Studies in the History of the English Language IV: Empirical and Analytical Advances in the Study of English Language Change (Topics in English Linguistics)

Studies in the History of the English Language IV: Empirical and Analytical Advances in the Study of English Language Change

Edited by Susan M. Fitzmaurice Donka Minkova

Mouton de Gruyter

Studies in the History of the English Language IV

≥

Topics in English Linguistics 61

Editors

Elizabeth Closs Traugott Bernd Kortmann

Mouton de Gruyter Berlin · New York

Studies in the History of the English Language IV Empirical and Analytical Advances in the Study of English Language Change Edited by

Susan M. Fitzmaurice Donka Minkova

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

Library of Congress Cataloging-in-Publication Data Studies in the history of the English language IV : empirical and analytical advances in the study of English language change / edited by Susan M. Fitzmaurice , Donka Minkova. p. cm. ⫺ (Topics in English linguistics ; 61) Includes bibliographical references and index. ISBN 978-3-11-020587-9 (alk. paper) 1. English language ⫺ History. 2. English language ⫺ Grammar, Historical. I. Fitzmaurice, Susan M. II. Minkova, Donka, 1944⫺ III. Title: Studies in the history of the English language 4. IV. Title: Empirical and analytical advances in the study of English language change. PE1075.S885 2008 420.9⫺dc22 2008040565

ISBN 978-3-11-020587-9 ISSN 1434-3452 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. ” Copyright 2008 by Walter de Gruyter GmbH & Co. KG, 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. Cover design: Christopher Schneider, Berlin. Typesetting: ptp ⫺ berlin gmbH, Berlin. Printed in Germany.

Richard Hogg was the UK ambassador to the first SHEL meeting at UCLA in 2000. As the grand-marshal of the just finished 6-volume Cambridge History of the English Language, he brought the accumulated wisdom, good will, and best wishes of the ICEHL community to North America. Later he continued his enthusiastic involvement with the SHEL series as a presenter and reviewer, and only his sudden death in September 2007 stopped him from attending SHEL-5. As the invited commentator for the Old English section of the SHEL-4 collection he was characteristically prompt, incisive, and generous in supporting the enterprise. We dedicate this volume to his memory with deep appreciation of his lasting contributions to our field, of his collegiality and friendship.

Table of contents

Dedication to Richard M. Hogg

v

Tabula Laudatoria

ix

Introduction: Heuristics and evidence in studying the history of the English language Susan Fitzmaurice and Donka Minkova

1

Triggering events William Labov

11

What’s new in Old English? Richard M. Hogg

55

Coding the York-Toronto-Helsinki Parsed Corpus of Old English Prose to investigate the syntax-pragmatics interface Elizabeth Closs Traugott and Susan Pintzuk

61

Anglian dialect features in Old English anonymous homiletic literature: A survey, with preliminary findings R.D. Fulk

81

The elusive progress of prosodical study Thomas Cable Fidelity in versification: Modern English translations of Beowulf and Sir Gawain and the Green Knight Jennifer Anh-Thu Tran Smith

101

121

Response to Tom Cable’s comments Jennifer Anh-Thu Tran Smith

153

Metrical evidence: Did Chaucer translate The Romaunt of the Rose? Xingzhong Li

155

Trochees in an iambic meter: Assumptions or evidence? Xingzhong Li

181

viii

Table of contents

“Ubbe dubbede him to knith”: The scansion of Havelok and Middle English -es, -ed, and -ed(e) Christina M. Fitzgerald

187

A response to Tom Cable Christina M. Fitzgerald

205

Patterns and productivity David Denison

207

Borrowed derivational morphology in Late Middle English: A study of the records of the London Grocers and Goldsmiths Chris C. Palmer

231

Fixer-uppers and passers-by: Nominalization of verb-particle constructions Don Chapman

265

Words and constructions in grammaticalization: The end of the English impersonal construction Graeme Trousdale

301

Variation in Late Modern English: Making the best use of ‘bad data’ Joan C. Beal

327

English/French bilingualism in nineteenth century Louisiana: A social network analysis Connie Eble

337

Taking permissible shortcuts? Limited evidence, heuristic reasoning and the modal auxiliaries in early Canadian English Stefan Dollinger

357

‘What strikes the ear’: Thomas Sheridan and regional pronunciation Raymond Hickey

387

Author index

413

Subject index

423

Tabula Laudatoria

Richard Hogg Cindy Allen Leslie Arnovick Joan Beal Ricardo Bermúdez-Otero Douglas Biber Derek Britton Thomas Cable Don Chapman Claire Cowie Anne Curzan David Denison Teresa Fanego Maurizio Gotti Jonathan Hope Raymond Hickey Patrick Honeybone Marianne Hundt

Dieter Kastovsky Christian Kay Ans van Kemenade Peter Kitson Bill Kretzschmar Ian Lancashire Katie Lowe Emma Moore Lynda Mugglestone Christian Mair Derek Pearsall Malcolm Richardson Nikolaus Ritt Jeremy Smith Robert Stockwell Graeme Trousdale Ilse Wischer Wim van der Wurff

Introduction: Heuristics and evidence in studying the history of the English language*

Susan Fitzmaurice and Donka Minkova This volume continues the project of initiating and energizing the conversations among historians of the English language fostered by the series of conferences on studying the history of the English language (SHEL), begun in 2000 at UCLA. It follows in the footsteps of three high-profile SHELbased collections of peer-reviewed research papers and point-counterpoint commentaries. In the current volume, we invited our contributors to reflect upon their approaches and practices in undertaking historical studies, focusing particularly on the methods deployed in selecting and analyzing data. The essays in this volume represent interests in the study of linguistic change in English that range across different periods, genres, and aspects of the language and show different approaches and use of evidence to deal with the subject. They also represent the current state of research in the field and the nature of the debates in which scholars and historians engage as regards the nature of the evidence adduced in the explanation of change and the robustness of heuristics. We approach the history of the English language from different perspectives. One of us (DM) works on phonology, morphology, and meter, principally in Old and Middle English, while the other (SF) works on grammatical and semantic-pragmatic change, principally in seventeenth and eighteenth century English. Despite these different orientations, however, we share a strong interest in examining the evidence that informs and grounds research in our fields at the same time as interrogating the heuristics employed by our colleagues for the histories they present. The contributions to the volume give expression to these interests. Our first contribution is an essay by SHEL-4 plenary speaker William Labov that explores the nature of what he calls ‘triggering events’. His principal concern is to identify the immediate ‘triggers’ of changes in vowel systems that tend to be represented as chain shifts. His empirical foundation rests both on historical sound changes in English and on data drawn from speakers whose dialects have provided the basis for the

2

Susan Fitzmaurice and Donka Minkova

detailed account of regional accents provided in the Atlas of Northern American English (ANAE). At issue in this article is the nature of ‘causes’ as opposed to triggers: the difference, in Labov’s view, is that “there are bends in the chain of causality at which triggering events are located. Around the bend there are further chains of causality, but they are often orthogonal to the question that drives the original search”. The “triggering events”, as we understand them, can be system-internal, driven by factors such as dispersion or functional load, and externally motivated. The paper demonstrates ANAE’s potential to reveal new relationships between ongoing sound changes and throw new light on the long-standing discussion of the distinction between “proximate mechanisms” and “causal explanations” (Minkova 1999), a central concern for historical linguists. The rest of the volume is organized into four sections, partly along chronological lines, and partly following the topics of the contributions. For each section, we invited a colleague whose own work is related to the topics and approaches represented by the contributions to write an introduction to the papers. The result is interesting conversation within each section with quite different outcomes in terms of our contributors’ responses to the section introductions and the shape and interaction of the essays in that section. Section 1 concerns developments in the study of Old English. In the last completed publication before his untimely death, Richard Hogg’s introduction poses the question, ‘What’s new in Old English studies?’ and finds much in the two studies in the section to help him ponder the answer and conclude optimistically that “[the contributions on Old English] create new avenues to explore and I hope that my own comments have suggested yet other avenues. Old English is alive and well.” For example, the availability of the York-Toronto-Helsinki Parsed Corpus of Old English Prose (YCOE) has significantly enhanced work on Old English syntax. Recent interest in the syntax-pragmatics interface, most especially with respect to information structure, led Elizabeth Traugott and Susan Pintzuk to address the challenge of building on the syntactically parsed corpus to make frequency counts of these factors possible. Their ultimate aim is to enable as detailed an account as is feasible of Old English information-structure, including the function, status, and frequency of clause-initial elements. Their more immediate purpose in their contribution is to discuss the method used in coding for information status and to exemplify some of the potentials of adding this kind of information. Their methodology builds on the notion of antecedence (Prince 1992, Birner and Ward 1998), which identifies referents that are “discourse-old” as opposed to “hearer-old” rather than just “given”. This

Introduction

3

allows for a considerably more fine-grained account of Topicalization and Left-dislocation than has been used in the past, and paves the way for expanding the coding to other nominals and for further detailed comparisons of information structure not only with other stages of English but also with other dead languages available to us only through manuscripts. The premise of Rob Fulk’s paper “Anglian dialect features in Old English anonymous homiletic literature: A survey, with preliminary findings” is the observation that there is much disagreement and confusion about the cause of the admixture of seemingly Anglian dialect features frequently found in Late West Saxon prose. The phenomenon has been attributed alternately to diatopic variation within Wessex, stylistic considerations, and Anglian origins for the greater part of the corpus of late Old English prose (the last possibility perhaps having implications for the dating of the relevant texts). The stylistic explanation, which continues to hold considerable sway, is very plausible only in regard to homiletic works, in which the elevated tone associated with Anglian dialect characteristics would be appropriate. Homilies (including saints’ legends), however, comprise most of the relevant corpus of late anonymous prose, and so uncertainty in regard to these texts is the greatest obstacle to explaining the phenomenon. As a preliminary step in the research required to resolve these issues, then, the corpus of anonymous homilies, amounting to about a quarter of a million words, is surveyed to determine the incidence and distribution of 54 distinctive Anglian (or non-West Saxon) features, including 26 items of vocabulary, and the results are presented for each of the nearly 150 texts. The results justify the preliminary conclusions that (1) reliable Anglian features are commoner in anonymous homilies than has generally been recognized, (2) their incidence shows considerable variability from one text to the next, and (3) features of different linguistic types (phonological/graphemic, morphological, syntactic, and lexical) invariably co-occur in texts that display more than a few Anglian characteristics. Settling the larger question about the cause of seeming dialect mixture will depend upon the close analysis of individual texts. The three studies in Section 2 bring the evidence from older verse into focus. Meter can be a friend or a foe; verse evidence is sometimes deliberately disregarded in syntactic reconstruction. Smith, Li, and Fitzgerald show how verse can be a friend in the undergraduate classroom, in scholarly debates about authorship, and in reconstructing language change. Thomas Cable surveys these contributions with a critical eye; his essay “The Elusive Progress of Prosodical Study” is an excellent example of how

4

Susan Fitzmaurice and Donka Minkova

problematic the area of English historical metrics can be. Here a wellestablished scholar takes issue with both the technical details and the larger conceptual problems in the work of younger scholars who inhabit a parallel theoretical universe. The field of English historical metrics is notoriously divisive, and Cable outlines clearly the historical demarcation between literary and generative prosodists; this in itself is a good lesson for the next generation of English historical metrists. Although the combative tone has gone from our exchanges and we do not start our statements with “I repeat regretfully, respectfully, but peremptorily and irrevocably, that it is impossible to argue with persons who say that …” (Saintsbury 1923: 145, n.), the passions and sometime unbridgeable differences are still there. As volume editors, we take a detached stance; we believe that all four contributions in this section add much to our understanding of the interaction between meter, language, and literature. They offer insights into the advantages and the problems of developing new research methodologies in historical reconstruction. There is clearly both room and need for more work that will further and deepen our knowledge of the older verse traditions, their literary and linguistic settings, and their integration into linguistic theory. Jennifer Smith’s paper “Fidelity in Versification: Modern English Translations of Beowulf and Sir Gawain and the Green Knight” attempts to assess the extent to which modern translators have been faithful in their reproduction of the elements of Old and Middle English alliterative verse. In place of the sense of holistic fidelity that most scholars use to evaluate translations of Old and Middle English verse, she adopts an approach comparing frequencies. Through the examination of six present day English translations of Beowulf and seven present day English translations of Sir Gawain and the Green Knight she is able to conduct an objective or independent assessment of the fidelity of one aspect of verse translation. Her method of analysis consists of measuring how much the scansion of the Modern English poetry violates or deviates from the Old and Middle English standards. She defines alliterating violations as failure to alliterate and alliteration in the wrong position; positional violations are defined as deviations from the four (or five) stressed line of alliterative verse. Ultimately, what she finds is that the attitude of the translators, their perspective on the text, what they see as most salient, be it the sound or the pacing, the alliteration or the epic themes, permeate the Modern English versions entirely. In exploring how modern translators have made choices in their own translations, she seeks to clarify some of the prosodic qualities of OE, ME, and ModE, as well as to contribute to the awareness that the collective power of

Introduction

5

many translations are in and of themselves powerful forces of literary work both in their ability to mediate their source text and in their ability to revitalize linguistically inaccessible texts. Xingzhong Li’s study in this section re-examines the long-standing controversy about the authorship of the ME poem The Romaunt of the Rose translated from the original French poem the Roman de la Rose. Over a century, Chaucerian scholars have used a range of evidence of diction, rhyme, grammar of Chaucerian English, and other information in determining the authorship. The accepted view, dating back to 1900, has been that Chaucer translated only the first 1,705 of the 7,692 lines of the poem; this view has dominated the literature on The Romaunt authorship. Li’s study adopts a comparative approach and exploits new and independent metrical evidence to test the authorship as well as the claims of earlier scholars. The findings strongly support the accepted view that Fragment A of the poem is by Chaucer while Fragment B is not, but they disagree with the centennial hypothesis on the authenticity of Fragment C. In the fourth study involving metrical evidence, Christina Fitzgerald tackles the question of when Middle English -es, -ed, and -ede ceased to be syllabic. Following G. V. Smithers’ (1983) study of scansion and the use of ME -en and -e in Havelok the Dane, Fitzgerald uses scansion of the same early fourteenth century poem – a text with authorial origins in Lincolnshire and soon after copied in Norfolk – to elicit information about the pronunciation of -es, -ed, and -ede. Her study is concerned with scansion and metrical stress as tools to unlock linguistic information rather than with reconstructing all of the prosodic features of the poet’s language. Her analysis shows that 32.67 % of regular and unambiguous metrical environments of -es, -ed, and -ede in the poem show syncopation. Though the pronunciation of inflectional endings outweighs their syncopation by a ratio of approximately 3 : 2, the numbers make it clear that syncopation is by no means an infrequent or isolated phenomenon. Fitzgerald concludes that in the early fourteenth century East Midlands dialect of the Havelok poet, the inflectional endings -es, -ed, and -ede had begun to lose their syllabicity, though they could still be used for metrical purposes in alternating stress poetry. Section 3 of the volume represents the diversity of current approaches to morphosyntactic change in English. David Denison’s introduction “Patterns and Productivity” explores how the three contributions approach the question of the patterns that speakers draw upon in lexical and morphosyntactic innovation, assessing the relative productivity of particular patterns.

6

Susan Fitzmaurice and Donka Minkova

His essay raises important questions about the extent to which researchers can make reasonable inferences about patterns on the basis of different kinds of evidence – from the immensely rich source of the internet to the surviving records of small groups of speakers in particular historical periods to selected examples from the Old English corpus. The three essays themselves adopt rather different perspectives to this question. The methods adopted by the group are more speculative in some respects than they are analytical – Chapman in terms of treating the internet as an open-ended corpus, Palmer in terms of extrapolating from a very small dataset, and Trousdale in terms of being truly speculative. Denison’s essay provides an insightful and energetic critique of their effectiveness and their differences. Chris C. Palmer’s paper “Borrowed Derivational Morphology in Late Middle English: A Study of the Records of the London Grocers and Goldsmiths” seeks to characterize the morphological status of several derivational suffixes borrowed from French and Latin within the records of two communities in the fifteenth century. It compares the use of native nominal affixes (-ness, -ship, and -hood) with borrowed, potential affixes (-cion, -ance, -ity, -age, and -ment) throughout the English portions of these multilingual texts. Attempting to locate evidence of the naturalization of these forms – the process by which these endings become derivational morphemes in the general English lexicon – Palmer develops the notion of local productivity. This measure combines both quantitative and qualitative data to show that, even in smaller corpora, historical linguists can find evidence of the morphological status of different potential affixes for communities within particular historical moments. Palmer finds that despite variation in the use of -age between the two communities the majority of borrowed potential affixes were in the early stages of naturalization. They had limited productivity within a restricted subset of the lexis, and speakers were beginning to see them as individual units. Ultimately, Palmer argues that diachronic studies should consider such data to better understand the social and linguistic mechanisms that may have led to the increasing productivity of borrowed derivational affixes in the English language. The -er suffix has been one of the most productive derivational suffixes in English, regularly forming nominalizations of verbs like farmer, teacher and writer. The verb + particle construction, like wake up or find out has also been long established in English. But the combination of these two constructions, that is the -er suffixation of a verb + particle construction has not been as well established. There are four possibilities, namely patterns like on-looker, looker-on, looker-onner, and look-onner, and in his study

Introduction

7

“Fixer-uppers and passers-by: Nominalization of verb-particle constructions”, Don Chapman examines the distribution of each of these in the history of English. This sort of examination is difficult to conduct using standard corpora because these nominalizations retain an ad hoc or ludic feel to them and rarely show up in published writings. The Oxford English Dictionary (OED) and the internet present two tools for finding less established constructions like this one. The results of examining the distribution of these nominalizations in the OED are that the pattern looker-on has been the most widely used, followed by the pattern on-looker. Both however dip sharply in the twentieth century, the same century that the pattern lookeronner began to appear. The results of Chapman’s internet search confirm the trend of the twentieth century in the OED, namely that looker-onner has increased in usage while on-looker and looker-on have decreased. He observes that looker-onner appears to be the most productive form by far, but looker-on continues to be used, while on-looker has become only minimally productive. The form that did not occur in the OED, look-onner, is used as much as looker-on in the internet searches. Chapman speculates that if this represents a trend, perhaps one day look-onner will supplant looker-onner. This trend would present a progression of looker-on > looker-onner > look onner, such that the suffix gradually moves from the verb where it properly belongs to the end of the term, where suffixes usually go, with an intermediate stage in which the suffix occurs in both positions. Graeme Trousdale’s article “Words and constructions in grammaticalization: the end of the English impersonal construction” is concerned with the remnants of the impersonal construction in early Modern English, and specifically with the role of grammaticalization in the development of both this construction and the transitive construction in English. An informative outline of the history of the impersonal construction up to and including the early Modern period is followed by a discussion of impersonal and transitive constructions using some of the theoretical apparatus from cognitive linguistics in general, and Cognitive Grammar in particular. Finally, he accounts for the change in terms of patterns of grammaticalization, developing some of the arguments of Meillet (1958 [1912]) whose work is traditionally cited in discussions of the grammaticalization of lexical items, but who in fact also considered changes in word order and phrase structure as potential instances of grammaticalization. The article proposes that constructional accounts of language structure can both inform and be informed by grammaticalization theory.

8

Susan Fitzmaurice and Donka Minkova

Section 4 concludes the volume. In this section, three authors provide different perspectives on aspects of variation in late modern English (LModE). The title of Joan Beal’s introduction “Variation in Late Modern English: making the best use of ‘bad data’” echoes Roger Lass’ (1990) famous “How to do things with junk”. She provides a cogent research context for examining the complex question of how to extract answers from evidence that is scanty and holey. The papers by Eble and Dollinger point up the problem very clearly as they deal with varieties of English that have not been studied from a sociohistorical perspective. In contrast, Hickey scrutinizes the evidence for assessing the extent to which the pronouncements of an influential commentator such as Sheridan might shape language history. Connie Eble’s essay “English/French bilingualism in nineteenth-century Louisiana: A social network analysis” applies the notion of social network to an archive of family papers to explore why French disappeared as the language of public life in northwestern Louisiana by the time of the Civil War. The social network itself and the language practices of a cohesive group of white creoles living in rural Natchitoches Parish during the nineteenth century are inferred from the writings preserved in the Prudhomme Family Papers. For a half century, dense and strong ties of kinship preserved French as the language of the descendants of the founder population, while restricting it increasingly to the domains of personal life and religious practice. At the same time, weaker connections favoured the addition of English for communication outside the creole network that aided economic prosperity. Sending members of the younger generation beyond the local area to learn to speak, read, and write English made knowledge of English a property of their ties to each other and made them conduits of linguistic change to their creole network. Eble finds that by the beginning of the twentieth century, French had disappeared entirely from the Prudhomme Family Papers and presumably from the lives of the creole network of the area. Stefan Dollinger asks provocatively: “Progressive colonial English?” The paper deals with a ‘bad’ data problem specific to colonial Englishes in the Late Modern English period. Considering the complex sociolinguistic situation of newly-formed colonial varieties in the LModE period, Dollinger argues, in light of the present suboptimal resources, for the adoption of heuristic methods of approximation, of ‘good-enough’ estimates that have proved useful in other disciplines. These methods may provide a feasible shortcut for English historical linguistics in general, but particularly for the

Introduction

9

characterization of colonial varieties. The approach is illustrated by the semantic development of CAN and MAY in Ontario English in terms of their progressive, respectively conservative, behaviour in comparison to British English, before being applied to a larger set of modal auxiliaries. In the LModE colonial context, the limitations of statistical testing are discussed and a solution is suggested by combining LModE findings, based on limited data, with twentieth-century findings on the modal auxiliaries, which allows the assessment of features of colonial varieties with a certain degree of confidence. Raymond Hickey’s paper “‘What strikes the ear’: Thomas Sheridan and regional pronunciation” concludes the volume. The study turns our attention to the contemporary description of late modern English by considering the role of the elocutionist and grammarian Thomas Sheridan in the rise of sociolinguistic censure. He examines Sheridan’s attitude to non-standard features in both southeastern British English and Irish English in the late eighteenth century to track how prescriptive notions of language use seemed to be fleshed out during this time. He looks in some detail at what present researchers might glean about the nature of regional pronunciations in the late modern period from Sheridan’s negative comments on the speech of his fellow Irishmen. Finally, he also considers the possible influence of Sheridan’s strictures on the development of Irish English during the nineteenth-century. We close with acknowledgments and thanks to the individuals and organizations that helped the progress of the conversations captured in this volume. The expertise of our reviewers (named in the Tabula Laudatoria) informed the work performed by new and established scholars alike in developing their contributions.

Notes *

The term heuristics is understood in different ways in humanities. In our understanding, heuristics refers to methods of discovery, which may include empirical, analytical and speculative methods.

10

Susan Fitzmaurice and Donka Minkova

References Birner, Betty J., and Gregory Ward 1998 Information Status and Non-Canonical Word Order in English. (Studies in Language Companion Series 40.) Amsterdam/Philadelphia: John Benjamins. Lass, Roger 1990 How to Do Things with Junk: Exaptation in Language Evolution. Journal of Linguistics 26.1: 79–109. Meillet, Antoine 1958 Linguistique historique et linguistique général. Paris. Champion. [1912] Minkova, Donka. 1999 Proximate mechanisms vs. causal explanations, Zeitschrift für Sprachwissenschaft 18.2: 226–230. Prince, Ellen F. 1992 The ZPG letter: Subjects, definiteness, and information-status. In Discourse Description: Diverse Analyses of a Fundraising Text, William C. Mann and Sandra A. Thompson (eds.), 295–325. (Pragmatics and Beyond, New Series 16.). Amsterdam/Philadelphia: John Benjamins. Saintsbury, George 1923 A History of English Prosody. Vol. I: From the Origins to Spenser. London: Macmillan and Co. Limited. Smithers, G.V. 1983 The scansion of Havelok and the use of ME –en and –e in Havelok and by Chaucer. In Middle English studies presented to Norman Davis in honour of his seventieth birthday, Douglas Gray and E.G. Stanley (eds.), 195–234. Oxford: Clarendon Press.

Triggering events

William Labov There is general agreement that the heart of the study of language change is the search for causes.1 It is what we generally mean by the explanation of change. And while we would like to apply to this search the universal principles that govern grammar as a whole, it is also understood, following Meillet (1921), that no universal principles can account for the sporadic course of change, in which particular changes begin and end at a given time in history. The actuation problem (Weinreich, Labov and Herzog 1968) demands that we search for universals in particulars. However, the pursuit of the causes of any given change might on further reflection involve us in an unsatisfactory and endless recursion. It goes without saying that any given state of a language is the outcome of a previous state of that language, and so on backward in time as far into the past as our knowledge can carry us. The title of this chapter then needs some justification if it refers to linguistic events. In such an endless chain of causes, every state of the language is a triggering event for the one that follows. Even if there is no change in a given system, it has a cause: the state of equilibrium that was reached in the preceding period. And when there is change, as Martinet (1955) has argued, the evolving system reflects a series of earlier readjustments that spiral backward in time. I would like to defend the concept of “triggering event” by arguing that this sequence of preceding causes is not a smooth and uniform sequence. Rather, there are bends in the chain of causality at which triggering events are located. Around the bend there are further chains of causality, but they are often orthogonal to the question that drives the original search (Figure 1). A nonlinguistic example may illustrate the point. We are all interested in the pre-history that gave rise to mammalian evolution, and in this causal sequence we encounter the extinction at the K-T boundary between the Cretaceous and Tertiary periods of the dinosaurs, along with plesiosaurs, mosasaurs, and a majority of all other existing families. What caused this massive extinction? The most strongly supported theory is that of Luis

12

William Labov

and Walter Alvarez, originally proposed in the 1980s: that the K-T extinction was the result of the impact of a large meteor with the earth. While the exact killing mechanisms may or may not yet have been identified, all the data – including the rate of extinction, the nature of the recovery, and the patterns of survivorship – are concordant with the hypothesis of extinction by asteroid impact. (Fastovsky and Sheehan 2004).

What were the causes of this intersection of asteroid and earth? It is an important question for the future of the human race, which would be profoundly influenced by a major impact of this kind. The answer would involve calculations of celestial mechanics that are not immediately relevant to the later history of biological evolution. The hypothesis of a meteor impact, if it continues to be supported, provides a satisfactory answer to the question, what was the triggering event that gave rise to mammalian predominance in the evolutionary sequence? The linguistic triggering event that we are looking for may also be the result of a variety of factors concatenated by historical accident.

Figure 1. A bend in the chain of causality

Chain shifts are a natural subject for the study of causal sequences, and the search for triggering events. Table 1 lists six chain shifts studied in the Atlas of North American English (Labov, Ash and Boberg 2006; henceforth ANAE). The most recent events are listed on the right, and the events preceding them in the two columns on the left. Some shifts link two or three events, one shows five. In each case we are led to ask, what was the triggering event that was responsible for this shift? We might think, again following Martinet, that this triggering event must be an external event impinging on the linguistic process, like the Norman invasion or World War II, outside of the realm of autonomous lin guistic explanation. For some shifts, this is the case.2 But in others, it will

Triggering events

13

Table 1. Six chain shifts described in ANAE. The Canadian Shift (Ch. 15) The Pittsburgh Shift (Ch. 19) The Northern Cities Shift (Ch. 14) The Southern Shift (Ch. 18) The Back Upglide Shift (Ch. 18) Back Chain Shift before /r/ (Ch. 19)

Entering → /e/ → /æ/ → /¡ / → 

/o/

Leaving→ /o/

→

→

→

/¡ / →

/oh/ →

/o/

/iy/ →

/ey/ →

/ay/ →

/ah/

/nw/ →

/aw/ →

/ahr/ →

/ohr/ →

/e/



→

/æ/ →

appear that there are linguistic bends in the chain of causality. I will argue that there are triggering events of a purely linguistic character. Their explanation calls upon a different set of principles than those that operate on the changes they initiate. First however I’d like to show that bends in the linguistic chain are essential characteristics of chain shifts. In fact, without such shifts of direction it will be difficult to defend the very concept of a chain shift. Consider the simplest kind of chain shift (1)

BĺAĺ

A is the leaving element and B is the entering element following the notation of Table 1.3 A causal connection might be said to exist if A moves away because B approached A, reducing the margin of security, or if A moved away, increasing the margin of security, and B consequently moved in the direction of A. However, such chain shift events are subject to an alternative interpretation. The movement of A may be generalized to B, just as the change of a front vowel may be generalized to the corresponding back vowel without any relevant change in margins of security. In (2), if A is a vowel /e/ moving in the vowel space from mid to high, and B is a low vowel /æ/ that moves from low to mid behind it, one could argue that the movements of A and B are causally related. But this can also be conceived as a single expression (3) in which all front vowels undergo a loss of one

14

William Labov

degree of openness. Thus whatever factor C acted on /e/ to make it less open came to act equally on /æ/ so that the causal relationship is seen as (4) rather than (1). (2)

(3)

A

eĺi

B

æĺe

Į open ĺ Į -1 open / _____ [+ant]

(4)

C

B

A

Option (4) is not available if A and B are different kinds of linguistic processes. Thus in the Southern Shift, A is the monophthongization of /ay/ and B is the lowering and centralization of the nucleus of /ey/ (ANAE: Ch. 18), as represented in (5). In A, /ay/ is a vowel from the subsystem of front upgliding vowels that moves to the system of long and ingliding vowels, while B is an adjustment entirely within the set of front upgliding vowels. (5)

A

ay ĺ ah4

B

ey ĺ ay

Here we must accept a chain shift of type (1) since there is no single process that can be generalized to unite the behavior of A and B. The causal relationship seems clear: the removal of /ay/ from the front upgliding system led to a readjustment by the well recognized principle of maximal dispersion – the tendency of vowels to achieve equidistant positions within a subsystem (Martinet 1955, Liljencrants and Lindblom 1972, Disner 1980, Lindblom 1988). Figure 2 sums up the characterization of these two situations: generalizable shifts within a subsystem, and sequential shifts across subsystems. The type of causal explanation applied to chain shifts is not in question here. In this search for triggering events, one may take a teleological posi-

Triggering events

15

Figure 2. Generalizable and sequential chain shifts across subsystems.

tion, like Martinet (1955) or Jakobson (1972), and argue that speakers shift their vowels to minimize misunderstanding. Or one can attribute these linked movements to the mechanical effects of misunderstanding on the probability matching of the language learner (Labov 1994, Ch. 20). Evidence for the causal link may come from temporal sequencing, geographic nesting, or internal correlation (ANAE Chs. 14, 18). However, the order of events is crucial to the present discussion: whether we are dealing with a drag chain or a push chain will be decisive in the search for triggering events. Subsystems of English vowels. Much of the logic of chain shifting involves movements out of and into subsystems. The binary notation of ANAE Ch. 2, is designed to characterize these subsystems in a coherent and systematic manner. Table 2 shows the four subsystems of North American English: short vowels, front upgliding vowels, back upgliding vowels, and the smaller set of long and ingliding vowels.5 In r-pronouncing dialects, this fourth subset consists primarily of the /ah/ class in father, ma, pa, pajama, etc., and /oh/ in law, talk, off, cloth, etc. In r-less dialects, the marginal members of this subset /ih/, /eh/, /uh/ are greatly expanded to include beer, bare, boor, etc. The notation does not describe the set of contrasts in any one dialect, but rather the initial position from which present-day dialects can be derived. In that sense, the individual units are historical word classes comparable to the lexical key words presented in Wells (1982).6 The principles of maximal dispersion and maintenance of margins of security developed in Martinet (1955) operate within subsystems (see also Liljencrants and Lindblom 1972, Lindblom 1988). Labov (1994, Ch. 9) presented data from misunderstandings in spontaneous speech which show that confusions occur primarily within members of a subsystem, rather than across subsystems. There is for example more confusion between /i/ and /e/

16

William Labov

than between /e/ and /ey/, and more between /ey/ and /ay/ than between /ay/ and /aw/. Table 2. ANAE notation for four subsystems of vowels of North American English, with type words. LONG SHORT

Front upgliding

Back upgliding

Vy

Vw

V nucleus high

mid

low

1.

front

back

Ingliding

Upgliding

front

back

front

Vh back

/i/

/u/

/iy/

/iw/

/uw/

bit

put

beat

suit

boot

unrounded

rounded

/e/

/¡/

/ey/

/oy/

/ow/

/oh/

bet

but

bait

boy

boat

bought

/æ/

/o/

/ay/

/aw/

/ah/

bat

cot

bite

bout

balm

The Canadian Shift

The first of the North American chain shifts in Table 1 is the Canadian Shift as in (6). (6)

/e/ → /æ/ → /o/ →

This chain shift was first described by Clarke et al. in 1995 on the basis of word lists read by 16 college students, and has since been confirmed by several further studies of Canadian English (ANAE Ch.16; Boberg 2004). It is the most consistent marker of the Canadian English dialect in ANAE, and it is the basis for the isogloss defining the Canada region (including all points in Canada outside of the Atlantic Provinces). Figure 3 compares the Canada dialect region with the combined means for all others for the vowels involved in the shift. There is no notable difference for /i/. But the Canadian /e/ is significantly lower than the general mean, and an even greater difference appears for /æ/ in both F1 and F2. One can also observe that Canadian /o/ is well back of the general mean.

Triggering events

17

Figure 3. Mean values of vowels in the Canadian Shift for the Canada region [N = 25] and all other dialects combined [N = 414]. Source: ANAE Ch. 15]

It was clear from the outset that the Canadian shift of the short front vowels was a response to the low back merger of /o/ and /oh/ in cot and caught, Don and dawn, etc., well established in Canada. (6a) is therefore a more complete representation of the Canadian Shift. /oh/ is not a leaving element, but collapses with /o/. (6a) /e/ → /æ/ → /o/ → /oh/ To which subsystem do we assign the collapsed vowel phonologically? The decision is dictated by phonological facts. While the original short-o was a checked vowel, which cannot occur in stressed word-final position, the merged vowel occurs in free as well as checked position: that is, the vowel of cot is now an allophone of the vowel of caw. Though both vowels may shift position in the course of the merger,7 it is /o/ that moves to the long and ingliding subsystem rather than /oh/ to the short subsystem. Figure 4 embeds the Canadian Shift in the acoustically defined phonological space characteristic of the Germanic language family, with a peripheral region enclosing a nonperipheral region.8 By the principles of chain shifting developed in Labov (1994, Ch. 5–6), tense or long vowel nuclei rise along the peripheral track and lax or short nuclei fall along the nonperipheral track. A shift from a short to a long subsystem appears as a movement from a nonperipheral to a peripheral track, as indicated in Figure 4. The remaining

18

William Labov

short vowels then readjust their positions along the nonperipheral track to achieve maximal dispersion.

Figure 4. Shift of subsystems in the Canadian Shift

The temporal relations of the low back merger and the Canadian shift are consistent with the causal assignment to the merger as prior. As noted above, the first report of the shift of /e/ and /æ/ date from 1995. The low back merger in Canada was firmly documented well before then (Scargill and Warkentyne 1972, Gregg 1957). Chambers (1993: 11–12) cites literary sources for the merger in the middle of the 19th century. The geographic distribution of the Canadian Shift and the low back merger are also consistent with the causal connection inferred; here we encounter the nesting relation that plays an important role in the application of dialect geography to historical sequencing. Figure 5 maps the distribution of ANAE subjects who satisfy the acoustic criteria for the Canadian Shift (grey symbols), and the isogloss that defines the region in which these symbols predominate. The homogeneity of this isogloss – the proportion of speakers within the area who satisfy the criteria – is .84. Twenty-one of the 25 Canadians within the isogloss do so, producing a more reliable definition of the Canadian dialect than Canadian raising, the best known stereotype of Canadian English (ANAE, Ch. 15). However, consistency – the proportion of speakers who show the trait who are within the isogloss – is quite low, since the same forces are operating wherever the low back merger is found. The implicational relation between the Canadian Shift and the low back merger is evident in that only three of the 60 speakers who show the Canadian Shift have /o/ and /oh/ distinct. The important geographic relation is that the Canadian Shift isogloss is strictly contained within the low back merger isogloss (the oriented line on Figure 5). The low back merger extends to a much wider territory, covering the West, Western Pennsylvania and Eastern New England. A total of 123 speakers

Triggering events

19

produced /o/ and /oh/ the same in minimal pair tests, and only 60 showed the back shifting of /e/ and /æ/. At the same time, the Canadian Shift does appear among a minority in other low back merger areas: twelve in the West, five in Western Pennsylvania, four in Texas, where the merger is reported in progress (Bailey et al. 1991); and seven in the Midland where the merger is generally in transition. However only two grey symbols appear within the dashed isoglosses: these outline the areas of greatest resistance to the merger: in the Inland North, the Mid-Atlantic States and the South.

Figure 5. Nesting of Canadian Shift within the Low Back Merger isogloss

Both temporal and spatial evidence reinforce the general principles of chain shifting to indicate that the low back merger creates the conditions for the backing of /æ/ and accompanying backing and lowering of /e/. In removing /o/ from the subset of short vowels,9 it acts as the triggering event for the Canadian Shift.

20

2.

William Labov

The Pittsburgh Shift

ANAE reports for the first time a chain shift in the city of Pittsburgh, as indicated in (7). As in (6a), we add the third element /oh/, indicating the low back merger that is missing in Table 2. (7) /¡ / → /o/ → /oh/ 

Figure 6 presents the Pittsburgh Shift in the same framework as Figure 4. The low back merger is solidly entrenched in Pittsburgh, as it is in Canada. But in Pittsburgh, the phoneme /¡/ moves downward on the nonperipheral track from mid, back of center position, while /æ/ remains in place in the low front area. Figure 7 provides a detailed view of this downward movement in the vowel system of a 35-year-old man from Pittsburgh, interviewed in 1996. On the left, the short-a vowels follow the nasal system: words with nasal codas are raised to mid and upper-mid position, while all others are in a tight cluster in low front position. In the back, /o/ is clearly merged with /oh/ in the same lower mid back position as in Canada. Between /æ/ and /o/~/oh/ are located the majority of the tokens of /¡ /. Words with /¡ / before /n/ are particularly low (sun, fun, months), but the token of duck is regularly judged to be dock by speakers of other dialects. Figure 8 places this Pittsburgh development against the mean values of the low vowels of Canada and 18 other North American dialects.10 It can be seen that the mean /æ/ of Canada is the furthest back of all dialects, while Pittsburgh /æ/ is in normal low front, unraised position.11 At right, both Canada and Pittsburgh show the merger of /o/ and /oh/ in lower mid back position (the two Canada tokens practically coincide). In the center, the Pittsburgh mean for /¡ / is much lower than any other dialect, not far from the general /o/ distribution. 





Figure 6. Shift of subsystems in the Pittsburgh Shift

Triggering events

21

Figure 7. The Pittsburgh Shift in the vowel system of Kenneth K., 35 [1996], TS 545.

Figure 8. Mean positions of low vowels for 20 ANAE dialects, with Canadian Shift labeled for Canada [CA], Pittsburgh Shift labeled for Pittsburgh [PI] and Northern Cities Shift labeled for the Inland North [IN].

22

William Labov

The low back merger is evidently the conditioning event for the Pittsburgh Shift, just as it is for the Canadian Shift. Here, however we have the same cause with two different effects. In the search for causes of linguistic change, it seems reasonable to expect that the same cause will have similar or comparable effects. Why is it that /¡ / instead of /æ/ moved into the empty space created by the back shift of /o/ and its merger with /oh/? Among North American sound changes, there are other cases of two neighboring phonemes competing to fill the empty space in the pattern.12 One might say that these are two equally likely possibilities, and it is a matter of chance which was realized. But these choices are not equiprobable: there are 60 communities which show evidence of the Canadian Shift and only one city with the Pittsburgh Shift.13 To account for the unique Pittsburgh development, it is not unreasonable to turn to the other unique feature of the Pittsburgh dialect: the monophthongization of /aw/. The Pittsburgh long monophthong in down, town, south, out and house is located in low central position, partially overlapping with /¡ /. There is no danger of confusion between /¡ / and /aw/, however, since monophthongized /aw/ (now /ah/) has twice the length of /¡ /, so that typically the longest /¡ / is shorter than the shortest /aw/ (ANAE p. 273). One hypothesis is that the lowering of /¡ / is the result of a change in the organization of the vowel system of Pittsburgh speakers in which /¡ / is re-analyzed as /a/, the short counterpart of /ah/. This would oppose the long and short pairs down ~ dun, about ~ but, howl ~ hull as /dahn ~ dan, ‹baht ~ bat, hahl ~ hal/. If further evidence supports such an abstract re-analysis, then both the low back merger and the monophthongization of /aw/ appear to be triggering events for the Pittsburgh Shift. Both are movements of word classes into the long and ingliding subsystem from other subsystems. 













3.

Causes of the low back merger

Given our understanding of the effect of the low back merger on other linguistic events, the question that naturally arises is: what are the causes of the low back merger? Herold (1990, 1997) has provided a convincing social account of the actuation of the low back merger in Eastern Pennsylvania – the influx of large numbers of immigrants from Eastern Europe into coal-mining communities. However, no linguistic mechanism for a substratum effect has yet been staked out, and the inquiry we are conducting here calls for a much more general solution. We must account for the linguistic

Triggering events

23

antecedents of the collapse of /o/ and /oh/ in more than half of the North American continent with a variety of vowel systems, and in Scotland as well. Why then is the distinction between /o/ and /oh/ so likely to collapse? If there is a linguistic answer to this question, then the low back merger is not the triggering event we are looking for, but it is only a link in the causal chain. A first thought about the cause of a merger is the functional load of the distinction. In the case of /o/ and /oh/, there is no problem in finding minimal pairs. We can generate sizeable numbers in the style of (8). (8)

cot rot tot sot cotter dotter Don yon pond fond hock stock

caught wrought taught sought caught her daughter dawn yawn pawned fawned hawk stalk

cock tock odd nod cod mod sod Sol moll collar holler odd ability

caulk talk awed gnawed cawed Maud sawed Saul maul caller hauler audibility

This proliferation of minimal pairs masks, however, the odd skewing in the distribution of /o/ and /oh/ that appears in Table 3. Almost all of the contrast between /o/ and /oh/ is before a set of five apical consonants /t,d,s,n,l/ and one non-apical /k/, as indicated by the bold lines. There is no contrast before labials or palatals. Occurrences of /o/ before /z/ are limited to special lexical items and words in which intervocalic /s/ is voiced. In the lower half of Table 3, there are six environments where /oh/ is not represented at all, and one – final position – where /o/ does not appear. Three sets of /oh/ words in Table 3 are italicized. These are /o/ words that are tensed in American English before front voiceless fricatives and nasals, the same core phonetic conditioning that operates in the tensing of short-a in the Mid-Atlantic region and broad-a in Britain (Ferguson 1975, Labov 1989).14 This tensing process typically proceeds by lexical diffusion, but does not substantially increase contrast between /o/ and /oh/. There are a total of six environments in which one side or the other is represented by a small number of learned, colloquial or specialized vocabulary, so that in

24

William Labov

twelve environments, contrast is marginal; monosyllabic and minimal pairs are not to be found.15 Table 3. ANAE notation for four subsystems of vowels of North American English, with type words. /o/

/oh/

APICALS t

cot, tot, hot, got, dot

caught. bought, taut, fought

d

odd, hod, god, sod

awed, hawed, gaud, sawed

s

toss, moss, floss, cost, loss

sauce, exhaust, caustic

z

(Oz, positive)

cause, clause, hawser, pause, paws

n

don, Ron, pond

dawn, awn, yawn, lawn

l

doll, moll, collar

all, tall, maul, caller

NON-APICALS p

hop, pop, top, sop

b

rob, hob

F 

Scotch, botch, watch

----------

j

lodge, dodge, Roger

----------

g

log. hog, cog, dog

(auger, augment, augur, August)

k

stock, hock, clock

stalk, hawk, talk

f

(boff, toff)

→ off, doff, scoff (cough, trough)

7

(Goth)

→ cloth, moth

6

(gosh, bosh)

(wash)

'

(bother)

] 

----------

m

bomb, Tom, prom

1

(pong, Kong)

#

----------

---------(daub, bauble)

---------------------------→ strong. song, wrong, strong law, saw, flaw, thaw, claw

Triggering events

25

Figure 9. Historical development of the long open-o word class

In order to see how this bizarre distribution came about, it may be helpful to review the historical formation of this word class, as summarized schematically in Figure 9. Proceeding from left to right, the diagram shows 1. an original /aw/ diphthong in Old English (thaw, straw, claw) 2. additions to Old English /aw/ through Old English sound changes a. breaking and rounding of strong verb preterits before velars in complex codas (fought, taught) b. vocalization and rounding of /l/ in complex codas (talk, call, all) c. vocalization of coda /g/ (maw, saw, draw) 3. additions to /aw/ in Middle English through vocalization of /v/ (hawk, laundry) accretion of new /aw/ forms from Old French loan words a. original OF back upgliding diphthongs (applaud, because) b. collapse of bisyllabic /a + u/ words to single syllables (pawn, brawn) c. denasalization and rounding of nasal vowels (lawn, spawn) 4. smoothing (monophthongization) of /aw/ to /oh/.

26

William Labov

5. lengthening of /o/ to /oh/ in Early Modern English before voiceless fricatives and velar nasals (cloth, off, loss, lost, strong, song, wrong, long) 6. lexically irregular rounding of /a/ after /w/ (water, warrant, walrus) The O.E. /aw/ class traced here is not a reflex of PGmc /aw/, which is realized in Old English as ƝƗ. It was cobbled together by a series of conditioned sound changes so that its shape is a matter of historical accident. The general sound change that set the stage for the low back merger was the smoothing of ME /aw/ to /oh/.16 It must have taken place before the shift of Middle English /o/ to /oh/ by compensatory lengthening in thought and brought. We can also argue that it must have also preceded the completion of the Great Vowel Shift in the back vowels, by which ME u: diphthongized to /aw/. The smoothing of /aw/ created the juxtaposition of /o/ and /oh/ – two lower back mid vowels differentiated only by length,17 which is unstable on two counts. First, it is well established that length distinctions without accompanying differences in vowel quality tend to collapse, in English and many other languages (Chen and Wang 1975). Second is the asymmetrical distribution of Table 3. Given this situation, the merger of the opposition is a likely outcome unless qualitative differences develop to support it. Such qualitative differentiation of /o/ and /oh/ did develop in three areas outlined by the dotted isoglosses of Figure 5 (ANAE Ch. 11): (1) the unrounding and fronting of /o/ in Western New England and New York State;18 (2) the raising of /oh/ to upper mid position (east coast dialects from Providence to Baltimore); (3) restoration of the back upglide of /oh/ in the South.19 Outside of these areas, the low back merger is either complete or in transition. It follows that the juxtaposition of long and short open-o by the smoothing of /aw/ was the triggering event of the low back merger. What is the relationship of the other events of Figure 9 to the low back merger? The /aw/ class originated in final position, where it could not contrast with short open /o/. The changes that followed were largely conditioned by the vocalization of /l, g, x/ in complex codas before /k/, /l/, /t/. They created the limited contrasts which resisted the merger to a certain extent; however, one would have to say that it was the absence of sound changes conditioned by other consonants that favored the merger. If the smoothing of /aw/ was the triggering event for the low back merger and ultimately the Canadian Shift and the Pittsburgh Shift, we must ask if it in turn had a relevant predecessor. I argued that it must have pre-

Triggering events

27

ceded the completion of the Great Vowel Shift on the assumption that it was a drag chain. But it is also possible that a push chain was involved, and that the descending diphthong [8u] → [ԥu] → in out, south, down, etc. reduced the margin of security of /aw/ realized as [#u] in a way that promoted the shift to [n:]. If that is the case, we would have to push our inquiry into the triggering event of the Great Vowel Shift, a question that has been much discussed (Luick 1903, Martinet 1955, Stockwell and Minkova 1997). There is not enough evidence to pursue this connection here, except to emphasize the possibility of a chain of linguistic triggering events receding into the past. In any case, there is no reason to believe that any one external event intervened to produce these chain shifts.

4.

The fronting of /uw/

In the two cases just studied, the low back merger was seen to set the conditions for subsequent changes in the vowel system, responding to the tendency of subsystems to maintain equidistant spacing or maximal dispersion. We will now consider a sound change that appears to be inconsistent with previous explanation based on these principles. This is the fronting of /uw/, an ongoing shift that covers 90% of the North American continent. The various phonetic forms involved are shown in (9). (9)

u

7

u

Œ

üu

ü

7A +

ü

u

Martinet (1955) advanced an explanation for what is now recognized as a general principle of chain shifting: that back vowels move to the front.20 He argues that the repeated fronting of /u/ and /o/ is the result of the fact that even though there is a strong tendency to front-back symmetry in the vowel system, there is physically less room in the back than in the front. Such fronting is then the result of pressure to relieve overcrowding among the back vowels. Specifically, this happens when through one linguistic process or another, a vowel system develops four degrees of height among the back vowels. Haudricourt and Juilland (1949) applied this logic to a wide range of sound changes in western Europe and confirmed Martinet’s prediction in every case. Labov (1991), defining three major dialects of English, argued that the third dialect, characterized by the low back merger, would be sta-

28

William Labov

ble, and resist the fronting of /uw/ and /ow/ that is predominant in the South and the Midland. Figure 10 shows that the completed ANAE data does not satisfy this expectation. The grey symbols identify speakers for whom /uw/ after coronal consonants – in do, dew, too, two, soon, noon, etc. – is front of center, that is, mean F2 is greater than the midpoint of 1550 Hz in this normalized system. This includes 89% of the population: there are only 49 of the 439 ANAE subjects for whom this is not the case. Furthermore, these 49 are concentrated in two narrowly circumscribed areas: New England and the Minnesota-Wisconsin. In general, Eastern New England is a conservative area in regard to the fronting of /uw/ and /ow/, and its behavior is consistent with what we would expect from the low back merger in that area. The Minnesota-Wisconsin area shows considerable variation in regard to the low back merger. But the conservative character of the vowel system, with back /uw/ and /ow/ often monophthongal, must be accounted for by a strong Scandinavian and German substratum (Allen 1973).

Figure 10. Fronting of /uw/ after coronal consonants. Grey symbols = Second formant > 1550 Hz.

Triggering events

29

Once we have dispensed with these two areas, we are faced with the fact that /uw/ is fronted without exception in all other regions: in the Midland, the Mid-Atlantic States, in the South, and most importantly in three areas where the low back merger is complete: Canada, The West, and Western Pennsylvania. It is not possible to account for this massive, continentalwide fronting as a response to overcrowding among the back vowels. Although the structural approach to the causes of /uw/ fronting in North America seems to fail in this case, we can open an inquiry into the causes of this phenomenon from another structural direction. Because /uw/ fronting is so widespread in North America, it is unlikely that we will find a specific population movement like the migration of Slavic coal miners into Eastern Pennsylvania identified by Herold (1990). The antecedent event must be one of great generality. One clue to the problem may be found in the extraordinary difference between the fronting of /uw/ after coronal consonants, examined in Figure 10, and the same word class after non-coronal consonants in roof, boots, coop, food, move, etc. While 390 ANAE subjects shifted /uw/ after coronal consonants front of center, only 130 did so for the non-coronal class. Table 4 (columns 2 and 3) demonstrates this extraordinary effect of coronal onset in a regression analysis of all 4,747 tokens of /uw/ measured acoustically. The age coefficient in Table 4 indicates vigorous change in progress in apparent time. The figure –101 in column 1, row 1 indicates that the expected value of F2 for speakers 25 years older than the mean is 101 Hz less than the general mean of F2 for /uw/, all other things being equal. For the generation 25 years younger than the mean age, the fronting of /uw/ is advanced by 101 Hz. As in most sound changes in progress, women lead: in this case by the effect of half a generation. Among the internal constraints, the effect of a preceding coronal stands out at 480 Hz, more than twice the effect of any other. This means that for the average speaker with a mean F2 for /uw/ after coronal consonants of 1800 Hz, the value of /uw/ after noncoronals is around 1300 Hz, half way between a back and a center vowel. This preponderant effect of preceding coronals is a striking exception to the general rule that English vowels are influenced by the following environment much more than the preceding one.21 It is not difficult to explain the tendency for preceding coronals to promote the fronting of /uw/, which is a widespread effect. It appears strongly in Lennig’s (1978) analysis of sound change in progress in Paris. The F2 locus of apical consonants ranges closely around 1800 Hz, so that when a following back /uw/ requires

30

William Labov

Table 4. Regression coefficients for F2 of /uw/ and /ow/ for all of North America. Vowels before /l/ excluded.a Age * 25 years represents the age coefficient times 25. /uw/ [N=4747] Coefficient Constant

Probability

1547

/ow/ N=6736] Coefficient

Probability

1386

SOCIAL Age * 25

–101

picker-upper > pick-upper. As a multi-word verb is more strongly regarded as a single term, the less need there would be to use a suffix on the verb part of the term at all, giving us the pickupper pattern that appears to be productive today. Perhaps one day pickupper will supplant picker-upper, meaning that the suffix gradually will have moved from the verb to which such suffixes have historically belonged, to the end of the term, where suffixes usually go, with an intermediate stage in which the suffix occurs in both positions. Whether this will happen remains to be seen. But for now it appears that both picker-upper and pick-upper are productive patterns. So even though it means putting a suffix in two places, or even though it means putting a deverbal suffix on a particle, speakers nowadays do not seem to mind – at least in speech. It will be interesting to see how far this development will go.

5.

Conclusion

So the -er nominalization of multi-word verbs has been around for a long time and continues to be used today. Nevertheless it is hard to find in published writing, and examining it has required using unorthodox corpora, namely the OED and the internet. Using these corpora requires care in characterizing the diachronic and synchronic use of these nominalizations, but we can tentatively note several trends. Insofar as the citations from the OED accurately represent historical stages of English, we can first note that the picker-up pattern has historically been the most common nominalization of multi-word verbs, but that its use dropped off in the twentieth century. Second, the by-stander pattern occurred fairly commonly in past stages of English, but declined steadily through the centuries. Third, the picker-upper pattern appears to be a recent innovation, probably of the twentieth century. It is hard to say much about the pick-upper pattern, since it does not occur in the OED; its exclusion could mean that the construction is recent or that it is too unconventional to have shown up in the OED citations. Insofar as the internet type counts accurately represent present-day

280

Don Chapman

usage, we can note that the picker-up pattern continues to be used today, but not as much as the picker-upper pattern, which is the most popular pattern. The pick-upper pattern is robust today, as it occurs as much as the picker-up pattern. Perhaps usage is moving from historically favoring the picker-up pattern to eventually favoring the pick-upper pattern, with the picker-upper pattern occurring as a middle stage.

Appendix 1 Nominalizations from the OED, given by first occurrences of each type.12 1400 picker-up pattern setter-forth bringer-out putter-out rattler-out finder-up holder-up raiser-up 1400 on-looker pattern forth-bringer in-holder up-holder out-putter out-rider 1500 picker-up pattern bearer-about carrier-about groper-after looker-after pothunter-after evidencer-against inveigher-against player-against rebeller-against rebutter-against

biter-at player-at pleader-at shifter-away stealer-away passer-by stander-by writer-down labourer-for mourner-for provider-for pourer-forth departer-from flitter-from dealer-in dweller-in hanger-on looker-on finder-out layer-out rooter-out searcher-out seeker-out shredder-out thruster-out weeder-out bringer-up

lifter-up setter-up taker-up encroacher-upon goer-upon partaker-with player-with 1500 on-looker pattern by-hanger by-walker in-dweller out-cryer out-dweller out-stander up-setter up-taker 1600 picker-up pattern discontinuer searcher-after thirster-after contriver-against disputer-against practicer-against trespasser-against writer-against

Fixer-uppers and Passers-by aimer-at carper-at driver-at nibler-at plodder-at practicer-at scoffer-at snatcher-at waiter-at caller-away seller-away slipper-away drawer-back goer-between interceder-between intervener-between batterer-down knocker-down angler-for fisher-for stander-for suiter-for holder-forth setter-forward putter-forwards dissenter-from distracter-from raiser-from bringer-in delighter-in fetcher-in usherer-in diver-into breaker-off cutter-off shaker-off carrier-on depender-on doter-on

drawer-on fastener-on leader-on player-on puller-on putter-on inquirer-out letter-out pourer-out sifter-out squarer-out stretcher-out striker-out sweater-out thrower-out gatherer-together layer-together worshipper-towards binder-up layer-up puffer-up rooter-up screwer-up shorer-up shutter-up spinner-up stirrer-up treasurer-up worker-up wriggler-up essayer-upon runner-upon speaker-upon complier-with consulter-with converser-with sider-with

281

1600 on-looker pattern by-stander forth-putter in-bringer in-comer in-creeper on-looker on-setter out-comer out-lier over-looker over-ruler up-bearer up-giver up-lifter 1700 picker-up pattern enquirer-after poacher-after trotter-after pursuer-against repiner-against attender-at enterer-at frowner-at gambler-at grasper-at knocker-at stayer-at putter-down thrower-down arguer-for runner-forth eloper-from wanderer-from believer-in whipper-in equirerer-into bearer-off

282

Don Chapman

clawer-off putter-off rubber-off borderer-on creeper-on decider-on layer-on spurrer-on treader-on looker-out layer-over adviser-to setter-to writer-to putter-together filler-up gatherer-up locker-up picker-up stringer-up tider-up tucker-up winder-up intruder-upon invader-upon performer-upon trampler-upon usurper-upon sharer-with trifler-with stayer-within 1700 on-looker pattern down-hauler in-hauler off-putter out-hauler out-setter over-runer

up-striker 1800 picker-up pattern hanger-about idler-about rusher-across burrower-after hunter-after inquirer-after luster-after seeker-after striver-after conspirer-against insurer-against offender-against striver-against journeyer-along trudger-along beginner-at boarder-at flincher-at needler-at refuser-at winner-at worker-at scarer-away taker-away wiper-away bringer-back caster-back snapper-back tapper-back saunterer-by boiler-down cougher-down hander-down hewer-down pusher-down screwer-down

smoother-down softener-down talker-down atoner-for carer-for clamourer-for pleader-for struggler-for sufferer-for putter-forth deliverer-from escaper-from ouster-from sufferer-from breaker-in cutter-in dabbler-in dropper-in feeder-in giver-in licker-in lier-in looker-in mover-in putter-in roper-in shutter-in stroker-in swearer-in taker-in twister-in unbeliever-in leader-into researcher-into caller-off knocker-off leader-off marker-off taker-off

Fixer-uppers and Passers-by warder-off wetter-off bringer-on commenter-on getter-on helper-on holder-on hooker-on hounder-on lingerer-on runner-on stander-on tryer-on walker-on blotter-out caller-out chucker-out cutter-out diner-out piecer-out pusher-out sitter-out wiper-out turner-over listener-to speaker-to teacher-to gluer-together screwer-together builder-up caller-up caster-up chipper-up crumpler-up digger-up drawer-up getter-up helper-up knocker-up

licker-up maker-up patcher-up penner-up plodder-up preacher-up putter-up roller-up runner-up scratcher-up seller-up shover-up springer-up sticker-up stubber-up swallower-up tier-up tosser-up vamper-up wakener-up washer-up writer-up meditater-upon performer-with prevailer-with sporter-with dweller-within lodger-within 1800 on-looker pattern down-comer down-looker down-puller forth-comer forth-speaker in-breather in-gatherer in-layer in-lier

283

in-player off-bearer off-comer off-scourer on-bearer on-hanger out-fitter out-goer out-pourer out-speaker out-sticker over-lier up-shutter up-turner 1900 picker-up pattern gawper-after safener-against flier-at guesser-at hanger-back slower-down quitter-from chipper-in drawer-in listener-in mucker-in sitter-in weigher-in listener-into jacker-off seer-off washer-off coupler-on passer-on researcher-on setter-on worker-on doler-out

284

Don Chapman

hander-out noser-out opter-out stamper-out tailer-out boiler-over looker-round slinker-round muddler-through pointer-towards backer-up cleaner-up drier-up jacker-up lighter-up louser-up mopper-up opener-up pepper-up

ringer-up scuffer-up shaper-up snapper-up sparker-up speeder-up sucker-up tidier-up toucher-up warmer-up whooper-up thinker-upon 1900 picker-upper pattern tearer-downer helper-outer builder-upper looker-upper

maker-upper mucker-upper opener-upper pepper-upper picker-upper tidier-upper waker-upper warmer-upper washer-upper 1900 on-looker pattern forth-teller forward-looker in-setter in-taker on-goer over-setter over-cutter up-converter

Fixer-uppers and Passers-by

Appendix 2 Nominalizations found in Google Nominalization Bring about Come about Go about Hang about Mess about Set about Worry about Come along Go along Get around Go around Hang around Lie around Look around Move around Run around Turn around Walk around Arrive back Bring back Come back Cut back Date back Draw back Drive back Fall back Fight back Fly back Get back Glance back Go back Hit back

Picker-up pattern X

Picker-upper pattern

Pick-upper pattern

X X X

X

X X X X X X

X

X X

X X

X

Bystander pattern

285

286

Don Chapman

Nominalization Hold back Hurry back Lay back Lean back Look back Move back Pay back Pull back Push back Put back Report back Run back Send back Sit back Smile back Stand back Stare back Step back Take back Throw back Trace back Turn back Walk back Welcome back Bend down Bog down Break down Bring down Burn down Calm down Climb down Close down Come down Cut down Drop down Face down

Picker-up pattern

X

X

X

X

X

X X

Picker-upper pattern X X X X X X X X X X X

Pick-upper pattern

X

X

X

X X X X X

X

X X X X X X X X X X

X X X

X X X X

Bystander pattern

Fixer-uppers and Passers-by Nominalization Fall down Get down Glance down Go down Hand down Jump down Kneel down Knock down Lay down Let down Lie down Look down Play down Pull down Put down Run down Set down Settle down Shoot down Shut down Sink down Sit down Slide down Slow down Stand down Stare down Step down Take down Track down Turn down Walk down Write down Break in Bring in Build in Call in

Picker-up pattern X

X

X

X X X X X X X X X

X

Picker-upper pattern X X

Bystander pattern

X X X X X X X X X X X X X X X

X X X X X

X X X X

X X X

X X X

X X X X X X X X X

X

X

X X

Pick-upper pattern X

X X X X X

287

288

Don Chapman

Nominalization Cash in Climb in Close in Come in Fall in Fill in Fit in Get in Give in Go in Join in Live in Move in Put in Send in Step in Stir in Swear in Take in Throw in Walk in Break off Call off Come off Cut off Drive off Fall off Finish off Get off Go off Kill off Make off Move off Pay off Pull off Put off

Picker-up pattern X

X X X X X X X X

Picker-upper pattern X X X X X X X X X X X X X

X X X X

Pick-upper pattern X

X X X

X X X

X X X X X

X X

X

X

X

X

X

X X X X X X X

X X X

X

Bystander pattern

Fixer-uppers and Passers-by Nominalization Run off Sell off Send off Set off Shake off Show off Spark off Start off Switch off Take off Trail off Turn off Walk off Ward off Write off Bring on Carry on Come on Follow on Get on Go on Hang on Hold on Keep on Live on Move on Pass on Press on Put on Stay on Switch on Take on Walk on Wear on Work on Bear out

Picker-up pattern X

X

Picker-upper pattern X X

Pick-upper pattern X

X X X

X

X

X X X

X X X

X X X

X

X

X X

Bystander pattern

X

X X

X X X

X X X

X X X X X X X

X X X

X X X X X X X X X

X

X

289

290

Don Chapman

Nominalization Blow out Blurt out Break out Bring out Burn out Burst out Call out Carry out Catch out Check out Climb out Come out Cry out Cut out Draw out Dry out Fall out Figure out Fill out Find out Fly out Get out Give out Go out Hand out Hang out Help out Hold out Jump out Keep out Knock out Lash out Lay out Left out Let out Look out

Picker-up pattern

X

X

Picker-upper pattern X X X X X X X X X X X

Pick-upper pattern

X X X

X X X

X X X X X X X

X X X

X X

X

X X

X

X X X X X X X X X X X X

X X

X X

Bystander pattern

Fixer-uppers and Passers-by Nominalization Lose out Make out Map out Mark out Meet out Miss out Move out Opt out Pay out Phase out Pick out Play out Point out Pour out Pull out Put out Reach out Read out Roll out Rule out Run out Seek out Sell out Send out Set out Share out Shoot out Single out Slip out Sort out Speak out Spell out Spread out Stand out Stare out Start out

Picker-up pattern X X

Picker-upper pattern X X

X

X

X

X X X X

Pick-upper pattern X X

X

X X

X

X X

X X

X X X X X X X X X X X X X

X X

X X

X X X X X

X X

X X X X X X X

X

X

Bystander pattern

291

292

Don Chapman

Nominalization Stay out Step out Stick out Stretch out Strike out Take out Throw out Try out Turn out Walk out Watch out Wear out Wipe out Work out Write out Come over Fall over Give over Go over Hand over Lean over Leave over Look over Pass over Roll over Run over Take over Turn over Voice over Walk over Come round Get round Glance round Go round Look round Spin round

Picker-up pattern

X X X X

X

Picker-upper pattern X X X X X X X X X X X X X X

Pick-upper pattern X

Bystander pattern

X X

X X X X

X

X X

X X X X

X X X X

X

Fixer-uppers and Passers-by Nominalization Swing round Turn round Come through Get through Go through Add up Back up Beat up Bind up Blow up Break up Bring up Build up Call up Catch up Clean up Clear up Climb up Come up Conjure up Cover up Curl up Divide up Draw up Dress up Dry up End up Fill up Finish up Follow up Gaze up Get up Give up Glance up Go up Grow up

Picker-up pattern

Picker-upper pattern

X X X

X

X

X X X X X X X X X X X X X X X X X X X X X X X X X

X

X

X

X X X X X

Pick-upper pattern X

X X X X

X X X X X X X X X X X X X

X X

X X

X X

X

Bystander pattern

293

294

Don Chapman

Nominalization Hang up Hold up Hurry up Join up Jump up Keep up Leap up Light up Line up Link up Lock up Look up Make up Meet up Mix up Move up Open up Pack up Pay up Pick up Pile up Prop up Pull up Put up Reach up Ring up Rise up Roll up Run up Screw up Set up Show up Shut up Sign up Sit up Snap up

Picker-up pattern X

X X

X X

X

X X X

X X

Picker-upper pattern X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X

Pick-upper pattern X X X X

X X

X X X

X

X X X X X X X X X X X X X X X

Bystander pattern

Fixer-uppers and Passers-by Nominalization Speak up Speed up Split up Spring up Stand up Stare up Start up Stay up Step up Stir up Straighten up Sum up Swallow up Take up Team up Throw up Tidy up Tie up Total up Turn up Use up Wake up Walk up Warm up Wash up Weigh up Wind up Wrap up

Picker-up pattern X X X

X

X

Picker-upper pattern X X X X X X X X X X

X

X

X X X X X X

X X X X X X X X X X

Pick-upper pattern X X X X X X X X X X X X X X X X X X X X

X X

X X

Bystander pattern

295

296

Don Chapman

Notes 1. Though sometimes -er has created nominalizations that are patients or fulfill other roles, such as keeper, meaning a fish that is big enough to keep when fishing. Cf. Ryder 2000: 295. 2. This construction has been given many names, such as phrasal verbs, multiword verbs, verb-particle combinations, etc. This paper will use the terminology from Biber et al. 1999. 3. Besides, these patterns do not represent the same kind of nominalizations of the verb-particle construction. For picker (1: 4), the original verb-particle construction will seldom be recoverable. Perhaps listeners can relate listener to listen to or waiter to wait on, but can turn on be recovered from turner or get out from getter? Such relations would have to become entirely conventionalized, as waiter < wait on has. For drop-in (1: 6), the range of meanings is much broader than forms with the -er suffix. A drop-in can be personal (e.g. someone who drops in), but it is more likely to be impersonal (e.g. the action of dropping in, an occurrence of someone dropping in) (Sørensen 1986: 278–79). 4. Finding a good cover term for these different kinds of nominalizations is a problem. Word-formation studies often speak of these four different nominalizations as being different word-formation types. But since this study analyzes the occurrence of these nominalizations in corpora, it will be necessary to use the sense of type (as opposed to token) from corpus studies. To avoid confusing these two senses of type, I refer to the word-formation types as wordformation patterns, of course realizing that pattern can also mean different things in different studies. 5. Štekauer et al. 2005: 50. 6. It was, in fact, this question that led me to this topic. Pavol Štekauer, a native Slovakian speaker, had designed the first draft of this question without any of the double suffixes. As I collaborated on the draft, I told him that we needed to include butter-inner and cutter-inner. He expressed some surprise and incredulity at such a double-marked construction, and I could shed little light on the construction, since that was the first time that I had really noticed it. But at least the respondents agreed with my intuitions. 7. I used regular expression searches to find the various -er patterns for each corpus, and then inspected the over-generated list to find instances of multiword verb nominalizations. It is possible I could have inadvertently excluded some nominalizations, but not many. 8. Normalizing the data is not completely straightforward. For this study, the normalized figures represent the number of new types that occur per million words in each century. The total size of the corpus for each century represents the number of words in the citations, not the number of head-word entries. Normalizing per X (e.g. million) words is typical in corpus studies, but is used

Fixer-uppers and Passers-by

9.

10.

11. 12.

297

here for less typical reasons. In a typical corpus study, we assume that the instances that we are studying occur at a constant rate. The rate, however, that new words will appear for the first time in a corpus is not constant, but instead logarithmic (cf. Baayen 2001). For a typical corpus, a normalized distribution would be misleading for analyzing the occurrence of new types. But the OED is not a typical corpus, and we are not measuring the likelihood that a new word (type) will appear in the corpus as the number of words increases. Instead we are measuring the likelihood that the OED editors will record a new word, as the number of words increases. The rate of recording new words will be constant if the occurrences of the nominalizations are largely limited to citations illustrating the nominalization as a head-word (instead of occurring in citations illustrating some other word), and if the number and size of citations is the same from entry to entry. The first condition is so, and if we assume the second, we are justified in using a words-per-million normalization for measuring the rate at which new words show up in the OED. In short, it is because the OED adds citations at about the same rate as it adds new head-words that we can assume that the number of new words will increase as the number of total words increases. Of course many other multi-word verbs undoubtedly occur with some kind of separation between the verb and the particle. Such have not been included in the internet searches. Admittedly, this procedure works better for the picker-upper and bystander patterns, where the percentage of true positives is relatively high in all rounds, but even for the picker-up and pick-upper patterns the low number of additional types in rounds four and ten suggests that these patterns probably would not produce many new types in the other rounds available, but unexamined (rounds 5–9). The phrasal particle over does not, but that may owe to the phonological shape of over-speakers resist putting two identical sounds together. The nominalizations are spelled with hyphens in this appendix simply for the sake of uniformity in presentation. They certainly did not all occur in the OED spelled with a hyphen.

298

Don Chapman

References Ackema, Peter, and Ad Neeleman 2000 M-selection and phrasal affixation. In UCL Working Papers in Linguistics 12. UCL. C. Iten and A. Neeleman (eds.), 307–342. Adams, Valerie 1973 An Introduction to Modern English Word-formation. London: Longman. Andersen, Gisle 2002 Corpora and the double copula. In From the COLT’s Mouth … and others’. Leiv Egil Breivik and Angela Hasselgren (eds.), 43–58. Amsterdam, Rodopi. Bauer, Laurie 1983 English Word-Formation. Cambridge: Cambridge University Press. Baayen, R. Harald 1992 On frequency, transparency and productivity. Yearbook of Morphology: 181–208. 2001 Word Frequency Distributions. Dordrecht: Kluwer. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan 1999 The Longman Grammar of Spoken and Written English. London: Longman. Brewer, Charlotte 2000 OED Sources. In Lexicography and the OED. Lynda Mugglestone (ed.), 40–58. Oxford: Oxford University Press. Brinton, Laurel J. 1988 The Development of English Aspectual Systems: Aspectualizers and post-verbal particles. Cambridge: Cambridge University Press. Brinton, Laurel J. and Elizabeth Closs Traugott 2005 Lexicalization and Language Change. Cambridge: Cambridge University Press. Claridge, Claudia 2000 Multi-word Verbs in Early Modern English: A Corpus-based Study. Amsterdam: Rodopi. Clark, Eve V., Barbara Frant Hecht, and Randa C. Mulford 1986 Coining complex compounds in English: Affixes and word order in acquisition. Linguistics 24: 7–27. Chapman, Don and Royal Skousen 2005 Analogical modeling and morphological change: The case of the English adjective negative prefix. English Language and Linguistics. 9: 333–357.

Fixer-uppers and Passers-by

299

Fraser, Bruce 1976 The Verb-Particle Combination in English. New York: Academic Press. Jespersen, Otto 1942 A Modern English grammar on Historical Principles. Part VI: Morphology. Heidelberg: Carl Winter. Kwon, H.-S. 1997 Negative prefixation from 1300 to 1800: A case study in in-/unvariation. ICAME Journal 21: 21–42. Lindelöf, Uno 1938 English verb-adverb groups converted into nouns. Commentationes Humanarum Litterarum 9. 1–41. Marchand, Hans 1969 Categories and Types of Present-day English Word-Formation. 2nd ed. Munich: C. H. Beck’sche. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik 1985 A Comprehensive Grammar of the English Language. London: Longman. Ryder, Mary Ellen 2000 Complex -er nominals: Where grammaticalization and lexicalization meet? In Between Grammar and Lexicon. Ellen Contini-Morava and Yishai Tobin (eds.), 291–331. Amsterdam: Benjamins. Sørensen, Knud 1986 Phrasal verb into noun. Neuphilologische Mitteilungen 1986 (87): 272–283. Štekauer, Pavol, Don Chapman, Slavka Tomascikova, and Stefan Franko 2005 Word-formation as creativity within productivity constraints: Sociolinguistic evidence. Onomasiology Online 6: 1–55. Véronis, Jean 2005 Web: Google’s counts faked? Technologies du Langage. Weblog. http://aixtal.blogspot.com/2005/01/web-googles-counts-faked.html.

Words and constructions in grammaticalization: The end of the English impersonal construction*

Graeme Trousdale 1.

Introduction

This article provides a new analysis of the loss of a construction which has been the topic of frequent discussion in work on English historical syntax (for instance Elmer 1981, Fischer and van der Leek 1983, 1987, Denison 1990, 1993, Allen 1995, and Anderson 1997a), namely the impersonal construction. The example under (1) below is often cited to provide evidence of the impersonal construction in Old English (OE): (1) him ofhreow þæs mannes 3SM-dat pity-3SPast the-gen man-gen to-him was-pity because-of-the-man ‘He pitied the man’ Or: ‘The man caused pity in him’ (ÆCHom I XIII.281.12) In the development of English, the impersonal construction disappears; processes which historically were coded in such constructions come to be construed personally. Examples such as (1) are described in the literature as ‘impersonal’ since they have the following morphosyntactic properties: within the clause there is no nominative argument, and the verbal inflection is third person singular, irrespective of the number of either argument. The present aim is to outline aspects of the history of the impersonal construction in English to show how the demise of the construction is associated with the emergence, via a process of grammaticalization, of the transitive construction (TrnCxn, on which see further section 2 below), and to show how constructional taxonomies are implicated in such grammaticalization processes. Grammaticalization has been defined as “the change whereby lexical items and constructions come in certain linguistic contexts to serve grammatical functions and, once grammaticalized, continue to

302

Graeme Trousdale

develop new grammatical functions” (Hopper and Traugott 2003: 18). It is this second stage of grammaticalization on which I intend to focus here: specifically, I argue that (grammatical) constructions may be subject to further grammaticalization as they are used in different ways by speakers to code events in particular ways. Such grammaticalization may not result in new grammatical functions, but may instead involve the emergence of more schematic constructions, which in turn sanction instances which deviate yet further from a prototype instance of that construction. The place of constructions in grammaticalization is considered with reference both to early work on grammaticalization by Meillet (1958 [1912]), and later work more explicitly concerned with Radical Construction Grammar (e.g. Croft 2001, Traugott 2006, forthcoming a, b). The main thesis developed in this article is that constructions do not necessarily solely constitute the context within which lexemes can be grammaticalized; they may also be subject to grammaticalization themselves. The development of the TrnCxn is an instance of category strengthening, a phenomenon which is well attested in the history of English syntax, in the development of the tense auxiliary do, and the modals (as argued by Hudson 1997), and this claim is supported by corpus work on transitivity in present-day American English (Thompson and Hopper 2001). The structure of the remainder of this paper is as follow. Section 2 provides the general framework for a constructional account of grammaticalization, while section 3 provides a discussion of relevant aspects of the history of the impersonal construction, up to and including the early Modern period. Section 4 considers aspects of transitivity and subjecthood in Modern English. The central thesis, developed in section 5, is that the loss of the impersonal construction is tied in with the increased productivity and schematicity of the transitive construction, and that this constitutes an instance of grammaticalization.

2.

Constructions and grammaticalization

In this paper, I adopt the model of a linguistic construction as presented in Croft (2001). Constructions are symbolic form-meaning pairings (see figure 1); constructions, and parts of constructions, form a network, in which elements are related by taxonomies. These taxonomic links connect both constructions and parts of constructions (see figure 2).

Words and constructions in grammaticalization

syntactic properties

303

CONSTRUCTION

morphological properties FORM

phonological properties Symbolic correspondence (link)

semantic properties pragmatic properties

(CONVENTIONAL) MEANING

discourse-functional properties

Figure 1. The symbolic structure of a construction (Croft 2001: 18)

SbjArg

IntrSbj

Kick1

Pred

IntrVerb

kick

TrSbj

Kick2

TrVerb

Kiss1

kiss

TrObj

Kiss2

Figure 2. A partial constructional taxonomy (Croft 2001: 56)

This constructional taxonomy is highly relevant to grammaticalization. Grammaticalization has traditionally been seen to affect lexical items within constructions, but not constructions themselves. In fact, ‘construction’, in traditional accounts of grammaticalization, is often simply another

304

Graeme Trousdale

term for the context for grammaticalizing lexemes. The word ‘construction’ in such accounts does not mean a form-meaning pairing; it means the other members of the string of words in which the grammaticalizing form appears. However, as has been reported elsewhere in the constructional literature, lexical items are constructions, in the technical Radical Construction Grammar sense, since constructions operate on, or are defined at, all levels of grammatical structure; this is echoed in another variant of Construction Grammar when Goldberg (2006: 18) writes “it’s constructions all the way down” (emphasis original). What distinguishes a lexical construction from a syntactic construction is that a lexical construction is both substantive and atomic (a syntactic unit, fully specified in both form and meaning) while a syntactic construction is both schematic and complex (a combination of syntactic units in which form and meaning are less specific, see Croft 2001: 16–17). Given the parallel between lexical and syntactic constructions, therefore, we might expect syntactic constructions also to undergo grammaticalization (particularly secondary grammaticalization, as defined by Hopper and Traugott 2003: 18, as noted in the introduction). As lexical constructions grammaticalize, they become more schematic, in both form and meaning, and develop new grammatical functions. We would therefore expect syntactic constructions to behave in a similar way. Traugott (2006, forthcoming a, b) has shown that NP of NP constructions (such as a bit/lot of) have grammaticalized as they developed from Partitive Constructions to Degree Modifier Constructions. The question to be addressed here is the extent to which argument structure constructions can also legitimately be said to grammaticalize. In what follows, the notion of constructional prototype is also relevant, and this too ties in with the idea of a constructional taxonomy outlined above. Each level in the constructional taxonomy is sanctioned by a more schematic construction at a higher level. Thus in Figure 2, the [SubjArg]-[Pred] construction sanctions the Intransitive Construction; similarly, at a lesser degree of schematicity, the Transitive Construction sanctions the [Kiss1][kiss]-[Kiss 2] construction. At all levels in the taxonomy, within each category, not all instances will be of the same status; some instances come closer to the conceptual core of the category, while others represent extensions from that prototype. The more an instance deviates from the prototype, the more likely it is that the instance may be recategorised (Labov 1973). Since constructions display prototype effects, we might expect examples whereby non-prototypical constructions (Cx) of category A share

Words and constructions in grammaticalization

305

features with non-prototypical constructions (Cy) of category B; and Cx and Cy may indeed be ambiguous in their category membership. Having outlined features of constructional taxonomies, I now review aspects of the history of the impersonal construction in English.

3.

Aspects1 of the history of the English impersonal construction

3.1. The situation in Old English Example (1) in the introduction – repeated here as (3) for convenience – represents a prototypical instance of the OE impersonal construction: (3) Him ofhreow þæs mannes 3SM-dat pity-3SPast the-gen man-gen to-him was-pity because-of-the-man ‘He pitied the man’ Or: ’The man caused pity in him’ (ÆCHom I XIII.281.12) Many impersonal constructions of this type form a subgroup of a more general category of experiencer constructions (see Allen 1995) in which one of the arguments is the experiencer of some sort of psychological state or process. An instance of an experiencer construction (hereafter ExpCxn) which is not an instance of the impersonal construction is given as (4): (4) hu him se sige gelicade how 3SM-dat the-nom victory-nom please-3Spast how to-him the victory pleased ‘how the victory pleased him’ Or: ’how the victory caused pleasure in him’ (Orosius 156. 25) Here although there is an experiencer (him), the presence of a nominativemarked nominal encoding the Source of the pleasure, means this is not an instance of the impersonal construction. Nonetheless, it is common to relate the impersonal construction to other ExpCxns which have nominative arguments. Such constructions appear under a number of different classificatory labels in, for example, Allen (1995), Denison (1993), Elmer (1981),

306

Graeme Trousdale

and Fischer and van der Leek (1983). I follow Elmer and Allen in their classification, which can be summarised as follows under (5): (5) 2-NP Experiencer constructions in Old English Type N Case of Experiencer: Dative or Accusative Case of Source: Genitive Example: him ofhreow þæs mannes 3SM-dat pity-3SPast the-gen man-gen to-him pitied because-of-the-man ‘He pitied the man’ Or: ‘The man caused pity in him’ (ÆCHom I XIII.281.12) Type I Case of Experiencer: Dative Case of Source: Nominative Example: him ne ofhreow na þæs deofles hryre 3SM-dat not pity-3SPast not the-gen devil-gen fall to-him not pitied not the devil’s fall ‘He did not pity the devil’s fall’ Or: ‘The devil’s fall did not bring about pity in him’ (ÆCHom I XIII.281.14) Type II Case of Experiencer: Nominative Case of Source: Genitive ofhreow Example: se mæssepreost þæs mannes the-nom priest-nom the-gen man-gen pity-3SPast the priest because-of-the-man pitied The priest pitied the man Or: ‘The priest felt pity because of the man’ (COE) ÆLS (Oswald 262) The same verb can appear in more than one ‘type’ of ExpCxn in Old English. We can see this in (3), in that ofhreowan appears in Type N, Type I and Type II constructions. Allen (1995: 85) has helpfully collated these variations in one table, which I give in modified form as Table 1 under (6):

Words and constructions in grammaticalization

307

(6) Table 1. Variation of the 2-NP constructions with the Old English experiencer verbs (Allen 1995: 85)

I only

N only

II only

I and N

II and N

All

losian

lystan

behofian

ofþyncan

sceamian

ofhreowan

gelician

langian

wilnian

þyncan

tweonian

mislician

giernian

hreowan

oflician

reccan

lician eglian gehreowan laþian Another type of construction to be considered is exemplified by the clause in (7): (7) he acwealde þone dracan 3MS-nom kill-1/3SPast the-acc dragon-acc ‘he killed the dragon’ (Ælfric Homilies (Supp.), XXI, 455) Here the NP functioning as agent is case-marked as nominative, and the NP functioning as patient is case-marked as accusative. The verb denotes an action involving the transfer of physical energy. Since case marking on nominals in any given clause in OE “is correlated at least in part with the perspective taken on the state of affairs described” (Traugott 1992: 211), I consider such case marking to be an instance of grammatical organization symbolizing aspects of conceptual content. I classify the example in (7) as a prototypical instance of another type of construction, Type T; prototype instances of Type T involve action verbs. We are now able to construct a partial taxonomy for these 2-NP constructions in OE. This is given in Figure 32. I take Type N constructions to be the prototypical instances of the more schematic ExpCxn category (illustrated in Figure 3 by the emboldened box around the Type N construction), while Type I and Type II are extensions sanctioned by the ExpCxn schema. Type N is maximally distinct from

308

Graeme Trousdale NP1

ExpNP

ExpV erb

Pred

NP2

SouNP

AgNP

AcVerb

PatNP

TYPE T Exp Dat

Exp Verb

TYPE N

Sou Gen

Exp Dat

Exp Verb

Sou Nom

TYPE I

Exp Nom

Exp Verb

Sou Gen

TYPE II

Figure 3. A partial constructional taxonomy for 2-NP construction in OE

Type T in terms of verbal morphology (lack of agreement) and case marking. The extensions are non-prototypical members of the category, since the NPs have different case markings; such case marking shows an overlap with members of the Type T category3. The highest construction in this partial taxonomy is what I will call the Transitive Construction (TrnCxn). It is this superordinate level construction which ultimately sanctions all of the others, and is schematic in both form and meaning. We have seen above how variation in case marking indicates the formal variation in subtypes of the TrnCxn; I turn now to other ways in which the various subtypes of the TrnCxn vary, using some of the parameters for transitivity as defined by Hopper and Thompson (1980), given in (8) below. (8) Table 2. Some parameters for transitivity (adapted from Hopper and Thompson 1980: 252)

KINESIS ASPECT PUNCTUALITY VOLITIONALITY AGENCY AFFECTEDNESS OF O

HIGH action telic punctual volitional A4 high in potency O totally affected

LOW non-action atelic non-punctual non-volitional A low in potency O not affected

Words and constructions in grammaticalization

309

Such parameters allow us to identify significant differences between instances of Type T and Type N in OE. For example, prototypical instances of Type T are telic, volitional and denote action; prototypical instances of Type N are atelic, non-volitional and often denote states rather than actions. 3.2. The later history of the impersonals Patterns relating to the late history of the impersonal construction have been established by the account provided by Allen (1995), in her detailed history of the experiencer verbs as a whole. Allen provides evidence of the increasing subject assignment for the Experiencer, an analysis based on the syntactic criteria that Preposed Dative Experiencers (PDEs)5, and Experiencers which have nominative inflection if pronominal, should be classified as subjects, though the number of examples in the corpora used in standard accounts are small. Thus rewen6, which never appears in Type II constructions in OE, could have an unambiguous nominative subject experiencer in Middle English, as in (9): (9) We schold rew þat sore (?1325 Swet Iesus v2, l6) The general pattern then is for Experiencers rather than Sources to become coded as subject. While this pattern is true generally, the development was a very gradual one, and we can notice what seem to be different trajectories for individual lexemes, such as ail and behoove. Allen (1995: 271) notes that both ail and behoove resist nominative experiencers: “we find examples of PDEs in the late fifteenth century in texts which contain no examples with nominative experiencers.” The construction in which ail appears is largely fixed, a reported question of the type what DAT ail. Such fixed expressions are not characteristic of behove, witness (10): (10) Hym behoued non other lighte (Caxton, 77.13) which, by the seventeenth century, had evolved into the non-impersonal (11): (11) those Persons that rob me, are not fit to be Evidences against me, because it behoves them that I be convicted, to prevent their being indicted for Felony. (Lisle)

310

Graeme Trousdale

Similarly, in the sixteenth century, there is evidence of the continuation of Type I constructions with PDEs, in the writings of Sir Thomas More, exemplified by examples (12) through (14): (12) Thou shalte labour to thy power by juste and trewe busynesse, to geate that thee and thyne behoueth (More 1557 [c.1522] 90) (13) And sone after one hower betwene .x. & .xi. he returned into ye chamber among them, al changed with a wonderful soure angrye countenaunce, knitting the browes, frowning and froting and knawing on hys lippes, and so sat him downe, in hys place: al the lordes much dismaied & sore merueiling of this maner of sodain chaunge, and what thing should him aile (More 1513, 47) (14) And for asmuch to her seemed the Cardinall more redy to depart (More 1513, 40) Finally, the examples in (15) show that variation with like continued well into the seventeenth century. (15) a. these two, trauelinge into east kent, resorted vnto an ale house there, being weried with traueling, saluting with short curtisey, when they came into the house, such as thei sawe sitting there, in whiche company was the parson of the parish; and callinge for a pot of the best ale, sat down at the tables ende: the lykor liked them so well, that they had pot vpon pot (Harman 1567) b. her name is Buckle, a Sharpsheare woman: if you like of it, I would thinke of haueing of her; for I haue no body aboute me (Harley 1633) c. That in the mean tyme yt wyll lyke ye to wryte to my lord how carefull and myndfull I have byn of him, shall doe me a great pleassur (Dudley 1585) d. yf my cosin like it, I will send him more. (Cornwallis 1664) What we see, therefore, is an interplay between lexical variation and constructional variation – some of the early Modern data show specific behaviour of particular tokens (e.g. ail, behoove), and some show behaviour of

Words and constructions in grammaticalization

311

particular types (e.g. Type I constructions). This issue has been noted by Allen: The reader might reasonably object that in saying that PDEs lastest [sic] longest in particular constructions because they had always (or at least, for some time) been most common in those constructions is begging the question of why they were more popular in some constructions than in others. (Allen 1995: 282–283, emphasis added)

The question, then, is why some constructional variants of the ExpCxn are more resistant to change than others. The development, and the subsequent loss of PDEs, is usually explained as a change from lexical to structural case marking: “[t]he disappearance of the PDE as a structural possibility is best seen as a gradual disfavouring of the option of case-marking subjects lexically” (Allen 1995: 347). However, I propose an alternative account, based on grammaticalization and constructional taxonomies as outlined in section 27. Following Bybee, Perkins and Pagliuca (1994), Langacker (2005) and Traugott (forthcoming b) on constructional change generally, I argue that the TrnCxn in the history of English has become more productive (sanctioning a wider range of subject types, for instance), more general (encoding a wider range of thematic relations between the verb and its arguments), and less compositional (since some variants of the TrnCxn have become less clearly telic, for instance). Increased productivity, increased generalisation, and decreased compositionality, taken together, are evidence of the TrnCxn having grammaticalized over time. This structuralization (on which see further Thompson and Hopper 2001, and section 4 below) is related to the strengthening of the TrnCxn as a superordinate category in English. Evidence from this comes in part from an analysis of the TrnCxn in present-day English, which is discussed in the next section.

4.

Transitivity and subjecthood in Modern English

Like the impersonal construction, transitivity in English has also been the subject of a number of analyses. In an article cited earlier (see (8) above), Hopper and Thompson (1980) attempted to provide a set of attributes which characterised clauses in terms of high and low transitivity. In a more recent paper, Thompson and Hopper (2001) reconsider the concept of transitivity in light of evidence from spoken corpora. They suggest that clauses which

312

Graeme Trousdale

are canonically ‘transitive’ – i.e. telic, punctual, volitional on the part of the agent, and so on – are surprisingly infrequent in naturally-occurring conversation, and also that the findings of the study support the claim, made by Hopper (1987: 150), that “the more useful a construction is, the more it will tend to become structuralized, in the sense of achieving cross-textual consistency, and serving as a basis for variation and extension”. Their data suggest that clauses with two participants were not only rare in themselves (most of the clauses had only one participant), but also very low in transitivity: for instance, the clauses were typically atelic (86 %); non-punctual (98%); and had a non-affected object (84 %). This phenomenon appears to be true not only of English: Thompson and Hopper also cite other studies which suggest that this occurs cross-linguistically. Thompson and Hopper also discuss predicate meanings with reference to Roland and Jurafsky (2002), where they state that: “the sense of a verb or predicate is related to the grammatical schemas that it can occur in” (Thompson and Hopper 2001: 44), suggesting that such meanings: can only be understood as including a vast range of semantic and pragmatic associations regarding the sorts of activities, states, and participants that can invoke their use … these ‘meanings’ are actually generalizations from many repetitions of hearing predicates used in association with certain types of human events and situations over the course of a person’s lifetime. What appears to be a fixed ‘structure’ is actually a set of schemas, some more ‘entrenched’ (Bybee 1985, 1998; Langacker 1987) than others, arising out of many repetitions in daily conversational interactions. (Thompson and Hopper 2001: 47).

This view also accords with that proposed in construction grammars of various kinds (e.g. Goldberg 1995, 2006), where the claim is that usagebased grammar is essentially organised around meaning. The generalisations or extensions that are witnessed typically relate high type frequency verbs (such as verba dicendi) to high type frequency constructions (such as the ditransitive). Thompson and Hopper also suggest that verbs of high frequency (get, think, see, want and mean, for instance) have argument structures which are difficult to categorize – they tend to appear more frequently in fixed (lexicalised) expressions – while verbs of low frequency (such as spray, swarm, or load) have a relatively fixed argument structure. Tomasello (1998: xviii) has explicitly linked this to a prototype analysis of transitivity in English, as follows:

Words and constructions in grammaticalization

313

All constructions, whether composed of one word, or of many words in specific orders with specific markers and intonations, and at whatever level of abstraction, derive from recurrent events or types of events, with respect to which the people of a culture have recurrent communicative goals. Because they are formed this way historically, they exhibit prototypical structure just like other cognitive categories and schemas. Thus in English, the basic transitive construction has as its prototype utterances such as He broke the vase in which an animate actor does something to cause a change of state in an undergoer (Hopper and Thompon 1980). But the construction over historical time has been extended to other, less prototypical situations in which the ‘force dynamics’ are not so clear or are only metaphorical, as in, for example, John entered the room and The car cost $400. In German … the transitive construction has stayed much closer to the prototype.

And this is illustrated by Taylor (1998: 188) with examples such as those in (16): (16) a. The hotel forbids dogs b. The tent sleeps six c. *Das Hotel verbietet Hunde d. *Das Zelt schläft sechs Constructions which do not instantiate the prototype of the category are not necessarily any ‘less grammatical’ than those which do; rather, the transitive construction is organised as a network (cf. the earlier discussion of constructional taxonomies), with certain properties of the more central members of the network displaying grammatical properties which do not characterise those at the periphery (i.e. those extensions which, while sanctioned by the schema, may display properties of other categories). Such a network analysis is applicable too to the category of subject. Taylor (1998) gives examples of the kinds of constituents which can function as subjects, in addition to typical noun phrases, as in (17): (17) PP subject: Over there is fine Clausal subject: That Horace should eat meat is astonishing Existential subject: There are six people in that car All are clearly morphosyntactic subjects, since they can undergo raising, as in (18):

314

Graeme Trousdale

(18) Over there seems to be fine and have a regular tag question formation, as in (19): (19) That Horace should eat meat is astonishing, isn’t it? But the use of such subjects is constrained (see Anderson 1997a, b). Locative PPs as subjects of passives can be problematic, witness the dubious grammaticality of (20): (20) ?Over there was sniffed by the dog Clausal subjects do not appear in interrogatives, as illustrated by the ungrammaticality of (21): (21) *Is that Horace should eat meat astonishing? And the syntax of there as a subject (as Lakoff 1987 has demonstrated) shows various odd patterns of behaviour as exemplified by the clauses in (22): (22) There came a dense fog over Dalkeith *There left a delegation for Dalkeith Such patterns can also be accounted for in terms of constructional taxonomies. This is illustrated in Figure 4. Just as we witnessed extensions from the prototype of the ExpCxn in OE (see Figure 3), so in Modern English we see extensions sanctioned by the TrnCxn, to allow for instances such as This tent sleeps six. These must be considered as extensions because they do not display grammatical properties associated with central transitives (e.g. no passive variant *Six are slept by this tent, see Taylor 1998). The next section therefore considers two related issues: how did the ExpCxn come to be coded ‘personally’ rather than impersonally; and how are we to understand the parallel between extensions to the ExpCxn which are attested in OE on the one hand, and extensions to the TrnCxn which are attested in Modern English on the other? I argue that a grammaticalization account explains both these issues.

Words and constructions in grammaticalization Subj

TrnSubj

TrnVerb

Pred

315

Comp

TrnObj

C op s ub j

be

Predicative

e.g. Over there is fine Exp Subj

Trn Verb

Sou Obj

e.g. Bob likes Sue

Agent Subj

Trn Verb

e.g. He killed the dragon

Pat Obj

Locative Subj

Trn Verb

Theme Obj

e.g. This tent sleeps six

Figure 4. A partial constructional taxonomy illustrating subjects and transitivity in Modern English

5.

A constructional account of grammaticalization

There is a particular phrase in Meillet’s 1912 article which is very frequently cited as a definition of grammaticalization, namely that grammaticalization is: (23) l’attribution du caractère grammatical à un mot jadis autonome (Meillet 1958 [1912]: 131) [the attribution of grammatical character to a word which once was autonomous (my translation)] But as Traugott (2006, forthcoming a, b) has noted, Meillet allowed for the possibility that grammaticalization affected not only lexical items per se, but groups of lexical items: (24) Les mots ne sont du reste pas seuls à être sujets à devenir des éléments grammaticaux; la façon de grouper les mots peut aussi devenir un procédé d’expression grammaticale (Meillet 1958 [1912]: 147) [Words are not the only things which can become grammatical elements. The manner of grouping words may also become a pathway for grammatical expressions (my translation)]

316

Graeme Trousdale

His topic was specifically the development of certain word order patterns in the evolution of Late Latin into some of the Romance languages. But the critical issue is that of “la façon de grouper les mots” – the development of groups of words into grammatical items lends itself to a constructional account of linguistic structure. Indeed, Meillet suggests that, in terms of the actual processes which characterise grammaticalization, the behaviour of lexical items on the one hand, and groups of such items on the other, is the same: (25) “Le phénomène est de même ordre que la ‘grammaticalisation’ de tel ou tel mot; au lieu que ce soit un mot employé en groupe avec d’autres qui prenne le caractère de ‘morphème’ par un effet d’habitude, c’est une manière de grouper les mots” (Meillet 1958 [1912]: 148) [The phenomenon is of the same order as the grammaticalization of this or that word; rather than a word, working in concord with others, which takes one the character of a morpheme through habitual use, it is a way of grouping words (my translation)] Such a view predates, but accords with, cognitive and constructional accounts of grammaticalization. For instance, Bybee (2003: 153) defines one aspect of grammaticalization as the “process of automatization of frequently occurring sequences of linguistic elements”, or in Cognitive Grammar terms, grammaticalization is increased entrenchment (cf. Langacker 1987, 1991). But what, precisely, is entrenched? It is not the instances of the schema (which we expect in low-type, high-token cases), but rather the schema itself. Given that such schemas typically form superordinate categories, we might expect to find, following Croft and Cruse (2004) in their discussion of features that distinguish basic level from superordinate categories, that the transitive construction will show a high degree of differentiation in comparison with other schematic, superordinate categories (such as the copula construction illustrated in Figure 4 above) but a low degree of mutual similarity between members of that category, which is indeed the case in Modern English. For instance, while the copula allows non-NP subjects (Over there is fine), few8 verbs in the TrnCxn do (mutual distinctiveness with respect to other superordinate categories); on the other hand, not all instances of the TrnCxn allow passives (low degree of mutual similarity between members of the category). How has this come about? I suggest that this is one of the consequences of grammaticalization – that as a construction type grammaticalizes, it becomes more schematic, and thereby

Words and constructions in grammaticalization

317

automated. On the relationship between linguistic constructions and other cognitive processes, Tomasello (1998: xvii) argues that the schematic constructions are typical of cognitive processes in general; they are “relatively automatized procedures that operate on a categorical level”. Moreover, these constructions, schematic though they may be, are not ‘meaningless’: “each of these abstract linguistic schemas has a meaning of its own, in relative independence of the lexical items involved” (ibid). This links into language as a creative process, where speakers extend their use of constructions by inserting different words into that frame. They form part of the “inventory of symbolic resources that language users control” (Tomasello 1998: xviii). This too is expected if we assume that constructions which are abstract or schematic are not unique to language, but share properties with schemas in other cognitive domains. It is then necessary to ask ‘how schematic is schematic’? Schematicity, just like entrenchment, is a gradient phenomenon: Langacker (1987, 1991) has repeatedly suggested that the meaning of grammatical elements is usually quite schematic, but given the gradient nature of schematicity, it would not be unwarranted to assume that some grammatical elements are more schematic than others, and that the process by which grammatical elements become more schematic is known more generally in linguistic theory as grammaticalization. This can be illustrated concretely by examining the relationship between impersonal and transitive constructions. Note that I am not discussing the grammaticalization of particular impersonal phrases (as in methinks), although this has been proposed by Palander-Collin (1997). Specifically, the development of the Types N, I and II constructions between Old English and Modern English illustrates a kind of grammaticalization. The development of the transitive and impersonal constructions in the history of English is an example of specialization (in the sense of Hopper 1991), whereby a narrower range of construction types license a larger range of specific tokens, though the grammatical meaning indicated by such construction types becomes more general: specifically, the transitive construction has emerged as a superordinate category (on the criteria listed, for example, by Croft and Cruse 2004). Thus in OE, the basic level constructions were: (i) the ExpCxn (which sanctioned the prototypical Type N, and its extensions, Type I and Type II) and (ii) Type T. In ME, structural case assignment becomes the only option, following the demise of lexical case assignment, but the gradual loss of the various subtypes of the ExpCxn is more consistent with a grammaticalization account which involves the gradual disfavouring of a set of constructional subtypes (cf. Fischer 1992 for a slightly

318

Graeme Trousdale

different argument regarding the gradualness of the change) and an increase in productivity for the remaining types. By Modern English, the range of distinctive constructional types has reduced, and the TrnCxn licenses an increasingly larger range of tokens. This change is facilitated by the ambiguous – and sometimes nonexistent – force dynamics of the processes associated with impersonal constructions. On stative mental relations, Croft (1998) notices significant cross-linguistic variation regarding the assignment of experiencer and stimulus (or source) to the grammatical roles of subject and object. In these mental states, we cannot appeal to force dynamics, since the participants do not enter into such a relationship: Hence, in some languages (such as English), the experiencer is normally made subject; in other languages (such as Russian and languages of southcentral Asia), the experiencer is normally made (indirect) object; in still other languages, the experiencer and stimulus are both encoded as either subjects (Japanese ‘double-ga’ constructions) or as objects (Eastern Pomo) (Croft 1998: 85).

The selection of either the experiencer or the source as subject (as in the OE Type II and Type I respectively) is warranted by the fact that both roles share features associated with prototypical subjects: experiencers are typically animate and human; sources are associated with causation, like agents.

6.

Future prospects and concluding remarks

Denison (this volume) has observed that the present paper does not provide any information regarding the relative frequency of the various constructions involved in the change. It is my intention to develop the framework established in this piece of research by carrying out a collostructional analysis (see Stefanowitsch and Gries 2003; Gries and Stefanowitsch 2004) of the standard parsed corpora that Denison lists, in order to establish the frequency with which certain verbs appear in the various subconstructions identified here. Such an analysis will also shed light on the process of constructional blending mentioned in footnote 3.

Words and constructions in grammaticalization

319

In summary, we can see how the demise of the impersonal construction relates to grammaticalization by considering the three steps for grammaticalization suggested by Croft (2001): 1. Extension of function of construction type A to construction type B This stage involves “alternative conceptualisations of the same experience”, that is, variation in construal, but specifically whereby “the new construction at least partially imposes the conceptualization of its original structure and function” (Croft 2001: 127). In our case, this is where experiencers become coded as more agentive (and have grammatical properties associated with agents, such as subject marking.) 2. Marginalization of the old construction The prototypical impersonal construction (Type N) is already rare in Old English. Many experiencer verbs already have either the source or the experiencer coded as nominative, and increasingly such constructions involve NPs with distinctive subjectlike properties (e.g. raising, verbal concord.) 3. The new construction undergoes shifts in grammatical behaviour. At this stage, we expect to see ‘transitives’ with an increasingly wide scope of ‘subject’ types, and fossilisation of the remnants of the old construction (as in methinks, and possibly if you like), which, while they are a feature of the language of modern English speakers, could not be considered as evidence of a productive grammatical process. Given the prominence he has been accorded elsewhere in this article, it is fitting that the final quotation should come from Meillet: (26) A chaque fois qu’un élément linguistique est employé, sa valeur expressive diminue et la répétition en devient plus aisée (Meillet 1958 [1912]: 135) [Each time a linguistic element is used, its expressive value diminishes, and the repetition becomes easier (my translation)] Meillet’s work on grammaticalization has significant resonances for those working on constructional change in English. His views on lexical grammaticalization are well known; his views on the place of constructions in grammaticalization, less so. But the development of impersonal and transitive

320

Graeme Trousdale

clauses in the history of English provides a compelling argument for a constructional account of grammaticalization9. Grammaticalization can be seen both as increased entrenchment and as increased schematization (Bybee 2003), or perhaps more succinctly, as the entrenchment of schemas.

Notes *

This article is based on a paper presented at the Studies in the History of the English Language 4 conference held at Northern Arizona University in Flagstaff in September 2005, and on a presentation at Roehampton University’s English Language and Linguistics research seminar series in December 2005. I am grateful to the audiences at both venues for their helpful and insightful comments, and to Nik Gisborne, Dirk Noël, Lene Schøsler and especially to Elizabeth Traugott, for a number of particularly useful discussions on grammaticalization and constructions, and to David Denison, for very helpful discussions based on his reactions to the paper (see Denison, this volume). I would also like to thank the two anonymous reviewers of an earlier version of this piece; their comments were extremely useful. All remaining shortcomings are my own. I am also very grateful to the British Academy and to the University of Edinburgh Development Trust for grants (British Academy Overseas Conference Grant 41410 and University of Edinburgh Small Project Grant 2081) which paid for my travel costs to attend SHEL 4 and to present the original paper. 1. As Denison (this volume) rightly notes, this paper presents only a partial history of some variants of the impersonal and transitive constructions (but see also footnote 3). A planned larger study will consider not only the kinds of complexities raised by the kinds of constructional blends discussed in footnote 3 and by Denison, but also the crucial issue of word order as discussed in detail by Allen (1995), which is also not addressed here. 2. In Figure 3, for the central constructional variants under discussion, I give the thematic role played by each NP, and the case marking on the NP, but do not ascribe a (grammatical) function such as subject or object; in Figure 4, however, I do not give the case marking, but do give the grammatical function. This is meant to reflect what I see as a change in the organization of the grammar of English. The grammaticalization of the transitive construction involves the grammaticalization of the category of subject in English (on which see further Anderson 1997b). It is not clear to me that we can be confident on subjecthood in relation to many instances of the OE ImpCxn (cf. Denison 1993: 61–2), but we can be clear(er) on the roles played by participants in any

Words and constructions in grammaticalization

3.

4.

5.

6.

7.

8.

9.

321

given clause. The category of ‘subject’, as a result of its grammaticalization, has strengthened as it has evolved in English. Some verbs (e.g. OE lufian) show yet a different kind of constructional blending, which is not covered by Types N, I, II or T. With lufian, the experiencer verb has a nominative experiencer argument and an accusative source argument. Denison (this volume) provides further examples of clauses which show yet further kinds of constructional blending. I would classify his examples (22)–(24) as non-prototypical extensions of the Type T construction in OE. His suggestion that such examples might be analysed using a multiple inheritance model of constructional taxonomies (see Goldberg 2006) is clearly worth pursuing, and I have argued elsewhere (see Trousdale 2008) that such multiple inheritance models can provide a principled account of the development of some light verb constructions in the late Modern English period. In Hopper and Thompson (1980), the authors follow Dixon (1979) in using A and O to denote the two participants in a two-participant clause, Agent and Object respectively; but the authors “make no claims about the grammatical relations that the NP arguments referring to these participants might bear to the verb” (Hopper and Thompson 1980: 252, footnote 1). Preposed Dative Experiencers occur in clauses in which “a non-nominative Experiencer is fronted, without regard to the case of the Theme” (Allen 1995: 21–2) It is possible that the example in (9) is a development of OE ofhreowan, which did occur in Type II constructions, though the form arewen was also common in the fourteenth century. The change from lexical case to structural case is an important factor in the loss of the PDEs, but I see this development as part of the grammaticalization process; this view is consistent with that of Meillet (1958 [1912]) on grammaticalization. In an earlier version, I suggested that no verbs in the TrnCxn appear with nonNP subjects, but the observations made by Denison (this volume) show that this was too strong a claim. However, note that the example that he gives (Over there suits me fine) is a marginal (i.e. non-prototypical member) of the transitive category: there is, for instance, no passive variant (*I am suited fine by over there). See Noël (2007) for a discussion of the problems in bringing together grammaticalization and constructional accounts of language.

322

Graeme Trousdale

References Primary sources ÆCHom I = Clemoes, P. (ed.) 1997 Ælfric’s Catholic Homilies: the first series. Oxford: Early English Texts Society. ÆHom (Supp.) = Pope, J. (ed.) 1968 Homilies of Ælfric: a supplementary collection, volume II. Oxford: Early English Texts Society. Caxton = Blake, N.F. (ed.) 1970 The History of Reynard the Fox Translated from the Dutch Original by William Caxton. Oxford: Early English Texts Society. Cornwallis = Griffin, R. (ed.) 1842 The Private Correspondence of Jane, Lady Cornwallis, 1613–44. London: S. & J. Benson, Wilson and Fley, 1842. Dudley = Bruce J. (ed.) 1844 Correspondence of Robert Dudley, Earl of Leycester, during his government of the Low Countries, in the years 1585–1586. Camden First Series. Harley = Lewis, T.T (ed.) 1854 Letters of the Lady Brilliana Harley, wife of Sir Robert Harley, of Brampton Bryan, Knight of the Bath. Camden First Series. Harman = Viles, E. and F.J. Furnivall (eds.) 1937 Thomas Harman [1567, 3rd ed.] A Caveat or Warening for Commen Cursetors Vulgarely called Vagabones. London: Early English Texts Society. Lisle = The Trial of Lady Alice Lisle In R. Hargreave (ed.), A Complete Collection of State Trials and Proceedings for High Treason and other Crimes and Misdemeanours from the reign of King Richard II to the end of the reign of King George I. London, 1730. ME Sermons = Woodburn O. Ross (ed.) 1940 Middle English Sermons. London: Early English Texts Society. More 1513 = Sylvester, Richard S. (ed.) The Complete Works of Sir Thomas More volume 2: The History of King Richard III. New Haven and London: Yale University Press. More 1557 [c.1522] = Facsimile text of The Four Last Things by Sir Thomas More in Campbell, W.E. and A.W. Reed (eds.) 1931 The English Works of Sir Thomas More, volume 1. London: Eyre and Spottiswoode.

Words and constructions in grammaticalization

323

Orosius = Sweet, H. (ed.) 1883 King Alfred’s Orosius. Oxford: Early English Texts Society. Swet Iesus = ‘Swet Iesus, hend and fre’ [The Poem of Michael Kildare] in Heuser, W. (ed.) 1904 Die Kildare-gedichte. Bonner Beträge zur Anglistik 14. I have also made use of Healey and Venezky (1980)’s Microfiche Concordance to Old English, listed in the secondary sources below.

Secondary sources Allen, Cynthia 1995 Case Marking and Reanalysis: Grammatical Relations from Old to Early Modern English. Oxford: Oxford University Press. Anderson, John M. 1997a Subjecthood and the English impersonal. In Language History and Linguistic Modelling, Volume 1. Raymond Hickey and S. Puppel (eds.), 251–263. Berlin/New York: Mouton de Gruyter. 1997b A Notional Theory of Syntactic Categories. Cambridge: Cambridge University Press. Bybee, Joan 1985 Morphology: A Study of the Relation between Meaning and Form. Amsterdam/Philadelphia: Benjamins. 1998 The emergent lexicon. CLS 34: The Panels, 421–435. 2003 Cognitive processes in grammaticalization. In The New Psychology of Language: Cognitive and Functional Approaches to Language Structure, volume 2. Michael Tomasello (ed.), 145–167. Mahwah, New Jersey: Lawrence Erlbaum. Bybee, Joan L., 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. Croft, William 1998 The structure of events and the structure of language. In The New Psychology of Language: Cognitive and Functional Approaches to Language Structure, volume 1. Michael Tomasello (ed.), 67–92. Mahwah, New Jersey: Lawrence Erlbaum. 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.

324

Graeme Trousdale

Denison, David 1990 The Old English impersonals revived. In Papers from the Fifth International Conference on English Historical Linguistics. Sylvia Adamson, Vivien Law, Nigel Vincent and Susan Wright (eds.), 111– 140. Amsterdam/Philadelphia: Benjamins. 1993 English Historical Syntax. London: Longman. Dixon, Robert M.W. 1979 Ergativity. Language 55: 59–138. Elmer, Willy 1981 Diachronic Grammar: the History of Old and Middle English Subjectless Constructions. Tübingen: Niemeyer. Fischer, Olga 1992 Syntax. In The Cambridge History of the English Language volume II, 1066-1476 Norman Blake (ed.), 207–408. Cambridge: Cambridge University Press. Fischer, Olga and Frederike van der Leek 1983 The demise of the Old English impersonal construction. Journal of Linguistics 19: 337–368. 1987 A “case” for the Old English impersonals. In Explanation and Linguistic Change (Current Issues in Linguistic Theory 45). Willem Koopman, Olga Fischer, and Frederike van der Leek (eds.), 79–120. Amsterdam/Philadelphia: Benjamins. Goldberg, Adele 1995 Constructions: a Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press. 2006 Constructions at Work: The Nature of Generalizations in Language. Oxford: Oxford University Press. Gries, Stefan Th. and Anatol Stefanowitsch 2004 Extending collostructional analysis: a corpus based approach to ‘alternations’. International Journal of Corpus Linguistics 9.1: 97–129. Healey, A. and R. Venezky 1980 A Microfiche Concordance to Old English. Dictionary of Old English Project: University of Toronto. Hopper, Paul, J. 1987 Emergent grammar. BLS 13: 139–157. 1991 On some principles of grammaticalization. In Approaches to Grammaticalization volume 1. Elizabeth Closs Traugott and Bernd Heine (eds.), 1–35. Amsterdam/Philadelphia: Benjamins. Hopper, Paul and Sandra Thompson 1980 Transitivity in grammar and discourse. Language 56: 251–299. Hopper, Paul J. and Elizabeth Closs Traugott 2003 Grammaticalization, 2nd ed. Cambridge: Cambridge University Press.

Words and constructions in grammaticalization

325

Hudson, Richard A. 1997 The rise of auxiliary DO: verb-non-raising or category strengthening? Transactions of the Philological Society 95: 41–72. Labov, William 1973 The boundaries of words and their meanings. In New Ways of Analyzing Variation in English. C. J. N. Bailey and R. Shuy (eds.), 340– 373. Washington, DC: Georgetown University Press. Lakoff, George 1987 Women, Fire and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press. Langacker, Ronald 1987 Foundations of Cognitive Grammar volume 1: Theoretical Prerequisites. Stanford: Stanford University Press. 1991 Foundations of Cognitive Grammar volume 2: Descriptive Application. Stanford: 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 : Walter de Gruyter. Meillet, Antoine 1958 L’évolution des formes grammaticales. In Linguistique historique et [1912] linguistique générale. Antoine Meillet, 130–148. Paris: Champion. Noël, Dirk 2007 Diachronic construction grammar and grammaticalization theory. Functions of Language 14.2: 177–202. Palander-Collin, Minna 1997 A medieval case of grammaticalization, methinks. In Grammaticalization at Work: Studies of Long-term Developments in English. Matti Rissanen, Merja Kytö and Kirsi Heikkonen (eds.), 371–403. Berlin/ New York: Mouton de Gruyter. Roland, Douglas and Daniel Jurafsky 2001 Verb sense and verb subcategorization probablilities. In The Lexical Basis of Sentence Processing: Formal, Computational, Experimental Issues. Suzanne Stevenson and Paolo Merlo (eds.), 325–346. Amsterdam/Philadelphia: Benjamins. Stefanowitsch, Anatol and Stefan Th. Gries 2003 Collostructions: investigating the interaction of words and constructions. International Journal of Corpus Linguistics 8.2: 209–243. Taylor, John 1998 Syntactic constructions as prototype categories. In The New Psychology of Language: Cognitive and Functional Approaches to

326

Graeme Trousdale

Language Structure, volume 1. Michael Tomasello (ed.), 177–202. Mahwah, New Jersey: Lawrence Erlbaum Thompson, Sandra and Paul Hopper 2001 Transitivity, clause structure, and argument structure: Evidence from conversation. In Frequency and the Emergence of Linguistic Structure. Joan Bybee and Paul Hopper (eds.), 27–60. Amsterdam/Philadelphia: Benjamins. Tomasello, Michael 1998 Introduction. In The New Psychology of Language: Cognitive and Functional Approaches to Language Structure, volume 1. Michael Tomasello (ed.), vii–xxiii. Mahwah, New Jersey: Lawrence Erlbaum. Traugott, Elizabeth Closs 1992 Syntax. In The Cambridge History of the English Language volume 1: the beginnings to 1066. Richard M. Hogg (ed.), 168–289. Cambridge: Cambridge University Press. 2006 Constructions and language change revisited: Constructional emergence from the perspective of grammaticalization. Paper presented at Directions in English Language Studies. University of Manchester, April 6–8 2006. Forthcom. The grammaticalization of NP of NP constructions. In Constructions a. and Language Change. Alexander Bergs and Gabriele Diewald (eds.). Berlin/New York: Mouton de Gruyter. Forthcom. Grammaticalization, emergent patterns, and the notion of ‘newb. ness’. Plenary paper presented at the High Desert Linguistics Society. University of New Mexico, November 9–11 2006. German translation to appear as ‘Grammatikalisierung und emergente Konstruktionen’ in Konstruktionsgrammatik und grammatische Konstruktionen. Anatol Stefanowitsch and Kerstin Fischer (eds.), Tübingen: Stauffenburg. Trousdale, Graeme 2008 Constructions in grammaticalization and lexicalization: evidence from the history of a composite predicate in English. In Constructional explanations in English grammar. Graeme Trousdale and Nikolas Gisborne (eds.), 33–67. Berlin/New York: Mouton de Gruyter.

Variation in Late Modern English: Making the best use of ‘bad data’

Joan C. Beal 1.

Overview

The three papers in this section all deal in different ways with what Labov (1994) identified as the key challenge for historical linguists: that of “making the best use of bad data”. The researcher engaged in a study of contemporary English can design a balanced, representative sample and has access to information about the social factors and external influences that may impinge on linguistic variation, but, as Stefan Dollinger points out in his paper “Taking permissible shortcuts? Limited evidence, heuristic reasoning and the modal auxiliaries in early Canadian English”, those dealing with historical data have a much more difficult task. They have to rely on evidence that “survive[s] by chance not design” (Labov 1994: 11), which is often in what Labov terms a “normative dialect”, and for which, as Nevalainen (1999) points out, the metadata is lacking. For scholars researching the Late Modern period, these problems are less acute than for medievalists, since more data survives, including documents from lesseducated, lower-class writers, and the social and historical background is easier to reconstruct. On the other hand, as Dollinger points out here, the “colonial” varieties which emerge during this period present their own challenge to the researcher: the impossibility of finding sufficient data from the large number of input varieties which influence the formation of e.g. Canadian English. Dollinger here proposes a heuristic approach to the “bad data” problem and tests this out in a study of the semantic development of modal auxiliaries in early Canadian English. Connie Eble’s paper “English/French bilingualism in nineteenth-century Louisiana: a Social network analysis” deals with a different “bad data” problem: that of accessing information about, and documentation of, language use in a bilingual community of the Late Modern period. Working within the social network framework first applied to sociolinguistics by James and Lesley Milroy in the 1970s (Mil-

328

Joan C. Beal

roy 1987), and applied to historical linguistics in the collection of papers edited by Ingrid Tieken-Boon van Ostade, Terttu Nevalainen and Luisella Caon (2000), Eble analyses an archive of family papers to discover the reasons for the shift from French to English between the late eighteenthand the early twentieth century. Ray Hickey’s paper “‘What strikes the ear’: Thomas Sheridan and regional pronunciation” tackles what is perhaps the most intransigent “bad data” issue of all: the lack of information concerning the pronunciation of any variety of English before the invention of the phonograph. Taking what we might call a historiographical or even philological approach, and taking as his primary source texts which might easily have been dismissed as “normative”, the works of the Irish-born elocutionist Thomas Sheridan, Hickey discusses the evidence for IrishEnglish pronunciation in the eighteenth century.

2.

Eble on bilingualism in nineteenth-century Louisiana

Social network studies of Late Modern English have hitherto tended to be based on the analysis of letters, since these provide both samples of language and evidence of network connections in writers and recipients themselves and third parties referred to in the letters. There are now several corpora of letters, such as the Corpus of Early English Correspondence (Raumolin-Brunberg and Nevalainen (2007)); the Corpus of Scottish Correspondence (Meurman-Solin 2007); the Corpus of Late Eighteenth-century Prose (van Bergen and Denison (2007)); and the corpus of nineteenthcentury Scottish correspondence compiled by Marina Dossena (2004). Other corpora of Late Modern texts, such as the Network of Eighteenthcentury Texts (Fitzmaurice 2007) contain substantial subcorpora of letters. The material analysed by Eble is slightly different in that it constitutes the whole of a family archive, Prudhomme Family Papers (#613), held in the Southern Historical Collection of the University of North Carolina at Chapel Hill. Whilst personal and business correspondence makes up part of this archive, there is also a diversity of other materials including “baptismal, marriage and death certificates, diaries, inspirational verse, personal correspondence, school lessons and composition books, scrap books and autograph books, greeting cards and religious cards, invitations, newspaper clippings about social and civic activities, and family genealogies … plantation journals, ledger books, inventories, promissory notes, acts of sale, and bills” (p. 338). Whilst much of this will have provided little in the way

Variation in Late Modern English: Making the best use of ‘bad data’

329

of linguistic data, and Eble herself admits that it was at first “disappointing”, it is invaluable for piecing together the details of family life: socioeconomic circumstances, kinship, friendship and business ties, etc. which are vital for a social network analysis. Whilst this data may be “bad” in the sense that it provides no direct evidence for linguistic variation and/ or change, nor any overt comment on the use of French or English, it does allow Eble to gain insight into the position of the Prudhomme family and other white creoles in plantation society and to demonstrate how the changes in their social networks led to the shift from French to English. The descendants of early French settlers such as the Prudhommes formed an elite whose bilingualism separated them from both the later (Acadian) French-speaking settlers and the English-speaking incomers, plantation managers and slaves. In the early nineteenth century, their position in Louisiana society was quite isolated, so their networks of kinship, marriage and business were dense and multiplex. Certain members of the family would have weak ties to other groups and individuals with whom business was conducted. Eble notes that, at this stage, French was used within the dense, multiplex family network, and English outside it: Phanor Prudhomme (1807–1865) kept his journal and conducted family correspondence in French, but his business and legal correspondence was in English of nativespeaker standard. Eble comments that he “functioned in the Englishspeaking world of the Americans” (p. 344). The next generation, exemplified by Phanor’s son Alphonse (1838–1919) used English for most business purposes, including the plantation journal, with French now relegated to being used for “personal and family reasons” (p. 345). By the next generation, French had become a foreign language learnt at school. Eble argues convincingly that the shift from French to English in the generation of Alphonse Prudhomme was facilitated by the weak network ties created as a result of the young people of this generation being sent out of Louisiana for their education, which was conducted in English. This paper provides a good example of the insights to be gained from applying the social network model to historical material. It also demonstrates the value of collections of family papers as providing, not only linguistic corpora to be mined for evidence, but also rich evidence of the social and historical background against which these texts were produced. There must be many more collections like the Prudhomme Family Papers yet to be discovered by historical linguists: Eble’s paper provides a useful model for their exploitation.

330

3.

Joan C. Beal

Dollinger on modals in Early Canadian English

Stefan Dollinger’s paper directly confronts the issue of “bad data” and puts forward the idea of using a heuristic method to overcome this. Although Dollinger deals with variation and change in modal verbs in early Canadian English, this is presented as a case study and the main thrust of the paper is methodological. Dollinger cites the definition of heuristic as: A process, such as trial and error, for solving a problem for which no algorithm exists. A heuristic for a problem is a rule or method for approaching a solution. (Blackburn 2005: s.v. heuristic, cited p. 358).

He points out that heuristic principles have been used extensively in other disciplines, such as engineering, computer sciences and social sciences, but, perhaps surprisingly not in historical linguistics, where they might usefully be applied to the analysis of “bad data”. Historical studies of Canadian English are few, and Dollinger, as a pioneer in this field, rightly argues that, if we were to wait until corpora of a size large enough to provide statistically sound evidence for influence from all the varieties which could have provided input to early Canadian English, progress would be very slow indeed. Instead, he suggests that we take what data we have and compare results from this with those from the richer and more extensive corpora of Present-day English, in order to measure the relative conservativeness or progressiveness of Canadian English from different genres and at various points in history. Dollinger’s analysis of modal verb usage is based on the comparison of data from the Corpus of Early Ontario English, preConfederation section (CONTE-pC); A Representative Corpus of Historical English Registers, version 1 (ARCHER-1), which provides American and British English data, and the Corpus of Late Eighteenth-Century Prose (CL18P), which he uses for data from the North-west of England. The Canadian, American and British data is subdivided into three genres: letters, diaries and newspapers. The data are also divided into four periods: 1776– 1799, 1800–1824, 1825–1849 and 1850–1899, but here the “bad data” problem becomes apparent, as only the British English component of the Archer corpus provides data for all periods. However, such comparisons as can be made from this incomplete dataset do suggest trends in the development of “root possibility” and “permission” senses of CAN in early Canadian English. For root possibility uses of CAN, Canadian English is closer to American English in those periods and genres for which comparable data is available. Compared with British English usage, Canadian Eng-

Variation in Late Modern English: Making the best use of ‘bad data’

331

lish appears more progressive in its use of “root possibility” CAN in diaries and more conservative in letters and newspapers. In the later periods, Canadian and British usage in newspapers converges, perhaps suggesting “a stylistic change towards more BrE norms in Canadian newspapers” (p. 367). With regard to “permission” uses of CAN, Dollinger’s analysis shows an increase in all genres in both British and Canadian English, suggesting a parallel development. Whilst the data suggests these interpretations, Dollinger admits that “no change of CAN is statistically significant” because of the low number of tokens involved. However, using heuristic reasoning, he is able to use the knowledge of CAN’s status as the unmarked variant, gained from more statistically robust studies of 20th-century English such as Coates (1983), to extrapolate an interpretation of his incomplete data as “a LModE parallel development in both CanE and BrE leading up to the PDE distribution” (p. 368). Dollinger then goes on to apply the same methodology to a comparison of the other “core” modals across the same genres, varieties and periods. His conclusion is that, whilst in the first period Canadian English usage is close to American English usage for six out of the nine modals, confirming the Loyalist base of early Anglophone Canadian society, overall, Canadian English can be seen as developing its own patterns rather than simply following British or American usage, which tallies with previous findings in the area of vocabulary. Readers accustomed to the statistical rigour of corpus linguistics might find Dollinger’s account of developments in the modal system of early Canadian English unconvincing, but, as he points out, the task of creating a corpus of sufficient size representing all the input varieties would be considerable. In the meantime, Dollinger’s “best guess” approach at least provides us with a set of working hypotheses. In fact, accustomed as we are to “making the best use of bad data”, historical linguists have always employed heuristic methods, but in recent times we have increasingly felt guilty about it. Dollinger’s paper provides a rationale for the heuristic approach, and at the same time a spur to the collection of “better” data which would allow us to test out his hypotheses more rigorously.

4.

Hickey on Thomas Sheridan and regional pronunciation

For those of us who are interested in the pronunciation of Late Modern English, the data is not so much bad as non-existent. For any period before the invention of the phonograph (1877), there is no possibility of finding

332

Joan C. Beal

“data” as such, and even for the first half of the 20th century surviving recordings are sparse and their preservation and subsequent availability often a matter of luck. The ONZE corpus (Gordon et.al. 2007) is a rare example of systematically collected data from this period (collected between 1946 and 1948, but including speakers born as early as 1851) being preserved intact and made available to linguists. Given the paucity of “real” data, historical phonologists have to reconstruct the pronunciation of earlier periods from the accounts of orthoepists, rhymes, puns and “incorrect” spellings in letters and diaries. I have discussed the relative merits of these sources elsewhere (Beal 1999: 38–47) and pointed out that, although Dobson (1957: I, 311) dismissed the eighteenth century as producing no writers to compare with the sixteenth- and seventeenth-century orthoepists who provided the sources for his magnum opus, the elocutionists and authors of pronouncing dictionaries whose works began to appear towards the end of the eighteenth century constitute a very rich source of evidence for variation and change in Late Modern English pronunciation. The relative neglect of this period which caused Charles Jones to refer to the eighteenth and nineteenth centuries as “the Cinderellas of English historical linguistic study” (1989: 279), has to a great extent been rectified, not least by Jones’s more recent (2005) work, which is to some extent a “Dobson” for the Late Modern period. Nevertheless, there is still a great deal to be said about the influence of individual authors – elocutionists rather than orthoepists in this period – on the development of varieties of English in the Late Modern period. Raymond Hickey focuses on one of the most influential elocutionists of the late eighteenth century, Thomas Sheridan. Like his ultimately more successful rival John Walker, Sheridan was an actor who turned his hand to elocution at a time when there was a demand, particularly from the upwardly-mobile and especially those living in Scotland, Ireland and parts of England remote from the capital (“the provinces”) for clear and explicit guides to “proper and polite” pronunciation. As I have pointed out (Beal 2004) the normative pronouncements of eighteenth-century authors can provide evidence for regional pronunciations, especially those which were most salient and hence most likely to be condemned. Hickey here takes up this suggestion and examines Sheridan’s comments both as a source of evidence for Irish-English pronunciation in the eighteenth-century and in order to determine the extent of Sheridan’s influence on the subsequent development of this variety.

Variation in Late Modern English: Making the best use of ‘bad data’

333

Hickey tells us that Sheridan was born and educated in Ireland and was the godson of Jonathan Swift. However, whilst Swift was “held up as a paragon of English style” by authors such as Lowth, Sheridan, as an Irish Catholic, albeit with impeccable connections, was subject to criticism. Dr Johnson questioned his credentials for teaching correct pronunciation to the English on the grounds that he was Irish. As Hickey points out here, the disparity between Swift’s reception and Sheridan’s could be due to the fact that the former was an Anglo-Irish Protestant clergyman and therefore respectable, whereas Sheridan was “a Catholic actor” with the temerity to be “a self-appointed arbiter of correct English usage” (p. 389). Hickey bases his account on Sheridan’s Rhetorical Grammar of the English Language (1781) in which he inserted a series of “Rules to be observed by the natives of Ireland in order to attain a just pronunciation of English” which had first appeared in his General Dictionary of the English Language (1780) and was later to be taken over wholesale by John Walker in his Critical Pronouncing Dictionary (1791). Hickey notes that Sheridan only mentions the features of Irish English that “struck his ear” (p. 394), and that he has nothing to say about the consonantal features which are so salient in present-day Dublin English. Nevertheless, Hickey’s comparison of Sheridan’s observations with later studies of Irish English does provide evidence for the earlier history of these features. Perhaps the most familiar of these to scholars in the field of English historical phonology is the use of /e:/ rather than RP /i:/ in words such as least, deal and beet. Sheridan condemns this as the chief “mistake” of the Irish, but goes on to discuss what is clearly a hypercorrection, where the “well-educated natives of Ireland” whose speech he describes, pronounce words such as prey, convey and bear, with /i:/, even though these words would not have this vowel in educated London English. This suggests that the stereotypical Irish-English /e:/ in, for instance Jaysis!, was already heavily stigmatised in the eighteenth century. This is corroborated by Maria Edgeworth who writes of Irish ladies who are “ashamed of their country” and “betray themselves by mincing out their abjuration, by calling tables teebles and chairs cheers” (1802, cited in Crowley 2000: 136). A similar instance of hypercorrection is suggested by Sheridan’s remark that the Irish used the vowel which he represented as , corresponding to /¡/ in IPA, in words such as pulpit, pudding, bush, bull, push and pull, where both present-day RP and Irish English would have /7/. We know from other sources such as Kenrick (1773) and Walker (1791) that the use of /7/ in e.g. cup, blood was already stigmatised as a Northern English feature, and that there was evidence of a similar

334

Joan C. Beal

hypercorrection amongst northerners (Beal 1999: 135), so Hickey’s inference is well supported. Hickey concedes that Sheridan’s evidence is piecemeal, and that he “appears to have been selective in his censure” (p. 405): the centralised diphthong in PRICE which was to become an “Oirish” stereotype in the nineteenth century is not mentioned, and, of course, where Irish pronunciation coincided with the “correct” usage of London, as was the case with the lengthened pronunciation of the vowel in cloth, off, etc. Sheridan makes no comment. Even here, though, negative evidence, the “dog that didn’t bark in the night” is useful in pointing to the social evaluation of variants in eighteenth century Britain. Hickey also discusses the extent to which Sheridan’s “rules” actually had any effect on the pronunciation of the “natives of Ireland”. Whilst his strictures appear to have had no immediate effect, some of the pronunciations which he condemned, such as /Ε/ in catch and /Ο:/ in psalm have since disappeared from Irish English. Whether they became stigmatized because of Sheridan’s censure, or were noticed by Sheridan because they were already shibboleths of Irish speech, is impossible to tell. Hickey’s paper is a good example of the ways in which eighteenthcentury normative texts can be mined for evidence of non-standard pronunciation, making accounts of the histories of these varieties possible, and providing time-depth to synchronic studies of variation and change. There are many more such texts available electronically in Eighteenth-century Collections Online, so I look forward to seeing much more work in this vein.

5.

Conclusion

The three papers in this collection provide different solutions and methodologies for dealing with “bad data”, and give pointers to future work on variation in Late Modern English. Eble demonstrates how family archives can be useful for social network studies, even when the texts themselves are largely written in standard varieties; Dollinger challenges the naysayers who might be disheartened by a lack of sufficient data for statistically significant results, advocating a heuristic approach which can extract workable hypotheses from incomplete datasets; and Hickey makes use of a normative text to provide evidence for the pronunciations which its author sought to eradicate. Trudgill (2002: 44) asserts that there is “no excuse” for confining

Variation in Late Modern English: Making the best use of ‘bad data’

335

accounts of Later Modern English to studies of British and US English: the papers in this section demonstrate that the problem of “bad data” is likewise no excuse for neglecting the study of variation in this period.

References Beal, Joan C. 1999 English Pronunciation in the Eighteenth Century: Thomas Spence’s “Grand Repository of the English Language”. Oxford: Clarendon Press. Beal, Joan C. 2004 “Marks of disgrace”: Attitudes to non-standard pronunciation in 18th-century pronouncing dictionaries. In Methods and Data in English Historical Dialectology. Marina Dossena and Roger Lass (eds.), 329–49. Bern: Lang. Beal, Joan C., Karen P. Corrigan and Hermann L. Moisl (eds.) 2007 Creating and Digitizing Linguistic Corpora. Basingstoke: Macmillan. Blackburn, Simon 2005 The Oxford dictionary of philosophy. 2nd ed. Online version. Oxford: Oxford University Press. Coates, Jennifer. 1983 The semantics of the modal auxiliaries. London: Croom Helm. Crowley, Tony 2000 The Politics of Language in Ireland, 1366–1922. A Sourcebook. London: Routledge. Dobson, E. J. 1957 English Pronunciation 1500–1700. Oxford: Clarendon Press. Dossena, Marina. 2004 Towards a corpus of nineteenth-century Scottish correspondence, Linguistica e Filologia 18: 195–214. Fitzmaurice, Susan. 2007 Questions of standardization and representativeness in the development of social networks based corpora: the story of the Network of Eighteenth-century English Texts. In Beal, Corrigan and Moisl (eds.), 49–81. Gordon, Elizabeth, Margaret Maclagan and Jennifer Hay 2007 The ONZE corpus. In Beal, Corrigan and Moisl (eds.), 82–104. Jones, Charles 1989 A History of English Phonology. London: Longman.

336

Joan C. Beal

Jones, Charles 2005 English Pronunciation in the Eighteenth and Nineteenth Centuries. Basingstoke: Palgrave Macmillan. Kenrick, William 1773 A new dictionary of the English language. London: John and Francis Rivington. Labov, William 1994 Principles of linguistic change. Volume 1: Internal factors. Oxford: Blackwell. Meurman-Solin, Anneli. 2007 The manuscript-based diachronic corpus of Scottish correspondence. In Beal, Corrigan and Moisl (eds.), 127–147. Milroy, Lesley 1987 Language and Social Networks. Oxford: Blackwell. Nevalainen, Terttu. 1999 Making the best use of “bad” data. Neuphilologische Mitteilungen. 100 (4): 499–533. Raumolin-Brunberg, Helena and Terttu Nevalainen. 2007 Historical sociolinguistics: The corpus of early English correspondence. In Beal, Corrigan and Moisl (eds.) 148–171. Sheridan, Thomas. 1780 A General Dictionary of the English Language. London: R & J. Dodsley, C. Dilly and J. Wilkie. Sheridan, Thomas. 1781 A rhetorical grammar of the English language calculated solely for the purpose of teaching propriety of pronunciation and justness of delivery, in that tongue. Dublin: Price. Tieken-Boon van Ostade, Ingrid, Terttu Nevalainen and Luisella Caon (eds.) 2000 Social Network Analysis and the History of English. Special issue of the European Journal of English Studies (EJES) 4.3. Trudgill, Peter. 2002 The history of the lesser-known varieties of English. In Alternative Histories of English. Richard Watts and Peter Trudgill (eds.), 29–44. London: Routledge. Van Bergen, Linda and David Denison 2007 A corpus of late eighteenth-century prose. In Beal, Corrigan and Moisl (eds.), 228–246. Walker, John 1791 A Critical Pronouncing Dictionary and Expositor of the English Language. London: G.G.J. and J. Robinson and T. Cadell.

English/French bilingualism in nineteenth century Lousiana: A social network analysis

Connie Eble Dialects of French remained the ordinary spoken language of southern Louisiana west of the Mississippi River until World War II – almost a century and a half after the area had become part of the United States by the Louisiana Purchase of 1803. However, French had been the ancestral language in other parts of the state as well, including the farmlands drained by the Red River, whose center of population was Natchitoches, the oldest European town in the Louisiana Purchase territory, about 250 miles northwest of New Orleans. In that part of the state, French disappeared as the language of public life by the time of the Civil War. This essay applies the notion of social network to an archive of family papers to explore why. It describes the changing use of written French and English during the nineteenth century by a cohesive group of white creoles1 living in rural Natchitoches Parish in northwestern Louisiana, between what are now the cities of Shreveport and Alexandria. The social network itself and the language practices of its members are inferred from the letters, journals, scrapbooks, and other writings preserved in the Prudhomme Family Papers (#613), held in the Southern Historical Collection of the University of North Carolina at Chapel Hill.2 The Prudhomme Family Papers (#613) occupy 41 linear feet of shelf space and comprise more than 16,000 items dating 1765–1997. Preserving on paper the life stories of a single family living in the same parish for more than two centuries and in the same house for 170 years, they include baptismal, marriage and death certificates, diaries, inspirational verse, personal correspondence, school lessons and composition books, scrap books and autograph books, greeting cards and religious cards, invitations, newspaper clippings about social and civic activities, and family genealogies. The family’s financial condition for seven generations is shown by business correspondence, plantation journals, ledger books, inventories, promissory notes, acts of sale, and bills. The bulk of the surviving papers of the de-

338

Connie Eble

scendents of Jean Baptiste Prudhomme, the first family member born in Louisiana, in 1736, form the Prudhomme Family Papers (#613) in North Carolina. Smaller collections are held at Watson Memorial Library of Northwestern Louisiana State University in Natchitoches and the Historic New Orleans Collection. For a linguist, this immense archive is at first disappointing. It reveals nothing directly about language change or variation or contact between French and English, or even opinions or perceptions about language. I have not yet found one comment on the replacement of French by English. Yet the shift in public life from the language of the founding population to the language of the newcomers must have been meaningful to this family. This essay aims to extract information about language shift by the people whose lives are represented in this archive by thinking of them as a social network. A social network approach to language analysis takes as its premise that social connections create networks that serve as mechanisms for transmitting linguistic influence. Observing who speaks to whom, how often, and about what can be particularly useful for examining linguistic variation that cannot be due to major sociolinguistic factors like socio-economic class, ethnicity, educational level, or regional identity, i.e., variation among speakers who do not differ by these measures. For example, John Gumperz (1982) appealed to the notion of social network to explain the shift from Slovenian/German bilingualism to German in a small, rural Austrian village with little social stratification. Social network analysis was established as a tool of current synchronic sociolinguistics by the work of James and Lesley Milroy in three urban communities in Belfast, Northern Ireland, in the 1970s (Milroy 1980, 1987).3 Briefly and broadly, here are the main components of the Milroys’ network approach. A social network is all of the relationships that an individual has contracted with others, taken together. The point of entry is the individual. The patterns of interactions among individuals constitute systems that can be characterized and measured by type and strength. Networks can differ in scope: an individual’s relationships can be confined to a community (closed) or extended outside the community (open). The number of people to whom an individual is connected and who are connected to each other determines the density of the network. In the most dense network, everyone knows everyone else. The number of different ways in which members of a network are linked is the measure of multiplexity. Discovering the social features that constitute meaningful multiplex links

English/French bilingualism in nineteenth century Lousiana

339

for a network is a challenge to investigators, for properties that are important may differ from one community to another and may change. Other characteristics of linkages that contribute to the strength of ties and might elucidate the social dynamics are frequency, duration, and reciprocity. Dense and multiplex linkages can form sectors or clusters, e.g., a workplace cluster or a kinship cluster. An individual can be central to a network or peripheral. In the urban populations that the Milroys were studying, dense, multiplex networks – in which most of the speakers were linked to each other by several measures – tend to be norm-enforcing and to conserve language features. In other words, strong network ties resist change. Loose ties, on the other hand, provide the route for innovation and change. Mobility and migration often place speakers in situations in which their ties to a cohesive, well established group are loose (or weak), and their position is marginal. The weak ties of these individuals tend to be the conduits to localized networks of new information, ideas, and linguistic forms (Milroy 2000: 217–219). Lippi-Green (1989) showed that social network analysis could help explain a phonological change in progress in an entirely different setting – in a small, rural German-speaking community in western Austria. A workshop in conjunction with the 1998 International Conference on English Historical Linguistics took as its aim the application of social network analysis to older stages of the English language. Seven papers presented at the workshop were published together in the European Journal of English Studies in 2000, along with an introductory essay by Lesley Milroy. The papers span a millennium, from the monastic community of the Benedictine Reform of late 10th century England to the contemporary elite of Dublin. They reveal a range of problems in extending a tool developed for contemporary urban sociolinguistics to imperfect linguistic and sociological data from the past. A major problem is determining which ties were important to the individuals and groups in their lifetime. In general, the studies do support the generalizations that strong network ties favor conservatism, whereas loose ties serve as the agents for innovation. Deducing social networks from surviving written records also requires a good understanding of the historical context in which the documents were produced. It is against the following historical backdrop that the language practices preserved in the Prudhomme Family Papers must be examined. French-speaking settlers came to what is now the state of Louisiana in the early decades of the eighteenth century. The first colonial town, Natchi-

340

Connie Eble

toches, grew up around Fort Jean Baptiste, the military outpost established in 1714 on the banks of the Red River on the route to Texas. It marked the western extent of France’s claim. Just fifteen miles to the west, the mission, and later fort, Los Adaes marked the eastern boundary of Spanish Texas. The first Prudhomme came as a French soldier to Natchitoches in 1716 and stayed as a merchant and trader, marrying a Parisienne who came with a shipload of young French women to provide wives for the colonists. Their son Jean Baptiste Prudhomme prospered and became the King’s Physician under the French and a landowner. In 1763 as the result of the French and Indian Wars, France ceded Louisiana to Spain. Although Spanish officials took over the administration of the colony, few Spanish speaking colonists migrated to Louisiana, and the culture and language remained essentially French. Jean Baptiste’s son, Jean Pierre Emanuel Prudhomme, in 1789 secured a land grant from the Spanish governor for a tract on the banks of the Red River thirteen miles south of Natchitoches, called Ile Brevelle (now Bermuda, Louisiana). There, relying on the labor of slaves, he successfully cultivated tobacco and indigo and was perhaps the first planter in Louisiana to grow cotton. In 1821 he moved his family into a newly-constructed, fiveroom plantation house that is now called Oakland. For six generations his direct descendants lived in the house – his grandson, great-grandson, and great-great-grandson each celebrating fifty years of marriage there. Jean Pierre Emanuel’s plantation has been recognized by the United States government as one of two “bicentennial farms” west of the Mississippi River, having been owned and managed by the same family for 200 years. The Prudhomme Family Papers begin during the lifetime of Jean Pierre Emanuel, when Louisiana was a colony of Spain, and extend to 1997, when the house and its remaining outbuildings and land were acquired by the National Park Service as part of the Cane River Creole National Historic Park. Under both French and Spanish rule, the vast Louisiana territory was sparsely populated by European colonists and their descendants, except for the most southern portion around the port of New Orleans and the bayou and prairie regions to its south and west where Acadian immigrants displaced from Canada had settled. In addition, there were a few important enclaves, like St. Louis in Missouri and the Natchitoches area, which included an independent and prosperous community of free people of color who spoke French and were Catholic (Mills 1997). Regardless of their ethnic origins, the language of colonials in Louisiana and their creole descendants was mostly some variety of French. Understanding issues of language

English/French bilingualism in nineteenth century Lousiana

341

contact in Louisiana requires recognizing the continuum of varieties of French in use historically. A conventional description locates three points on the continuum: a variety close in vocabulary and grammar to standard continental French and used in writing; Acadian French based on provincial dialects brought to Canada in the seventeenth century and influenced in Louisiana by contact with other languages; and a creole developed by slaves and their descendants (Breton 1979: 22–23). Even before the Louisiana Purchase, English-speaking Americans saw opportunities afforded by the vast Louisiana territory and crossed the Mississippi River into the Spanish colony, exchanging nominal allegiance to the Spanish crown and the Catholic Church for land grants or administrative positions. The acquisition of Louisiana by the United States removed all impediments to their westward migration. During the nineteenth century, the influx of “Americans” overwhelmed the creoles established along the banks of the Red and Cane Rivers. The newcomers were quite different from the ancienne population: their forebears had come from the British Isles, they spoke varieties of English, and they were Protestants. One study of the settlers of the hill country of northern Louisiana maintains that 73.3% of them came from the five states of Georgia, Alabama, South Carolina, Tennessee, and Mississippi (Trout 1996: 473). In northern Louisiana, the development of an agricultural economy based on cotton also contributed to the region’s rapid Americanization, as it allied the commercial interests of the area to those of other southern cotton-producing states and increased the need for slave labor, a need that could be filled only by acquiring slaves from states east of the Mississippi. Linguist Michael Picone (2003) thinks that the large number of slaves required by cotton production forced the importation of monolingual English-speaking slaves from other parts of the South and was an important factor in the shift to English in the region. To this day, the cultural divide between northern and southern Louisiana is apparent, with northern Louisianans more similar to the people of Mississippi, Alabama, Georgia, and the Carolinas in folkways, religion, speech, politics and culture than they are to southern Louisianans. Beginning about 1830, the Prudhomme family became part of the linguistic and cultural minority of northern Louisiana. Many political leaders in the United States had opposed the Louisiana Purchase of 1803 because of the religious, cultural, and linguistic incompatibility of the people of the United States and the people of the Louisiana territory. To assure the rights of the non-francophone minority of the residents of Louisiana in 1812 when it became the eighteenth state of the Un-

342

Connie Eble

ion, the first state constitution was published simultaneously in both French and English. It provided that all laws passed by the legislature be promulgated in “the language in which the constitution of the United States is written” and that other legal documents be preserved and made available in English (Smith-Thibodeaux 1977: 32–33). The provision for public documents in French was not specified, as French was then the language of the majority of voters and elected officials. The state constitutions of 1845 and 1852 specifically required that the constitution and the laws of the state be published in English and in French (33). However, in 1847 twenty parishes in northern Louisiana were allowed to use English as the sole language of official publication, except in instances when residents of the parish required French (Trépanier 1988: 30). In the same year, public elementary schools were established in the state, with no specification of the language of instruction (29). The “Carpet Bag Constitution” of 1868 mandated English only, but the constitution of 1879 permitted laws to be published in French also, and in heavily French-speaking areas some provision for French in schools was possible (Smith-Thibodeaux 1988: 36–37). The constitution of 1921 is generally regarded as the deathblow to the French language in Louisiana, eliminating it as the medium of instruction in schools and reducing the numerous Louisiana-born monolingual speakers of French to the status of second-class citizens. The social network in which the members of the Prudhomme family led their lives and made their linguistic choices was embedded in the historic context sketched above and was influenced by it. The natives of Natchitoches Parish whose lives can be inferred from the Prudhomme Family Papers formed a dense, multiplex network. They identified themselves as creoles and, for the most part, distinguished themselves from Frenchspeaking Acadians, most of whom lived on subsistence farms or trapped and hunted for a living. Laura Locoul Gore (granddaughter of Phanor Prudhomme’s second wife), in her Memories of the Old Plantation Home, recalled, “The Acadians were the poor, uneducated whites. They seemed to have no ambition to learn …” (63). The white creoles of the Red River area constituted an elite. The names of the most prosperous of them show up in the Prudhomme Family Papers. They were related by birth or marriage or both to creoles in this area and to others along the Mississippi River and in New Orleans and St. Louis. They were Catholics. The basis of the family’s wealth was cotton. They owned slaves. In the most prosperous times, some owned townhouses in Natchitoches as well as the plantations on which their livelihood depended. Although they were wealthy, there were constant

English/French bilingualism in nineteenth century Lousiana

343

financial anxieties, as the entire local economy was subject to the vagaries of weather, flooding, and the price of cotton in Europe. The archive gives abundant evidence of lending, borrowing, buying, and selling among members of the network, both male and female. (Creole women were privy to financial matters and often controlled their own inheritances or managed the family corporation (Marmillion 2001: 135–137).) “In the isolation of the cotton fields around Natchitoches, the Creoles were content to live among themselves in year-long schedules of family dinners, outings, religious devotions and inter-marriage” (144). For most of the nineteenth century, they were bilingual in French and English. The relationships of an individual who figures prominently in the Prudhomme Family Papers can illustrate some of the characteristics of the Natchitoches Parish creole network in the period from the Louisiana Purchase to the Civil War. Pierre Phanor Prudhomme’s life (1807–1865) spans roughly the entire antebellum period. (Except for official documents, most of the time he signs his name as Phanor Prudhomme.) Most evident in the documents and difficult to untangle are the relationships of kinship. Phanor was the youngest of six children; his mother was a Lambre. His siblings married into the local St. Anne, Metoyer, and Lecomte families. Phanor himself married two granddaughters of his aunt Susanne, who were daughters of his cousin Marie Aurore Lambre and Benjamin Metoyer. Another of his aunts married Phillipe DuParc of St. James Parish on the Mississippi River. The daughter of Phanor’s second wife (and his niece) married the grandson of his aunt DuParc. Their infant daughter Laura Locoul Gore and her mother lived with Phanor during the Civil War. After the war, when grown, Laura participated in creole society in New Orleans. Phanor’s older son married the daughter of Ambroise Lecomte, whose plantation was near Phanor’s and who shared Phanor’s love of horse racing. Phanor’s younger son married a Buard, the cousin of his first daughter-in-law. Phanor’s daughters married young men of the Brazeale family. Phanor’s brother-inlaw Felix Metoyer shows up often in papers concerning inheritance, and he served for a while as executor of Phanor’s estate. Other connections of kinship give Phanor ties to the remainder of the landholding creole families of the area, who bear the surnames Bossier, Cloutier, Deblieux, Hertzog, Rachal, and Sompayrac. The network was clearly dense – not only did everyone know everyone else, they were all related by blood or marriage. Perhaps more important, many of the ties were multiplex. Phanor was related to Ambroise Lecomte, for instance, not only more than one way by kinship but also as a neighbor, fellow cotton planter, fellow thoroughbred horse

344

Connie Eble

breeder and member of the Jockey Club, fellow chef de famille expected to support financially the building of a new Catholic church, and fellow financial backer of the Confederate military unit from Natchitoches Parish. Phanor’s ties extended beyond this cluster, however – to cotton factors in New Orleans, to English-speaking overseers of his plantation, to his slaves and hired hands, to school masters and mistresses, to inventors of farm machinery, to government officials, and so forth. Although the family papers contain little on this topic, Phanor also served in the Louisiana legislature and was a delegate to the 1844 state constitutional convention. Within the kinship cluster, his language was almost entirely French. Outside that cluster, he used both French and English. Until the 1830s, the language of the Prudhomme Family Papers and, by inference, the language of their social network, was French. The first generation of Prudhommes after American acquisition preferred French in both personal and business affairs but used English beyond their creole circle in correspondence with government officials and other non-creoles with whom they did business. This is the generation of Pierre Phanor Prudhomme, 1807–1865. The plantation journals that he kept 1837–1841 are in French, mainly a list of dates followed by an event or observation.4 [1837] [1839]

22 mars

Commencé a planter le Coton 25 juillet Beau temps après beaucoup de pluie Novembre 10 donné les Souliers aux Domestiques de la Maison. Decembre 15 donné les Souliers aux Nègres du Champs (Folders 268–269)

Phanor Prudhomme’s Cotton Book for 1836, however, was kept by J. F. Culbertson, probably an overseer. It opens, “Began picking cotton Monday August the 29th 1836.” Although the book is mostly numbers tracking the amount each slave picked, the few notations are in English, for example, “on this side”, “over the river”, “total”, and “lbs” ‘pounds’ (Folder 267). The Cotton Book is signed and dated by both Culbertson and Phanor Prudhomme on November 15, 1836. Virtually all of Phanor Prudhomme’s personal correspondence is in French, and so is a large portion of his business correspondence, mostly with cotton factors in New Orleans. But in 1850 when he wanted to dispute a claim against his father’s estate and to inquire about installing a new kind of cotton press (Folder 143), he wrote in English with the usage and fluency of a native speaker. He functioned in the English-speaking world of the Americans: he read Southern Cultivator and Scientific American, took

English/French bilingualism in nineteenth century Lousiana

345

his daughters and nieces to New York City, and enrolled his sons in schools in Connecticut, Virginia, North Carolina, and Washington, DC. Although the writings of Phanor Prudhomme usually show no mixing of the languages, a letter to Ambroise Lecomte from New Orleans dated 28 Mai/60 reports a conversation with an American owner of racehorses. The letter is in French, but Phanor gives the words of the conversation in English (Folder 902). Although the documents cannot reveal the phonology of his French or English, Phanor Prudhomme appears to have been comfortably fluent in both languages. Phanor Prudhomme’s older son and heir, James Alphonse (1838–1919), left writings that show that he too knew both French and English well. English was his primary written language, and he seems to have written French only for personal and family reasons. When he was away at school, Alphonse wrote his diary entirely in English. As a student at the University of Virginia and then the University of North Carolina in the late 1850s, he would certainly have spoken English in all school courses and activities. Following his father’s death in 1865, Alphonse assumed the responsibility of running the family plantation. In contrast to his father, who ordinarily corresponded with longtime cotton factors in New Orleans in French, Alphonse conducted business with them in English. By 1870 both the plantation journals kept on site and communications with outsiders concerning plantation operations were in English. Alphonse is representative of the generation of creoles who grew up in the vicinity of Natchitoches during the 1840s and 1850s and came to maturity on the eve of the Civil War. When they were not away at school, where their lives were ordered and disciplined, the large group of sisters and brothers, cousins, and cousins-of-cousins led a merry life of visiting, attending dances and parties, and preparing for the next social event. The plantation area around Bermuda became known throughout French creole society as the Côte Joyeuse. These young adults all knew the standard variety of French taught in school, as shown by school notebooks of lectures on French literature and history and by school compositions written in French. And they probably understood and spoke some variety of French with older family members and with black creole household servants. However, the generation of J. Alphonse Prudhomme corresponded with each other in English and kept their diaries in English. Fulbert Cloutier’s brief Civil War diary, written in 1864, is in English except for a short French poem, “Le veritable Amour,” copied onto the first page (Folder 102). The notebook kept during the Civil War by Elise Lecomte, Alphonse’s future wife, is

346

Connie Eble

almost entirely in English, with petitions to God sometimes in both languages, for example, “Protect my loved Phonse from all dangers” and “Mon Dieu, ayez pitié de nous” (Folder 891). Strong, dense network ties usually favor conservatism in language. For this old, established network that found itself in the linguistic minority visà-vis newcomers, the close-knit ties did support the retention of French but retracted it into the private domains of family and religious practice. Further, the type of French they wrote was a standard variety instilled by education that gives no information about their spoken French and no hint of the local colloquial varieties used by monolingual black and white speakers that they came in contact with in Natchitoches Parish. Economic prosperity for members of this network required going outside the closed creole cluster and forming loose network links with monolingual English speakers. Weak links peripheral to a dense multiplex network provide the route for various kinds of innovation, including linguistic innovation, and the Prudhomme papers give evidence of this. The Red River creoles seem to have deliberately extended their network in an important way – in the education of their children. Although some among the wealthy in Louisiana employed tutors for their children, beginning early in the nineteenth century the Prudhommes appear to have sent their children – both boys and girls – to boarding schools. The earliest generations of the French creole network developing around Natchitoches in the eighteenth century were not all literate. As late as 1809 and 1810, two of them, Marie Louise Leconte and Louis Rachal, signed promissory notes with a “marque ordinaire d’une croix,” i.e., with an X (Folder 141). At the same time, others in the network were literate and perhaps even educated. By 1828, young Henri Herzog was enrolled at St. Joseph’s in Bardstown, Kentucky, and writing home to his parents.5 Mon cher Papa, J’ai recu votre derniere letter du 24 aout vous me dite croire que quelqu’un m’ede a ecrire mes lettres mais je vous assure que je les aicrit par ma propre main je suis etonné moi meme que j’aie apris si vite le stil de mes lettre je ne pouvait apeine ecrire une lettre quand je suis arivé au colege mas mon cher pere je vous assure par mon honneur que personne ne m’aide pas … . je ne fait qu’apprendre l’anglais et le français c’est parce qu’il faut que l’on aprenne bien l’anglais pour apprendre le latin et l’espagnole … (Folder 887)

English/French bilingualism in nineteenth century Lousiana

347

Henri’s native tongue is French, but, as the errors in his letter show, he is just beginning to learn how to write. It is also clear that he is expected to learn English before he will be allowed to study Latin, a compulsory subject at that time for entrance to higher education. Because of the schools established by two French orders of nuns, the Ursulines and the Religious of the Sacred Heart, many creole girls remained in Louisiana for their schooling, though the difficulties of travel usually meant that they returned home only once a year, and sometimes not that often. A beautiful handwritten arithmetic book in the collection attests that Lise Metoyer (the daughter of Benjamin Metoyer and first wife of Phanor Prudhomme) attended the Ursuline school in New Orleans. It is inscribed “Cahier d’Arithmétique fait par Lise Metoyer au couvent des Ursulines de la Nouvelle Orléans 14 septembre 1832” (Folder 111). At Ursuline schools in America, morning lessons were conducted in French and afternoon lessons in English (Aycock 1993: 213). Lise learned arithmetic in French. Although there is no record of how or where Phanor Prudhomme was educated, his “cahier D’Arithmétique[s]” also survives. This notebook contains a basic, structured course in arithmetic entirely in French. The handwritten book appears to have been either copied or written down from dictation from an established text, opening “On entend par mathématiques une science qui traite des grandeurs ou quantités … . L’Arithmétique est la science des nombres” (Folder 130). Both Lise and Phanor seem to have acquired their education in a Frenchspeaking context. In 1821 the Religious of the Sacred Heart, a French order of nuns, established a school for girls in Grand Coteau, Louisiana, followed by one sixty miles upstream from New Orleans on the Mississippi River in 1825, and one in Natchitoches itself in 1847 (Calan 1937: 775–780). By the time of the Civil War there were also Sacred Heart academies in St. Louis and in Manhattanville, New York, and at least one of the young women writers in the Prudhomme collection wrote a composition at Manhattanville. All of the Religious of the Sacred Heart were under one superior general, and all of the private academies that they established followed the same Plan of Studies and held many of the same religious and cultural activities. Phanor Prudhomme’s daughters and the daughters of his second wife attended the academy in Natchitoches in the 1850s. The 1851 report card of his stepdaughter, Désirée Archinard Locoul, listed as subjects Lecture française, Lecture anglaise, Grammaire française, Grammaire anglaise, Traduction anglaise, and Traduction française (Gore:144).6 Désirée’s daughter, Laura Locoul Gore, writing her memories for her children in 1935 and 1936, re-

348

Connie Eble

called about her mother, “She sang and played delightfully and had perfect command of both English and French” (24). During the nineteenth century, the Ursuline and Sacred Heart schools in Louisiana preserved and transmitted French language and culture and Catholic devotional practices while preparing the girls in their charge to live as English-speaking American women. Many of the young cousins of the Côte Joyeuse of James Alphonse Prudhomme’s generation were sent beyond Louisiana for their education. A number went to school in Bardstown, Kentucky, where the first Catholic diocese in United States territory west of the Alleghenies had been established and where boys could board at St. Joseph’s and girls at Nazareth Female Seminary, founded by a new American order of nuns, the Sisters of Charity of Nazareth. The 1850 census for Nelson County, Kentucky, lists 120 females nineteen years and younger living at the Nazareth Female Seminary. Of those, 29 (24%) show Louisiana as their birthplace (US GenWeb Census Project). At least three of the surnames – Bossier, Lambre, and Lecomte – are part of the Prudhomme family. The 1850 census reports 130 students in residence at St. Joseph’s for boys, fifty-five of them (42%) listed with Louisiana as their birthplace (US GenWeb Census Project). At both schools most of the other students had Irish, English, or German surnames, and most came from Kentucky, Mississippi, and Alabama. In Catholic schools outside of Louisiana, creole children learned English well and associated with the children of other prosperous Catholics, often from other parts of the South. Although items in the Prudhomme papers show that French grammar, literature, and history were subjects of study, in the Bardstown schools in the 1850s, the language of instruction was English, and English composition was emphasized. Pauline Bossier’s composition book of 1852 included essays on “Filial Affection,” “Who Approaches Nearest to Happiness on Earth?” and “Is Happiness Necessary to the Enjoyment of Life?”. Under the title “The Pleasures of Memory,” one of Pauline’s compositions uses President George Washington as an example of someone whose noble behavior must have yielded joyful memories: How much joy must not the noble, the illustrious Washington have experienced, when his memory took him back to the time, when he attained the height of his glory, when he promoted to his country’s good, by subjecting himself in every thing for so noble a cause. (Folder 96)

English/French bilingualism in nineteenth century Lousiana

349

Pauline’s last entry is a personal note: “We have read our last compositions for the year 1852. Great many of the sisters were there and Mother C. did something or at least said something I did not like at all. Oh! I am glad to think I will have six weeks to rest in, I mean I will not be bothered by those abominable compositions …” (Folder 96). At least two of the boys represented in the Prudhomme collection went to school in New England, Henry Herzog – the Henri who had to learn English before tackling Latin at Bardstown a decade earlier – and J. Alphonse Prudhomme. A year after his mother’s death, Alphonse, then fifteen years old, was attending the Commercial Institute of New Haven in Connecticut. A letter from Wm. H. Russell of New Haven to Phanor Prudhomme, dated October 18, 1853, acknowledges receipt of payment and adds, “Your son, I am happy to say, is in good health. We wish all our pupils were as amiable and correct as he” (Folder 144). In 1856 Alphonse went from New Haven to the University of Virginia in Charlottesville for two years, then to the University of North Carolina in Chapel Hill for two years, where he graduated with a Bachelor’s degree in Science in 1860. His course notebooks are entirely in English. One, filled cover to cover with chemical descriptions of candles, inks, gunpowder, pottery, and so forth, inside the back cover contains – as if a doodle – two instances of Sault Salt, elaborately written with flourishes (Folder 116). Perhaps in an idle moment his thoughts turned to Louisiana and brought the French word to mind. The year before the outbreak of the Civil War, Alphonse’s younger brother attended Georgetown College in Washington, D.C. When the girls finished school, they returned to Natchitoches Parish to await marriage, spending their time reading, playing the piano, doing needlework, making scrapbooks, and keeping in touch with each other for news of the latest parties or budding romances. They often spent several days at a time visiting each other’s plantations or meeting at their grandmother’s house in Natchitoches. In those days, a trip from one plantation to another or from a plantation in Bermuda to town thirteen miles away could not be undertaken on a whim. The young women kept in touch by mail or by the kindness of friends and family willing to deliver their notes and letters. The Prudhomme Family Papers contain a series of about thirty letters dated 1858 and1859 from Coralie Buard to her cousin Fanny Bossier. They are all in English, with just a handful of French sentences and phrases. The code-mixing in the September 12, 1858, letter from Coralie to Fanny is typical.

350

Connie Eble

There were no Creole gentlemen present, excepting our dear cousins (Herzog) and Felix. The party was gay and went off better than we had expected. John C.[loutier] was anxious to go, but on account of his brother’s death he was obliged to sacrifice his pleasure. He was seen at the door and window during the party. Ainsi, tu vois que son envie etait forte. As for the fishing party last Thursday, we also enjoyed it better than we had expected in the company of American girls and gentlemen. (Folder 1)

The content shows that despite their schooling away from their Frenchspeaking home and their use of English with each other, the young women still considered themselves culturally distinct from the monolingual English speakers that by then formed the majority of the population of Northern Louisiana where they lived – still in 1858 referring to them as American. For the young Red River creoles of Alphonse’s generation, their boarding school experiences made English their primary language. At the same time, at home in Natchitoches Parish, their parents – also bilingual – were becoming accustomed to an English-speaking environment. Many of their slaves spoke English, as did the overseers they hired to run their plantations and the people they dealt with producing and marketing their cotton. By March 1860, even the flier announcing the opening of the Natchitoches Races at the Prudhomme Course was in English only, despite the fact that the 114-member Natchitoches Jockey Club that sponsored the races was a bastion of old line creoles. Among the surnames on the subscription list were those that appear in the Prudhomme Family Papers: Bossier, Breazeale, Buard, Cloutier, Deblieux, Hertzog, Lambre, Lecomte, Metoyer, Prudhomme, Rachal, and Sompayrac. After 1870, the Prudhomme Family Papers contain almost no items written in French. The exceptions are a handful of letters and a few poems, prayers, and good wishes on the occasion of baptisms, weddings, and anniversaries. The generations following continued to study both French and English at school. Desirée Archinard Locoul’s daughter Laura attended Mrs. Cenas’ School for Young Ladies in New Orleans about twenty-five years after her mother had attended the Sacred Heart Academy in Natchitoches. Laura took almost the same subjects as her mother had, but Laura’s 1877 report card was in English and listed the subjects in English. Laura earned a perfect grade of 10 in Personality, in Manners and Conduct, and in Music. She scored 9 in English Recitation and in English Written Exercises and 8½ in English Elocution. Her grades in French were slightly lower, with 8 in

English/French bilingualism in nineteenth century Lousiana

351

French Grammar, in French Written Exercises, and in French Elocution and 7 in French Recitation (Gore 2001: 72). By 1920 when Alphonse and Elise’s granddaughter Adele Prudhomme was a student at St. Mary’s Academy conducted by the Sisters of Divine Providence in Natchitoches, French was still a required subject. But school notebooks show that the grammatical explanations of French were in English. In an eight-stanza poem that Adele wrote describing the school day, she does not mention French but gives English a complete stanza: And then of course there’s English A study all our own you know, Because it is our own language. Like French is for the French and so. (Folder 120)

In the twentieth century, English was the language of the white creoles of the Red River area. As shown by Adele’s verse, they learned French as a foreign language, as the language of a country across the Atlantic Ocean, just as other Americans did. The Prudhomme Family Papers tell the story of at least a half century of sustained English/French bilingualism only indirectly. In the documents, the two languages are kept separate, with almost no interspersing of the words of one language into the other. The collection contains few examples of grammatically faulty English or French, with the exception of stray miswritings and variable spellings. Through a large part of the nineteenth century, bilingualism in standard varieties of French and English was an important property of the network of white creoles of privilege in the Red River area. Knowledge of French differentiated them from the American immigrants who quickly outnumbered them after the Louisana Purchase. Knowledge of English differentiated them from the uneducated monolingual speakers of French like the Acadians and black creoles. Their bilingualism served their status as elites. Within a decade after the end of the Civil War, the Prudhomme Family Papers contain French only in devotional materials, phrases in memory books, and a very small number of communications between the generations in the creole network. In northern Louisiana in the public sphere, French had disappeared. After about 1870, whatever remained of either spoken or written French leaves no traces in the Prudhomme Family Papers.

352

Connie Eble

The documents left to posterity by families, like the Paston and Cely letters of fifteenth century England, have long been valued as evidence of past forms of the language. This study uses a collection of nineteenth century family papers from Natchitoches, Louisiana, not for analysis of the linguistic forms but for evidence of shift from French to English. The archive is treated as a network of the individuals whose lives are represented within its documents. The dominant cluster within the white creole network thus constructed, which is both dense and multiplex, is one of kinship. The original language of the network was French, which was retained within the network long after French had been displaced by English as the sole language of the majority of the population of northern Louisiana and the language of the public sphere. But by the end of the nineteenth century, members of the close-knit creole network also wrote only English. The strength of the ties within the kinship network had probably not diminished. However, weaker ties to individuals outside the cluster had facilitated the importation of English as an addition to French. The weaker ties were formed for economic reasons, i.e., to ensure the continued prosperity and privileged status of the members of the creole network, the founder population, after English-speaking newcomers became the majority in the area. In particular, sending creole children away to schools where they were taught to speak, read, and write English put the members of the younger generation in a position to serve as agents of innovation to their dense and strong kinship network in Louisiana. The English language became one of the properties of the linkages among the young members of the creole network and set them apart from their French-speaking contemporaries in Natchitoches Parish who were less privileged. By the beginning of the twentieth century, the generations of creoles for whom bilingualism had been a mark of class had passed away, and the creoles joined the monolingual English majority.

Notes 1. In this essay, creole means ‘native to the colony or descended from natives to the colony’, i.e., in Louisiana before 1804. 2. Because archives are organized according to the principle of provenance, all the documents saved by a particular family and entrusted to a library or museum are ordinarily kept together rather than dispersed by subject matter. This principle of organization does not facilitate research on a particular topic, e.g., the vocabulary of disease in 19th century America or thorough-bred racehorses

English/French bilingualism in nineteenth century Lousiana

3.

4.

5.

6.

353

in the antebellum South – though the writings preserved within the Prudhomme Family do hold such information. However, collections of papers saved within a family and transmitted to succeeding generations can be a good source for reconstructing past networks of who communicated with whom, how often, and about what. Furthermore, the documents saved by a family often include writings by individuals not linked by kinship, making the network represented by the papers larger than the biological family who preserved them. This is the case with the Prudhomme Family Papers. In the discussion period following the presentation of this paper at the SHEL4 conference, Ray Hickey suggested that local court documents such as depositions would likely give a better glimpse of the French used in the area. He is undoubtedly right. However, because this essay is in large part an attempt to explore whether or not family archives are amenable to a network approach, I have confined this study specifically to the network delineated by this one family’s papers. Lesley Milroy’s 1980 (2nd ed. 1987) Language and Networks is the foundational document of this approach. Chambers (1995) and a 20-page essay by Milroy (2002) survey the scholarly work in social network analysis of the past two decades in the context of the major issues in sociolinguistic theory. 22 March Began to plant cotton 25 July Good weather after much rain November 10 Gave shoes to the house servants December 15 Gave shoes to the field Negroes My dear Papa, I received your last letter of August 24. You told me that you believed that someone helped me write my letters. But I assure you that I wrote them with my own hand. I am surprised myself that I have so quickly learned the shape of my letters. I could scarcely write one letter when I arrived at college. But, my dear father, I assure you on my honor that no one helped me … . I am studying only English and French. That’s because one has to know English well to study Latin and Spanish. Reading in French, Reading in English, French Grammar, English Grammar, English Translation, and French Translation

References Aycock, Joan Marie 1993 The Ursuline School in New Orleans, 1727–1771. In Cross crozier and crucible: a volume celebrating the bicentennial of a Catholic diocese in Louisiana. Glenn R. Conrad (ed.), 203–218. New Orleans: the Archdiocese of New Orleans and the Center of Louisiana Studies.

354

Connie Eble

Breton, Roland J. L. 1979 Géographie du français et de la francité en Louisiane.Quebec: International Center for Research on Bilingualism. Calan, Louise 1937 The Society of the Sacred Heart in America. London: Longmans, Green, and Co. Chambers, J. K. 1995 Sociolinguistic Theory. Oxford: Blackwell. Gore, Laura Locoul 2001 Memories of the old plantation home. Vacherie, LA: The Z Company, Inc. Gumperz, John J. 1982 Social Network and Language Shift. In Discourse Structures. John J. Gumperz (ed.), 38–58. Cambridge: Cambridge U Press. Lippi-Green, Rosina 1989 Social network integration and language change in progress in a rural alpine village. Language in Society 18: 213–234. Marmillion, Norman and Sand 2001 A creole family album. In Memories of the old plantation home. Laura Locoul Gore, 115–166. Mills, Gary B. 1977 The Forgotten People: Cane River’s Creoles of Color. Baton Rouge: Louisiana State University Press. Milroy, Lesley 1987 Language and social networks. 2nd ed. Oxford: Blackwell. 2000 Social network analysis and language change: introduction. European Journal of English studies 4: 217–223. 2002 Social Networks. In The Handbook of Language Variation and Change. J. K. Chambers, Peter Trudgill, and Natalie Schilling-Estes, (eds.), 549–572. Oxford: Blackwell. Picone, Michael 2003 Anglophone slaves in francophone Louisiana. American Speech 78: 404–433. Smith-Thibodeaux, John 1977 Les francophones de louisiane. Paris: Editions Entente. Trépanier, Cécyle 1988 French Louisiana at the threshold of the 21st century. PhD. dissertation, Department of Geography, Pennsylvania State University. Trout, Robert O. 1996 The origin of the pioneer population of north central Louisiana hill country. In A refuge for all ages: immigration in Louisiana, The Louisiana Purchase Bicentennial Series in Louisiana History,

English/French bilingualism in nineteenth century Lousiana

355

Vol. 10. Carl A. Brasseaux (ed.), 470–477. Lafayette: Center for Louisiana Studies, University of Southwestern Louisiana. Prudhomme Family Papers (#613) 1765–1997. Southern Historical Collection, Wilson Library, University of North Carolina at Chapel Hill. [collection description: www.lib.unc.edu/mss/inv/htm.00613.html] USGenWebCensusProject. 1850. Federal Census, Nelson County, Kentucky, District 1. www.rootsweb.com/~census (Accessed on September 10, 2005).

Taking permissible shortcuts? Limited evidence, heuristic reasoning and the modal auxiliaries in early Canadian English*

Stefan Dollinger 1.

Introduction

The present study focusses on the ‘other’ North American variety of English, Canadian English (CanE) from a diachronic perspective. While CanE is a relative newcomer to the field of linguistics, diachronic studies in realtime, as Brinton and Fee (2001: 426) point out, have been almost entirely missing. Much of what is known about historical CanE is in the field of lexicology, as a direct result of the work on the Dictionary of Canadianisms on Historical Principles (Avis et al. 1967), which has become seriously out of date.1 In the present study, an attempt is made to illustrate a kind of “bad” data problem that is of particular relevance to the study of Late Modern English (LModE) varieties in colonial contexts. The modal auxiliaries in Ontario English (OntE) in the period from 1776 to 1849 shall serve as a test case here and are surveyed in relation to British English (BrE), with some comparisons to American English (AmE). One of the problems diachronic linguists are confronted with in many colonial contexts, such as in early Ontario, is a large array of input varieties and only exemplary real-time evidence that is available in machine-readable data formats. In this paper, an approach is suggested that aims to circumvent these data limitations in colonial contexts for the period between 1700 and 1900. The focus of the present paper is therefore not on the variables and their developments as such (see Dollinger 2008 in this respect), but on what is best called a heuristic method to overcome, at least in part and as an approximation, a) the lack of readily-available data, b) low token discourse frequencies and c) genre-specific variation that makes the characterization of linguistic variables within national varieties a particular challenge.

358

Stefan Dollinger

After a case study on the modals CAN and MAY that illustrates the approach adopted here, the results of this kind of reasoning will be presented for eleven modals and semi-modals in OntE in relation to BrE and an attempt will be made to assess their behaviour in the light of later developments on a cline from conservative to progressive behaviour when compared to BrE. It will become clear that we will inevitably have to rely on some kind of heuristic reasoning in order to arrive at generalizations. A heuristic process can be described as A process, such as trial and error, for solving a problem for which no algorithm exists. A heuristic for a problem is a rule or method for approaching a solution. (Blackburn 2005: s.v. heuristic).

For our purposes, we can equate the term “algorithm” with fully-fledged, empirical, corpus-based study, which is firmly based in the philological tradition that has traditionally been centred on a standard variety. Heuristic reasoning seeks to make the best of limited data and aims to allow statements on the development of colonial varieties when not all data are available. Heuristic principles, while used pervasively in other disciplines, such as in engineering, computer science, social sciences (see Michalewicz and Fogel 2004, Tversky and Kahneman 1982), have not yet been explicitly applied to English historical linguistics (while its principles may have played a role) because, perhaps, they do seem to be incompatible with the standard of description that is usually found in English historical linguistics. The test case presented here, however, will make it clear that the time may be ripe for embracing “good enough” solutions and approximations more directly even in this discipline, and thus follow the lead of the “harder” sciences. It will be shown that heuristic means are a kind of reasoning that may provide fairly reliable shortcuts to answers to research questions that would otherwise be years, if not decades, away. Two kinds of “bad data” scenarios are usually discussed in historical linguistics. While both points are not new, I will refer to two more recent contributions to this discussion. First, the problem that historical documents “survive by chance, not by design”, often containing “a normative dialect” and not the vernacular (Labov 1994: 11) is something one has to deal with. This problem becomes less pervasive in the LModE period, when data produced by minimally-schooled writers become available, which has proven to be a rich resource. Second, Nevalainen (1999) addresses the problem of acquiring social information on the informants and points out that in Early

Taking permissible shortcuts?

359

Modern English (EModE), and to an increased extent in LModE, the chances to reconstruct the social milieu of a writer increase dramatically. In the colonial context, however, I would like to suggest a third “bad” data problem which is revealed in scenarios of new-dialect formation (Trudgill 2004) and the complex nature of input varieties in colonial Englishes. Put in a nutshell, the problem is the sheer variety of input varieties that came to form new colonial Englishes. To exemplify the problem let us review briefly the external language history of the variety for the sociohistorically most important variety of Canadian English, Ontario English. 1.1. Late Modern English, colonial English, Ontario English The Late Modern English period (LModE) has only fairly recently become the object of linguistic enquiry. Although Jones (1989: 279) referred to the eighteenth and nineteenth centuries as “the Cinderellas of English historical linguistic study” not long ago, the late 1990s saw a considerable increase of research activity.2 The LModE period coincides with an accelerated spread of colonial Englishes and entails a focus on what Clyne (1992) refers to as “non-dominant” varieties of English, i.e. varieties other than British and American English. Indeed, Trudgill (2002: 44) proclaims that for the period after 1700 there is “no real excuse” to confine one’s focus to the two dominant varieties of English in more general works of English historical linguistics. Little pre-1900 historical work has been done on the historically most important variety of CanE, namely OntE, which provided the model for Standard Canadian English as an urban middle class dialect (cf. Chambers 1998: 252). Thomas (1991) is a notable exception. Based on linguistic atlas data from the 1950s and employing the apparent-time design, he traces the development of Canadian Raising back to the late nineteenth century. Chambers’ (1981, 1993, 2004) work on language attitudes in Victorian Ontario adds another important layer to the available documentation, as does Hultin (1967). M. Bloomfield (1948) and Scargill (1957), the two classic contributions on the origins of CanE, are, however, based on settlement history alone and are devoid of linguistic data. Only recently, a realtime study of pre-Confederation OntE on the modal auxiliaries from the viewpoint of new-dialect formation (Trudgill 2004) has been completed (Dollinger 2006b, 2008).

360

Stefan Dollinger

1.2. External history The external history of Canada is divided into four distinct settlement waves (Chambers 1991, 1998). Because of the temporal focus on preConfederation (pre-1867) OntE in the present paper, settlement waves I and II are of importance here. Wave I is comprised of early immigration from the United States into Ontario, from roughly around 1776 to 1793, with a small trickle continuing until 1812, when war with the United States broke out. This early immigration resulted in the first peopling of Ontario and the steady transformation of the “primeval forest west of Montreal” that had previously been inhabited by only “a few hundred English-speaking” people (Orkin 1970: 52f), most of whom were transient fur traders. This first wave, however, was a rather heterogeneous mix of “American loyalists”. Besides native English speakers, most of them of early AmE, there were significant elements of disbanded German soldiers (including some Swiss German speakers), large contingents of Scottish Gaelic speakers and sizeable Dutch elements, which were joined by some Scots speakers, Irish English speakers and French speakers (Dollinger 2008: 64–78). As those minorities came early, they are expected to have had an influence on early CanE via language contact phenomena, which included L2 varieties that waned in the late 1800s. Wave II is comprised of post-1815 immigration. After the War of 1812 had come to an end (in 1814) and the Napoleonic wars had ended in Europe, the European population surplus was dumped on the colonies. This wave is characterized by massive immigration from the British Isles (although sizeable immigration from the German states is reported prior to 1850, Bausenhart 1989). The second wave included Scottish Gaelic speakers and Scottish English speakers, speakers of Northern English dialects, large numbers of Southern Irish and Ulster Scots (from Northern Ireland and also from the USA) and only a minority of Welsh, southern English dialect speakers and AmE speakers. In 1812 the Ontarian (then Upper Canadian) population was somewhere in the vicinity of 83,000 inhabitants (Gourlay 1822: 612), the majority of whom were Americans (cf. Akenson 1999: fn 117). This demographic make-up was changed by wave II. Figure 1 shows the relative population input to Ontario for the years 1829–1859, giving an impression of the numerical swamping of the American base with immigrants from the British Isles. For instance, immigration in 1829–31 alone outnumbered the entire population of the province in the

Taking permissible shortcuts?

361

year 1812, with almost 16,000 coming in 1829, 28,000 in 1830, and 50,000 in 1831, and so it continued, with ups and down, in that fashion: Arrivals at Port of Quebec from overseas

Percentage of all immigrants

80 70 60 50 40 30 20 10

England %

Ireland %

Scotland %

Europe %

1859

1857

1855

1853

1851

1849

1847

1845

1843

1841

1839

1837

1835

1833

1831

1829

0

Maritime

Figure 1. Arrivals at the Port of Quebec 1829–1859 (raw data from Cowan 1961: 289, table II; classification based on ports of departure)3

We have some evidence from recent socio-historical studies that the overwhelming majority of these second wave immigrants did, contrary to one’s intuition, not settle in the established centres founded by the American loyalists. Akenson (1999: 36) reasons that for Irish immigrants, who represented the bulk in the later years in figure 1, “one has to conclude that the overwhelming majority of Irish migrants to Ontario settled in the countryside”. Indeed, there was plenty of unsettled land in Ontario at the time, just waiting to be cleared (Wood 2000: 29–31). This finding has important implications from a sociolinguistic point of view, as it contradicts the otherwise logical reasoning that most of these immigrants would have settled “in towns and villages founded by the Loyalists” (Chambers 1998: 252). It also implies that AmE varieties – both native and L2 varieties – would have tended to be used in Ontario’s centres

362

Stefan Dollinger

and that dialects from the British Isles would have been evident in the countryside, before they eventually would have merged (except in the case of Ontario’s linguistic enclaves such as the Ottawa Valley or Peterborough). Linguistically, however, the second wave has usually been considered from a twentieth-century perspective as having had only a highly limited influence on OntE (Avis 1978: 4, Chambers 1998: 263). Real-time data show some patterns that do not necessarily corroborate the limited influence of BrE variants for all areas of modal usage (Dollinger 2008: 235–243, data on SHALL and WILL). 1.3. Data and variables The data come from three corpora: the Corpus of Early Ontario English, pre-Confederation section (CONTE-pC) (see appendix 3), provides the CanE material, A Representative Corpus of Historical English Registers, version 1 (ARCHER-1)4, the AmE and BrE data, and the Corpus of Late Eighteenth-Century Prose (CL18P)5 is used for data from NW England. Table 1 summarizes the varieties, corpora, genres and periods for the available data: Table 1. Overview of corpora, periods and genres (shaded areas: nonexistent corpus data)

Genres Period 1 Period 2 Period 3 Period 4+5

AmE BrE CanE6 (CONTE-pC) (ARCHER-1) (ARCHER-1) letters, diaries, newspapers 1776–1799 1750–1799 1750–1799 1800–1824 1800–1849 1825–1849 1850–1899 1850-1899 1850–1899

NW-BrE (CL18P) letters 1761–1790

While most findings are based on periods 1–3, period 4+5 was used as an additional benchmark in AmE in one instance (item no. 6 in Table 3). Period 2+3, which is referred to later on, comprises the years 1800–1849. Grey shadings in Table 1 indicate data that are not available. While these gaps are apparent, the bigger gaps are not shown in Table 1 as they concern the input varieties for which we do not even have corpora. When aiming to characterize regional varieties (or national varieties),7 we would wish for corpus data from all input varieties. As we know from the external history

Taking permissible shortcuts?

363

in Section 1.3, this is a formidable task. Given the immensity of data needed to compare colonial varieties with their input, there is of course the question whether we may be in the position to produce findings at all. In order to fully address questions such as the influence of input varieties on newly-formed dialects or the conservatism or progressivism of a variety in relation to another one, we would need a complete set of input data corpora, which would include a number of varieties in the Ontarian context. For the pre-1800 period, besides AmE, (standard) southern BrE and north-western regional BrE, we would need to have Irish English data, Scottish English, L2 varieties of Scots Gaelic speakers, German, Dutch, French speakers and First Nations speakers, among others. For the post1800 period, we would require data from Northern British English, southern BrE, both Southern as well as Northern Irish English (Ulster Scots), AmE, Scottish and Scots Gaelic L2 speakers, AAVE and First Nations in Canada and L2 varieties from immigrants (differences in the genres in different varieties, e.g. personal letters, further complicate the picture and are almost beyond the control of the researcher). One way towards problem solving in heuristics is to consider the available data carefully, which is easily illustrated by examples from logic (see Michalewicz and Fogel 2004: 9). In historical linguistics the data may be – within the limitations of “bad” data outlined above – “available” in libraries and archives, but not readily accessible. Often this catch is interpreted as a deficit of the researcher, and not as a result of the vastness of the task of data mining. While it is always a good idea to increase one’s baseline data, I would argue that we should seriously consider heuristic approaches in historical linguistics and make use of “good enough” solutions that have proven useful in other disciplines. Related to the data question is the problem of quantitative vs. qualitative studies. Concerning the former, the cut-off point of what constitutes eligible frequencies (e.g. at least 10 tokens) is, of course, inspired by the available corpus sizes. The modal auxiliaries have usually tended to yield, on the whole, acceptable frequencies in the standard corpora. But what are we to do if conventional corpus sizes do not produce the expected token counts? I would argue, again thinking heuristically, that there may be an alternative to going straight back to the archives in search of more data. This line of reasoning, making the proverbial best of “bad” data, will be illustrated in the quantitative treatment – and statistical testing – of limited data that, combined with Present Day English (PDE) findings, may prove

364

Stefan Dollinger

just enough to characterize the LModE development of the modals in the Ontarian context. Even in the LModE period the modal auxiliary complex undergoes considerable changes in usage and most of those changes tend to be stylistic rather than categorical in nature (Denison 1998: 165, 1993). In Section 2, CAN and MAY serve as examples for the kind of heuristic reasoning used for an assessment of whether CanE showed progressive or conservative behaviour in the modal auxiliaries in relation to BrE. In Section 3, the overall picture for the modals in early OntE will be presented.

2.

A close-up: CAN and MAY in early OntE

The early development of CAN and MAY is well documented. In OE mæg, the formal ancestor of MAY, is prevalent in the sense ‘to be able to’. Following a pragmatic/semantic cline of grammaticalization, this ability use was extended to denote possibility and was subsequently developed into permissive uses. Permissive uses can be occasionally found in OE (Traugott 1972: 71f), but the “full performative use of may, as in You may go ‘I permit you to go’, did not gain wide currency until the sixteenth century” (Traugott 1972: 118). OE cann, on the other hand, originally meant ‘to know, be acquainted with, know how to’. Once CAN acquired the meaning of ‘to be able to’, MAY gave up the meaning of ‘to be able, be strong’ (Kytö 1991: 65). After this semantic change, it was only a question of time before the meanings of ‘possibility’ and finally ‘permission’ developed by implication, i.e. pragmatic factors in the context favouring these interpretations rather than ‘ability’ readings (Kytö 1991: 65). CAN and MAY, revolving around the notions of ability, permission and possibility, have been studied at the synchronic level and for older language stages up to EModE (e.g. Coates 1983; Facchinetti 2003, 2002, 2000; Kytö 1991: 81–258, Warner 1993: 176–178, Denison 1993: 292– 325), while in LModE general developments have been identified in the area of permission (e.g. Traugott 1972: 170f; Kytö 1991: 65; Denison 1993: 303). In comparison to the long-term development from OE to EModE, the study in LModE has not been pursued as rigorously. Simon-Vandenbergen (1984) seems to be the only quantitative study of CAN and MAY that includes semantic categories as well as a sizeable proportion of LModE texts (based on personal and official letters and drama data). She shows (1984: 364) that root uses of CAN, in the sense of permission, first occur in nine-

Taking permissible shortcuts?

365

teenth-century plays and encroach on the territory of root MAY. A number of examples are found in the three corpora that illustrate this shift of functions of CAN, as shown in example 1(a–c): (1) a. ability (usually with animate subject): I fancy Kitty can do nothing better … (ARCHER-1, BrE, 1750–99, letter section) b. neutral, root possibility (ambiguous case): I should like to by it if I can by it two Advantage (CONTE-pC, 1825–49 [a minimallyschooled letter writer]) c. permission: We have decided upon my sitting with Mamma every night during tea and Minnie during dinner as then I can read Mamma's prayers to her. (CONTE-pC, 1825–1849; a teenaged diary writer) [the girl is allowed to read to her sick mother then] As pointed out by Coates (1983: 139 for 1960s BrE), 1 (b, c) illustrate the core functions of MAY, which means that CAN was beginning to compete with MAY in the latter’s core domain. This long-term developmental scenario provides the background for the identification of progressive and conservative forms – CAN as progressive, and MAY as conservative – with CAN conveying more informal undertones that MAY (Coates 1983: 103). Table 2 shows the areas of competition for CAN and MAY in the late eighteenth and nineteenth centuries and gives examples of their uses. The grey shaded cells highlight the semantic areas in which CAN and MAY compete. The example in MAY/permission can be paraphrased in PDE by CAN, and it is possible to substitute CAN/permission with MAY without a significant change in meaning. MAY/root possibility can again be rephrased with CAN, while CAN/root possibility with MAY is more of a borderline case, but acceptable in some varieties. We are attempting to trace changes in these two areas, which expanded in LModE, in CanE in relation to BrE. As MAY had become obsolete as a marker of ability in ME and CAN was not used for epistemic possibility, the competition is limited to the areas of root possibility and permission. From a twentieth-century perspective, we know that CAN was to gain the upper hand in both permission and, to some extent, in root possibility meanings, as a result of influences from informal genres (see Coates 1983: 106f) and is now showing first uses in epistemic readings (Coates 1995 cites first examples of 1990s

366

Stefan Dollinger

AmE for epistemic CAN, Facchinetti 2000 finds first examples in early 1990s BrE). Table 2. CAN and MAY – semantic categories and examples Prototype FORM MAY

CAN

NOTIONAL FUNCTION Permission

EXAMPLE

If he show any Disposition to write me a penitential Letter, you may encourage it; not that I think it of any Consequence to me, but because it will ease his Mind and set him at rest. (BrE-1) Root possibility Any person wishing to purchase may depend upon getting a great bargain (CanE-2) Epistemic … and my ideas upon the several points which possibility may, between this and then, occur to me (AmE-1) Ability Permission

I fancy Kitty can do nothing better … (BrE-1) I have your Certificate that the land is not leased or vacant of course none [of the settlers] can be located without the sanction of the Lt. Gov. as in other cases. (CanE-2) Root possibility Nothing can be more satisfactory than the readiness and unanimity with which the Legislature have applied to meet the emergencies (CanE-3)

Figures 2 and 3 show the results for CanE and BrE in periods 1 and 2+3 and for AmE in period 1. All instances of CAN are shown in relation to MAY; the functions of ‘permission’ and ‘root possibility’ are shown by genre, as the choice of CAN and MAY seems to have been influenced by text type (See note8 for abbreviations used), but we will need to summarize the usage across different genres later on. First, we will focus on root possibility uses of CAN, as shown in Figure 2. By comparing the CanE with the BrE data, we can see a picture that is largely parallel in incidence in the diary and letter genres. Only in the newspaper genre the pattern diverges. The AmE data are closer to the CanE values across all genres, which indicates the loyalist base of CanE. Viewed in the larger diachronic context, CanE seems more progressive in its use of CAN in diaries (i.e. closer to 1960s BrE usage, where CAN is predominantly used in possibility readings, Coates 1983: 101), but more conservative in letters and newspapers than BrE. The apparent tendency for an increase of CAN in diaries and letters is, however, not confirmed by statisti-

Taking permissible shortcuts?

367

cal testing, which yields no significance for an increase in CAN for the changes in Figure 2 at the 95% level (Appendix 1). From our background knowledge, these data seem to suggest, by heuristic reasoning, a drift scenario in diaries and letters, even though statistical testing does not help us here. The different developmental patterns in CanE and BrE newspapers could either be a result of chance, or alternatively, as it shows the statistically strongest change (Appendix 1, n–1 and n–2+3), a stylistic change towards more BrE norms in Canadian newspapers. CanE uses MAY more often than CAN in letters, but not in diaries, which indicates a more formal style in CanE letters than in British ones. It is probably best to interpret these figures as reflecting instances of directional drift and stylistic variation in CanE, with more formal, possibly more conservative tendencies in the letter and newspaper genres. We can say that ‘root possibility’ in newspapers tended to be expressed by MAY in CanE, as opposed to BrE, but with both varieties converging in the 30% range in period 2+3 (Figure 2). Overall, this change at the time can be perceived as a change from above the level of consciousness, as the use of CAN is found be to documented – and castigated – by eighteenth-century grammarians (Sundby et al. 1991: 211). For the second function in which CAN and MAY overlap, permission uses, a change has been reported for LModE. Figure 3 shows the development for CAN denoting permission and we see an increase of CAN in both CanE and BrE in all genres. While differences in percentages appear in part to be considerable, again, no change of CAN is statistically significant (cf. appendix 2), which is partly a result of low token incidences. In 1960s BrE, informal texts tend to show more uses of permission CAN (Coates 1983: 101), but are generally used less frequently than possibility readings of CAN. Using a different classification system, Facchinetti (2002: 239) shows for early 1990s BrE that 5% of all uses of CAN are deontic readings, which are one core of our permission uses. What should we make of these results? It is clear that CAN is expanding its use in both CanE and BrE in the areas of permission and, for diaries and letters, root possibility uses. But neither the differences between CanE and BrE in each period, nor their increases from period 1 to 2+3 are significant, and statistical testing (neither chi-square nor Fisher’s Exact) does not help us much to confirm a long-term change corroborated in previous studies. The change appears to be too slow to reach the 95% significance level in 25-50 year periods in the data. We know by hindsight, however, that in twentieth-century BrE, CAN is usually the default variant and not MAY,

368

Stefan Dollinger

which “is marked for formality” (Coates 1983: 103 for BrE, Ehrman 1966: 12 notes the use of permission CAN in 1961 AmE dialogue data). With this knowledge, we would interpret the diachronic changes in Figures 2 and 3 as a LModE parallel development in both CanE and BrE leading up to the PDE distribution, despite their lack of statistical significance in our data. This ‘heuristic’ conclusion could easily be tested in a quantitative framework. As Appendices 1 and 2 show, the token frequencies are low. In BrE diaries (Appendix 1, first table), which shows a solid relative increase of root possibility CAN from period 1 to 2+3, the chi-square value is 1.04, and therefore short of the 3.84 required to reach significance at the 95% confidence interval. If we quadrupled the token frequencies, we would reach a value of 4.18, and this gives us an idea of how much data are needed. We would need at least four times as much as we have now, or c. 500,000 words for the late eighteenth and early nineteenth century to be able to base our study on firm statistical grounds. And here, even CONCE (Kytö, Rudanko, Smitterberg 2000: 89), the biggest corpus of nineteenthcentury BrE to date, would provide too little data for our three genres to meet the baseline data criterion (although the c. 250,000 words of letter data are a substantial body of evidence). Clearly, the data mining needed to test our heuristic reasoning on strict quantitative terms would be a project on its own. The examples show that changes which appear to be instances of slow moving drift or parallel development do not necessarily reach levels of significance in the data. We have also said, however, that based on our knowledge of the further development, it is justifiable to interpret the data as exhibiting a parallel development, since five out of six instances of a change point in the same direction (all except for the CanE and BrE newspaper data). This approach is best described as heuristic reasoning. However, we do not have AmE data from the early nineteenth century to confirm our reasoning of drift in the last instance, and we are at a loss for other input varieties of OntE (such as Scottish English, Ulster English or Irish English, or post-1800 Northern English). We do know, however, that in twentieth-century BrE, CAN has become the default form for root possibility meanings (cf. Coates 1983: 101) and more and more in permission uses, despite its condemnation by prescriptive grammarians. The method applied here is hardly new as it has been implicitly applied in many diachronic studies. What is new, though, is the claim that we cannot, and should not, expect studies to include data that are, strict sensu,

Taking permissible shortcuts?

369

'Root possibility' CAN (vs. MAY) in three varieties 100 80.0 80

60.0

percent

29.4 60

70.4

55.6

60.9 52.9

50.050.0

51.6 42.9

39.5 33.8

40 20.0

CanE BrE AmE

22.7

20

0 d1

d2+3

l1

l2+3

n1

n2+3

genres & periods

Figure 2. Diachronic development of CAN and MAY coding for ‘root possibility’ in CanE and BrE (AmE-1 for comparison) 'Permission' CAN (vs. MAY) in three varieties 100.0 100.0

100.0 100 80.0

75.0

80

percent

60.0 60

CanE

50.0

50.050.0

BrE 40

33.3

AmE

15.014.3

20 0.0

0.0

0.0

0 d1

d2+3

l1

l2+3

n1

n2+3

Figure 3. Diachronic development of CAN and MAY coding for ‘permission’ in CanE and BrE (AmE-1 for comparison)

370

Stefan Dollinger

needed to make any generalization about colonial varieties, but should embrace the educated guessing applied in heuristics. The method is an approximate comparison of the situation in early OntE with BrE (and AmE varieties) and it needs to be stressed that the method cannot rule out reverse developments. However, it would be somewhat unlikely for CAN to move backwards given clines of grammaticalization (e.g. Traugott 1989). As the method makes use of the data available and applies other sources and benchmarks, its heuristics provides us with insights into otherwise uncharted waters. Section 3 shows the results of this heuristic reasoning for eleven modals in a total of 19 contexts in early OntE.

3.

The bigger picture: eleven modals in early OntE

In this section, the nine core modals, CAN/MAY, COULD/MIGHT, SHALL/WILL, SHOULD/WOULD, MUST, plus OUGHT TO, and semimodal HAVE TO, will be presented in terms of progressive or conservative behaviour in CanE in relation to BrE. For this purpose, the results for each period (the diachronic development as well as the empirical base) have been considered, gauged and assigned to one category in the manner illustrated in the previous section. The assessment is carried out on a 5-tier grid, classifying each function into “conservative”, “neutral” and “progressive”, and “towards conservative” or “towards progressive” for intermediate cases. Clearly, this kind of attempt to classify the overall behaviour across the three genres also requires some form of heuristics, i.e. the consideration of the available data and a principle to synthesize the variables into one measure. As the empirical base for permission uses of CAN (cf. Appendix 2) is especially slim in CanE diaries and newspapers (CanE-d, CanE-n), preference is given to the letter data, which are slightly more conservative than BrE data. The overall assessment for permission uses of CAN in CanE is perhaps best described as “towards conservative” (item no. 1 in Table 3). Root possibility CAN, with diaries more progressive, letters more conservative and newspapers conservative (but approaching BrE values), is also best characterized as “towards conservative” (item no. 2). The remaining 17 contexts were assessed in a similar manner, resulting in Table 3. Table 3 shows that pre-Confederation OntE does not lean heavily in any direction. If anything, it appears to be slightly more progressive in its overall use of the modals than BrE. Interestingly, in six out of nine cases it patterns with AmE data from period 1, which linguistically confirms the loyalist base

Taking permissible shortcuts?

371

hypothesis for AmE input. The data from NW England (NW-BrE) in period 1, however, are rather distant from both CanE and AmE where available (items no. 6 and 15). Item no. 6, use of SHALL and WILL in the first person, is subject to massive change from period 1 to period 2+3, which is the reason for its occupying two slots (see note 9). Without more BrE data we cannot decide on the two categories, but it seems clear that the change was the result of regional BrE influence (Dollinger 2008). The overall pattern shown in Table 3 suggests that early CanE did not simply follow AmE or BrE usage, but was beginning to show idiosyncratic developments in the modal auxiliary complex and most likely elsewhere. For the area of vocabulary, the unique Canadian character has been long accepted (Lovell 1955: 5) and researched (Avis et al. 1967; de Wolf et al. 1997; Barber 2004). Table 3 provides some indicators for developments specific to CanE in the use of modal auxiliaries. For item no. 18, COULD in epistemic uses, for instance, the loyalist base theory of American input cannot account for this behaviour, as AmE and CanE are placed at opposite ends of the spectrum. In others areas, such as the use of first person WILL (item no.6), CanE may well have been more progressive than AmE, although this remains a hypothesis until post-1800 AmE data can be found. An overall assessment such as that represented in Table 3 allows new insights into the behaviour and genesis of colonial varieties and would warrant the element of “educated guesswork” inherent in heuristic approaches, of which the summarization in Table 3 is the result.

4.

Possible conclusions: evidence and reasoning

The overall results in Table 3 show that CanE modal use appears to have been slightly on the progressive side when compared with BrE. While largely matching AmE use, early OntE shows some developments that may have been specific to the variety. As LModE varieties are a very recent area of inquiry, and LModE dialectal variation largely remains to be studied (but see for instance Watts and Trudgill 2002), the present situation may be comparable to the state of knowledge of EModE variation in the 1980s (Görlach 1988). If one wishes to make generalizations along the lines suggested in Table 3 between a colonial variety and BrE – with some reasoning for AmE and regional BrE – some approximations and heuristic ways of reasoning are necessary

372

Stefan Dollinger

Table 3. Eleven modals in CanE in comparison to BrE (data for AmE-1 and NWBrE-1 is indicated in relation where applicable) (adapted from Dollinger 2008: 276) VARIABLES

CAN & MAY

OUGHT TO

SHALL & WILL

function / context 1)

permission

2)

root poss.

3) 4)

affirmative contexts negative contexts

5)

overall

6)

1st person

Conservative

Towards cons.

Neutral

Towards progr.

Progressive

(AmE-1) (AmE-1)

9

(NW-BrE)

(AmE-1)

nd

7)

2 person

8)

3rd person

9)

inanimate subjects

10) passive structures SHOULD & WOULD

11) hypotheticals 12) non-hypotheticals 13) root uses

MUST & HAVE TO

COULD & MIGHT

14) epistemic uses 15) ep. MUST NECESS. 16) affirmative contexts 17) negative contexts 18) epistemic uses 19) non-epistemic uses

(AmE-1) (AmE-1) (AmE-1)

(NW-BrE) AmE-1) (AmE-1)

(AmE-1)

Taking permissible shortcuts?

373

in the light of gaps in the historical corpus inventory. Given the usual production phase of a historical corpus of about three to four years and the multitude of input varieties for LModE colonial varieties, we are still some time off from more complete data sets of input varieties which were, as is well known, mainly regional dialects (Hickey 2004a: 1). We have seen in Section 1.3 that in the Ontarian context, at least ten groups in wave I and nine groups in wave II prior to 1850 would need to be considered. Of the required 19 corpora, we have just one regional corpus for period 1 (NWBrE), one AmE for period 1 and two BrE text collections (ARCHER-1 and CONCE), the latter two of which include texts that are closer to standard varieties and may not be the type of material one would ideally wish for.10 In some respects, we are 16 corpora short of being able to reach hard and fast conclusions. Moreover, given current copyright practices, we may be a couple of generations of researchers away from the necessary research tools. Circumventing this lack of resources, the approach laid out in Sections 2 and 3 outlines a method that synthesizes the OntE data with existing realtime resources in combination with studies of the variables in later periods. While this approach is somewhat compromised by the possibility of retrograde developments in the nineteenth and twentieth centuries, it offers a more precise scenario of language use in colonial varieties and their relations to dominant varieties of English.11 All that is needed is one reliable corpus of the variety to be surveyed, such as CONTE-pC, and access to the available corpora of input varieties. Findings gleaned from studies such as the present one are, given the lack of resources, likely to hold for many years to come. While the current approach employs both percentage changes in the distributional frequency of features and statistical testing, it does not necessarily rely on levels of significance in the interpretation of its findings. Given the short period intervals of 25 years, which are necessary to tap into the process of new-dialect formation (Trudgill 2004), the customary sample sizes in historical linguistics do not easily produce statistical significance at the 95% level even for relative high-frequency items like the modals. However, the percentage distributions appear to show a clear long-term trend from a present-day perspective. The biggest challenge for LModE variationist linguistics is the lack of specialized corpora for regional English dialects, and this is compounded by the lack of sufficiently large corpora for historical linguistic analysis using corpus linguistics techniques. We have seen that – in theory – at least

374

Stefan Dollinger

four times the customary sample sizes would be the base line needed to reach significance levels for the modal auxiliaries in semantic studies. However, as I have tried to show in the discussion of CAN and MAY, one can arrive at meaningful interpretations of even limited data sets by exploiting the relative temporal proximity of LModE to twentieth-century English, using Coates’, Ehrman’s and Facchinetti’s studies. As a result, given the limited resources for corpus compilation, the production of smaller, specialized corpora of more diverse LModE varieties, rather than the production of a bigger-sized corpus a single variety, should remain a priority. Logically, the “three main source regions of extraterritorial varieties of English” in the British Isles (Hickey 2004b: 33), English English (with a differentiation of north and south [Trudgill 2004]), Scottish English and Irish English (both Ulster Scots and Southern Irish English), from around 1700 to 1900, would be the first choices for new research tools for the spread of English in the LModE period. In the Ontarian context, 19th-century AmE data is another prime desideratum. While we would wish to have access to complete dialect lineages of all input dialects involved in the formation of a given variety (as called for by Wolfram and Schilling-Estes 2004: 182), corpora of LModE Scottish and Irish English would tremendously facilitate the line of heuristic reasoning proposed in Sections 2 and 3. Ultimately, only complete data sets can prevent us from drawing tentative conclusions about influences, independent developments and questions of colonial lag that may not quite stand the test of time. However, by taking a more explicit, exploratory heuristic point of view, we do not necessarily have to call it quits until these resources materialize.

Appendices: Statistical testing (95% level) For the following contingency tables, both the Chi-Square Test scores (the standard in much of corpus linguistics when statistical testing is applied) and Fisher’s Exact Test scores are provided. Please note that Fisher’s Exact is to be preferred in all cases where one of the cells contains less than five instances (Vogt 2005: 122; Oaks 1998: 25, cf. Woods et al. 1986: 144,) (which is the case in all but five tables). In all tables shown here, both tests arrive at the same conclusion about statistical significance, which supports Woods et al.’s assessment – given the “bad” data situation in historical linguistics – to “go ahead and carry out the chi-squared test even if some

Taking permissible shortcuts?

375

expected frequencies are rather too small” (1986: 145). Provided one is aware that the chi-square values tend to be rather larger than they ought to be and one would consider this in one’s interpretations, this seems to be acceptable practice. Chi-square calculations (Ȥ2) follow the test statistic used in Nelson, Wallis and Aarts (2002: 264–7) and are calculated to answer the question whether the choice of CAN is affected by the independent variable (periods 1 and 2+3, BrE and CanE respectively) and not whether the entire grammatical choice (table) is affected (see Nelson, Wallis and Aarts 2002 for an account). The latter method, usually offered by online tools (e.g. Georgetown Linguistics Web Chi Square Calculator), reaches significance more easily but says nothing about which of the two variables (CAN or MAY) is significant. Fisher’s Exact p-values are calculated with Preacher and Briggs’ (2001) online tool and show the two-tailed p-values (p), which produces values in between two more extreme values (one of which might be chosen according to the expected distribution of the variables, which was not applied in the following calculations). Appendix 1: Root possibility uses of CAN BrE-d 1 2+3

CAN

MAY

5 5

12 4

CanE-l 1 2+3

CAN

MAY

18 39

16 25

CanE-n 1 2+3

CAN

MAY

2 4

2 1

Ȥ2 = 1.04 < 3.84, not significant p = 0.23 > 0.05, not significant

Ȥ2 = 0.24 < 3.84, not significant p = 0.52 > 0.05, not significant

Ȥ2 = 0.3 < 3.84, not significant p = 0.52 > 0.05, not significant

CanE-d 1

BrE-n 1

n-1 CanE

CAN 2

MAY 8

2+3 25 49 Ȥ2 = 0.52 < 3.84, not significant p = 0.49 > 0.05, not significant

CAN 16

MAY 15

2+3 15 23 Ȥ2 = 0.56 < 3.84, not significant p = 0.34 > 0.05, not significant

CAN 2

MAY 8

BrE 16 15 Ȥ2 = 1.72 < 3.84, not significant p = 0.14 > 0.05, not significant

376 BrE-l 1

Stefan Dollinger CAN 24

MAY 16

n-2+3 CanE

2+3 19 8 Ȥ2 = 0.27 < 3.84, not significant p = 0.44 > 0.05, not significant

CAN 25

MAY 49

BrE 15 23 Ȥ2 = 0.2 < 3.84, not significant p = 0.68 > 0.05, not significant

Appendix 2: Permission uses of CAN CanE-d CAN MAY 1 1 1 2+3 1 0 2 Ȥ = 0.25 < 3.84, not significant p = 1.0 > 0.05, not significant

CanE-n CAN MAY 1 0 0 2+3 5 0 2 Ȥ cannot be calculated p = 1.0 > 0.05, not significant

BrE-l CAN MAY 1 1 6 2+3 3 2 2 Ȥ = 1.83 < 3.84, not significant p = 0.22 > 0.05, not significant

CanE-l CAN MAY 1 2 11 2+3 8 16 Ȥ2 = 1.01 < 3.84, not significant p = 0.44 > 0.05, not significant

BrE-d CAN MAY 1 2 2 2+3 4 2 Ȥ2 = 0.3 < 3.84, not significant p = 1.0 > 0.05, not significant

BrE-n CAN MAY 1 4 4 2+3 6 2 Ȥ2 = 0.4 < 3.84, not significant p = 0.61 > 0.05, not significant

l-1 CAN MAY CanE 2 11 BrE 1 6 Ȥ2 = 0.37 < 3.84, not significant p = 1.0 > 0.05, not significant

l-2+3 CAN MAY CanE 8 16 BrE 3 2 Ȥ2 = 0.78 < 3.84, not significant p = 0.34 > 0.05, not significant

Taking permissible shortcuts?

377

Appendix 3 Textual sources of the Corpus of Early Ontario English: periods, texts and sample sizes (number of words). CONTE-pC size: 125,000 words (CONTE size: ca. 225,000 words). period

newspapers

diaries

1 1776– Upper Canada Gazette, 1799 ca. 2,800 Canada Constellation, ca. 1,700

Benjamin Smith, ca. 1,800 Anne Powell, ca. 6,200

sum: 4,500 words 2 1800– Upper Canada Guard1824 ian, ca. 8,200 Upper Canada Gazette, ca. 5,000 Kingston Gazette, ca. 1,400 sum: 14,600 words 3 1825– Upper Canada Gazette, 1849 ca. 8,500 Niagara Argus, ca. 4,500 Gore Gazette, ca. 3,000

sum: 8,000 words Benjamin Smith, ca. 8,500 Ely Playter, ca. 8,500 (Eleanora Hallen, ca.3,700 not incl. in pC version)12 sum: 17,000 words Sophia MacNab, ca. 11,400 Charlotte Harris, ca. 9,200

sum: 16,000 words TOT genre total: 35,100 125,700

sum: 20,600 words genre total : 45,600

(semi-)official letters

61 letters (by 48 authors)

sum: 15,000 words 65 letters (by 48 authors)

sum: 15,000 words 77 letters (by 64 authors)

sum: 15,000 words genre total : 45,000

Period

SIN-speakers

Lower Class

Middle Class

1 2 3

2,200 1,700 1,000

3,200 1,000 3,100

11,700 14,000 12,000

Social (and regional stratification) within CONTE-pC: sample sizes for Scottish-Irish-Northern English speakers (SIN) and social class subsections.

378

Stefan Dollinger

Notes *

1. 2.

3.

4. 5.

6. 7.

The research for this paper was funded by the Austrian Academy of the Humanities and Sciences, Österreichische Akademie der Wissenschaften, DOC grant 21701. I would like to thank three anonymous reviewers for their feedbacks on an earlier version of this paper. Recently, a revision process has begun. See www.dchp.ca (21 Dec. 2006) for more details. The best indicators are probably the founding of the Late Modern English Conference series (LMEC) in 2001 and the appearance of special collections within bigger English historical linguistics conferences (Bueno Alonso et al. eds. 2007, Dalton-Puffer et al. eds. 2006), the appearance of textbooks and reference guides (Bailey 1996, Romaine 1998, Görlach 1999, 2001, Beal 2004) and major projects such as Tieken-Boon van Ostade’s ”The codifiers and the English language“ project, see http://www.ulcl.leidenuniv.nl/index.php3?m=9&c=122 (31 Jan. 2006). Inner-British migration is a factor to be considered here. While a departure port in England or Scotland does not necessarily mean that the immigrant was English or Scottish (Liverpool as a gateway for the Irish is a prime example, or Glasgow and Greenock in Scotland), “relatively few Britishers [English and Scottish] sailed from Irish ports” (Akenson 1999: 14). Cowan (1961: 287) stresses the fact the that Irish were using British ports, especially once steamboat transportation from Ireland to Britain had become available (probably around 1825) reducing the costs. This would cause the Irish element to be underrepresented in figure (1), further increasing their share of the total (later) immigration. I am indebted to Christian Mair at Freiburg University for granting me access to ARCHER-1 (compiled by Douglas Biber et al.) Compiled by David Denison, Linda van Bergen and Joana Soliva Proud. My thanks go to David Denison for granting me online access to the corpus (cf. http://lings.ln.man.ac.uk/info/staff/dd/papers/newcastle_late_18c.pdf, 24 Jan. 2006). OntE is referred to as CanE in the context of BrE and AmE, unless stated otherwise. In the past two decades, genre-specific analyses have made big strides (Biber 1988, Kytö and Rissanen 1983), and while it is self-evident that variables differ between genres, regional provenance will be foregrounded as the prime independent variable. We will therefore need to devise a means how to assess the behaviour of a linguistics variable across the genres used. In this respect, we aim to complement detailed genre analysis with statements of the overall behaviour of a variable in a given variety in relation to other varieties.

Taking permissible shortcuts?

379

8. Both the diagrams and the appendices use the following abbreviations: d = diaries, l = letters, n = newspapers, periods 1, 2+3 as defined in table (1). CanE-l would therefore mean “Canadian English letters” as used in the appendices. 9. 1st person WILL poses challenges for the 5-tier grid, without referring to the diachronic development, and shows the limitations of this, admittedly rough, assessment. 10. Note that Southern AmE corpora, SPOC and BLUR (see Schneider 2007: 355, 358) would not constitute input varieties of OntE. 11. However, given the steady, long-term development of the modals along clines (Traugott 1989, Abraham 2002), it is somewhat unlikely that reversals would have occurred. 12. CONTE-pC is the first part of the CONTE corpus (1776–1899, see Dollinger 2006a) and will be made accessible to reseachers as soon as copyright has been fully cleared and the manual has been completed (at the time being, please contact the author for more information). Cf. Dollinger (2006a: 25) for the provisional CONTE design (full version).

References Abraham, Werner 2002 Modal verbs: epistemics in German and English. In Modality and its interaction with the verbal system (Linguistik Aktuell, 47). Sjef Barbiers, Frits Beukema and Wim van der Wurff (eds.), 19–50. Amsterdam: Benjamins. Akenson, Donald H. 1999 The Irish in Ontario. A study in rural history. 2nd ed. Montreal: McGill-Queen’s University Press. Avis, Walter S. 1978 Problems in the study of Canadian English. In Walter S. Avis: essays and articles. Selected from a quarter century of scholarship at the Royal Military College of Canada, Kingston. Thomas Vincent, George Parker and Stephen Bonnycastle (eds.), 3–12. Kingston: Royal Military College of Canada. Avis, Walter S., Charles Crate, Patrick Drysdale, Douglas Leechman, Matthew H. Scharill, and Charles L. Lovell (eds.) 1967 A dictionary of Canadianisms on historical principles. Toronto: Gage. Barber, Katherine (ed.)2 2004 Canadian Oxford dictionary. 2nd ed. Toronto: Oxford University [11998] Press.

380

Stefan Dollinger

Bausenhart, Werner 1989 German immigration and assimilation in Ontario, 1793–1918. New York: Legas. Beal, Joan C. 2004 English in modern times 1700–1945. London: Arnold Hodder. Biber, Douglas 1988 Variation across speech and writing. Cambridge: Cambridge University Press. 2004 Modal use across registers and time. In Studies in the history of the English language II. Unfolding conversations. Anne Curzan and Kimberly Emmonds (eds.), 189–216. (Topics in English Linguistics, 45) Berlin: Mouton de Gruyter. Biber, Douglas, Susan Conrad and Randi Reppen 1998 Corpus linguistics. Investigating language structure and use. Cambridge: Cambridge University Press. Blackburn, Simon 2005 The Oxford dictionary of philosophy. 2nd ed. Online version. Oxford: Oxford University Press. Bloomfield, Morton W. 1948 Canadian English and its relation to eighteenth century American speech. Journal of English and Germanic Philology 47: 59–66 [reprinted in Chambers (1975), 3–11]. Brinton, Laurel J. and Margery Fee 2001 Canadian English. In The Cambridge History of the English Language. Vol. VI. English in North America. John Algeo (ed.), 422– 440. Cambridge: Cambridge University Press. Bueno Alonso, Jorge L., Dolores González Álvarez, Javier Pérez Guerra, and Esperanza Rama Martínez (eds.) 2007 ‘Of varying language and opposing creed’: New Insights into Late Modern English. Bern: Peter Lang (= Linguistic Insights). Chambers, J. K. (ed.) 1975 Canadian English. Origins and structures. Toronto: Methuen. Chambers, J. K. 1981 ‘Lawless and vulgar innovations’: Victorian views of Canadian English. Toronto Working Papers in Linguistics 2: 13–44. 1993 “Lawless and vulgar innovations”: Victorian views on Canadian English [rev. version] In Focus on Canada. Sandra Clarke (ed.), 1–26. (Varieties of English Around the World, G11) Amsterdam: Benjamins. 1998 English: Canadian varieties. In Language in Canada. John Edwards (ed.), 252–272. Cambridge: Cambridge University Press.

Taking permissible shortcuts? 2004

381

‘Canadian Dainty’: the rise and decline of Briticisms in Canada. In Legacies of colonial English. Studies in transported dialects. Raymond Hickey (ed.), 224-241. (Studies in English Language) Cambridge: Cambridge University Press. Clyne, Michael (ed.) 1992 Pluricentric languages. Differing norms in different nations. Berlin: Mouton de Gruyter. Coates, Jennifer 1983 The semantics of the modal auxiliaries. London: Croom Helm. 1995 The expression of root and epistemic possibility in English. In Modality in grammar and discourse (Typological Studies in Language 32). J. Bybee and S. Fleischman (eds.), 55–66. Amsterdam: Benjamins. Cowan, Helen 1961 British immigration to British North America. The first hundred years. Rev. and enl. ed. Toronto: University of Toronto Press. Dalton-Puffer, Christiane, Dieter Kastovsky, Nikolaus Ritt, and Herbert Schendl (eds.) 2006 Syntax, style and grammatical norms: English from 1500–2000 (Linguistic Insights, 39). Bern: Peter Lang. Denison, David 1993 English historical syntax: verbal constructions. (Longman Linguistics Library) London: Longman. 1998 Syntax. In The Cambridge history of the English language. Vol. IV: 1776–1997. Suzanne Romaine (ed.), 92–329. Cambridge: Cambridge University Press. Dollinger, Stefan 2006a Oh Canada! Towards the Corpus of Early Ontario English. In The changing face of corpus linguistics (Language and Computers, 55). Antoinette Renouf and Andrew Kehoe (eds.), 7–25. Amsterdam: Rodopi. 2006b The modal auxiliaries HAVE TO and MUST in the Corpus of Early Ontario English: gradience and colonial lag theory in Early Canadian English. Canadian Journal of Linguistics 51/2&3: 189–210. 2007 The importance of demography for the study of historical Canadian English: three examples from the Corpus of Early Ontario English. In J. L. Bueno Alonso et al. (eds.), 105–136. 2008 New-dialect formation in Canada: evidence from the modal auxiliaries (Studies in Language Companion Series 97). Amsterdam: Benjamins.

382

Stefan Dollinger

Ehrman, Madeline E. 1966 The meanings of the modals in present-day American English. The Hague: Mouton. Facchinetti, Roberta 2000 Can and could in contemporary British English: a study of the ICEGB corpus. In New frontiers of corpus research (Language and Computers. Studies in Practical Linguistics, 36). Pam Peters, Peter Collins and Adam Smith (eds.), 229–246. Amsterdam: Rodopi. 2002 Can. In Variation in central modals. A repertoire of forms and types of usage in Middle English and Early Modern English. Maurizio Gotti, Marina Dossena, Richard Dury, Roberta Facchinetti and Maria Lima (eds.), 45–65. Bern: Lang. 2003 Pragmatic and sociological constraints on the functions of may in contemporary British English. In Modality in contemporary English (Topics in English Linguistics, 44). Roberta Facchinetti, Manfred Krug and Frank Palmer (eds.), 301–327. Berlin: Mouton de Gruyter. Georgetown Linguistics Web Chi Square Calculator. Georgetown Linguistics Web Chi square Calculator. Programmed by Catherine N. Ball, Jeffrey Connor-Linton, Kathi Taylor. http://www.georgetown.edu/faculty/ballc/webtools/web_chi.html (28 Dec. 2006). Görlach, Manfred 1987 Colonial lag? The alleged conservative character of American English and other ‘colonial’ varieties. English World-Wide 8 (1): 41–60. 1988 The study of early Modern English variation - the Cinderella of English historical linguistics. In Historical dialectology. Regional and social. Jacek Fisiak (ed.), 211–228. (Trends in Linguistics. Studies and Monographs, 37) Berlin: Mouton de Gruyter. Gourlay, Robert 1822 Statistical account of Upper Canada. Compiled with a view to a grand system of emigration (CIHM 35937). London: Simkin & Marshall. Hickey, Raymond 2004a Introduction. In Legacies of colonial English. Studies in transported dialects. Raymond Hickey (ed.), 1–30. (Studies in English Language) Cambridge: Cambridge University Press. 2004b Dialects of English and their transportation. In Legacies of colonial English. Studies in transported dialects. Raymond Hickey (ed.), 33– 58. (Studies in English Language) Cambridge: Cambridge University Press.

Taking permissible shortcuts?

383

Hultin, Neil C. 1967 Canadian views of American English. American Speech 42: 243– 260. Jones, Charles 1989 A history of English phonology. London: Longman. Kytö, Merja 1991 Variation in diachrony, with early American English in focus. Studies on CAN/MAY ad SHALL/WILL. Bern: Lang. Kytö, Merja, Juhani Rudanko and Erik Smitterberg 2000 Building a bridge between the present and the past: a corpus of 19thcentury English. ICAME Journal 24: 85–97. Kytö, Merja and Matti Rissanen 1983 The syntactic study of Early American English. The variationist at the mercy of his corpus? Neuphilologische Mitteilungen 84: 470– 490. Labov, William 1994 Principles of linguistic change. Volume 1: Internal factors. Oxford: Blackwell. Lovell, Charles J. 1955 Lexicographic challenges of Canadian English. Journal of the Canadian Linguistic Association (Canadian Journal of Linguistics) 1/(1) (Mar.): 2–5. Marckwardt, Albert H. 1958 American English. New York: Oxford University Press. Michalewicz, Zbigniew and David B. Fogel 2004 How to solve it: modern heuristics. 2nd ed. Berlin: Springer. Nelson, Gerald, Sean Wallis and Bas Aarts 2002 Exploring natureal language. Working with the British component of the International Corpus of English. Amsterdam: Benjamins. Nevalainen, Terttu 1999 Making the best use of ‘bad’ data. Neuphilologische Mitteilungen. 100 (4): 499–533. Oaks, Michael P. 1998 Statistics for corpus linguistics. Edinburgh: Edinburgh University Press. Orkin, Mark M. 1970 Speaking Canadian English. An informal account of the English [1971] language in Canada. [Reprint]. New York: McKay Company. Preacher, Kristopher J. and Nancy E. Briggs 2001 Calculation of Fisher’s Exact Test: an interactive calculation tool for 2 x 2 tables. Avail. from http://www.quantpsy.org (20 Dec. 2006).

384

Stefan Dollinger

Scargill, Matthew H. 1957 Sources of Canadian English. Journal of English and Germanic Philology 56: 611-614 [reprinted in Chambers (1975), 12–15]. Schneider, Edgar W. 2007 MY BABY LOVES ME, SHE LOVE ME: verbal -s variability in the history of black and white dialects of the southern United States. In Tracing English through time: explorations in language variation (Austrian Studies in English, 95). Ute Smit, Stefan Dollinger, Julia Hüttner, Gunther Kaltenböck and Ursula Lutzky (eds.), 345–358. Vienna: Braumüller. Simon-Vandenbergen, A. M. 1984 Deontic possibility: a diachronic view. English Studies 65 (4): 362– 365. Sundby, Bertil, Anne Kari Bjørge and Kari E. Haugland 1991 A dictionary of English normative grammar 1700–1800. Amsterdam: Benjamins. Thomas, Eric R. 1991 The origin of Canadian Raising in Ontario. Canadian Journal of Linguistics 36: 147–170. Traugott, Elizabeth Closs 1972 A history of English syntax. A transformational approach to the history of English sentence structure. New York: Holt, Rinehart & Winston. 1989 On the rise of epistemic meanings in English: an example of subjectification in semantic change. Language 65 (1): 31–55. Trudgill, Peter 2002 The history of the lesser-known varieties of English. In Alternative histories of English. Richard Watts and Peter Trudgill (eds.), 29–44. London: Routledge. 2004 New dialect formation. The inevitability of colonial Englishes. Edinburgh: Edinburgh University Press. Tversky, Amos and Daniel Kahneman 1982 Judgment under uncertainty: heuristics and biases. In Judgment under uncertainty: heuristics and biases. Daniel Kahnemann, Paul Slovic and Amos Tversky (eds.). Cambridge: Cambridge University Press. Vogt, W. Paul 2005 Dictionary of statistics & methodology. Thousand Oaks, CA: Sage. Warner, Anthony R. 1993 English auxiliaries. Structure and history. Cambridge: Cambridge University Press.

Taking permissible shortcuts?

385

Watts, Richard and Peter Trudgill (eds.) 2002 Alternative histories of English. London: Routledge. de Wolf, Gaelan Dodds, Robert J. Gregg, Barbara P. Harris, and Matthew H. Scargill (eds.)5 1997 Gage Canadian Dictionary. 5th ed. Toronto: Gage. [11967] Wolfram, Walt and Natalie Schilling-Estes 2004 Remnant dialects in the coastal United States. In Legacies of colonial English. Studies in transported dialects. Raymond Hickey (ed.), 172–202. (Studies in English Language) Cambridge: Cambridge University Press. Wood, J. David 2000 Making Ontario. Agricultural colonization and landscape recreation before the railway. Montreal: McGill-Queen’s University Press. Woods, Anthony, Paul Fletcher and Arthur Hughes 1986 Statistics in language studies. Cambridge: Cambridge University Press.

‘What strikes the ear’ Thomas Sheridan and regional pronunciation*

Raymond Hickey Judging by the number of publications by different authors during the eighteenth century and by the success they achieved with their writings, this was the period in which public concern with correctness in language as a social qualification takes a firm foothold in English, Scottish and Irish society. The general concern with putative correctness has its roots in earlier debates about standard English, above all among the writers of the Augustan period. Prominent among these was Jonathan Swift, to whom later writers refer and to whose authority they appeal. Robert Lowth in the preface to his famous grammar of 1762 writes: “Swift must be allowed to have been a good judge of this matter (i.e. the imperfect state of our language – RH); to which he was himself very attentive, both in his own writings, and in his remarks upon those of his friends: he is one of the most correct, and perhaps the best of our prose writers” (Lowth 1995 [1762]: vi).1 But Augustan writers like Swift appear to have been concerned with changes in English because these separated the language from that of their predecessors and would render their own writings linguistically obsolete to later generations. Lowth is somewhat different in his stance. In his grammar he talks of expressing oneself “rightly” but he also stresses the notion of “propriety” which was definitely a social concept. The idea is foregrounded that incorrect, i.e. non-standard, grammar is offensive to educated, middle-class ears, especially those of the capital London and its environs. With that two issues become topical: (i) regional pronunciations of English and (ii) uneducated, or “vulgar” usage. In earlier centuries, we have indications of the regional origin and accents of public figures. For instance Sir Walter Ralegh (1554–1618) was from Devon and we know from contemporary remarks that he spoke with a southwest English accent. What is new in the eighteenth century is the social censure of regional accents which were seen as signs of poor breeding and education and generally to be avoided.

388

Raymond Hickey

There is an added issue here which concerns the attitude and reaction of people from the regions to centralist notions of linguistic propriety. The first major reaction comes from Ireland; others from Scotland and the north of England were to follow. It might at first sight seem strange that an Irishman should be among the first to adopt, in print, an openly prescriptive and corrective stance on the English language. But this is only an apparent paradox. Given that eighteenth-century Dublin was a major city of Great Britain but an outpost at the same time, it is understandable that some Dubliners may well have internalised centralist notions of correctness and indeed, in a rush of over-assimilation, sought to be more English than the English themselves.

1.

Thomas Sheridan

The Irish individual in question is one Thomas Sheridan who was born in 1719, grew up in Dublin and died in London in 1788. Sheridan enjoyed a diverse career as actor, lecturer and writer. Significantly, he was the godson of Swift and produced The Works of Swift with Life (18 volumes) in 1784. As a dramatist Sheridan is known for one play, Captain O’Blunder or The Brave Irishman (1754), which he wrote in 1740 as an undergraduate. He was also manager of the Smock Alley theatre in Dublin (Sheldon 1967) where he worked for some years. However, the plays of his son Richard Brinsley Sheridan (1751–1816) were better known in his day and were regularly produced in London at Covent Garden and at the Drury Lane theatre. The interest which Thomas Sheridan latterly showed in correctness of language goes back at least to his book British Education: Source of Disorders, which appeared in 1756 shortly after he had left Ireland for England (see remarks below), and in his 1761 book A Dissertation on the Causes of the Difficulties which Occur, in Learning the English Tongue. This combined interest in education and in correcting what he perceived as unacceptable usage may well be something which he inherited from this father. Thomas Sheridan senior (1687–1738) was a clergyman and educator who established his own school in Capel St., Dublin where pupils were trained in the classical languages. In the present context, Thomas Sheridan junior is important as the author of A Rhetorical Grammar of the English Language (1781) which contains a section on the Irish pronunciation of English. He is also the author of a successful General Dictionary of the English Language (1780, 2 vol-

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

389

umes) and an earlier A Course of Lectures on Elocution published in the same year as Lowth’s grammar – 1762 – and by the same publishers – R. and J. Dodsley in London. Probably on the grounds of these linguistic interests, Sheridan developed a close friendship with the lexicographer Samuel Johnson. The latter’s opinion of the work of Sheridan throws an interesting light on English-Irish attitudes of the time. Dr. Johnson saw it as a distinct disadvantage for Thomas Sheridan to be Irish and pronouncing on English (Beal 2004b: 331). This contrasts strongly with the English attitude towards Swift, who was so often held up as a paragon of English style. But Swift was a Protestant dean who had worked in England and only occasionally wrote on specifically Irish matters. Sheridan, on the other hand, was a Catholic actor, minor playwright and self-appointed arbiter of correct English usage.

2.

Elocution

Sheridan’s interest in language surfaces in his middle years with the publication of his lectures on elocution. There is a connection between this initial concern and his later pronouncements which will be the subject of discussion below. Let me show this by considering what was meant by elocution in late 18th century Britain and Ireland. To start with one can locate elocution in classical Latin writings as one of the five so-called “offices”, or sections, of rhetoric: (1) invention, analyzing a topic and collecting material for it, (2) disposition, arranging the material for a speech, (3) elocution, finding appropriate words for the topic, the speaker and the audience, (4) pronunciation, oral delivery of a speech and (5) memory, committing the contents of the speech to memory. Elocution seems to have been mainly the domain of oratory and concerned with good style and expression, but its meaning altered over time, as recorded in the definition given in the Oxford English Dictionary: elocution /elԥ/kju$ԥn/ n. Late Middle English [L elocutio(n-), f. elocut- past participle stem of eloqui: see ELOQUENT, -ION.] 1 Oratorical or literary expression; literary style as distinguished from matter; the art of appropriate and effective expression. Late Middle English. 2 Eloquence, oratory. Latin 16c to Latin 18c. 3 The art of (public) speaking, esp. of pronunciation, delivery, gesture, and voice production. English 17c. 4 Manner or style of speaking. English 17c. M19.

390

Raymond Hickey

There is a discernible shift in emphasis by the 17th century. Elocution as the third office of rhetoric came to refer to effective public speaking: pronunciation, gesture and voice quality are singled out here. It is not surprising then that elocution came to be associated with actors given their professional concern with speaking in public and the fact that Sheridan himself was originally an actor. Consider also cases like that of Sarah Siddons (1755–1831), a successful tragic actress and acquaintance of Sheridan’s dramatist son, Richard Brinsley, who engaged her in Drury Lane in 1782. The following year she was engaged to teach elocution to the children of the royal family.2 Returning to Sheridan, we can see that he was concerned with establishing serious academic underpinnings for elocution. He found in the work of the 17th century philosopher John Locke assumptions which improved the credentials of elocution. Locke believed that “words are the signs of ideas, tones, i.e. spoken language, the signs of passions”. As elocution was taught by reciting previously composed material in public, it was clearly associated with ideas and hence to be taken more seriously. The opinion was adopted that the way to teach individuals to speak well was to train them in reciting written material. This stance characterised elocution within educational systems in the English-speaking world and still does inasmuch as it is still a school subject.

3.

Sheridan’s influence

It is known that Sheridan travelled widely throughout the British Isles, lecturing on elocution and “correct” English, notably in Scotland. The question of language, specifically of the differences between Scottish and southern English usage, had become an increasingly sensitive issue there, something which is apparent in the works of Buchanan, such as his British Grammar (1762) and his Essay Towards Establishing a Standard for an Elegant and Uniform Pronunciation of the English Language, Throughout the British Dominions (1766), where the term “British” carries distinct political overtones, all the sharper given the union of the English and Scottish parliaments in 1707 (Beal 2004a: 96). Sheridan had a considerable influence on authors in Britain, notably on his main rival, London-born John Walker who in 1791 published A Critical Pronouncing Dictionary of the English Language much in the vein of Sheridan’s works and in which he compares his own dicta to those of his

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

391

Irish colleague (see the many discussions in this respect in Beal 2004a, e.g. p. 129). Sheridan’s influence on Walker should not be underestimated. Not only did the latter include Sheridan’s admonitions to the Irish, Scottish and Welsh (see his reprint of Sheridan’s original comments, Walker 1815: 13– 19) but he adopted wholesale the prescriptive remarks which the former made about regional pronunciation in general. Walker seems to have had the greater success3 and his dictionary was reprinted repeatedly until the beginning of the twentieth century (over 100 times between 1791 and 1904 to be precise, Beal 2004a: 171). Walker was very detailed in his discussion of pronunciation variants in English (far more so than Sheridan) and the 15th edition of his dictionary contains a preliminary section, ‘Principles of English Pronunciation’, which is nearly 70 pages long (Walker 1815: 21–90). Many authors quickly recognised that among the rising middle classes there was a market for works on elocution and they jumped on the bandwagon with alacrity. A good example is Stephen Jones who produced a work somewhat conceitedly entitled Sheridan Improved: A General Pronouncing and Explanatory Dictionary of the English Language which reached its 3rd edition by 1798. Sheridan had other rivals apart from John Walker who produced works of the same type as he did, indeed in one case with the same title, and almost at the same time. For instance, William Kenrick produced A New Dictionary of the English Language in 1773 in London and brought out A Rhetorical Grammar of the English Language in 1784 only three years after Sheridan’s work of the same name. There were not just imitators but detractors as well; for instance, there is an anonymous book entitled A Caution to Gentlemen who use Sheridan’s Dictionary which, given the title, scarcely needs comment. John Walker also had his imitators and improvers. In 1836 B. H. Smart brought out Walker Remodelled. A New Critical Pronouncing Dictionary of the English Language, and followed this with The Practice of Elocution in 1842.

4.

Correcting the natives

Elocution is not just about teaching what is “correct” pronunciation but also about censuring what is deemed by authority to be “base”, “vulgar”, “rude” or just “provincial”. The censures of the elocutionists can be used by linguists today to glean information about regional and/or colloquial pronunciation of their time. In his A Rhetorical Grammar of the English Language (1781)

392

Raymond Hickey

Sheridan inserted an appendix entitled “Rules to be observed by the natives of Ireland in order to attain a just pronunciation of English” (Sheridan 1781: 137–55). This also contains a number of remarks on the “mistakes” which the Irish make when speaking English. At one point he offers a list consisting of two columns with Irish and English pronunciations respectively. Here there can be no doubt about Sheridan’s prescriptive intentions. Sheridan does not hold back with criticism of regional accents in parts of Britain either: “With regard to the natives of Scotland as their dialect differs more, and in a greater number of points, from the English, than that of any others who speak their language, it will require a greater number of rules, and more pains to correct it” (Sheridan 1781: 146). He is unashamedly censorious, but kindly offers suggestions for curing the unacceptable speech habits of the Scots. Here the goal is quite clear: provincial pronunciations are to be abandoned and those of educated southeast England are to be adopted. The Welsh are taken to task as well, this time for the devoicing of consonants in initial and medial position, e.g. fice for vice, seal for zeal, ashur for azure, etc. and again the goal is to weed out such provincialisms from their speech and so make it acceptable in educated circles in and around the capital. Before looking at Sheridan’s comments in detail, we might well ask what basis for making such pronouncements he may have had, given that he was born and reared in Ireland, a place where exposure to educated southeastern English usage would have been minimal. Despite his Irish birth and upbringing, Sheridan can claim to have had exposure to educated southeastern English usage of the mid 18th century. After attending his father’s school in Dublin, he went to Westminster College in London before returning to Ireland to attend Trinity College Dublin. He later became actively involved in the management of Smock Alley Theatre in Dublin, but in 1744 he travelled to London and competed as an actor with David Garrick, whom he persuaded to come to Dublin for the season in 1745–6. In the following years he managed Smock Alley but following riots in 1747 and 1754 he retired from acting and management and went to England where he began to publish his works on education and elocution.

5.

Sheridan’s strictures and 18th century Irish pronunciation

In the appendix to his grammar Sheridan compares the pronunciation of Irish English with southern British English. He is fairly accurate in his de-

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

393

scription and representation of vowel quality. The Irish English forms which he quotes are those in which “well-educated natives of Ireland differ from those of England” (1781: 146). The only group in Ireland that could be described as “well-educated” at the end of the 18th century was the Protestant middle and upper class. At this time the middle class of the city was prospering (see Moody and Martin 1967: 232ff. and Johnston 1994 [1974]: 77ff., for descriptions of Dublin at this time). And more than ever before this social group enjoyed independence from England (up to the ‘Act of Union’ of 1801, Johnston 1994 [1974]: 164ff., Ó Tuathaigh 1972: 29ff.). At this time Dublin enjoyed a certain political freedom (Craig 1969) and this would have minimised the direct influence of southeastern British English on the Irish pronunciation of English. Although one can identify Sheridan’s group of speakers fairly easily, it is not equally apparent just what he regarded as the standard English that he refers to and who is supposed to have spoken it. In the preface to his grammar he talks of “our pronunciation” (1781: xxii) and refers to Johnson with regard to spelling (1781: xxiii). He further notes that the pronunciation of English by the people in Ireland, Scotland and Wales can deviate from a standard, without offering any more specific information on what he regards this standard to be. His praising remarks (1781: xix) on the correct pronunciation of the “Augustan Age” in England (Sheridan was after all the godson of Swift, Croghan 1990), are of little help. From this one can conclude only that Sheridan was a prescriptivist and assumed educated southeastern English, the variety of “polite” society in London and the Home Counties, as a yardstick with which to compare Irish English of his time. 5.1. Sheridan’s system of pronunciation In the following paragraphs the system which Sheridan used (Sheldon 1946) for indicating regional pronunciations of English is explained and the extent to which it can be viewed as a window on Dublin English in the 18th century is discussed. However, it would be misleading to interpret his references as relating solely to educated Dublin usage of his time. Many of his strictures concern pronunciations that were common in Britain at the time. For instance, the lowering of /e/ to /a/ before /r/, the raising of /æ/ to /ȏ/, especially after velars, and the realisation of short vowels before /r/ are all matters that are relevant to varieties of English in Britain during his time.

394

Raymond Hickey

Notwithstanding this general relevance, Sheridan’s motive in his description was clearly to show the Irish just how wrong their pronunciation of English was. In consequence, on several occasions the detail which one would like concerning local features is not given because it does not seem to have been pertinent to Sheridan’s goal of demonstrating to the Irish the wrongness of their phonetic ways. Furthermore, what Sheridan mentions is obviously what “struck his ear” which does not mean that he registered all non-standard features of late 18th century Dublin English. He is singularly silent about non-standard consonantal features, of which there are many in present-day conservative Dublin English (Hickey 2005: 34–45) and so probably in that of his time as well. In order to describe the special features of Dublin English Sheridan employs a system of notation in which the five vowel graphemes (and y) are used together with a number (from 1 to 3) as a diacritic to denote possible vowel values in Irish English (Sheridan 1781: 151). In the original the number is actually placed over the vowel. For technical reasons, it is placed here as a superscript digit to the right of the vowel letter. In his text, Sheridan organises the vowel values into three columns as follows: (1) a e i o u y

First ha1t be1t fi1t no1t bu1t love-ly1

Second ha2te be2ar fi2ght no2te bu2sh ly2e

Third ha3ll be3er fi3eld no3ose blu3e

Sheridan also offers some notes in which he explains his transcriptional system: All improper diphthongs, or, as I have called them digraphs, I mean where two vowels are joined in writing, to represent any of the simple sounds in the scheme, are changed in the second column into the single vowels which they stand for; as thus bear

be2re

hear

he3re

fourth

fo2rth

door

do2re

(Sheridan 1781: 152)

The ‘final mute e’ is added to certain words in Sheridan’s pronunciation system in order to disambiguate the transcriptions.

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

395

It needs to be stressed here that Sheridan was working within an existing tradition of describing and classifying vowels. For instance, Samuel Johnson in his dictionary of 1755 used the same system to describe the three vowels which he recognised, “the slender, open and broad”, i.e. the sounds as in face, father and all respectively (Johnson’s examples, see discussion in “A grammar of the English tongue”, Johnson 1755 [no page numbers]). Sheridan’s table of vowels is assumed to represent values from educated southeastern English usage of his time against which he then compared Irish (and Scottish) pronunciations. The phonetic values of the above vowels in educated southeastern English usage are not known in all cases. For instance, the vowel a3 cannot be decided upon precisely. It is true that /au/ before /l/ in the early modern English period (Dobson 1968: 603ff.) went to /o:/, but it is not certain just how close this vowel was, it may well have been closer to / and chides others for recommending this pronunciation in some words (Kenrick 1773: 40f.). This vowel value showed considerable variation, cf. Johnson (1755 [no page numbers]) who mentions in the preface to his dictionary that “Ea sounds like e long, as mean; or like ee, as dear, clear, near”, a remark which shows that for him mean was still [me:n]. While such mid-front pronunciations were replaced by a high front value soon afterwards in England, the older pronunciation survived much longer in Ireland and still exists in rural varieties and as a vernacular option for urban speakers. Some of the words quoted by Sheridan seem to be lexicalised exceptions to general English sound shifts. For instance, Sheridan quotes ‘drought’ as having the pronunciation dro3th /dru:t/ in Ireland, but there is no further evidence for the non-diphthongisation of Middle English /u:/ to /au/ at this late stage of early modern Irish English. When discussing late 18th century pronunciation and when comparing Irish and English usage at this time it is necessary to be explicit about what varieties are being referred to. On the Irish side, the matter is relatively simple, as Sheridan clearly states that the strictures he articulates refer to the speech of “the gentlemen of Ireland”. This a fairly clear reference to educated Irish usage centred around Dublin as the main English-speaking city and centre of education in late 18th century Ireland. On the English side, however, the matter is not quite so simple. To determine just what Sheridan thinks is the English pronunciation worthy of emulation, one can best let him speak for himself. In the preface to his Rhetorical Grammar of

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

397

1781, Sheridan is at some pains to be explicit about what his yardstick of pronunciation actually is. It must be obvious, that in order to spread abroad the English language as a living tongue, and to facilitate the attainment of its speech, it is necessary in the first place that a standard of pronunciation should be established, and a method of acquiring a just one should be laid open. That the present state of the written language is not at all calculated to answer that end, is evident from this; that not only the natives of Ireland, Scotland, and Wales, who speak English, and are taught to read it, pronounce it differently, but each county in England has its peculiar dialect, which infects not only their speech, but their reading also. (Sheridan 1781: xvii)

After this lament for the lack of uniformity in spoken English across the regions of Britain and Ireland Sheridan continues to remark on the variation to be found in social groups which enjoy higher positions in society. It is well known, that there is a great diversity of pronunciation of the same words, not only in individuals, but in whole bodies of men. That there are some adopted by the universities; some prevail at the bar and some in the senate-house. (Sheridan 1781: xix)

The above statements are a preamble to his main objective, that of specifying the yardstick of pronunciation which he favours, namely that of the early 18th century, specifically that of the reign of Queen Anne (1702– 1714), the last of the Stuarts. There was a time, and that at no very distant period, which may be called the Augustan age of England, I mean during the reign of Queen Anne, when English was the language spoken at court; and when the same attention was paid to propriety of pronunciation, as that of French at the Court of Versailles. This produced a uniformity in that article in all the polite circles; (Sheridan 1781: xix–xx)

This provides him with a baseline from which to criticise the changes in English which had taken place between that time and his own. Many pronunciations, which thirty or forty years ago were confined to the vulgar, are gradually gaining ground; and if something be not done to stop this growing evil, and fix a general standard at present, the English is likely to become a mere jargon, which every one may pronounce as he pleases. (Sheridan 1781: xx)

398

Raymond Hickey

In keeping with the Augustan tradition, Sheridan is concerned with fixing the language and protecting it from the pernicious effects of variation and change. When we reflect, that no evil so great can befall any language, as a perpetual fluctuation both in point of spelling and pronouncing, it is surely a point to be wished, that a permanent and obvious standard of both should at some certain period be established; and if possible, that period should be fixed upon when probably they were in the greatest degree of perfection. (Sheridan 1781: xxiii)

Against this background it is now possible to examine some of the pronunciations listed by Sheridan which have changed since his time and which offer evidence for what he regarded as acceptable usage and what was subject to his disapproval. 18c IrE /u:/ In the late Middle English period there was some fluctuation between /o:/ and /u:/ (Dobson 1968: 681ff.). Some words which later appear with /o:/ show /u:/ in Sheridan’s list. These are instances of the failure of /u:/ to diphthongise before /r/ (Dobson 1968: 688ff.). /du:r/ ‘door’ (4) a. do3re /flu:r/ ‘floor’ b. flo3re c. co3urse /ku:rs/ 18c IrE /e:/ The lack of raising of ME <ea> is a well-known characteristic of vernacular Irish English, both north and south, and is still found in present-day vernaculars (Milroy and Harris 1980). This feature and the use of /a:/ in words like patron, matron (see below) is pointed out by Sheridan as the chief “mistake” in the Irish pronunciation of English. From the point of view of earlier English, i.e. late Middle English, one has a collapse of the distinction between /e:/ and /e:/ but no further shift to /i:/: least /le:st/, deal /de:l/, beet /be:t/. Sheridan’s group of speakers would seem to have been aware of this feature of Irish English and it is understandable that in attempts to avoid the local pronunciation they engaged in hypercorrection by shifting all instances of /e:/ to /i:/, even where this was not justified etymologically.

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

(5) a. prey b. convey c. bear

ĺ ĺ ĺ

pree convee beer

399

(Sheridan 1781: 143)

This hypercorrection was noted by other authors too. For instance, Maria Edgeworth in her Essay on Irish Bulls (1802) remarked that “There are Irish ladies, who, ashamed of their country, betray themselves by mincing out their abjuration, by calling tables teebles, and chairs cheers!” (Crowley 2000: 136). 18c IrE /v/ Sheridan indicates that the /v/ vowel was to be found in the Dublin of his time in words like the following. (6) a. pu1lpit /pvlpit/ b. bu1sh /bv$/ d. pu1sh /pv$/

b. pu1dding c. bu1ll e. pu1ll

/pvdin/ /bvl/ /pvl/

All but the first of these words have /u/ in (southern) Irish English and British English today. Bearing in mind that Sheridan’s speaker group consisted of “well-educated natives” one can safely assume that the above instances are cases of hypercorrection parallel to the unconditioned shifting of /e:/ to /i:/ mentioned in the previous paragraph. 18c IrE /a:/ Similar to the lack of shift for historical /e:/ to /i:/ one also finds that /a:/ is not always raised to /e:/: patron /pa:trqn/, matron /ma:trqn/. Sheridan remarks that forms such as those just quoted have the same vowel as in father. Unshifted ME /a:/ was obviously a salient feature of 18th century Irish English and appears abundantly in parody literature, e.g. in George Farquhar’s The Beaux’ Stratagem (1707) as seen in the following extract.

FOIGARD Ireland! No, joy. Fat sort of plaace (= [pla:s]) is dat saam (= [sa:m]) Ireland? Dey say de people are catcht dere when dey are young.

Swift’s rhymes also show that /a:/ was an educated Irish pronunciation in the early 18th century. The censure of this pronunciation probably followed on the raising of the /a:/ to /e:, e:/ in southern England some considerable time before Sheridan. Swift was likely to have maintained a pronunciation, at least in rhyming verse, which was conservative even in his time.

400

Raymond Hickey

How by heroes of old our chiefs are surpass’d In each useful science, true learning, and taste. Verses on the Battle of the Books Why should the first be ruin’d and laid waste, To mend dilapidations in the last? Ode to Doctor William Sancroft This realisation has been lost everywhere in Ireland as a result of the later superimposition of more standard pronunciations. Occasionally, one has spelling pronunciations like status [sta:tqs] or data [da:tq] which is not a remnant of the former situation, however. 18c IrE /i:/ References to an /i:/ pronunciation in cases where English has /ai/ are not numerous in Sheridan. For instance, he gives the following form: Mi3kil /mi:kԥl/, compare English Mi2kil /maikql/ which he also quotes. However, this may be a transfer of the vowel value in the Irish form of the name, Mícheál //mji:h:t/

Eng wra3th /ro:2/ Eng wro1th /r>2/

Here the difficulties with Sheridan’s English reference accent are most apparent. He would seem to favour a conservative southeastern variety, that of “polite society” as he specifies in his preface, but just what group in England had retraction, rounding and lengthening of /a:/ after /w/ in wrath along with a short vowel before /2/ in wroth is unclear. Lack of rounding of /a/ after /w/ The rounding and retraction of short /æ/ in a position after /w/ or /hw/, but not before velars, e.g. wander, what but wag, probably started in the 16th century and was adopted into more standard forms of English in the 17th century (Prins 1974: 149f.). Sheridan shows that this rounding had not occurred in Dublin English by the end of the 18th century: squa1dron /skwædrqn/ = squo1dron /skw>drqn/ and is still not found in popular Dublin English.

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

403

Cluster simplification In the relatively short word list Sheridan offers two words, already attested in earlier Irish English, which show a simplification of clusters: lenth /lent/ ‘length’ and strenth /strent/ ‘strength’. The typical instance of such simplification in present-day English, namely nasal and stop in word-final position, as in ground, band, is not given. It may well be that reduction of clusters in this position was too salient a feature of popular Dublin English and hence avoided by his speaker group. Final cluster reduction is, however, well-attested before the 18th century and is still a marked feature of popular Dublin English. Dentalisation before /r/ Of all the features of late 18th century Dublin English that of dentalisation before /r/ is explained most clearly by Sheridan. In his remarks on /t/ (1781: 35) he notes that the Irish “thicken t (and d) so that they say betther for better and utther for utter”. It is clear subsequently that he is talking of dentalisation: “this faulty manner arises from the same cause that was mentioned as affecting the sound of the d, I mean the protruding of the tongue so as to touch the teeth.” He furthermore recognises that there is a morphological alternation of alveolar and dental stops in the comparison of adjectives: “thus though they (the educated Dublin Irish, RH) sound the d right in the positives loud and broad, in the comparative degree they thicken it by an aspiration; and sound it as if it were written loudher, broadher” (1781: 29). This pronunciation has been and still is very widespread in vernacular forms of English throughout all of Ireland. 5.4. Word stress From Sheridan’s system of vowel quantity notation there would seem to have been a correlation between quantity and stress placement in early modern Irish English. Consider first the manner in which accent is noted by Sheridan: The accent is placed throughout over the letter on which it is laid in pronunciation; over the vowel, when the stress of the voice is on the vowel; over the consonant when the stress is on that as thus – Consonant subscript 1 stu r/ 1 lu v/ 1 bi z/zy 1 1 la f /tu r

Vowel subscript 2 be /re 3 he /re 2 gro /ne 2 so /shal

404

Raymond Hickey

The syllables of words are divided according to the mode of pronouncing them; that is, all letters which are united in utterance in the same syllable, are here kept together also in writing, and separated from the rest; ... (Sheridan 1781: 155)

Statements such as the above have to be interpreted in modern terms to make sense. Sheridan would appear to view all consonants which follow on a short vowel as belonging to the same syllable. His superscript stroke indicates the end of the syllable. This device is usually superfluous, however. With the following forms, for instance, the superscript digit indicates a long vowel and an interpretation via syllable structure is not necessary. (9) a. IrE: ze2alous /ze:lqs/ b. IrE: ze2alot /ze:lqt/

Eng: ze1llus /zelqs/ Eng: ze1llut /zelqt/

Long vowels occur in Sheridan’s list in final position where English has a short unstressed vowel, cf. cla1/mour /kla:mu:r/ – cla1/mur /klæmqr/. Furthermore, one also finds that words of three syllables with a short stressed middle vowel have a long vowel in Sheridan’s speaker group (Sheridan 1781: 146). (10) a. IrE: ende2avour b. IrE: mali3cious

/en/de:vu:r/ /mq/li:$qs/

Eng: ende1v/ur /en/devqr/ Eng: mali1sh/us /mq/li$qs/

This vowel length is a conservative trait as is the lack of accent shift to the front of words which is also attested by Sheridan. (11) IrE: mischi3/evous /mis/t$i:vu:s/

Eng: mi1s/chivous //mist$ivԥs/

Some of his transcriptions show deviations in the supposedly English forms. For instance, he contrasts long and short vowels in Irish and English pronunciation which in the latter only have long vowels today. (12) a. IrE: che3ar/ful b. IrE: fe3ar/ful

/t$e:rful/ /fe:rful/

Eng: che1r/ful Eng: fe1r/ful

/t$e(r)ful/ /fe(r)ful/

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

405

5.5. Features not proscribed by Sheridan In dictating modes of pronunciation, Sheridan may not have been very successful. We know that his famous playwright son, Richard Brinsley, had a distinct Dublin accent. In her reminiscences of famous individuals she knew, Fanny Burney (1752–1840) imitates the Irish accent of the playwright by referring to his pronunciation of kind as [kԥind], indicated orthographically as koind. But Sheridan appears to have been selective in his censure. The centralised diphthong of the PRICE lexical set would appear to have been a feature of educated middle-class usage in 18th century Dublin. But it apparently only became a stereotype in the following century. By the end of the 19th century, English authors, such as Kipling, when trying to indicate an Irish accent in writing, used the oi spelling to indicate the centralised diphthong onset as in woild Oirland. There is an important generalisation here: elocutionists like Sheridan seemed to have proscribed pronunciations which were salient for them (Hickey 2000). They ignored those that were not. Naturally they also ignored features where regional pronunciations coincided with educated southeastern English usage. What happened occasionally is that the latter changed but the regional pronunciation did not, leaving it distinct from later standard usage. A case in point is the length of the vowel in the SOFT lexical set. In southern forms of English both early modern /æ/ and />/ were lengthened before voiceless fricatives, thus giving familiar pronunciations today like path, staff, pass all with long /:ft], [k>:st], [>:f] and non-Dublin [s>ft], [k>st], [>f]. There are a couple of other pronunciation features which Sheridan did not censure. One of these would appear to be traceable to the effect of the Irish language on English in Ireland. Sheridan favoured the assibilation of /tj/ to [t$] as in tune [t$u:n] whereas it is condemned by his English counterparts. In this, as Beal (2004: 147) rightly notes, Sheridan was favouring a particularly Irish pronunciation. 5.6. Summary By and large one can say that Sheridan’s prescriptive comments on late 18th century Dublin English provide a useful glimpse of what pronunciation must have been like at the time. Especially in the area of vowels, Sheridan provides information that corroborates findings for present-day local Dublin English. The area of consonants was not dealt with in any great detail by Sheridan – something which has been typical of the elocutionary tradition since then – so that confirmation or refutation of such typical present-day features as t-lenition is unfortunately not forthcoming from Sheridan’s rhetorical grammar.

6.

Developments in the 19th century

It is interesting to consider the long-term effects of the proscriptions embodied in Sheridan’s works.4 To assess these one must consider the situation in 19th century Ireland and see what has happened to the pronunciations of his time which he chose to censure. Here one can distinguish different situations. There are features which have disappeared entirely. An example of this is ‘CATCH raising’, i.e. the pronunciation of the word as /ket$/. The raised pronunciation, especially after velars, was common in both Britain and Ireland in the later 18th and the 19th century, indeed well into the 20th century when it disappeared after World War Two. (The vowel raising has, if anything, been reversed so that a pronunciation closer to /a/ is to be found in RP nowadays, Bauer 1994: 110–21.) There are many attestations of this raising up to the early 20th century when Sean O’Casey used it to indicate local Dublin pronunciation. Sheridan (1781: 144) illustrates this by writing gather as gether and catch as cetch. A similar raising

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

407

would seem to have applied to low back vowels in certain environments. Sheridan remarks that words like psalm, qualm, balm are pronounced as if written psawm, quawm, bawm (Sheridan 1781: 141), a feature attested elsewhere in historical documents of Irish English. Both types of raising have disappeared without trace or comment from Irish English. Another feature, destined to disappear in the course of the 19th century, is ‘SEARCH lowering’, indicated in Sheridan by sa1rch /sa:rt$/ which he explicitly contrasts with English se1rch /se:rt$/ (Sheridan 1781: 145). This lowering before /r/ was a general feature of Irish English up to the 19th century. It is also found in many dialects of English and can be seen in names like Derby, Berkshire, Hertfordshire and is responsible for pronunciations like dark and barn in present-day English. The lowering has been entirely removed from Irish English by the superimposition of later more standard pronunciations and cannot be used as a diagnostic for present-day Irish English. Sheridan showed a preference for some words with long /a:/ before /r/, e.g. merchant which Walker thinks should be pronounced with /e:/ (Walker 1815: 330). He remarked that /ar/ in words like service and servant is “still heard among the lower order of speakers” (Walker 1815: 30). However, these words survive with /ar/ into the late 19th century in Ireland as seen in many dramas such as those by Dion Boucicault (1820–1890) writing in the 1860s and 1870s, see Hickey (2007, Chapter 5). Beal (2004a: 131) discusses Walker’s attitude to Sheridan on this point. Walker rightly recognises that Sheridan was old-fashioned. His preference for the low vowel in such words is an instance of lag where Dublin pronunciation in the latter half of the 18th century was simply not keeping up with developments in England where a raised vowel, /e:/, was diffusing rapidly through the lexicon. In addition, as Beal (2004a: 132) rightly notes, Walker was likely to have been an innovator on the fringe of polite society and seeking to become part of it. If a feature was not removed from Irish English in the course of the 19th century, it may instead have been relegated to a vernacular mode with a more standard pronunciation representing default usage. Consider the case of what I call ‘BOLD diphthongisation’ (see 5.3 Conditioned realisations above). Sheridan examples – cowld /kauld/ ‘cold’ and bowld /bauld/ ‘bold’ as well as others – were still quite common well into the 19th century. However, what happened in supraregional forms of Irish English is that the standard English pronunciation with /o:/ was adopted and the forms with /au/ were confined to vernacular varieties. In addition a lexical split took

408

Raymond Hickey

place whereby a world like old with the diphthong, i.e. [aul], implies personal attachment, affection as in I’m afraid the aul car has finally cracked up.

7.

Conclusion

The aim of Thomas Sheridan in the late 18th century was to point out what he regarded as unacceptable features in provincial accents. How these were determined is not entirely clear, but certainly features which had become salient, the object of negative social comment and linked to vernacular varieties, were certainly among the preferred targets for criticism by Sheridan and other elocutionists who followed. A high degree of arbitrariness would seem to have been typical then and now. There are no clearly stated reasons why Sheridan should have denounced some features but failed to mention others. The writer Thomas Sheridan was in a way like John Walker, his British counterpart: he was an outsider to polite English society and just the sort of individual who would adopt emerging pronunciations in the language around him. Compared with Walker he was, however, more conservative given his Irish background and in some cases, as in the example of merchant with older [ar] rather than more recent [er], he recommended pronunciations which Walker did not. In common with Walker and other prescriptivists who followed him, Sheridan shared a strong conception of acceptable English, a socially preferred variety, based on non-local southeastern English usage with which vernacular forms were unfavourably compared (Mugglestone 2003). This is a specific development of the 18th century and can be seen by comparing the age of Swift, at the beginning of this century, with that of Sheridan towards the end of the century. In the intervening decades a sea change had taken place: concerns about the immutability of the English language had given way to concerns about what social groups spoke in what way. The avenue of sociolinguistic assessment and censure had been opened up and was never to close again.

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

409

Notes *

1.

2.

3.

4.

My thanks go to two anonymous reviewers who provided a number of useful and justified criticisms of an earlier version of this article, specifically regarding the need to view Sheridan and his pronouncements in a wider English context. Shortcomings in the final version are, as always, my own responsibility. Lowth also owes a debt to earlier grammarians, those of the previous century who were concerned with general principles of language and he states: “Grammar in general, or Universal Grammar, explains the Principles which are common to all languages” (Lowth 1995 [1762]: 1). Here we have an echo of 17th century authors such as Wilkins and their notions of a philosophical language (Barber 1975: 137–41) though Lowth is much more practical and concrete in his approach. Later a link between quality education and elocution became explicit and elocution became an established academic discpline. Again consider a case like the American Hallie Quinn Brown (1850–1949), who was an educator and elocutionist and instrumental in the setting up of women’s clubs for African Americans. In 1893 she was appointed professor of elocution at Wilberforce University, Ohio where she had studied as a young woman. One could speculate why Walker’s rather than Sheridan’s pronunciation guide was to survive into the 19th century (Sheldon 1947). One reason could well be that educated individuals in Victorian England (Phillipps 1984) were not inclined to use the work of an Irishman as a yardstick of correctness in their own language. See Mugglestone (2003: 44f.) for some comments on the possible influence of Walker and Sheridan on the pronunciation of English by the natives of Ireland.

References Barber, Charles 1997 Early Modern English. 2nd ed. Edinburgh: University Press. Bauer, Laurie 1994 Watching English Change. London: Longman. Beal, Joan C. 2004a English in Modern Times 1700–1945. London: Edward Arnold. 2004b “Marks of disgrace”: Attitudes to non-standard pronunciation in 18th-century pronouncing dictionaries. In Methods and Data in English Historical Dialectology. M. Dossena and R. Lass (eds), 329–49. Craig, Maurice 1969 Dublin 1660–1860. A Social and Architectural History. Dublin: Allen Figgis.

410

Raymond Hickey

Croghan, Martin J. 1990 Swift, Thomas Sheridan, Maria Edgeworth and the evolution of Hiberno-English. In The English of the Irish. Irish University Review 20:1. Terence P. Dolan (ed.), 19–34. Dublin: n.p. Crowley, Tony 2000 The Politics of Language in Ireland, 1366–1922. A Sourcebook. London: Routledge. Dobson, E. J. 1968 English Pronunciation 1500–1700. 2nd ed. Oxford: Clarendon Press. Hickey, Raymond 1986a Possible phonological parallels between Irish and Irish English, English World-Wide 7, 1: 1–21. 1986b Length and frontness with low vowels in Irish English, Studia Linguistica 39, 2: 143–156. 2000 Salience, stigma and standard. In The development of standard English 1300–1800. Theories, descriptions, conflicts. Laura Wright (ed.), 57–72. London: Cambridge University Press. 2001 The South-East of Ireland. A neglected region of dialect study. In Language Links: The Languages of Scotland and Ireland. Belfast Studies in Language, Culture and Politics, 2. John Kirk and Dónall Ó Baoill (eds.), 1–22. Belfast: Queen’s University. 2003 How and why supraregional varieties arise. In Insights into Late Modern English. Linguistic Insights, Studies in Language and Communication, Vol 7. Marina Dossena and Charles Jones (eds), 351– 373. Frankfurt: Peter Lang. 2005 Dublin English. Evolution and Change. Amsterdam: John Benjamins. 2007 Irish English. History and Present-Day Forms. Cambridge: University Press. Hogan, James Jeremiah 1927 The English Language in Ireland. Dublin: Educational Company of Ireland. Ihalainen, Ossi 1994 The dialects of England since 1776. In The Cambridge History of the English Language. Vol. 5: English in Britain and Overseas: Origins and Development. Robert W. Burchfield (ed.), 104–119. Cambridge: Cambridge University Press. Johnson, Samuel 1755 A dictionary of the English language. London: J. Strahan. Reprinted 1967 by The Ams Press, New York.

‘What strikes the ear’: Thomas Sheridan and regional pronunciation

411

Johnston, Edith Mary 1994 Eighteenth Century Ireland. The Long Peace. The New Gill History [1974] of Ireland, Vol. 4. Dublin: Gill and Macmillan. Johnston, William 1764 Pronouncing and spelling dictionary. London: The author. Jones, Stephen 1798 Sheridan improved: A general pronouncing and explanatory dictionary of the English language. 3rd ed. London: Vernor and Hood. Joyce, Patrick Weston 1979 English as We Speak it in Ireland. London: Longmans, Green & Co. [1910] Kenrick, William 1773 A new dictionary of the English language. London: John and Francis Rivington. 1784 A rhetorical grammar of the English language. London: Cadell and Longman. Kirk, John and Dónall Ó Baoill (eds) 2001 Language Links: The Languages of Scotland and Ireland. Belfast Studies in Language, Culture and Politics, 2. Belfast: Queen’s University. Lowth, Robert 1995 A Short Introduction to English Grammar. Reprint. London. [1762] Milroy, James and John Harris 1980 When is a merger not a merger? The MEAT/MATE problem in Belfast vernacular, English World-Wide 1: 199–210. Moody, Theodore and Francis X. Martin (eds) 1967 The Course of Irish History. Cork: Mercier. Mugglestone, Lynda 2003 ‘Talking Proper’. The rise of accent as social symbol. 2nd ed. Oxford: University Press. Ó Tuathaigh, Gearóid 1972 Ireland before the Famine 1798–1848. The Gill History of Ireland, Vol. 9. Dublin: Gill and Macmillan. Phillipps, K. C. 1984 Language and class in Victorian England. Oxford: Blackwell. Prins, Anton 1974 A History of English Phonemes. Leiden: University Press. Ray, John 1674 A collection of English words not generally used with their significations and original in two alphabetical catalogues. London: C. Wilkinson.

412

Raymond Hickey

Romaine, Suzanne 1998 The Cambridge History of the English Language. Volume IV 1776– 1997. Cambridge: University Press. Sheldon, Esther K. 1946 Pronouncing systems in eighteenth-century dictionaries, Language 22: 27–41. 1947 Walker’s influence on the pronunciation of English, Proceedings of the Modern Language Association of America 62: 130–46. 1967 Thomas Sheridan of Smock Alley. Princeton, NJ: University Press. Sheridan, Thomas 1761 A dissertation on the causes of the difficulties which occur, in learning the English tongue. London: R. and J. Dodsley. 1781 A rhetorical grammar of the English language calculated solely for the purpose of teaching propriety of pronunciation and justness of delivery, in that tongue. Dublin: Price. 1784 The life of the Rev. Dr. Jonathan Swift. London. Reproduced Garland 1974. 1967 A general dictionary of the English language. 2 vols. Menston: The [1780] Scolar Press. 1970 A course of lectures on elocution. Hildesheim: Georg Olms. [1762] Smart, B.H. 1836 Walker Remodelled. A New Critical Pronouncing Dictionary of the English Language. London: T. Cadell. 1842 The Practice of Elocution. 4th ed. Lond: Longman, Brown, Green and Longmans. Swift, Jonathan 1712 A proposal for correcting, improving and ascertaining the English language. Wakelin, Martyn F. 1977 English Dialects. An Introduction. 2nd ed. London: Athlone Press. [1972] Walker, John 1791 A Critical Pronouncing Dictionary and Expositor of the English Language. Reprinted by the The Scolar Press (Menston). 1815 A Critical Pronouncing Dictionary and Expositor of the English Language. 15th ed. London: A. Wilson. Wełna, Jerzy 1978 A Diachronic Grammar of English. Part One: Phonology. Warsaw: Panstwowe Wydawnictwo Naukowe.

Author index

A

Aarts, Bas, 375, 383 Abraham, Werner, 379 Ackema, Peter, 278, 298 Adams, Valerie, 266, 298 Adamson, Sylvia, 324 Akenson, Donald H., 360, 361, 378n, 379 Alexander, Michael, 123, 131, 132, 137, 141–143, 151 Allen, Cynthia L., 227, 301, 305, 306, 307, 309, 311, 320n, 321n, 323 Allen, Harold, 28, 50 Álvarez, Dolores González, 380 Andersen, Gisle, 275, 298 Anderson, Eric, 121 Anderson, John M., 301, 314, 320n, 323 Ash, Sharon, 48, 53 Attridge, Derek, 117 Avis, Walter S., 357, 362, 371, 379 Aycock, Joan Marie, 347, 353 B

Baayen, Harald R., 208, 210, 227, 236, 237, 270, 297n, 298 Bailey, Charles J.N., 325 Bailey, Guy, 19, 50, 51 Ball, Catherine N., 382 Baranowski, Maciej, 38, 48, 53 Barber, Katherine, 371, 379 Barbier, Sjef, 379

Barber, Charles, 409 Barton, Michael, 49, 51 Bately, Janet, 97n, 98 Bauer, Laurie, 214, 225n, 227, 235, 266, 298, 406, 409 Bausenhart, Werner, 360, 380 Beal, Joan C., 8, 229, 378, 389, 390, 391, 405, 406, 407, 409 Beckman, Mary E., 63, 76 Benskin, Michael, 58, 82, 97n, 98 Benson, Larry D., 156, 163, 175, 176, 177 Bergs, Alexander, 260n, 262, 326 Beths, Frank, 60, 78, 80 Beukema, Frits, 379 Biber, Douglas, 261n, 262, 267, 296n, 298, 378n, 380 Biewer, Carolin, 222, 229 Birner, Betty J., 2, 10, 65, 66, 71, 76 Bjørge, Anne Kari, 384 Bjork, Robert E., 151 Blackburn, Simon, 330, 335, 358, 380 Blake, Norman, 324 Bliss, A. J., 103, 118, 153 Bloomfield, Morton W., 359, 380 Boberg, Charles, 16, 46, 48, 51, 53 Bonnycastle, Stephen, 379 Borroff, Marie, 123, 134, 136, 139, 144–146, 151 Boucicault, Dion, 407 Bradley, David, 47, 51

414 Author index

Breton, Roland J. L., 341, 354 Brewer, Charlotte, 270, 273, 298 Briggs, Nancy E., 375, 384 Brinton, Laurel J., 277, 278, 298, 357, 380 Brogan, T. V. F. , 112, 118 Brunner, Karl, 57, 59 Bueno Alonso, Jorge L., 378 Bülbring, Karl, 55, 57, 81, 98 Burchfield, Robert W., 410 Burney, Fanny, 405 Burnley, David, 233, 234, 236, 247, 249, 259n, 262 Bybee, Joan, 255, 256, 259, 261n, 262, 311, 316, 320, 323, 381 C

Cable, Thomas, 3, 101–119, 125, 126, 133, 134, 151, 153, 159, 177, 205 Calan, Louise, 347, 354 Cameron, Angus, 56, 98 Campbell, Alistair, 55, 59, 82, 97n, 98, 115, 118 Caon, Luisella, 328, 336 Carver, Craig M., 44, 45, 51 Chambers, J. K., 18, 51, 353, 354, 359, 360, 361, 362, 380, 381, 384 Chapman, Don, 6, 7, 207, 210– 215, 222, 223, 227, 265–299 Chen, Matthew, 26, 51 Chickering, Howell D., 103, 118, 121, 123, 131, 132, 135, 137, 141–143, 151, 153 Chomsky, Noam, 104, 118 Claridge, Claudia, 267, 269, 278, 298 Clark, Eve V., 278, 298

Clarke, Sandra, 16, 51, 380 Clyne, Michael, 359, 381 Coates, Jennifer, 331, 335, 364, 365, 367, 368, 374, 381 Collins, Peter, 381 Connor-Linton, Jeffrey, 382 Conrad, Susan, 262, 298, 380 Contini-Morava, Ellen, 230 Corrigan, Karen P., 335 Cowan, Helen, 361, 378, 381 Cowie, Claire, 208, 227, 232, 235, 240, 241, 243, 244, 261n, 262 Craig, Maurice, 393, 410 Crate, Charles, 379 Croft, William, 220, 227, 302, 303, 304, 316, 317, 318, 319, 323 Croghan, Martin J., 393, 410 Crowley, Joseph Patrick, 96n, 98 Crowley, Tony, 333, 335, 399, 410 Cruse, D. Alan, 316, 317, 323 Curzan, Anne, 260n, 263, 380 D

Dalton-Puffer, Christiane, 232, 240, 241, 244, 246, 247, 262, 263, 378, 381 Delfs, Lauren, 72, 78 Denison, David, 5, 207–230, 259, 301, 305, 318, 320n, 320n, 324, 328, 336, 364, 378n, 381 Diewald, Gabriele, 326 Disner, Sandra, 14, 51 Dixon, Robert M.W., 321n, 324 Dobson, E. J., 332, 335, 395, 398, 410 Dolan, Terence P. 410

Author index 415

Dollinger, Stefan, 8, 327, 330– 331, 357–385 Dossena, Marina, 328, 335, 382 Drout, Michael, 122 Drysdale, Patrick, 379 Duggan, Hoyt, 105, 126, 133, 134, 151 Dury, Richard, 382 E

Eble, Connie, 8, 327, 328–329, Eckert, Penelope, 38, 49, 51 Edgeworth, Maria, 333, 399 Edwards, John, 380 Ehrman, Madeline E., 368, 374, 382 Elenbaas, Marion, 226n, 228 Elmer, Willy, 301, 305, 324 Elms, Ford, 51 Emmonds, Kimberly, 380 F

Facchinetti, Roberta, 364, 366, 367, 374, 382 Fasold, Ralph, 38, 51 Fastovsky, David E., 12, 51 Fee, Margery, 357, 380 Ferguson, Charles A., 23, 51 Finch, Casey, 123, 134, 136, 139, 144–146, 151 Finegan, Edward, 262, 298 Fischer, Kerstin, 326 Fischer, Olga, 76, 301, 306, 317, 324 Fisiak, Jacek, 382 Fitzgerald, Christina M., 3, 5, 101, 102, 113–116 Fitzmaurice, Susan M., 1, 201, 222, 228, 327, 335

Flasdieck, Hermann, 97n, 98 Fleischman, Suzanne, 232, 249, 259, 263, 381 Fletcher, Paul, 385 Fogel, David B., 358, 363, 383 Frank, Roberta, 82, 98 Franko, Stefan, 299 Frant Hecht, Barbara Frascarelli, Mara, 63, 77 Fraser, James B [Bruce], 212, 228, 266, 299 Fretheim, Thorstein, 64, 77 Fulk, Robert D., 3, 55, 56, 57, 58, 59, 81, 97n, 98, 103, 118, 153 G

Gadde, Frederik, 232, 263 Gisborne, Nik, 320, 326 Givón, Talmy, 64, 67, 77 Glass, Gene V., 158, 177 Glassie, Henry, 43, 44, 45, 52 Gneuss, Helmut, 97n, 98 Godden, Malcolm, 82, 98 Godfrey, J., 62, 77 Goldberg, Adele E., 220, 228, 304, 312, 321n, 324 González Álvarez, Dolores, 380 Gordon, Elizabeth, 332, 335 Gordon, Matthew J., 37, 52 Gore, Laura Locoul, 342, 343, 347, 351, 354 Görlach, Manfred, 371, 378, 382 Gotti, Maurizio, 382 Gourlay, Robert, 360, 382, Greenbaum, Sidney, 299 Gregg, Robert J., 18, 52, 381 Gregory, Michelle L., 62, 65, 77 Gries, Stefan Th., 226n, 228, 318, 324, 325

416 Author index

Guerra, Javier Pérez, 380 Gumperz, John J., 338, 354 Gundel, Jeannette, 62, 64, 65, 69, 77 H

Haeberli, Eric, 60, 78 Halle, Morris, 117, 118, 159, 161, 177 Hardy, Ashley K., 97n, 98 Harris, Barbara P., 381 Harris, John, 398, 411 Harrison, Keith, 123, 134, 136, 144–146, 151 Haudricourt, A. G., 27, 52 Haugland, Kari E., 384 Hay, Jennifer, 255, 259, 263, 335 Hayes, Bruce, 117, 118, 162, 177 Healey, Antonette diPaolo, 323, 324 Heaney, Seamus, 121, 123, 131, 132, 137, 138, 139, 141–143, 151 Hedberg, Johannes, 97n, 99 Hedberg, Nancy, 62, 64, 69, 77 Heikkonen, Kirsi, 325 Heine, Bernd, 324 Herold, Ruth, 22, 29, 42, 52 Herring, Susan C., 62, 77 Herzog, Marvin, 54 Hickey, Raymond, 8, 9, 328, 331– 334, 353, 373, 374, 381, 382, 387–413 Hilpert, Martin, 226n, 228 Hinterhölzl, Roland, 63, 77 Hogan, James Jeremiah, 402, 410 Hogg, Richard M., 2, 55, 58, 59, 76, 326 Holliman, E., 62, 77

Hopkins, B. R., 158, 177 Hopkins, Kenneth D., 158, 177 Hopper, Paul, J., 220, 302, 304, 308, 311, 312, 313, 317, 321n, 324, 325 Horn, Laurence J., 77 Hudson, Richard, 214, 215, 228, 302, 325 Hughes, Arthur, 385 Hultin, Neil C., 359, 383 Hundt, Marianne, 222, 228 Hutcheson, B.R., 131, 151 Hüttner, Julia, 384 I

Ihalainen, Ossi, 230, 401, 410 J

Jakobson, Roman, 15, 52 Jespersen, Otto, 31, 32, 52, 266, 272, 299 Johansson, Stig, 262, 298 Johnson, Samuel, 389, 393, 395, 396, 411 Johnston, Edith Mary, 393, 411 Johnston, William, 411 Jones, Charles, 332, 335, 336, 359, 383 Jones, Stephen, 391, 411 Jordan, Richard, 97n, 99 Joyce, Patrick Weston, 400, 411 Juilland, A. G., 27, 52 Jurafsky, Daniel, 312, 325 K

Kahneman, Daniel, 358, 384 Kaltenböck, Gunther, 384 Kaluza, Max, 155, 156, 178 Kastovsky, Dieter, 381

Author index 417

Kaufman, Terrence, 247, 264 Kehoe, Andrew, 381 Keller, Frank, 222, 229 Kemenade, Ans, van, 60, 61, 76, 77, 78 Kennedy, A. G., 212, 229 Kennedy, Charles W., 123, 125, 129, 131, 132, 136, 141–143, 151 Kenrick, William, 333, 336, 391, 396, 411 Kenyon, John, 32, 52 Ker, N. R., 88, 89, 90, 91, 92, 93, 94, 95, 99 Keyser, Samuel J., 117, 118, 159, 161, 177 Kiparsky, Paul, 159, 178 Kirk, John, 411 Kiss, Katalin É., 62, 63, 64, 77 Klaeber, F., 103, 118, 153 Kleinman, Scott, 97n, 99 Kniffen, Fred B., 43, 44, 45, 52 Knott, Thomas, 32, 52 Koch, John, 156 Kohonen, Viljo, 72, 73, 78 Koopman, Willem, 60, 76, 78, 324 Kroch, Anthony, 61, 72, 78 Krug, Manfred, 382 Kurath, Hans, 32, 52 Kwon, H.-S., 272, 299 Kytö, Merja, 325, 364, 368, 378, 383

Lakoff, George, 314, 325, Lambrecht, Knud, 62, 64, 65, 66, 78 Langacker, Ronald, 311, 316, 317, 325 Lapata, Maria, 222, 229 Lass, Roger, 8, 10, 189, 204, 335 Law, Vivien, 324 Lecoy, Félix, 155, 178 van der Leek, Frederike, 301, 306, 324 Leech, Geoffrey, 262, 298, 299 Leechman, Douglas, 379 Lehmann, Ruth P., 102, 104, 118, 123, 131, 137, 141–143, 151 Lennig, Matthew, 29, 53 Li, Xingzhong, 3, 5, 101, 106– 112, 155–179 Liberman, Mark, 159, 178, 225 Liljencrants, Johan, 14, 15, 53 Lima, Maria, 382 Lindblom, Bjorn, 14, 15, 53 Lindelöf, Uno, 55, 266, 299 Lippi-Green, Rosina, 339, 354 Lloyd, Cynthia, 225n, 229 Locke, John, 390 de Lorris, Guillaume, 155 Los, Bettelou, 61, 77 Lovell, Charles J., 371, 379, 383 Lowth, Robert, 333, 387, 389, 409n, 411 Luick, Karl, 27, 53, 55 Lutzky, Ursula, 384

L

M

Labov, William, 1, 2, 11, 17, 23, 27, 37, 38, 39, 42, 45, 47, 52, 53, 304, 325, 327, 336, 358, 383 Laing, Margaret, 58, 60

Maclagan, Margaret, 335 Mair, Christian, 378 Marchand, Hans, 225n, 226n, 229, 231, 232, 245, 263, 266, 299

418 Author index

Martin, Francis X., 393, 411 Marckwardt, Albert H., 383 Martinet, André, 11, 12, 14, 15, 27, 49, 53 Martínez, Esperanza Rama, 380 Masayuki, Ohkado, 73, 76, 78 McCone, Kim, 47, 53 McDaniel, J., 62, 77 McDavid, Jr., Raven I., 32, 52 McIntosh, Angus, 58, 60, 201n, 204 McIntyre, Andrew, 226n, 229 McKelvey, Blake, 41, 42, 53 Meillet, Antoine, 7, 10, 11, 53, 302, 315, 316, 319, 321n, 325 de Mendoza Ibáñez, Francisco J. Ruiz, 325 Menner, Robert J., 97n, 99 Merlo, Paolo, 325 Merwin, W.S., 105, 106, 123, 124, 134, 136, 139, 144–146, 150n, 151 Meurman-Solin, Anneli, 328, 336 de Meun, Jean, 155, 156, 157 Michaelis, Laura A., 62, 65, 77 Michalewicz, Zbigniew, 358, 363, 383 Miller, D. Gary, 232, 247, 263 Mills, Gary B., 340, 354 Milroy, James, 327, 338, 339, 354, 398, 411 Milroy, Lesley, 238, 263, 327, 336, 338, 339, 353, 354 Minkova, Donka, 1, 2, 10, 27, 54, 116, 118, 126, 152, 159, 178, 201n, 202n, 204, 205, 206 Mitchell, Bruce, 55, 56, 60, 99 Mondorf, Britta, 230 Moisl, Hermann L., 229

Moody, Theodore, 393, 411 Mossé, Fernand, 115, 118, 198, 204 Mugglestone, Lynda, 408, 411 N

Nearey, Terrance M., 48n, 54 Neeleman, Ad, 278, 298 Nelson, Gerald, 375, 383 Nesselhauf, Nadja, 222, 229 Nevalainen, Terttu, 79, 230, 327, 328, 336, 358, 383 Nightingale, Pamela, 239, 259, 260n, 263. Noël, Dirk, 320, 321n, 325 Nurmi, Arja, 79 O

Ó Baoill, Dónall, 411 O’Casey, Sean, 406 O’Dowd, Elizabeth M., 224, 229 Oakden, J. P., 119 Oaks, Michael P., 374, 383 Orkin, Mark M., 360, 383 Ó Tuathaigh, Gearóid, 393, 411 Ourioupina, Olga, 222 P

Pagliuca, William, 311, 323 Palander-Collin, Minna, 317, 325, Palmer, Chris C., 5, 207, 208–210, 222, 229 Palmer, Frank, 382 Parker, George, 379 Perkins, Revere, 311, 323 Peña Cervel, M. Sandra, 325 Pérez Guerra, Javier, 380 Peters, Pam, 382 Peterson, Paul W., 97n, 99

Author index 419

Phillipps, K. C., 409n, 411 Picone, Michael, 341, 354 Pierrehumbert, Janet, 63, 76 Pintzuk, Susan, 2, 55, 56, 57, 60, 61, 62, 76, 78, 79 Pope, John C., 97n, 99, 103, 119, 153 Pounder, Amanda, 250, 263 Pratt, Lynda, 221, 229 Preacher, Kristopher J., 375, 384 Prévost, Sophie, 72, 78 Prince, Alan, 159, 178 Prince, Ellen F., 2, 10, 62, 65, 66, 79 Prins, Anton, 401, 402, 411 Q

Quirk, Randolph, 217, 266, 299 R

Raffel, Burton, 104, 123, 131, 132, 135, 136, 140, 141–143, 150n, 152 Rama Martínez, Esperanza, 380 Raumolin-Brunberg, Helena, 328, 336 Ray, John, 401 Reddaway, T. F., 239, 264 van Reenen, Pieter, 62, 77 Reinhart, Tanya, 64, 79 Renouf, Antoinette, 381 Rissanen, Matti, 325, 378 Ritt, Nikolaus, 381 Roberts, Jane, 83, 97n, 99 Rohdenburg, Günter, 213, 229, 230 Roland, Douglas, 312, 325 Romaine, Suzanne, 228, 378, 381 Rosenbach, Anette, 222, 229

Rothwell, William, 238, 264 Rowling, J.K., 121 Rudanko, Juhani, 368, 383 Ryder, Mary Ellen, 211, 230, 276, 296n, 299 S

Saintsbury, George, 4, 10 Samuels, M. L., 58, 60 Sand, Lori, 50, 51 Santorini, Beatrice, 72, 78 Scargill, Matthew H., 18, 54, 359, 381, 384, Schabram, Hans, 97n, 99 Scharill, Matthew H., 379 Schendl, Herbert, 381 Schilling-Estes, Natalie, 374, 385 Schlüter, Julia, 62, 79, 226n, 230, Schneider, Edgar W., 378 Schøsler, Lene, 62, 77, 320 Schultink, Henk, 235, 264 Scragg, Donald G., 97n, 99, Sheehan, Peter M., 51 Sheldon, Esther K., 388, 393, 409n, 412 Sheridan, Thomas, 328, 331, 332– 334, 336 Shuy, Roger, 325 Siegel, Dorothy, 201 Simon-Vandenbergen, A. M., 364, 384 Sisam, Kenneth, 97n, 100 Skeat, Rev. Walter W., 155, 156, 165, 174, 175, 176, 177, 178 Skousen, Royal, 272, 299 Slovic, Paul, 384 Smart, B.H., 391, 412 Smit, Ute, 384 Smith, Adam, 382

420 Author index

Smith, Jeremy J., 82, 100 Smith, Jennifer Anh-Tran, 3, 4, 101, 102–106, 121–152, 153– 154 Smith, M. Estellie, 51 Smithers, G. V., 5, 10, 101, 102, 113, 114, 115, 119, 187, 188, 189, 198, 201, 202n, 204, 205 Smith-Thibodeaux, John, 342, 354 Smitterberg, Erik, 368, 383 Sørensen, Knud, 266, 296n, 299 Spevak, Olga, 72, 79 Speyer, Augustin, 61, 62, 79 Sprockel, C., 81, 100 Stefanowitsch, Anatol, 226n, 228, 318, 324, 325 Steiner, Richard, 37, 39, 50, 53 Štekauer, Pavol, 296n, 299 Stevenson, Suzanne, 325 Stockwell, Robert, 27, 54, 118, 126, 152 Stone, Brian, 123, 134, 136, 144– 146, 152 Sundby, Bertil, 367, 384 Suzuki, Seiichi, 103, 119, 153 Svartvik, Jan, 299 Swift, Jonathan, 333, 387, 388, 389, 393, 395, 399

U

T

Ukaji, Masatomo, 215, 230 Upchurch, Robert K., 97n, 100

Tarlinskaja, Marina, 107, 119, 159, 178 Taavitsainen, Irma, 230 Taylor, Ann, 60, 61, 72, 78, 79, 80 Taylor, John, 313, 314, 326 Taylor, Kathi, 382 Thomas, Eric R., 359, 384 Thomason, Sarah Grey, 247, 264

Thompson, Sandra, 220, 302, 308, 311, 312, 313, 321n, 324, 326 Tieken-Boon van Ostade, Ingrid, 222, 230, 328, 336, 378 Tillery, Jan, 51 Tobin, Yishai, 230 Tolkien, J.R.R., 105, 119, 122, 123, 133, 135, 136, 137, 144– 146, 152 Tomascikova, Slavka, 299 Tomasello, Michael, 312, 317, 326 Trager, George L., 42, 54 Traugott, Elizabeth Closs, 2, 55, 56, 57, 62, 76, 80, 226n, 277, 298, 302, 304, 307, 311, 315, 320, 325, 326, 364, 370, 379, 384 Trépanier, Cécyle, 342, 354 Trousdale, Graeme, 6, 7, 207, 215–221, 222, 230, 301–326 Trout, Robert O., 341, 355 Trudgill, Peter, 32, 43, 54, 334, 336, 359, 371, 373, 374, 384, Tuggy, David, 255, 264 Tversky, Amos, 358, 384

V

Vachek, Josef, 36, 49, 54 Vallduví, Enric, 62, 80 Van Bergen, Linda, 328, 336 Vantuono, William, 123, 134, 135, 136, 139, 144–146, 152 Venezky, R., 324 Véronis, Jean, 275, 299

Author index 421

Vincent, Nigel, 324 Vincent, Thomas, 379 Vleeskruyer, Rudolf, 81, 82, 83, 97n, 100 Vogt, W. Paul, 374, 384 W

Wakelin, Martyn F., 402, 412 Walker, John, 332, 333, 336, 390, 391, 405, 407, 408, 409n, 412 Walker, Lorna E. M., 239, 264 Wallis, Sean, 375, 383 Wang, William S.-Y., 26, 51 Ward, Gregory, 62, 65, 66, 71, 76, 77 Warkentyne, H.J., 18, 54 Warner, Anthony, 60, 79, 364, 385, Watts, Richard, 336, 371, 385 Weinreich, Uriel, 11, 54 Wells, John C., 15, 48, 54 Wełna, Jerzy, 401, 413 Wenisch, Franz, 97n, 100 Whitton, Laura, 61, 80

Wikle, Tom, 50, 51 Wilson, 82 Wimsatt, W. K., 117, 119 de Wolf, Gaelan Dodds, 371, 385 Wolfram, Walt, 374, 385 Wood, J. David, 361, 385 Woods, Anthony, 374, 385 Wright, Laura, 410 Wright, Susan, 324 van der Wurff, Wim, 61, 76, 80, 379 Y

Yaeger, Malcah, 37, 39, 50, 53 Yerkes, David, 86, 100 Youmans, Gilbert, 158, 159, 161, 179 Youssef, Amani, 51 Z

Zacharski, Ron, 62, 64, 69, 77 Zelinsky, Wilbur, 42, 54 Zink, Gaston, 198, 204 Zwicky, Arnold, 225n

Subject index

A

Ælfric, 57, 63, 74, 76n, 81, 82, 86, 91, 94, 96n Catholic Homilies I, 72, 73, 74, 75, 76n, 96 Catholic Homilies II, 74, 75, 76n Lives of Saints, 74, 75, 76n, 96 Æthelwold, 81, 96n actuation, 22, actuation problem, 11 affix, 208, 209, 225, 231–237, 240–258 detachability of affixes, 254, 258, 259 age, 6, 208–209, 231–259 age coefficient, 29, 30, 31, 34, 46, 50n aggregation, 207, 208, 240, 241, 244, 245, 246, 248, 254, 255 Alfred the Great, 82 algorithm, 40, 330, 358 Alliterative Revival, 127 alliterative verse, 123, 124, 126, 135 Middle English –, 122, 126, 129, 133, 139, 150n Old English –, 122, 127, 131, 132 alliteration, 4, 103, 105, 125, 127– 135, 140–145, 153 American English (AmE), 11–54, 220, 302, 330–331, 358, 362, 363, 366–370

Americanization, 341 Atlas of Northern American English (ANAE), 2, 12–18, 20, 22, 24, 26, 28, 31–32, 34, 37, 39, 41, 44, 46, 48n anapest anapestic feet, 105, 107, 171, 172, 173, 174, 176 anapestic substitution, 107, 171 Anelida and Arcite, 158 animacy, 67, 76n Anglian dialect features, 3, 58, 59, 81, 83, 84, 86, 88 antecedence, 64, 65, 67 apocopation, 188, 189, 190, 205 articulatory ease, 31 Augustan period, 387, 393, 397, 398 authorship, 5, 155, 174, 176 B

bad data, 8, 327, 328, 330, 331, 334, 357–359, 363, 374 Beowulf, 4, 101, 102, 117, 121, 122–125, 130–133, 136, 137 bilingualism, 8, 327, 328, 329, 337–352 Blickling homilies, 84, 89, 90, 91, 92, 93 Boethius, 63 bob and wheel, 124 Bodleian MS Laud Misc.108, 115, 187

424 Subject index

borrowing, 231–255 British English (BrE), 9, 330–331, 358, 359, 362–363, 366–370, 395, 399 British National Corpus (BNC), 211, 222, 267, 269, 270, 274, 275, 276 Brown Corpus, 56, 268 business writing, 328, 337, 344 Byrhtferth of Ramsey, 94 C

caesura, 138 CAN, 9, 358, 364–372, 375–376 root possibility sense, 330– 331, 365–369 permission sense, 330–331, 364–368 Canadian English (CanE), 16, 327, 330–331, 357–379, Standard Canadian English, 359 Canadian Raising, 359 Canadian Shift, 13, 16, 17, 18, 19, 21, 22, 26, 39 case marking, 217, 307, 308, 311, 320n category, basic level, 219, 223, 316, 317 category strengthening, 221, 222, 302, 311 Catholicism, 340, 341, 342, 344, 348, 389 causality, chain of, 11–13, 46 Cawdrey, Table Alphabetical, 235 chain shifts, 12–16, 26 Chaucer, Geoffrey, 5, 101, 106, 112, 155–176, 256 Canterbury Tales, The, 158

Book of the Duchess, The, 107, 158, 176 House of Fame, 107, 158, 176 Romaunt of the Rose, 5, 106, 117, 155–176 Fragment A (RomA), 111, 155, 156, 163–176 Fragment B (RomB), 111, 155, 156, 163–176 Fragment C (RomC), 111, 155, 156, 157, 163–176 Troilus and Criseyde, 158 children, 278, 347, 348, 352 Prudhomme children, 343, 346 coding system for corpora, 62, 64, 65, 66, 71, 72, 76n Cognitive Grammar (CG), 7, 255, 316 coinage, 209, 210, 234–236, 248 colloquial language, 269, 270, 276, 346 colloquial pronunciation, 391 colonialism, 8, 327, 357, language of colonials, 340, 358 colonial varieties, 358, 363, 370, 371, 373 composition, 140 oral, 134 visual, 134, 135, 138 compound words, 278 Construction Grammar, 207, 215, 218, 221, 222, 312 (Grammatical Construction Theory), 217 constructional taxonomy, 301– 305, 308, 311, 313–315

Subject index 425

corpus, corpus linguistics, 260n, 331, 358, 373, 374 corpora, 55, 56, 57, 61, 62, 208, 222, 236, 265, 268, 279, 297– 298n, 309, 328, 329, 330, 362, 373–374 annotated, 56, 66 historical, 55, 56, 57, 208, 222, 233, 237, 265, 270, 328, 362–365, 373–374 ICAME corpora, 268, 269, 274 multilingual, 237, 247 parsed, 56, 57, 61, 72, 318 spoken, 311 A Representative Corpus of Historical English Registers, version 1 (ARCHER-1), 236, 330, 362, 365, 369, 372–373 Corpus of Early English Correspondence (CEEC), 268, 328 Corpus of Early Ontario English, pre-Confederation section (CONTE-pC), 330, 362, 365, 373, 377, 379n Corpus of Late EighteenthCentury Prose (CL18P), 212, 328, 362 Corpus of Nineteenth-century English (CONCE), 368, 373 Corpus of Scottish Correspondence, 328 Helsinki Corpus of Historical English Texts, 56, 232, 236, 241, 244, 247, 268 Penn Parsed Corpora of OE, ME and eModE, 221,

Network of Eighteenth-century English Texts (NEET), 328, 335 York-Toronto-Helsinki Parsed Corpus of Old English Prose (YCOE), 57, 61, 62, 66, 72, 76n counts, see frequency creativity, self-conscious, 209, 221, 225n, 235, 250, 269, 272 creole society, 8, 329, 343, 345 D

derivational morphology, 6, 231– 259 borrowed derivational morphology, 231, 232, 246, 249, 254, 257, 258 diachrony, diachronic, 232, 235, 240–241, 252, 258, 276, 357, 366, 368–370 dialect geography, 18 regional English dialects, 373 provincial dialects, 341 dialect mixture, 56, 57, 81, 82, 83 diaries, 328, 330, 331, 332, 337, 345, 362, 366, 367, 368, 370, 377 Dictionary of Old English, 56, 96n Dictionary of Old English Corpus in Electronic Form, 85 diffuseness test, 175, 176 directional drift, 367, 368 double marking, 211, 213, 214, 215, 226n, 278, 279, 296n

426 Subject index E

-e, final, 105, 107 see also Middle English -er, 6, 207, 211, 213–215, 222, 224, 225, 265–297 Early English Books Online (EEBO), 237 Early Modern English (EmodE), 26, 218, 221, 225n, 232, 235, 302, 310, 359, 364, 371, 395 educated south-eastern English usage, 9, 392, 393, 395, 402, 405, 408 education, 329, 338, 346, 347, 348, 388, 390, 392, 396, 409n Eighteenth-century Collections Online (ECCO), 334 elicitation tests, 210 elision, 107, 113, 114, 115, 188, 189, 190, 198, 205 elocution, 332, 389–392, 409n elocutionists, 9, 328, 332, 391, 405, 408, 409n English fourteenth century, 187, 188, 198, 201, 232, 237, 238 fifteenth century, 6, 232, 237, 238, 239, 258, 272, 309 sixteenth century, 215, 272, 310, 332 seventeenth century, 215, 235, 272, 309, 310, 332, 390, 400 eighteenth century, 8, 328, 332, 359, 365, 367, 368, 387, 389, 392, 395, 405–406, 407 nineteenth century, 8, 9, 235, 272, 329, 359, 364–365, 368, 373, 401, 406–407

twentieth century, 272, 276, 279, 331, 332, 365, 373, 391, 406 twenty-first century, 276 English usage compared with Irish usage, 333, 395, 408 entrenchment, 316, 317, 320 Erie Canal, 40, 41, 42, 46, 47, 50n experiencer, 305, 306, 309, 318, 319, 321n Experiencer Construction (ExpCxn), 216, 217, 219, 220, 221, 305, 306, 307 F

foot, 107, 114, footed, 201, 202n also see iambic tetrameter French, 5, 8, 101, 198, 231, 237, 238, 245, 249, 327, 328, 329, 340, 345–348, 397 standard continental French, 341, 345, 346, 351 Acadian French, 337–338, 341–342, 351 frequency, 208, 210, 220, 221, 232, 234, 238, 244–245, 255, 318, 363, 373 token, 208, 210, 211, 240, 242, 255, 266, 270, 274, 363, 368 type, 208, 210, 220, 240, 241, 243, 251, 257, 270, 275, 278, 279, 312 focus, 62, 63, 64, 65, 66, 69, 70, 71, 73, 75 functional load, 23

Subject index 427 G

givenness, 62, 64, 69, 73 genre, 62, 74, 238, 245, 252, 260n, 330–331, 362, 363, 365, 366–369, 370, 377, 378n Goldsmiths’ Company, London, 239 Goldsmiths’ Mistery of London, Wardens’ Accounts and Court Minute Books of, 232, 239 Google, 210, 211, 214, 224, 225, 226n, 274, 275, 285 grammaticalization, 7, 207, 217, 218, 221, 222, 277, 278, 301– 304, 311, 314, 315–320, 364, 370 Great Vowel Shift, 26, 27 Grocers’ Company, records of, (see also London Grocers), 237, 245 H

Halle-Keyser Theory, 162 hapax legomena, 208, 210, 236, 240, 257, 266, 270 haplology, 213 Havelok the Dane, 5, 101, 114, 115, 116, 117, 187–200, 205– 206 headless line, 167, 176, 177n, 188, 190, 197 hemistich, 103, 104, 105, 125, 126, 138, 159, 161, 164, 167, 169, 170, 171 headless hemistich, 167 heuristics, 1, 8, 9, 55, 222, 330, 334, 363, 370

heuristic reasoning, 327, 330– 331, 357–358, 364, 367, 368, 371, 374 homilies, homiletic prose, 81, 83 anonymous homilies, 83, 84 horror aequi, 213 hybrid, 208, 236, 246–259, hypercorrection, 333, 334, 396, 398, 399 I

iamb, iambic phrase, 168 iambic pentameter, 117, 158, 168, 173, 174 iambic tetrameter, 101, 157– 160, 170, 173, 174, 176, 189 footed, 159 unfooted, 159 ICAME, see corpora immigration, 239, 360, 378n, immigrants, 22, 42, 340, 351, 360–363 impersonal construction, 7, 207, 215, 216, 217, 218, 220, 221, 301–305, 309–319 information structure, 61, 62, 63, 64, 74, 75 Inland North (IN), 19, 21, 37, 38, 39, 40, 41, 45, 46, 50n internet, 5, 207, 211, 213, 222, 223, 226n, 270, 274–279, 297n Irish usage, compared with English, 392, 393 Irish English, developments in the 19th century, 9, 334, 401, 406–408 /iw/ vs. /uw/ contrast, 32, 33, 34, 35, 36, 49n

428 Subject index J

Dr. Johnson, 333, 389, 393, 395, 396 K

Kaluza’s Law, 104, 153 Kentish, 84, 85, 87, 90, 91, 92, 93, 94, 95, 96 koiné, 43, 46 /Kuw/ class, 33, 34 L

Lancaster-Oslo-Bergen Corpus (LOB), 56, 268 language contact, 338, 341, 346, 360 language shift, 328–329, 338, 341, 352 Late Modern English (LModE), 8, 321n, 327–336, 357, 358, 359, 364, 365, 367, 368, 371, 373, 374 pronunciation of, 331–334 Latin, 72, 231, 237, 238, 316, 347, 349, 353n, 389 Latinate bases, 208, 235, 243 Left-dislocation, 61, 62, 65, 67, 70, 72 letters, 328, 330, 331, 332, 337, 349, 350, 352, 362–364, 366, 367, 370, 377 lexical isogloss, 43, 44, 45 linguistic propriety, 387, 388, 397 London Grocers, 6, 231, 232, 238, 239, 242–259 London, Port of, 238 long and ingliding subsystem, 17, 22

Louisiana, 8, 327, 328–329, 337– 353 low back merger, 17, 18, 29, 36, 42, 47, 48n lowering of /e/ before /r/, 393, 401–402 M

margins of security, 13, 15, 27, 37, 38 maximal dispersion, 14, 15, 18, 27, 47 MAY, 9, 358, 364–372, 375–376 Mercian features, 81, 82, 84, 85, 88, 90, 95, 96n, 97n meter, 3, 101–104, 121, 187–189, 196, 201n accentual-syllabic, 101 Middle English, 101, 102, 105 Modern English, 104 Old English, 101, 102, 104 methodology, 124, 153, 157, 234, 240, 244, 258, 265–266, 275, 331–334, 369–370, 373, 375 metrical beat, 101, 108, 110–113, 117 off-beat, 189, 191, 194 on-beat, 189, 194 silent beat dip, strong, 126, 128, 129, 133, 134, 144–146, 150n dip, unstressed, 213, 214 deviation, 102, 158–161, 169, 176 elision, 107, 113, 114, 115 evidence, 110, 155–174 foot, 107, 159, 160, 163, 164, 170

Subject index 429

lift, 125, 126, 127–132, 134, 135, 141–146 parameters, 161 position, 103, 104, 126, 131 prototype, 158, 159–161, 163, 165, 171, 176 resolution, 128, 161, 170, 171, 172 style, 160, 162, 166, 170 saliency, 104, 160, 161, 164, 168, 169 metrics, English historical, 101, 205 Middle English (ME), 25, 26, 58, 122, 124, 207, 215, 222, 231– 264, 309 French loan words in, 161, 190, 198, 199, 203n, 231, 238 genitives, 187, 206 inflectional endings, 188, 189–201 plurals, 187, 198, 202n, 206 verse, 136–140, 153 weak past participles final –ed(e), 5, 113, 187, 189, 190, 191, 192, 194, 196–201, 205, 206 final –es, 5, 113, 115, 197, 205 final –en, 113, 187, 189, 190, 205 final –e, 105, 113, 115, 187, 188, 189, 197, 198, 201, 205 Middle English Compendium, 237 Middle English Dictionary, 260n minimal pairs, 23, 24, 32 Missprachen, 58, 59

modal auxiliary verbs, 9, 302, 327, 330–331, 357–385 Modern English (ModE), 218– 220, 302, 311–319 monophthongization of /aw/, 22, 25, 49n monophthongization of /ay/, 15 morphosyntax, 207, 217 multiple inheritance, 220, 221, 321n N

nativization, 235 naturalisation, 6, 207, 208, 232– 237, 246–249, 254, 255, 258 neologism, 243, 244, 257, 260n newspapers, 330, 331, 362, 366, 367, 370, 377 nonperipheral track, 18, 20, 39 nominalization, 6, 208, 210, 211, 222, 223, 224, 225, 226n, 235, 237, 253, 265–297 North American English, see American English Northern Breaking, 39 Northern Cities Shift (NCS), 13, 21, 37, 38, 44, 45, 46, 48n O

Old English corpus, 6, 55, 59 Oxford English Dictionary (OED), 7, 207, 209, 210, 213, 221, 222, 226n, 259n, 260n, 269–280, 389 Old English, 25, 55, 58, 60–64, 66, 69, 72, 301, 305–307, 309, 317 anonymous homilies, 81, 82, 83, 84, 85, 86, 97n, 207, 215

430 Subject index

composite homilies, 95, 97n smoothing, 84, 87, 91, 93 syntax, 2, 55, 61, 64, 75, 207, 215–220, 301, 305–307, 319 verse, 126, 130–132, 153 Old French, loan words, 25 Ontario English (OntE), 9, 330, 357, 358, 359, 360, 362, 364, 368, 370–371, 373 Optimality Theory, 226n, Orm, 110 Ormulum, 159 Orosius, 74, 75, 76n, 305 orthoepists, 332 /ow/, fronting of, 28, 31, 33 P

particle, 211, 212, 213, 224, 226n, 265–297 Paston Letters, 212, 260n, 352, pattern, 71, 207, 210, 212, 215, 217, 265–280, 309, 314, 316, 366–367 coining pattern, 210, 212, 236 information pattern, 62, 74, prosodic pattern, 63 peripheral track, 17, 39, 45, personal writing, 328–329, 337, 344, 349 Piers Plowman, 117 phonemic length, 104 phonology, 235, English historical phonology, 333 phonotactic constraints, 191, 194, 196 Pittsburgh Shift, 13, 20, 21, 22, 26, 38 playfulness, 208–209, 225n, 231, 235, 259n

poetry, 104, 105, 112, 113, 116 poset (partially ordered set), 65, 66, 69, 71 pragmatic features, 57, 66, 303, 364 preposed dative experiencer (PDE), 309–311, 321n Present Day English (PDE), 62, 63, 64, 69, 72, 73, 231, 232, 235, 311–315, 330, 331, 363, 365, 368, 403, 407 PRICE lexical set, 405 probability matching, 15 productivity, 5, 208, 210, 225n, 232, 234, 235–258, 265, 266, 270, 274, 302, 311, 318 lexically restricted, 235, 236 local, 6, 207, 208, 233, 240, 241, 246, 248, 254 measure, 208, 236, 240 morphological, 236, 237, 248, 252 pronunciation features not proscribed, 405–406 prosody, 62, 71, 101, 108, 123, 189 Prototype Theory, 217 Prudhomme Family Papers, 8, 328–329, 337–355 pyrrhic, 162 R

r-less dialects, 15 r-pronouncing dialects, 15 Radical Construction Grammar, 302, 304 raising of /æ/, 38, 39, 44, 45, 393, 400, 406 raising of ME <ea>, lack of, 398

Subject index 431

regional accents, criticism of, 332, 387, 392 regional variation, 81, 82, 84 register, 74, 82, 210, 231, 232, 235, 238, 276 religion, 341 Rhetorical Grammar, 333, 388, 391, 396, 406 rhyme, rime, 153, 156, 157, 175, 176, 190 imperfect rhyme, 175, 176 rhythm, 153, 214 rhythmic alternation, 226n rhythmic template, 212, 213 rising middle classes, 391, 393 Riverside Chaucer, The, 156, 175 Roman de la Rose, 5, 155, 156, 157, 175 S

sample, 327, 328 random, 158 size, 158, 210, 237, 373, 374, 377 systematic, 157, 158 scansion, 124, 125, 128, 130, 153 rules of, 5, 125, 126, 128–129, 130–134, 138, 154, 187, 189, 199 schema, 307, 312, 313, 316, 317, 320 schematicity, 302, 304, 317 Sensitivity Principle of Trochees, 166 Sheridan, Thomas, 8, 9, 328, 331– 334, 388–413, Sheridan’s influence, 332, 334, 390–391, 409n

Sheridan’s system of pronunciation, 393–395 shift Back Chain Shift before /r/, 13 Back Upglide Shift, the, 13 Sievers, 103, 125, 126 Type A, 125 Type A3, 128 Type D, 125 Type E, 125 Sir Gawain and the Green Knight (SGGK), 4, 102, 121, 122, 124, 126, 136–140 smoothing of /aw/, 25, 26, 49n social networks, 8, 221, 222, 238, 259n, 260n, 327, 329, 337– 339, 342, 344, 353n density, 238, 329, 338, 339, 342, 343, 346, 352 multiplexity, 238–239, 329, 338, 339, 342, 343, 346, 352 network ties, family, 329, 342, 346 kinship, 329, 339, 343–344, 352 economic, 239–240, 329, 346, 352 social status, 342, 351, 352 SOFT lexical set, 405 Southern Shift, the, 13, 14 standard English, 334, 387, 393, 396, 400, 402 stress, 101, 124, 132, 153 linguistic, 101, 108, 112, 159 metrical, 111, 126 primary, 103, 212 secondary, 103, 125, 161 patterns, 110, 160, 161, 167, 172, 173

432 Subject index

doublets, 107, 161 peak, 162 phrasal, 110 sentential, 212 style Chaucerian, 176, non-Chaucerian, 169, 176 subject, 309, 311, 313–316, 318– 319, 320n, 321n subjecthood, 302, 311, 320n subsystems, 14, 15, 16, 18, 22, 26, 47, 48n suffix, 108, 208–210, 214, 225, 231–258 noun-forming, 207 supraregional varieties, 396, 401, 407 Swift’s rhymes, 395, 399 syllabicity, 187, 194, 200, 201, 205 syllable, 107, 110, 135 end-stressed syllable, 165 extrametrical syllable, 105, 107, 161, 169, 170, 171, 173, 176, 189, 198 missing syllable stressed syllable, 162 unstressed syllable, 128, 133, 134, 135, 144–146, 172, syncopation, 5, 114, 116, 187– 201, 205 syncope, 84, 87, 114, 190, 196, 205 syntax-pragmatics interface, 61 T

text types, 81, 261n, 363 topic, 62, 63, 64, 65, 67, 70, 71, 72

pragmatic topic, 64, 65, 67, 70, 71, 75, 76n syntactic topic, 64, 72, 75 topic persistence, 64, 70, 71, 73, 76n topicalization, 61, 62, 64, 65, 66, 72, 73 transitive construction (TrnCxn), 7, 207, 216, 219–221, 301, 302, 304, 308, 311, 313, 314– 318, 320n, 321n Type N, 216–218, 220, 226n, 306–309, 317, 319 Type T, 216–220, 307–309, 317, 321n transitivity, 302, 308, 311–312, 315 translation, 81, 86, 92, 96n, 101, 103–104, 121–140, 153, literal, 139 poetic, 121, 139 personal, 139 translatorship, 157, 159, 161 triggering event, 2, 11–15, 19, 22, 23, 46, 47, 48n triple marking, 213, 222, 226 trochee, 108, 111, 162, distribution of trochees, 107, 163, trochaic makeup, 107, 161, 165, 176, trochaic substitution, 105, 107, 112, 161, 162, 163, 164, 176, trochees of disyllables, 166, trochees of monosyllables, 162, 166 /Tuw/ class, 33, 34

Subject index 433 U

/uw/, fronting of, 27, 28, 29, 47 V

verb, impersonal, 216, 220 intransitive, 267, 272, 273 multi-word, 265–270, 272, 274–279, 297n phrasal, 207, 211, 212, 223, 224, 226n, 267, 272, 273, 277, 278, 296n prepositional, 207, 211, 223, 224, 226n, 267, 272, 273, 277 transitive, 220, 267, 272, 273 weak, 114, 115 verse, 105 a-verse, 105, 133, 153 b-verse, 105, 133, 153 on-verse, 125–131, 133–135, 141–145 off-verse, 125, 127–131, 133, 135, 141–145 hypermetric verses, 128, 131 Vercelli homilies, 84, 88, 89, 90, 92, 93, 94, 96, Vespasian Psalter, 90, 96n violations, positional, 4, 127–129, 131, 132, 141–148, alliterating, 4, 127, 128, 131– 132, 134, 141–148

vowels, North American English vowels, 16, 24, 32, 40 short vowels, 15, 18, 19, 47, 393, 395, 402 upgliding vowels back upgliding, 15, 16, 34, 35 front upgliding, 14, 15, 16 vowel lengthening before voiceless fricatives, 402 W

weighted score, 127, 129, 134 West Saxon, Late, 3, 57, 58 West Saxon Gospels, 74, 75, 76n Whitbread Book of the Year, 121 word formation, 235, 236, 240, 250, 251, 253, 255, 261n, 266, 267, 269, 270, 272, 296n word order, 72, 75, 218, 316, 320n OSV, 72, 73, 74, 75 OVS, 72, 73, 74, 75 word stress, 208, 210, 226n Wulfstan, 90, 93, 94, 95 Y

Yiddish Movement, 66, 71, 73, 76n /yuw/ class, 31, 32, 36, 47