more grammaticalized counterparts. While accepting the existence of counterexamples, Haspelmath concludes that unidirectionality is still a useful generalization, as the main goal of historical linguistics ought to be an understanding of language change, not merely describing individual changes, and that for this reason it is important to look for universals of language change. Many of these universals take the form of directionality constraints, including the assumed unidirectionality of grammaticalization. Looking at the phenomenon of grammaticalization from a traditional historical linguistic point of view Brian Joseph doubts that grammaticalization theory has added much to our understanding of the processes and mechanisms that play a role in language change. Joseph shows, first of all, that traditional, Neogrammarian methods were not only concerned with the phonetic/phonological level, as is often thought, but also with the grammatical level. In other words, it is not the case that the theory of grammaticalization fills a gap in the traditional historical methods. Similarly, he indicates that grammaticalization theory makes use of notions, such as phonetic reduction, analogy and reanalysis, that were well understood in earlier theories. The only difference is that within grammaticalization theory these notions or mechanisms are linked together so as to form one continuous process of change rather than a series of separate steps, not necessarily connected, within a particular development. Joseph’s main point, however, is that this combining of mechanisms, which is so typical for grammaticalization theory, often leads to an ahistorical way of looking at changes in that there is a tendency not to take full account of all the different steps present in any particular development. There is a real danger within grammaticalization studies to give greater weight to typological or cross-linguistic tendencies than to the details of the language undergoing the development in question. This may result in a neglect of details, both of a historical and a synchronic nature. Concerning the latter, Joseph shows that grammaticalization studies are indeed often also asynchronic in that they do not take into consideration the synchronic system within which a particular construction functions and develops. In her paper on the historical development of the English s-genitive, which is often quoted as a counterexample to the unidirectionality hypothesis (among



TSL[v.20020404] Prn:13/04/2004; 15:31



F: TSL5901.tex / p.4 (4)

Olga Fischer, Muriel Norde, and Harry Perridon

others, by Haspelmath in this volume), Anette Rosenbach seems to pay heed to Joseph’s admonitions. Following Allen (1997), she assumes that this development proceeded in several steps and that it is closely interrelated with other developments within the NP, in particular the loss of NP-internal agreement and the rise of phrasal N+N constructions such as brother-in-law. She furthermore shows that the development towards a phrasal possessive marker is closely interwoven with the new function of -s as a definite determiner. As to the status of the Modern English s-genitive, Rosenbach argues that -s is a ‘clitic-like’ element – it is clearly no longer an inflectional suffix, as it was in Old English, but not (yet) a full clitic either. In order to answer the question of whether this development ought to be regarded as a case of degrammaticalization, Rosenbach introduces the distinction between ‘type grammaticalization’ and ‘token grammaticalization’ (based on Andersen’s 2001 distinction between ‘change schema’s’ and ‘historical changes’). As far as the s-genitive is concerned, the relevant change schema is Hopper and Traugott’s (2003[1993]) ‘cline of grammaticality’. According to Rosenbach (contra Norde 2001), English -s did not move back along this cline from affix to clitic, but instead became part of the newly emerging article system. Following the type-token distinction, Rosenbach argues that -s looks like a case of degrammaticalization on the token level, but since she defines unidirectionality as a property of types (i.e. clines), and since a token can form part of several clines, she concludes that unidirectionality on the type level is preserved. She stresses, however, that the interpretation of the history of the s-genitive depends heavily on the definition of (de)grammaticalization, and that therefore priority should always be given to the data, which may give rise to different interpretations. Another frequently discussed possible counterexample to the unidirectionality of grammatical change is the topic of Martine Taeymans’ paper on the English marginal modals dare and need and the semi-modals dare to and need to. Drawing data from the British National Corpus World Edition (BNC), which contains approximately 10 million words of spoken and 90 million words of written presentday British English, she shows how DARE and NEED oscillate between main verb and modal verb morphology and syntax. Their histories are totally different, however – whereas NEED is originally a main verb which acquired modal characteristics, DARE derives from an Old English preterite-present which acquired main verb characteristics at a later stage. For this reason, the history of DARE is commonly regarded as an example of degrammaticalization (e.g. in Beths 1999). From the data on NEED it becomes evident that semi-modal need to has become the predominant form, at the expense of modal need. Though this appears to violate the unidirectionality hypothesis, Taeymans argues that unidirectionality is observed on the semantic level, in that need to changes its meaning from internal to external necessity (previously a function of modal need). According to van der Auwera and Plungian (1998) such a shift is in accordance with claims about semantic unidi-

TSL[v.20020404] Prn:13/04/2004; 15:31

F: TSL5901.tex / p.5 (5)

Introduction

rectionality. The author’s frequency analysis of DARE, on the other hand, does not yield conclusive results. Assuming that a difference between written and spoken PDE reflects a diachronic development, it would seem that modal use is increasing (52 per cent in the spoken parts of the BNC versus 35.6 per cent in the written ones, excluding the petrified expressions How dare you and I dare say). Taeymans points out, however, that an increase in modal dare seems counter-intuitive in the light of two facts. First, modal verbs tend to be replaced by periphrastic expressions and secondly, modal dare has become very rare in American English, which is generally considered less conservative than British English. Taeymans is therefore unable to identify a clear direction in which DARE may be considered to be heading. As a result, it cannot be established whether DARE is grammaticalizing or degrammaticalizing.

. Semantic change In her article on a seemingly counterdirectional change in the meaning and use of the Chinese modal dei/dé Debra Ziegeler arrives at the same conclusion as Taeymans, viz. that the hypothesis is not at risk if unidirectionality is defined at the semantic level. In Mandarin Chinese dé was originally a main verb, meaning ‘to obtain’, later it developed a permissive meaning when followed by a verb (phrase), which in its turn gave rise to the meaning of obligation. In present-day Mandarin the modal can be used epistemically, but, what is more surprising, also as a (new) main verb meaning ‘to need’, which need not be followed by another verb but can also take a noun (phrase) as its complement. If category change is automatically defined as grammar change we are faced here with a clear counter-example to the unidirectionality hypothesis, since one and the same item turns from a main verb (‘to obtain’) into an auxiliary (‘be allowed’, ‘have to’) back into a main verb (‘to need/require’) again. Assuming that grammaticalization begins with conceptual changes Ziegeler suggests that semantic change is unidirectional in the sense that a meaning A can change into meaning B, but that this meaning B cannot change back to meaning A again. These changes may occasionally lead to less specific meanings, in which case they may or may not be accompanied by changes in form. A series of such changes may give rise to a grammaticalization chain. But the endpoints of such chains may be subject to new changes, which eventually may lead to a new lexification of the item in question.1 Unidirectionality of semantic change as defined by Ziegeler is also assumed by Foong HaYap, Stephen Matthews and Kaoru Horie in their paper on the ways in which (pro)nominalizers in three unrelated Asian languages (Japanese, Mandarin Chinese and Malay) turn into pragmatic markers. Japanese no is originally



TSL[v.20020404] Prn:13/04/2004; 15:31



F: TSL5901.tex / p.6 (6)

Olga Fischer, Muriel Norde, and Harry Perridon

a genitive marker: Taroo no hon ‘Taro’s book’, which turns into a kind of pronoun when the head noun is not expressed: Taroo no ‘Taro’s one, ‘the one that is Taro’s; in a later stage a clause may take the place of the possessor noun: Taroo-ga kat-ta no ‘[Taro bought]’s one, the one Taro bought’. It developed further into a complementizer, which presents the nominalized sentence as a fact. Used without a matrix verb, no functions as a stance marker, expressing the unchallengeable status of the information encoded in the sentence preceding no. Chinese de has approximately the same range of meanings, but parts of its history are different. Originally a noun meaning ‘bottom, end’, it came to be used as a demonstrative and an interrogative, and later on as a relative, and as a nominalizing element. From this pronominal use evolved its function as a genitive marker. As a stance marker de has the same value as Japanese no. Malay empu-nya was originally a noun phrase consisting of the noun empu ‘master’ and a third person clitic, which became a verb (‘to own’), and later on a genitive marker, a nominalizing pronoun and finally a stance marker. The various paths taken by Japanese no, Chinese de and Malay (em)punya all lead to the same goal, which raises some doubts as to the universality of the direction of semantic change. Yap et al. try to save the notion of unidirectionality in semantic change by claiming that it holds for changes across ontological domains, but not for changes within one and the same domain. Since both the relative, the pronominal and the genitive uses of the elements in question belong to the same domain of first-order entities (persons, things and the like, cf. Lyons 1977) grammaticalization may (and does) proceed in one way or the other. Changes from uses with third order entities (propositions) to uses with second or first order entities, e.g. from complementizer to relative pronoun, or genitive marker, are on the other hand not to be expected. A more modest view on unidirectionality in semantic change is expressed by Jacqueline Visconti. In a detailed description of the development of the verb suppose from an objective, lexical verb via a subjective, evaluative verb into a conditional connective (or conjunction), she shows how the path of development is an instance of subjectification similar to the process described for modals, deictic elements, discourse markers etc. in such works as Sweetser (1990), Kuteva (2001) and Traugott and Dasher (2002). In each case the original lexical element first operates in the propositional domain, and moves via the textual domain into the epistemic domain. Visconti shows that in such a subjectification process, it is the semantic pragmatic change that leads the way, while the grammatical developments follow. The process is made visible by indicating the shifts in contexts the verb undergoes (suppose appears more and more often in interrogative and imperative contexts, and contexts containing ‘related free adjuncts’), the way in which the verb may be coordinated with other verbs and adverbs, and the changes in ‘control’ and in the operation of scope from the narrative subject to the Speaker/Writer (who need not be the syntactic subject of the clause). Visconti indicates the generality of this

TSL[v.20020404] Prn:13/04/2004; 15:31

F: TSL5901.tex / p.7 (7)

Introduction

development by comparing the process in English with developments involving the cognates of the verb suppose in French and Italian. In all cases the stages are the same. Visconti argues, like Ziegeler, that certain types of semantic change (in this case subjectification) may be unidirectional. This semantic unidirectionality should not be seen as a principle of this type of grammaticalization but rather as a strong tendency that may serve as a useful heuristic tool. Serious doubts as to the assumed unidirectionality of semantic change are expressed by Anastasios Tsangalidis, who shows that the diachronic development of modal markers in Modern Greek poses a number of problems for the ‘unidirectionality’ principle. Applying Bybee et al.’s (1994) path of development suggested for grammaticalization within the domain of modality, to the Greek case, he comes to the conclusion that the suggested pathway, which involves a conceptual cline (from ‘agent-oriented’ via ‘speaker-oriented’ and ‘epistemic’ to ‘subordinate’) as well as a formal one (from lexical modal elements to particles and affixes) is not always followed in Greek. The first problem encountered is the imperative mood, which is expressed inflexionally but does not seem to have been derived from earlier lexical material, as predicted by the pathway. Also problematic is the development of tha from a lexical verb expressing ‘desire’ into a pure future marker, where the formal reduction of thela to tha does not neatly follow the conceptual development. The same is true for the development of na, which grammaticalized from a purposive conjunction (hina ‘so that’) to a general purpose subordinator (‘that’) and to a particle expressing various modalities, and was even reduced to semantically zero. Again the formal changes do not historically correlate with the conceptual changes. Especially troublesome is the fact that subordinating uses of na are historically earlier than the main clause uses, which according to Bybee et al.’s pathway should be the other way around. Like Tsangalides, Ulrich Detges questions the idea that the forces which trigger grammaticalization are primarily cognitive in nature, and more especially the suggestion of theorists like Bernd Heine and Tania Kuteva that grammaticalized structures for e.g. tense and aspect derive from a small number of basic ‘source concepts’, and that the grammaticalization development from source to target is pretty similar across languages.2 In his article on the development of the Catalan perfet perifràstic (va cantar, ‘he goes sing’, i.e. ‘he sang’), Detges first of all shows that in Catalan the go (+to) infinitive does not follow the ‘normal’ route from sourcepath-goal to ‘future’ but instead develops into a past tense marker. Secondly, he argues that the driving forces are discourse-pragmatic rather than cognitive, showing how the perfet peripfàstic (the past tense) in Catalan developed in narrative texts.3 In his discussion, Detges indicates that a number of general narrative principles play an important role in Catalan, but also in other Romance languages, such as the desire to use inchoative verbs in order to mark an event as more dynamic and spectacular and thus to increase listeners’ involvement. In a later stage



TSL[v.20020404] Prn:13/04/2004; 15:31



F: TSL5901.tex / p.8 (8)

Olga Fischer, Muriel Norde, and Harry Perridon

this discourse marker may come to be used to mark the ‘turning point’ in narrative sequences, and indeed any major (sub)episode boundary. At this stage the marker still occurs with a restricted number of verbs. When its use becomes more frequent, it tends to lead to ‘overuse’; the marker being used with more and more verbs, so that it becomes conventionalized. This last stage was reached only in Catalan, and not in the other Romance languages. Detges suggests that this may well be due to the absence of a written standard in Catalan (between the 16th and 19th centuries) so that a more specifically oral device (the marker was characteristic in story-telling) was able to grammaticalize before the standard was developed. Detges also shows how the same device – i.e. the use of an incipient marker to enhance the narrative event – could lead to lexicalization in the case of the inchoative suffix -sc- in the Romance languages. In the case of -sc- the device became eventually restricted to a small number of verbs, while in the case of Catalan anar ‘go’ + infinitive, it became applied to any kind of event.

. Standardization The suggestion that the development of a standard (written) language may play a role in grammaticalization cases, is an interesting one, which needs further investigation. Jim Miller’s contribution on the grammaticalization of the English perfect also looks at the influence of the standard language but he is especially interested in this aspect from a methodological point of view. He shows in his analysis of perfect and perfect-resultative constructions in spoken and non-standard English that the grammaticalization path described for the perfect that appears in written standard English is very much an idealization. The standard written perfect looks perfectly fixed, protected as it is by grammars of standard English and editorial practices. The situation is very different in spoken English and non-standard forms of English, where the perfect still competes with the past tense and with the resultative construction out of which it arose (i.e. I have my work done). Even more interesting are some developments of the perfect with adverbs, which in fact show a kind of anti-grammaticalization in that these constructions are again becoming more specific rather than more general (the latter would be usual on the grammaticalization cline). Miller’s article indeed calls into question the status of standard English as a yardstick against which to measure grammaticalization processes. Miller’s analysis is different from that of linguists like Tagliamonte (2000) and Winford (1994), who see the variety of perfect types present in non-standard varieties of English as a range of variations available before the perfect grammaticalizes, and indeed see these variants as stages towards the ‘idealized’ perfect of the standard. Miller, on the other hand, treats the perfect-variants in non-standard varieties not as part of

TSL[v.20020404] Prn:13/04/2004; 15:31

F: TSL5901.tex / p.9 (9)

Introduction

a ‘perfect’-pathway or cline but as independent structures, which may each go their own way. In other words, it is not the ideal pathway that gives shape to the development of these variants, but the system of the non-standard or spoken languages themselves, in which these variants are used. Lea Laitinen in her contribution also emphasizes the importance of language external factors such as standardization and codification. Such factors are particularly relevant in Finnish historical linguistics, since Finnish has a relatively short written history and its standard form was not created until the late 19th century. Laitinen hypothesizes that in the process of standardization the more context-bound, indexical functions tend to be rejected. The changes discussed are the grammaticalization of modal necessity verbs (NEC constructions), the emergence of discourse particles from logophoric pronouns and the development of the non-agreeing negation word (NEG). In the case of NEC-constructions, the more grammaticalized variant is avoided in standard Finnish. The change from logophoric pronoun to discourse particle, on the other hand, had been overlooked by normative grammarians and hence the more grammaticalized variant has not been excluded from the written standard. The negation marker is discussed in most detail. Laitinen shows that the development of a non-agreeing negation particle was interrupted in the written language (but not in certain contemporary dialects) because the less grammaticalized, agreeing negative auxiliary was felt to be more genuinely Finnish. Ilona Herlin and Lari Kotilainen stress, as do Miller and Laitinen, the significance of language-external factors. Their main aim is to show that seemingly straightforward examples of grammaticalization may be far more complex than grammaticalization textbooks with their emphasis on universal grammaticalization paths would lead us to believe. One of these paths is the development from temporal to causal meaning, of which the Finnish subordinating conjunction kun ‘as, when, while’ is said to be a typical example (discussed e.g. in Traugott & König 1991; Heine & Kuteva 2002). Herlin and Kotilainen outline the historical development of this conjunction and the converb essa. They argue that changes in grammatical meaning are not necessarily the result of universal cognitive processes and pragmatic principles. In the course of standardization, kun was assigned temporal and causal functions while its one-time comparative and similaritive functions were transferred to kuin, originally a variant form of kun. The authors believe that this artificial distinction was made because Indo-European languages usually have different conjunctions for these functions as well. Spoken Finnish still has kun in all four functions. In addition, only written Finnish may use kun in contrastive constructions, most probably under the influence of Indo-European cognates (English while, Swedish medan) as well. A further point they note is that the converb essa developed causal, conditional and contrastive uses in written standard Finnish only. The history of kun and essa thus shows that synchronic polysemy does not



TSL[v.20020404] Prn:13/04/2004; 15:31



F: TSL5901.tex / p.10 (10)

Olga Fischer, Muriel Norde, and Harry Perridon

necessarily reflect language-internal diachronic developments, but may be due to language-external factors such as standardization and language contact.4

. Reanalysis, polysemy and homonymy An important question in grammaticalization studies has always been whether grammaticalization leads to polysemy (one and the same item has different functions) or to homophony (different items share the same form). In the latter case a lexical split must be assumed to have taken place that separated the evolving grammatical item from its lexical mother (i.e. what Hopper 1991 terms ‘divergence’). In most cases the grammaticalizing item will then belong to another word class than the lexical item it derives from. In English, for example, the indefinite article and the numeral one go back to one and the same word, viz. OE a¯ n. In unaccented position the vowel was weakened, the nasal was lost before a consonant: a¯ n > a(n), and the meaning was ‘bleached’ to that of an indefinite article. The original item a¯ n split in this way into what is commonly assumed to be two different items, viz. a numeral (which it always had been) and an indefinite article (a new item in the lexicon). Others would, however, argue that the article and the numeral are still one and the same word, with different forms in different (stress) positions. In that case the meanings of the polysemous item would be distributed in a complementary way: article when unstressed, numeral when stressed. In her paper on demonstratives in copula-like constructions in the NorthAmerican language Passamaquoddy, Eve Ng assumes that a lexical split has taken place. The language possesses a large number of demonstratives, which are marked for three distances, two numbers, animacy, obviation (only animate forms), and absentativity. In verbless constructions with nominal predicates, distal demonstrative forms seem to function as copulas, which raises the question whether such forms should be analysed as copulas or as occurrences of the distal demonstrative. A careful analysis of the agreement phenomena in the various verbless constructions in which the demonstrative form seems to function as a copula, shows according to Ng that it is almost impossible to argue that the forms used in the verbless constructions belong to the same word class as the demonstratives they derive from. In her paper on the uses and meanings of the Kambera word wàngu Marian Klamer defends a more radical solution to the problem of homonymy or polysemy. Her view on the matter is that ‘bleaching’ not only affects the semantics of a given lexical item, but also its word class membership. The outcome of a grammaticalization process may hence be an underspecified, category neutral lexical item. The different interpretations of the word, then, are induced by the context in which it

TSL[v.20020404] Prn:13/04/2004; 15:31

F: TSL5901.tex / p.11 (11)

Introduction

occurs. In this way the unity between form and meaning is restored. One of the main arguments Klamer adduces for her theory is that in a large number of cases it is not evident (or rather: it does not matter) in exactly what function wàngu is used. In the less ambiguous cases it is the context alone that determines how the word is to be interpreted (as a main verb, meaning ‘to use’, as a preposition (‘with’), as part of a compound verb, or as a matrix verb). Whether or not the term ‘multifunctional item’ as used by Richard Wong (and also by Umberto Ansaldo and Lisa Lim, see below) is synonymous with what Klamer calls an underspecified item is not entirely clear, but at least Wong’s paper on the acquisition of the various uses of bei2 by Cantonese children seems to suggest that this indeed might be the case. Wong addresses in this paper the question whether the various stages in the diachronic development of this morpheme are recapitulated in language acquisition. The lexical meaning of the word bei2 is ‘to give’, but it is also used as a dative marker (‘to’, recipient), as a matrix verb in permissive constructions (‘to let/allow’) and in passive constructions. The first three uses of bei2 share according to Wong the same core meaning ‘transfer’, whereas the passive is related to the permissive by way of a shared syntax. The lexical use of bei2 is also basic in the sense that it does not require (or even permit) another verb to be present in the construction, whereas the three derived uses all require a second verb. In order to investigate the hypothesis that there are at least three different stages in the acquisition of bei2 (main verb (‘transfer’) > dative, permissive > passive) Wong looked at occurrences of bei2 in (a modified version of) CANCORP, consisting of the transcripts of recordings from a longitudinal study of eight Cantonese-speaking children, aged between 22 and 42 months. The findings corroborate the hypothesis that syntactically simpler constructions (bei2 as a main verb) are acquired first, followed by those constructions that are semantically closest to the original meaning. Passives are rare in Cantonese child language, and appear relatively late. It seems thus safe to conclude that the order of acquisition of the various uses of a lexical item that has been grammaticalized mirrors the diachronic development of the item in question, at least in those cases where the lexical item can still be used in its original meaning.

. Form Another question that is still in need of a definite answer is what constitutes a change in form. How, for instance, is phonetic reduction, which is often one of the main indicators that a grammaticalization process has occurred, to be asserted in isolating tonal languages, which do not allow the kind of reduced syllables that are found elsewhere? Umberto Ansaldo and Lisa Lim claim that phonetic reduction



TSL[v.20020404] Prn:13/04/2004; 15:31



F: TSL5901.tex / p.12 (12)

Olga Fischer, Muriel Norde, and Harry Perridon

in fact does take place in these languages, but is to be found in the suprasegmental features. In order to prove this hypothesis they investigated the intonational behaviour of some multifunctional items in two Sinitic languages, Cantonese (C) and Hokkien (H), in both non-grammaticalized and grammaticalized contexts, viz.: (C) gwo33 ‘to pass/cross’ > comparative marker ‘-er than’; (C) dou33 ‘to arrive’ > resultative marker; (H) khi21 ‘to go’ > aspectual (perfective) marker; and (H) ho33>21 ‘to give’, which like the words for ‘to give’ in many other Sinitic languages can be used as a causative marker, as a permissive marker and as a passive. When used in their original lexical sense, the morphemes had in all cases a significantly longer duration than when used as function words. There were, moreover, clear differences in vowel quality between the lexical and (more or less) grammatical uses of the items in question. Finally, the grammaticalized items tended to have lower fundamental frequency and intensity. Ansaldo and Lim conclude from these results that semantic bleaching is accompanied by phonetic reduction in isolating tonal languages. The phonetic reductions are probably due to the fact that when grammaticalized, the morpheme tends to be in a prosodically weak position, most often between two content morphemes. Another question pertaining to the reduction of form is posed by Sergey Say who wonders whether the fixing of word order constitutes a case of grammaticalization. He makes a distinction between word order shift (such as SOV to SVO) and word order fixation in that the former cannot be considered a case of grammaticalization since there is no reduction in variety (as is typical of grammaticalization processes) nor is there an increase in the abstractness of the rule. The case he looks at is the fixing of the genitive-noun word order within the NP in modern Lithuanian. He investigates the history of this construction and finds evidence that there used to be more variety in the position of the genitive in older Lithuanian texts. However, some of the evidence is obscured by the fact that many Old Lithuanian texts are rather literal translations from Polish. Say also suggests that there does not seem to be a semantic difference between pre- and postposed genitives in these texts. In other words, if this case represents an instance of grammaticalization, it did not lead to a loss of semantic meaning (this in contrast to Fischer (2000), who, when investigating a similar case of grammaticalization involving fixing of word order, found that the fixing of adjective-noun order within the NP in the history of English involved a loss of meaning and hence presented some sort of semantic bleaching). Say explains the direction of the change from pre- and postposition to preposition only (with one, clearly defined exception) to the fact that the Lithuanian genitive covers a wide range of functions (wider than in most languages) and that the genitive may have become preposed since one of the important functions of the genitive in Lithuanian is adjectival, and adjectives were generally preposed in Lithuanian. A harmonic word order tendency may also have been at issue but Say

TSL[v.20020404] Prn:13/04/2004; 15:31

F: TSL5901.tex / p.13 (13)

Introduction

stresses that it is the synchronic factors rather than general diachronic processes that play an important role here.

. Outlook The seventeen papers in this book show that there is not yet a body of knowledge that might be a called a ‘theory’ of grammaticalization. The initial optimism created by the renewed interest in grammaticalization in the eighties and nineties of the last century has somewhat vanished, and it has become clear that the laws that one liked to think were discovered in processes of language change are not laws, but tendencies at best. Not all changes go in the same direction, there are exceptions to the assumed unidirectionality of change in grammar, which remain exceptions even if one rejects the majority of the suggested counterexamples as mere ‘lexifications’. Not all changes have to originate in cognition, formal changes do not need to follow changes in conception, in some cases it might even be the other way round, as perhaps is the case with the indefinite article in English.5 Borrowing, standardization, competition between various forms and constructions of different dialects and/or styles, all these factors may play a role in the history of grammatical forms and constructions in a given language. When confronted with a ‘grammatical’ and a ‘lexical’ use of a given form or construction, one cannot be absolutely certain that the grammatical meaning derives from the lexical meaning. It is still a probability, and as such any theory of grammatical change would have to account for this. Similarly, the circumstance that the mechanisms of grammaticalization (e.g. phonetic reduction, semantic bleaching, category change etc.) so often occur together is something that a theory of grammatical change needs to explain.6 We believe that certainty about or a better understanding of the nature of grammaticalization can only be obtained by a careful analysis in each particular instance of the history of the forms or constructions in question. Grammaticalization studies thus find themselves in the same position as other branches of historical linguistics: tendencies may be discovered but each word and each construction also has its own history.

Notes . It seems to us that, if unidirectionality is redefined in this purely semantic way, the relexification of the Mandarin modal dé/dei, or indeed any semantic change, does not then constitute a threat to the unidirectionality hypothesis.



TSL[v.20020404] Prn:13/04/2004; 15:31



F: TSL5901.tex / p.14 (14)

Olga Fischer, Muriel Norde, and Harry Perridon . Indeed the grammaticalization Lexicon provided by Heine and Kuteva (2002) is based on this idea, i.e. that there are regular paths between source concepts (the initial lexical stage) and targets (grammaticalized structures), and that the range of both is restricted. . Note, however, that Kuteva (2001) emphasizes both the general cognitive forces as well as the discourse-pragmatic ones. . Note that this has serious implications for diachronic claims based on polysemies in languages without a written history, such as most (all?) African languages, on which the work of Heine and others crucially rests. . It has also been suggested in Fischer (1994) that in the grammaticalization of the English have + to infinitive, the formal changes forced the semantic changes and not the other way around. . Plank’s (1995) study on the properties of grammaticalization and degrammaticalization is very interesting in this respect because he shows that grammaticalization involves “ein schrittweise[s] Ablauf ” (p. 200, ‘a stepwise decline’), i.e. a gradual, orderly loss of thematically linked properties along a cline of properties which are part of the construction in question. In cases of degrammaticalization, this cline of properties has been disturbed by other (synchronic) factors (‘Systemstörung’), with the result that it is not the properties themselves that get re-ordered (they cannot be because the cline itself has been disturbed) but it is the formal expression that gets re-ordered or changed in such a way that it fits the left-over, disturbed properties. What his study clearly implies is that there are implicational relations between the various mechanisms involved in grammaticalization but not in degrammaticalization. Building on this important distinction between grammaticalization and degrammaticalization, Norde (2003) concludes that, unlike grammaticalization, degrammaticalization is not an autonomous process but the result of other changes, usually of the kind that Plank had identified as ‘Systemstörung’. Crucially, degrammaticalization always involves a single shift from right to left on the cline of grammaticality, in other words, there are no known cases in which a grammatical item gradually moves all the way up the cline, passing though the same intermediate stages a grammaticalizing item passes through, in the reverse order.

References Allen, Cynthia (1997). “The Origins of the ‘Group Genitive’ in English”. Transactions of the Philological Society, 95, 111–131. Beths, Frank (1999). “The History of dare and the Status of Unidirectionality”. Linguistics, 37, 1069–1110. Bybee, Joan L., William Pagliuca, & Revere D. Perkins (1994). The Evolution of Grammar: Tense, Aspect and Modality in the Languages of the World. Chicago: University of Chicago Press. Campbell, Lyle (Ed.). (2001). Grammaticalization. A Critical Assessment (Special issue of Language Sciences, 23, 2–3).

TSL[v.20020404] Prn:13/04/2004; 15:31

F: TSL5901.tex / p.15 (15)

Introduction

Fischer, Olga (1994). “The Development of Quasi-Auxiliaries in English and Changes in Word Order”. Neophilologus, 78, 137–164. Fischer, Olga (2000). “The Position of the Adjective in Old English”. In Ricardo BermúdezOtero, David Denison, Richard M. Hogg, & C. B. McCully (Eds.), Generative Theory and Corpus Studies. A Dialogue from 10 ICEHL (pp. 153–181). Berlin: Mouton de Gruyter. Fischer, Olga, Anette Rosenbach, & Dieter Stein (Eds.). (2000). Pathways of Change. Grammaticalization in English. Amsterdam: John Benjamins. Giacalone Ramat, Anna & Paul J. Hopper (Eds.). (1998). The Limits of Grammaticalization. Amsterdam: John Benjamins. Haspelmath, Martin (1999). “Why is Grammaticalization Irreversible?” Linguistics, 37, 1043–1068. Heine, Bernd, Ulrike Claudi, & Friederike Hünnemeyer (1991). Grammaticalization. A Conceptual Framework. Chicago: University of Chicago press. Heine, Bernd & Tania Kuteva (2002). World Lexicon of Grammaticalization. Cambridge: Cambridge University Press. Hopper, Paul J. (1991). “On Some Principles of Grammaticization”. In Traugott & Heine (Eds.), Vol. I (pp. 17–35). Hopper, Paul J. & Elizabeth Closs Traugott (2003[1993]). Grammaticalization. Cambridge: Cambridge University Press. Kuteva, Tania (2001). Auxiliation: An Enquiry into the Nature of Grammaticalization. Oxford: Oxford University Press. Lass, Roger (2000). “Remarks on (Uni)Directionality”. In Fischer et al. (Eds.), 207–227. Lehmann, Christian (1995[1982]). Thoughts on Grammaticalization. München & Newcastle: Lincom Europa. Lyons, John (1977). Semantics. Volume 2. Cambridge: Cambridge University Press. Newmeyer, Frederick J. (1998). Language Form and Language Function. Cambridge, MA: MIT Press. Norde, Muriel (2001). “Deflexion as a Counterdirectional Factor in Grammatical Change”. Language Sciences, 23, 231–64. Norde, Muriel (2003). “Degrammaticalization – Process or Result?” Paper presented at the XVIth International Conference on Historical Linguistics, University of Copenhagen, August 11–15 2003. Plank, Frans (1995). “Entgrammatisierung – Spiegelbild der Grammatisierung?” In Norbert Boretzky, Wolfgang Dressler, J. Orešnik, K. Teržan, & Wolfgang U. Wurzel (Eds.), Natürlichkeitstheorie und Sprachwandel (Essener Beiträge zur Sprachwandelforschung.) (pp. 199–219). Bochum: Brochmeyer. Sweetser, Eve E. (1990). From Etymology to Pragmatics. Metaphorical and Cultural Aspects of Semantic Structure. Cambridge: Cambridge University Press. Tagliamonte, Sali (2000). “The Grammaticalization of the Present Perfect in English. Tracks of Change and Continuity in a Linguistic Enclave”. In Fischer et al. (Eds.), 329–354. Traugott, Elizabeth C. & Richard B. Dasher (2002). Regularity in Semantic Change. Cambridge: Cambridge University Press. Traugott, Elizabeth C. & Bernd Heine (Eds.). (1991). Approaches to Grammaticalization. Vols. I, II. Amsterdam: John Benjamins.



TSL[v.20020404] Prn:13/04/2004; 15:31



F: TSL5901.tex / p.16 (16)

Olga Fischer, Muriel Norde, and Harry Perridon

Traugott, Elizabeth C. & König, Ekkehard (1991). “The Semantics-pragmatics of Grammaticalization Revisited”. In Traugott & Heine (Eds.), Vol. I (pp. 189–218). van der Auwera, Johan & Vladimir A. Plungian (1998). “Modality’s Semantic Map”. Linguistic Typology, 2, 79–124. Winford, Donald (1994). “Variability in the Use of Perfect have in Trinidadian English: A Problem of Categorial and Semantic Mismatch”. Language Variation and Change, 5, 141–188. Wischer, Ilse & Gabriela Diewald (Eds.). (2002). New Reflections on Grammaticalization. Amsterdam: John Benjamins.

TSL[v.20020404] Prn:13/04/2004; 15:41

F: TSL5902.tex / p.1 (17)

On directionality in language change with particular reference to grammaticalization* Martin Haspelmath Max-Planck-Institut für evolutionäre Anthropologie, Leipzig

.

Introduction

The issue of the directionality of grammaticalization has attracted considerable attention in recent years, as illustrated by works such as as Plank (1995), Frajzyngier (1997), Newmeyer (1998: Ch. 5), Haspelmath (1999a) (and the ensuing debate: Geurts 2000a, b; Haspelmath 2000b), the papers in the recent special issue of Language Sciences (Campbell (Ed.) 2001), Traugott (2001), Kim (2001), van der Auwera (2002), Heine (2003), and quite a few others. In this paper, I would like to put this issue in a somewhat broader perspective, discussing not only the unidirectionality of grammaticalization, but also unidirectionality in other areas of language change. But the main focus will be on grammaticalization, and after defending the unidirectionality claim for grammaticalization against several criticisms, I will examine a substantial number of alleged exceptions to the unidirectionality and show that only very few of them can be accepted as real exceptions. The paper is divided into three main sections, in which I will make the following larger points: i.

If we want to understand language change, we need to identify universals of language change. Directionality constraints are among the strongest universals of language change (Section 2). ii The unidirectionality of grammaticalization is the most important constraint on morphosyntactic change (Section 3). iii. Most cases of “degrammaticalization” that are cited in the literature do not show the reversal of grammaticalization (or ‘antigrammaticalization’), but something else (Section 4).

TSL[v.20020404] Prn:13/04/2004; 15:41



F: TSL5902.tex / p.2 (18)

Martin Haspelmath

. Universals of change and directionality constraints Most of the time, historical linguists are occupied with the business of describing language change, which is quite a challenging task in itself, given that change is so difficult to observe. But ultimately we would also like to understand language change to the extent possible, or in other words, we want to answer why-questions. In particular, we want to know why language structure changes in the way it does. There are other why-questions about language change, such as the question why languages change at all, the question why the social propagation of an initial innovation can often be described by an S-curve, and so on. These will not be addressed here. Linguists working on particular languages are also often interested in particular why-questions such as the question why the Romance languages lost the Latin case inflections. But unfortunately, particular why-questions of this kind are for most practical purposes unanswerable. The number of factors affecting language change is so enormous and we can control only so few of them that most change events must appear to us as historical accidents. Latin could have kept its cases, even with all the phonological erosion that made them difficult to distinguish, simply by applying morphological changes serving to preserve the case contrasts. Or Latin speakers could even have developed more cases the way Hungarian and Finnish speakers did. It so happened that Latin lost its cases, and trying to understand this unique historical event typically leads to frustration. In general, understanding requires that we identify non-accidental phenomena, and for understanding language change, this means that we have to find universals of language change. To illustrate what I mean by this, a few random examples of proposed universals of language change (of different degrees of generality) are given in (1). (1) a.

Survival of the Frequent (‘Unmarked’) (e.g. Winter 1971; Wurzel 1994) When a grammatical distinction is given up, it is the more frequent category that survives. (E.g. plural forms survive when dual/plural distinction is lost.) b. Sound Alternations Result from Sound Change (phonetics > phonology; *morphology > phonology) c. From Space to Time (e.g. Haspelmath 1997b) (spatial > temporal marker; *temporal > spatial marker) d. From ‘Something’ to ‘Nothing’ (Haspelmath 1997a: 230) indefinite pronouns ‘something’ > ‘nothing’ (*‘nothing’ > ‘something’) e. From Esses to Aitches: s > h (*h > s) (Ferguson 1990)

TSL[v.20020404] Prn:13/04/2004; 15:41

F: TSL5902.tex / p.3 (19)

On directionality in language change

These are all general laws which we can potentially explain,1 and when we have such an explanation, we can apply it to individual instances of these universals. For example, we might want to say that the universal ‘Survival of the Frequent’ is explained with reference to the cognitive notion of frequency-induced entrenchment (cf. Bybee 1985: 119): A frequent linguistic unit is remembered better because frequency of exposure leads to greater memory strength. When a distinction is given up, only the most entrenched category survives. Now let us take an individual instance of the Survival of the Frequent, say, the fact that when the Classical Greek dual/plural distinction was given up, only the plural forms survived. The plural was more frequent than the dual (cf. Greenberg 1966: 31–37), so this change is in line with the universal, and if we want to know why the plural rather than the dual survived in Greek, we can appeal to the explanation that we just gave. So in this sense we can say that a particular change was explained after all; but of course the explanation of the particular change has nothing particular about it. We cannot explain why this changed happened in Greek but not, say, in Slovene (where the old dual survived), and we cannot explain why it happened two and a half millennia ago rather than a thousand years later or a thousand years earlier. So wherever we can understand structural change, it is really universals of structural change that we understand. But unless we know whether a given instance of change is part of a larger trend, we do not know whether there is anything to explain. Now when we look at reasonably robust universals of language change, we see that many of them take the form of directionality constraints. Of the five examples in (1) four have the form “X can change into Y, but Y cannot change into X”. Especially in phonology, it is easy to find cases of this type, and I list a few more in (2). (2) a. b. c. d. e. f.

[k] > [tw] (*[tw] > [k]) [p] > [f] (*[f] > [p]) [u] > [y] (*[y] > [u]) [z] > [r] (*[r] > [z]) [ts] > [s] (*[s] > [ts]) [l] > [w] (*[w] > [l])

So quite a few sound changes appear to be unidirectional, but there are of course also bidirectional sound changes, such as those in (3). Some of these changes are more likely in some positions than in others, and maybe a more fine-grained description of the type of change would reveal a directionality tendency in some of these cases as well. (3) a. [t] > [θ] and [θ] > [t] b. [o] > [a] and [a] > [o] c. [i] > [6] and [6] > [i]



TSL[v.20020404] Prn:13/04/2004; 15:41



F: TSL5902.tex / p.4 (20)

Martin Haspelmath

d. [au] > [o] and [o] > [au] e. [b] > [v] and [v] > [b] Thus, it is an empirical question whether a type of sound change is unidirectional or not. I am not aware of any extensive discussion of this issue in the theoretical literature on phonological change, but as Ferguson (1990) observes, every linguist with some experience in diachronic phonology has the intuition that there are often directionality constraints at work: One of the most powerful tools in the armamentarium of linguists engaged in the study of diachronic phonology is the often implicit notion that some changes are phonetically more likely than others. Thus if a linguist finds a systematic correspondence between [g] and [dŠ] in two related language varieties, it will be reasonable to assume that the stop is the older variant and the affricate the younger one until strong counter evidence is found. The linguist makes such an assumption because experience with many languages has shown that the change of [g] to [dŠ] is fairly common and tends to occur under certain well-documented conditions whereas the reverse change is unusual and problematic. (Ferguson 1990: 59–60)

Ferguson goes on to observe that this powerful tool of directionality constraints is not generally covered in textbooks or handbooks of phonology or historical linguistics. These typically include taxonomies of attested sound changes and introduce technical terms like lenition, assimilation, syncope and epenthesis, but they usually do not say what an impossible change is, or which changes are ubiquitous and which ones are exceedingly rare. For synchronic universals in phoneme systems, we have Maddieson’s (1984) handbook with inventories of 317 languages. Diachronic phonology, whether theoretically oriented or primarily interested in reconstructing particular protolanguages, would profit enormously from having a handbook of attested sound changes in the world’s languages. Such a handbook would make it possible to identify constraints on possible sound changes, and many of the most interesting constraints will no doubt be directionality constraints. After all, that [u] presumably never changes to [a] in one step, or that [l] never changes to [b], is not surprising, whereas the unidirectionality of the [u] > [y] change and the [l] > [w] change is much harder to explain. There are also some clear tendencies of lexical semantic change (e.g. ‘cup’ can change to ‘head’ and ‘head’ can change to ‘chief ’, but the opposite changes are extremely unlikely). Once we are confident that we have a universal directionality constraint in some domain, the question arises as to how it should be explained. If the source structure and the target structure are similar enough so that one can change into the other gradually and often imperceptibly, why can’t they change in either direction? The historical-linguistic literature is full of proposals accounting for specific

TSL[v.20020404] Prn:13/04/2004; 15:41

F: TSL5902.tex / p.5 (21)

On directionality in language change

cases, appealing to a variety of structural and external factors. For phonology, some authors have proposed that sound change is by and large reductive in nature (Mowrey & Pagliuca 1995; Bybee 2001), and that the unidirectionality of changes like [p] > [f] and [ts] > [s] can be accounted for in this way. The explanation for unidirectionality is also beginning to be addressed by researchers working in the area of grammaticalization (e.g. Lehmann 1993; Haspelmath 1999a), and this discussion could profit from analogous discussions in the other subfields of linguistics. In this paper, I will not say anything about the correct explanation of unidirectionality in grammaticalization, because at present I have nothing to add to my earlier proposals. I will, however, address a number of criticisms and counterexamples that can be found in the recent literature.

. Unidirectionality of grammaticalization . How important is unidirectionality? Although it is very difficult to quantify language change, it seems to me that it is undeniable that the unidirectionality of grammaticalization is by far the most important constraint on morphosyntactic change, simply because grammaticalization changes are so ubiquitous. As far as I can see, the only serious competitor of unidirectionality is the diachronic universal ‘Survival of the Frequent’ (see (1a)). This universal seems to hold not only when categorial distinctions break down, but also in analogical leveling in inflectional morphology. For example, when a stem alternation such as dream/dreamt is leveled, it is the more frequent present-tense stem that survives (so that we get dream/dreamed, not *drem/dremmed). This is a fairly important universal for morphological change, but it seems to be much less important for syntactic change. Grammaticalization, by contrast, is of paramount importance both for syntactic change and for morphological change. A rough estimate is that two thirds of the papers on diachronic syntax published in recent volumes such as van Kemenade and Vincent (1997) and Pintzuk et al. (2000) deal with grammaticalization changes (even if they rarely mention the term ‘grammaticalization’). The relatively high number of non-grammaticalization papers in these volumes has to do with the fact that word order change is so salient in some European languages, especially of course word-order change having to do with verb-second phenomena. But we know that verb-second word order and the changes related to it are highly unusual phenomena that are hardly found outside of Europe. My guess is that if we were able to study syntactic change on all continents, grammaticalization would play an even greater role in diachronic syntax. Of course, this is not more than an impres-



TSL[v.20020404] Prn:13/04/2004; 15:41



F: TSL5902.tex / p.6 (22)

Martin Haspelmath

sionistic statement, but I challenge anyone to come up with a long list of interesting syntactic changes that are unrelated to grammaticalization. Like unidirectionality in sound change, unidirectionality in grammaticalization is very important in practical terms for the historical-comparative linguist. Suppose we have two related languages with no historical documentation, and one of them has a future-tense affix that looks similar to a future-tense auxiliary of the other language. If both directions of change were equally likely, we would not know what to reconstruct for the ancestor language. But because grammaticalization is overwhelmingly irreversible, the historical linguist can safely reconstruct the future auxiliary for the protolanguage in this case. Moreover, unidirectionality helps us assess the likelihood of competing etymologies even if older stages are attested. For instance, historical linguists of Indo-Aryan have long debated the etymology of the Hindi/Urdu ergative-case clitic =ne. Quite a few linguists in the 20th century traced this element back to Sanskrit -ina, an instrumental suffix that would be a very plausible source from a semantic point of view. In a recent contribution to this debate, Butt (2001: 114) has pointed out that such a change would constitute a counterexample to unidirectionality and is hence very unlikely (one would have to postulate phonological expansion from [na] to [ne:] and a change from affix to clitic). This, among other reasons, leads Butt to reject this etymology and look for some other possible source of =ne in a full lexical item.2 Now despite the theoretical importance of grammaticalization studies for understanding language change and their practical importance for historical linguistics, there have been a number of critical voices in recent years. In the remainder of this section I would like to address some of these points of criticism and show that while some are well-taken, others are quite unfounded.

. Partially valid criticism Three points made by grammaticalization critics that I regard as partially justified are listed in (4). (4) a. Unidirectionality is not exceptionless b. ‘Grammaticalization theory’ is not one theory c. ‘Pathways of morphemes’ must be linked to speakers’ actions Unidirectionality was apparently first stated explicitly as an important universal property of grammaticalization in Lehmann (1995a [1982]: 16–19). Lehmann coined the term degrammaticalization for a phenomenon that he believed did not exist, the reverse of grammaticalization. But now the phenomenon had a name, and it seems that Lehmann’s strong initial claim plus his nice neologism spurred

TSL[v.20020404] Prn:13/04/2004; 15:41

F: TSL5902.tex / p.7 (23)

On directionality in language change

linguists to look for actual examples of degrammaticalization. And indeed, a number of good exceptions were found (see Newmeyer 1998; Campbell 2001; Janda 2001; Norde 2001), and this has generally been acknowledged by grammaticalization researchers. The counterexamples did not pose a serious threat to the original generalization, but a presumed absolute universal had to be weakened to a statistical universal. What are the consequences of this for the theory of grammaticalization? One might say that now that we know that unidirectionality has exceptions, it has become somewhat less interesting. This is true insofar as stronger generalizations are more surprising than weaker generalizations, but the reverse is also true: Because unidirectionality is so interesting, we know about the exceptions. If someone proposes an uninteresting universal, we may never discover the exceptions because nobody bothers to look for them. The basic generalization of unidirectionality stands unchallenged as long as nobody shows that degrammaticalization is as common as grammaticalization. If one is interested in generalizations rather than arbitrary facts, one must put aside the exceptions, because unless they can be subsumed under some further generalization, they cannot be explained. Harris and Campbell (1995: 338) say in this context: “An adequate theory must account for infrequent phenomena, not merely for the most common patterns.” This is of course right if by ‘theory’ they mean ‘descriptive framework’: We need terminology also for rare phenomena. But if by ‘theory’ we mean understanding and explanation, this is not right, because exceptions cannot be understood by definition; they are the residue that resists explanation.3 Thus, although it is true that unidirectionality is not exceptionless, this does not make it any less intriguing and important for our understanding of language change. The second point of criticism that I find partially justified is Newmeyer’s (1998) claim that what linguists commonly call ‘grammaticalization theory’ is not a theory in the sense of a well-defined system of interconnected falsifiable hypotheses. What unites researchers in the area of grammaticalization is not that they subscribe to a single monolithic theory, but that they see a large class of semantic and morphosyntactic changes as sharing similarities and as theoretically interesting. There are a fair amount of quite different theoretical ideas and hypotheses concerning grammaticalization changes, and some of them are probably not compatible. Thus, ‘grammaticalization theory’ is more like ‘evolutionary theory’, which is not one single monolithic system either, but describes a range of related approaches and basic issues in the area of historical biology. It would perhaps be more accurate to say ‘theorizing about grammaticalization’ (instead of ‘grammaticalization theory’), and to some extent the use of the term ‘theory’ may be motivated by its prestige. The prestigious term ‘theory’ has experienced a rather inflationary development in recent decades in linguistics. Like grammaticalization, inflationary processes are generally irreversible (cf. Dahl 2002), so it seems unlikely that



TSL[v.20020404] Prn:13/04/2004; 15:41



F: TSL5902.tex / p.8 (24)

Martin Haspelmath

the term ‘grammaticalization theory’ will be abandoned. But if we want to be clear about what we are doing, the term ‘theory’ is not particularly useful. I prefer to talk about the goal of understanding, or explaining, or answering why-questions. These are terms from our everyday language that everyone understands, and our endeavors can be accurately characterized with them. The third point of criticism is Janda’s (2001) reminder that it is impossible to understand language change phenomena if we see them as divorced from the speakers.4 If we talk about a morpheme traveling along a pathway, we should be aware that this is a very abstract metaphor that may invite all kinds of unwarranted inferences. We need to be careful with metaphors, and we should make more efforts to go down to the micro-level of individual speakers and derive the observed constraints on structural changes from known constraints on speakers’ linguistic behavior. But at the same time it is clear that we cannot do linguistics without abstract metaphors, and so far at least concepts like ‘grammaticalization path’ have done far more good than damage. We would know far less about possible and impossible changes if we had not started drawing diagrams of grammaticalization paths and semantic maps. Recently, some linguists have stated precise rules for interpreting a diagram showing a semantic map with grammaticalization paths (see, e.g., van der Auwera & Plungian 1998; Croft 2001; Haspelmath 2003). Consider, for example, the semantic map in Figure 1. This figure embodies the following two claims: i.

Synchronic: Polysemous forms cover adjacent nodes (i.e. nodes linked by a line or arrow); ii. Diachronic: A linguistic form may extend its range of functions on the map in any direction, but not against the direction of an arrow. Thus, it is predicted that a direction marker such as Latin ad ‘to’ can become a recipient marker (as in French à), but it cannot then go on to become a beneficiary marker. Thus, we can see the path metaphor and the diagrams based on it as analogous to tree diagrams in syntax: These are probably not literally in people’s mental grampredicative

external

possessor

possessor

direction

recipient

beneficiary

purpose

experiencer

judicantis

Figure 1. A semantic map of typical dative functions (from Haspelmath 2003: 234)

TSL[v.20020404] Prn:13/04/2004; 15:41

F: TSL5902.tex / p.9 (25)

On directionality in language change

mars either, but if syntacticians drew no tree diagrams, we would know much less about syntax.

. Invalid criticism Three points made by grammaticalization critics that I regard as unjustified are listed in (5). (5) a.

Unidirectionality implies a fully isolating prehistoric state and thus contradicts uniformitarianism. b. Unidirectionality is built into the definition of grammaticalization and is hence not an empirical claim. c. There is nothing unique about the kinds of changes that are associated with grammaticalization. (Newmeyer 1998; Campbell 2001; Janda 2001)

The first point, about the contradiction to uniformitaranism, was recently brought up by Roger Lass (see also Hoenigswald 1991: 25): The claim that all grammatical material is ultimately lexical means that there was a time when all human languages were ‘isolating’ (in the days of Homo erectus or whatever everybody spoke Vietnamese). . . [This] is counteruniformitarian, and so methodologically inadmissible. (Lass 2000: 216)

But first of all, the claim that “all grammatical material is ultimately lexical” does not follow from unidirectionality, because it may well be that some elements such as demonstratives or interrogative pronouns are never created by grammaticalization from full lexical items, and have simply always been demonstratives or interrogative pronouns. Moreover, at least since Meillet (1912) it has generally been recognized that analogy is another important source of grammatical items, besides grammaticalization. But even if one were to make the speculative claim that all grammatical material is ultimately lexical, there would be no methodological problem because the principle of uniformitarianism does not require the assumption that early hominids such as Homo erectus (if they already had some kind of language) had languages of the same type as modern humans. If we allow ourselves speculation about the distant past, we can easily imagine that the first modern humans inherited part of their lexicon from the cruder languages of earlier hominids and added more lexical differentiation and grammatical elaboration. But since language has been around for tens of thousands of years and we know next to nothing about its origin, we really do not have to worry about the consequences of diachronic universals for prehistory.5 The second point of criticism is that unidirectionality is built into the definition of grammaticalization and hence represents a tautology (Campbell 2001: 124;



TSL[v.20020404] Prn:13/04/2004; 15:41



F: TSL5902.tex / p.10 (26)

Martin Haspelmath

Janda 2001: 294). This is very easy to counter: Yes, it is true that unidirectionality is built into the definition of grammaticalization. My current definition of grammaticalization is given in (6). (6) A grammaticalization is a diachronic change by which the parts of a constructional schema come to have stronger internal dependencies.6 This describes a unidirectional process, so saying that “grammaticalization is unidirectional” is strictly speaking tautologous. The point is, of course, that the easily imaginable reverse of this process does not occur (apart from a few exceptional instances). So this is not a substantive point at all, and one wonders why one hears it repeated so often. The third point of criticism is that there is nothing special or unique about grammaticalization changes. Campbell (2001) expresses it as follows: Grammaticalization has no independent status of its own; it merely involves other kinds of changes and mechanisms of change which are well understood and are not limited to cases involving grammaticalization: sound change, semantic change, and reanalysis. (Campbell 2001: 117)

And Janda (2001: 266) maintains that grammaticalization “is actually an epiphenomenon which results from the intersection and interaction of other, independently motivated domains” (see also Newmeyer 1998: 237ff.). Somehow these authors seem to think that grammaticalization is wrongly regarded as a primitive concept, although I know of no claim to this effect. On the contrary, studies of grammaticalization such as Lehmann (1995a [1982]), Heine and Reh (1984), Hopper and Traugott (1993) are quite explicit in listing the various low-level changes that are associated with grammaticalization, such as phonological erosion, desemanticization, reanalysis, decategorialization, and so on. The claim that these authors and other have made is that grammaticalization is a particularly interesting concept, because it is largely irreversible and because we observe strong correlations between phonological, syntactic and semanticpragmatic changes. It is a macro-level phenomenon which cannot be reduced to the properties of the corresponding micro-level phenomena. Campbell’s, Janda’s and Newmeyer’s criticism is similar to an objection against sociological studies of social classes on the grounds that social class is not a primitive concept, but an epiphenomenon which results from the interaction of human individuals. Most of the subject matters studied by linguists are epiphenomenal in the sense that they are complex higher-level phenomena involving the interaction of a multiplicity of lower-level phenomena.7 Campbell says in the above quotation that sound change, semantic change and reanalysis are “well understood”, but unless he refers to the terminology and really means “well defined”, I find this far too optimistic. Diachronic phonologists and di-

TSL[v.20020404] Prn:13/04/2004; 15:41

F: TSL5902.tex / p.11 (27)

On directionality in language change

achronic semanticists have not even begun collecting the systematic cross-linguistic data that would allow us to arrive at empirically well-founded universals of sound change and universals of lexical semantic change. Whereas for grammaticalization we now have Heine and Kuteva’s (2002) World Lexicon of Grammaticalization, we are still waiting for a World Lexicon of Sound Change and a World Lexicon of LexicalSemantic Change. In diachronic syntax, all we have is the handbook by Harris and Campbell (1995) with an exhaustive classification of syntactic changes and ample cross-linguistic exemplification, but few constraints and thus little explanation. In all these areas we are far from really understanding language change. Newmeyer (1998: 259) urges linguists not to “invite the conclusion that some dynamic is at work in grammaticalization that cannot be understood as a product of [certain independently occurring semantic and phonetic changes].” But this is exactly what is claimed by grammaticalizationists. Even if there is no single universally accepted explanatory architecture for grammaticalization yet, we have made a lot of progress in understanding the dynamic of grammaticalization. At the very least we have thorough cross-linguistic documentation, and a strong generalization, unidirectionality.

. Antigrammaticalization and “degrammaticalization” My third main point is that most cases of “degrammaticalization” that are cited in the literature do not show the reversal of grammaticalization, but something else. I will discuss a fairly large number of changes that have been mentioned in the literature, and I will classify them into various types. There is no space here to describe the changes in any detail, so I must refer the reader to the earlier literature. The purpose of this section is twofold. On the one hand, I want to show that exceptions to the unidirectionality universal are not “rampant” (as Newmeyer 1998: 263 claims), but are quite rare. Although probably around a hundred cases of degrammaticalization have been mentioned in the literature, the number of real exceptions is much lower. On the other hand, since the phenomena called “degrammaticalization” are so heterogeneous, it seems useful to identify various subclasses of “degrammaticalization”. I do not think that these cases have anything in common, so that we do not really need a term like “degrammaticalization” for them, and I only use this term in quotes.

. Antigrammaticalization: The reversal of grammaticalization One important new term that I want to introduce here is antigrammaticalization. By this I mean a change that leads from the endpoint to the starting point of a



TSL[v.20020404] Prn:13/04/2004; 15:41



F: TSL5902.tex / p.12 (28)

Martin Haspelmath

potential grammaticalization and also shows the same intermediate stages. For instance, a change from a case suffix to a free postposition with the intermediate stage of a postpositional clitic would be an antigrammaticalization. This implies that the change occurs in a construction which can be seen as preserving its identity before and after the change, as in grammaticalization, where we also have a gradual change of the properties of a construction, but we do not get a new construction. In this characterization of grammaticalization and antigrammaticalization I presuppose that grammaticalization changes modify the constructions they affect but preserve their identity. Admittedly this makes my characterization somewhat vague, because there is not always a consensus on which changes just modify a construction and which changes lead to totally new constructions. In Haspelmath (1998), I have argued at some length that grammaticalization changes are gradual, preserve the identity of the construction, and occur in ordinary language use, whereas reanalysis (and likewise analogy) is abrupt, leads to new constructions (or subsumes an existing unit under a completely different construction), and occurs in language acquisition. For example, a change whereby an erstwhile relational noun turns into a preposition (e.g. German wegen “because of ”, from the dative plural of Weg “way”) is said to preserve its identity because it consists of numerous imperceptible changes with no radical break at any single point. In particular, there is no reason to postulate an abrupt reanalysis of the noun Wegen as a preposition wegen (unless one assumes that speakers only have a small fixed set of innately specified categories at their disposal and cannot internalize a grammar with items that are intermediate between nouns and adpositions). It should be noted that my definition of antigrammaticalization is intended to cover types of changes, not tokens. Janda (2001: 295) and Norde (2001: 236) seem to interpret the term ‘reversal’ as ‘token reversal’, so that irreversibility would only mean that once a structure A has changed into a structure B, it does not change back to A. This claim, that token reversal does not occur (or is very unlikely), is of course not particularly interesting. My term antigrammaticalization is intended to cover any type of change that goes against the general direction of grammaticalization (i.e. discourse > syntax > morphology). Armed with this new term, we can now say that only antigrammaticalizations are exceptions to unidirectionality (cf. Lehmann 1995b: 1256), whereas other kinds of “degrammaticalization” are not necessarily expected to be rare or exceptional. This is not a weakening of the unidirectionality claim, because at least the way it was originally formulated (in Lehmann 1995a [1982]: 16–19), it is clear that only antigrammaticalizations were supposed to be ruled out, not any kind of change from grammar to lexicon. Janda (2001) has made a similar terminological distinction between reversibility of grammaticalization and counterability (however, I am not aware that anyone ever claimed that grammaticalization should be not only irreversible, but also ‘non-counterable’).

TSL[v.20020404] Prn:13/04/2004; 15:41

F: TSL5902.tex / p.13 (29)

On directionality in language change

Let us now look at some antigrammaticalizations in this sense. Eight cases are listed in (7). (7) attested antigrammaticalizations a. English and Mainland Scandinavian genitive suffix -s > clitic =s (Janda 1980; Norde 1997; Newmeyer 1998: 266, 256; etc.; but see Börjars 2003) b. Irish 1st person plural subject suffix -muid > independent pronoun muid (Bybee et al. 1994: 13–4; Roma 1999) c. Japanese adverbial subordinator -ga “although” > free linker ga “but” (Matsumoto 1988) d. Saame abessive suffix *-ptaken > clitic =taga > free postposition taga (Nevis 1986a) e. Estonian question marker -s > =es > free particle es (Nevis 1986b; Campbell 1991: 290–292) f. English infinitive prefix to- > proclitic to= (Fischer 2000; Fitzmaurice 2000) g. Modern Greek prefix ksana- “again” > free adverb ksana “again” (Méndez Dosuna 1997) h. Latin rigid prefix re- “again” > Italian flexible prefix ri- (e.g. ridevo fare “I must do again”) For me, these cases are real exceptions, which means that they do not fall under any other generalization, and I cannot say more about them. This does not mean that more could not be said about them in the future. For instance, Bybee et al. (1994: 13–14) say about the case of the Irish personal pronoun muid that there was “strong paradigmatic pressure” that facilitated the change. It could be that we will eventually be able to identify further factors such as ‘paradigmatic pressure’ that make antigrammaticalization possible, but until we have a solid generalization, any attempt at explaining these cases away seems premature. All other cases of “degrammaticalization” that I have found in the literature are not antigrammaticalizations, as I will now show.

. Delocutive word-formation from function words and affixes A first type of change that has been called “degrammaticalization” is delocutive word-formation from function words and affixes. A delocutive lexeme is one that was derived by some regular word-formation process from another lexeme whose use in speech somehow determines the meaning of the derived lexeme. For instance, Latin negare “deny” is said to derive from the negative marker nec, so it literally means “say not”, and French tutoyer is derived from the pronoun form tu



TSL[v.20020404] Prn:13/04/2004; 15:41



F: TSL5902.tex / p.14 (30)

Martin Haspelmath

and means “use tu as address form”. An example of a delocutive noun would be a hello in English (meaning “an act of saying hello”), as in I heard many hellos and few good-byes, or the noun yes in I never know whether her yes is really a no. Now my claim is that expressions such as ifs and buts are delocutive nouns of the same type, and iffy is a delocutive adjective. A few further cases are listed in (8). (The symbol “