Research Guide on Language Change 9783110875379, 9783110120462

Citation preview

Research Guide on Language Change

Trends in Linguistics Studies and Monographs 48

Editor

Werner Winter

Mouton de Gruyter Berlin - New York

Research Guide on Language Change

Edited by

Edgar C. Polome

Mouton de Gruyter Berlin · New York

1990

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

© Printed on acid-free paper which falls within the guidelines of the ANSI to ensure permanence and durability.

Library of Congress Cataloging in Publication

Data

Research guide on language change / edited by Edgar C. Polome. p. cm. — (Trends in linguistics. Studies and monographs ; 48) Includes bibliographical references. ISBN 0-89925-579-5 (alk. paper) 1. Linguistic change. 2. Reconstruction (Linguistics) I. Polome, Edgar C. II. Series. P142.R47 1990 417 —dc20 89-13909 CIP

Deutsche Bibliothek Cataloging in Publication

Data

Research guide on language change / ed. by Edgar C. Polome. — Berlin ; New York : Mouton de Gruyter, 1990 (Trends in linguistics : Studies and monographs ; 48) ISBN 3-11-012046-1 NE: Polome, Edgar C. [Hrsg.]; Trends in linguistics / Studies and monographs

© Copyright 1990 by Walter de Gruyter & Co., D-1000 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. Typesetting: Arthur Collignon GmbH, Berlin. — Printing: Gerike G m b H , Berlin. — Binding: Lüderitz & Bauer, Berlin. — Printed in Germany.

Preface

Habent sua fata

libelli.

This volume was planned in the early 1980s by the Subcommittee on Language Change of the Modern Language Association of America at the initiative of a number of historical linguists who wanted to provide a general guide for the student and scholar to the various aspects of language change. A group of prospective participants was contacted and manuscripts were collected by Edgar C. Polome to whom the task of editing the volume had been entrusted. In 1983, an eye operation paralyzed the activity of the editor for several months, and with the pressure of other commitments and requested changes in the editing procedure, continued work on the collected contributions was considerably delayed. When the editorial work could finally be restarted in 1987 with the active help of Linda Giles, the need to update the papers appeared evident. This was done in the course of the academic year 1987 — 1988, and it is this updated version which is now presented here, providing a wide coverage of the various wellestablished aspects of language change and the relevant methodology. We wish to thank all the authors most heartily for their extreme patience and most responsive collaboration with us in revising and completing this work. We also express our gratitude to Cynthia Sanches, Eric Dwyer, and Kathleen Linnes for their work in retyping manuscripts and assistance with data verification during the editing process. Personally, I would like to extend my deepest thanks to Linda Giles without whose gracious help, scrupulous attention, and indefatigable work the careful editing of this work would have been impossible. University of Texas at Austin

Edgar C. ΡοΙονηέ

Contents

I. Introduction Edgar C. Polome Language change and the Saussurean dichotomy: Diachrony versus synchrony

3

Werner Winter Linguistic reconstruction: The scope of historical and comparative linguistics

11

II. Aspects of Language Change Franklin C. Southworth Synchronic manifestations of linguistic change

25

Irmengard Rauch Evidence of language change

37

Joseph Salmons The context of language change

71

Methods to Study Language Change Thomas Cable Philology: Analysis of written records 97 Bridget Drinka The chronology of phonological change 107 Edgar C. Polome Linguistic paleontology: Migration theory, prehistory, and archeology correlated with linguistic data 137 Raven I. McDavid, Jr. Linguistic geography and language change 161 Carlota S. Smith Psycholinguistics: A research review 175

viii

Contents

John A. Rea Lexicostatistics Theoretical Models of Change Kurt R. Jankowsky The Neogrammarian hypothesis Archibald A. Hill A structural view of sound-change Robert D. King The transformational-generative model Other Approaches Peter Trudgill Dialect geography John Baugh Social stratification of language Franklin Southworth Contact and interference

217

223 241 249

257 273 281

III. Types of Language Change Phonological Change Elmer H. Antonsen Phonetic, phonemic, and phonotactic change Robert Kyes Evidence Elmer H. Antonsen Structuralist interpretation Robert J. Harms Synchronic rules and diachronic "laws": The Saussurean dichotomy reaffirmed Frans van Coetsem and Susan McCormick Morphophonology

297 303 307

313 325

Philip Baldi and William R. Schmalstieg Morphological change

347

Winfred P. Lehmann Syntactic change

365

Contents

Lexical Change Ladislav Zgusta Onomasiological change: Sachen-change reflected by Wörter John Algeo Semantic change John Algeo Borrowing Edgar C. Polome Etymology Change of Languages Henry M. Hoenigswald Language families and subgroupings, tree model and wave theory, and reconstruction of protolanguages Thomas L. Markey The development of standard language (koine) and dialect: Language split and dialect merger P. Sture Ureland Contact linguistics: Research on linguistic areas, strata, and interference in Europe Ian Hancock Creolization and language change Nicole Domingue Bi- and multilingualism: Code-switching, interference and hybrids Subject Index Language Index Author Index

ix

389 399 409 415

441

455

471 507

527 535 548 553

I. Introduction

Language change and the Saussurean dichotomy: Diachrony versus synchrony Edgar C. ΡοΙονηέ

It is obviously a truism to state that ever since man directed his attention to language, he was puzzled by the problem of its diversification — hence the story of the tower of Babel. As the idea of the original unicity of language prevailed for centuries, a sound reaction against referring every term to an Hebraic etymon only came in the eighteenth century, when Turgot (1756: 103) scoffed at those who would compare the Greek perfect passive participle ήμμενος of απτω 'fasten to' with Hebr. hemmen 'arranged, joined'. Change was ascribed to "an internal necessity", and it was assumed to be regular, respecting the internal organization of the language (Ducrot — Todorov 1979: 7). Scholars in the nineteenth century believed, with Jacob Grimm, that language attains its highest degree of development at the eve of the historical period when its process of decay starts. As the Old Indie religious language was assumed to be still very close to the protolanguage, this view was illustrated by referring to the richness in forms (Formenfülle) of Vedic, contrasted with the deflection of Modern English. As for the elaboration of the "ideal" protolanguage, August Schleicher assumed that it went successively through three phases characterized by 1) isolated monosyllables, 2) agglutination, 3) inflections (cf. Arens 1969: 253). The positivistic approach of the Neogrammarians changed these views in the last quarter of the nineteenth century by stressing that linguistic change could not simply be described: its causes had to be explained. They concentrated on sound change, as the "phonetic laws" operated mechanically — exceptions being accounted for by hitherto undiscovered "laws" or by psycholog-

4

Edgar C. Polome

ical explanations, especially the association of ideas involved in the analogical process. Having provided a magistral synthesis of the Neogrammarian achievements in the field of Indo-European vocalism and laid the foundation for the "laryngeal theory" at the age of 21 with the Memoire sur le systeme primitif des voyelles dans les langues indo-europeennes (Leipzig 1879), Ferdinand de Saussure felt in his later career the need to question the time-honored principle that the former state of the language is the underlying grammatical structure of the new state, and that the phonetic laws, linked with the needs of communication, are responsible for the progressive breaking up of the system. It is indeed assumed that a change like an accent shift placing a strong dynamic accent on the root syllable will entail the progressive loss of the case and verbal endings, and this deflection will, in turn, promote a change from a synthetic (i. e., inflectional) to an analytical syntactic pattern (with "function words" like prepositions or auxiliaries), as it can be exemplified, for example, in the diachronic development of Germanic or Indo-Aryan. After probing into the nature of language in the first part of his Cours de linguistique generale (Paris 1916), Saussure came to the conclusion that there was an inner duality in linguistics illustrated by the strictly synchronic approach of the Port Royal Grammar describing the state of French under Louis XIV and the method of comparative Indo-European philology attempting to reconstruct the past. For him (1959: 83), "the opposition between the two viewpoints... is absolute and allows no compromise". To demonstrate this, he resorts to the case of the German plural with umlaut: (a) in Old High German, masculine -/-stems like gast 'guest' would form their plural in -i, but this -i would produce umlaut of the a of the stem, hence: *gasti —> gesti, later geste —> modern German Gäste; (b) in modern German, a set of terms forms its plural on the pattern of Gast, Gäste. Synchronically, the alternation a : ä in Modern German is simply a mechanism to indicate the plural. On the other hand, in Old High German, the change of a to e under the influence of the i of the following syllable is a phonological process which has nothing to do with the pluralization of nouns, since the same umlaut occurs, e. g., in the 3rd person singular present in tragit —* tregit —> Mod. German trägt. Obviously, diachronic facts have not changed the system: "modification does not affect the arrangement but rather its elements" (1959: 84).

Language change and the Saussurean

dichotomy

5

Having strongly stressed the dichotomy between synchronic and diachronic linguistics, Saussure remarks at the very beginning of the third part of his Cours (1959:140): "What diachronic linguistics studies . . . i s relations between successive terms that are substituted for each other in time." Broadening this statement to mean "successive grammars" one comes indeed very close to the way generative grammar considered language change. However, one should avoid hasty generalization. Robert King (1969: 85) presents the generative model of language change through three generations schematically as in Figure 1. Generation 1

Grammar

Generation 2

Language Acquisition Device

I Optional Grammar

\

Speech Output

Speech Output A

+

\

Innovations

I Adult Grammar Generation 3

Speech Output Β

Language Acquisition Device

Figure 1. King's model of language change

Saussure (1959: 149) remains rather skeptical about the transmission of "faulty pronunciations" through children, but what he has in mind are sporadic changes, and not the process of rule simplification implied by King's matrix. Saussure (1959: 98) states explicitly that "each change is launched by a certain number of individuals before it is accepted for general use", but not all innovations receive this recognition! As long as they remain individual, the linguist may safely ignore them, since they only actually enter his field of investigation after adoption by the whole community. Therefore, Winfred Lehmann (1968: 15) considers that, for Saussure, the "speaker" is ineffective in initiating and even controlling language change. A typical illustration of this view is Saussure's analysis of the process of analogy as "the chance product of an isolated speaker", when he shows that it actually depends on the linguistic mechanism linking productive forms stored by the language and "arranged according to their syntagmatic and associative relations". This

6

Edgar C. ΡοΙοπιέ

contrasts clearly with the position of Jerzy Kurylowicz (1966: 171), who stresses that the innovation is due to causes external to the linguistic system and that its start and spread depend on the social context. 1 For Saussure, language is a social institution that has its own internal arrangement. He frequently compares the latter to a game of chess: in language, one moves from one synchronic state to the next, as one moves only one piece at a time on the chessboard. But, just as one move can change the whole game, and sometimes the chessplayer may not even foresee all its effects, one change in language may have a repercussion throughout the whole system. However, "each move is absolutely distinct from the preceding and the subsequent equilibrium" (1959: 89): whichever way a state in the game has been reached is irrelevant, and someone looking at the chessboard at that very moment can assess the situation without having to know what happened just seconds before. Similarly, as speaking operates on a synchronic level, the changes that affected successive states of the language are irrelevant to it. Thus, the reference to parole (speaking) stresses the radical difference between diachrony and synchrony in Saussure's mind. He constantly points out that any individual creation in language is doomed unless it is taken over by the speech community, at which time the individual loses all control over it. Thus, "a language never exists apart from the social fact" (1959: 77), and while society insures its continuity, language is, like all social institutions, subject to change in time. But when it comes to specifying the external social forces responsible for change in language, Saussure doubts the effects of political and economic upheaval on linguistic stability: for him (1959: 150), the changes that Latin underwent during the turbulent period of the Germanic invasions are "self-generated", and not ascribable to external conditions. In the light of Pulgram's views (1958: 317 — 23) on spoken and written Latin, Saussure is obviously right in assuming that the relative stability of classical Latin is due to external factors, such as the cultural influence of the Roman intelligentsia on the imperial court, on the administration, on law and education. With the breakdown of this influence and the accession of different social strata to power, the spoken language which had developed unhampered by external factors since the early Republic became prominent and was ultimately the source of written "Romanic" and of the "Romanic" vernaculars — an evolution which is clearly illustrated by the texts presented by Ernst Pulgram (1978). Saussure's doubts about the belief

Language change and the Saussurean

dichotomy

1

that turbulent periods in the history of a nation precipitate linguistic evolution, are accordingly justified, and they would deserve further investigation, together with his statement (1959: 150) that what actually takes place is a return of the language to its "free" (— uncontrolled) state and its natural evolution. An important point that Saussure makes in this context is that "it is impossible to cite a single period — even among those where language is in a deceptive state of immobility — that has witnessed no phonetic changes" (1959:151). Language is in constant flux through time, modifying and renewing itself. It is, therefore, rather unlikely that the language pattern of ancient Europe would have remained practically unchanged between 4500 B. C. and 1000 B. C., as it has been assumed to accommodate some archeologically based diffusion theory (cf. Renfrew 1987: 164). But it has long been recognized that geographical spread favors diversification through time: Saussure himself devotes the fourth part of his Cours to this topic and recognizes (1959: 208) that, within a unilingual coherent whole, geographical continuity goes together with perpetual differentiation, each innovation spreading over its particular area. The importance of linguistic geography for historical linguistics has indeed been abundantly illustrated since the formulation of the "wave theory" by Johannes Schmidt in 1872 (cf., e.g., Anttila 1972: 3 0 4 - 0 9 ; Bynon 1977; 192-95). However, as Lehmann (1973: 137) rightly points out, the study of dialect geography has shown that a bidimensional model, reflecting the areal subdivisions of a language and the social and occupational groups speaking it, is inadequate, even when supplemented by a time dimension. The reason for the inadequacy of the models Saussure used is the awareness of the extreme complexity of the social conventions involved in language that sociolinguistic studies have given us. The pioneering work of William Labov (1966, 1972) has provided a new insight into the social stratification of language and, by a proper evaluation of the linguistic variables, it has broken through the Saussurean dichotomy of synchrony and diachrony to demonstrate the synchronic reflection of historical change 2 and analyze the mechanics of language change within its social setting. Apropros of the latter, Hans H. Hock (1986: 661) makes a very important observation: to the extent that the Labovian studies on analogical and syntactic change are focused on change in progress, they show a direct correlation between social factors and linguistic change — which, again, runs contrary to the

8

Edgar C. ΡοΙοητέ

Saussurian contention that external events have only an indirect influence on the internal development of the language. I would conclude, with Lehmann (1968: 19 — 20), that, while recent work clearly departs from Saussure's views and his sharp dichotomies, his achievements cannot be disregarded, and his challenge has definitely triggered responses providing a better insight into the problems of language change 3 .

Notes 1. The problem of analogy has remained a disputed issue: the principles presented by Kurytowicz in his 1949 article (1966) are summarized and supplemented in his book on The inflectional categories of Indo-European (Heidelberg: Carl Winter, 1964), 38 — 55. For a discussion of Kurytowicz' views, cf. Anthony Arlotto's chapter on "Analogy" in his Introduction to Historical Linguistics (Boston: Houghton, Mifflin, 1972, 130 — 148). A different approach is presented by Witold Mariczak in his "Tendances generates des changements analogiques" (in Lingua, 7 [1958], 298 — 325, 3 8 7 - 4 2 0 ) . On Kurytowicz versus Manczak, cf. Anttila 1977: 7 6 - 8 0 . About the treatment of "analogy" in transformational grammar, cf. King 1969: 1 2 7 - 1 3 4 ; Anttila 1977: 8 7 - 1 1 0 ; Hock 1986: 2 3 8 - 2 7 9 . About analogy in general, see Anttila 1977; Hock 1986: 167 — 279. 2. That synchronic grammar may recapitulate history was recognizcd by transformational-generative grammar, with the reservation that it might not do so in rule ordering, and that historical evidence should not be considered as directly relevant to the evaluation of synchronic grammars (cf. King 1969: 101 — 104). 3. Unless we subscribe to the rather extreme critical skepticism of Roger Lass in his challenging book On explaining language change (Cambridge Studies in Linguistics, vol. 27; Cambridge [U.K.]: Cambridge University Press, 1980)!

References Anttila, Raimo 1972 1977

An introduction to historical and comparative linguistics. New York: MacMillan/London: Collier — MacMillan. Analogy (Trends in Linguistics: State-of-the-Art Reports 10). The Hague/ Paris/New York: Mouton.

Arens, Hans 1969 Sprachwissenschaft. Der Gang ihrer Entwicklung von der Antike bis zur Gegenwart (2nd thoroughly revised edition). Freiburg/München: Karl Alber. Bynon, Theodora 1977 Historical linguistics (Cambridge Textbooks in Linguistics). Cambridge/ London/New York: Cambridge University Press.

Language change and the Saussurean dichotomy

9

Ducrot, Oswald — Tzvetan Todorov 1979 Encyclopedic dictionary of the sciences of language. Trans. Catherine Porter (Originally published as the Dictionnaire encyclopedique des sciences du langage, Paris: Editions du Seuil, 1972). Baltimore/London: The Johns Hopkins University Press. Hamp, Eric P. — Fred W. Householder — Robert Austerlitz (eds.) 1966 Readings in linguistics, II. Chicago: University of Chicago Press. Hock, Hans Henrich 1986 Principles of historical linguistics. Berlin/New York/Amsterdam: Mouton de Gruyter. King, Robert D. 1969 Historical linguistics and generative grammar. Englewood Cliffs, N.J.: Prentice-Hall. Kurylowicz, Jerzy 1966 "La nature des proces dits 'analogiques', in: Hamp —Householder —Austerlitz (eds.), 158 — 174. (Originally published in Acta Linguistica 5 [1945 — 49]: 121-138). Labov, William 1966 The social stratification of English in New York City. Washington, D. C.: Center for Applied Linguistics. 1972 Sociolinguistic patterns (Conduct and Communication 4). Philadelphia: University of Pennsylvania Press. Lehmann, Winfred P. 1968 "Saussure's dichotomy between descriptive and historical linguistics", in: Lehmann — Malkiel (eds.), 3 — 20. 1973 Historical linguistics: An introduction (2nd edition). New York: Holt, Rinehart, and Winston. Lehmann, Winfred P. —Yakov Malkiel (eds.) 1968 Directions for historical linguistics: A symposium. Austin/London: University of Texas Press. Pulgram, Ernst 1958 The tongues of Italy: Prehistory and history. Cambridge, Mass.: Harvard University Press. 1978 Italic, Latin, Italian 600 B. C. to A. D. 1260: Texts and commentaries. Heidelberg: Carl Winter. Renfrew, Colin 1987 Archaeology and language: The puzzle of Indo-European origins. London: Jonathan Cape. (Published by Cambridge University Press, New York, 1988, in the U.S.). Saussure, Ferdinand de [1959] Course in general linguistics. Trans. Wade Baskin (Originally published as Cours de linguistique generale. Charles Bally and Albert Sechehaye (eds.). Paris 1916). New York: Philosophical Library. Turgot, Anne Robert 1756 "Etymologie", in: Encyclopedic ou Dictionnaire raisonne des Sciences, des Arts et des Metiers. Vol. VI. Ed. Diderot and d'Alembert. Paris: Briasson/ David/Le Breton/Durand, 1756, 98 — 111. ([compact] reprinted New York/ Paris: Pergamon Press, 1969. Vol. 1, pp. 1283-86).

Linguistic reconstruction: The scope of historical and comparative linguistics Werner Winter

As long as there has been a scientific study of language, there has been an interest on the part of linguists in two major aspects of the field. One may be expressed by the simple question: What is a language like? — the other, slightly more involved, by: What made a language like it is now? The first question covers all kinds of synchrony-oriented investigations, no matter whether they are concerned with rather simple matters, for example, the inventory of certain words or parts of paradigms, or whether they involve a high degree of abstractness and sophistication, say, regarding preconditions or purposes of some facets of language use. The second question is directed toward the history and prehistory of a language. Simple observation tells us (though usually in rather marginal contexts) that languages do not remain stable through time, but that they change. It is thus not trivial to ask the second question; and as was the case with synchronic investigation, diachronic study of language is very complex in itself. First of all, there is the painstaking analysis of information derivable from attestations of the language under consideration that date from the past. Except for very recent data, all attestations will be through the medium of writing. The written record is never an exact rendering of its spoken counterpart; a writing system imposes limitations on what can be recorded of the properties of a language. Different writing systems are characterized by different degrees of appropriateness for the expression of features of spoken language; some of such features are almost always suppressed even in otherwise quite reliable tran-

12

Werner

Winter

scription systems, mainly because the competent speaker-reader can be expected to supply these features when converting written language back to its spoken form. No person studying the texts of a language of the past now can be a fully competent native speaker of this language. Historical study of a language thus requires, to be sure to varying degrees, an interpretation of the data even if one's immediate goal has been set fairly low — if, for instance, one is only concerned with the question: What may this written text have sounded like? The more demanding one's questions, the more difficult one's analysis of written texts becomes, for all texts from times past have one property in common: not only are they imperfect as renderings of actually spoken language, but beyond that, they are always incomplete as far as their coverage of relevant phenomena is concerned, and no gap in the pattern of information can be closed by resorting to the knowledge of a competent native speaker. Therefore, not only the interpretation of what is found in the texts requires the development of frequently rather elaborate hypotheses, but one's goal of discovering and evaluating the patterns of this early language can only be attained to the extent that one can come up with sound hypotheses concerning the unattested parts of these patterns. Direct observation is out of the question, and so is the use of an informant; instead one has to use one's own understanding of the properties observed to extrapolate assumptions about the unobserved and unobservable. Work on data extracted from written documents of the past one may call historical linguistics in the narrow sense of the word. To start with, such work need not be, and often is not, diachronically oriented; like living languages, dead ones may be described and analyzed in their own right. But whenever a dead language can be connected with a living one, there will be a strong tendency to look at the two together, particularly if the dead one can be considered an antecedent of the living language. When this is done, diachronic analysis replaces its synchronic counterpart. For generations of linguists, such diachronic analysis constituted the first and most important task of all scientific linguistic activity. The reason behind this attitude is to be seen in the conviction that to know the origins and the subsequent stages in the development of a phenomenon meant to understand and therefore to be able to explain it. The fact that for later generations there was a complete reversal of attitude need not concern us here — the reason being that since languages could be observed to be fully functioning here and now,

The scope of historical and comparative linguistics

13

used most successfully by speakers totally unaware of any aspect of the earlier history of their language, there was obviously no need for bringing in history as a basis for explanation of linguistic phenomena, the more so as there were very many languages indeed without a written record of past stages in their development. By now, a certain modus vivendi has been reached: Both points of view are seen to have their own merit — the one helps elucidate problems the other cannot accommodate, and a judicious use of both approaches contributes much more than an obstinate fixation on one of the two could ever achieve. Very early in the development of linguistics as a science, it was noticed that the domain of diachrony-oriented linguistics could be significantly, even dramatically, extended by studying not only the history of a language as attested in written documents, but its prehistory as well. On first consideration such as step seems preposterous to propose and to take. Human speech is a most perishable commodity, and what evidence could there possibly be that would have preserved for us even mere traces of earlier properties of a language if we were to reach out beyond the realm of written materials? However, there is one very important characteristic of human language that enables us to extend the domain of linguistic research to preliterate periods of the past. It can be stated as a general rule that the sign inventories of all natural languages are made up essentially of entities that are arbitrary, that is, only very rarely are properties of what is referred to in the world around us reflected in the form of linguistic signs (even where onomatopoeia and iconicity are involved one observes relatively strong variation from one language to the other). For a language to function as a means of communication, it is necessary that the basically arbitrary signs be conventionalized within each speech community — no speaker can be permitted to use only signs of his own choosing, or else he could not have partners who would understand him and whom he would understand. If signs are arbitrary to start with, then agreement, complete or partial, between signs denoting the same feature or features of the universe in two or more different languages must be in need of an explanation. Apart from the relatively few cases of only partial arbitrariness, all instances of agreement or similarity require languageinternal explanations. There are essentially three types of explanation:

14

Werner

Winter

a) Similarity may be due to chance. Chance agreement does occur, but only very rarely; agreement of this type is virtually nonexistent between forms of some complexity. All in all, once we find a moderately high number of cases of complete or even only partial agreement, chance can be ruled out as a possible cause. b) Similarity may be due to borrowing. If signs in two languages show a very high degree of similarity, and if the two languages are known to be, or to have recently been, in contact, borrowing is very likely to provide the explanation for the agreement. However, the more remote in time a borrowing process took place, the more difficult it becomes to prove borrowing, as all components of a language, whether they were borrowed or not, are affected by language change as time goes on, and differences between foreign and native elements tend to be obscured. Moreover, once reliable historical documents are no longer available, we cannot claim with ultimate assurance that the contact condition for borrowing really can be met. c) Similarity of items in two or more languages compared, finally, may be due to these items deriving from a source which represents an earlier stage in the development of the languages compared, common to all of them. By showing such properties, languages are said to indicate genetic relationship, and the shared features are called inherited. The use of these terms from life sciences, as is so often the case with the adoption of an outside terminology, should be taken with a grain of salt: A parent language and a daughter language are not individual entities clearly set off against each other, but more or less arbitrarily selected subcontinua in an unbroken chain of development; features proper to this chain are not inherited, but among the features acquired by a learner in his language-acquisition process, these happen to be the ones that originated within the language (or language group) itself, while borrowed items are those taken over by means of an interlanguage acquisition process. (It goes without saying that the picture drawn here is slightly idealized — in most instances, borrowed items will have been part of the language of the learner for a while, so that what really happens here is also an intralanguage acquisition process.) The task of the prehistorian of language is the conversion into a coherent picture of observations about similarity that can only be explained by positing an earlier common source for the "inherited" components of the languages under investigation. To do this, the

The scope of historical and comparative

linguistics

15

prehistorian has to go through a number of carefully planned and executed steps. His immediate goal is to replace the vague notion of similarity or loose partial agreement (a notion which forms the basis of work in lexicostatistics, glottochronology, and large-scale comparison aimed at establishing distant relationships among languages and language groups) by the much more stringent one of regular correspondence. Ideally, this will imply the discovery of regularities of the type "If, in language L,, a feature X in an environment Zj is matched, in language L 2 , by a feature Y in an analogous environment Z 2 , then the agreement will recur in all other occurrences of the environmental class Ζ , , Z 2 ". Any deviation from this regularity will have to be explained, either by determining deviant syntagmatic conditions or by proposing the influence of different paradigmatic conditions. Insistence on systematic, recurrent agreement not only does away with the impressionistic notion of similarity but can also afford to go much beyond this by allowing for regular correspondence between overtly divergent features, thus eliminating a recourse to similarity altogether except as a crude heuristic device. The steps to be taken are as follows: First is a collection of data. Incidences of recurrent agreement are brought together in lists, with all variation, however slight, given proper attention: the application of the principle "Only one set of correspondences per list" is likely to result in a fairly large number of such lists. Next comes an evaluation of the status of the lists. A close study of the environments is carried out, aimed at determining whether some lists can be interpreted as syntagmatically conditioned variants of other lists, thereby reducing the number of lists of the first order. Ideally, a further reduction should take place by the identification of paradigmatically determined variants; as a rule, however, this task cannot be properly fulfilled, since that presupposes a rather thorough knowledge of the grammatical and semantic paradigms of the languages investigated. Usually, one has to be content with suspecting paradigmatic interference whenever syntagmatically conditioned deviation can be ruled out and when — an important practical consideration — a deviant correspondence chain is much less frequently attested than its suspected regular counterparts are. Once the number of correspondence chains of the first order has been reduced to manageable proportions, each of the chains which are not thought to be variants of others, is labeled. Each label is to be

16

Werner

Winter

used only once. In principle, anything unambiguous would do for a label; thus it would be perfectly conceivable to use numbers for labels. However, a very useful and generally accepted convention rules out this possibility: the label chosen is to be maximally similar to the parts of the chain it identifies. Therefore, if the features subsumed under a chain are sounds, the label should be the symbol for a sound. The sound denoted by the label must not deviate more than absolutely necessary from the sounds contained in the chain. Hence, if all sounds in a correspondence chain are identical, the sound indicated by the label must be identical with the sounds forming part of the chain. If the members of the chain are not identical, the sound denoted by the label should still be maximally similar to all of them, that is, the difference in terms of distinctive features between member sounds and label sound should be kept to the bare minimum. Obviously, with a mixed chain, there can be more than one choice for a label; however, a limitation arises immediately from the requirement that each label only be used once. Thus, the obligatory selection of labels made for unmixed chains will automatically reduce the number of alternatives available for mixed chains; a further reduction will become obvious as we proceed further in our discussion. The requirements of biuniqueness and of maximal similarity of chain labels to the members of their respective chains are sufficient for an enumeration of all chains observed in an unambiguous and reasonable manner. However, were we to stop at this point, we would only have described regular correspondences existing between languages observable and observed; we would not have taken a very decisive step beyond achrony. This step rests on the introduction of a most important assumption: If we think that the large-scale recurrent agreement we observe is the result of a common descent of the languages from which we have drawn our data from some (possibly quite remote) "parent language", then our labels could be taken to be representations of entities to be ascribed to this parent language. This assumption immediately introduces another requirement to be made for the choice of our labels: Not only must they be used only once and be maximally similar to the members of the chains they stand for, but they must also be relatable to one another in such a way that they can be interpreted as parts of one reasonable pattern. The evaluation criteria, at this point, are typological: the more outlandish a pattern which has resulted from the labeling process, the more likely

The scope of historical and comparative

linguistics

17

it is that some of the labels need adjustment; which may mean that rejected alternative labelings now have to be reintroduced in order to make the overall pattern appear more natural. With the last two steps we will have passed the threshold between "mere" comparison (which would be on a level with typological and with contrastive-confrontative comparison) and comparative reconstruction. We no longer merely describe what is there in one language in relation to what is found in its "sister languages", but we now make claims as to how what is there got to be there, and as to what was there at an earlier stage in the development of this language group that would help account for what is met in the directly observable languages. We have left the domain of formulas expressing regular correspondences and reached the point where achronic formulas are taken to be reflections of facts about a stage or stages of a language of times long past. This protolanguage is considered similar to observable languages in that it is endowed, through our reconstruction, with properties that seem fitting for natural languages, although we can never be quite sure that the results of our reconstruction will not lead us here or there beyond the confines of what is known from languages observed; still, information about languages in actual use provides an extremely helpful yardstick against which to measure our reconstructions. We should not disregard the fact that there are several limitations to our ability to reconstruct languages of the past in all their richness. First of all, the comparative method can be applied only if there is something to compare. Features of the protolanguage lost in the course of later developments in all daughter languages are irretrievable in this context; even where we find an isolated phenomenon without a match in the related languages, we cannot apply comparative techniques, and we cannot, again only in the present context, determine whether it is a lone survivor of the past, a recent development in this particular language, or a borrowing from a source not identified — the latter two possibilities having no bearing upon the reconstruction of the protolanguage. It is in some ways even more troublesome, when we find features attested only in languages of which we have reason to suspect that they were spoken in adjacent territories at some time between the periods of the protolanguage and of the observed languages, so that we cannot be sure whether the features shared were inherited or were taken over by intrafamily borrowing. Thus, ideally, only widely attested forms from languages nonadjacent to one another

18

Werner

Winter

should be used as a basis for reconstruction. In the case of ProtoIndo-European, no major difficulties arise from the application of this principle as long as we are interested in the reconstruction of the phonological system and paradigmatic morphology of late Proto-IndoEuropean (although even in the realm of paradigmatic morphology, different interpretations of the sets of facts are possible), but what can be reconstructed of the lexicon, rich as it appears to be, seems to be only a sadly deficient fragment of the Proto-Indo-European vocabulary. This state of affairs is, of course, not surprising at all, since we know from the closer inspection of languages with a long written tradition the extent to which the vocabulary of a language is subject to change, and how much of an old vocabulary is affected not only by morphological reshaping, but also by outright loss. It is therefore not unusual that our endeavors in comparative reconstruction end up with highly incomplete results. Worse, time and again we find ourselves confronted with findings that contradict each other. At this point, the question becomes important: Is it possible to transcend the limits of comparative reconstruction? Or, to put it differently: Are there ways to reconstruct aspects of earlier stages of a language or a language group other than comparative reconstruction? Areas of applicability for such reconstruction can easily be named: There are observable language isolates, that is, languages with no known genetically related counterpart. There are languages too distantly related with one another to permit a normal application of the comparative method. Finally, in applying the comparative method we will sooner or later reach an impasse where the result of our reconstruction becomes an isolate itself; it would be highly desirable to be able to extend reconstruction further, both in time depth and by the elimination of conflicting results of comparative reconstruction. Comparative reconstruction, given the proper data, is easily enough performed whenever either no change has occurred or the change has been along regular lines; deviations caused by syntagmatic conditioning can also be fairly well accommodated. Changes without regular occurrence, such as paradigmatically conditioned ones ("analogical" changes) as well as largely abrupt changes, such as dissimilation or metathesis, permit a comparative reconstruction only in terms of probability — such and such an item is unlikely to show a regular development; it is likely that the form attested was influenced by such and such a property of the language in which the change occurred. This type of argumentation is possible whenever information about

The scope of historical and comparative

linguistics

19

the regular development to be expected can be extracted from the comparison with related languages. Reconstruction within the bounds of one language is possible both by internal comparison and by interpretation of paradigmatic data. The technique mentioned first is constantly applied in the production of grammars and dictionaries. Once the regularities have been discovered in related paradigms, there is generally little reluctance about projecting attested features into gaps in surviving paradigms. Neither the complete paradigm of Latin amäre 'to love' nor that of Greek paideuein 'to educate' can be lifted from surviving Latin and Greek texts; still, there is generally no objection to the use of the filled-up paradigms as standard parts of elementary Latin and Greek grammars. The reasoning implied can probably be stated as follows: "The forms attested make Latin amäre and Greek paideuein seem perfectly regular verbs; it is therefore extremely likely that the forms missing should agree with actually attested forms from parallel regular paradigms". There are, of course, limits to this type of projection. Once attested forms cannot be unambiguously assigned to just one paradigm, all claims about missing forms can only be made with proper caution, indicating either that a reconstructed form is only one of several possible ones or, to preferably, listing alternatives explicitly. In other cases, internal properties of paradigms give rise to hypotheses about their earlier shape. The following examples will illustrate this point. In present-day German, two nominal suffixes correspond to English -hood, namely, German -heit and -keif, both form abstract nouns from adjectives. German -keit has a more limited distribution than -heit: -keit is found after adjectives ending in -bar, -ig, -lieh, -sam (plus some in -el and -er)·, -heit is used elsewhere. The identity of function and the similarity of form make it tempting to propose that -heit and -keit are just variants of the same suffix, with -keit being the conditioned form. A change from -h- to -k- is not regular; however, an assimilatory process is conceivable. There is only one environment for -keit in which assimilation could have taken place, namely in the position after the suffix -ig /-ik/. -keit would then have spread to other environments so that it is now no longer a phonologically conditioned variant, but one that requires an enumeration of conditioning morphological elements. The only fault to be found with the above hypothesis is that it has been misdated: the development posited did not occur in Modern High

20

Werner Winter

German as assumed, but in Middle High German, and even the spread of -keii is to be assigned to this period. While in the case of German -heit : -keit recent variation can be reduced to older identity, another example will show that assumptions can also be made about a state of affairs prior to present uniformity. Taking again Modern German data, we find that the stems of strong preterite forms are identical in singular and plural, but that they show two types of vocalism — either with an -a- or without it. We thus find er sang : sie sangen 'he sang : they sang', er ritt : sie ritten 'he rode : they rode'. In some cases both -a- and n o n - a - vocalism are found with the same verb: er ward : sie warden :: er wurde : sie wurden 'he became : they became' {ward : warden is more archaic and no longer used in everyday language). An inspection of the data shows that, while it seems likely that at an earlier stage in the development of German there was an alternation between -a- and non—a- stem in the strong past tense, details of the distribution can no longer be easily determined. However, some archaisms still in use provide help: While 'he sang' is always er sang, we find for 'they sang' a variant of sie sangen in a rhymed proverbial saying Wie die Alten sungen, so zwitschern die Jungen 'As the old ones used to sing, so the young ones twitter'. Protected by the rhyme, sungen could survive longer than in normal speech; its survival gives us a chance to posit -α—less forms as belonging originally to the plural, while the domain of -a- forms was the singular of the strong preterite. To prove this state of affairs, it would have been much easier to consult older Germanic languages and to work through comparative reconstruction; the present argument merely serves to show that the absence of older outside data would not have made an internal reconstruction impossible. When dealing with the earliest proto-language reconstructed for an entire language family (or, for that matter, with any other language isolate), internal reconstruction is the only means enabling us to penetrate into a more remote past (typological evidence can at best be suggestive). By positing earlier patterns, we, firstly, try to resolve difficulties arising from conflicting reconstructions (as, for instance, when Indo-European words for 'fire' showed forms ending either in -r or in -n, a "heteroclitic" rjn stem was reconstructed long before actual evidence for such an alternation in this word became available in Hittite). Secondly, we attempt to simplify our statements by subsuming overtly differing phenomena under one common formula which may or may not require positing directly unattested elements condi-

The scope of historical and comparative linguistics

21

tioning the difference. A prime example of the latter approach is de Saussure's proposal to reinterpret Proto-Indo-European full-grade long vowels as results of the coalescence of a nuclear short vowel with a following lengthening consonant subsequently lost. This analysis had the tremendous advantage that now all subtypes of ablaut in ProtoIndo-European could be treated alike and that canonical forms of roots could be set up; along with this, however, went the disadvantage that the Proto-Indo-European system of vowels apparently had to be reduced in an unreasonable way. The example mentioned last shows some of the problems which internal reconstruction of otherwise inaccessible stages of languages has to face: A hypothesis is introduced because it provides for a more reasonable pattern than that found in a language isolate either observed or reconstructed. The hypothesis can only be evaluated in terms of what it achieves — greater simplicity, greater consistency, greater naturalness of patterns than those that provided the input for positing the hypothesis. Outside confirmation can only be found where internal reconstruction is applied to nonisolates. Hence, we will not be able to get beyond the point where acceptance or rejection of an internal reconstruction depends on whether the hypothesis offered is considered plausible or not. We have to conclude that, while important tasks of linguistic reconstruction can only be tackled by resorting to internal reconstruction, its results are often handicapped by the fact that there can be no proof from outside the original chain of argument. It is not surprising that important proposals made in this context, such as that of de Saussure concerning what later came to be called the laryngeals, keep meeting with much more scepticism than is usually reserved for findings of the comparative method. This cannot just be the result of a greater novelty of the method of internal reconstruction (because it is not new at all), but must be due to the different quality of the two types of reconstruction: Claims by comparativists can be falsified by data from observable languages, while claims of scholars applying the methods of internal reconstruction to language isolates can only be subjected to the test of plausibility, and this test is a highly subjective test indeed.

II. Aspects of Language Change

Synchronic manifestations of linguistic change Franklin C.

Southworth

Linguistic change manifests itself synchronically as variation, which can be defined in a somewhat general way as a relationship between different ways of expressing the "same" thing. Traditional historical linguistics recognized that many of the changes noted between successive (historically attested) linguistic stages appear to be the results of processes which can be observed in contemporary variation, e. g., /elm/~/ebm/ (cf. Latin Periclum —> periculum, Sturtevant 1961: 63, see also Lass 1980: 11). From the time of Saussure until the late 1960s, it was assumed by the great majority of linguists that the "omnipresence of ongoing change" (Bailey 1973: 33) which is observable in all real-life linguistic data must be ignored in synchronic description (apart from the possibility of allowing "free variation"; see Bloch —Trager 1942: 42). This picture has now substantially changed, as a result of numerous careful studies of linguistic variation, most of which have been directly or indirectly inspired by the work of William Labov. (There are still, one should add, many linguists who work in the Saussurean framework.) The new framework, which Bailey (1973: 21 — 35) refers to as the "dynamic paradigm" in contrast to the older "static paradigm", assumes that variation is inherent in all natural languages, and that any attempt to eliminate it by restricting the field of observation to a limited class of speakers, or to limited contexts of speech, will lead inevitably to a failure to accomplish the linguist's goal of describing linguistic competence. The reason for this is that grammars of idiolects, or of single-style speakers, or of arbitrarily selected groups of individuals, can only be incomplete and unintelligible fragments of a larger whole which Labov calls the "grammar of the speech community"

26

Franklin C.

Southworth

(1972 a: 247). Such a grammar would presumably describe what Bailey (1973: 27) calls "panlectal competence" — a type of competence which allows for variation in one's own speech as well as between different individuals, and which enables speakers of different varieties of a language to interact with each other (Labov 1972 a: 223 — 237; Bailey 1973: 3 4 - 3 5 ; Weinreich - Labov - Herzog 1968: 187-188; Sankoff 1978). The older view of linguistic description seeks to describe some ideal form of a language. 1 Its practitioners seek "good speakers" as sources of data, and are often obliged to fall back on themselves as informants. The new view, as described above, differs mainly in that it substitutes a more complex ideal. Bailey, for example, speaks of individuals' linguistic competence "asymptotically approaching" panlectal competence over their lifetimes (1973: 27). There can be no doubt, however, that this new and more complex view accounts for many phenomena which cannot be dealt with by the older one. What does all this have to do with linguistic change? First, workers within the "dynamic paradigm" would claim that all change (or almost all change, at any rate) begins as variation, and the studies referred to below (e.g., those in Labov 1980) provide a strong justification for such a claim. Second, it is claimed that the processes and the causes of linguistic change can never be understood by looking at old documents or by reconstructing proto-languages, but only by observing ongoing change — i. e., contemporary linguistic variations, their trajectories, and their social concomitants. 2 Not unexpectedly, linguists working within this framework are developing descriptive models which can be applied both synchronically and diachronically — thus breaking another Saussurean taboo. Bailey, for example, claiming that "the function of time in defining synchronic language patterns cannot be ignored in valid descriptions of language" (1973: 32, his italics), proposes the development of "dynamic or time-based models" which will be "suitable for either historical or descriptive analysis" (1973: 31). Various applications of this proposal are represented by works cited below. The following paragraphs contain brief discussions of the principal types of linguistic variation discussed in the literature, arranged in order of (the author's) convenience, with comments on their diachronic implications. Since most of these phenomena are treated in greater detail elsewhere in this guide, no attempt is made to provide exhaustive references. 3

Synchronic

manifestations

of linguistic change

27

Dialect variation (including both regionally and socially differentiated varieties), particularly phonological variation, has been the main focus of recent sociolinguistic literature. (For recent summaries see Chambers — Trudgill 1980, Hudson 1980.) The recent development of quantitative methods for the study of phonological variation (Labov 1969, 1971, 1972 a: ch. 8, 1980; Cedergren - Sankoff 1974; Trudgill 1974 c; Anshen 1978) has been motivated by a concern for developing more precise models of linguistic change, as well as more accurate descriptive statements and a better understanding of synchronic variation in all its aspects, structural as well as social. The approach to describing variation favored in most recent work involves adaptations of the generative model (Chomsky —Halle 1968), modified to include variable rules (Labov 1969; Bickerton 1971; Cedergren — Sankoff 1974; Bailey 1973; Bailey-Shuy 1973; F a s o l d - S h u y 1975; Labov 1971; Romaine 1981).4 Social dialects (as opposed to purely regional varieties) have occupied a central role in recent work on variation, the black English vernacular being perhaps the most-studied non-standard variety (see Labov et al. 1968; Labov 1972 b; Wolfram 1969; Fasold 1972; among others). Studies of this have been concerned with questions of change (see, for example, the studies of copula deletion by Stewart 1967, 1968; Labov 1969; Baugh 1980) as well as descriptive problems. Labov (1980: 253-255), following Kroch (1978), points out that a number of studies of linguistic change in progress in Western societies have identified the upper working class or the lower middle class as the innovating groups in the process of phonological change. Studies of social dialects in South Asia indicate that upper and lower dialects innovate independently (Gumperz — Ferguson 1960, see also Labov 1972 a: 296 — 297). Ramanujan (1968) suggests that in societies characterized by caste divisions, high-caste groups may innovate as a means of avoiding lower-caste behavior. The opposite is also true: Lowercaste groups may innovate to avoid the accusation of imitating highercaste speech (Southworth 1975: 192).5 Differences between male and female speech are mentioned by various authors (Labov 1972 a: 243, 301-304; Trudgill 1974 a; Chambers - Trudgill 1980: 7 1 - 7 4 , 9 7 - 9 8 ; Lakoff 1973, 1975; C r o s b y Nyquist 1977). Chambers and Trudgill note that women tend to use more forms conforming to the standard (in Western societies) than men do (1980: 97), which often puts them in an innovating role vis-ävis men of their own class (Labov 1972 a: 303), though in traditional

28

Franklin C.

Southworth

societies they are usually found to be more conservative linguistically than men of their own group (Trudgill 1974 a). (See also Steckler — Cooper 1980.) Variation in the speech of different age groups in a population is significant as an indication of change — the presumption being that older speakers retain earlier linguistic stages, as a result of their usage (or rather, their capacity to restructure) having become "frozen" at some time in their teens (see Hockett 1950). Thus, variation in age is often referred to as "change in apparent time", as opposed to change in real time. The relationship between age and variation is complex, however; Labov notes, for example, that lower-middle-class speakers "even in middle age ... tend to adopt the latest prestige markers of the younger upper-middle-class speakers" in formal contexts (1972 a: 134).6 Payne (1980) makes it clear that the "freezing" of usage in youth is also complex, and does not happen all at once: She suggests the age of eight as the "cut-off point" for acquisition of phonetic conditioning rules of a new dialect (i.e., children moving into a new area after that age have a lower chance of completely acquiring the new patterns). Further, "children do not freely restructure and/or reorganize their grammars up to the age of 14 but ... do have the ability to add lower level rules" (1980: 175).7 Stylistic or register variation, often linked to a particular standard form of a language, is important not only for its own sake, but also for the ways in which it interacts with other kinds of social variation (Labov 1972 a: chs. 4 — 5). Labov (1972 a: ch. 3) has presented a method for obtaining reliable and consistent data on stylistic variation, in order to resolve what he calls the "observer's paradox" (1972 a: 209 — 210). More informal levels are generally found to be more innovative phonologically, lexically, and grammatically than the more formal levels, though the latter are often more innovative in the use of foreign words, using more of them and showing a better approximation to the original phonetics (like the elite dialects which they are based on — see above). By the same token, the "higher" (H) forms of diglossia are more conservative phonologically than the "lower" (L) forms (Ferguson 1959) — though from another point of view, those varieties which are consciously archaized or resurrected (e. g., Hebrew, classical Tamil, and to a lesser extent Hindi and Urdu) could be considered innovative. 8 The focus in most of the works cited above is primarily on phonological change. Examples of discussions of the relevance of variation

Synchronic manifestations

of linguistic change

29

in the study of syntactic change are to be found in Labov 1972 c; Poplack 1980; Klein 1980; Steever et al. 1976; Ashby 1981; Naro 1981; Sankoff-Cedergren 1976. The language of professional and other social subgroups — such as the military (Jonz 1975), criminals (Maurer 1940; Yenne 1927), drug addicts, prostitutes (James 1972), entertainers, college students (Morse 1927), teenagers, cowboys, sportsmen (Craven 1980), homosexuals (Farrell 1972) — is often distinguished by innovative uses of lexical items (most commonly imaginative metaphors and restrictions of the semantic range of a term), some of which find their way into the speech of other groups and occasionally into the standard dialects. (See also Belasco 1979; Chaika 1980.) Change resulting from language contact may appear as differences between the speech of bilinguals (and those who interact with them) and others lacking such contact. Such changes are usually traceable if the language contact situation is known, and even in many cases where it is not — though Polome (1981) cautions against the creation of "ghost languages" on the basis of presumed non-indigenous vocabulary in ancient languages. Weinreich's (1974) general discussion of language contact phenomena has not been superseded, though there have been many individual treatments of these phenomena, some of them with important theoretical implications. 9 Pidgin-creole continua also offer evidence of change when basilectal and acrolectal forms are compared. The original changes which created the pidgins are still poorly understood, but the process of decreolization involving the interaction of basilect and acrolect has recently been studied in considerable detail (see discussion in Hancock, this volume). Certain types of lexical relationships which are not usually considered under the heading "variation" do nevertheless conform to the definition given at the beginning of this section, and may therefore be briefly mentioned here. Under this heading can be included most metaphors, extensions of meaning, and narrowing of meanings (especially if they are "live" in the sense that both the old and the new senses coexist synchronically). Recent contributions to the study of metaphor include Bickerton 1969, Abraham 1975, Sapir — Crocker 1977, and Stross 1975. For other recent discussions of variation and semantic change, see Lehrer 1978, Kroskrity 1978, Kay 1975, Kristol 1980, Maurer 1980. The types of variation mentioned above are part of the process of change. Another type of variation, long familiar to historical linguists,

30

Franklin C. Southworth

involves synchronic irregularities which are a residue of earlier change — such as keep-kept, foot-feet, as well as cases like twitch-tweak, dwarftwerp, etc. See Drinka (this volume) for further discussion of this type of variation.

Notes 1. See Chomsky 1965: 3 for a very strong statement of this ideal. 2. See Labov 1972 a: ch. 9, 1980; Weinreich - Labov - Herzog 1968. For the significance of the study of variation to historical linguists, see Lass 1976; Anttila 1972: 4 7 - 5 6 , 191-193; Anderson 1973: 173-178. 3. It may be noted here that not all variations imply change; for example, some do not "go to completion" — see Nunberg 1980 for an example. Furthermore, many cases which fit the above definition of variation do not involve change in the sense that variant A replaced variant Β in some context, e. g., paraphrases like aunt and father's/mother's sister. Even in a more restricted sense of change, in which A —• Β (i. e., A "organically becomes" B), certain cases would not fit, for example suppletive alternants. The broad definition of variation is used here merely to provide greater scope for this discussion. 4. See King 1969: 2 9 - 3 2 , and Chambers - Trudgill 1980: 3 8 - 4 5 for discussions of the need to modify the earlier proposal of Weinreich (1954) for describing dialect variation. 5. For the effects of ethnic group membership on linguistic change, see Labov 1972 a (esp. chapters 1, 4, 5, 7, 8, and pages 296-298); Labov 1972 b; and Laferriere 1979. 6. See other entries under age in index (Labov 1972 a: 337); also Hudson 1980: 15 — 18, 148-152; Chambers-Trudgill 1980: 8 9 - 9 0 , 165-167, 179-180. 7. See also Labov's comments (1980: xviii) on Payne; and Hinton 1980. 8. See Ferguson — Gumperz 1960; Das Gupta — Gumperz 1968; and Southworth 1975 for discussion of some relevant South Asian cases. 9. See, for example, Bokamba 1977; Caskey-Sirsons — Hickerson 1977; Midgett 1970; Thananjayarayasingham 1973; Ts'ou 1975; Southworth 1982: 2 8 - 3 3 .

References Abraham, Werner 1975

A linguistic approach to metaphor. Lisse, Netherlands: Peter de Ridder Press. Anderson, James M. 1973 Structural aspects of language change. London: Longman Group. Anshen, Frank 1978 Statistics for linguists. Rowley, Mass.: Newbury House. Anttila, Raimo 1972 An introduction to historical and comparative linguistics. New York: MacMillan.

Synchronic manifestations of linguistic change

31

Ashby, William J. 1981 "The loss of the negative particle ne in French: A syntactic change in progress", Language 57: 674—687. Bailey, Charles-James N. 1973 Variation and linguistic theory. Washington, D. C.: Center for Applied Linguistics. (Reviewed by E. Traugott, Language 52 (1976): 502 — 506). Bailey, Charles-James N. — Roger W. Shuy (eds.) 1973 New ways of analyzing variation in English. Washington, D. C.; Georgetown University Press. (Reviewed by R. Lass, Language in Society 5 (1976): 219-228.) Baugh, John 1980 "A reexamination of the black English copula", in: William Labov (ed.) 83-106. Belasco, Simon 1979 "Derived and modified nouns in French slang", Lingua 48: 177 — 192. Berman, Ruth A. 1980 "Child language as evidence for grammatical description: Preschoolers' construal of transitivity in the verb system of Hebrew", Linguistics 18: 677-701. Bickerton, Derek 1969 "Prolegomena to a linguistic theory of metaphor", Foundations of Language 5: 34 — 51. 1971 "Inherent variability and variable rules", Foundations of Language 7: 457 — 492. 1975 Dynamics of a creole system. Cambridge: Cambridge University Press. Bloch, Bernard — George L. Trager 1942 Outline of linguistic analysis. Baltimore: Linguistic Society of America. Bloomfield, Leonard 1933 Language. New York: Henry Holt & Co. Bokamba, Eyamba 1977 "The impact of multilingualism on language structures: The case of Central Africa", Anthropological Linguistics 19: 181—202. Bynon, Theodora 1977 Historical linguistics. Cambridge: Cambridge University Press. Caskey-Sirsons, Leigh Α. —Nancy P. Hickerson 1977 "Semantic shift and bilingualism: Variation in the color terms of five languages", Anthropological Linguistics 19: 358 — 367. Cassano, Paul V. 1973 "The substrate theory in relation to the bilingualism of Paraguay: Problems and findings", Anthropological Linguistics 15: 406 — 426. Cedergren, Henrietta J. —David Sankoff 1974 "Variable rules: Performance as a statistical reflection of competence", Language 50: 333 — 355. Chaika, Elaine 1980 "Jargons and language change", Anthropological Linguistics 22: 77 — 96. Chambers, J. K. —Peter Trudgill 1980 Dialectology. Cambridge: Cambridge University Press.

32

Franklin C. Southworth

Chomsky, Noam A. 1965 Aspects of the theory of syntax. Cambridge, Mass.: M . I . T. Press. Chomsky, Noam A. — Morris Halle 1968 The sound pattern of English. New York: Harper & Row. Conrad, James — William W. More 1976 "Lexical codes and sub-cultures: Some questions", Anthropological Linguistics 18: 2 2 - 2 6 . Craven, Robert R. 1980 "Billiards, pool, and snooker terms in everyday use", American Speech 55: 9 3 - 1 0 0 . Crosby, Faye —Linda Nyquist 1977 "The female register: an empirical study of LakofPs hypotheses", Language in Society 6: 313 — 322. Das Gupta, Jyotirindra—John J. Gumperz 1968 "Language, communication control in North India", in: Fishman —Ferguson—Das Gupta (eds.), 151 — 166. Dietrich, W . - H . Geckeler (eds.) 1981 Semantics. Vol. 3 of Logos Semantikos. Berlin: de Gruyter/Madrid: Gredos. Dingwall, William O. (ed.) 1971 A survey of linguistic science. College Park: University of Maryland. Farrell, Ronald A. 1972 "The argot of the homosexual subculture", Anthropological Linguistics 14: 9 7 - 1 1 0 . Fasold, Ralph W. 1972 Tense marking in black English. Arlington, Virginia: Center for Applied Linguistics. Fasold, Ralph W . - R o g e r W. Shuy (eds.) 1975 Analyzing variation in language. Washington, D. C.: Georgetown University Press. Feagin, Crawford 1979 Variation and change in Alabama English. Washington, D. C.: Georgetown University Press. (Reviewed by Butters in Language 57 (1981): 735 — 738.) Ferguson, Charles A. 1959 "Diglossia", Word 15: 325-340. Ferguson, Charles A.— John J. Gumperz 1959 "Linguistic diversity in South Asia", International Journal of American Linguistics 26, No. 3 (1960), part III. Fishman, Joshua A.— Charles A. Ferguson—Jyotirindra Das Gupta (eds.) 1968 Language problems of developing nations. New York: John Wiley. Gumperz, John J. —Dell Hymes 1972 Directions in sociolinguistics: The ethnography of communication. New York: Holt, Rinehart and Winston. Haber, Lyn R. 1976 "Leaped and leapt: A theoretical account of linguistic variation", Foundations of Language 14: 211—238. Hancock, Ian F. (ed.) 1979 Readings in creole studies. Ghent: Story-Scientia. (Reviewed by P. Mohan in Language 57 (1981): 904-911.)

Synchronic manifestations of linguistic change

33

Hinton, Leanne 1980 "When sounds go wild", Language 56: 320-344. Hockett, Charles F. 1950 "Age grading and linguistic continuity", Language 26: 449 — 457. Hooper, Joan Bybee 1979 "Child morphology and morphophonemic change", Linguistics 17: 21 — 50. Hudson, R. A. 1980 Sociolinguistics. Cambridge: Cambridge University Press. Hymes, Dell 1971 Pidginization and creolization of languages. Cambridge: Cambridge University Press. James, Jennifer 1972 "Two domains of streetwalker argot", Anthropological Linguistics 14: 172-181. Jonz, Jon G. 1975 "Situated address in the United States Marine Corps", Anthropological Linguistics 17: 68 — 77. Kay, Paul 1975 "Synchronic variability and change in basic color terms", Language in Society 4: 257-270. King, Robert D. 1969 Historical linguistics and generative grammar. Englewood Cliffs, New Jersey: Prentice-Hall. Klein, Flora 1980 "A quantitative study of syntactic and pragmatic indications of change in the Spanish of bilinguals in the U.S.", in: William Labov (ed.), 6 9 - 8 2 . Kristol, Andres M. 1980 "Color systems in southern Italy", Language 56: 137—145. Kroch, Anthony S. 1978 "Toward a theory of social dialect variation", Language in Society 7: 17 — 36. Kroskrity, Paul V. 1978 "Aspects of syntactic and semantic variation within the Arizona Tewa speech community", Anthropological Linguistics 20: 340 — 350. Labov, William 1966 The social stratification of English in New York City. Washington: Center for Applied Linguistics. 1969 "Contraction, deletion, and inherent variability of the English copula", Language 45: 715 — 762. 1971 "Methodology", in: William O. Dingwall (ed.), 4 1 2 - 4 9 7 . 1972 a Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press. 1972 b Language in the inner city. Philadelphia: University of Pennsylvania Press. 1972 c "Negative attraction and negative concord in English grammar", Language 48: 773-818. Labov, William (ed.) 1980 Locating language in time and space. Vol. I. New York: Academic Press.

34

Franklin C. Southworth

Labov, William —Paul Cohen — Clarence Robins —John Lewis 1968 Α study of the non-standard English of Negro and Puerto Rican speakers in New York City. Philadelphia: U. S. Regional Survey. Laferriere, Martha 1979 "Ethnicity in phonological variation and change", Language 55: 603 — 617. Lakoff, Robin 1973 "Language and woman's place", Language in Society 2: 45 — 79. 1975 Language and woman's place. New York: Harper Colophon Books. Lass, Roger 1980 On explaining language change. Cambridge: Cambridge University Press. (Reviewed by E. Itkonen, Language 57 (1981): 688-697.) 1976 Review of Bailey-Shuy (eds.) 1973. Language in Society 5: 219-228. Lehmann, Winfred P. — Yakov Malkiel (eds.) 1968 Directions for historical linguistics. Austin: University of Texas Press. Lehrer, Adrienne 1978 "Structures of the lexicon and transfer of meaning", Lingua 45: 95 — 123. Maurer, David W. 1940 The big con. Chapter 9 "The con man and his lingo". Indianapolis: BobbsMerrill. Maurer, David W., assisted by Ellessa Clay High 1980 "New words — where do they come from and where do they go?", American Speech 55: 264 — 277. Midgett, Douglas 1970 "Bilingualism and linguistic change in St. Lucia", Anthropological Linguistics 12: 158 — 170. Morse, William R. 1927 "Stanford expressions", American Speech 2: 275 — 279. Mougeon, Raymond 1976 "Bilingualism and language maintenance in the Gaspe Peninsula, Quebec, Canada", Anthropological Linguistics 18: 53 — 69. Naro, Anthony J. 1981 "The social and structural dimensions of a syntactic change", Language 57: 6 3 - 9 8 . Nunberg, Geoffrey 1980 "A falsely reported merger in eighteenth-century English; A study in diachronic variation", in: William Labov (ed.), 221—250. Payne, Arvilla C. 1980 "Factors controlling the acquisition of the Philadelphia dialect by out-ofstate children", in: William Labov (ed.), 143-178. Polome, Edgar 1981 "Lexical data and cultural contacts: A critique of the study of prehistoric isoglosses and borrowings", in: W. Dietrich —H. Geckeler (eds.), 505 — 513. Poplack, Shana 1980 "The notion of the plural in Puerto Rican Spanish: Competing constraints on (s) deletion", in: William Labov (ed.), 55 — 68.

Synchronic manifestations of linguistic change

35

Ramanujan, A. K. 1968 "The structure of variation: A study in caste dialects", in: Singer—Cohen (eds.), 4 7 - 6 3 . Romaine, Suzanne 1980 "The relative clause marker in Scots English: Diffusion, complexity, and style as dimensions of syntactic change", Language in Society 9: 221 — 247. 1981 "The status of variable rules in sociolinguistic theory", Journal of Linguistics 17: 9 3 - 1 1 9 . Sankoff, David 1978 Linguistic variation: Methods and models. New York: Academic Press. Sankoff, David — Henrietta J. Cedergren 1976 "The dimensionality of grammatical variation", Language 52: 163 — 178. Sapir, J. David — Christopher Crocker (eds.) 1977 The social use of metaphor: Essays on the anthropology of rhetoric. Philadelphia: University of Pennsylvania Press. Singer, Milton — Bernard S. Cohn (eds.) 1968 Structure and change in Indian society. New York: Wenner Gren Foundation. Southworth, Franklin C. 1975 "Sociolinguistic research on South India: Achievements and prospects", in: B. Stein (ed.), 181-205. 1982 "Substratum influence on early Indo-Aryan", Paper presented at the Third ICSAL, Mysore, India, January 1982. Steckler, Nicole Α.—William E. Cooper 1980 "Sex differences in color naming of unisex apparel", Anthropological Linguistics 22: 373 — 381. Steever, Sanford E. —Carol A. Walker—Salikoko S. Mufwene (eds.) 1976 Papers from the parasession on diachronic syntax. Chicago Linguistic Society 12th Regional Meetings 22 April 1976. Chicago: Chicago Linguistic Society. Stein, Burton (ed.) 1975 Essays on South India. Honolulu, Hawaii: The University Press of Hawaii. Stern, Gustaf 1931 Meaning and change of meaning, with special reference to the English language. Bloomington: Indiana University Press. Stewart, William A. 1967 "Sociolinguistic factors in the history of American Negro dialects", Florida Foreign Language Reporter 5/2: 11, 22, 24, 26. 1968 "Continuity and change in American Negro dialects", Florida Foreign Language Reporter 6/1: 3ff. Stross, Brian 1975 "Metaphor in the speech of Tzeltal children", Anthropological Linguistics 17: 305-324. Sturtevant, Edgar H. 1961 Linguistic change: An introduction to the historical study of language. Chicago: Phoenix Books. (Originally published Chicago: University of Chicago Press, 1917.)

36

Franklin C. Southworth

Thananjayarayasingham, S. 1973 "Bilingualism and acculturation in the Kuravar community of Ceylon", Anthropological Linguistics 15: 276 — 280. Trudgill, Peter 1974 a Sociolinguistics: An introduction. Harmondsworth: Penguin Books Ltd. 1974 b The social differentiation of English in Norwich. Cambridge: Cambridge University Press. 1974 c "Linguistic change and diffusion: Description and exploration in sociolinguistic dialect geography", Language in Society 3: 215 — 246. T'sou, Benjamin K. 1975 "On the linguistic covariants of cultural assimilation", Anthropological Linguistics 17, No. 9: 445—465. Weinreich, Uriel 1954 "Is a structural dialectology possible?" Word 10: 388-400. 1974 Languages in contact: Findings and problems. The Hague-Paris: Mouton. Weinreich, Uriel — William Labov —Marvin I. Herzog 1968 "Empirical foundations for a theory of language change", in: Lehmann — Malkiel (eds.), 9 5 - 1 8 8 . Wolfram, Walt A. 1969 A sociolinguistic description of Detroit Negro speech. Washington: Center for Applied Linguistics. Yenne, Herbert 1927 "Prison lingo", American Speech 2: 280-282.

Evidence of language change Irmengard

Rauch

I. Linguistic evidence: inviolable Data are untouchable. No amount of manipulation can alter evidence; it is what it is. However, its use in linguistic inquiry has varying purposes. So, for example, Toon (1976: 71), while observing that "the study of the texts containing the earliest recorded English has a long and distinguished history", appears to lament that "interest in abstract grammars has shifted the attention of English historical linguists farther and farther from the manuscript data, the most interesting of which was relegated to the waste pile of free variation and optional rules". In point of fact, abstract grammarians do not set aside manuscript data, but by appealing to the principle of Ockham's Razor they tend to shift the focus from the data to the hypothesis relating to the evidence. Thus, for example, Botha (1981: 284) writes: Data that can be explained on the basis of a hypothesis therefore do not prove or demonstrate the truth of the hypothesis. However, the data explained by a hypothesis may be regarded as evidence for the hypothesis. Data constitute evidence for a hypothesis if these data — as presented in the minor premiss(es) of one or more arguments — indicate inconclusively that the hypothesis could possibly be true.

It is to be noted that the shift in focus does not contradict Cohen and Nagel's (1934: 394) statement that "no single proposition dealing with matters of fact is beyond every significant doubt. No proposition is so well supported by evidence that other evidence may not increase or decrease its probability", since the key word is "other" evidence not "additional" evidence. Published research on language change accordingly exhibits both what we may label as philological (here meaning overtly textbound)

38

Irmengard Rauch

and abstract approaches. They are obviously not mutually exclusive; indeed, they are mutually dependent, and all researchers employ both approaches but in varying proportions. Botha (1981: 302) nicely defines internal linguistic evidence as consisting of "linguistic data about objects within the grammarian's linguistic reality. These linguistic data concern the utterances of a language and the linguistic competence underlying these utterances." The sources of internal linguistic evidence are then as follows: a) the native speaker's intuitions ("intuitive evidence"), b) the linguist's grammatical hypotheses ("hypothetical evidence"), and c) the grammarian's intuitions ("theoretical intuitive evidence") (1981: 302 — 303). The philological methodologists assume as a priori to their work these three types of evidence, while the abstract methodologists certainly recognize, for example in their "theoretical intuitive evidence", the link to the philological methodologist's category of grammarian's evidence or the filiation of reconstruction to evidence from hypotheses. The linguistic hypothesis that "languages change", which underlies both methodologies, is deceptively simple and would lead one to believe that evidence in language change is by and large empirical and belongs to empirical or scientific theories. Burks (1963: 37) distinguishes between evidence which "usually relates to particular events not observed under controlled conditions and not repeatable" and "scientific evidence [which] generally relates to laws and repeatable events". Indeed, linguists have long been intent on discovering the "laws" of language change; the Neogrammarian hypothesis was a landmark breakthrough in this direction. However, what does the linguist do with Toon's (see above) "waste pile of free variation and optimal rules", in particular where semantics and pragmatics play dominant roles? Stephens (1968: 80) tells us that Statements that are framed, and offered, in such a form so that they can be related to evidence, we might term "empirical discourse" or "scientific discourse". Statements ... which cannot be related to evidence if one does not first grossly change or even distort them, we might term "humanistic discourse".

The linguist may, nevertheless, prefer to cling to his scientific bonds, citing perhaps Cohen and Nagel (1934: 394): Science is thus always ready to abandon a theory when the facts so demand. But the facts must really demand it. It is not unusual for a theory to be modified so that it may be retained in substance even though "facts" contradicted an earlier formulation of it. Scientific procedure is therefore a mixture of a willingness to change, and an obstinacy in holding on to theories apparently incompatible with facts.

Evidence of language change

39

What is eminently clear from any and all arguments is that despite one's predilection for method, whether philological or abstract, whether scientific or humanistic, the evidence always withstands individual conviction. Language change is more readily traceable from evidence in past changes than from evidence of contemporary change. It follows, then, that more research is available on the analysis of historical change even though, as Samuels (1972: 4) notes: "for diachronic purposes our choice of evidence is far more limited: we must for the most part reconstruct from written records, bearing in mind continually the differences that exist today between spoken and written media." Moreover, evidence from historical times may contain errors which can hardly be considered with certainty as being falsifications. Swadesh (1971: 117) advises that No one can deny that it is best to have accurate information, but since the best of manmade reports may contain some errors, and since the data we have for languages are not always the best, the most that we can require is that scholars use the most reliable evidence available, interpreting it as they proceed. The only legitimate objection that can be made to a body of evidence is that the amount and type of incorrect information is so great or so serious as to invalidate it. Properly, we should distinguish among degrees of reliability: (a) occasional errors, (b) frequent errors, (c) mostly wrong information, (d) invented data. Situations (a) and (b) do not invalidate a case, especially if the scholar makes allowance for the possibility of errors. Even situation (c), if handled with sufficient care, may permit effective analysis; if the correct information is carefully separated from the incorrect information, it cannot be seriously contaminated. In fact, some of our most important linguistic information from remote times and places comes to us mixed with errors, and scholars recognize that it can be used.

Finally, our respect for the inviolability of linguistic evidence turns to the fact that such evidence has a life outside of linguistic method. Kahane, Kahane, and Ash (1979: 67 — 68) observe that Linguistic evidence is, in principle, threefold: (a) When other forms of evidence are scarce, particularly in prehistoric times, language (plus or minus archeological support) is often decisive for hypotheses of reconstruction ... (b) Frequently historical events and linguistic data supplement and confirm each other ... While the documentary sources reveal the circumstances of their transmission, an etymological, semantic, distributional analysis of the words provides many details which round out the chronicler's reports ... (c) Often linguistic data function as an analogue for historical events, stimulating a hypothesis even though absolute proof is unavailable ... Recorded linguistic behavior becomes, in short, a model for unrecorded social behavior.

Linguistic evidence of language change thus ultimately evidences history, society, and above all, man, in spite of himself.

40

Irmengard Rauch

2.1. Datable documents "Indo-European is the family best known to us at present, because texts in the various languages have been most thoroughly studied, and because these texts extend through a period of almost four thousand years ..." (Lehmann 1978: 405). How well the Indo-European languages are documented can be understood by reading the Introduction "Der indogermanische Sprachstamm im Allgemeinen und seine Verzweigung" in Volume 1, Part 1 of Brugmann's Grundriss der vergleichenden Grammatik der indogermanischen Sprachen. Merritt Ruhlen has compiled information on documentation of the languages of world in his resourceful Guide to the languages of the world, which also provides information on genetic and typological structures of many languages. Valdis Zeps, building on Chapter IV of Bloomfield's Language (1933), "The languages of the world", has provided us with an updated Languages of the world: A selective survey (1966). From it we learn that the oldest documentation for Insular and Continental West Germanic dates from the eighth and ninth centuries, North Germanic inscriptions from the fourth century, and the East Germanic Bible translation of Wulfila from sixth century manuscripts (1966: 1—2). Celtic is documented in manuscripts from the eighth century, but the Ogam inscriptions date from much earlier (1966: 3). The Baltic family has late documentation, i.e., sixteenth century, while Slavic is recorded from the ninth century Old Church Slavonic (1966: 3—4). The Italic group is documented in the ancient Oscan-Umbrian inscriptions and Latin documentation predates 300 B.C. (Devoto 1968: 71). Albanian has late documentation, i. e., fifteenth century (Baldi 1983: 87). Hellenic documentation is found in inscriptions from the seventh century B. C., the Homeric poems 800 B. C., and Mycenean Linear Β of 1450 B. C. (Zeps 1966: 4 — 5). Armenian documents date from the fifth century A. D., Iranian Old Persian rock inscriptions from the sixth to the fourth centuries, and Avestan sacred texts from 600 B.C. (1966: 5). In the Indie family the Rig-Veda dates from 1200-1000 B.C. Tocharian is documented in the seventh century A. D., while Anatolian yields cuneiform tablets from the eighteenth century B. C. (Neu 1974: 2). Gray (1939: 297) identifies as the oldest documented languages Sumerian 4000 B.C., Egyptian 3500 B.C., Semitic 2800 B.C., and Chinese 1500 B.C. On the other hand he notes that documentation for Australian, Malayo-Polynesian, African, and American Indian is found only in the past few centuries (1939: 298).

Evidence of language change

41

A study such as Cowgill's "A search for universals in Indo-European diachronic morphology" (1963) is based on datable documents: Greenberg's (1960) evidence from Classical Sanskrit, pre-fourteenth century A. D. Hitopadesa; Alfred's Old English Boethius, c. 900 A. D.; Modern English New Yorker, 1952; Modern Persian, 1889. To these four Cowgill adds evidence from the Rig-Veda, 1200-1000 B.C.; Early Middle Indie Asoka's "Rock Edicts", third century B. C.; Bengali of 1862, Kaliprasanna Sinha's Hutom penchär nakshä; Old Persian of Darius I's Bisotun inscription, the last two decades of the sixth century B. C.; cuneiform Hittite of the thirteenth century B. C. in The apology of Hattusilis·, Homeric Greek in the eighth century B.C. Iliad; New Testament Greek of the first century A. D. Luke; Modern Greek of the 1888 Khamena logia; the Gothic Bible of the fourth century Visigothic Bishop Wulfila, transmitted in the sixth-century Ostrogothic Codex Argenteus; Old Church Slavonic of the ninth century A. D. Penzl (1970) investigates the phonology of Old High German via the datable Old High German Isidor of 800, Otfrid of 860, Tatian of 830, Notker of 1000, Exhortatio ad plebem christianam of 850, and the Otloh's Gebet of 1070. Dietrich (1973) traces periphrastic verb formations in datable documents ranging from Ancient Greek, through Vulgar Latin, and up to Modern Romance (list of documents used: 1973: 329 — 336). Datable documents for linguistic change within a particular language are frequently inventoried as part of the grammar of a given language; see, for example, Gray's (1971: 4 — 6) discussion of Semitic documentation. The Bible serves as one of the most common documents of comparison over time and known geographical distribution. There are numerous bilingual presentations such as that of Streitberg's (1908) Greek and Gothic Bible. From the Greek Old and New Testament also came the Armenian version (fifth century). In the ninth century we have the Slavic translation of the Bible by Cyril and Methodius (Gray 1939: 427). Tschirch (1969) offers eight versions side by side: New Testament Greek, the Vulgate, the Old High German Tatian 830, the Mentel Bible of 1466, the Evangelienbuch of 1343, Luther's Bible of 1522, the Zinzendorf Bible of 1739, and the Menge Bible of 1926. Another source of evidence across time and space is provided by etymologies and glossaries: Isidore of Seville's (636) Latin Origines sive Etymologiae (cf. 4.2. below), which parallels the Greek Etymologicum Magnum and the Etymologicum Gudianum; the glossaries of Hesychios (fifth century), which includes not only Greek and Latin,

42

Irmengard Rauch

but also Egyptian, Akkadian, Galatian, Indian, Lydian, Persian, Phrygian, Phoenician, Scythian, Parthian; the Glossae Latinograecae et Graecolatine (second — sixth centuries); the Gaulish Endlicher Glossary (fifth century); the Anglo-Saxon Epinal, Erfurt, Corpus and Leyden glossaries (700); the Althochdeutschen Glossen (850); the Romance Reichenau and Cassel glosses (eighth — ninth centuries) (Gray 1939: 425-426). 2.2. Documents as sources of evidence Phonological evidence is gleaned from orthography with a search for misspellings, overcorrections, inverse spellings, occasional spellings, archaic and obsolete spellings, and very importantly rhyme and metrical evidence. Repeatedly in the literature the English Great Vowel Shift is used to exemplify evidence for phonological change. So, for example, there is ample rhyme evidence for the late coalescence of ME e and έ into NE [Τ], Chaucer (1400) rhymed mean with clean but not with keen, queen, or green (Bloomfield 1933: 295). Dryden (1700) rhymed dream and sea with shame and obey, respectively. Pope (1744) rhymed weak and eat with take and state, respectively. And Swift (1745) rhymed seat and meat with weight and say't respectively (Lehmann 1973 b: 162). Scott (1967: 7 7 - 7 8 ) is interested in establishing the consistent spelling distinction between < e and < ξ. The source of his evidence is the Oxford English Dictionary, which yields beetle 1589, deem 1581, deep 1560, flee 1600, free 1580, geese 1577, green 1578, seek 1590 beside cheap 1567, clean 1568, deal 1534, lead 1569, leap 1580, sea 1555, speak 1535. Although phonetically distinct at this period, the sounds remain orthographically the same into New English. Accordingly, only "rhymes and occasional spellings" can attest to their later coalescence, "since the value of using regularly maintained orthographic contrast has now been exhausted". Penzl (1957: 204) cites rhymes from the Old High German Otfrid quad : sprah 'speak' (third person singular past), and ward 'become' (third person singular past): tharf'need' (third person singular present), as evidence for the spirantal pronunciation of from WGmc. f). Suprasegmental evidence is also appealed to by Lehmann (1955: 107) in determining continuant articulation (cf. 5.2. below) for the IndoHittite laryngeals. He observes that in Homeric meter the laryngeals make position just as σ and F before ρ and λ. A line of alliterative verse can reveal information about both vowels and consonants. Leh-

Evidence of language change

43

mann (1973 b: 67) cites from the Old English Beowulf: Da com of more under misthleopumj Grendel gongan, Godes yrre baer 'Then came from the moor under cover of darkness/Grendel walking, God's anger he bore'. One can extrapolate the length of the ο of more since it is the word which alliterates, while the ο of of is not in an alliterating word, metrically non-prominent and therefore short. Further one can assume that the alliterating g's of the second line must have been similarly pronounced. Occasional spellings such as Lat. Caesar written in Greek texts as kaisar point to the retained stop pronunciation of Lat. k (Bloomfield 1933: 296; cf. 3.2. below). Early Modern English and American English occasional spellings like nex 'next', husbon 'husband', myssomer 'midsummer', wrytyn 'writing', orphants 'orphans', meten 'meeting', and of 'have' show "loss (or addition) of phonemes in fast, unstressed forms" (Penzl 1957: 202). Inverse spellings such as Spenser's whight for white tell us that gh after vowel is no longer pronounced (Lehmann 1973 b: 67; cf. 3.2. below). The overcorrection of Lat. thesaurus 'treasure' in Late Latin documents to thensaurus with unhistorical intrusive η lets us infer that the Late Latin scribe was aware that the Latin cluster -ns- of, for example, mensa 'table' or sponsa 'bride' was reduced to s in Pro to Romance (Hall 1964: 292). Archaic spellings point to language change. Jeffers — Lehiste observe the high functional load of i in Modern Greek. They write: Because we are fortunate to have a long history of written records for Greek, we know that Modern Greek i merged unconditionally ... if Modern Greek did not use archaic spellings, the extraordinary statistical preponderance of i within Greek would be the only hint that the original state of affairs might have been different (1979: 47).

Essentially, documentary evidence for the morphological, syntactic, and semantic levels of language is pragmatically based as well. The appearance of morphemes, phrases, and sentences in context and discourse, and their alternation in form or syntagm serve as evidence (cf. 3.1. and 5.1. below). Polome (1966: 60 — 61) uses both age of document and ability to rhyme as support for the suggestion that Skt. umä 'flax', which occurs in the Sätapatha Brähmana and rhymes with younger ksumä 'flax', may be derived from Chinese rather than being of Indo-European origin. Li —Thompson (1974) speak of the emergence of compounds in early Archaic Chinese (tenth — eleventh centuries B.C.), where they were rare, in late Archaic Chinese (third — fourth centuries), where they increased, and during the Han dynasty,

44

Irmengard Rauch

where they became yet more frequent. In Modern Chinese, compounding is a productive process leading toward agglutination and the tendency of SVO to SOV in Mandarin Chinese. Traugott (1972: 93) observes the alternation between inflected and non-inflected past participle in Old English periphrasis with 'be', thus, we wceron cumene 'we were come' and we wceron cumen 'we had come' as indicative of the change from aspect to tense. In interpreting the documentary material of whatever linguistic level, the time-honored principle called variously that of biuniqueness, isomorphism, or iconicity, namely, that of one form: one meaning, is a fundamental working principle (Anttila 1972: 17), as can be observed in the above discussions of evidence. 3.1. Loans and grammatical reflexes of contact The interpretation of loan material is a manifold operation. Penzl (1972) distinguishes between loans transmitted orally, for example quotidian terms, and those transmitted in documents. A transliteration or transference of a word from language X into language Y can be indicative of the pronunciation of language Y. This assumes thorough knowledge of the graphemics and phonology of the donor language (1972: 6.5 a). If the document can be dated and identified for dialect, accuracy of the analysis is heightened and the information can also be used to shed light on the phonology of the source language. Contact data are occasionally attributed to causation of linguistic change (1972: § 9.4), whether through adstratum, superstratum, or substratum. Penzl further distinguishes between imitation and substitution of the loan material. With reference to phonology he writes: Lautnachahmung ist die Wiedergabe des Lautes der Ursprungssprache durch den nächststehenden Laut in der entlehnenden Sprache. 'Nächststehend' kann auch hier nicht rein phonetisch bestimmbar sein, sondern wird schon phonologisch-strukturelle Faktoren einschließen müssen. Lautersatz kann darüber hinaus oft unter Verzicht auf phonetisch möglichst nahe Wiedergabe verschiedene Grade der 'Nostrifizierung' mit phonotaktischen Umstellungen, Vereinfachung von Lautgruppen, Morphemersatz durch 'Volksetymologie', d. h. Angleichung an vorhandene Morpheme der Entlehnsprache enthalten (1972: § 6.5 a).

Penzl's imitation and substitution parallel somewhat Haugen's substitution and importation, respectively (Haugen 1969: 60 — 61). Haugen correlates these two types of borrowing to a "scale of adoptability" whereby, following Whitney's insight that "whatever is more formal or structural in character remains in that degree free from the intrusion

Evidence of language change

45

of foreign material" (1969: 72), certain elements of language are more likely to be borrowed than others. Borrowing can lead to language change in any component of the grammar, but it is generally agreed that phonological change is least affected by contact evidence, while lexical change is most affected. The interpretation of contact evidence in the phonic, grammatical, and lexical components is systematized by Weinreich (1964) in identifying the type of interference exhibited by the loan data, and also explaining the structural and non-structural stimuli and resistance factors involved. On the phonic level, (a) underdifferentiation occurs when two sounds of the source language are not distinguished and therefore confused in the target language. The opposite occurs in (b) overdifferentiation. Reinterpretation of distinctions (c) occurs when a relevant feature in the donor language is not relevant in the borrowing language, but relevance is signaled by another, possibly concomitant, feature in structural conformity with the borrowing system. Phone substitution (d) refers to the borrowing of a structurally equivalent phoneme adapted to the borrower's pronunciation (1964: 18 — 19). On the grammatical level Weinreich considers (a) the transfer of both bound and free morphemes. Interference from the recipient's language can reinforce a morpheme by adding the functionally corresponding native morpheme to the borrowed morpheme (1964: 33). As an example of (b) transfer of grammatical relations, Weinreich cites interference by word order, sentence intonation, and congruence rules from the recipient to the donor language (1964: 37 — 39). In a third case (c), change takes place in the function of a borrowed morpheme or grammatical category due to the borrower's misidentification (1964: 39 — 40). By (d), abandonment of obligatory categories, Weinreich refers to loss of grammatical markers of the source language by the target language (1964: 4 2 - 4 3 ) . Lexical evidence is categorized by Weinreich as 1) outright transference of words, 2) semantic extensions, and 3) phonic adjustment without effect on the content (1964: 47 — 50). Weinreich sums up the effects of lexical evidence on the recipient language thus: Except for loanwords with entirely new content, the transfer or reproduction of a foreign word must affect the existing vocabulary in one of three ways: 1) confusion between the content of the old and the new word; 2) disappearance of the old word; 3) survival of both the new and the old word, with a specialization in content (1964: 54).

46

Irmengard Rauch

3.2. Loan and contact evidence Graphic interference is rarely considered with regard for its own system, i. e., as orthographic change or development. Traditionally the orthographic emergence of writing systems in Europe is traced from Western Semitic (Phoenician), to Greek, into such European languages as Germanic, Slavic, and Romance. A study of writing (Gelb 1952) is a thorough presentation of the evolution of writing and of comparative writing systems. Johannes Friedrich's Extinct languages (1957) provides the history and methods of the deciphering of ancient scripts such as the cuneiform Hittite of Bogazköy, one of the spectacular finds of the twentieth century. See also George Trager's convenient overview in his "Writing and writing systems" (1974). Similarly, the adoption of, for example, German and into Finnish, or that of the Runic thorn into Old English, are hardly interesting for language change unless they have phonological consequences, such as serving as evidence of their pronunciation in German and Runic, respectively. Lehmann (1973 b: 226) represents another case in point in the replacement of Runic/Old English thorn by English upsilon for the definite article the in fifteenth-century printing: thus ye was used in, for example, ye olde gifte shoppe entailing a change in pronunciation. Similarly, the introduction of unhistorical early into Eng. autor borrowed from Fr. autor yields the spellingpronunciation /00dr/ (Hall 1964: 272). The vast amount of spelling evidence of borrowed words, then, properly belongs to phonological evidence. Phonological evidence consists principally of two types: that which introduces a new feature into the borrowing language and that which serves as corroboration and/or confirmation of the persistence of a feature already in the borrowing language. Weinreich (1964) gives as new phonological sequences produced through borrowing, word-initial v- and z- in English, final -ng in French, and word-initial dl- in Yiddish. Entirely new phonemes through contact "may be created: /g/ as distinct from /k/ in Czech, /λ, ji/ as distinct from /l, n/ in Yiddish, /o/ as distinct from /uo/ in Lettish, /f/ as distinct from /xv/ in dialectal Russian, jtj as distinct from jdj in Mazateco ..." (1964: 27). The reintroduction of the initial ^-cluster into English came via Scandinavian borrowings such as sky, skin, skirt after English had shifted [sk] to [s] as in OE scöh which is NE [su] 'shoe' (Jeffers - Lehiste 1979: 149). Other phonological loan material can be applied to yield evidence either of the source or the target language. Most commonly borrowed

Evidence of language change

47

matter accommodates the structural tendencies of the target language. Thus Chinese words such as chow mein or kowtow lose their tones in English. This tells us only that English is a non-tone language and nothing about sound change per se. Quite otherwise is the case of Swedish spoken in Finland, where loss of Swedish tone is due to the Finnish adstratum. On the other hand, in N E Don Quixote, [h] resembles Spanish pronunciation, while its derivative adjective quixotic [ks] is integrated into English structure. The non-integrated Quixote is of interest on the sociolinguistic level, displaying retention of a foreignism as a possible learned or prestige pronunciation. The [h] changes nothing in the structure of English phonology, but the points to the fact that when first borrowed Quixote was probably pronounced according to English habit [ks] (Arlotto 1972: 186). Penzl (1957: 207) points to the Old High German rendering of Old Slovenian s, ζ by O H G and of Old Slovenian s, ζ by OHG as indicative of a sibilant pronunciation for OHG ζ but a shibilant pronunciation for O H G s. Interestingly, a change in spelling took place from Middle English deleite, a loan from Old French, to later delight. The pronunciation remained [ai] and the newer spelling serves as evidence for the loss of the velar spirant in such words as light and eight (Bloomfield 1933: 294). Phonological change in the source language is revealed by the German borrowing of Lat. cellarium as Ger. Keller 'cellar' but Lat. cella as Ger. Zelle 'cell', showing the Latin change of stop to affricate about the fifth to the seventh centuries. Consequences of morphological borrowing are also twofold. Lehmann writes (1973 b: 218) "borrowings generally take on the patterns of native elements". He cites the borrowing of the Old Norse reflexive bäda sik 'bathe oneself into Old English as bask, wherein the reflexive pronoun is not perceived by the English speaker. Morphological blends occur in the case of the borrowing of Chinese kenkyuu 'study' or English taipu 'type' into Japanese, which supplies its native forms suru 'do' or shita 'did' for conjugation. Arlotto (1972: 187) describes the borrowing of Arabic qadi 'mayor' into Spanish as el alcade, in which al-, the Arabic definite article, is not perceived by the Spanish and accordingly Spanish supplies its native morpheme. The opposite occurs in the case of English apron, borrowed from Old French naperon, by reinterpreted juncture after the initial nasal, giving the semblance of the English indefinite article preceding the noun. Similarly, Whiteley (1967: 138) describes the integration of English loans with initial ma- into Swahili, for example,

48

Irmengard Rauch

maching'oda 'marching order', as Swahili plurals due to the ma- plural marker; a singular ching'oda is then derived. Weinreich's (1964: 32) discussion of the transfer of the Bulgarian first and second persons singular verb desinences into Meglenite Rumanian is an instance of outright transfer, i. e., transfer without modification to the native system. Thus, Bulgarian -um (-äm) and -is replace older -u, -i; for example, aflum, aflis Ί , you find'. He explains the success of this change in the Rumanian dialect as dependent upon "congruent grammatical structures and a priori similar vocabularies" (cf. Penzl's discussion of loan imitation and substitution, 3.1. above). On the other hand, Jeffers and Lehiste (1979: 149 — 150) cite an instance in which the morphological acceptance by the target language is only partially completed. Russian pal'to 'overcoat' is borrowed from French paletot. The French masculine noun assumes neuter gender in Russian under the influence of neuter nouns ending in -o. However, pal'to resists inflection in all cases and numbers and accordingly is still perceived as a loan. Syntactic interference on the word level is noted by Lehmann (1973 b: 221) in the case of English compounds such as attorney general and malice aforethought, which reflect the noun adjective structure of a VO language such as French, from which English borrowed these collocations. The Russian habit of signalling possession by the preposition u 'at, by', rather than through a verb of possession, for example, Rus. u menja dengi 'at me money' = Ί have money', reflects transfer of Finno-Ugric syntax, thus Hungarian van nekem egy könyr 'there is to me one book' = Ί have a book', according to Arlotto (1972: 194). Arlotto further cites the borrowing of the Persian izafet, the morpheme linking adjective and noun, together with Persian NA order in Turkish, for example, Per. mardan -e xub 'men -izafet good' = "good men' beside Tur. donanma -i hümayun 'fleet -izafet imperial' = 'the imperial fleet' (1972: 195). Topicalization in the English spoken in Ireland is affected by the Celtic substratum. Jeffers and Lehiste (1979: 155) observe the possible variations of the sentence I'm going to Dublin tomorrow in HibernoEnglish as "a) It's me that's going to Dublin tomorrow, b) It's going that I am to Dublin tomorrow, c) It's to Dublin that I'm going tomorrow, d) It's Dublin that I'm going to tomorrow, e) It's tomorrow that I'm going to Dublin". Another syntactic example is the basic word order of Amharic, an Ethiopic Semitic language, which is attributable to enduring contact with Cushitic (Givon in Hyman 1975: 115).

Evidence of language change

49

When contact results in the direct phonological assumption of morphemes, it belongs to morphological contact (cf. above). When contact results in the partial or total assumption of the meaning of a word from the source language with or without phonological resemblance, we speak of semantic change in the target language. Thus, contact between Old Norse and Old English led to the change of OE dwellan 'lead astray' to 'dwell' due to ON dvelja 'abide'; it further led to the change of OE eorl 'brave warrior' to 'earl' due to ON jarl 'nobleman' (Lehmann 1973 b: 217). Bloomfield (1933: 455) cites the Christianization of the concepts Go. daupjan 'dip', OE fulljan from fullwihjan 'to make fully sacred', and ON skirja 'to make bright' by borrowing from Greek and Latin the meaning 'baptize'. So-called loan translations work with native morphology and thus belong to semantic contact change. For example, Gk. suneidesis composed of sun 'with' and eidenai 'to know' = 'conscience' is rendered by the corresponding native components: Lat. con 'with' and scientia 'knowledge' = conscientia\ Go. mip 'with' and wiss- from *witan 'know' = mipwissi; SI. so 'with' and vest' 'knowledge' = sovest' (Bloomfield 1933: 456). 4.1. Earlier grammarians The importance of grammarians in the history of Sanskrit is unequalled anywhere in the world. Also the accuracy of their linguistic analysis is unequalled until comparatively modern times. The whole of the classical literature of Sanskrit is written in a form of language which is regulated to the last detail by the work of Pänini and his successors (Burrow 1955: 47).

The fifth century Indie linguistic treatise Nirukta of Yäska and the some four thousand rules of the Sütras of Pänini (fourth century B. C.), together with their elucidation in the Mahäbhäsya of Patanjali (150 B. C.), represent India's unmatched grammatical legacy. In Greece Plato's (347 B. C.) Kratylos is an early grammatical document, but Aristotle (322 B.C.) is regarded as "the father of grammar in the Occidental world" (Gray 1939: 423). He is followed by the Stoic Chrysippos (206 B. C.). The first Greek grammar transmitted to posterity intact is that of Dionysius Thrax (second century B. C.). The Latin grammarian Varro (27 B.C.) transmitted to us his De lingua Latina, of which six out of twenty-four books remain. Roman grammarians are followed by the Byzantine Priscian (560), who wrote Grammatical Commentaries based on the Alexandrine Apollonius Dyskolos and Herodian. The Graeco-Roman grammarians dominate

50

Irmengard

Rauch

grammatical thought in the Middle Ages. Noteworthy are the two treatises on grammar: De grammatica of St. Anselm (1109) and De modis significandi, sive grammatica speculativa of Duns Scotus (1308). Especially based on Priscian are the Graecismus of Eberhard of Bethune (thirteenth century) and the school grammar Doctrinale puerorum of Alexander de Villa Dei (1199). Grammars dealing with non-classical languages emerge, such as Aelfric's Grammatica latino-saxonica, which dates from the eleventh century, and the Welsh Dosparth Ederym Day od Aur of Ederyn the Golden-Tongued from the thirteenth century. The Irish Auraicept na n-eces is of unknown author and date. The four Old Icelandic treatises date from the twelfth to the fourteenth centuries. And Dante's De vulgari eloquentia was written in the early fourteenth century. This trend continues into the fifteenth and sixteenth centuries and includes grammars of languages such as that of Spanish Arabic, Arte para ligeramente saber la lengua aräviga (1505) by Pedro de Alcala; that of Hungarian, Grammatica Hungaro-Latina in usum puerorum (1539) by Erdösi; that of Ethiopic, Chaldaeae sive Aethiopicae linguae institutiones (1548) by Victorius; that of Mexican, Totonak, and Huastek, Grammatica et lexicon linguae Mecicanae, Totonacae et Huaxtecae (1555 — 60) by Olmos. In the seventeenth century, grammarians turn to Tagalog, Arte y reglas de la lengua Tagala (1610) by Joseph; to Malay, Spieghel van de maleysche taal (1612) by Ruyll; to Finnish, Linguae Fennicae institutio (1649) by Petraeus; to Moxo, Arte y vocabulärio de la lengua morocosi (1699) by Marban. (Cf. Gray 1939, Chapter 13). Malmberg (1964) gives an account of earlier grammarians in his opening chapter and Pedersen (1931) is an excellent source for eighteenth and nineteenth century grammarians. Kukenheim's (1932) study of the history of Italian, Spanish, and French grammar during the Renaissance offers a listing of grammarians and grammars from Ancient Greek and Latin through the fifteenth, sixteenth, and seventeenth centuries of these Romance languages (1932: 218 — 229). 4.2. Grammarians as witnesses for evidence Graphemics among other parts of the grammar is discussed by an earlier grammarian in the Middle Irish Auraicept na n-eces (Calder 1917: 89):

Evidence of language change

51

Now Fenius Forsaidh is the same man that discovered these four alphabets, to wit, the Hebrew, Greek, and Latin alphabets, and the Beithe Luis Nin of the Ogham, and it is for this reason the last, to wit, the Beithe is more exact because it was discovered last. There were in the school twenty-five that were noblest among them, and these are their names, which are upon the Beithe Luis Nin both vowels and consonants ...

The unknown grammarian of the Old Icelandic First grammatical treatise (Haugen 1950: 13) also speaks of orthography as part of his phonology: I have composed an alphabet for us Icelanders as well ... I have used all the Latin letters that seemed to fit our language well and could retain their proper sound, as well as some other letters that seemed needful to me... Some of the consonants of the Latin alphabet were rejected, and some new ones added. No vowels were rejected, but a good many were added, since our language has the greatest number of vowel sounds.

Ancient evidence of sign representation might be Tacitus' (first century A.D.) observation: The Egyptians first represented concepts by means of animal figures; these oldest monuments of human memory may still be seen engraved in stone. They claim to be the inventors of writing. From them the Phoenicians are said to have brought the script to Greece because they ruled the seas, and they received the credit for inventing what they only took over (Annals II: 4, as quoted in Lehmann 1973 b: 62).

Although also intent on reforming orthography, Louis Meigret of Lyons in his sixteenth-century Gramm§re provides eyewitness evidence for the pronunciation of French vowels, for example: A is the one which undeniably requires a wider mouth opening than any other [vowel]; that for open e ... is thereafter greater than any of the others, and somewhat less than for a, as may be seen in jamgs 'never', or in as '(thou) hast', ps vb. 'art', §t 'is'. Then comes close e, the other [e] being more open (Shipman 1953: 19).

Also in the sixteenth century, but for English, John Hart describes the pronunciation of consonants: "de fou · r pe · ra huit£ ar ma · d uid a stoping brej): tu ui · t b, ρ : d, t: g, k : and d3, t?· den d'uder J)ri · J>rulei bre · dd pe · rs, tu ui · t d, p : ν : f and z, s · den de 5 semiuoka · Is I, m, n, r, and L, and de tu · bre · ds t and h " (Jespersen 1907: 22). For sixteenth-century Italian, Claudio Tolomei observes: ... and I would be so bold as to say that in the original and pure speech of Tuscans, this was a universally valid rule [i.e., that Latin pi- gave Italian pi-], and that all those words which are now used and written differently, such as plora: 'he weeps', implora: 'he implores', splende : 'it is resplendent', plebe: 'populace' and the like, were not taken from the middle of the town squares of Tuscany, i.e., from everyday speech, but were set up by writers, and by someone who wishes to enrich the language, preferring to use them in the form in which he found them written in

52

Irmengard Rauch Latin, without giving them the form of Tuscan speech [i. e., without substituting pifor pi-] ... because without a doubt the common usage of earlier times would, had it inherited these words, have said piora, impiora, spiende and pieve, and we have manifest evidence of this latter in that in the vernacular we call pieve a church devoted to the religious services of the common people (Hall 1964: 301).

Tolomei thus exhibits working knowledge of the principle of the regularity of sound change and of the value of explaining its exceptions. Penzl (1957: 204) relates Valentin Ickelsamer's negative attitude, in his Teutsche Grammatica of 1537, toward the cluster spelling : "Since neither sound is completely heard in Engel angel franck but rather a fusion ('zusammen schmeltzung')". This attests to the change of η + g to [η]. With reference to the English Great Vowel Shift (cf. 2.2. above), grammarians contribute considerable evidence, for example, Wyld (Scott 1967: 75) cites occasional spellings such as agryd and symed to demonstrate the raising of ME e. And a non-grammarian native speaker, Queen Elizabeth I, is credited with writing " 'biquived' for bequeathed" (Lehmann 1973b: 162-163). Evidence from grammarians (cf. 4.1.) spans centuries and countries. Hall (1964: 293) relates an incident recorded by the Latin historian Suetonius in which the Emperor Vespasian was criticized by the grammarian Florus for using substandard /plo:stra/ 'wagons' instead of prescribed /plaustra/. In turn, Vespasian called Florus Flaurus. Hall interprets this as evidence of Proto-Romance tense ο for such words as /ko A da/ 'tail' from /kauda/. Hall further relates (1964: 293-294) the satirizing by the Roman poet Catullus of the /?-lessness of an upstart by the name of Arrius. Catullus overcorrected him by adding unhistorical h to Arrius, thus Harrius, commodo, thus chommoda, which Hall interprets as evidence for an Α-less Proto-Romance compared with Classical Latin. Other components of the grammar are also described by earlier grammarians. Polome (1983 b) explains Lambert ten Kate's systematization of verb conjugation and the morphophonemic alternations involved in ten Kate's Gemeenschap tussen de Gottische Spraeke en de Nederduytsche of 1710 and his Aenleiding tot de Kennisse van het verhevene Deel der Nederduitse Sprake of 1723. Polome notes: The criterion for this classification [of verbs] are the vowel alternations as ten Kate himself points out in his letter to Α. V. ( = Verwer, his form teacher): 'De Verba vond ik onderscheyden in soorten, waerom ik die gene, welke in de loop hunner veranderinge eenerley rooy hielden, onder eene zelfde Classis bracht; en deze wederom elk in zyn soorte byzonder afdeelde: aldus vind 'er Uw Ε ... sesderley van de tweede (Classis), sevenderley van de derde, tweederley van de vierde ... welker elk in klank of verwisselinge iets verschilt' (Polome 1983 b, Fn. 6; cf. also 1983 a).

Evidence of language change

53

Rohlfs (1970: 27) explains material from the Appendix Probi (thirdfourth centuries A. D.), in which an unknown author called Probus deplores the Vulgar Latin loss of inflection on street names; thus Probus prescribed "vico capitis Africae non vico caput Africae". The foundations of Greek syntax are found in the writings of Apollonius Dyscolus (second century A. D.). Apollonius raises the question of why irregularities occur in syntax: II faut done exposer avec soin ce qui produit l'irregularite, et pour cela il ne faut pas se borner au vain travail de rassembler des examples de figures, comme ont fait quelques auteurs, proclamant tres-haut des solecismes, mais sans en dire la raison; car si vous ne faites pas voir cette raison, tous vos exemples sont en pure perte (Egger 1854: 223).

Appolonius considers both celui-ci, the masculine pronoun with reference to a female, and ainsi parlerent la multitude, plural verb with singular subject, as irregularities, while some of his peers argue that the first can only be agrammatical if the referent is visually observed. He refutes this explanation saying: Conclusion ridicule; car les solecismes tombent sous le sens de l'ou'ie et se reconnaissent au desaccord des mots rapproches dans une meme phrase; ils sont done sensibles meme pour un aveugle, puisque l'aveugle ne manque pas du sens auquel la voix s'adresse, je veux dire de l'ou'ie (Egger 1854: 224—225).

Visser (1973) ascribes the actual change in English from the 'be' auxiliary to 'have' with intransitive partly to the negative influence of eighteenth- and nineteenth-century grammarians. He thus quotes John Burn, A practical grammar 1766: Neuter [sc. intransitive] verbs signifying some sort of motion, change of place, of posture, etc., admit of a passive form without losing their neuter signification; as, He is come, He is gone, I am set, etc., but the propriety of the neuter verb, in the passive forms, is doubtful in the following examples, 'how very far he is departed from the purity and holiness of the divine nature' (Dickinson's Letters). — Better thus...'has departed'. Ί am informed by the tower-guns that the place is surrendered' (Spectator no. 165); 'the rules of our holy religion, from which we are infinitely swerved' (Tillotson, Sermon 27) ... (Visser 1973: 2043).

Visser further quotes L. Brittain in his 1778 Rudiments of English grammar as considering as "Errors" the use of "be" in the following: " 'The obligation was ceased' (Hume); 'whose number was amounted to 300' (Swift); 'half the men are deserted' (Addison) ..." (Visser 1973: 2043). Finally Visser cites William Cobbett's A grammar of 1841: 'The noble Earl, on returning to town, found that the noble Countesse was eloped with his Grace.' I read this very sentence in an English newspaper not long ago. It should be had eloped', for was eloped means that somebody had eloped the Countess·,

54

Irmengard Rauch it means that she had received, or endured, from some actor, the act of eloping, whereas, she is the actress, and the act is confined to herself (Visser 1973: 2044).

Semantics among the early grammarians concentrates heavily on etymology. Among the most well-known older etymological works are the twenty books of the Origines sive Etymologiae of Isidore of Seville (636). In Book XI he renders his classical etymology of man thus: Man, homo, is so called because he was made from the soil, humus, as is [also] said in Genesis [2: 7]: Et creavit Deus hominem de humo terrae, 'And God created man from the dust of the earth'. By an improper usage, however, it is announced that the whole man [was made] from two substances, that is, from the union of body and soul. For properly 'man', homo, is from 'soil', humus (Sharpe 1964: 38).

Another very intriguing etymology is rendered in the Old High German Notker Labeo's (1022) introduction to his translation of Martianus Capella's De Nuptiis Philologiae et Mercurii. In the opening sentence Notker, referring to Remigius Autissiodorensis (908) as his source, explains the significance of the name Martianus Mineus Felix Capella. In particular, Notker notes the etymology of Capella which he glosses as 'sharp mind', related to Latin capra (sic) 'she-goat' which is dorcas 'gazelle' in Greek, etymologically 'seeing', that is, 'bright-eyed'. To be sure, Ernout and Meillet (1959) show the gloss dorkas for caprea, but both Frisk (1960) and Pokorny (1959: 513) give as the Indo-European etymon for dorkas IE *iork 'Tier aus der Gruppe der Rehe', and not IE*derk — 'blicken' (1959: 213). Notker reads: "Remigius leret unsih tisen auctorem in alenämen uuesen geheizenen Martianum, unde Mineum umbe sina fareuua, Felicem limbe heilesod, Capellam ümbe sinen uuässen sin, uuända capra apud Grecos dorcas a videndo geheizen ist" (Braune 1962: 65). 5.1. Reconstructed and typological materials Reconstruction is based on both concrete and hypothesized evidence. The latter depends heavily on typological materials. The type of reconstruction known as internal reconstruction exploits paradigmatic alternations within a given language in an effort to explain their origin. Hoenigswald (1960: 68 — 69) enumerates a wide variety of inferences that can be drawn from alternates: Sound change, syncretism, the functioning of analogy, dialect borrowing, obsolescence, homonymy. In particular, morphophonemic alternations have been studied to reconstruct two types of alternations from primary split, one in which the one alternate merges with an already existing phoneme, the other

Evidence of language change

55

in which the alternate merges with zero. Alternations resulting from secondary split differ in that they point to a previous merger (1960: 102 — 104). Hoenigswald points out, however, that with alternations one does not necessarily know the direction of the split, that is, which is the original stage and which the final. Other evidence such as isolated forms and doublets indicate the direction. He writes (1960: 111): "The order of trustworthiness then is: indeterminate morphemes ('isolated forms'), doublets, alternating morphemes, non-alternating morphemes." Kurylowicz, master of internal reconstruction, claims that the object of comparison of alternations, e. g., in the Germanic consonant shift, should be innerlinguistic rather than crosslinguistic (1964: 11). Since the methods of reconstruction aim to recover earlier stages of alternations, they yield insights as to the relative chronology of these stages (cf. below ordering of rules 5.3.). Much of Kurylowicz' data is semantically based, i. e., alternations reflecting a split must be identified as reflecting the primary or secondary function of the original form. He observes: The hierarchy of functions within a semantic zone has a panchronic character, in principle independent of the individual languages ... The more grammatical (i.e., general) a given category, the less differentiated the secondary functions, the easier to determine the semantic split and the relative chronology (1964: 26). (See 5.2. for examples.)

Inherent in the Kurylowicz observation is the fact that the methods of internal reconstruction intersect with those of reconstruction by typology. Kurylowicz' (1964: 29) several panchronic laws which he derives from his data on internal reconstruction read as typological laws "of functional shift such as: iterative > durative (present) > (general or indetermined) present; static verb > perfect > indetermined past > narrative tense; desiderative > future; adverb > 'concrete' case > 'grammatical' case; sex (in substantive) > gender (in adjective); collective > plural". In turn, linguistic typology in language change is closely allied with the study of "drift", tendencies, and universals. Typologies have been constructed of the phonological, morphological, syntactic, and semantic levels of language. They serve as models against which data can be compared and reconstructions can be inferred and/or predicted. Hockett (1955) provides a large array of vowel typologies based on two, three, or four contrasting heights. For consonant typologies he uses the number of manner of articulation contrasts as the primary classi-

56

Irmengard Rauch

fication of types. He then compares skewed systems against his models, observing (1955: 89): "Uncertainties of classification, such as for Votyak, and holes in otherwise neat patterns, such as the absence of /ü/ in varieties of standard Korean which have an /ö/, are normal enough". Precisely these sorts of uncertainties yield evidence for reconstruction, and Hockett's use of the term "hole" strongly reminds of Martinet's (1955) "push" and "drag chains" for inferring directions of language change. Phonological typologies can also be constructed on the basis of acoustic distinctive features (Jakobson — Fant — Halle 1951). The use of typology as evidence in morphological change per se appears more restricted in the literature than that for phonology and syntax. To a certain extent, this is attributable to the fact that morphological change is frequently evidenced in phonological or syntactic data. Greenberg (1960), building on Sapir's (1921: 142-143) classification, sets up ten morphological indices: An index of synthesis, agglutination, compounding, derivation, gross inflection, prefixing, suffixing, isolation, pure inflection, and concord (1960: 187). He claims his indices yield evidence of "the general direction of historical changes in languages over an extended period" (1960: 194). Greenberg, too, is responsible for the early construction of typologies for syntax (see Greenberg 1963). He proposes some forty-five universale, twenty-six of which deal directly with the six basic order typologies, and nineteen of which concern morphological correlations. Evidence of one universal in a typological network is predictive of the entire network. As Jakobson (1962: 526) writes: "Typology discloses laws of implication which underlie the ... structure of language: the presence of A implies the presence (or on the contrary the absence) of B. In this way we detect in the languages of the world uniformities or near-uniformities As is obvious from Greenberg (above), typological evidence readily includes universals; the statistics, the degrees to which they occur, and their correlations play roles in defining specific typologies, but a universal in and of itself obviously "cannot lead to a typology" (Greenberg 1973: 184). Thus, for example, Ullmann, in his discussion of "Semantic universals" (1963), considers that four of five universals of descriptive semantics, viz. " — motivation, generic versus specific terms, polysemy, and homonymy — may, if studied on a suitable scale, yield criteria for linguistic typology. All four criteria are statistical: they are concerned with relative frequencies". Ullmann excludes synonymy, about which, although he is able to derive some general

Evidence of language change

57

principles, he expresses doubts as to its validity. He cautions that "certain semantic phenomena are not precise enough to be amenable to rigorously statistical analysis ..." (1963: 174). Ullmann nicely distinguishes universals, in turn, from tendencies which are not necessarily universal. The evidence from tendencies and universals according to Ullmann "help[s] the etymologist and the comparatist in two ways. Firstly, it would tell him what kind of changes to expect, and whether a particular change suggested by his data would be common or infrequent, normal or exceptional. Secondly, it would enable him to choose between alternative explanations" (1963: 196). Building on the Saussurean analogy of language change maneuvers similar to chess moves and Jakobson's concept of therapeutic processes maintaining the equilibrium within a language system, van Ginneken (1956: 576) notes a series of "incompatible or hostile linguistic tendencies" and a series of " 'sworn comrades', i. e., favorable tendencies". That these are not necessarily universal is seen in van Ginneken's observation that Jakobson's explanation of Slavic phonological structure, as a fight between the tendency of soft and hard consonant syllables and the incompatible tendency of musical accent, is not at all the case in Japanese where consonant harmony and musical accent are compatible tendencies (1956: 577). With the study of tendencies as evidence, we move into the realm of Sapir's "drift", which Malkiel (1981: 544) argues becomes "almost indistinguishable from 'tendency'" but is to be clearly distinguished from "slope" and "slant". Malkiel thus reduces the power generally accorded Sapirean drift by limiting it to "a single, isolated, undisturbed evolutionary strain or streak ... a slowly-occurring change" (1981: 566). Vennemann (1975: 274) interprets Sapir's concept drift, writing "Sapir explains certain past and present morphological, syntactic, and lexical changes of English by revealing that they are consequences of certain psychological tendencies of speakers of this language — which he calls 'drifts'". Concerning its use as evidence in language change, Vennemann continues, "Since such tendencies remain alive over long periods of time, he (Sapir) feels safe to predict certain further, similar changes for the future of English ..." We accordingly perceive in the use of tendencies, drifts, typologies, and reconstructed materials evidence with predictive power for language change. Certainly, universals are considered most powerful in their predictive and explanatory capabilities. Lieb (1978) offers an axiomatic description of the use of universals in linguistic explanation.

58

Irmengard

Rauch

With particular reference to diachronic materials, Greenberg (1978) shows how diachronic principles can help explain synchronic factors. This is consistent with the use of synchronic evidence in reconstruction of historical language change. For example, insights into a marked feature can be gotten by understanding its origin relative to its corresponding unmarked feature. Further Greenberg writes (1978: 89): "Diachronic principles are involved in the explanation of both low and higher level synchronic generalizations. In so doing they often explain exceptions". One of the most provocative proposals for treating evidence in language change is the ordering universal of "proper inclusion precedence" of Koutsoudas, Sanders, and Noll (1974), whereby rules are ordered according to the inclusion of the structural description of one in the other. Evidence from typological materials follows in 5.3. 5.2. Evidence from reconstruction Phonological split can readily be reconstructed from morphophonemic alternations. The Modern German interchange of voiced obstruents intervocalically with voiceless obstruents in final position, e.g., Bund 'alliance' ~ Bundes (genitive singular) where [-t] alternates with [-d-] and merges with [t] of bunt (adjective) 'brightly colored'. In Ancient Greek a split of s to 0 and to s can be reconstructed from such pairs as geneos (genitive singular) 'race' and genessi (dative plural), respectively. The s of the root genes was lost intervocalically. However, words such as ambrosia 'elixir of life' which has a cognate in t, ambrotos 'immortal' indicate the reintroduction of s intervocalically before -/, pointing to a second split (Jeffers — Lehiste 1979: 38 — 40). The relative chronology can be established as counter-feeding since the shift of t > s must follow the shift of s > 0 if an s is to be preserved intervocalically. Hoenigswald (1960: 102, 111) gives evidence from Latin to demonstrate the reconstruction of a primary split with conditioned merger. The alternation of s with r in, e. g., est (third person singular present) 'be' with erimus (first person plural future) points to intervocalic voicing of s to r in which the latter merges with an already existing phoneme. The occurrence of the isolated Lat. cräs 'tomorrow' shows that the non-intervocalic s stayed s. In the case of Eng. kite beside knight, the primary split of k > k and 0 before η can be inferred (1960: 92). The fact that in Sanskrit c occurs before palatal vowels

Evidence of language change

59

and k before velar vowels points to a conditioned split which, however, is secondary since both c and k occur before a, from which a previous merger of ο and e to a can be inferred (1960: 102—103). Kurylowicz (1964: 11) predicts the Germanic shift of IE *t > Gmc. *th on the basis of the coalescence of t, e.g., IE *esti (third person singular present) 'be' with d of IE *nisdos 'nest', whereby t becomes marked in the cluster st. Martinet (1955: 182) views the aspiration in Gmc. *th as impetus for the devoicing of IE *d in Germanic in a drag chain. His reconstructions rely heavily on general language tendencies (cf. 5.1. and 5.3.). On the morphological level Kurylowicz observes the syncretism of the ablative and genitive singular cases except in the o-stems in Sanskrit. In the o-stems a new ablative built on pronominal morphology was formed, for example, vrkät with primary function ablative, beside bähoh with primary function genitive, secondary function ablative. Kurylowicz concludes that the original function of bähoh can be inferred as ablative, and that the split to genitive took place in the transformation from adverbal partitive to adnominal partitive as reflected, for example, in the sentence Lat. hostem occidit > occisio hostis 'he kills the enemy > the killing of the enemy' (1964: 18 — 19). Semantic evidence at the morpho-syntactic level is the shift in Sanskrit from aspect to tense in the aorist. Kurylowicz (1964: 23) reconstructs the Indo-European present form as having the primary function of imperfectivity and the secondary function of simultaneity with the moment of speaking; the Indo-European imperfect form as having the primary function of imperfectivity and the secondary function of simultaneity with a past moment; the Indo-European aorist form as having the primary function of perfectivity and the secondary function of anteriority referring to the moment of speaking. In narration the aorist becomes identified with the imperfect; accordingly it loses its primary function, perfectivity, and is limited to its secondary function of anteriority. In turn, the imperfectivity of the imperfect too becomes secondary. Kurylowicz compares the distinction between Sanskrit aorist and imperfect as that between Eng. I have written a letter and I wrote a letter. In fact a similar development from primary aspect to primary tense can be traced for the Germanic languages. The Modern English alternations I have a letter written and I have written a letter give clues of this earlier shift (cf. Rauch 1982). Other than inferring original states, related changes, and relative chronologies, evidence in reconstruction can point to intermediate

60

Irmengard Rauch

states in language change. Thus Penzl (1971: 153 — 154) postulates a pre-Old High German stage of aspirated voiceless stops from Germanic as an intermediate stage in the Second Sound Shift. A further stage is the affrication of these aspirated stops where their final stage results in geminate fricatives. Where their final stage is identical with the affricated stage Old High German evidence shows alternates minimally constrained after vowel, e. g., uuize [s] (third person singular present optative) 'punish' beside uuizze [ts] 'knowledge'. As a final case of evidence in reconstruction, let us consider the paradigm example of the reconstruction of the Indo-Hittite laryngeals. Saussure (1879), prior to the discovery of the Hittite evidence, relied entirely on reconstructed Indo-European alternations. The ablaut alternations between normal grade e or ο and lengthened grade e, δ served as evidence for Saussure of a split of an original short vowel to a vowel constrained by laryngeal. Similarly, his observation of the canonical syllable shape of the Indo-European root as consonantvowel-consonant beside roots with only one consonant, for example IE *bher 'bear' and IE *es 'be', respectively, gave evidence of a split resulting in zero in the case of the monoconsonantal roots. Postulating the Indo-Hittite hypothesis, Sturtevant (1942) provides thorough evidence of laryngeals in Hittite, and of the effects of laryngeals in IndoEuropean. Evidence for the articulation of the Indo-Hittite laryngeals is derived from within Indo-European by Lehmann (1955: 105 — 108) by observing the effects of laryngeals in the later dialects. He infers that they had continuant articulation, three of four were voiceless, and they patterned like fricatives, i.e., they had relatively open articulation. Lehmann's evidence consists of such changes as the Old French loss of voiceless s with compensatory lengthening, e. g., OF testa > N F tete 'head'; the voiceless reflex of Proto-Indo-European resonants constrained by another continuant, s, in Greek; and the patterning of the laryngeal near the peak, the most open part of the syllable, e. g., Lat. lävit 'washed' < leXw-. Non-generic evidence is adduced by Sapir from contemporary language in reconstructing the Indo-Hittite laryngeals quasi-typologically from Nootka and Kwakiutl evidence in his 1938 article "Glottalized continuants in Navaho, Nootka, and Kwakiutl (with a note on Indo-European)". Martinet (1957) combines both Lehmann's internal evidence and Sapir's external evidence; for the latter he appeals to Modern Arabic and Abaza.

Evidence of language change

61

5.3. Evidence from typology Asymmetry in the stop inventory of Old Irish, where ρ is virtually lacking except for its occurrence in loan words, allows the inference to be drawn that IE *p was lost in the early history of Irish (Jeffers — Lehiste 1979: 47). On the other hand, "l'attraction du systeme" is accountable, according to Martinet (1955: 179), for the "pushing" of *T to [z] in Old Castilian due to the change of /ll/ > [T]. Martinet's insights are not mechanical, but rather they rely on general tendencies. He writes (1955: 182): "Inspiration d'une serie non-voisee entraine en general un devoisement gradual concomitant de la serie voisee: si, dans une correlation de voix, /t/ tend vers [th], /d/ tendra vers [4]". Jakobson (1962: 528 — 529) offers an array of evidence from typology which reads in part as tendencies: there is no support in the recorded languages of the world for the one-vowel Proto-Indo-European structure. In no language does a /dh/ develop beside /t/ and /d/ without a /th/ also occurring. Lastly, Jakobson notes that "views prior or opposed to laryngeal theory, which assign no jhj to IE, disagree with typological experience" (1962: 528). Building on the tendency of languages which distinguish the pairs voiced: voiceless and aspirate: nonaspirate to have an /h/, Jakobson observes that Slavic, Baltic, Celtic, and Tocharian make no distinction between aspirates and non-aspirates, while Greek, Indie, Germanic, and Armenian treat the two series differently and also have /h/. It is on the basis of typological evidence or paradigmatic congruity, in particular the rarity (near or alleged absence) of b, that the voiced stops come under possible reinterpretation as glottalized (voiceless) stops in Proto-Indo-European (cf. Szemerenyi 1985: 4 - 1 5 ) . Distinctive feature typologies are considered by Jakobson, Fant, and Halle (1951: 33). The primary tonality feature of both consonants and vowels can determine a square or a triangular pattern. When we compare, for example, the Czech, Slovak, Serbocroatian, Hungarian square voiceless stop structure with a grave: acute opposition among compact consonants to the lack of this opposition in the triangular English and French systems, we may infer, with the help of some general tendencies, changes relative to the common Indo-European origin for the two systems. Anttila (1972: 340) gives evidence from Tai where word-final η of Siamese corresponds to both word-final η and / of Saek. Whereas an η can be reconstructed for the η correspondences, one hesitates to reconstruct a final / for the I: η contrasts since, as

62

Irmengard Rauch

Anttila writes: "such an / goes heavily against the 'areal grain', because no language has, or is known to have had, a final -/ in the immediate typological area". Typological evidence from morphology is adduced by Lightfoot (1979: 160) in his observation that New English compounds such as windowcleaner, meathook reflect an earlier OV syntactic word order which is still productive as evidenced in the more recent car-wash and snowplough. Lehmann (1973 a), working with sentence qualifier morphemes, observes a correlation between OV languages in which such qualifiers as the negative, interrogative, reflexive, potential or causative follow the verb and are characteristic of agglutinating languages. Conversely, in VO languages these qualifiers precede the verb and tend to appear in synthetic languages. He cites evidence from several languages including the paradigm OV Japanese (1973 a: 52): yomu ka 'does he read?', yoma-nai 'he does not read', yom-eru 'he can read', as compared with the paradigm VO Hebrew (1973 a: 54): halca: tab se .per 'did he write a book?', lo: ka: taB 'he did not write', hiKti: Β 'he caused to write'. Deviations from these correlations may point to language change. In the evidence provided by Quirk (1970) for the ongoing morphological change of the irregular weak verbs in American English, Rauch (1978) finds a correlation between root-alternating verbs and voiceless dental suffix, e. g., dream: dreamt, and non-alternating-root verbs and voiced dental suffix, e. g., spill: spilled. The syntactic typology based on the principle of natural serialization (Bartsch — Vennemann 1972), whereby order of modifiers to their heads emulates essentially the order of the object to its verb, is used as a paradigm against which inconsistencies can be studied as evidence of earlier change or as evidence for ongoing or future change (cf., however, Lightfoot 1983). For example, Lehmann (1971: 22 — 23), studying German through time, notes that it is evincing SOV characteristics such as adjective-noun order, which was less frequent in Old and Middle High German. Middle High German phrases such as kiinec guot 'good king' is reflected in a relic or frozen phrase such as N H G Karl der Große 'Charles the Great'. Lehmann addresses the extended adjective innovation of New High German as further evidence for German becoming OV (cf. otherwise Esau 1973). Ullmann (1963: 188) indicates how universale interdigitate with typology. There is a direct correlation between typological criteria and context sensitivity. Evidence from French contrasted with German, viz. Fr. aller versus Ger. gehen 'go', reiten 'ride', fahren 'drive', shows

Evidence of language change

63

the French to be by far the more context sensitive. Ullmann writes: "It follows that languages where generic terms, polysemy, and mononymy are prevalent will be relatively 'context bound'; French is a classic example of this type of semantic structure". Evidence from tendencies may be grouped with that from drift (cf. 5.2. above). Ullmann (1963: 190 — 193) cites the expanded use of the word shrapnel for beans by French soldiers during World War I; the anthropomorphic extension of body parts to non-human entities, e. g., the heart of the matter; the extension of concrete to abstract meanings, e. g., light in to throw light on or to enlighten·, and the synesthetic combinations such as cold voice. The most notable evidence in the literature for metaphorical tendencies comes from Sapir (1921: 156): Who(m) did you seel He attributed the loss of the accusative morpheme to three major drifts: "the familiar tendency to level the distinction between the subjective and the objective" (1921: 163); "the tendency to fixed position in the sentence, determined by the syntactic relation of the word" (1921: 166); and "the drift toward the invariable word" (1921: 168; cf. also Bynon 1977: 150-154). Where tendencies or statistical universals end and universale begin is often impossible to determine. Thus, for example, Benveniste's statement (1968: 92 — 93) that "auxiliation is a syntactic process generously used in the widest range of languages" appears non-universal, yet he builds universals upon the tendency. He writes: "A syntagm involving an auxiliary universally exhibits certain common traits". Benveniste's evidence comes from two American Indian languages, Tunica and Aztec, in which he observes as one of his universals the fact that they enjoy non-auxiliary or independent existence. Very prominent universals in syntax are Behaghel's (1932: 4) Law I, whereby semantically associated concepts are syntactically adjacent, and Law II, whereby topical material precedes new information in the sentence. Rauch (1981) appeals to Law I to explain the predominant verb-first order evidence in Old Saxon. Universals, avidly studied in the past quarter-century for all components of the grammar with synchronic and diachronic evidence, have also been viewed with special reference to language change. Thus Hoenigswald asks the question "Are there universals of linguistic change?" (1963), which he answers in the affirmative. Not the least evidence comes from the exceptions to sound change, allowing the universal that (1963: 29) "sound change 'is' indeed regular everywhere on earth". Greenberg (1978: 86) gives evidence of diachronic factors

64

Irmengard Rauch

in explaining exceptions to synchronic generalizations. Japanese, for example, has no final voiceless vowels, which might be expected due to the generalization that word-internal voiceless vowels imply wordfinal voiceless vowels. The diachronic factor of final voiceless vowel proceeding to zero explains the Japanese evidence. Similarly, long vowels are universally marked. If the number of long vowels is equal to or larger than that of short vowels in a given system then diachronic change is implied. The method in applying this evidence is like that of reconstruction (cf. 5.2.); however, here reconstruction is effected through known universale. Finally, of all the work done on the ordering of rules, the "proper inclusion precedence" of Koutsoudas, Sanders, and Noll (1974) reads as a universal (cf. 5.1.). Evidence for the application of this rule is to be found in the distinct treatment in Latin American Spanish of / which becomes a glide except in final position, where it is depalatalized as are all I in Castilian Spanish. They posit the two bleeding and counter-bleeding rules: (a) T—> I / # , (b) Γ—> y, which yield Latin American akel 'that' and akeyos 'those', but Castilian akel and akelos. Koutsoudas — Sanders — Noll write: This principle (proper inclusion precedence) determines the correct application of the two Latin American Spanish rules (a before b) . . . , since the structural description of the former is the string / # and that of the latter is the string T, and Τ φ properly includes I. There is no need therefore, for any language-specific constraints on the application of these rules (1974: 9).

References Anderson, John M. —C. Jones (eds.) 1974 Historical linguistics. Amsterdam: North Holland. Anttila, Raimo 1972 Introduction to historical and comparative linguistics. Boston: Houghton Mifflin. Arlotto, Anthony 1972 Introduction to Historical Linguistics. Boston: Houghton Mifflin. Bailey, R. W . - L . Matejka-P. Steiner (eds.) 1978 The sign: Semiotics around the world. Ann Arbor: University of Michigan. Baldi, Philip 1983 An introduction to the Indo-European languages. Carbondale: Southern Illinois University Press. Bartsch, Renate —Theo Vennemann 1972 Semantic structures: A study in the relation between semantic and syntax. Frankfurt: Athenäum.

Evidence of language change

65

Behaghel, Otto 1932 Deutsche Syntax. Vol. 4. Heidelberg: Winter. Benveniste, Emile 1968 "Mutations of linguistic categories", in: Lehmann — Malkiel (eds.), 83 — 94. Birnbaum, Henrik—Jaan Puhvel (eds.) 1966 Ancient Indo-European dialects. Berkeley: University of California Press. Bloomfield, Leonard 1933 Language. New York: Holt, Rinehart, Winston. Botha, Rudolf P. 1981 The conduct of linguistic inquiry: A systematic introduction to the methodology of generative grammars. The Hague: Mouton. Braune, Wilhelm 1962 Althochdeutsches Lesebuch (14th revised edition, revised by Ernst Ebbinghaus on the basis of the 13th edition (1958) revised by Karl Helm). Tübingen: Niemeyer. Brugmann, Karl 1930 Vergleichende Laut-, Stammbildungs- und Flexionslehre der Indogermanischen Sprachen. Grundrisse der vergleichenden Grammatik der indogermanischen Sprachen I, 1. Berlin: de Gruyter. Burks, Arthur W. 1963 Chance, cause, reason: An inquiry into the nature of scientific evidence. Chicago: University of Chicago Press. Burrow, Thomas 1955 The Sanskrit language. London: Faber and Faber. Bynon, Theodora 1977 Historical linguistics. Cambridge: Cambridge University Press. Calder, George (ed.) 1917 Auraicept na n-Eces: The scholar's primer. Edinburgh: John Grant. Christie, William M., Jr. 1976 Current progress in historical linguistics. Amsterdam: North Holland. Cohen, Morris R. —Ernest Nagel 1934 An introduction to logic and scientific method. New York: Harcourt, Brace. Cowgill, Warren 1963 "A search for universals in Indo-European diachronic morphology", in: Joseph H. Greenberg (ed.) 1963: 9 1 - 1 1 1 . Devoto, Giacomo 1968 Geschichte der Sprache Roms. Heidelberg: Carl Winter. Dietrich, Wolf 1973 Der periphrastische Verbalaspekt in den romanischen Sprachen (Beihefte zur Zeitschrift für romanische Philologie 140). Tübingen: Niemeyer. Egger, Emile 1854 Appolonius Dyscole: Essai sur l'histoire de theories grammaticales dans l'antiquite. Paris: Auguste Durand. Ernout, A. — Antoine Meillet 1959 Dictionnaire etymologique de la langue latine. Paris: Klinksieck. Esau, Helmut 1973 "Order of the elements in the German verb constellation", Linguistics 98: 20-40.

66

Irmengard Rauch

Friedrich, Johannes 1957 Extinct languages. New York: Philosophical Library. Frisk, Hjalmar 1960 Griechisches etymologisches Wörterbuch. Heidelberg: Winter. Gelb, Ignace Jay 1952 Α study of writing (Revised edition). Chicago: University of Chicago Press. Gray, Louis H. 1939 Foundations of language. New York: Macmillan. 1971 Introduction to Semitic comparative linguistics. Amsterdam: Philo Press. Greenberg, Joseph H. 1960 "A quantitative approach to the morphological typology of language", I J AL 26: 178-194. 1963 "Some universale of grammar with particular reference to the order of meaningful elements", in: Joseph H. Greenberg (ed.), 58 — 90. 1973 "The typological method", in: Thomas A. Sebeok et al. (eds.), 149-193. 1978 "Diachrony, synchrony and language universals", in: Joseph H. Greenberg et al. (eds.), 6 1 - 9 1 . Greenberg, Joseph H. (ed.) 1963 Universals of language. Cambridge, Mass.: MIT Press. Greenberg, Joseph H. et al. (eds.) 1978 Universals of human language. Stanford: Stanford University Press. Hall, Robert Α., Jr 1964 Introductory linguistics. New York: Chilton Books. Halle, Morris et al. (eds.) 1956 For Roman Jakobson: Essays on the occasion of his sixtieth birthday. The Hague: Mouton. Haugen, Einar 1950 First grammatical treatise: The earliest germanic phonology. (Language Monograph 25). Baltimore: LSA. 1969 "The analysis of linguistic borrowing", in: Roger Lass (ed.), 58 — 81. Hockett, Charles F. 1955 A manual of phonology (UAL Memoir 11). Baltimore: Indiana University Publications in Anthropology and Linguistics. Hoenigswald, Henry M. 1960 Language change and linguistic reconstruction. Chicago: Univ. of Chicago Press. 1963 "Are there universals of linguistic change?" in: Joseph H. Greenberg (ed.), 23-41. Hyman, Larry M. 1975 "On the change from SOV to SVO: Evidence from Niger-Congo", in: Charles N. Li (ed.), 113-147. Jakobson, Roman 1962 "Typological studies and their contribution to historical comparative linguistics", in: Roman Jakobson, 523 — 532. Jakobson, Roman 1962 Selected writings. Vol. I. The Hague: Mouton. Jakobson, Roman —C. Gunnar—M. Fant —Morris Halle 1951 Preliminaries to speech analysis. Cambridge, Mass.: The MIT Press.

Evidence of language change

67

Jeffers, Robert J. —Ilse Lehiste 1979 Principles and methods for historical linguistics. Cambridge, Mass.: The MIT Press. Jespersen, Otto 1907 John Hart's pronunciation of English (1569 and 1570). Heidelberg: Winter. Kahane, Henry —Renee Kahane—Roberta Ash 1979 "Linguistic evidence in historical reconstruction", in: Irmengard Rauch — Gerald F. Carr (eds.), 6 7 - 1 2 1 . Koutsoudas, Andreas — Gerald Sanders — Craig Noll 1974 "The application of phonological rules", Language 50: 1 —28. Kukenheim, L. 1932 Contributions ά l'histoire de la grammaire italienne, espagnole et fran^aise ά l'epoque de la Renaissance. Amsterdam: Ν. V. Noord-Hollandsche Uitgeversmaatschappij. Kuryfowicz, Jerzy 1964 "On the methods of internal reconstruction", in: H. G. Lunt (ed.), 9 — 36. Lass, Roger (ed.) 1969 Approaches to English historical linguistics. New York: Holt, Rinehart, Winston. Lehmann, Winfred P. 1955 Proto-Indo-European phonology. Austin: University of Texas Press and Linguistic Society of America. 1971 "On the rise of SOV patterns in New High German", in: K. G. Schweistal (ed.), 1 9 - 2 4 . 1973 a "A structural principle of language and its implications", Language 49: 47-66. 1973 b Historical linguistics: An introduction (2nd. edition, revised). New York: Holt, Rinehart, Winston. 1978 "Conclusion: Toward an understanding of the profound unity underlying language", in: Winfred P. Lehmann (ed.), 395—432. Lehmann, Winfred P. (ed.) 1978 Syntactic typology: Studies in the phenomenology of language. Austin: University of Texas Press. Li, Charles N. (ed.) 1975 Word order and word order change. Austin: University of Texas Press. Li, Charles N. —Sandra A. Thompson 1974 "Historical change of word order: A case study in Chinese and its implications", in: J. M. Anderson and C. Jones (eds.), 199 — 217. Lieb, Hans-Heinrich 1978 "Universale and linguistic explanation", in: Joseph H. Greenberg et al. (eds.), 157-202. Lightfoot, David W. 1979 Principles of diachronic syntax. Cambridge: Cambridge Univ. Press. 1983 "On reconstructing a proto-syntax", in: Irmengard Rauch and Gerald F. Carr (eds.), 128-142. Lunt, Horace G. (ed.) 1964 Proceedings of the Ninth International Congress of Linguists. The Hague: Mouton.

68

Irmengard Rauch

Malkiel, Yakov 1981 "Drift, slope, and slant: Background of, and variations upon, a Sapirian theme", Language 57: 535 — 570. Malmberg, Bertil 1964 New trends in linguistics (Trans. E. Carney). Stockholm: Holmiae. Martinet, Andre 1955 Economie des changements phonetiques. Bern: Francke. 1957 "Phonologie et 'laryngales'", Phonetica 1: 7 - 3 0 . Neu, Erich 1974 Der Anitta-Text (Studien zu den Bogazköy Texten, V. 18). Wiesbaden: Harrassowitz. Pedersen, Holger 1931 The discovery of language: Linguistic science in the nineteenth century. Bloomington: Indiana Univ. Press. Penzl, Herbert 1957 "The evidence for phonemic changes", in: Ernst Pulgram (ed.), 193 — 208. 1971 Lautsystem und Lautwandel in den althochdeutschen Dialekten. München: Hueber. 1972 Methoden der germanischen Linguistik. Tübingen: Niemeyer. Pokorny, Julius 1959 Indogermanisches etymologisches Wörterbuch. Bern: Francke. Polome, Edgar C. 1966 "The position of Illyrian and Venetic", in: Henrik Birnbaum—Jaan Puhvel (eds.), 5 9 - 7 6 . 1983 a "Lambert ten Kate's significance for the history of the Dutch language", Tijdschrift voor Nederlands en Afrikaans 2: 2 — 14. 1983 b "Netherlands contributions to the debate on language change: From Lambert ten Kate to Josef Vercoullie", in: Irmengard Rauch and Gerald F. Carr (eds.), 163-173. Pulgram, Ernst (ed.) 1957 Studies presented to Joshua Whatmough on his sixtieth birthday. The Hague: Mouton. Quirk, Randolph 1970 "Aspect and variant inflection in English verbs", Language 46: 300 — 311. Rauch, Irmengard 1978 "Distinguishing semiotics from linguistics and the position of language in both", in: R. W. Bailey-L. M a t e j k a - P . Steiner (eds.), 328-334. 1981 "Inversion, adjectival participle, and narrative effect in Old Saxon", Niederdeutsches Jahrbuch 104: 22 — 30. 1982 "Uses of the Germanic past perfect in epic backgrounding", Journal of Indo-European Studies 10: 301 — 314. Rauch, Irmengard — Gerald F. Carr (eds.) 1979 Linguistic method: Essays in honor of Herbert Penzl. The Hague: Mouton. 1983 Language change. Bloomington: Indiana University Press. Rauch, Irmengard — Charles Τ. Scott (eds.) 1967 Approaches in linguistic methodology. Madison: University of Wisconsin Press.

Evidence of language change

69

Rohlfs, Gerhard 1970 From Vulgar Latin to Old French: An introduction to the study of the Old French language (Trans. V. Almazan and L. McCarthy). Detroit: Wayne State Univ. Press. Ruhlen, Merrit 1975 A guide to the languages of the world. Stanford: Merritt Ruhlen. Samuels, M. L. 1972 Linguistic evolution: With special reference to English. London: Cambridge University Press. Sapir, Edward 1921 Language. New York: Harcourt, Brace. 1938 "Glottalized continuants in Navaho, Nootka, and Kwakiutl (with a note on Indo-European)", Language 14: 248 — 274. Saussure, Ferdinand de 1879 Memoire sur le systeme primitif des voyelles dans les langues indo-europeennes. Leipzig: Teubner. Schweistal, Κ. G. 1971 Grammatik, Kybernetik, Kommunikation: Festschrift für Alfred Hoppe. Bonn: Dümmler. Scott, Charles T. 1967 "On the dating of NE ee and ea spellings from ME e and §", in: Irmengard Rauch-Charles Τ. Scott (eds.), 7 3 - 7 9 . Sebeok, Thomas A. (ed.) 1974 Current trends in linguistics. Vol. 12. The Hague: Mouton. Sebeok, Thomas A. et al. (eds.) 1973 Current trends in linguistics. Vol. 2. The Hague: Mouton. Sharpe, William D. 1964 Isidore of Seville: The medieval writings (Transactions of the American Philosophical Society, New Series 54, No. 2). Philadelphia: The American Philosophical Society. Shipman, George Raymond 1953 The vowel phonemes of Meigret. Washington, D. C.: Georgetown University Press. Stephens, William N. 1968 Hypotheses and evidence. New York: Thomas Y. Crowell. Streitberg, Wilhelm 1908 Die gotische Bibel. Heidelberg: Winter. Sturtevant, Edgar H. 1942 The Indo-Hittite laryngeals. Baltimore: Linguistic Society of America. Swadesh, Morris 1971 The origin and diversification of language. Chicago: Aldine-Atherton. Szemerenyi, Oswald 1985 "Recent developments in Indo-European linguistics", Transactions of the Philological Society 1985: 1 - 7 1 . Toon, Thomas E. 1976 "The variationist analysis of Early Old English manuscript data", in: William M. Christie, Jr. (ed.), 7 1 - 8 1 .

70

Irmengard Rauch

Trager, George L. 1974 "Writing and writing systems", in: Thomas A. Sebeok (ed.), 373—496. Traugott, Elizabeth Closs 1972 A history of English syntax: A transformational approach to the history of English sentence structure. New York: Holt, Rinehart, Winston. Tschirch, Fritz 1969 1200 Jahre deutsche Sprache in synoptischen Bibeltexten (2nd edition). Berlin: de Gruyter. Ulimann, Stephen 1963 "Semantic universals", in: Joseph H. Greenberg (ed.), 172-207. van Ginneken, J. 1956 "Roman Jakobson pioneer of diachronic phonology", in: Morris Halle et al. (eds.), 574-581. Vennemann, Theo 1975 "An explanation of drift", in: Charles Ν. Li (ed.), 169-305. Visser, F.T. 1973 An historical syntax of the English language. Vol. Ill, 2. Leiden: Brill. Weinreich, Uriel 1964 Languages in contact: Findings and problems. The Hague: Mouton. Whiteley, W. H. 1967 "Loanwords in linguistic description: A case study from Tanzania, East Africa", in: Irmengard Rauch and Charles T. Scott (eds.), 125 — 143. Zeps, Valdis J. 1966 Languages of the world: A selective survey. Madison: College Printing and Typing Co.

The context of language change Joseph Salmons

0. Introduction This chapter reviews research on four points of intersection between language change and sociolinguistics:1 1) 2) 3) 4)

the sociolinguistic variable social factors in language change language contact change areal aspects of language change.

Historical linguists have increasingly utilized sociolinguistic insights to understand language change. 2 Even linguists long and intimately identified with mechanistic theories of language change now acknowledge that social factors play a crucial role in language change. P. Kiparsky, a key figure in transformational-generative (TG) historical linguistics, wrote (1980: 415): Language acquisition and language use cannot be idealized away from the theory of change as they can from formal grammar, nor smuggled in disguised as structure, as has happened in generative work. The theories of grammar, language acquisition, and language use have to be considered as interacting subsystems in the explanation of change.

R. D. King, another central TG historical linguist into the early 1970s, has recently written (1980: 414, cf. also King 1973) that a linguist "who rigidly separates the internal and external aspects of language history inevitably pays the ultimate price for his linguistic 'purity': what he says is a shallow imitation of linguistic reality". In short, it is widely agreed at present that social context and language change are inextricably connected.

72

Joseph Salmons

1. Sociolinguistic variation The modern study of language change in progress, pioneered by Labov and others, rests on the study of sociolinguistic variation. 3 What is a sociolinguistic variable? Sankoff (1972: 58; 1980) defines it broadly: "Whenever there are options open to a speaker, we can infer from his or her behavior an underlying set of probabilities". This notion is so broad as to cover variation which carries no social or stylistic meaning. Lavandera (1977: 16) calls for a distinction between "variables" which are elements of the language and the carriers of social and stylistic significance, and "variables" which are simply heuristic devices to group alternates and subjcct them to quantitative analysis.

While both should be studied, she makes clear that the restrictive sense, i. e., as a carrier of social and stylistic significance, is much more useful. Labov and Lavandera agree that two items must be "referentially identical" in order to constitute variables, a formulation rejected by Romaine (1984b: 413-415; 1982a: 3 2 - 3 3 ) as too vague. Romaine replaces this with the requirement that the items possess "constancy of cognitive or descriptive meaning". These issues come up once the study of variation moves beyond phonology and morphology. 4 Sociolinguistic variation has been found in most areas of language: 1) Phonological variables, including Labov's New York work. In that study (1972: 43 — 69), Labov found that sales clerks at more prestigious stores pronounce /r/ after vowels more frequently: 62% of Saks employees in contrast to 20% of Klein's employees (the lowest prestige group) used postvocalic /r/ in some or all instances. 2) Morphological variation, Romaine (1984 a) on djt deletion for example (that is, d ~ t —• 0/C *; past tense missed vs. mist). 3) Morphosyntactic or morpholexical variation, such as Sankoffs study of #we-deletion in Montreal French (1982). Que is widely used as a complementizer in Montreal and Sankoff was able to correlate use or deletion of que with social status. 4) Sociolinguistic variation in syntax is more problematic; Romaine (1984 b) finds that current sociolinguistic and syntactic theories are not adequately developed to explain syntactic variation yet. Romaine's greatest concern is probably with equivalence of meaning between two syntactically different utterances. 5 In 1971, Callary applied Labovian methods to study syntactic variation by socio-economic status, finding

The context of language change

73

for instance more transpositions (example b. below) among higher class speakers than lower class speakers in formal settings: a. She drinks coffee in the morning. b. In the morning she drinks coffee. The adverb anymore in midwestern American English currently shows clear dialectal syntactic differences and is in a state of flux, although the sociolinguistic ramifications are not so clear. Youmans (1986: 69) found that in Missouri anymore in sentence initial position ranges from "divided usage" (roughly equal numbers of speakers would and would not accept the sentence, cf. sentence a. below) to "rejected" (sentence b.): a. Anymore those are worthless. b. * Anymore do they sell those? While data may still be scarce, syntactic sociolinguistic variables may well exist. 5) The study of variation has now been initiated at the discourse level, cf. Dines (1980) and Lavandera (1984). Ohmann (1986) treats social class and discourse differences. Cf. also the issue of Linguistics devoted to variation and discourse edited by Dittmar (1987).6

2. Social factors and language change This section looks briefly at work in two areas: 1) social/demographic categories and language change, using gender and language as a more detailed example; 2) approaches to prestige and language change. Social correlates of linguistic change include ethnicity (e. g., Dorian 1980; Laferriere 1979; and LePage - Tabouret-Keller 1985), class (especially the works of Labov), social networks (Gal 1979, BortoniRicardo 1985) and gender, among others.7 Age has been used to create an apparent time depth in the study of language change in progress. That is, features found in the speech of older members of a community are taken to reflect an earlier stage while different features found in the speech of younger people are taken to be innovations. A cornerstone for much later work on gender and language was Lakoff's Language and woman's place (1975). The book's stated purpose was not to become a canon, but to act "as a goad to further research" (1975: 5), an aim certainly fulfilled.8 One facet of gender and language change lies in conscious attempts to rid language of sexism and to establish guidelines for non-discrim-

74

Joseph Salmons

inatory usage (cf. Frank —Anshen 1983: 107 — 114). Cannon and Roberson (1985: 33), working with dictionaries, conclude that this can indeed be accomplished: "conscious, quick change in vocabulary can be effected". Perhaps the most prominent diachronic question has been the "semantic degeneration" of nouns referring to women (cf. Schulz 1975 and Bebout 1984, as well as Eakins - Eakins 1978, Miller-Swift 1976, Smith 1985: 47 — 48). Many pairs of words ostensibly synonymous except for gender have developed negative (or less positive) connotations for the feminine forms: mister sir lord bachelor

mistress madam lady spinster

In German, parallel developments have taken place with several words for 'women, girl', with Dime 'girl' becoming 'prostitute' and Weib 'woman' becoming 'broad, bitch'. In some varieties of Spanish, madre 'mother' has developed negative and even obscene connotations for some speakers. The general tendency underlying such changes may be that any linguistic difference can be exploited in situations of social inequality, whether for sexist, racist, or classist ends. Rates of language shift may also vary by gender. Sole (1975) finds more rapid shift to English among Mexican-American women in Texas than among men, who tend both to use more Spanish and to codeswitch more often. Sole's conclusions for some Texas Chicanas are supported by other research, for example, Patella and Kuvelsky (1971) also working with Texas Spanish and Moore (1980) for slighter differences among Hill Country Texas Germans. Williamson and van Eerde (1980: 74), on the other hand, find gender a very minor factor in the maintenance of various European minority languages (Friulian, Rhaetoromansch, and Scots Gaelic). In a more general historical study, Shapiro (1985) traces the evolution of some key vocabulary of the women's movement through the 1960s: chauvinism, sexism, liberation, feminism. 9 Moving on to prestige and language change, prestige has long been seen as a major factor in linguistic change, e. g., Sturtevant (1917: 66). It was indeed central to the Neolinguistic concept of language change: The main factor in the triumph of a language or of a linguistic innovation (which is the same thing) is its prestige. This is not only military, political, or commercial; it is also, and much more, literary, artistic, religious, philosophical (Bonfante 1947: 357, cf. also Bärtoli 1925).

The context of language change

75

The classic American definition is perhaps Labov's statement (1972: 308) that it "must be defined in terms of the people using it and the situations in which it is used". Smith (1979: 113) works along these same lines, writing that "prestige" cannot be used interchangeably with "standard" in sociolinguistics, for the linguistic varieties that are socially advantageous (or stigmatized) for one group may not be for another. That is, the evaluative connotations of speech cannot be assessed independently of the people that use them.

The best solution in many cases may be simply to drop the word prestige and replace it with the many far more specific factors that make up prestige. Weinreich (1968: 77 — 80, originally published 1953) already gave a breakdown of such factors under the heading "relative status of languages", including "emotional involvement", "function in social advance", and "literary-cultural value". Gloy (1980: 366) also elaborates the notion of prestige into a set of criteria including greater frequency of occurrence. For additional discussion of prestige, especially work from Great Britain, cf. Hudson (1980: 191 - 2 1 4 ) . Kahane (1986) provides a detailed description of factors leading to the rise and fall of "prestige languages" in historical western European examples, from Greek onward. An early and classic example of prestige comes from Kloeke's 1927 investigation of Dutch dialects. In the thirteenth or fourteenth century, some Dutch dialects developed /u:/ into /y:/ and later — especially after 1585 — this front-rounded vowel became [ce:i] , spreading out from southwest Brabants. It came into Holland with Brabanders as a prestige feature among higher classes, while lower classes retained /y:/· 10 Kloeke's dialect map revealed that in three areas the vowel of huis 'house' reflected the newer form while muis 'mouse' still showed the older form. These words would seem likely to develop in parallel fashion: they differ only in the initial consonant, both are everyday items, etc. Kloeke (1927: 190 — 191) suggested that [ce:i] was so closely connected with prestige that 'mouse' remained unchanged because of its lower frequency of occurrence. 11 Bloomfield (1933: 328 — 330) takes a slightly different tack: the southerners, many of them officials, would have had occasion to talk about houses in their official capacities but not about mice, providing a model for the former and not for the latter. The autochthonous speakers then would have picked up the new diphthong for some words, but not have extrapolated to muis.

76

Joseph

Salmons

While more prestigious varieties are more likely to influence less prestigious ones, numerous counterexamples show how complex prestige is. Nadkarni (1975) finds that Konkani has taken over relative clause syntax from Kannada, although Kannada "does not enjoy any social prestige among the Saraswat Brahmins" (i.e., those Konkani speakers who use Kannada). Yet he says that Kannada is "functionally dominant". Irvine reports that some Wolof speakers regard "correct" grammar as a sign of lower class. Incorrect grammar, especially in the noun classes, or slurred speech "may show extremely high rank or it may show incompetence" (1982: 9). She also goes on to suggest that language change within the community appears to be in the direction of the perceived incorrect speech. Black speech has repeatedly influenced Anglo-American speech and continues to do so at present in lexicon and idiom, cf. Dillard (1973: 186 — 228). Even syntactic features closely identified with Black English may have been borrowed by southern Anglos. Feagin (1984: 269) finds it probable for instance that invariant be (/ put her up till I be sure) has passed into Anglo speech, where it occurs only occasionally, from Black speech as well as perhaps the remote present perfect uses of been (such as I been knowin' him all his life.) Borrowing frequently goes in both directions. Spanish, in many ways a less prestigious language than English in Texas, has loaned many words into English. Atwood (1962: 125 — 129) for instance catalogues over twenty. The relative prestige of varieties can also change over time. Mainland North Frisian has long been considered the language of da latje mään 'the little man'. Walker (1979: 142) writes that when North Frisian speakers strove to establish themselves in the German middle class, High German was the clear language of prestige. The correlation of Frisian with the lower class and High German with the middle class thus encouraged language shift. Today however enough North Frisians have established themselves in the middle class that the association of Frisian with da latje mään is breaking down. Many families have become aware of their Frisian heritage and begun to value it overtly, although this is not intrinsically connected with language maintenance. Other examples of change in prestige can be more dramatic: as Cooper (1982: 11, 25; following Macnamara) put it, the Irish adopted English "in spite of Irish antipathy towards the English". 12 Again, as with North Frisian, prestige has shifted back to favor the threatened language, in the case of Irish with considerable institutional support, but

The context of language change

77

without great impact on language maintenance and shift. For the United States, Fishman (1984 and 1985) has dealt at length with the "ethnic revival" and language shift in American immigrant languages, analyzing census data and institutional uses of immigrant languages since I960. 13 Taboo represents another example of change in social structure leading to linguistic change. 14 While prestige depends on vertical social structure, taboo tends to rely on religion, superstition, etc. When basic attitudes and values change, taboos change as well. Spalding (1973), for instance, traces the shift from strict Christian values and taboos to Enlightenment era values and taboos. Since the late Middle Ages Christianity has gradually lost its central role in European society, although its strictest taboos remain stigmatized in many contexts, e. g., English goddamn. A newer set of values has grown up since the Enlightenment, a set including more democratic and egalitarian values. Eventually, Spalding argues, these values brought new taboos to reflect attempts to make society more egalitarian. For instance, "unattractive professions" (i. e., low paying or low prestige jobs) have often been renamed to reflect a more egalitarian social order: Bauer ('farmer', sometimes with the connotation "peasant, dirt farmer") became Landwirt ('someone who manages the land'). Socially disadvantaged groups were renamed: poor people became underprivileged; Fremdarbeiter ('foreign workers') became Gastarbeiter ('guest workers'), etc. Spalding sees this shift away from strict religious values toward more democratization as an attempt by society to ease the burden of these groups. The distinction between overtly claimed values and actual societal values may however be overlooked here. A taboo mirrors what society says should be central values. Germans may feel better about how Gastarbeiter are treated than about how the Fremdarbeiter were, but the great West German need for foreign labor in the 1960s motivated a certain rationalization of the situation. That this linguistic change neither signaled nor reflected any fundamental change in popular German attitudes and values is apparent from the increasing tensions between many Germans and their guests. When treating questions of prestige one must also mention the related issues of standardization, literacy, and prescription, factors which frequently slow the rate of linguistic change. They are surveyed in Zengel (1962), Bright (1976), and most extensively in Milroy and Milroy (1985).

78

Joseph Salmons

3. Language contact change Uriel Weinreich's Languages in contact represented the beginning of a new era in language contact studies. Within a taxonomic structuralist work, Weinreich introduced a tremendous amount of socio-culturally relevant information, formulating questions that have reappeared throughout the literature ever since. More recent guides to language contact include Clyne's state of the art report (1975); Jongen et al. 1983; Meid and Heller 1981 on language contact and language change; Neide 1983; Trudgill 1986; Ureland 1979; Ureland 1982; Ureland and Clarkson 1984 on Scandinavian language contact; Amastae and EliasOlivares 1982 on Spanish in the United States; and Schach 1980 on various languages in the Great Plains. The most accurate general statement on language contact and language change may be SilvaCorvalän's (1986: 602) that language contact leads to an acceleration of language change. Language contact studies have evolved in several directions. Ureland (1979) traces among them: 1) the Indo-Europeanist tradition, from the Sprachmischung hypothesis of Paul and Schuchardt to a newer substrate tradition including Pisani and Kuhn. Substratal influence remains controversial (cf. Neumann 1971). 2) Sprachbund work (especially of the Prague School), including Jakobson and Trubetzkoy and more recently Lewy, Wagner, and Haarmann, tracing features across language and language family boundaries. 3) American structuralism, including Sapir and Haugen. Ureland associates these scholars with the attempt to establish a hierarchy of borrowed elements. Other groups include ethnolinguistic, typological (as in Lehmann's recent work, e.g., 1973), and pidgin and Creole research. More recently, Bickerton (1981,1984) seeks the direction of language contact change in an innate structure of the brain. Certain structures, for example, the relationship between causal and non-causal or specific and non-specific, are learned easily and early by children and are distinctions made somehow in almost all languages. Bickerton claims that certain kinds of contact situations lead to a stripping away of less essential rules, taking us back to the most fundamental and innate distinctions. Therefore, he claims, Creoles tend to mark these key features in similar ways, even though he denies any genetic relationship among the Creoles he treats. A more general view of language contact change (as opposed to creolization as in his book) would thus be built

The context of language change

79

around language contact bringing out features of the "bioprogram", a process which is not necessarily steady (Bickerton 1981: 291). Bickerton's views, with such profound implications for language contact and language change, have naturally brought considerable response. Cziko (1986) seeks support for the bioprogram in language acquisition studies, including a recent bibliography on the bioprogram hypothesis. See also the important reviews of Muysken (1983) and Mühlhäusler (1984). The search for universale has also been applied to language contact. Moravcsik (1978) surveys language contact research, distilling a set of several rules for universal constraints on borrowing. Rule 1, for instance, states that words are borrowed before any other feature (syntactic, morphological, etc.). Such rules form a plausible set of working hypotheses for language contact study, but only restricted insight can be gained from purely language-internal criteria. Precisely in the study of languages and cultures in contact, context becomes a central consideration. Moravcsik's set of rules attempts to transcend context, a tempting move given how little is understood about the impact of social setting on language contact change. In spite of the publication of numerous basic works and the development of several distinct approaches, fundamental ambiguities remain for language contact studies at even a definitional level. Diglossia, for example, has undergone constant redefinition since Ferguson's 1959 article on the subject. Ferguson originally used the term to describe language contact settings with the following characteristics: 1) a stable bilingual situation; 2) a "superimposed" variety (high or H), more codified and associated with writing, formal usage, etc., contrasted with a home variety used informally (low or L). Ferguson's examples all involved closely related varieties: Classical versus Colloquial Arabic, Swiss versus High German, etc. Later, Fishman (e.g. 1972: 91 — 100) extended the term (as had Gumperz and Rubin) to situations where two either related or unrelated languages are used systematically by a population according to situation. Fishman then extends the term far beyond this to describe "diglossia without bilingualism", i. e., situations where two distinct speech communities exist, two languages socially stratified without bilingualism. The elite in pre-World War I Europe often spoke a different language from that of the "masses", Fishman says, thus constituting an elite Η variety never used by lower classes and an L variety not used by the elite. Such a definition bears little resemblance to Ferguson's.

80

Joseph Salmons

Britto (1986), in the most complete and up-to-date treatment of diglossia (including extensive bibliography), has synthesized Ferguson's and Fishman's definitions into the following system:15 1) pseudo-diglossia, essentially Fishman's diglossia without bilingualism; 2) user-oriented diglossia, in which only some members of the society use only Η and others use both Η and L varieties; 3) use-oriented diglossia, where everyone uses both, i. e., by situation. The term "triglossia" has also found its way into the literature. Knowles (1980) uses it to describe the linguistic situation in Luxembourg. Certainly Letzebuergesch, German, and French are spoken by most members of society in distinct situations, but Letzebuergesch is difficult to characterize as a classic L variety. It is used in the House of Commons, has standardized orthography, and is used in newspapers, etc. The functional divisions here run much differently from in most diglossic settings. Abdulaziz (1972) makes a clearer case for a three-way social hierarchy in Tanzania with English-Swahili-local languages (90% of them Bantu). A triglossic division of function exists with English as H, Swahili taking a middle ground, and local languages serving as L. This is reflected in three distinct levels of standardization, range of vocabulary, etc. He notes (1972: 212) that shift is underway from a three level system to a two level system — with English and Swahili sharing Η functions and local languages as L — as Swahili becomes more prestigious and more widely used in official domains. Other important terminological problems are "pidginization" and "creolization", which have been used to describe a broad range of language contact processes. Mühlhäusler (1986: 6 — 12) divides the myriad definitions of Creole into three major groups: 1) mixed languages evolved from cultural mixture; 2) pidgins that have developed native speakers; 3) Bickerton's bioprogram as discussed above. The first group is vague enough to include various contact varieties that often fall under other rubrics. "Creoloid", for instance, exists as a name for partially creolized languages and has found its way into much recent literature on language contact (e.g., Trudgill 1983: 102 — 107, 127; Mühlhäusler 1986). "Koine" is used by Trudgill to indicate languages that have undergone mixing, leveling, and simplification (1986: 106—110).16 The second group is rejected by Mühlhäusler since boundaries between pidgin and Creole or between first and second languages cannot be drawn clearly. Finally, by Bickerton's definition a creole is a language that

The context of language change

81

1) Arose out of a prior pidgin which had not existed for more than a generation. 2) Arose in a population where not more than 20 percent were native speakers of the dominant language and where the remaining 80 percent was composed of diverse language groups (1981: 4).

The first point excludes many languages that virtually all other creolists would accept without question, including Tok Pisin. The second would exclude Reunion Creole, on account of the size of its French element. Bickerton's examples are highly stratified social settings with European superstrate languages. Other pidgins and Creoles are virtually ignored in this view although they are numerous, cf. Heine (1970 and 1973) for African examples and Kaye (1985) for sources on Arabic pidgins and Creoles. Görlach (1986: 332 — 333) notes the following traditional criteria: 1) the acquisition of native speakers, 2) maintaining the linguistic stability of the pidgin, 3) simplification, and 4) unintelligibility to speakers of the related standard variety. 17 Things become even more complex with the application of these terms to historical or prehistorical situations. Domingue (1977), Bailey and Maroldt (1977), and Görlach (1986) all consider whether Middle English was creolized. Proto-Germanic has also intermittently been taken to be a Creole since Feist's work in the 1930s. Leveling and simplification have not reached nearly the levels in Middle English that one finds in classic pidgin and Creole languages, maintaining for example far more verbal morphology than Creoles. Proto-Germanic has maintained much of the normally reconstructed Indo-European morphology and has even been regarded as one of the most archaic Indo-European languages (Lehmann 1953; Polome 1982, who also deals with the question of creolization in early Romance and IndoEuropean). Ultimately, virtually everyone invents his or her own definition or simply acknowledges that the situation is terribly confused. 18 As Mühlhäusler stresses repeatedly, pidgins and Creoles are extremely "leaky" and "open" systems that defy application of theories and models designed to describe closed systems.

4. Areal linguistics and language change Chambers (1982: 1) writes that three important areas have emerged for understanding the "geographical dispersions of linguistic elements":

82

Joseph

Salmons

1) urban dialectology, 2) variation studies, 3) application of insights from human geography to language. This section illustrates points two and three briefly. Chambers (1982) introduces the use of variable rules to geographical variation, taking a set of data from the Survey of English dialects which appears completely chaotic: consonant cluster simplification in northern England. Chambers constructs four variable rules to describe the phonological processes of simplification: 19 (R2) C > (0)/C *C. (R3) C > (0)/C * a C, bV, where a > b. (R4) C > (0)/Son **. (R5) C > (0)/cSon, d Obs **, where c > d. These rules can be arranged hierarchically or implicationally, or they may be written as one rule: (R6) C > (0)/ c Son, d Obs_* a C, b V, where a > b > c > d. When the application of these rules is complete, the previously chaotic data can be clearly mapped (cf. Map 1.) and a chart can illustrate the hierarchy of environments: 20 Rule _C _V Son Obs. R2 a 0 0 0 R3 a b 0 0 R4 a b c 0 R5 a b c d Thus the use of variable rules can clarify geographical relationships among linguistic variants. Chambers also observes (1982: 16) that isoglosses become irrelevant: areal relationships become visible without artificial divisions based on features. In addition to Chambers' work, numerous collections have recently been published on areal linguistics and linguistic variation, including Shores and Hines (1977); Allen and Linn (1986); and Montgomery and Bailey (1986). Methods have also been adopted and adapted from human geography to explain the spread of linguistic features over space. Expansion or spread of linguistic features tends to take place in one of two basic ways, according to Trudgill (1974, reprinted in Trudgill 1983: 52 — 87). First, in the "neighborhood effect" a given feature advances, as it were, from house to house. Second, a feature may jump from one urban center to the next, skipping over the rural areas lying in between. The second pattern is apparently found for uvular r in western Europe

The context of language change

Map 1.

83

A composite view of the geographical dispersion of (R2) —(R5) in northern England (Chambers 1982). Apart from two isolated lects in Yorkshire, the lects with the less general rule tend to be bounded by lects with the next most general rule in the set.

84

Joseph Salmons

(cf. Map 2.).21 Trudgill (1974: 219-221) follows the change from Paris across central European centers to Scandinavia. From some urban centers, uvular r has spread into general usage throughout a country, e. g., most of France and Denmark. McDavid (1948, reprinted in McDavid 1980) correlates history and linguistic change in South Carolina, tracing the use or deletion of postvocalic r to the geography of the plantation system in the ante-bellum Southeast. Where plantations were profitable, i. e., where proper soil conditions were present, a culture developed much different from other areas. For example, the non-plantation areas were settled more heavily by non-slave-holding small farmers. Areas not involved in plantation farming generally showed constriction, which then spread inland aided by the prestige of the "old plantation caste". 22 One view of language change was built around areal data. Beginning in the 1920s, the Neolinguistica of Italy developed a school of linguistics in direct opposition to the Neogrammarians and their mechanistic conception of language. Bärtoli (1925, 1945), Bonfante (1947 and elsewhere) and others evolved a set of "norms", i.e., tendencies, in contrast to the "laws" of the Neogrammarians. Among their norms are these (from Hall's summary 1946: 274, building also on Bonfante's reply to Hall, 1947): 1) A stage of linguistic development attested earlier is usually older. 2) The earlier stage is usually preserved in more isolated areas. 3) Lateral areas usually show older stages of development, unless the central area is the more isolated. 4) A larger area tends to preserve an earlier stage, unless the smaller area is more isolated or consists of lateral areas. 5) The earlier stage is usually preserved in an area settled later. 6) If a stage of development is obsolete or dying out, that stage is usually older. These principles, while not always warmly received (cf. Hall 1946), represent general trends of areal linguistic change, as still often acknowledged today (e.g., by Chambers - Trudgill 1980: 181).23

5. Conclusion The variety of issues and approaches discussed or mentioned here makes a straightforward summary impossible. One does, however, find a red thread running through much of the above discussion. As social

The context of language change

Map 2.

Distribution of uvular /r/ (Chambers-Trudgill 1980, after Trudgill 1974).

86

Joseph Salmons

context is being increasingly integrated into every facet of the study of language change, many basic concepts, terms, and points of method have lacked coherent and generally accepted definitions; only recently have some of these problems been addressed in detail. For example, Labov has claimed that "social values are attributed to linguistic rules only when there is variation" (1972: 297), raising the sociolinguistic variable to a central position in sociolinguistics and historical linguistics. Yet only more recently does one find critical discussions of what constitutes a sociolinguistic variable. Ways of analyzing or describing variation depend upon answers to this question. Likewise, the definition and ramifications of "prestige" have not yet been explicated and the term often retains a peculiarly Western, industrial/post-industrial tint. Descriptive terms for key language contact situations — like diglossia — and products of language contact situations — like Creoles — continue to be used vaguely and with numerous distinct meanings. On the other hand, tremendous progress is clearly being made in integrating social context into the study of language change as both new data and theoretically-oriented works appear.

Notes 1. The massive recent literature on these questions obviously precludes more than a passing reference to even major works. Thanks for comments and suggestions especially to Marianne DiPaolo (University of Utah) and Margie Berns (Purdue University). 2. For a survey of the historical relationship between historical linguistics and sociolinguistics, cf. Lehmann (1981). 3. Variation does not, however, always signal language change in progress, ingjin', h versus Α-less dialects, and other features represent stable variables in English. Cf. especially Romaine (1984 a). 4. The question of discreteness of levels in sociolinguistic variation has been treated by Kerswill (1987), working with three: lexical, phonological, and "connected speech processes". 5. See also Garcia (1985), who regards variationalist studies as a "theoretical blind alley". 6. Ryan and Giles (1982) deal with attitudes toward linguistic variation. 7. The anthology of Giles and Scherer (1979) covers various social markers, though not specifically with respect to language change. 8. General bibliographies on language and gender include Key (1975); Frank—Anshen (1983); and Smith (1985). 9. Henley (1987) treats sexism and language change, but was not yet available when this manuscript was completed.

The context of language change

87

10. /y:/ users were in turn prestige speakers when they moved form urban Amsterdam to the more rural areas of the Wadden Islands, Friesland, and the land around the Zuiderzee. Here local people took over the /y:/ by that time out of fashion farther south. For the sociolinguistic situation in the seventeenth century, cf. Polome (1985). 11. Kloeke's explanation was not universally accepted and a debate raged in the Low Countries into the 1930s. Cf. van Loey (1964) for a summary. 12. Recent work on the sociolinguistic situation of Irish can be found in Edwards (1984). 13. A "flip-flop" of prestige at a micro-level — in the pronunciation of words such as tune ([tun] versus [tjun]) — is treated by Pitts (1986). Pitts shows that the southern and northern dialects in the United States have reversed their evaluations, with young southerners now favoring the second form. 14. Maledicta brings forth a steady stream of heavily data-oriented work on linguistic taboo, including occasional pieces on historical developments. Havers (1946) reviews literature on Indo-European linguistic taboo. Cf. also Jespersen's classic article (1970). 15. Another recent work on diglossia, Costello (1986), traces the demise of a diglossic situation with Pennsylvania German. 16. The following chart, developed by Mühlhäusler (1978) and discussed by Trudgill (1983: 110) and Görlach (1986: 333-334), attempts to outline the relationships among various language contact varieties: lingua franca pre-pidgin pidgin Creole post-creole creoloid

reduction + + + — — —

simplification + + + + — +

stability + + + — +

unintelligibility — — + + — —

17. For another side of this terminological debate, cf. Thomason (1983) for a discussion of whether Chinook Jargon constitutes a pidgin. 18. Other sociolinguistic terms also present definitional difficulties, including "speech community" (cf. Romaine 1982 b). 19. Superscript indicates the probability of occurrence, so that an option marked a has a greater likelihood of occurrence than b: a > b. 20. This hierarchy is also supported by data from New York and Detroit studies of Black English consonant cluster simplification. 21. Some question the existence of apical r in Germanic dialects and thus reject a spread of uvular r from French to German, cf. Runge (1973) and Howell (1985). 22. See also Rickford (1986) on socio-historical evidence on the development of some Black English features. 23. See Campbell (1985) for recent general discussion of areal linguistic implications for historical linguistics.

References Abdulaziz Mkilifi, Μ. H. 1972 "Triglossia and Swahili-English bilingualism in Tanzania", Language in Society 1: 1 9 7 - 2 1 3 .

88

Joseph Salmons

Allen, Harold — Michael Linn (eds.) 1986 Dialect and language variation. Orlando, Fla.: Academic Press. Althaus, H. P. et al. (eds.) 1980 Lexikon der germanistischen Linguistik (2nd edition). Tübingen: Niemeyer. Amastae, Jon —Lucia Elias-Olivares (eds.) 1982 Spanish in the United States: Sociolinguistic aspects. Cambridge: Cambridge University Press. Atwood, E. Bagby 1962 The regional vocabulary of Texas. Austin, Tex.: University of Texas Press. Bärtoli, Matteo 1925 Introduzione alia Neolinguistica: Principi — Scopi — Metodi. Biblioteca dell' "Archivum Romanicum", Series II, vol. 12. Geneve: L. S. Olschki. Bailey, C. J . - K a r l Maroldt 1977 "The French lineage of English", in: Jürgen Meisel (ed.), 21 — 53. Bailey, C. J . - R . Shuy (eds.) 1972 New ways of analyzing variation in English. Washington, D. C.: Georgetown University Press. Baumann, Richard — Joel Sherzer (eds.) 1982 Case studies in the ethnography of speaking. Austin, Tex.: University of Texas Press. Bebout, Linda 1984 "Asymmetries in male-female word pairs", American Speech 59: 13 — 30. Bickerton, Derek 1984 a Roots of language. Ann Arbor, Mich.: Karoma Press. 1984 b "The language bioprogram hypothesis", Behavioral and Brain Sciences 7: 173-188. Bloomfield, Leonard 1933 Language. New York: Holt. Blount, Ben G. — Mary Sanchez (eds.) 1977 Sociocultural dimensions of language change. New York: Academic Press. Bonfante, Giuliano 1947 "The Neolinguistic position", Language 23: 344—375. Bortoni-Ricardo, Stella Maris 1985 The urbanization of rural dialect speakers: A sociolinguistic study of Brazil (Cambridge Studies in Linguistics, supplementary volume). Cambridge: Cambridge University Press. Bright, William 1976 "Sociolinguistic variation and language change", in: Anwar Dil (ed.), Variation and change in language: Essays by William Bright. Stanford: California University Press, 37 — 56. (Reprinted from the Proceedings of the Ninth International Congress of Linguists. (1962) Ed. H. Lunt. The Hague: Mouton 1964). Britto, Francis 1986 Diglossia: A study of the theory with application to Tamil. Washington, D. C.: Georgetown University Press. Callary, Robert 1971 "Syntactic correlates of social stratification", Diss. Louisiana State University.

The context of language change

89

Campbell, Lyle 1985 "Areal linguistics and its implications for historical linguistics", in: Jacek Fisiak (ed.), 2 5 - 5 6 . Cannon, Garland —Susan Roberson 1985 "Sexism in present-day English: Is it diminishing?" Word 36: 23 — 35. Chambers, J. K. 1982 "Geolinguistics of a variable rule", Discussion Papers in Geolinguistics 5. North Staffordshire: North Staffordshire Polytechnic, Dept. of Geography and Sociology. Chambers, J. K. —Peter J. Trudgill 1980 Dialectology. (Cambridge Textbooks in Linguistics.) Cambridge: Cambridge University Press. Clyne, Michael 1975 Forschungsbericht Sprachkontakt. Kronberg: Skriptor Verlag. Cooper, Robert 1982 "A framework for the study of language spread", in: Robert Cooper (ed.), 5-36. Cooper, Robert (ed.) 1982 Language spread: Studies in diffusion and social change. Bloomington, Indiana: University of Indiana Press. Costello, John R. 1986 "Diglossia at twilight: German and Pennsylvania 'Dutch' in the mid-19th century", in: Werner Enninger (ed.), 1—24. Cziko, Gary A. 1986 "Testing the language bioprogram hypothesis: A review of children's acquisition studies", Language 62: 878 — 898. Dahlstedt, Karl-Hampus (ed.) 1973 The Nordic languages and modern linguistics. Vol. 2. Stockholm: Almqvist and Wiksell. Dillard, J. R. 1973 Black English: Its history and usage in the United States. New York: Vintage. Dines, E. R. 1980 "Variation in discourse — 'and stuff like that'", Language in Society 9: 13-31. Dittmar, Norbert (ed.) 1987 Variation and discourse. Special issue of Linguistics 25, No. 1. Domingue, Nicole Z. 1977 "Middle English: Another Creole?" Journal of Creole Studies 1: 8 9 - 1 0 8 . Dorian, Nancy 1980 "Language lag as an ethnic marker", Language in Society 9: 33 — 42. Eakins, B. — G. Eakins 1978 Sex differences in human communication. Boston: Houghton-Mifflin. Edwards, John 1984 "Irish and English in Ireland", in: Peter Trudgill (ed.), 4 8 0 - 4 9 8 . Enninger, Werner (ed.) 1986 Studies in the languages and verbal behavior of the Pennsylvania Germans (Zeitschrift für Dialektologie und Linguistik, Beiheft 51) Wiesbaden: Fr. Steiner.

90

Joseph Salmons

Faarlund, Jan Terje (ed.) 1985 Germanic languages: Papers from a symposium at the University of Chicago. Bloomington, Indiana: Indiana University Linguistics Club. Feagin, Crawford 1984 "Southern White in the English language community", in: Harold Allen — Michael Linn (eds.), 259-283. Ferguson, Charles 1959 "Diglossia", Word 15: 325-340. Fishman, Joshua 1971 Sociolinguistics: A brief introduction. Rowley, Mass., Newbury House. 1972 The sociology of language: An interdisciplinary social science aproach to language in society. Rowley, Mass.: Newbury House. 1984 "Mother tongue claiming in the United States since 1960: Trends and correlates related to the 'revival of ethnicity'", International Journal of the Sociology of Language 50: 21—99. 1985 The rise and fall of the ethnic revival. Berlin: Mouton. Fisiak, Jacek (ed.) 1985 Papers from the Sixth International Conference on Historical Linguistics (Current issues in Linguistic Theory 34) Amsterdam: John Benjamins. Frank, Francine — Frank Anshen 1983 Language and the sexes. Albany: SUNY Press. Gal, Susan 1979 Language shift: Social determinants of linguistic change in bilingual Austria. New York: Academic Press. Garcia, Erica C. 1985 "Shifting variation", Lingua 67: 189-224. Gloy, Klaus 1980 "Sprachnormen", in: H. P. Althaus et al. (eds.), 363-367. Görlach, Manfred 1986 "Middle English — a creole?" in: Dieter Kastovsky—Aleksander Szwedek (eds.), 329-344. Greenberg, Joseph (ed.) 1978 Universals of human language. Vol. 1. Stanford: Stanford University Press. Hall, Robert A. 1946 "Bartoli's 'Neolinguistica'Language 22: 2 7 3 - 2 8 3 . Haugen, Einar 1950 "The analysis of linguistic borrowing", Language 26: 210 — 231. Havers, Wilhelm 1946 Neuere Literatur zum Sprachtabu (Abhandlungen der Akademie der Wissenschaften Wien. Philosophisch-historische Klasse. Sitzungsbericht, Bd. 223, Heft 5) Vienna: Rudolf Rohrer. Heine, Bernd 1970 Status and use of African lingua francos. Munich: Weltforum. 1973 Pidgin-Sprachen im Bantu-Bereich (Kölner Beiträge zur Afrikanistik 3). Berlin: Reimer. Henley, Nancy M. 1987 "This new species that seeks a new language: On sexism in language and language change", in: Joyce Penfield (ed.), 3 — 27.

The context of language change

91

Herzog, Marvin et al. (eds.) 1980 The field of Yiddish. 4th collection. Philadelphia: Institute for the Study of Human Issues. Howell, Robert 1985 "On the origin of uvular r in the Germanic languages", in: Jan Terje Faarlund (ed.), 1 - 9 . Hudson, Richard A. 1980 Sociolinguistics (Cambridge Textbooks in Linguistics). Cambridge: Cambridge University Press. Irvine, Judith 1982 "Wolof speech styles and social status", in: Richard Baumann —Joel Sherzer (eds.), 2 - 1 3 . James, Jennifer 1972 "Two domains of streetwalker argot", Anthropological Linguistics 14: 172-181. Jespersen, Otto 1929 "Veiled Language", Society for Pure English, Tract 33. (Reprinted in Linguistica. College Park, Md.: McGrath, 1970: 4 0 9 - 4 1 9 . ) Jongen, Rene —Sabine DeKnop —Peter Neide — Marie-Paule Quix (eds.) 1983 Mehrsprachigkeit und Gesellschaft (Linguistische Arbeiten 134). Tübingen: Niemeyer. Kahane, Henry 1986 "A typology of the prestige language", Language 62: 495 — 508. Kastrovsky, Dieter—Aleksander Szwedek (eds.) 1986 Linguistics across historical and geographical boundaries. In honor of Jacek Fisiak on the occasion of his fiftieth birthday. Vol. I. The Hague: Mouton. Kaye, Alan S. 1985 "On the importance of pidgins and Creoles for historical linguistics" (review article), Diachronica 2: 201 — 230. Kerswill, Paul E. 1987 "Levels of linguistic variation in Durham", Journal of Linguistics 23: 25 — 49. Key, Mary Ritchie 1975 Male/female language, with a comprehensive bibliography. Metuchen, N. J.: Scarecrow Press. King, Robert D. 1973 "Integrating linguistic change", in: Karl-Hampus Dahlstedt (ed.), 47 — 69. 1980 "The history of final devoicing in Yiddish", in: Marvin Herzog et al. (eds.), 3 7 1 - 4 3 0 . Kiparsky, Paul 1980 "Concluding statement", in: Traugott et al. (eds.), 4 1 0 - 4 1 7 . Kloeke, G. G. 1927 De Hollandsche expansie in de zestiende en zeventiende eeuw en haar weerspiegeling in de hedendaagsche Nederlandsche dialecten. 's Gravenhage: M. Nijhoff. Knowles, Jane 1980 "Multilingualism in Luxembourg", in: Peter Neide (ed.), 2 4 3 - 2 5 3 .

92

Joseph Salmons

Labov, William 1966 The social stratification of English in New York City. Washington: Center for Applied Linguistics. 1970 "The study of language in its social context", Studium Generale 23: 66 — 84. 1972 Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press. 1973 "The social setting of linguistic change", in: Thomas Sebeok (ed.), 195 — 251. 1980 "The social origins of sound change", in: W. Labov (ed.), 251 —265. Labov, William (ed.) 1980 Locating language in time and space. New York: Academic Press. Laferriere, Martha 1979 "Ethnicity in phonological variation and change", Language 55: 603 — 617. Lakoff 1975 Language and woman's place. New Yoek: Harper and Row. Lavandera, Beatriz 1978 "Where does the sociolinguistic variable stop?" Language in Society 7: 171 — 183. Originally appeared as Working Papers in Sociolinguistics 40. Austin, Tex.: Southwest Educational Development Laboratory, 1977. 1984 Variaciön y significado. Buenos Aires: Hachette. Lehmann, W. P. 1953 "The conservatism of Germanic phonology", Journal of English and Germanic Philology 52: 140-152. 1973 "A structural principle of language and its implications", Language 49: 47-66. 1981 "Historical linguistics and sociolinguistics", International Journal of the Sociology of Language 31: 11—27. LePage, Robert — Andree Tabouret-Keller 1985 Acts of identity: Creole-based approaches to language and ethnicity. Cambridge: Cambridge University Press. Lewy, Ernst 1959/1960 "In atlantischer Landschaft. Spuren unerklärter Sprachbeziehungen im westeuropäischen Raum", Sprachforum 3: 219 — 221. McDavid, Raven 1948 "Postvocalic /r/ in South Carolina: Α social analysis", American Speech 23: 194-203. 1979 "Sociolinguistic Studies", review of Labov 1972, American Speech 54: 291-304. 1980 Varieties of American English. Stanford, Calif.: Stanford University Press. Macnamara, John 1971 "Successes and failures in the movement for the restoration of Irish", in: J. R u b i n - B . Jernudd (eds.), 6 5 - 9 4 . Meid, Wolfgang — Karin Heller (eds.) 1981 Sprachkontakt als Ursache von Veränderungen der Sprach- und Bewußtseinsstruktur. Innsbruck: Innsbrucker Beiträge zur Sprachwissenschaft der Universität Innsbruck 34.

The context of language change

93

Meisel, Jürgen (ed.) 1977 Pidgins — Creoles — languages in contact. Tübingen: Narr. Miller, C . - K . Swift 1976 Words and women: New language in new times. Garden City, Ν. Y.: Doubleday Anchor. Milroy, James —Lesley Milroy 1985 Authority in language: Investigating language prescription and standardisation. London: Routledge, Kegan Paul. Montgomery, Michael—Guy Bailey (eds.) 1986 Language variety in the South: Perspectives in black and white. University, Ala.: University of Alabama Press. Moore, Barbara J. R. 1980 "A sociolinguistic longitudinal study (1969 — 1979) of a Texas German community, including curricular recommendations", Diss. University of Texas, Austin 1980. Moravcsik, Edith 1978 "Language contact", in: Joseph Greenberg (ed.), 93 — 122. Mühlhäusler, Peter 1978 Pidginization and simplification of languages. Canberra: Pacific Linguistics. 1984 "Roots of language?" Review article on Bickerton 1982. Folia Linguistica 18: 2 6 3 - 2 7 7 . 1986 Pidgin and creole linguistics (Language in Society 11). Oxford: Blackwell. Muysken, Peter 1983 Review of Bickerton 1982, Language 59: 8 8 4 - 8 9 2 . Nadkarni, M.V. 1975 "Bilingualism and syntactic change", Language 51: 672 — 683. Neide, Peter H. (ed.) 1980 Sprachkontakt und Sprachkonflikt. Wiesbaden: Steiner. Neide, Peter Η. (ed.) 1983 Gegenwärtige Tendenzen der Kontaktlinguistik. Bonn: Dümmler. Neumann, Günter 1971 "Substrate im Germanischen?" (Nachrichten der Akademie der Wissenschaften in Göttingen, I, Philologisch-historische Klasse, 4). Göttingen: Vandenhoeck & Ruprecht. Ohmann, Richard 1986 "Reflections on class and language", in: Harold Allen —Michael Linn (eds.), 2 8 4 - 3 0 3 . Patella, Victoria — William Kuvelsky 1973 "Situational variations in language patterns", Social Science Quarterly 53: 855-864. Penfield, Joyce (ed.) 1987 Women and language in transition. Albany: SUNY Press. Pitts, Ann 1986 "Flip-flop prestige in American tune, duke, news", American Speech 61: 130-138. Polome, Edgar C. 1982 Language, society, and paleoculture. Stanford, Calif.: Stanford University Press.

94

Joseph Salmons 1985

"Seventeenth-century Dutch and the impact of immigration from the south: A sociolinguistic profile", NOWELE 6: 71 - 8 0 . Rickford, John 1986 "Social contact and linguistic diffusion: Hiberno-English and New World Black English", Language 62: 245-289. Romaine, Suzanne 1982 a Socio-historical linguistics: Its status and methodology (Cambridge Studies in Linguistics 34). Cambridge: Cambridge University Press. 1984 a "The sociolinguistic history of t/d deletion", Folia Linguistica Historica 5 (2): 221-255. 1984 b "On the problem of syntactic variation and pragmatic meaning in sociolinguistic theory", Folia Linguistica 18: 409 — 437. (Originally appeared as On the problem of syntactic variation. [Sociolinguistic Working Papers 82] Austin, Tex.: Southwest Education Development Laboratory, 1981.) Romaine, Suzanne (ed.) 1982 b Sociolinguistic variation in speech communities. London: Edward Arnold. Romaine, Suzanne — Elizabeth Traugott 1981 "The problem of style in sociolinguistics", paper presented at the 1981 Linguistic Society of America. Rubin, J. — B. Jernudd (eds.) 1971 Can language be planned? Honolulu: University of Hawaii Press. Runge, Richard 1973 Proto-Germanic r (Göppinger Arbeiten zur Germanistik 115) Göppingen: Kümmerle. Ryan, Ellen Bouchard — Howard Giles (eds.) 1982 Attitudes towards language variation: Social and applied contexts. London: Edward Arnold. Sankoff, Gillian 1972 "Above and beyond phonology in variable rules", in: C.J. Bailey — R. Shuy (eds.), 4 2 - 6 2 . 1980 The social life of language. Philadelphia: University of Pennsylvania Press. Sankoff, Gillian - P. Thibault 1977 "L'alternance entre les auxiliaires avoir et etre en fran^ais parle ä Montreal", Langue Fran$aise 34: 8 1 - 1 0 8 . (Also in Sankoff 1980). Schach, Paul (ed.) 1980 Languages in conflict: Linguistic acculturation on the Great Plains. Lincoln, Neb.: University of Nebraska Press. Scherer, Klaus —Howard Giles (eds.) 1979 Social markers in speech. Cambridge: Cambridge University Press. Schulz, Muriel 1975 "The semantic derogation of woman", in: B.Thorne —N. Henley (eds.), 64-75. Sebeok, Thomas (ed.) 1973 Current trends in linguistics (11th edition). The Hague: Mouton. Shapiro, Fred R. 1985 "Historical notes on the vocabulary of the Women's Movement", American Speech 60: 3 — 16.

The context of language change

95

Shores, David —Carol Hines (eds.) 1977 Papers in language variation. University, Ala.: University of Alabama Press. Silva-Corvalan, Carmen 1986 "Bilingualism and language change", Language 62: 587 — 608. Smith, Philip M. 1979 "Sex markers in speech", in: Scherer, Klaus —Howard Giles (eds.), 109 — 146. 1985 Language, the sexes and society (Language in Society 8). Oxford: Basil Blackwell. Sole, Yolanda 1978 "Sociocultural and sociopsychological factors in differential language retentiveness by sex", International Journal of the Sociology of Language 17: 29-44. Spalding, Keith 1973 "Sprachliche Aufwertung als neues Tabu", Muttersprache 83: 185 — 195. Stanley, David 1977 "Informal speech as performance in sociolinguistic studies", Journal of the Linguistic Association of the Southwest 2: 75 — 82. Sturtevant, Edgar H. 1917 Linguistic change. Chicago: University of Chicago Press. Thomason, Sarah Grey 1983 "Chinook Jargon in areal and historical context", Language 59: 820 — 870. Thorne, Β. —N. Henley (eds.) 1975 Language and sex: Difference and dominance. Rowley, Mass.: Newbury House. Traugott, Elizabeth C. et al. (eds.) 1980 Papers from the Fourth International Conference on Historical Linguistics, Stanford, California, 1979 (Amsterdam Studies in the Theory and History of Linguistic Science 14). Amsterdam: Benjamins. Trudgill, Peter 1974 "Linguistic change and diffusion: Description and explanation in sociolinguistic dialect geography", Language in Society 3: 215 — 246. 1983 On dialect: Social and geographical perspectives. New York: New York University Press. 1986 Dialects in contact. Oxford: Basil Blackwell. Trudgill, Peter (ed.) 1984 Language in the British Isles. Cambridge: Cambridge University Press. Ureland, Sture 1979 "Die Bedeutung des Sprachkontakts in der Entwicklung der Nordseesprachen", in: S. Ureland (ed.), 8 3 - 1 2 8 . Ureland, Sture (ed.) 1979 Sprachkontakte im Nordseegebiet. Akten des 1. Symposions über Sprachkontakt in Europa, Mannheim 1977. Tübingen: Niemeyer. 1982 Die Leistung der Strataforschung: Typologische Aspekte der Sprachkontakte. Akten des 5. Symposiums über Sprachkontakt in Europa, Mannheim, 1981. Tübingen: Niemeyer.

96

Joseph Salmons

Ureland, Sture —Iain Clarkson (eds.) 1984 Scandinavian language contacts. Cambridge: Cambridge University Press, van Loey, Adolphe 1964 Historisch grammatica van het Nederlands. Schets van de klankleer, vormleer en woordforming. Zutphen: W. J. Thieme. Walker, Alastair 1979 "Nordfriesisch: Eine sterbende Sprache?", in: S. Ureland (ed.), 129-149. Weinreich, Uriel 1968 Languages in contact. The Hague: Mouton. Williamson, Robert —John van Eerde 1980 "Subcultural factors in survival of secondary languages", International Journal of the Sociology of Language 25: 59—84. Zengel, Majorie 1962 "Literacy in language change", American Anthropologist 64: 132—139.

Methods to Study Language Change

Philology: Analysis of written records Thomas Cable

Historical linguistics would not be possible without written records, but the relationship between philology (the analysis of those records) and historical linguistics has not been a constant one. When Jespersen (1922) surveyed the scholarship of the preceding one hundred years, he observed that before the nineteenth century the study of language had played the role of handmaiden to philology. Linguistics, rudimentary as it was, had been a means to an end, and the end was philology — the understanding of a past culture primarily through its written literature. Between the discoveries of Rask, Grimm, and Bopp and Jespersen's review, linguistics had become an independent discipline and an end in itself. By the time of the surveys by Austerlitz (1975) and Anttila (1979), it was philology that was described as serving the handmaiden role and as needing defense from its imperialistic neighbor. Some of the most recent work on linguistic aspects of medieval manuscripts, however, shows a shift in emphasis again, away from this subservient role of philology, with implications that promise not only a better understanding of the written records as a goal in itself but also a more adequate basis for historical linguistics. Wells (1962) divided the development of the historical and comparative method into three stages: the literal, the phonetic, and the phonemic. To these one might now add a return to the literal. The first literal stage was characterized by Grimm's "Buchstabenlehre" and the view of letters as elements of language. The establishment in the second half of the nineteenth century of the spoken language as the foundation of linguistic study required a struggle, as reflected in spirited statements by Paul (1891: 433): "No philologist should ever disregard the fact that what is written is not language itself; that speech rendered into writing

98

Thomas Cable

always needs to be rendered back in speech before it can be dealt with"; and Sweet (1876: 471): "The first requisite is a knowledge of phonetics, or the form of language. We must learn to regard language solely as consisting of groups of sounds, independently of the written symbols, which are always associated with all kinds of disturbing associations, chiefly historical". 1 Structural and generative studies, while keeping the spoken language primary, introduced distinctions into the analysis of written records that could not have been discussed in Neogrammarian terminology. The most crucial of these is concerned with the level of linguistic representation at which a scribe writes. According to different approaches, letters represent units at the systematic phonemic level (the morphophonemic), the autonomous phonemic level, or the phonetic. Since scribes do not consistently write at any one level, extralinguistic explanations of departures from the ideal necessarily accompany every discussion. Thus, for King (1969) and O'Neil (1980), who approach the matter from generative grammar, the theoretically most satisfying representation would be morphophonemic, as in the familiar examples from Modern English, divine ~ divinity, electric ~ electricity, etc. However, as both studies point out, at the oldest stages of English and German, the representation is more nearly autonomous phonemic. O'Neil's Old English example dceg 'day' ~ dagas 'days' shows an alternation between /ae/ and /a/, which would have been unnecessary for the scribe to indicate because predictable by rule. Old High German rat 'wheel' ~ rades 'of the wheel', in King's example, shows an alternation between /t/ and /d/ which also would have been unnecessary to indicate. Yet the alternations are spelled out explicitly in both. O'Neil speculates that the autonomous phonemic qualities of Old English orthography derive from a situation in which foreigners without knowledge of the abstract phonology of Old English taught Englishmen to write. King speculates that similar qualities in Old High German orthography derive from a scribe's duties to teach others to spell in as unambiguous a way as possible. When we move from morphophonemic representation to phonemic and allophonic representation (in systems that permit such levels), the problem of distinguishing between levels in historical records becomes more difficult. For Penzl (1971 b) the ideal relationship between phoneme and grapheme is "biuniqueness", each phoneme being represented by one and only one grapheme. As Penzl acknowledges, there are numerous exceptions to this ideal, and some critics have found the

Philology:

Analysis

of written records

99

exceptions significant enough to undermine the heuristic value of the ideal (e.g., Jones 1979). A more restricted assumption, such as underlies Hockett (1959), is that scribes write representations only at the phonemic level, but they do not systematically show all phonemic contrasts; different spellings can refer regularly to the same phoneme with exactly the same phonetic realization in different morphemes. According to Hockett's assumption, then, different spellings cannot refer systematically to allophonic differences. This view was stated in opposition to Stockwell and Barritt (1951), who argued that the orthographic alternation of the digraphs ea ~ ce in Old English represented allophones of a single phoneme, e.g., healf ' h a l f , hcegl 'hail'. The "digraph controversy" inspired numerous articles, including a critique by Kuhn and Quirk (1953) which represents a traditional "philological" approach in opposition to the "linguistic" approach of Stockwell and Barritt (1951, 1961). An important point that emerged from the debate is that in attacking the problem as it has been posed, one must necessarily carry certain assumptions about scribal practice to the examination of the written documents, and these assumptions often cannot be confirmed or refuted by the documents themselves. By the traditional explanation of "breaking" in Old English, for example, ce [ae] changed to ea [aea] before I + a consonant, r + a consonant, and h (e. g., *hcerd > heard 'hard'). "Diphthongization by initial palatal" provided another source for these short diphthongs. However, the life of the Old English diphthongs in the language was relatively short because of the further sound change of "smoothing", and thus few examples of the diphthongs in unconditioned environments had a chance to develop. Kuhn and Quirk (1953) present a number of minimal pairs showing the spelling ea in contrastive distribution with the digraph ce, but Stockwell and Barritt dismiss these as orthographic analogies resulting from a strong spelling tradition. Thus, one is back to the assumptions one started with concerning what scribes tend to do, and parallels in support of these assumptions are drawn from any tradition that offers evidence. Hockett (1959), in arguing that scribes write phonemic rather than allophonic representations, cites the example of the Spanish conquistador Sahagün, who in training speakers of Nahuatl to write their language in Latin letters "virtually stood behind his scribes with a whip" to make them record distinctions that were phonemic in Spanish though not in Nahuatl. Stockwell and Barritt (1961), in arguing that scribes write allophonic rather than phonemic representations in the Old English digraphs, cite

100

Thomas Cable

allophonic representations in Quechua. Penzl (1971 b), in arguing that Old High German scribes write phonemic rather than morphophonemic representations, cites studies of the scribal practice of modern unsophisticated spellers. This reliance on assumptions formed from examples beyond the text in question is related to a strain that runs through many of the structurally oriented studies, as through the earlier Neogrammarian works: the presumed inadequacy of the medieval scribe in recording information in a form directly amenable to linguistic analysis. Unfortunately the conversion of the scribe's record into the linguist's system often results in a premature imposition of theory on data from widely varying sources. The historical evidence, which has been called "highly malleable and easily deformed" becomes obscured. All of the remaining studies in this survey share in one way or another a concern for returning to the physical manuscript and reading again as literally as possible the ink marks on vellum and paper. Generative phonology has inspired a new respect for conventional spelling, although its effect on linguistic paleography has been minimal; generative approaches work best for long-established orthographic systems with a history of major phonological changes occurring after the fixing of the orthography, as in Modern English. The theoretical basis of the most significant current work on the actual manuscripts has some of its roots in the Prague theory of written language and can be found in the writings of Uldall (1944), Vachek (1945, 1973 a, 1973 b), Haas (1970), and Venezky (1970). Especially in the articles of Mcintosh (1956, 1963, 1974, 1975), and Samuels (1963), and in their culminating work, A Linguistic Atlas of Late Medieval England (Mcintosh — Samuels — Benskin 1986) one can see the practical effect of new theory. Mcintosh and Samuels insist on studying the written language in its own right instead of simply as a mirror of the spoken language. Mcintosh (1956: 53) argues against "the false position of continually regarding a written manifestation of language as in some sense inferior in status to, and functionally dependent on its spoken equivalent." Both "the spoken language system" and "the written language system" may be regarded "as standing in some sort of filial relation to 'the language system'" (1956: 45). With this expanded concept of what counts as linguistic, Mcintosh and Samuels construct hundreds of "scribal profiles", including information that may or may not be relevant in an analysis of the spoken language. Thus, they plot erpe :

Philology:

Analysis

of written records

101

erthe and myjt : myght on their maps, though these forms are usually disregarded as orthographic variants implying no distinction in the spoken language. The Atlas editors also examine the physical shapes of letters as graphic equivalents of phonetic variation that may or may not be linguistically relevant. The concerns of the linguist and the paleographer draw closer together. The Toronto editors of the forthcoming new dictionary of Old English proceed with a similar mix of linguistic and philological methods. In a study related to their work on the dictionary, Cameron et al. (1981) examine the orthography, morphology, syntax, and vocabulary of Beowulf and attempt "to provide a description of part of the spelling system of Beowulf, not to make a phonological study". They count features "of little or no phonetic significance" and keep "interpretation of variant spellings to a minimum" (1981: 38). They are reluctant to draw conclusions until the manuscript has been squeezed of all the information that it will yield and the results then compared with similar analyses of other manuscripts. All this may change the traditional view that there is a simple connection between spelling systems and Old English dialects and that dialectal strata in a single work can be excavated by the usual methods (as in the classification of linguistic features by Klaeber [1950] into Early West Saxon, Late West Saxon, and Non-West Saxon). Preliminary comparisons by the Toronto group indicate that mixed spellings might be explained through scriptorial practice and convention at the time of the manuscript instead of through a long, dialectally mixed textual transmission. Studies by North American and European scholars working directly with the Toronto materials or exemplifying the spirit of that approach appear in Bammesberger 1985. In addition to recent studies that interpret familiar evidence anew by interpreting it minimally, several studies have scrutinized manuscript evidence that has been largely ignored in the past. Robinson (1973) examines syntactical glosses (in the form of letters, dots, and strokes) in Latin manuscripts from Anglo-Saxon England and demonstrates how these were used to indicate the word order that the text would assume if translated into Old English. Robinson argues that the word order so indicated can be considered neutral, unstyled, and straightforward, in contrast with the word order found in most other Old English texts, which invariably have at least some literary pretensions. Korhammer (1980) follows up on this line of investigation but concludes that the glosses do not provide a direct window onto neutral

102

Thomas Cable

Old English style. Kiernan's (1981) scrutiny of scribal proofreading in the Beowulf manuscript and his conclusion that the poem was as late as the eleventh century have stirred debate and a reexamination of other Old English manuscripts. Some of these studies, such as Kiernan 1984, have used technological advances in fiber-optic and ultra-violet lighting to reveal letters that had not been visible to modern scholars. For clues to Old English suprasegmentale, Clemoes (1952) and Thornley (1954) have studied punctuation and quasi-musical notation in the manuscripts, and Stevick (1968, 1975) has measured spaces between letters in the Beowulf codex as a step toward relating graphic features to linguistic and metrical features of the text. Mitchell (1985), who repeatedly reminds us that we can not know directly the intonational patterns of Old English, provides us in his masterly account of syntax with a basis from which intonational patterns may be inferred. For the suprasegmentale of medieval Latin, Treitler (1984) throws significant light from a musicological perspective by reexamining the relationship between Carolingian punctuation and the neumes of Gregorian chant. It would be inaccurate to describe recent manuscript work in historical linguistics as a return to the kind of literalness that characterized the period of Jacob Grimm; however, it is fair to say that there is a concern about the accumulated structures of a century and a half through which the primary sources are most often viewed. As Clemoes (1970: 88) states the problem with regard to the new dictionary of Old English, "the manuscript existed 1,000 years ago, it exists today, and it will exist tomorrow, unchanging". But "as soon as a manuscript is used for any purpose whatever, an inconstant element immediately enters; any use of a manuscript is ... essentially an essay in interpretation". For all the familiar oppositions between philology and historical linguistics, the two disciplines become, in practice, essays in interpretation, and the link between them remains intrinsic. One has only to compare the current state of analogous disciplines such as literary criticism and descriptive bibliography, which hardly share a common subject, to understand the coherence of a territory of inquiry delimited by the unchanging manuscript.

Philology: Analysis of written records

103

Notes 1. See also Osthoff — Brugmann (1878: iii—xx) as well as the account of historical phonetics in nineteenth-century England in Dowling (1982). 2. In Old High German, see the comments by Moulton (1971) on Penzl (1971 a); by Jones (1979) on Penzl (1971 a) and on Rauch (1967); in Old Saxon the comments by King (1965) on Twaddell (1963); and in Old English the comments by Kuhn and Quirk (1953) on Stockwell — Barritt (1951). Also see comments by Amos (1980) on numerous early twentieth-century works, including especially Barnouw (1902); and by Mitchell (1980) on punctuation in modern editions. 3. Fisher (1977) carries Samuel's ideas on Chancery English further in arguing that standard written English as it developed in the fifteenth century was independent of the spoken dialect of any region or class.

References Amos, Ashley Crandell 1980 Linguistic means of determining the dates of Old English literary texts. Cambridge, Mass.: Medieval Academy of America. Anttila, Raimo 1979 "Philology and metascience", in: Irmengard Rauch and Gerald F. Carr (eds.), 497-507. Austerlitz, Robert P. 1975 "Literary and folklore — philology and linguistics: Sacred Eurasian antiquity and profane worldwide modernity", in: Herbert H. Paper (ed.), 3-17. Bammesberger, Alfred (ed.) 1985 Problems in Old English lexicography: Studies in memory of Angus Cameron. Regensburg: Pustet. Barnouw, Adriaan Jacob 1902 Textkritische Untersuchungen nach dem Gebrauch des bestimmten Artikels und des schwachen Adjektivs in der altenglischen Poesie. Leiden: Brill. Cameron, Angus, et al. 1981 "A reconsideration of the language of Beowulf", in: Colin Chase (ed.), 33-75. Cameron, Angus — Roberta Frank—John Leyerle (eds.) 1970 Computers and Old English concordances. Toronto: University of Toronto Press. Chase, Colin (ed.) 1981 The dating of Beowulf. Toronto: University of Toronto Press. Clemoes, Peter 1952 Liturgical influence on punctuation in late Old English and early Middle English manuscripts. Cambridge: Department of Anglo-Saxon, Cambridge University. 1970 "The nature of manuscript evidence", in: Angus Cameron — Roberta F r a n k - J o h n Leyerle (eds.), 8 8 - 8 9 .

104

Thomas Cable

Dowling, Linda 1982 "Victorian Oxford and the science of language", PMLA 97: 160-178. Fisher, John H. 1977 "Chancery and the emergence of standard written English in the fifteenth century", Speculum 52: 870-899. Haas, Wfilliam] 1970 Phono-graphic translation. Manchester: Manchester University Press. Hockett, Charles F. 1959 "The stressed syllables of Old English", Language 35: 575-597. Hymes, Dell (ed.) 1974 Studies in the history of linguistics: Traditions and paradigms. Bloomington: Indiana University Press. Jespersen, Otto 1922 Language: Its nature, development and origin. London: Allen & Unwin. Jones, William J. 1979 "Graphemic evidence for the diphthongisation in Old High German", Neophilologus 63: 250-259. Kiernan, Kevin S. 1981 Beowulf and the Beowulf manuscript. New Brunswick, N.J. 1984 "The state of the Beowulf manuscript 1882-1983", Anglo-Saxon England 13: 2 3 - 4 2 . King, Robert D. 1965 "Weakly stressed vowels in Old Saxon", Word 21: 1 9 - 3 9 . 1969 Historical linguistics and generative grammar. Englewood Cliffs, N.J.: Prentice-Hall. Klaeber, Frederick (ed.) 1950 Beowulf and the fight at Finnsburg (3rd ed.). Boston: Heath. Korhammer, Michael 1980 "Mittelalterliche Konstruktionshilfen und altenglische Wortstellung", Scriptorium 34: 18 — 58. Kuhn, Sherman Μ. —Randolph Quirk 1953 "Some recent interpretations of Old English digraph spellings", Language 29: 143-156. Mcintosh, Angus 1956 "The analysis of written Middle English", Transactions of the Philological Society (1956): 2 6 - 5 5 . 1963 "A new approach to Middle English dialectology", English Studies 44: 1-11. 1974 "Towards an inventory of Middle English scribes", Neuphilologische Mitteilungen 75: 602 — 624. 1975 "Scribal profiles from Middle English texts", Neuphilologische Mitteilungen 76: 218-235. Mcintosh, Angus —Μ. L. Samuels — Michael Benskin 1986 A linguistic atlas of late medieval England. 4 vols. Aberdeen: Aberdeen University Press. Mitchell, Bruce 1980 "The dangers of disguise: Old English texts in modern punctuation", Review of English studies 31: 385—413. 1985 Old English syntax. 2 vols. Oxford: Clarendon.

Philology: Analysis of written records

105

Moulton, William G. 1971 Review of Penzl 1971 a. Phonetica 24: 2 3 8 - 2 5 2 . O'Neil, Wayne 1980 "English orthography", in: Timothy Shopen — Joseph M. Williams (eds.), 63-83. Osthoff, Hermann —Karl Brugmann 1978 Morphologische Untersuchungen auf dem Gebiete der indogermanischen Sprachen I. Leipzig: Hirzel. Paper, Herbert H. (ed.) 1975 Language and texts: The nature of linguistic evidence. Ann Arbor: Center for Coordination of Ancient and Modern Studies, University of Michigan. Paul, Hermann 1891 Principles of the history of language (2nd ed.). Trans. H. A. Strong. London: Longmans, Green. Penzl, Herbert 1971 a Lautsystem und Lautwandel in den althochdeutschen Dialekten. München: Hueber. 1971 b "Scribal practice, phonological change, and biuniqueness", German Quarterly 44: 305-310. Rauch, Irmengard 1967 The Old High German diphthongization: A description of a phonemic change. The Hague: Mouton. Rauch, Irmengard — Gerald F. Carr (eds.) 1979 Linguistic method: Essays in honor of Herbert Penzl. The Hague: Mouton. Robinson, Fred C. 1973 "Syntactical glosses in Latin manuscripts of Anglo-Saxon provenance", Speculum 48: 4 4 3 - 4 7 5 . Samuels, M. L. 1963 "Some applications of Middle English dialectology", English Studies 44: 81-94. Shopen, Timothy—Joseph M. Williams (eds.) 1980 Standards and dialects in English. Cambridge, Mass.: Winthrop. Stevick, Robert D. 1968 Suprasegmentals, meter, and the manuscript of Beowulf. The Hague: Mouton. 1975 Beowulf: An edition with manuscript spacing notation and graphotactic analyses. New York: Garland. Stockwell, Robert P . - C . Westbrook Barritt 1951 Some Old English graphemic-phonemic correspondences (Studies in Linguistics Occasional Papers, No. 4.). Norman, Okla.: Battenburg. 1961 "Scribal practice: Some assumptions", Language 37: 75 — 82. Sweet, Henry 1876 "Words, logic and grammar", Transactions of the Philological Society (1876): 470-503. Thornley, G. C. 1954 "The accents and points of MS. Junius 11", Transactions of the Philological Society (1954): 178-201.

106

Thomas Cable

Treitler, Leo 1984

"Reading and singing: On the genesis of occidental music-writing", Early Music History 4: 135-208. Twaddell, W. Freeman 1963 "Graphical alternation in Old Saxon suffixes", Monatshefte 55: 225 — 228. Uldall, H. J. 1944 "Speech and writing", Acta Linguistica 4: 11 — 16. Vachek, Josef 1945 — 49 "Some remarks on writing and phonetic transcription", Acta Linguistica 5: 8 6 - 9 3 . 1973 a "The present state of research in written language", Folia Linguistica 6: 47-61. 1973 b Written language: General problems and problems of English. The Hague: Mouton. Venezky, Richard J. 1970 The structure of English orthography. The Hague: Mouton. Wells, Rulon 1974 "Phonemics in the nineteenth century, 1876 — 1900", in: Dell Hymes (ed.), 434-453.

The chronology of phonological change Bridget

Drinka

Constructing an accurate chronology for linguistic change entails a considerable number of problems. Precise dates are seldom available, except in such rare instances as the the destruction of Pompeii in 79 A.D., the swearing of the Strassburg Oaths in 842, and the recording of events in the Anglo-Saxon Chronicle in Peterborough up to 1154. More often, judgments must be based on spelling clues, orthoepic commentary, borrowings, metrical evidence, and so on. When no documentary evidence is available at all, as is the case with most prehistoric sound changes, such methods as internal reconstruction must be used.

1. Documentary evidence 1.1. Orthography A change in spelling often indicates that a phonological shift has occurred. Latin provides an example of a well-documented shift which can be securely dated: rhotacism ( [z] > [r]/V V, e.g., *genes-es > generis; *flös-em > flörem). Cicero 1 says that L. Papisius Crassus, who was dictator in 339 B.C., was the first in his family to change his name to Papirius. In 312 B.C., Appius Claudius gave official approval to the use of the new in names (Meillet 1966: 142; Sommer 1977: 146). If it is true that names usually change more slowly than the rest of the language (Maniet 1957: 59), rhotacism must have been complete by around the mid-fourth century B.C., a fairly precise terminus ante quem.2

108

Bridget

Drinka

One of the first signs of phonological change often appears in the "naive spelling" of laymen, foreign scribes, and "poor spellers". Numerous examples can be found in informal written English: should of, gonna,3 etc. One of the best historical sources of such information is the graffiti of Pompeii, scratched or written on the walls of that city shortly before the eruption of Vesuvius in the first century A.D. This evidence gives us a rare specimen of casual, everyday Latin, found in the curses, announcements, eroticisms, and even shopping lists of the Latin-speaking people. It is extremely useful in determining the chronology of Latin sound changes, not only because the date of the eruption is known, but also because much of the city had been destroyed in an earthquake in 62 A.D. and had been recently rebuilt, so that a terminus a quo exists as well. Many of the changes documented at Pompeii demonstrate the surprisingly early origin of a number of Romance trends, and it will be instructive to examine a few of these in detail. For instance, a very early example of the use of the "prothetic vowel" before an initial /s/ + consonant is found in Pompeii: /SMVRNA ( = Gk. Σμύρνα). 4 The loss of m word-finally does not appear in standard orthography for another thousand years (Wright 1976: 180), yet it is extremely common in Pompeii: 5 IANVARIUS AMAT VENERIA _ 'Januarius loves Veneria' VINV I nON BIBIIT 6 'He won't drink wine' SIIRIINA ISIDORV FASTIDIT 'Serena can't stand Isidore' Syncope, too, was already evident in Pompeii, especially that of /u/ between a consonant and /l/: COLICLO 'small cabbage' ( < cauliculum)7 MASCLVS, MASCL(os) ( < masculusf Finally, the loss of /n/ before /s/ is also well-documented in Pompeii: COSIDERATE ( < considerate) MESA ( < mensa) INNOCES ( < innocens) The absence of the in the writing of authorities like Varro (mesa), Quintilian (cosul, dsa), and Cicero (foresia, Hortesia, Tjas for mfäns)

The chronology of phonological change

109

(Maniet 1957; Väänänen 1966: 68) as well as the use of hypercorrect forms (PARJENS for paries, FORMONSUS for formosus) indicates that this was an important, regular shift in the spoken language. 9 Phonological shifts can also be detected in the naive spelling of later Latin writers, especially among the less classically trained Christians. Even as early as the third century A.D., confusion had already arisen between /kj/ and /tj/, which were often interchanged in inscriptions: TERMINACIONES for TERMINATIONES DEFENICIONES for DEFINITIONES (See Slotty 1960 for examples) The Strassburg Oaths (842 A.D.), which consist of German and French translations of the original Latin oaths, contribute greatly to our knowledge of the development of French, and illustrate what kind of orthographic evidence can be gleaned from bilingual documents. The spelling used in the oaths is particularly interesting because the French oath is the first documented attempt by scribes to record spoken French for the benefit of the people. For example, the in ajudha, cadhuna may be an attempt to represent the fricative /δ/, as it appears to be in the German Oaths (bruodher, Ludhuuuig); Elcock (1960: 342) concludes from this that in Latin, intervocalic jij had developed in the following manner: -t- > -d- > -Ö- (later > 0) 10 This fresh attempt to link sound to symbol provides many of the same clues that naive spelling would. Another instance of phonological change can be seen in the spellings and , and in close proximity to one another. 11 These forms indicate that , , and had all come to represent /a/ word-finally (Elcock 1960: 341). Similar mergers are evident in other languages: Old English and had become interchangeable in northern and eastern England by Middle English times, suggesting that the sounds they originally represented, f\j and /y/, had also merged (Scragg 1974: 21). Since the mid-twelfth century, naive spellings such as (gresser) ('größer') and ('anzünden') have been found in the High German dialects, along with inverse spellings such as (bosser) ('besser') and ('Kirche'); /ö/ and /ü/ have evidently merged with their unrounded counterparts, /e/ and jij (Penzl 1968: 107). Orthographic evidence is often assumed to reflect pronunciation quite accurately, but this is not necessarily the case; interpretation is,

110

Bridget

Drinka

at any rate, not always straightforward. Intervention from external sources is often invoked as an explanation for poor phonemic-graphemic correspondence. For example, intervocalic lenition and voicing of voiceless stops must have taken place in Britonnic Celtic between the mid-fifth and the beginning of the sixth century A.D. according to Jackson (1953).12 But these processes were not marked in the Latin orthography at that time (1953: 68): Old Welsh bleuporthetic [bleuborOed'ig]

Old Cornish Petrocn [predrög]

Old Breton etbinam [edbinav] 14

Jackson explains this anomaly by suggesting that lenition and voicing had crept into the spoken Latin of the clergy. Their Latin was not subject to spelling reform, so that a tradition of using for jdj in the environment of a vowel may have been established and extended to Celtic forms. Sommerfelt (1958: 280) agrees that spoken Latin sound changes probably paralleled those of British Celtic, to some extent, but adds that some of the more "radical" changes must have been blocked by Latin spelling, such as /kt/ > /ith/ and /e/ > /ui/. Förster (1941: 258), on the other hand, sees such spellings as evidence that Latin writers were relying on rather archaic documents as their sources, rather than on living speakers of Celtic. While some phonemes appear to be inadequately represented by a spelling system, as in the above example, other phonemes seem to be over-represented: Kyes (1979) claims that Old Netherlandic 15 and (all derived from Proto-Germanic *aw), though distinguished in spelling, were apparently allophones of one phoneme. He explains this unequal distribution of orthography and phonology as due to the influence of Old High German scribal tradition. Germanspeaking scribes would have considered the Old Netherlandic contrasts [o:] and [ou] to be significant, since these were phonemes in their language. 16 Daunt (1968, first published in 1939) argues similarly that the influence of a foreign scribal tradition is evident in the orthography of Old English. She questions the traditional interpretation of and , and suggests that the second member of the diphthongs simply indicated "dark quality". She bases this assumption on evidence from Old Irish, where vowels were marked to show the pronunciation of following consonants, creating "false diphthongs" (1968: 400). Daunt suggests that this diacritical use of vowels by Irish scribes

The chronology of phonological

change

111

passed into the Old English tradition, and that Old English spelling must be interpreted accordingly. Kuhn and Quirk (1953), however, find a number of faults in Daunt's argument. Among the most serious are the logistic and chronological improbability of her theory, and the fact that evidence exists for the retention of a distinction between long and short diphthongs in the language. They point out, for example, that the English manuscripts which are most closely connected to Irish in time and place happen to be those which show the smallest number of digraphs (1953: 148). 17 Furthermore, some manuscripts indicate more than an allophonic status for and , and place names also suggest true diphthongs (1953: 150): -beaur (Cornwall 1324) > Estharabyar -fealw (Devon 1316) > Vialepitte -pearroc (Devon 1312) > Piarrecumbe With these and other arguments, Kuhn and Quirk convincingly demonstrate that the orthography of Old English must have, in this case, reflected phonetic reality. Schmalstieg (1981: 123) seems to have a much more pessimistic point of view than the authors discussed so far with regard to the possible relationship of sound to orthography. In noting the Latvian representation of /ua/ by , he goes so far as to say that different spellings may have no meaning at all or reflect only idiosyncracies of the scribe. If they do reflect differences in pronunciation they may merely show different styles of speech, speeds of production, or perhaps sound changes in statu nascendi.

Consistent spelling, he implies, is a recent and ill-represented creation in the written languages of the world, one which is not to be trusted for phonological analysis. 18 For example, he points out that Camden spelled "pity" as pity, pyty, pitie, pytie, pittie, and pyttye. But a brief look at Caxton's prologues and epilogues (collected in Crotch 1928) indicates that spelling alternation in English was not totally random even as early as the fifteenth century.19 Baugh and Cable (1978: 208) show, furthermore, that most writers in Shakespeare's time had developed some sort of stable spelling style, and that conclusions about their pronunciation can often be drawn. Besides, as we have seen, alternative explanations can sometimes be found for apparent poor correspondences. Schmalstieg's views may thus be somewhat overstated, but his general assessment is correct: spelling does not necessarily correspond to sound.

112

Bridget

Drinka

The source of information is also an important consideration: the naive spelling of semi-literates is not as trustworthy as that of grammarians, teachers, and other "good spellers". Pulgram (1958) concludes that a poor speller strives to imitate prestigious orthographic conventions, but fails because he has "a limited number of conventional letters to operate with and lacks the ingenuity to manipulate them as cleverly as does the expert" (1958: 316). Even grammarians are not a totally reliable source, however, since they have been known to pass down conventions, unaltered, for years (1958: 314). It would be dangerous to base a chronology on any of these sources alone. Likewise, too much importance must not be placed on conservative inscriptions, government documents, and names. Official proclamations are often quite archaic. For example, Latin words which frequently appear on monuments (e.g., idüs, hie, si) are still spelled with archaic in Pompeii (EIDVS, HEIC, SEI) even though the change in pronunciation had already occurred. 20 Latin interpretations of Celtic names were often drawn from official government documents, so that extremely conservative spellings were maintained: the apparent diphthongs in Alauna, Croucingo, and Leuca were probably all pronounced as /δ/ by 100 A.D. (and δ > ü by 300 A.D.), even though they were still spelled as diphthongs in the work of the anonymous geographer of Ravenna as late as c. 670 A.D. (Jackson 1953: 37). Jackson concludes: Consequently, providing that it appears to be textually sound, an early occurrence of a changed form in these Roman documents is of more value for dating than a late occurrence of an unchanged form. 21

Jackson (1953: 648) and Amos (1980: 19) note the conservative nature of Celtic and Anglo-Saxon names, respectively, in resisting syncope: Brit. *Düno-catus > Old Welsh Dinacat, Dincaf, Mid. Welsh Dinogat, Ding at OE Inguburg (from the Liber Vitae, c. 800) MW Dinogat was probably just copied from older sources, since stress shift in the eleventh century regularly caused reduced vowels to drop out. Likewise, in Old English, «-apocope had taken place centuries before elsewhere in the language, but u persisted in names. It is clear, then, that orthographic evidence can be extremely valuable, but must be used with due caution.

The chronology of phonological change

113

1.2. Orthoepie evidence Comments by grammarians — often prescriptively recommending the reversal of an already accomplished change — can also help determine a chronology. For example, the voicing of /t/ in a voiced environment is not a new phenomenon in English: Elphinston, who published in 1765 and 1787, noted forms like "proddestant" and "pardner" in the "vulgar" London speech of his time. Similarly, the German shift of /r|g/ > /η/ must have already taken place by Valentin Eckelsamer's time because in his book of 1537 he disapproves of the use of two letters to represent a single fused sound, as in Engel (Penzl 1968: 107 —108).22 An early example of an orthoepic source which is extremely valuable for phonological analysis is the Appendix Probi. The document was appended to the work of Valerius Probus, a Latin grammarian of the first century A.D., and includes a word list dating possibly from the end of the third century which advises readers as to the correct spelling of various questionable forms, 23 e.g.: Process indicated SPECULUM PRIDEM ANSA OCCASIO

non non non non

SPECLUM PRIDE ASA OCCANSIO

syncope loss of m / _ # n->0 /_s " " (hypercorrection)

That these corrections were felt to be needed indicates that the changes were already in progress. Many of these shifts, including those listed above, had already been documented in Pompeii at least 200 years before. The confusion of /b/ and j\j (BACULUS non VACLUS), the indecision concerning geminates (CAMERA non CAMMERA), and the apparent development of a [j] glide (VINEA non VINIA), were all prefigured in Pompeii, and were all carried out to various extents in the Romance languages. Such prescriptive lists can thus prove very useful in the analysis of the stages of phonological change. Orthoepic evidence, like orthographic evidence, is subject to different interpretations by different scholars. For example, a number of authors disagree in their reading of the English orthoepists of the sixteenth and seventeenth centuries, and their interpretations of the English Great Vowel Shift differ accordingly. Examining this question in some detail will provide an example of the complexities faced by historical linguists as they interpret this type of evidence.

114

Bridget Drinka

Lass (1978: 268) presents a fairly traditional explanation for the development of ME [ü] based on his interpretation of authors from Hart (1569) to Cooper (1687): u: 1 fflu: 1 OU

AU < I 9U I (an)

Stockwell (1972: 351), on the other hand, interprets the descriptions of Wilkins (1668) and Cooper to mean that the first element of the udiphthong was originally [λ] or even [i]: 1.

— —• Ty ey — aey —

Tw y \ e>y \

ey

\

\
y \

ae y ν ay

\

ME ai This approach has the advantage of characterizing the sound shift as a continuous progression, as opposed to the zig-zag path seen in both the "lowering first" and the "centralization first" approaches. Also to

116

Bridget

Drinka

their credit is the fact that Labov, Yaeger, and Steiner are able to synthesize both points of view to some extent by claiming that centralization of /iy/ did follow (and follow from) lowering, but was also influenced by a need to differentiate falling /iy/ and rising /aey/, as Dobson and Stockwell stress. Clearly, then, orthoepic evidence can provide fairly precise information about the chronology of a change, but can also lead to very different conclusions; like orthographic evidence, it requires careful analysis to be interpreted correctly. 26 1.3. Poetic evidence Meter, rhyme, alliteration, assonance, and other poetic devices are often used to help date changes in a language. Metrical evidence is extremely valuable in deciding the chronology of stress shifts or the loss of phonemic length. For example, it can be surmised that Greek authors of the second century A.D. were becoming unsure as to the quantity of their vowels, and that the dactylic hexameters of Nonnus, a poet around 400 A.D., were probably based on stress accent, rather than on pitch (Palmer 1980: 177). Rhyming evidence can also be useful: M H G /s/ and jzj must have merged word-finally around the thirteenth century: hüs 'house': üz 'out'; before that time, these forms were unacceptable as rhymes (Penzl 1968: 109).27 Kökeritz (1953: 31 ff.) points out that rhymes of the sixteenth and seventeenth centuries are not to be altogether trusted, since poets apparently passed down rhyming traditions from generation to generation, without necessary regard for pronunciation: love-prove, goodblood. These may well have been "eye-rhymes" by Shakespeare's time. A better source of information, according to Kökeritz (1953: 149), often proves to be the "homonymic pun", as seen in the following passage from Othello (3.1.6 ff.): Clown: Are these I pray you Winde Instruments? Musician: I marry they are sir. Clown: Oh, thereby hangs a tale. Musician: Whereby hangs a tale, sir? Clown: Marry sir, by many a Winde Instrument that I know. The humor of this pun relies on the homonymity of tail and tale. Kökeritz believes that the vowels of these two forms, /ai/ and /ä/ in Middle English, had merged into /ae:/ during the first half of the

The chronology of phonological change

117

fifteenth century, then raised to merge with /?:/ ( < ME [ε:]). He presents orthographic evidence to support his assumption that Shakespeare pronounced these three vowels, [a], [ai] and [?], the same: ME sale and vane are spelled as (saile) and , and puns are made on bait — bate — beat, mail—male —meal, and raisin —reason (1953: 139, 173 ff.): If Reasons were as plentie as Blackberries, I would give no man a Reason upon compulsion, I. (First Part of Henry IV, 2. 4. 260 ff.)

However, Cercignani (1981: 12 ff.) disagrees strongly with Kökeritz's analysis; he suggests, for example, that the word raisin had a variant pronunciation with [?], as opposed to Kökeritz's interpretation that the entire phoneme /ai/ had shifted to /?/, and finds little evidence for proposing an early merger of the other two phonemes, either. He notes, for instance, that Hodges, in his work, Special Help to Orthographie (1643) distinguished /ä/ from /ξ/. Cercignani goes on to make a more general criticism of Kökeritz's reliance on puns for evidence by showing that complete homophony is not guaranteed in a pun. For example, it seems unlikely that bate and beat would be homonyms in the following passage: That is, to watch her, as we watch these kites That bate and beat and will not be obedient. (Taming of the Shrew, 4. 1. 199)

Poetic evidence, then, can be quite informative, but it is also subject to controversy, and must be analyzed critically. 1.4. Contact evidence A great deal of information about the past can often be gained by examining loans and other vestiges of linguistic contact. For example, we can tell much about the history of /h/ in French by examining loans into English. Initial /h/ was lost in Vulgar Latin; therefore, the majority of French loans into Middle English from this source did not have initial /h/, except in spelling (e.g., honor).29. But later loans from Germanic into French retained initial /h/, and were passed into English with [h]'s intact: haste, hardy, etc. (Serjeantson 1935: 300; see also Fouche's extensive discussion, 1966: 578 — 585 and Kesselring 1973: 151 — 152). Thus, English evidence can be used alongside the French to judge which changes have occurred in the latter.

118

Bridget Drinka

In similar fashion, the fact that Latin jij was often transcribed as by the Greeks suggests the lowered quality of that vowel (Väänänen 1966: 21). Latin, in turn, also provides information about Greek: even though existed in their language, Latin speakers first transcribed the of Greek as

, as in Pilippus (Allen 1968: 20). Thus, was probably not a fricative but an aspirated stop at this point in time. This generalization would apply to and < / ) as well (cf. Lat. dracuma = δραχμα, tüs = θυος). By the first century A.D., however, the series had become fricative; Pompeii provides the earliest sure examples, such as Dafne. Contact information can also provide essential clues for reconstructing the phonological system and the chronological order of sound changes in the language of non-literate peoples. The following Latin loans in Old High German Oberdeutsch suggest that the second consonantal shift was already complete for /p/ in word-initial position at the time of the borrowing, while it had still not taken place for /t/ ad /k/ word-internally: pigmentum •1 biminza

impeltare Ί pelzon

portulaca picem Ί Ί purcella peh (Penzl 1972: 100)

Schrodt (1976: 69), while stressing that too much weight should not be placed on loanwords, suggests that the voiced series in general was the first to undergo the first consonantal shift, as possibly evidenced in the Germanized Celtic name *Gaisorigs—> Gmc. Caesorix. Schönfeld (1911: 59) agrees that this interpretation is possible, since both elements of the compound are attested Celtic words, but also mentions that the converse is conceivable, as well: The name could have been a Celticization of the Germanic form *Gaisaricus (cf. Goth. *Gaisa-reiks, OHG Ger-rich), which is well-attested in various forms (1911: 99-101). Fourquet (1948: 17) points out another problem in using Celt, rig —> rtk as evidence: the form could have been borrowed before the time of the shift, or, on the other hand, it could have been borrowed and immediately changed into the more acceptable form 29 (cf. Bach —> [bak] in English). One of the major problems with using loanwords, then, is interpreting the exact chronology of the loan; another serious problem is

The chronology of phonological change

119

knowing whether a given shift took place in the borrowing or the lending language. Sometimes the data can be sorted out successfully. For example, both Celtic and Latin underwent syncope, but in the following example, the non-Latin stress pattern of the original Brittonic form indicates that the syncope of /[/ took place in Celtic rather than in Latin: Brittonic episcöpus > Welsh esgob Cornish escop Breton eskop (Jackson 1953: 652) It is somewhat more difficult to separate the Latin from the Germanic in establishing the chronology of the partial merger of Gmc. /e/ —» /i/. Penzl (1972: 40 f., 99) studies Germanic names as they appeared in classical writings, such as Tencteri, Ingvaeones, Fenni, Segimerus, Sigimerus. He then sets up the following relative chronology: 1) 2) 3)

e —*•

i/

i)C /_CC / C0i (Penzl 1972: 99)

Penzl's hypothesis is supported by the fact that the earliest documented example of Sigi-jSegimerus contains : Σεγίμηρος (Strabo, first century B.C.). The next attestations seem ambiguous, since Velleius Paterculus (first century A.D.) writes Sigimerus, while Tacitus (roughly contemporary) uses the form, Segimerus. Penzl himself (1972: 40 f.) admits that such alternation may well have been the result of Latin rather than Germanic variation, and Schönfeld (1911: xviii) adds that Celtic Sego- could also have influenced the classical interpretation of Seg-. However, in defense of Penzl's chronology, it can also be argued that since Velleius Paterculus was an officer in the Roman army in Germany, his eye-witness account may be more accurate, whereas Tacitus might have been influenced by earlier Greek writers. 1.5. Sociolinguistic considerations Obviously, great care must be taken when setting up chronologies based on any kind of documentary evidence. Miscalculating whether the processes under consideration are linearly ordered or whether they coexist in separate dialects is one of the most serious hazards facing historical linguists as they attempt to establish a temporal framework. Some phonological innovations which appear at first sight to stem

120

Bridget

Drinka

directly from an earlier stage of the language must instead be explained as resulting from variation due to social stratification, dialectal contact, style shifting, etc. Variation found in texts is often discounted by historical linguists as unanalyzable or even uninteresting, but if it is admitted that languages of the past were as heterogenous as languages of the present, it should be possible to use regular variation found in documents to detect sound changes in progress, as sociolinguists have done for present-day material, and to trace the development and spread of these changes. Just such an attempt has been made by Toon (1983) in his thorough analysis of the Old English reflexes of WGmc. *a before nasals, alternating in spelling between and in texts of Mercian provenience or influence in the eighth and early ninth centuries A.D. Toon does not believe that the sound lay somewhere between [a] and [o], as has often been thought, but that a sound change was in progress. First of all, he observes that, in Mercian glosses, the conditioning environment of the shift widens as time goes by (1983: 105 f.):30 a —> / Ν Erfurt (c. 750 A.D.)

Corpus (shared)* Corpus (c. 800) (independent) .85 \ /m 1.00\ .80 \ / η .75 \ .50 \ / nCh .75 \ .33 — (not much data) \ q .66 / .00/ \ m C h .66/

*"Shared" glosses are those derived from the same archetype as the Erfurt glosses: "independent" glosses are derived from other sources; "h" refers to a homorganic consonant. By the time of the Vespasian Psalter (c. 825 A.D.), the conditioning environment had been completely generalized: WGmc. *a is represented only as when followed by any nasal (1983: 106 — 107). Now the tendency for *a to appear as in nasal environments is also seen in Kentish-Mercian charters of the same time; in fact, in such documents dating from 803 to 824 A.D., only appears (1983: 93). Toon attributes this fact to Mercian political influence in Kent, and cites extensive extralinguistic evidence to support his claim.31 Also important is the fact that Mercian orthographic and lettering styles

The chronology of phonological

change

121

appear predominantly in the Kentish documents which show (1983: 93 f.), and that other concurrent sound changes also point to Mercian influence in Kent. 32 Thus by recognizing variation as a manifestation of change in progress, rather than dismissing it as exceptional, Toon is able to trace the gradual development of a sound change; by correlating extralinguistic data with linguistic evidence, he is able to follow the diffusion of this shift into other speech communities. Social stratification and stylistic variation are also important considerations in the construction of chronologies. The importance of these factors is particularly evident in the complex development of the French diphthong , which Fouche (1958: 270 — 281) has examined in depth. A careful look at his analysis will illustrate how essential the consideration of social factors can be. Fouche sets up the following early chronology for non-western dialects of the "langue d'o'il": Lat. ei > φϊ > oi > oe > oe [we] > [w?]33 (by 1300) From c. 1300 on, there is abundant evidence of variation according to social class. The learned language tended to preserve the [w?], but among the Parisian bourgeoisie, the [w?] soon began to be replaced by [wa]34 as indicated in the rhyming patterns used by Francois Villon, the fifteenth century poet of pickpockets, prostitutes, and other lowlife characters of Paris: 35 poire (< pira) rhymes with quarre (= querre < quaerere) alongside maschouere (= maschoire) with chere The use of [wa] was frowned upon by grammarians, 36 but after the French Revolution, it became standard pronunciation for many forms, probably due to its popular connotations and to its Parisian origins. In the meantime, another variant of [w?], the form [?], was slowly spreading from the speech of the lower classes — the artisans and country folk — into that of the middle classes. [?] for [w?] was not yet evident in Villon's work (except after consonant clusters) 37 but it appears to have been adopted by the bourgeoisie soon after his time, towards the beginning of the sixteenth century: fret (= froid)

emplaye (= employe)

(from the seventeenth-century Mazarinades, songs and pamphlets against Cardinal Mazarin). The most pervasive and systematic use of the [$] was in the imperfect tense,

122

Bridget

Drinka

Mod.Fr. devait, OFr. -oit and in referring to inhabitants of certain countries and cities, Mod.Fr. frangais, OFr. -ois (cf., for example, the Modern French name Frangois). This pronunciation spread quickly to the courtly language. 38 Many grammarians of the time blamed the new development on Italian influence. 39 Italian may indeed have played some role in the shaping of courtly pronunciation, but the bourgeoisie (and ultimately, the lower classes) must have been more generally responsible for implementing the change, since many of the words which underwent the shift came from the vocabulary of everyday life: — — — —

claie 'wattle (of a sieve)' (OFr. cloie < Celt, cleta) cueillette 'gathering (a crop, etc.)' (OFr. coilloite < Lat. collecta) marais 'marsh' (OFr. marois < Lat.-Gm. *mariscu) rets 'net' (OFr. roiz < Lat. retes) (Fouche 1958: 274)

The lower class origins of [?] are also suggested by its connection with casual, non-eloquent style, as illustrated by the grammarian Chifflet's statement (1659): "II faut dire Frangais, Polonais, Anglais, Holandais dans la conversation, et non en parlant publiquement" (Fouche 1958: 280). The variants [wa] and [?], then, are two reflexes of one phoneme which emerged at about the same time, but which were connected with different social classes. As they each moved up the social scale, their histories intertwined in a complex fashion. Without reference to the social status of these forms, a clear characterization of the chronology of their development could not be made. Dressier (1973: 144) also shows how styles may have their own chronologies, causing great difficulties for historical linguists. He warns that we should not oversimplify the analysis of dead languages by assuming that documents can give a true picture of all styles of speech. Any kind of writing is formal, to some extent, 40 and furthermore, styles overlap and influence one another. He recommends that special attention be paid to the "negligent" styles and to key words which may indicate the style. OHG Keisur < Lat. Caesar is an example of a borrowing from the formal register, since monophthongization was occurring in more casual Latin at that time (Dressier 1973: 143). Väänänen (1966: 23) calls the shift [ae, ai] > [e] "rustic" — in any case, it was not the citified, sophisticated style.

The chronology of phonological change

123

Documentary evidence, then, is an extremely valuable source of information, but it must be analyzed with great care. It is seldom possible to find a simple, straightforward solution to a chronology problem without considering the entire system of the language and the society in which it functions.

2. Evidence from more abstract sources When documentary evidence is sparse, other methods must be relied upon to establish the chronology of linguistic change. Among the most frequently used methods are the analysis of synchronic rule ordering and internal reconstruction. 2.1. Evidence from synchronic rule order It has long been felt that the order in which rules appear in a synchronic grammar closely approximates the order in which they entered the language. Stockwell (1972: 355), in discussing the Great Vowel Shift in English, states that since "there has been no restructuring of the primes that participate in this shift, it should correspond synchronically with what happened historically." Hoenigswald (1960: 115) implicitly agrees: he gives a Latin example showing how vowel weakening in medial syllables must have preceded syncope because only this order produces attested forms like concutis: *

correct order" incorrect order"

1) 2) 1) 2)

weakening syncope syncope weakening

* caput is capitis caputis capitis

*conquatis conquitis concutis concutis *concitis

The presence of forms like πλάζω 'mislead, seduce' in Ancient Greek gives us a clue to the order of application of several rules, as well as an indication of the actual pronunciation of ( ζ ) (Allen 1968: 53 ff.; Palmer 1980: 210 — 211). The form derives from earlier *pländzö. The nasal is still preserved in the aorist infinitive πλάγξαι /plaqksai/ and the passive πλαγχθήναι /plaqk h t h enai/; the earlier affricate pronunciation is assumed based on similar spelling of forms like οζος 'branch, twig' < IE *osdos, (cf. Gm. Ast 'branch'). But Attic Greek, like later Latin, had a rule by which η > o/ [ + sib] (e.g., syn + stasis —>

124

Bridget

Drinka

systasis). For this rule to have applied in πλάζω, the < C > must have been pronounced [zd]. Therefore, metathesis must have occurred before the application of the nasal deletion rule. Thus, in a number of cases, it may indeed be possible to assume that the original rule order is preserved. However, there are at least two factors which must be taken into consideration when synchronic rules are used as diachronic evidence: a) rules do not exist at just one point in time; they may remain productive for centuries; b) rule order may be reanalyzed, i.e., reordered. a) Anderson (1978: 25) stresses that it is simplistic to expect rules to accumulate in linear order at the end of a grammar; he feels that they may overlap, applying synchronically in different orders. He gives an example from Modern Greek, in which high vowels continue to cause palatalization, even when new velars enter the language. Anderson assumes that where rule R is present in synchronic grammars over a given period, we would expect that when other rules enter the grammar during that period, as a result of other sound changes, rule R would give the appearance of applying both to the inputs and to the outputs of these later rules (1978: 27)

Stockwell (1978: 339) agrees that some rules and processes may remain productive for long periods. For example, he believes that the Great Vowel Shift is still going on in dialects like Southern American (ü > [aew]) and Cockney (ü > [aew] or [ew], Τ > [oy]). He feels that the process had been going on for a period of time (cf. diphthongization from the vocalization of g and w: OE wag, weg > ME wei, wai, way, wey; bugan > bow), before it became crystallized in the spelling. In fact, he sees the "out-glide simplification", already evident in WGmc. iu OE , > [i:] eu—• [ae:] as a case of the same kind of reduction seen in Southern [a:] 'eye', [hae:s] 'house', etc., that is, as part of the same vowel shift (1978: 341-42). The major difference between Anderson's and Stockwell's approaches, with regard to chronology, is that Anderson is concerned with the persistence of rules themselves, while Stockwell is referring to the perseverance of a trend: that of minimization of the contrast between vowel and "out-glide" (as well as maximization of vowel-in-

The chronology of phonological

change

125

glide contrast). Anderson would disapprove of using linear ordering as a method for setting up a chronology; Stockwell would consider rule order information essential, but would caution against pinning down a process to a narrow point in time. b) Closely related to this point is the fact that rule order may shift as time goes on: such reinterpretation is often considered to be a major source of phonological change. There are various explanations for why such a change would occur, among them the maximization of rule application towards feeding and counter-bleeding orders, paradigmatic unity, as well as transparency. Kiparsky (1978: 52) gives an example from early Iranian which shows how opacity can cause rule reordering. In older Avestan, the aspirated stops which had originally undergone Bartholomae's Law (e.g., drugh + ta—> *drugdha) had lost their aspiration but continued to cause progressive voicing assimilation, in contrast to the original unaspirated stops, which were still subject to regressive voicing assimilation: older Avestan /drug + taj —> drugda vs. /bag + taj —• baxtaAi ( < Indo-Iran. *dhrugh-) ( < Indo-Iran. *bkag-) 'deceived' 'allotted as a share'

At this point, Bartholomae's Law had become opaque, and so was lost in later Avestan: jdrug 4- ta/ —• druxta, like baxta. Thus, the synchronic rule order here does not preserve the rule's history, yet it is essential to our understanding of how the later history unfolded.

2.2. Evidence from internal reconstruction: Prehistory When little other evidence is available, it is still sometimes possible to set up a chronology of linguistic change based on the internal reconstruction of that language. Antonsen (1965: 20) stresses that a protolanguage, like any language, must go through stages, and cannot exist timelessly at one point; it must show "development in time, subphonemic variation, and gradual dialectal differentiation in contrast to the sharp splits presented by the family tree theory."

126

Bridget

Drinka

He sorts through a great deal of Germanic data to set up the following relative chronology (1965: 31): Gmc. consonant shift pre-Gmc. Gmc. accent shift Proto-Gmc. [i] > /e/ [e] > Ν Late Proto-Gmc. /e, a/ > 0 / _ # pre-NWGmc. /e1/ > /a/ NWGmc. (not yet defined) While Antonsen believes that the Germanic accent shift must have followed the consonant shift described by Grimm's Law, Lehmann (1968: 74) holds that the fixation of stress and Grimm's and Yerner's Laws should not be assigned to separate stages of Proto-Germanic, since "the sound change described by Verner affected the structure of Germanic only when the accent was fixed." Bennett (1968) agrees that the traditional chronology of consonant shift —• development of voiced allophones —> accent shift deserves a closer examination. He questions the reasoning which underlies the traditional order: Period 1. /p t k kw/ > (IE)

Period 2. /f χ xw s/ > (Gmc)

Period 3. /b d g g w z/ (Gmc)

Bennett believes, rather, that the voiced allophones arose in the same time frame as the voiceless ones, alternating with them when the preceding vowel was produced with "less effort", i.e., when stress did not immediately precede it. Ramat (1981: 391), too, believes that the voicing described by Verner's Law must have arisen at the time of the consonantal shift, since information about the placement of the accent in Indo-European would not be available after that time. Ramat thus rejects Verner's original chronology *p > (p h ) > f > b in favor of *P < f · 4 2 Any chronology arrived at by means of internal reconstruction must be carefully scrutinized, for it is as susceptible to error and reinterpre-

The chronology of phonological

change

127

tation as other methods are. For example, there are several possible ways to describe the development of Lithuanian palatals, but as Hock notes (1986: 550 — 555), each of them entails certain problems. Likewise, the exclusive use of internal reconstruction in analyzing the primacy of Sanskrit /k/ vs. /c/ would lead to an incorrect chronology: /c/, having a more restricted distribution, would be assumed to be the original phoneme, but this conclusion is contradicted by the comparative evidence (Miranda 1975: 300 f.). Internal reconstruction is a valuable tool for filling in gaps in a chronology, but is most reliably used in conjunction with other methods of analysis, whenever this is possible.

3. Conclusion Clearly, the establishment of a chronology of sound change is a complicated undertaking. Not only are precise dates seldom available, but documentary evidence is often scant and subject to variant interpretations. "Scribal errors" may be significant, but are not necessarily so. Documents, especially official ones, often reflect an earlier stage of the language, seldom the current vernacular. Even the most casual graffiti may well have been influenced by the written norm. In addition, sociolinguistic evidence, which is essential to any complete historical analysis, is often difficult to assemble, especially for prehistoric societies, and more abstract evidence, from rule-ordering, internal reconstruction, etc., can only be tentatively interpreted if the data is limited. Despite all these difficulties, it is still often possible to construct satisfactory chronologies. Just as it would be wrong to form hasty conclusions based on insufficient evidence, it would also be wrong to ignore the suggestive hints provided by grammarians, foreign scribes, poets, and unconventional spellers. Only by considering, in the most cautious and judicious manner possible, the entire array of available evidence can a fair representation of linguistic change be obtained.

128

Bridget Drinka

Notes 1. Epistulae ad Famiiiares IX, 21, 2. 2. The forms which apparently did not undergo the shift have various explanations (Palmer 1954: 37, 230):

3.

4.

5.

6.

7.

8. 9.

caesar: a loan from Sabine, which did not undergo rhotacism (cf. Sabine ausum = Roman aurum) Besides these, archaic forms like arbosem, lasibus persist. Meillet (1966: 85, 142) explains the retention of in such words as rosa, asinus, and miser as an indication that [s] had already become [z] intervocalically before these forms were borrowed, while Maniet (1957: 135) points out that a non-contiguous /r/ often precludes rhotacism elsewhere in the word (e.g., Pisaurum). See also Sommer 1977: 146 — 147. See, for instance, Stephenson (1967: 39—41) for examples from eighteenth-century North Carolina: hollar 'hollow', bedstids 'bedsteads', taffity 'taffeta', hypercorrect lining 'linen', and so on. While Classical Latin tolerated s + stop (cf. schola), Vulgar Latin speakers must have already begun to feel the need for a prothetic vowel, as seen also in its widespread use in the Romance languages. (Span, escuela, OF. escole, Ital. iscuola), and in later naive spellings in other Latin-speaking areas (e.g., ispose [< sponsae] on an African gravestone) (Elcock 1960: 25 — 27; Leumann 1977:104, with additional references). See Väänänen 1966: 74 for numerous examples. The havoc which the loss of this marker /-m/ created caused it to be restored in the formal writings of the classical age, but it was still frequently absent in Vulgar Latin, as evidenced by its nonoccurrence, outside of monosyllables, in Romance languages. marks the site of an expected but absent ; | represents a new line; the of nON was mutilated; the interpretation of II is complex: see Väänänen 1966: 22 f., 36. Hall (1976: 12) suggests that the alternation seen here between [au] and [o:] existed in Latin over a long period, from as early as Republican times, [au] forms are thus sometimes younger than [a:] forms: the [o~] in Proto-Romance /kcTda/ 'tail' can be reconstructed based on all the Romance languages, but the [au] seen in several Romance languages (e.g., Romanian) in forms like /auru/ 'gold' must have been influenced by Classical Latin. See also Allen's discussion (1965: 60 — 62), and that of Meyer-Lübke (1972: 171). Many Romance forms are based on such reductions (e.g., It. maschio, Sp. macho, Fr. male). The interplay of classical and vulgar Latin is probably significant here as well — very early inscriptions often did not show an g regularly in "western Roman", making it difficult to judge if this is truly a Celtic phenomenon. The transition of m > ν in Celtic is discussed in great detail by Förster (1941: 618 — 682). He adopts the chronology, am > aw > äw > äw > äv > αν (681), which does not include the somewhat doubtful nasalized [v] of Jackson's interpretation. He is referring to Old Low East Franconian, easternmost dialect of Dutch, which was spoken in parts of present-day Germany, as opposed to the dialect spoken nearer the coast, which is often referred to as Old Low West Franconian. Along similar lines, Gysseling (1960: 8) notes that Netherlands toponyms recorded in abbeys in High German territory often underwent the High German consonantal shift. Ancient Babylonian scribes likewise placed a new interpretation on a nondistinctive set of signs from Old Akkadian: and , which were used interchangeably in Old Akkadian to represent velar stops before u, came to represent a voiceless/voiced distinction in Babylonian syllabaries (Wilhelm 1983: 162). It is not even clear that the Irish were responsible for teaching the English to write English: Their instruction must have been long completed by the time entire documents appeared in English rather than in Latin, in the eighth century. The First Grammatical Treatise, written in Old Icelandic in the mid-twelfth century, represents a notable exception. The anonymous author displays a thorough understanding of the phonological system of his language as he attempts to reform its orthography (Haugen 1972; Benediktsson 1972). The fact that such reform was

130

19.

20.

21. 22.

23.

Bridget Drinka necessary and that his suggestions were not heeded, however, supports Schmalstieg's view. Caxton was fairly consistent in spelling "French" as frensshe or frenshe from 1475 — 1488 (with an interesting occasional divergence to ffrensshe, ffrensh(e) from 1475 — 1483). His spelling of "English" also shows a trend, though a greater amount of variation exists: 1475 — 1481 usually englissh 1477 englyssh ~ englissh (This alternation was found in one text — in the first and second editions of his translation of The Dictes or Sayengs of the Philosophres) 1481 —1490 englysshe greatly predominates Then a new trend seems to set in, towards single : 1489 english(e), englysh(e) Thus while a great many different spellings for a single word are attested, there is systematicity in the variation, and consistency within time periods. That this change was already accomplished is indicated especially by the unetymological use of in hypercorrections: VEIVANT (cf. Lat. viv- < IE* gwTw-) OCTOBREIS (Ts < *ms) (Väänänen 1966: 22). See also Pulgram 1958: 281 for a concurrent view. Orthoepists are not always reliable informants, of course. Russ (1986: 164ff.) presents several examples of imprecise or even mistaken judgments made by German orthoepists. See also Russ's examples of orthographic, poetic, and other evidence, much of which supports conclusions presented in this paper. See Väänänen 1981: 200 ff. for the complete list, along with additional bibliography and a discussion of the possibility of a later date for the Appendix. Elcock (1960: 29) believes that proper spelling, not pronunciation, is probably the goal of the work; he cites the following evidence: GYRUS non GIRUS CRISTA non CRYSTA VIRGO non VYRGO Elcock assumes that there was no distinction in pronunciation of these forms. However, as Leumann (1977: 51) notes, does tend to appear more frequently after labials, suggesting an /Ü/ pronunciation. Spelling is probably closely linked with pronunciation in most cases. Pulgram (1958: 317 ff.) adds that an important problem with such compilations is that it is very difficult to judge the frequency of a given "error".

24. To those who claim that /aw/ ( < ü) and /ay/ ( < I) existed but were simply not represented in the orthography, Wolfe counters that Robinson created a new set of symbols, so that he was evidently wary of orthographic influence, and probably avoided it. Besides, orthoepists before Hodges (1643) used /ey/, /ow/; only after Hodges did they begin to use /ay/ and /aw/. It seems unlikely, she suggests, that the /a/ suddenly became more perceptible. 25. They provide substantial evidence throughout the report that present-day languages do show such parallel shifting, and that languages are able to maintain extremely minute distinctions for long periods of time. 26. See also Labov —Yaeger—Steiner's interesting discussion of the inability of speakers to consciously distinguish certain sounds which they regularly distinguish in speech

The chronology of phonological change

27.

28.

29. 30.

31.

32.

33.

34. 35.

36. 37.

38. 39.

131

(chapter 6) and the implications this has for orthoepic interpretation (cf. especially p. 296). Penzl also believes that recent splits and mergers which result in rhymes are often considered unacceptable by speakers of the language if the traditional orthography is retained. For example, he says that the British do not usually make rhymes like water, quarter, and the French do not consider nous to rhyme well with loup. However, Martinon (1915: 268 — 69) does list nous and loup as rhymes. He mentions (pp. 38 — 39) an earlier reluctance by speakers to rhyme words ending in non-similar "silent consonants" (for example, and were permissible, but not and

), but it seems that this tendency could have been due to overconcern with spelling consistency rather than being a reflex of phonological change. This example illustrates once again the danger of relying too heavily on spelling for information about chronology: the grapheme was a restoration based on Classical Latin. According to his view, [k] and [γ] but not [g] existed in Germanic at the time of this borrowing. Toon writes the general variable rule as follows (p. 106): a —» / < + apical) [+ nas] Note that angle brackets are used here to indicate factors which covary in a variable rule; elsewhere in this paper, they have been used to represent orthographic, as opposed to phonological, representations. Examples include the close ties between the Mercian kings vEöelbald and Offa and the archbishops of Canterbury, the ability of Offa to grant land in Kent to his own archbishop in Lichfield, Offa's title as "King of all England" in a Kentish grant, and so on (p. 28 ff.). "Since Mercian kings made bishops and controlled the granting of land and other privileges, there must have been constant intercourse between Canterbury and Lichfield which no doubt included exchange of scribes, illuminators, and books as well as bishops." (p. 79 f.) For example, "second fronting" (the raising of [ae] to [e]) began to appear in Kentish charters at the time of Mercian dominance (p. 150 ff.; 159 f.), as did velar umlaut of [ae] (p. 156). Evidence for several of these stages still exists today: the Mod.Fr. spelling reflects the pronunciation of c. 1150; the form [w?] from c. 1300, is still used in Canadian French (Elcock 1960: 360). Fouche dates this change to the second half of the thirteenth century. For information about Villon's poetry and playful poetic devices, see Guiraud (1970); for his connections with the criminal elements of Paris, see the short biography in Bonner's translation (1960) of Villon's complete works. It was not until the end of the 18th century that Parisian grammarians accepted this pronunciaton, and even then they did so with reservations (Fouche 1958: 272). Spelling gives no clue here, one way or the other: Twenty conditional forms used by Villon (collected in Van Deyck and Zwaenepoel 1974: 628) are all still spelled . Guillaume des Autels (1631) noted that women wishing to imitate courtesans would say "il fait fret" (for froid) (Fouche 1958: 277). H. Estienne (1582) said, disapprovingly, that courtesans found the "Italian-style" words (with [?]) to be "de trop meilleure grace, pour ce qu'ils sont plus mignards et qu'ils ne faut pas que les dames ouvrent tant la bouche" (Fouche, loc. cit.).

132

Bridget Drinka

40. Pulgram (1958: 314) points out that even graffiti may not represent the spoken language very well, since at least some schooling is required to be able to write it, and the prestige of the written language may cause even graffiti-writers to try to "better" their spelling and style. 41. k > x/ C (Reichelt 1909: 43). 42. If one accepts the hypotheses put forward by Gamkrelidze and Ivanov (1984) or by Hopper (1973, 1982), which challenge the traditional characterization of the Proto-Indo-European phonological system, the operation of Verner's Law will, of course, be viewed quite differently.

References Allen, W. Sidney 1965 Vox Latina. Cambridge: University Press. 1968 Vox Graeca. Cambridge: University Press. Amos, Ashley Crandell 1980 Linguistic means of determining the dates of Old English literary texts. Cambridge, Mass.: The Medieval Academy of Mass. Anderson, James M.—JoAnn Creore (eds.) 1972 Readings in Romance Linguistics. The Hague: Mouton. Anderson, Stephen R. 1978 "Historical change and rule ordering in phonology", in: Jacek Fisiak (ed.), 23-41. Antonsen, Elmer H. 1965 "On defining stages in prehistoric Germanic", Language 41: 19 — 36. Augst, Gerhard (ed.) 1986 New trends in graphemics and orthography. Berlin/New York: Walter de Gruyter. Baugh, Albert C. —Thomas Cable 1978 A history of the English language. (3rd edition) Englewood Cliffs, N.J.: Prentice Hall, Inc. Benediktsson, Hreinn 1972 The first grammatical treatise. Reykjavik: Institute of Nordic Linguistics. Bennett, William 1968 "The operation and relative chronology of Verner's law", Language 44: 219-223. Bonner, Anthony (trans.) 1960 The complete works of Frangois Villon. New York: Bantam Books. Brosman, Paul W., Jr. 1979 "Low Franconian Τ in Old French", Zeitschrift fur französische Sprache und Literatur 89: 237 — 256. Cercignani, Fausto 1981 Shakespeare's works and Elizabethan pronunciation. Oxford: Clarendon Press. Chomsky, Noam—Morris Halle 1968 The sound pattern of English. New York: Harper & Row.

The chronology of phonological change

133

Coulmas, Florian — Konrad Ehlich (eds.) 1983 Writing in focus. Berlin/New York/Amsterdam: Mouton. Crotch, W. J. B. 1928 The Prologues and Epilogues of William Caxton. London: Oxford University Press. Daunt, Majorie 1968 "Old English sound-changes reconsidered in relation to scribal tradition and practice", in: Charles T. Scott—Jon L. Erickson (eds.), 394—414. (First published in Transactions of the Philological Society 1939.) Dingwall, W. O. (ed.) 1978 A survey of linguistic science (2nd edition). Stamford, Conn.: Greylock Publishers. Dobson, E. J. 1957 English pronunciation 1500 — 1700. Vols. I —II. Oxford: Clarendon Press. Dressier, Wolfgang 1973 "Pour une stylistique phonologique du latin ä propos des styles negligents d'une langue morte", Bulletin de la Societe de Linguistique de Paris 68: 129-145. Elcock, W. D. 1960 The Romance languages. New York: Macmillan Co. Fisiak, Jacek (ed.) 1978 Recent developments in historical phonology. The Hague: Mouton. Förster, Max 1941 Der Flussname Themse und seine Sippe. München: Verlag der Bayerischen Akademie der Wissenschaften. Fouche, Pierre 1958 — 1966 Phonetique historique du franfais. Vols. II —III. Paris: Librairie C. Klincksieck. Fourquet, J. 1948 Les mutations consonantiques du germanique. Paris: Les Belies Lettres. Gamkrelidze, Thomas V. — Vjaceslav V. Ivanov 1984 Indoevropejsky jazyk i indoevropejcy. Rekonstrukcija i istoriko-tipologiceskij analiz prajazyka i protokul'tury. 2 vols. Tbilisi: Izdatel'stvo Tbilisskogo Universiteta. Guiraud, Pierre 1970 Le Testament de Villon. Paris: Gallimard. Gysseling, Maurits 1960 Toponymisch woordenboek van Belgie, Nederland, Luxemburg, n. p.: Belgisch Interuniversitair Centrum voor Neerlandistiek. Hall, Robert Α., Jr. 1974 External history of the Romance languages. New York: American Elsevier Publishing Co., Inc. 1976 Proto-Romance phonology. New York: American Elsevier Publishing Co., Inc. Haugen, Einar 1972 First grammatical treatise. London: Longman Group Ltd. Hock, Hans Henrich 1986 Principles of historical linguistics. Berlin/New York/Amterdam: Mouton de Gruyter.

134

Bridget Drinka

Hoenigswald, Henry M. 1960 Language change and linguistic reconstruction. Chicago: University of Chicago Press. Hopper, Paul J. 1973 "Glottalized and murmured occlusives in Indo-European", Glossa 7: 141-166. 1982 "Areal typology and the early Indo-European consonant system", in: Edgar C. Polome (ed.), 121 - 1 3 9 . Jackson, Kenneth 1953 Language and history in early Britain. Edinburgh: University Press. Kesselring, Wilhelm 1973 Die französische Sprache im Mittelalter von den Anfängen bis 1300. Tübingen: Narr. Kiparsky, Paul 1978 "Historical Linguistics", in: W. O. Dingwall (ed.), 3 3 - 5 7 . Kökeritz, Helge 1953 Shakespeare's pronunciation. New Haven: Yale University Press. Kuhn, Sherman Μ. —Randolph Quirk 1953 "Some recent interpretations of Old English digraph spellings", Language 29: 143-156. Kyes, Robert L. 1979 "The phonological significance of orthographic variation: Proto-Germanic /AW/ in Old Netherlandic", in: Irmengard Rauch —Gerald F. Carr (eds.), 231-239. Labov, William — Malcah Yaeger—Richard Steiner 1972 A quantitative study of sound change in progress. (Report on N.S.F. Contract NSF-GS-3287) Philadelphia: University of Pennsylvania. Lass, Roger 1978 "Mapping constraints in phonological reconstruction: On climbing down trees without falling out of them", in: Jacek Fisiak (ed.), 245 — 286. Lehmann, Winfred P. 1968 "A definition of Proto-Germanic: A study in the chronological delimitation of languages", in: Charles T. Scott—Jon L. Erickson (eds.), 66 — 75. (First published in Language 37 [1961]: 6 7 - 7 4 ) . Leumann, Manu 1977 Lateinische Laut- und Formenlehre (Revised edition). München: Beck. Lommatzch, E. (ed.) 1918 Corpus Inscriptionum Latinarum (Revised edition) (referred to under the standard abbreviation CIL). Maniet, A. 1957 L'evolution phonetique et les sons du latin ancien. Louvain: Editions Nauwelaerts. Martinet, A. 1972 "La lenition en celtique et les consonnes du roman occidental", in: James M. Anderson —Jo Ann Creore (eds.), 228 — 261. (First printed in Economie des changements phonetiques. München: A. Francke 1955). Martinon, P. 1915 Dictionnaire methodique et pratique des rimes frangaises. Paris: Librairie Larousse.

The chronology of phonological change

135

Meillet, Antoine 1966 Esquisse dune histoire de la langue latine. Paris: Editions Klincksieck. Meyer-Lübke, Wilhelm 1972 Romanisches etymologisches Wörterbuch. Heidelberg: Carl Winter. Miranda, Rocky V. 1975 "Internal reconstruction: Scope and limits", Lingua 36: 289 — 306. Palmer, Leonard R. 1954 The Latin language. London: Faber & Faber, Ltd. 1980 The Greek language. Atlantic Highlands, N.J.: Humanities Press. Penzl, Herbert 1968 "The evidence for phonemic changes", in: Charles T. Scott — Jon L. Erickson (eds.), 97 — 112. (First published in Studies presented to Joshua Whatmough on his sixtieth birthday. Ed. Ernst Pulgram. The Hague: Mouton, 1957; 193-208). 1972 Methoden der germanischen Linguistik. Tübingen: Max Niemeyer Verlag. Polome, Edgar C. (ed.) 1982 The Indo-Europeans. Third and fourth millennia B.C. Ann Arbor: Karoma. Pulgram, Ernst 1958 The tongues of Italy. Cambridge, Mass.; Harvard University Press. Ramat, Paolo 1981 Einführung in das Germanische. Tübingen: Max Niemeyer Verlag. Rauch, Irmengard — Gerald F. Carr 1979 Linguistic method: Essays in honor of Herbert Penzl. The Hague: Mouton. Reichelt, Hans 1909 Awestisches Elementarbuch. Heidelberg: Carl Winter. Russ, Charles 1986 "Breaking the spelling barrier: The reconstruction of pronunciation from orthography in historical linguistics", in: Gerhard Äugst (ed.), 164 — 178. Schmalstieg, William R. 1981 "Old Prussian and Hittite orthographic variants: A parallel", PontoBaltica I: 119-124. Schönfeld, Μ. 1911 Wörterbuch der altgermanischen Personen- und Völkernamen. Heidelberg: Carl Winter. Schrodt, Richard 1976 Die germanische Lautverschiebung und ihre Stellung im Kreise der indogermanischen Sprachen (2nd edition). Vienna: Verlag Karl M. Halosar. Scott, Charles T. — Jon L. Erikson (eds.) 1968 Readings for the history of the English language. Boston: Allyn & Bacon, Inc. Scragg, D. G. 1974 A history of English spelling. Manchester: Manchester University Press. Serjeantson, Mary S. 1935 A history of foreign words in English. London: Kegan Paul, Trench, Trubner, & Co., Ltd. Slotty, Friedrich 1960 Vulgärlateinisches Übungsbuch. Berlin: de Gruyter.

136

Bridget

Drinka

Smith, M. Estelle (ed.) 1972 Studies in linguistics in honor of George L. Trager. The Hague: Mouton. Sommer, Ferdinand 1977 Handbuch der lateinischen Laut- und Formenlehre (4th revised edition, with Raimund Pfister.) Heidelberg: Carl Winter. Sommerfeit, Alf 1953 — 1958 Review of Language and history in early Britain, by Kenneth Jackson. Lochlann: A Review of Celtic Studies 1: 278 — 283. Stephenson, Edward A. 1967 "On the interpretation of occasional spellings", Publication of the American Dialect Society 48: 33 — 50. University of Alabama Press. Stockwell, Robert P. 1972 "Problems in the interpretation of the Great Vowel Shift", in: M. Estelle Smith (ed.), 3 4 4 - 3 6 2 . 1978 "Perseverance in the English Vowel Shift", in: Jacek Fisiak (ed.), 337 - 348. Toon, Thomas E. 1983 The politics of Early Old English sound change. New York: Academic Press. Väänänen, Veikko 1966 Le latin vulgaire des inscriptions pompeiennes. Berlin: Academie Verlag. 1981 Introduction au latin vulgaire. (3rd ed.). Paris: Editions Klincksieck. Van Deyck, Rika —Romana Zwaenepoel 1974 Fra^ois Villon, Oeuvres. (Vol. II). Saint-Aquilin-de-Pacy (Eure): LibrairieEditions Mallier. Wilhelm, G e m o t 1983 "Reconstructing the phonology of dead languages", in: Florian Coulmas —Konrad Ehlich (eds.), 157 — 166. Wolfe, Patricia M. 1972 Linguistic Change and the Great Vowel Shift in English. Berkeley, Los Angeles: University of California Press. Wright, Roger 1976 "Speaking, reading, and writing late Latin and early Romance", Neophilologus 60, No. 2: 1 7 8 - 1 8 9 .

Linguistic paleontology: Migration theory, prehistory, and archeology correlated with linguistic data Edgar C. Polome

From the very beginning of the development of linguistic science in the nineteenth century, scholars have been concerned with the problem of reconstructing the past on the basis of linguistic data. The way was paved by Rasmus Rask in his Undersögelse om det gamle Nordiske eller Islandske Sprogs Oprindelse (1818) in which he regrouped a sizable set of corresponding terms in different Indo-European languages according to semantic categories in order to show that these words belonged to their common core vocabulary. However, the foundation of linguistic paleontology was actually laid by Adalbert Kuhn with his book Zur ältesten Geschichte der indogermanischen Völker (1845), in which he attempts to describe the culture of the Indo-Europeans in correlation with the lexicon of the protolanguage. A long series of rather uncritical studies followed until it was recognized that the rather glamorous romantic depiction of the Indo-European past rested on fragile bases, as the enthusiastic authors failed to recognize the harsh realities of life in prehistoric societies and were often unable to identify wandering loanwords designating newly introduced cultural items within the inherited vocabulary of common Indo-European stock. The problem was approached with sound scepticism and sharp critical judgement by Victor Hehn in his Culturpflanzen und Haustiere in ihrem Übergang aus Asien nach Griechenland und Italien sowie dem übrigen Europa (1870), which had a considerable influence on the later research. But what affected the quality of further investigation most deeply was the tremendous impact of the "Neogrammarians" on Indo-

138

Edgar C. Polome

European studies. This is reflected, for example, by Otto Schrader's Sprachvergleichung und Urgeschichte (1883) and especially by his Reallexikon der indogermanischen Altertumskunde (1901), of which Alfons Nehring prepared a considerably improved second edition. In the early twentieth century, the problems of the homeland of the Indo-Europeans, their culture, and their migrations were still the object of lively discussion, and a couple of major syntheses were attempted before World War I, namely the two volumes on Die Indogermanen, ihre Verbreitung, ihre Urheimat und ihre Kultur by Herman Hirt, and Sigmund Feist's Kultur, Ausbreitung und Herkunft der Indogermanen (1913). For several decades, no new scholarly effort was attempted to integrate all the facts into a unified picture of the Indo-European past, as active research and sometimes bitter controversy went on about the location of the Indo-European homeland, about the socio-economic structure of their society, and about their level of culture. This is illustrated by two volumes of substantial essays published in 1936: Germanen und Indogermanen — Festschrift fur Herman Hirt, edited by Helmut Arntz, and Die Indogermanen- und Germanenfrage — the fourth volume of the Wiener Beiträge zur Kulturgeschichte und Linguistik — edited by Wilhelm Koppers. Unfortunately, in this period, views on the Indo-European past were strongly affected by political bias as a result of the racist policies of the German Third Reich. On the basis of the following statement of Tacitus: "Ipsos Germanos indigenas crediderim minimeque aliarum gentium adventibus et hospitiis mixtos" (Germania, chapter 2) [I would believe that the Germanic people themselves are autochthonous and have only been exposed to minimal admixture through immigration and friendly accommodation of other peoples], it was arrogantly assumed that the Germanic people alone, among all the Indo-European peoples, remained in the original homeland in northern Central Europe, and all kinds of arguments were then cleverly designed to back up this politically motivated assertion. (All that Tacitus presumably wanted to emphasize was the stability of the Germanic population; note no archeologically detectable population movement appears to have taken place in their area since Neolithic times). After World War II a strong need was felt to take stock of the whole question of Indo-European origins with the related problems of protoculture and diffusion, and to reexamine carefully the methodological approaches used to cope with these. To provide a background to this essential task, Anton Scherer (1968) compiled a most useful collection

Linguistic paleontology

139

of articles by leading archeologists and comparativists (originally published from 1892 to the early sixties). In a paper on "The comparative method for reconstruction in linguistics", written especially for Dell Hymes' Language in culture and society (1964: 585 — 586), the German Sanskritist Paul Thieme reviewed the evidence used to demonstrate genetic relationship, focusing on semantic correspondences to reconstruct the semantic system of Proto-Indo-European. His analysis of the terminology of the fauna and flora leads him to the conclusion that it is "possible to locate the area where our postulated language was spoken — within the domain of salmon rivers and tributaries, to the west of the 'beech line', outside Scandinavia". As the beech apparently did not grow beyond a line that runs from Kaliningrad (the former Königsberg), on the Baltic, to Odessa, on the Black Sea, this would situate the Indo-European homeland in the basins of the Vistula, Oder, and Elbe, where the ancestors of the Baits and the Slavs met those of the Germanic people (1964: 597). Meanwhile, the anthropologist Morris Swadesh, concerned about determining the time depth of the separation of genetically related languages from the common stock, suggested that the core vocabulary of a language, which he assumed not to be culturally conditioned — body parts, numerals, geographical features, etc. — is relatively stable and replaced at a constant rate through time, while the cultural vocabulary can change with contact, borrowing, diffusion of new techniques and products, or other reasons. He therefore claimed that the computation of the percentage of true cognates within the core vocabulary of two languages would make it possible to determine at what time they went their separate ways, provided unforeseeable factors like conquests, migrations, close social contact with alien elements, etc., did not unbalance the whole picture. His method, called "lexicostatics" or "glottochronology" was widely acclaimed in the 1950s and 1960s, and employed with genuine confidence by anthropologists and linguists working on languages without documented past (cf., e. g., Swadesh 1951,1959; Lees 1953; Gudschinsky 1956; Hymes 1960; Dyen, 1962 a/b; etc.). It had, however, some fundamental flaws: the basic word-list contains grammatical items like pronouns which could not readily be rendered in some of the languages investigated; it does not take into consideration that the nomenclature of parts of the body may be different in some cultures where, for example, the same term can be used for "hand" and "arm", or for "foot" and "leg;" it neglects some climatic considerations which make words like "freeze", "ice",

140

Edgar C. Polome

and "snow" irrelevant to definite language groups, while others might not be familiar with the concept of "tree". It fails to recognize some basic semantic principles such as the possibility of the existence of a number of specific words for different ways to hunt, while there is no particular term for hunting itself. Taboo and similar phenomena are apparently ignored, although they are likely to affect terms like "liver", "eye", or "tongue" in societies in which incantations, curses, magic spells, and divination play a prominent role. In spite of revisions of the basic wordlists, several problems remained, which a number of critical articles pointed out, e.g., Hoijer 1956; Rea 1958; Bergsland — Vogt 1962, etc. If, undoubtedly, the sharing of a common basic vocabulary is a probable indication of genealogical relationship, the hypothesis of a constant rate of retention from language to language remains unsubstantiated (Lehmann 1973: 106; Bynon 1977: 270), and, accordingly, Swadesh's method is to be considered as utterly unreliable in dating the separation of subgroups within a language family. This has been clearly illustrated for the Indo-European languages by the absurd results its careful application to the core lexicon of the protolanguage and its reflexes in the various dialects has yielded in two investigations conducted quite independently from each other by E. Polome (1964) and J. Tischler (1973). The lexicon was the basis of a different approach to relative chronology which was introduced by the Italian school of linguistics known as "Neolinguistica". Briefly presented in the 1920s by Matteo Bartoli in his Introduzione alia Neolinguistica (Principi — Scopi — Metodi) and in the Brevario di neolinguistica (which he wrote together with Giulio Bertoni in 1928), this new approach was based on the methodology and results of linguistic geography and focused on three problems: a) the chronological relation between stages of linguistic development; b) the centers of diffusion of innovations; c) the causes of innovations. The chronology of linguistic change ways, according to Bartoli: either on or through the geographical layout occur. To assess the importance and lished four norms (normi spaziali):

can only be established in two the basis of datable documents, of the areas in which the data meaning of the latter, he estab-

1. norma dell'area meno esposta ("norm of the less exposed area"), implying that a more isolated area will tend to be more conservative;

Linguistic paleontology

141

2. norma delle aree laterali ("norm of the lateral areas"), stating that peripheral areas are likely to be less (or not at all) affected by innovations taking place in a central area; 3. norma dell'area maggiore ("norm of the more extended area"), assuming that the oldest form will often tend to be spread over the largest area of the linguistic territory; 4. norma dell'area seniore ("norm of the older area"), pointing out that new territories opened to a language, e. g., colonies, may preserve older forms, lost by innovation in the homeland. Moreover, Bartoli also considered the cases where the old term had been completely ousted by a new one. His views were illustrated by examples from the development of the Romance languages, but in an appendix comparing his method with that of the Neogrammarians, he indicated that his approach could validly be applied to Indo-European. Although the weaknesses of these views were already stressed as early as 1931 by Hannes Sköld in his Beiträge zur allgemeinen und vergleichenden Sprachwissenschaft. I. Sprachgeographie und Indogermanistik, Italian Indo-Europeanists like Vittore Pisani (1933, 1940) espoused them enthusiastically. While Giuliano Bonfante remained fairly close to Meillet's presentation of the Indo-European dialects (1908, 1922) in his 1931 book, he became the forceful advocate of the Neolinguistica in the States around the middle of the century, presenting Bartoli's method (1945, 1946) and defending it, with Vidossi (1948), against the severe criticism of Robert A. Hall, Jr. (1946). Notwithstanding the rejection of the Neolinguistic approach by Hall (1963), it had a definite impact in the 1950s on research on the split of Indo-European into dialects and their dispersal: the criterion of the marginal areas is accepted unconditionally (einwandfrei) by W. Porzig (1954: 57), and W.J. Entwistle (1953: 90) approves of Bartoli's (1945: 1 - 3 1 ) application of his "norms" to assume the relative archaism of Latin within the Indo-European language family. However, L. R. Palmer (1954: 23 — 32) dismisses these views after a thorough critical examination. While linguistic geography can help comparative linguistics in establishing relationship models (cf., Anttila 1972: 304 — 307), areal linguistics also examines isoglosses, shared by a set of geographically contiguous languages, denoting features resulting from prolonged mutual contact rather than from common heritage: Meillet (1908: 21 — 39), following the tree diagram of derivation of the Indo-European dialects, assumes that Latin and Osco-Umbrian originally constituted

142

Edgar C. Polome

an Italic linguistic community after the breakup of the "Italo-Celtic unity". This view was disputed by A. Walde in his Über älteste sprachliche Beziehungen zwischen Kelten und Italikern (1917), and the Italian linguists, especially Giacomo Devoto (1931, 1940), followed suit, considering Latin and Osco-Umbrian as two separate Indo-European dialects whose common features resulted from convergence on Italian soil. Palmer (1954: 10 — 11) agrees with this view, but wants to account for it by examining the socio-cultural factors responsible for the convergence — a phenomenon that has received growing recognition in recent literature (cf., e.g., Hock 1986: 4 9 1 - 5 1 2 , 561, 565-566). M. Beeler (1966: 58), however, feels that the paucity of Osco-Umbrian materials prevents us from deciding against the "Proto-Italic" hypothesis, but he concedes that "the differences between the two languages [i. e., Latin and Osco-Umbrian] are deep and far-reaching", whereas Pulgram (1958: 134) sees their differentiation as a result of centuries of uninterrupted, slow Indo-European immigration and gradual superimposition on the aborigines of Mediterranean stock. Convergence areas are often referred to as Sprachbund when the languages involved are structurally distinct and not necessarily genetically related, but have developed their common features as a result of prolonged contact: a classic example is the case of the Balkan languages where Greek, Albanian, Rumanian, and the Balkan Slavic languages such as Bulgarian, Macedonian, and Serbo-Croatian share a number of characteristic correspondences (cf., e.g., Schaller 1975: 9 6 - 1 9 0 ; Solta 1980: 180-234). Typological studies help us assess these phenomena of convergence as they illustrate parallel features and patterns in genetically not directly related languages like Dravidian and Indo-Aryan in India (cf. Masica 1976; Shapiro — Schiffman 1981: 116 — 149): here, the process of convergence may have started with the arrival of the Aryans in the South Asian subcontinent about three and a half millennia ago (cf. Polome 1983: 13 — 19). As it has been assumed that the idiosyncracies of Hittite versus the other Indo-European languages were ascribable to the non-Indo-European languages they overlay or were in close contact with (Sommer 1947: 39 — 100; cf. Polome 1983: 10 — 13), one might wonder if some kind of Anatolian Sprachbund may not have developed in the Hittite kingdom? Similarly, when distant kinship is assumed between Indo-European and Uralic which were neighbors since prehistoric times, is it not a plausible to ascribe to convergence (as Hock [1986: 565 — 566] suggests) corre-

Linguistic paleontology

143

sponding structural features like inflectional and derivational suffixes, or SOV syntax? The basic concept dominating linguistic paleontology is, however, divergence and split into various subgroups from a common stock: as in the past, the main interest of scholars in the second half of the twentieth century was to reconstruct the protolanguage and, through it, the culture that it reflected, and to try to locate it in time and space, as well as to map the diachronic dispersion of its dialects. An ambitious new synthesis was provided in 1962 by Giacomo Devoto, who examines in detail the linguistic, anthropological, archeological, cultural, and historical problems relevant to the Indo-Europeans and their expansion before classifying and analyzing their common lexicon according to semantic categories. The work also contains a comparative wordlist covering 969 terms with their reflexes in (Old)Irish, Welsh, Latin, (Osco-)Umbrian, Old High German, Gothic, Greek, Hittite, Armenian, Lithuanian, Old Church Slavonic, Avestan, Sanskrit, and Tocharian A, on which he comments in the text. Correlating the archeological data with a Neolinguistic patterning of the linguistic material, Devoto assumes (1962: 90) that the original territory of the Indo-Europeans covered a vast area including Central Germany, Bohemia, Moravia, Slovenia, Hungary, Galicia, Silesia, Transylvania, Moldavia, and the Ukraine; he furthermore claims that Indo-European emerged as a speech community in the Neolithic period, but reached its full development in the Bronze Age with the expansion of the urnfields civilization. While Devoto's work was to a large extent summarizing the achievements of the recent decades (in archeology, he relied rather heavily on the writings of Pia Laviosa-Zambotti on the diffusion of agriculture [1943], earliest North-European cultures [1941], and the southward migrations of the Indo-Europeans [1950]), a different approach to the study of the Indo-European lexicon was initiated by Carl Darling Buck in his Dictionary of selected synonyms in the principal Indo-European languages (1949). Buck appropriately subtitled it: "A contribution to the history of ideas", as it is, in the words of Calvert Watkins, "a treasure house of words, word origins, expressions, and ideas in most of the great languages of the Indo-European family from India to Iceland and from their earliest attestation down to the present day". It is indeed the only semasiologically arranged volume that provides an adequate Wörter und Sachen treatment of over a thousand items

144

Edgar C. Polome

with the relevant etymological and semantic history, making it a most valuable source of information for research on Indo-European culture. While Devoto and Buck supply particularly useful lexical tools, Emile Benveniste (1969) investigates some semantic fields in depth within the context of early Indo-European societies, analyzing their economic terminology, for example, to define the procedures of trading, exchange, payment, lending and borrowing, etc.; examining the legal position of individuals within the societal context through the kinship terminology, the verbal specification of his/her function and status, the vocabulary describing the ruling power and its limitations, as well as the contractual obligations of members of the community, etc.; probing into ritual acts, divination, and the concept of the "holy". His brilliant seminal work has served as a model for later attempted syntheses of Indo-European culture such as Haudry's (1981) and Sergent's (1987), although some of his views on the Indo-European social structure have also been justifiably criticized (e. g., Zimmer 1987; cf. also Buti 1987). In the meantime, a new hypothesis on the origin of the IndoEuropeans had been formulated by the American archeologist Marija Gimbutas who ascribed the cultural changes taking place in Europe and the Near East in the fourth and third millennia B. C. to the "invasion of hordes of pastoralists from the North Caucasian — lower Volga steppe area" — the Kurgan or Barrow culture which she identified (1970) as Proto-Indo-European. In a number of publications (1973, 1977, 1980, 1985), she has since then detailed how the "waves" of invaders affected the populations and civilizations of "Old Europe", submerging their language, religion and culture. She distinguishes three phases in the penetration of the Kurgan people: 1) 4400—4200 B.C.; 2) 3400-3200 B.C.; 3) 3000-2800 B.C., and characterizes the Kurgans as semi-nomadic, living in seasonal settlements and transient villages as their economy is essentially pastoral, with subsidiary rudimentary agriculture, while their society is hierarchized and patriarchal. With the military superiority that the domestication of the horse and more advanced weaponry gives them, they easily subdue the peaceful, matrilineal/matrifocal, settled agriculturists of ancient Europe: the production of quality pottery with sophisticated symbolic decorations ceases, the cult of the mother-goddess is replaced by that of the warriorgod, chieftains and their retinue live in hill-forts, etc. This cultural transformation, which entails the Indo-Europeanization of Europe, implies, however, an intermediate "second homeland" for the Indo-

Linguistic paleontology

145

Europeans in the wake of phase II, when Central Europe becomes the staging area for further migrations after 3000 B. C. Linguistically, this secondary European homeland corresponds by and large for Marija Gimbutas (1986: 8 —9) to the area of the "Old European hydronymy" defined by Hans Krahe (1954: 48 — 71) and the territories of some trees whose Indo-European names Paul Friedrich (1970 a/b) has analyzed. As different groups detach themselves from this "European" core in the course of the Bronze Age (late Globular Amphora/early Corded Pottery extension, Tumulus expansion, Urnfield expansion, etc.), they constitute the various historical Indo-European dialects through hybridization with the substratum population. Although the views of Marija Gimbutas received wide recognition, they also met with strong opposition, namely from British archeologists (cf., e.g., Coles-Harding 1979: 6 - 8 ) , and in the early 1980s two Soviet scholars — Τ. V. Gamkrelidze and V. V. Ivanov (1985a/b/c) — proposed a radically different hypothesis: the homeland of the IndoEuropeans would be located on the northern periphery of Asia Minor south of the Transcaucasus as far as Upper Mesopotomia in the fifth — fourth millennia B. C., possibly connected with the Halaf culture (1985a: 31 —32). This implies that Anatolian and especially Armenian remained very close to the original homeland and puts the relations between Greece and Asia Minor in a new light. It also entails completely different routes of migration for the Indo-European languages of Europe, and the study of loanwords suggests a movement of the "Ancient European" dialects across Central Asia to the west, where they established an interim secondary "homeland" in the Black Sea and Trans-Volga areas, characterized linguistically by its uniform hydronymy (1985 b: 68 — 75). Here, the Soviet linguists join Marija Gimbutas, and even state that the culture reconstructed for Indo-European is compatible with the Pit —Grave or Kurgan in the third millennium B. C. in the territory from the northern Black Sea area to the Volga steppes and the Aral Sea. The challenging articles of Gamkrelidze and Ivanov, originally published in the Russian historical journal Vestnik drevnej istorii in 1980 — 1982, drew a rather strong reaction from the historian I. M. D'jakonov (1985), who questions the validity of an argument based on loanwords and the plausibility of locating the Indo-European homeland in an area where it does not appear as substratum or adstratum of the nonIndo-European languages (Hattic, Hurrian-Urartaean, North Semitic) spoken there in the third millennium B. C. D'jakonov's sharpest crit-

146

Edgar C. Polome

icism is, however, directed mainly against Gamkrelidze and Ivanov's assumptions concerning the routes of migration: as is obvious from his own views on the ethnogenesis of the Armenian people (1984), he cannot accept the hypothesis that they practically remained in the Proto-Indo-European original homeland. For D'jakonov (1985), this homeland should be located in the Balkan-Carpathian area (see Map 1), where domesticated animals and agriculture were introduced in the sixth/fifth millennia B.C. from Asia Minor {Qatal Hüyük), so that the local farmers and cattle-breeders could originally have come from the Near East. In a rejoinder, Gamkrelidze and Ivanov (1985 c) tried to meet D'jakonov's objections, but their rejection of the Balkan hypothesis rests on rather weak arguments, whereas this view deserves reconsideration in view of David W. Anthony's recent discussion of the Kurgan hypothesis (1986), in which ultimately he looks for the homeland of the Indo-Europeans in the Lvov-Kiev corridor that links Central Europe with the steppes across the Dnieper in the later part of the fourth millennium B. C., reaching conclusions close to those of V. G. Childe (1957), but basing them essentially on the use of the horse. The discussion of the views of Gamkrelidze — Ivanov in the early 1980s was, however, based on the limited information that they had provided in their papers; with the publication in Tbilisi a couple of years later of their magnum opus on Indo-European and the Indo-Europeans, the full extent of the materials used to build up their hypothesis has been made available. It consists of a comprehensive comparative grammar of Proto-Indo-European (428 pages), and a semasiologically organized lexicon of the protolanguage, with elaborate cultural commentary (400 pages), followed by a thorough discussion of the implications of borrowing and cultural contacts for the reconstruction of the dispersal of the Indo-Europeans (113 pages). The two-volume work also contains an extensive bibliography (142 pages) and indices of the terms analyzed in the various languages and the topics treated (196 pages). Not since the Neogrammarians and the cumulative work of Herman Hirt (1905-1907, 1921-1937, 1940) has such a compendium of IndoEuropean language and culture, incorporating both a comprehensive grammatical description and a concept-oriented presentation of the common lexical stock, been produced. Although many of its authors' views will undoubtedly be challenged (and, as a matter of fact, already have been; cf. Charachidze 1986, Polome 1987), it will be, for a long time, a basic work for the discussion of linguistic paleontology and

Linguistic paleontology

Map 1.

147

The Indo-European area in the fifth-fourth millennia B. C. (from D'jakonov 1985: 163) 1) Southern boundary of the deciduous forest and forest-steppe zone; 2) Southern boundary of the taiga zone and of the Finno-Ugric area; 3) Migration of the Hittito-Luwian dialects; 4) Migration of the Corded Ware people (Bandkeramiker) and related tribes, and opposite migration of European aborigines; 5) Assumed migration of the speakers of Tocharian dialects; 6) Spread of agriculture and cattle-breeding (by millennia B. C.); 7) Center of Common Indo-European area in the fifth millennium B. C. and its expansion.

148

Edgar C. Polome

the correlation of prehistory and archeology with linguistic data in regard to migration theory, as it makes many stimulating suggestions, for example, about the role of epidemics — the shift from burial to incineration being perhaps ascribable to the plague, which Gamkreldize and Ivanov also use as an argument for the intermediate stay of the ancestors of the "Ancient European" group of Indo-Europeans in the vicinity of the Aral Sea where they would have acquired their immunity against the bacterial disease. Recently, the whole problem of Indo-European origins and dispersal was taken up again by the British archeologist Colin Renfrew (1987), who strongly challenges the views of his predecessors. Rejecting the concept of "waves" of pastoralists invading Europe from the Pontic steppes, Renfrew asserts that Indo-Europeanization went hand in hand with the spread of farming and agriculture from the Near East, visualized as a steadily progressing territorial expansion through mostly peaceful migration generation after generation. As a better exploitation of the soil entails a substantial increase in population density, this "wave of advance model" implies the ultimate prevalence of the language of the farmers over that of the substratum population of huntergatherers. Referring to the results of modern anthropological investigation, Renfrew rightly stresses the fact that semi-nomadic pastoralism (which has been claimed to be the basis of the Indo-European economy) always requires the coexistence of agriculture, and he admits that, with the development of the technique of horse-riding and of a "chiefdom society", this may have created the pattern for domination by conquest. While Renfrew's thesis draws strength from the already previously indicated diffusion of agriculture from the Near East (compare Laviosa-Zambotti's 1955 map [Map 2] with Renfrew's [Map 3]) and the studies of Zohary — Hopf (1988) on the spread of domesticated plants in the Old World, the linguistic facts appear to be almost unreconcilable with his datings, for example, when he claims that Hittite is the direct continuation "in its own region of its very early Indo-European predecessor some five thousand years earlier", like its contemporary, Mycenean Greek, to the west (Renfrew 1987: 172). He does not account, indeed, for the fundamental differences at every grammatical level between Anatolian and Hellenic, on the one hand, and for the striking correspondences between the latter and IndoIranian, on the other, especially if these differences represent shared innovations, as many scholars now believe. Furthermore, he makes gratuitous assumptions to back up his extreme datations: since he

Linguistic

paleontology

149

Linguistic paleontology

151

needs "Indo-Europeans" as carriers of the farming civilization in the South Asian subcontinent as early as the sixth millennium B. C., he does not hesitate to claim that the people of the Indus valley culture in Mohenjo-Daro and Harappa were such Indo-Europeans! Just as unconvincing is his attribution of the glottogenesis of Celtic to a process of linguistic "crystallization", unifying (through the principle of "cumulative mutual Celticity") the three strands in the development of farming he assumes to constitute the fabric of the Celtic world — from Central Europe to northern Gaul (No. 3 on his map); from Italy to southern Gaul (No. 6 b), then bifurcating to North-Central Gaul, and hence to the British Isles (No. 9), and to Aquitaine (No. 8), the outrunner to Spain (Celtiberian) being left rather vague. Looking at the linguistic evidence one definitely feels more comfortable with the prudent statement of archeologists like Franz Fischer (1986) who agrees with linguists like Karl-Horst Schmidt in pointing out that a culturally clearly definable Celtic "ethnos" is not identifiable before the first half of the first millennium B. C., and Renfrew's statement that the process of "cumulative Celticity" started as early as 4000 B. C., after the Indo-Europeans had reached the later Celtic areas, remains, to say the least, very speculative. The projection of the Indo-European into a more distant past than is generally admitted compels Renfrew to associate the functional approach of Georges Dumezil to the interpretation of Indo-European religion with the later development of "heroic societies", in which sovereign power, closely associated with priesthood, rules over a turbulent military aristocracy and a rather indifferent mass of foodproducers and suppliers of other goods. Recent work on the development of society and state has certainly suggested that the strictly tripartite structure — priests; warriors; cattle-breeders/agriculturists — advocated by Dumezil since the late 1930s (cf. Littleton 1982), needs to be revised (cf. Polome 1982), but presumably in the direction of a closer analysis of the relations between the "technicians of the sacred" and the rest of the population. Even the "egalitarian peasants" of ancient Europe, described by Renfrew (1987: 253) as the first IndoEuropean speakers there, must have had among them some individual specialized in communication with the supernatural. Deities are needed to protect the crops and control the weather to ensure their steady growth, and these gods need to be placated if angered; someone has to be able to interpret their messages; etc. And, on an individual level, there must have been medicine-men able to heal with the help of

152

Edgar C. Polome

extraphysical powers! That all of this was organized into a system appears from the study of the Indo-European formulaic language (cf. Campanile 1977; Watkins 1982), and of Indo-European medicine (cf. Puhvel 1970), which was deeply grounded in their cosmogonic concepts (cf. Lincoln 1986). Therefore, Renfrew's rather negative assessment of comparative religion in correlation with linguistic paleontology is unjustified, and one had better refer to works like Jaan Puhvel's Comparative mythology (1987) for a fairer view of the subject. It is obvious that linguists and archeologists often deeply disagree when it comes to locating the homeland of the Indo-Europeans and dating the formation and breakup of the Indo-European speech community. The former try to reconstruct culture on the basis of definite sets of shared terms and to map linguistic contacts by means of isoglosses, with which they also attempt to establish their relative chronology, whereas the archeologists deal with some types of artifacts which can be dated and classified with a certain degree of precision, and correlated with specific cultures without defining them ethnically and/or linguistically. This explains why, using essentially the same materials, archeologists reach very divergent conclusions, as the recent critical survey of the approaches of mainly German scholars to the problem of Indo-European origins by Lothar Kilian (1983) may show. They expect linguists to offer them confirmation for their hypotheses, but often fail to take the limitations of linguistic reconstruction into consideration. Indeed, as Jürgen Untermann concludes, after a careful analysis of the terminology of the habitat: the only solid result of Indo-European comparative linguistics is that is implies a common origin for the different languages with mutual correspondences, which can best be represented as a shared protolanguage ... But linguistic comparison applied to prehistory is not an instrument to recognize historical reality or describe ethnic entities and their genesis (1985: 163).

As Jean Haudry (1981: 7) points out, all we can reconstruct are "probable models" — the way the Indo-Europeans pictured themselves through their vocabulary. This discussion of linguistic paleontology has focused on IndoEuropean, as the methodology of research has essentially been developed in this language family. Nonetheless, it is applicable to other families, and has been used in a similar way in that context, for example, for Bantu (cf. Polome 1977, 1980), a language family in which Renfrew (1987: 2 8 1 - 2 8 4 ) - wrongly (cf. Polome 1989) believes he can find a parallel to his views on Indo-European.

Linguistic paleontology

153

References Anthony, David W. 1986 "The 'Kurgan culture', Indo-European origins, and the domestication of the horse: A reconsideration", Current Anthropology 27: 291 — 313. Anttila, Raimo 1972 An introduction to historical and comparative linguistics. New York: Macmillan/London: Collier—Macmillan. Arntz, Helmut (ed.) 1936 Germanen und Indogermanen — Festschrift far Herman Hirt. 2 Vols. Heidelberg: Carl Winter. Bartoli, Matteo 1945 Saggi di linguistica spaziale. Turin: Rosenberg & Sellier. Bartoli, Matteo 1925 Introduzione alia Neolinguistica (Principi — Scopi — Metodi). Biblioteca dell' Archivum Romanicum, 2nd series, Vol. 12. Beeler, Madison S. 1961 "The interrelationship within Italic", in: Birnbaum — Puhvel (eds.), 51 — 58. Benveniste, Emile 1969 Le vocabulaire des institutions indo-europeennes: 1. economic, parente, societe; 2. pouvoir, droit, religion. Paris: Editions de Minuit. (English translation by Elizabeth Palmer under the title: Indo-European language and society (Miami Linguistic Series Vol. 12). Coral Gables, Florida: University of Miami Press, 1973). Bergsland, Knut —Hans Vogt 1962 "On the validity of glottochronology", Current Anthropology 3: 115 — 53. Bernhard, Wolfram — Anneliese Kandler-Pälsson (eds.) 1986 Ethnogese europäischer Völker. Stuttgart/New York: Gustav Fischer. Birnbaum, Henrik —Jaan Puhvel (eds.) 1961 Ancient Indo-European dialects. Proceedings of the Conference on IndoEuropean Linguistics held at the University of California, Los Angeles, 25 — 27 April 1960. Berkeley/Los Angeles: University of California Press. Bonfante, Giuliano 1931 I dialetti indoeuropei. Annali del R. Istituto Orientale di Näpoli IV) (Reprinted Brescia: Paideia, 1976). 1945 "On reconstruction and linguistic method", Word 1: 8 3 - 9 4 , 132-161. 1946 "Additional notes on reconstruction", Word 2: 155 — 156. 1947 "The Neolinguistic position: A reply to Hall's criticism of Neolinguistics", Language 23: 344-375. Bouquiaux, Luc (ed.) 1980 Lexpansion bantoue. Actes du Colloque International du CNRS, Viviers (France) 4 — 16 avril 1977. Paris: Societe d'Etudes Linguistiques et Anthropologiques de France. Buck, C. D. (ed.) 1949 Dictionary of selected synonyms in the principal Indo-European languages. Chicago: University of Chicago Press.

154

Edgar C. Polome

Buti, Giangabriella 1987 "The family and the tribe: Remarks on Indo-European social setting", in: Meid (ed.), 9 - 2 0 . Bynon, Theodora 1977 Historical linguistics (Cambridge Textbooks in Linguistics). Cambridge (U. K.): Cambridge University Press. Campanile, Enrico 1977 Ricerche di cultura poetica indoeuropea. Pisa: Giardini. Cardona. G. —H. Hoenigswald —A. Senn (eds.) 1970 Indo-European and Indo-Europeans: Papers presented at the Third IndoEuropean Conference at the University of Pennsylvania. Philadelphia: University of Pennsylvania Press. Charachidze, Georges 1986 Review of Indoevropejskij jazyk i indoevropejcy by Τ. V. Gamkrelidze and V. V. Ivanov (Tbilisi 1984). Bulletin de la Societe de Linguistique de Paris 81, 2: 9 7 - 1 1 2 . Childe, Vere Gordon 1957 The dawn of European civilization (6th, thoroughly revised edition of a work first published in 1925). London: Routledge & Kegan Paul. Coles, J. M. —A. F. Harding 1979 The Bronze Age in Europe: An introduction to the prehistory of Europe c. 2000- 700 B. C. New York: St. Martin's Press. Devoto, Giacomo 1931 Gli antichi Italici. Florence: Vallecchi (2nd edition 1951, 3rd edition 1969). 1940 Storia della lingua di Roma. Bologna (2nd edition 1944). (Updated German translation by Ilona Opelt: Geschichte der Sprache Roms. Heidelberg: Carl Winter, 1968). 1962 Origini indeuropee. ORIGINES (Studi e Materiali pubblicati a cura dell'Istituto Italiano di Preistoria e Protostoria) Florence: Sansoni. D'jakonov, Igor M. (also spelled: Diakonoft) 1984 The Pre-History of the Armenian people. Translated from the Russian by Lori Jennings. Delmar, Ν. Y.: Caravan Books. 1985 "On the original home of the speakers of Indo-European", Journal of Indo-European Studies 13, 1 - 2 : 9 2 - 1 7 4 . Dyen, Isidore 1962 a "The lexicostatistic classification of Malayopolynesian languages", Language 38: 38 — 46. 1962 b "The lexicostatistically determined relationship of a language group", International Journal of American Linguistics 28: 153 — 161. Entwistle, William James 1953 Aspects of language. London: Faber & Faber. Feist, Siegmund 1913 Kultur, Ausbreitung und Herkunft der Indogermanen. Berlin: Weidmann. Fischer, Franz 1986 "Die Ethnogenese der Kelten aus der Sicht der Vor- und Frühgeschichte", in: Bernhard —Kandier—Pälsson (eds.), 209 — 224. Friedrich, Paul 1970 a "Proto-Indo-European trees", in: Cardona —Hoenigswald —Senn (eds.), 11-34.

Linguistic paleontology 1970 b

155

Proto-Indo-European trees: The arboreal system of a prehistoric people. Chicago/London: University of Chicago Press. Gamkrelidze, Thomas V. — Vjaceslav V. Ivanov 1984 Indoevropejskij jazyk i indoevropejcy. Rekonstrukcija i istoriko-tipologiceskij analiz prajazyka i protokultury ("Indo-European and Indo-Europeans: A reconstruction and historical typological analysis of a protolanguage and a protoculture"). 2 vols., with a preface by Roman Jakobson. Tbilisi: Publishing House of the Tbilisi State University. 1985 a "The ancient Near East and the Indo-European question: Temporal and territorial characteristics of Proto-Indo-European based on linguistic and historico-cultural Data", Journal of Indo-European Studies 13,1 — 2: 3—48. 1985 b "The migrations of tribes speaking Indo-European dialects from their original homeland in the Near East to their historical habitations in Eurasia", Journal of Indo-European Studies 13, 1—2: 49 — 91. 1985 c "The problem of the original homeland of the speakers of Indo-European languages: In response to I. M. Diakonoffs article", Journal of IndoEuropean Studies 13, 1 - 2 : 175-184. Gimbutas, Marija 1970 "Proto-Indo-European culture: The Kurgan culture during the fifth, fourth, and third millennia B.C.", in: Cardona —Hoenigswald —Senn (eds.), 155-197. 1973 "Old Europe c. 7000-3500 B.C.: The earliest civilization before the infiltration of Indo-European peoples", Journal of Indo-European Studies 1, 1: 1 - 2 0 . 1977 "The first wave of Eurasian steppe pastoralists into Copper Age Europe", Journal of Indo-European Studies 5, 4: 277 — 338. 1980 "The Kurgan wave No. 2 (ca. 3400-3200 B.C.) into Europe and the following transformation of culture", Journal of Indo-European Studies 8, 3 - 1 : 273-317. 1985 "Primary and secondary homeland of the Indo-Europeans: Comments on the Gamkrelidze —Ivanov articles", Journal of Indo-European Studies 13, 1 - 2 : 185-202. 1986 "Remarks on the ethnogenesis of the Indo-Europeans in Europe", in: Bernhard — Kandier — Palsson (eds.), 5 — 20. Gudschinsky, Sarah C. 1956 "The ABC's of lexicostatistics (glottochronology)", Word 12: 175-210. Hall, Robert Α., Jr. 1946 "Bartoli's Neolinguistica", Language 22: 273-283. 1963 Idealism in Romance linguistics. Ithaca: Cornell University Press. Haudry, Jean 1981 Les lndo-Europeens. ("Que sais-je?" Vol. 1965.) Paris: Presses Universitäres de France. (2nd edition 1985). Hirt, Herman 1905 — 1907 Die Indogermanen. Ihre Verbreitung, ihre Urheimat und ihre Kultur. Straßburg: Karl Trübner. 1921 — 1937 Indogermanische Grammatik. 7 Vols. Heidelberg: Carl Winter. 1940 Indogermanica: Forschungen über Sprache und Geschichte Alteuropas. Edited posthumously by Helmut Arntz. Halle (Saale): Max Niemeyer.

156

Edgar C. Polome

Hock, Hans Henrich 1986 Principles of historical linguistics. Berlin/New York/Amsterdam: Mouton de Gruyter. Hoijer, Harry 1956 "Lexicostatistics: A critique", Language 32: 49—60. Hymes, Dell H. 1960 "Lexicostatistics so far", Current Anthropology I: 3—44. Hymes, Dell H. (ed.) 1964 Language in culture and society. New York/Evanston/London: Harper & Row. Koppers, Wilhelm (ed.) 1936 Die Indogermanen- und Germanenfrage (Wiener Beiträge zur Kulturgeschichte und Linguistik 4). Salzburg: Pustet. Krähe, Hans 1954 Sprache und Vorzeit. Europäische Vorgeschichte nach dem Zeugnis der Sprache. Heidelberg: Quelle & Meyer. Laviosa-Zambotti, Pia 1941 Le piü antiche civiltä nordiche e il problema degli Ugro—Finni e degli Indoeuropei. Milan. 1943 Le piü antiche culture agricole europee. Milan. 1950 "La successione delle gravitazioni indeuropee verso il Mediterraneo e la genesi della civiltä europea", Atti e Memorie dell'Accademia Toscana di Scienze e Lettere "La Golembaria" 16. Florence. Lees, Robert Β. 1953 "The basis of glottochronology", Language 29: 113-127. Lehmann, Winfred P. 1973 Historical linguistics: An introduction. 2nd edition. New York: Holt, Rinehart & Winston. Lehmann, Winfred P. (ed.) in press Proceedings of the IREX Conference on Comparative Linguistics and Typology at the Academy of Sciences of the U. S. S. R. in Leningrad, June 1988. Amsterdam: John Benjamins. Leveque, Pierre (ed.) 198-7 Peuples et Civilisations: 1. Les premieres civilisations des despotismes a la άίέ grecque. Paris: Presses Universitaires de France. Lincoln, Bruce 1986 Myth, cosmos, and society: Indo-European themes of creation and destruction. Cambridge, Mass./London: Harvard University Press. Littleton, C. Scott 1982 The new comparative mythology: An anthropological assessment of the theories of Georges Dumizil. 3rd edition. Berkeley/Los Angeles/London: University of California Press. Masica, Colin P. 1976 Defining a linguistic area: South Asia. Chicago: Chicago University Press. Meid, Wolfgang (ed.) 1987 Studien zum indogermanischen Wortschatz. Innsbruck: Institut für Sprachwissenschaft der Universität.

Linguistic paleontology

157

Meillet, Antoine 1908 Les dialectes indo-europeens (Collection linguistique, publiee par Societe de Linguistique de Paris Vol. 1). Paris: E. Champion. (Reprinted 1922 (with a new preface)). Palmer, Leonard R. 1954 The Latin language. London: Faber & Faber. Pisani, Vittore 1933 Studi sulla preistoria delle lingue indeuropee (R. Accademia Nazionale dei Lincei, Memorie, Ser. VI, Classe di scienze morali, Vol. IV, Fase. VI). Rome. 1940 Geolinguistica e indeuropeo (R. Accademia Nazionale dei Lincei, Memorie, Ser. VI, Classe di scienze morali, Vol. IX, Fase. II). Rome. Polome, Edgar C. 1964 "Considerations sur la valeur des donnees lexicostatistiques", in: Communications et Rapports du Premier Congres International de Dialectologie Genirale I, 29 — 36. Louvain: Centre International de Dialectologie. 1977 "Le vocabulaire bantou et ses implications culturelles", in: Paleontologia Linguistica. Atti del VI. Congresso Internazionale di Linguisti. Instituto Lombardo, Accademia di Scienze e Lettere e Sodalizio Glottologico Milanese, Istituto di Glottologia e Lingue Orientali, Universita di Milano, 181-201. Brescia: Paideia. 1980 "The reconstruction of Proto-Bantu culture from the lexicon", in: Bouquiaux (ed.), 779-791. 1982 "Indo-European culture, with special attention to religion", in: Polome (ed.), 156-172. 1983 "Bilingualism and language change as reflected by some of the oldest texts in Indo-European dialects", NOWELE 1: 9 - 3 0 . 1987 "Der indogermanische Wortschatz auf dem Gebiete der Religion", in: Meid (ed.), 201-217. 1988 "Linguistics and archeology: Differences in perspective in the study of prehistoric cultures", in: Lehmann (ed.). Polome, Edgar C. (ed.) 1982 The Indo-Europeans in the fourth and third millennia. Ann Arbor: Karoma. Porzig, Walter 1954 Die Gliederung des indogermanischen Sprachgebiets. Heidelberg: Carl Winter. Puhvel, Jaan 1970 "Mythological reflections of Indo-European medicine", in: Cardona — Hoenigswald — Senn (eds.), 369-382. 1987 Comparative mythology. Baltimore/London: Johns Hopkins University Press. Pulgram, Ernst 1958 The tongues of Italy: Prehistory and history. Cambridge, Mass.: Harvard University Press. Rea, John A. 1958 "Concerning the validity of lexicostatistics", International Journal of American Linguistics 24: 145 — 150.

158

Edgar C. Polome

Renfrew, Colin 1987 Archeology and language: The puzzle of Indo-European origins. London: Jonathan Cape. (Published in the U. S. in 1988 by Cambridge University Press, New York). Schaller, Helmut Wilhelm 1975 Die Balkansprachen: Eine Einführung in die Balkanphilologie. Heidelberg: Carl Winter. Scherer, Anton (ed.) 1968 Die Urheimat der Indogermanen (Wege der Forschung Bd. 166). Darmstadt: Wissenschaftliche Buchgesellschaft. Schräder, Otto 1917 — 1928 Reallexikon der indogermanischen Altertumskunde. Revised 2nd edition prepared by Alfons Nehring. (First published 1901). Berlin: W. de Gruyter. Sergent, Bernard 1987 "Les Indo-Europeens: genese et expansion d'une culture", in: Leveque (ed.), 4 7 1 - 6 0 4 . Shapiro, Michael C. — Harold F. Schiffman 1981 Language and society in South Asia. Delhi/Varanasi/Patna: Motilal Banarsidass. Sköld, Hannes 1931 Beiträge zur allgemeinen und vergleichenden Sprachwissenschaft. Vol. I. Lund: Gleerup. Solta, Georg Renatus 1980 Einführung in die Balkanlinguistik mit besonderer Berücksichtigung des Substrats und des Balkanlateinischen. Darmstadt: Wissenschaftliche Buchgesellschaft. Sommer, Ferdinand 1947 Hethiter and Hethitisch. Stuttgart: W. Kohlhammer. Swadesh, Morris 1951 "Diffusional cumulation and archaic residue as historical explanations", Southwestern Journal of Anthropology 7: 1—21. 1959 "Linguistics as an instrument of prehistory", Southwestern Journal of Anthropology 15: 2 0 - 3 5 . Tischler, Johann 1973 Glottochronologie und Lexikostatistik (Innsbrucker Beiträge zur Sprachwissenschaft Vol. 11). Innsbruck: Institut für Sprachwissenschaft der Universität. Untermann, Jürgen 1985 "Ursprache und historische Realität. Der Beitrag der Indogermanistik zu Fragen der Ethnogenese", in: Studien zur Ethnogenese. Abhandlungen der Rheinisch-Westfälischen Akademie der Wissenschaften Bd. 72: 133 — 164. Opladen: Westdeutscher Verlag. Vidossi, Giuseppe 1948 "Pro e contra le teorie di M. Bartoli", Annali della Scuola Normale Superiore di Pisa, Fase. I I I - I V : 204-219. Walde, Alois 1917 Über älteste sprachliche Beziehungen zwischen Kelten und Italikern. Innsbruck.

Linguistic paleontology

159

Watkins, Calvert 1982 "Aspects of Indo-European poetics", in: Polome (ed.), 104 — 120. Zimmer, Stefan 1987 "Indogermanische Sozialstruktur? Zu zwei Thesen Emile Benvenistes", in: Meid (ed.), 315-329. Zohary, Daniel —Maria Hopf 1988 Domestication of plants in the Old World: The origin and spread of plants in West Asia, Europe, and the Nile Valley. Oxford: Clarendon Press.

Linguistic geography and language change Raven I. McDavid,

Jr.

Dialect differences within speech communities and change with the passing of time were noticed long before linguistics became a science. But not until the later nineteenth century were the two considered together systematically. Perceptions of geographical variation, however fragmentary, came first: differences among neighboring villages are easy to notice, but the recognition of change requires historical records. Perhaps the best-known early story of the significance of an isogloss is that of the shibboleth (Judges 12.6), when Ephraimites were slaughtered at the fords of Jordan because they lacked the /§/ phoneme found in the speech of the Gileadites guarding the fords. By the fifth century B.C., regional varieties of Greek were recognized, and the dialects of Sparta and Thebes were exploited for comic effect — just as other literary artists, from Chaucer to Mark Twain, have exploited regional differences in the English-speaking world (Blair — McDavid 1982). By the end of the Renaissance it was realized that every living language had regional differences, though usually one variety (typically that of the capital) had more prestige than others. Dante's De vulgari eloquentia (1305) discussed fourteen regional varieties of Italian, still recognized by students of Romance dialects. Dante also recognized linguistic change, especially from Latin to Italian, as others had done before him. But the regularity of change was first stated specifically by Claudio Tolomei (1492—1555) in II Cesano (1935), as in the change of Latin initial clusters /pl-, kl-, fl-/ to Italian /pj-, kj-, fj-/, with exceptions explained by late borrowings from Latin and by dialectal evidence. Thus fiore 'flower' reflects regular change from Latin florem, but plebe 'people' is a direct borrowing

162

Raven I. McDavid,

Jr.

from Latin plebem, with pieve, which reflects the regular change, found as a dialectal term meaning 'a church for the common people'. During the seventeenth and eighteenth centuries, the study of language relationships became more systematic. Historical knowledge deepened as scholars became more attentive to past records. The Romantics sought in folk speech a more genuine usage than that of the literary languages. For various reasons, scholars sought not only to preserve past records but to describe unfamiliar languages and dialects of their own time; down through the nineteenth century, either traditional texts or translations of such familiar texts as the Lord's Prayer and the parable of the Prodigal Son were set down in all kinds of orthographies (Pop 1950). The announcement by Sir William Jones (1786) of the demonstrable kinship of Sanskrit, the classical tongues, and the languages of contemporary Europe was the catalyst which led to the postulation of the Indo-European family on the basis of systematic comparisons, but the postulation was derived from the work of many students with many interests. The scholars who made comparative Indo-European a subject for serious academic investigation were also seriously interested in folk speech and folk culture. Jacob Grimm, whose name is associated with one of the most important sets of systematic sound correspondences — the changes in consonant articulation that set off the Germanic group from other Indo-European languages 1 — is a striking example: however significant his contributions to comparative linguistics, he is most widely remembered as one of the brothers who collected German fairy tales. The techniques by which the Indo-European family was established were somewhat different from those developed later in linguistic geography. The purpose was different: to try to establish the ancestral form of the language from which the later tongues were derived. The materials were different: largely written records, since for some of the languages involved there was no evidence from living speakers. And, thanks to the accidents of cultural history, there was no consistent time-depth of the materials examined. Sanskrit, Greek and Latin provided copious evidence from several centuries before the Christian era; Germanic records dated from the fourth century A. D., with most of the evidence some centuries later; the earliest Slavic texts are still later. Nevertheless, one can say that the comparative Indo-Europeanists worked with linguistic geography on a grand scale, sorting out the interrelationships among languages spoken from the Ganges to Ice-

Linguistic geography

and language

change

163

land. Furthermore, as upon examination the relationships among IndoEuropean languages proved more complicated than originally suspected, the Indo-Europeanists were led into systematic linguistic geography as we now know it. For most of the nineteenth century, students of dialect were less concerned with systematic relationships. A great deal of evidence was gathered; and some, such as the work of A. J. Ellis in England (1889), sought to establish comparisons. But — and again particularly in England — much of the energy was devoted to the search for relic forms, odd words in odd places, such as surviving Celtic and Scandinavian loans. This emphasis persisted until the last quarter of the twentieth century. A century after linguistic geography had been established as a serious discipline on the Continent, and nearly half a century after it had been domiciled in North America, the tardily inaugurated Survey of English dialects (Orton —Dieth 1962 — 71) yielded a Word geography (Orton — Wright 1974) and a Linguistic atlas (Orton — Sanderson — Widdowson 1978) that failed to specify a single regional type of British English. Pending reexamination of the data from the Survey, the only comprehensive statement of the regional patterns of British speech is that of Ellis (1889), whose methodology antedates that of linguistic geography. The monumental English dialect dictionary (Wright 1898 — 1905) and English dialect grammar (Wright 1905) provide much information, but gathered by a variety of methods without serious editorial control. The reader must work out the regional patterning, form by form, and there is no cross-referencing for competing synonyms, for example, of names for the earthworm or the dragonfly. Moreover, the prefatory statement of Wright that he had recorded all the dialect words of English was not only naive but for half a century stood in the way of more systematic investigations. The change in the direction of dialect study was provoked by the success of the comparativists, especially the so-called "Neogrammarians" in Germany, with their mid-nineteenth-century emphasis on the Ausnahmslosigkeit (exceptionless nature) of sound change; typical was August Schleicher's Compendium der vergleichenden Grammatik (1861). The theory of Ausnahmslosigkeit of linguistic descent was challenged, and tested, by two areal approaches to language study: the wave theory of Johannes Schmidt (1872) and the investigation of living German dialects initiated by Georg Wenker (1876). Schmidt called attention to the fact that, within Indo-European, not only was the Germanic group set apart by the sound shift formulated by Grimm,

164

Raven I. McDavid,

Jr.

but there were other special features shared by a few of the language groups. The relationships indicated by one feature sometimes contradicted those indicated by others. For instance, one of the major divisions of Indo-European was that between centum and satem languages, based on the reflexes of the initial consonant of the postulated Indo-European word for "hundred", but other evidence suggested a close relationship between the Germanic centum group and the Slavic satem one. Various kinds of evidence led to the conclusion that the theory that each group of Indo-European languages had developed independently was far too simple. Perhaps groups sometimes shared with each other periods of common development, or, more likely, there was borrowing from one group to another through cultural contacts. The wave theory was a hypothesis; there was little documentation to support it. The work of Wenker2 and his successors3 was designed to test the hypotheses of the Neogrammarians against the massive evidence provided by a living language, in this instance the regional and local varieties of nineteenth-century German. Just as the first consonant shift had served to differentiate the German languages from the rest of the Indo-European family, so did a second and later High German consonant shift set off the High German dialects of southern Germany from other West Germanic languages as well as from Gothic and the Scandinavian languages. The question at issue was the consistency with which dialects had the sound changes identified with High German, the possibility that somewhere there was a boundary sharply setting off the varieties spoken in the North from those spoken in the South. Wenker devised a questionnaire of forty sentences of Standard German, which he sent to schoolmasters in German villages — first in the Rhineland, and ultimately to all the German-speaking areas of Europe — asking for conversion to the local dialects. Despite several shortcomings in the method, 4 the 44,251 questionnaires returned to the archives of the Sprachatlas at Marburg provide the finest-meshed network in any dialect survey and a basis for studies in greater depth and for follow-up investigations. The analysis of Wenker's evidence soon showed that there was no dramatic linguistic boundary setting off northern and southern Germany. In general, the isoglosses (i. e., the lines marking the limits of individual speech-forms) were fairly close together; but no two coincided exactly throughout their length. Furthermore, in the Rhineland there was a spreading out of these lines, in what has become known

Linguistic geography and language change

165

as the "Rhenish Fan" (Bloomfield 1933: 343-345). These divergent local boundaries, it turned out, reflected the political fragmentation of Germany from 1250 to 1800. The isoglosses follow the boundaries of the minuscule principalities liquidated when Napoleon formed the Confederation of the Rhine. The Rhineland thus appears as a "transition area". The Sprachatlas did not confirm the rigid interpretation of the regularity of sound change, but it did not repudiate it. Regularity of sound change is still the first assumption in historical linguistics, just as elementary physics sets up a basic formula for acceleration. When the evidence offers something not predicted by the basic formula, one seeks the cause — in physics the effect of friction, in historical linguistics the effects of cultural forces set in motion by the intricacies of human relationships. Where Wenker started from the Neogrammarian assumption of exceptionless sound change, Jules Gillieron, editor of the Atlas Unguis tique de la France, started from linguistic atomism, i.e., the belief that the history of each linguistic form is unique. His method, too, was different. Instead of using correspondence to gather data from every community within his area of investigation, he proposed intensive study of selected communities by a trained investigator, who used a longer questionnaire to interview in each community a selected local speaker, and who set down the responses on the spot, in fine-grained phonetics. The evidence is consonant with the social and cultural forces operating in France, just as Wenker's evidence illustrated the corresponding forces operating in Germany. In France, the first centralized state on the Continent, one finds as expected the powerful influence of the capital area around Paris. A good illustration is the development of words with Latin /ka-/: around the edges of the French-speaking area, one finds scattered "relic areas", where Latin /k-/ survives; around Paris is a compact but apparently expanding "focal area" where Parisian /s/, as in chanter, has been established; and inbetween are areas with jcj in the North and jtsj in the South, reflecting earlier changes. The jcj in Louisiana French, dating from the seventeenth century, surprises no one. In short, though Wenker and Gillieron started with different premises, their positions converged as they worked with their data. Dialectologists now start with the notion of regularity, and then seek explanations of the exceptions.

166

Raven I. McDavid,

Jr.

Linguistic geography in the contemporary sense is too recent in England for its findings to be brought to bear on problems of linguistic change. But historical linguistics in England has long been concerned with the implications of the dialect differences at earlier periods for an understanding of the way the standard language developed. There are copious Middle English records, from various regions, between the disappearance of standard West Saxon at the Conquest and the establishment of the Chancery standard of written English (Fisher 1977). West Saxon could not have been the ancestor of London Standard; records from the South and Southwest show divergent trends. On the other hand, the earliest London records lack a number of forms found in Standard English today. Regional texts show that some of the most significant grammatical features of this standard were imported into London from the North, for example, the pronouns they, their, them, and she; are as the plural present indicative of be; and the -s ending of the third singular present indicative. 5 Analysis of Orton's materials and the records from Guy Lowman's survey of southern England in 1937 —1938 will help in interpreting later linguistic developments. Since the primacy of London was established in the late Middle Ages, it has been the center from which innovations radiate (as in France they radiate from Paris), with the industrial Midlands often a secondary focus. Relics tend to survive in East Anglia and the south-western countries. The situation in North American English has no close parallel in Europe. In fact, the differences are so great that some critics, notably Pickford (1956), have questioned whether the methods of traditional linguistic geography can be adapted to the New World. Though ravaged by famine, pestilence, and war, the long-settled European populations suffered few major rearrangements before 1940. Englishspeaking North America was filled up in less than three centuries, simultaneously undergoing drastic social changes. Rarely could a New World community trace its roots to a single British source, and even the communities that could — like the Highland Scots settlement around Darien, Georgia — were soon contaminated with settlers of other origins. In addition to the general dialect mixture of Englishspeaking settlers, most communities had an early leavening of speakers of other languages. The early settlers represented only part of the British social spectrum — chiefly the lower middle class, the artisans, and the generally dissatisfied. Nobility, gentry, and peasantry knew their place and were content with it. Each new frontier, rural or urban,

Linguistic geography and language change

167

produced further mixture. The itchy American foot has taken a large number of each generation far from their birthplaces, and the opportunities of a fairly open society have brought to the White House the sons of sharecroppers, petty tradesmen, and impecunious ministers. Industrialization 6 has substituted the names of commercial products for those associated with home handicrafts; urbanization has juxtaposed speakers of diverse origins in situations where the rough edges of local pronunciations would be abraded; the commitment to general education, however haltingly practiced, has discouraged grammatical practices known or thought to be inelegant. Thus it is a commonplace that by European standards the United States has no regional dialects. 7 Recognition of the dynamics of American demography led to innovations in method: (1) sampling of urban speech as well as rural and (2) the investigation of common and cultivated speech as well as that of the folk. 8 Although this three-tiered sampling is sometimes misunderstood (as in Kurath's assumption [1939] that there are only three social classes in America), it provides a baseline for more intensive studies, particularly those that assay the changes taking place since the original surveys were conducted. Only one American study (Kurath — Lowman 1970) deals with British speech structurally, but it suggests future investigations, because the Orton materials have nearly five times as many records and almost four times as many items per record. 9 It shows that the coalescence of the ancestral forms of /i/ (see and sea), /e/ {hail and hale), and /o/ (ιroad and rode) is not only recent but geographically restricted. Looking at the American situation, one finds evidence of three of the phonological features often mentioned as setting off American or Canadian from British usage: the reestablishment and spread of postvocalic j-rj, the status of the "broad a" [a, a], and the loss of distinction between such pairs as cot and caught. Whatever the external forces for the rehabilitation of j-rj suggested by Labov (1966) for New York and Levine — Crocket (1977) for Hillsborough, North Carolina, one is on safer ground starting with the suggestion of R. McDavid (1948) for South Carolina: 10 the migration of j-rj speakers and their improving economic and educational opportunities. The 1941 Linguistic atlas records by Lowman show that j-rj was not unknown in New York, even at the highest social level. Hillsborough lies at the convergence of four dialect areas — Virginia Piedmont, Albemarle Sound-Neuse, Cape Fear-Peedee, and western North Carolina — and only in the first was j-rj lacking in the 1930s at all social levels. For the other

168

Raven I. McDavid,

Jr.

three areas, the loss of /-r/ at that time was a mark only of the most cultivated. For the "broad a" the Atlantic Seaboard provides two different status patterns. In rural New England and in the South Atlantic States, it was preserved sporadically as an older folk form in such words as hammer and Saturday. In eastern New England and occasionally elsewhere among the elite (eastern Virginia, New York City, Philadelphia), it appeared as a prestige form, borrowed from British English. The story of the cotjcaught homonymy is more complicated. The phenomenon is very widespread: north-eastern New England, including Boston (but not Providence: see Moulton 1968; R. McDavid 1981); the Pittsburgh area; most of Canada; and sporadic areas, not fully delineated, in the Upper Midwest, the Rockies, and the Pacific Coast. Preliminary analysis of the records from the North Central States suggests this same homonymy in areas along the Ohio and Mississippi between Cincinnati and St. Louis; records from the South Atlantic Coast and from South-western England suggest as the cause regional unrounding of the vowels in Britain and subsequent dialect mixture in North America. Historically, Mary, merry, and marry were distinct, with /e, ε, ae/ respectively. But by 1940 Mary and merry were homonyms in Pennsylvania, and in the Middle West all three are now usually pronounced alike. A detailed analysis of the records on both sides of the Atlantic is needed to work out the history of this coalescence. English has seen several changes in the distribution of vowels before post-vocalic /-r/ and its reflexes. Horse and hoarse are homonyms in British Received Pronunciation; Orton's evidence shows (R. McDavid 1968) that the homonym has spread since the Lowman survey of 1937 — 1938. Found in both Pennsylvania and the Hudson Valley, and characteristic of many of the more sophisticated informants interviewed in New England, this homonymy is spreading in America. Its chance of being generalized depends on the fate of another coalescence, of card and cord. In the Rockies, this later homonymy is popularly attributed to the Mormons, the shibboleth being the sentence "Were you born in a barnV It is also common in Texas, Oklahoma, and Western Louisiana; like the cot/caught homonymy, it occurs along the Ohio and Mississippi from Cincinnati to St. Louis; in the 1930s it was noted by Lowman for many Tidewater speakers in Virginia and Maryland. Investigators should be alert to the possibility that some idiolects, at least, may have only one back vowel before /-r/.

Linguistic geography

and language

change

169

This development would parallel what has happened in the front vowels. Most dialects contrast such pairs as air and ear. An older stage, represented by some southern American varieties,11 has a midfront vowel in ear, a lower one in air. In the majority of dialects, however, ear has moved to the high-front range, so that the Southerner's beer sounds like the Middle Westerner's bare. In some parts of the South Atlantic States, especially the South Carolina-Georgia Low Country, there is no contrast. 12 Some of Shakespeare's word play depends on this homonymy; the Orton materials should be examined for its distribution in England. Some New World innovations are apparent. In three regions — Canada, eastern Virginia, and the South Carolina-Georgia Low Country — /ai/ and /au/ have clearly defined positional variants, with a centralized beginning before voiceless consonants, an alternation not found in a single Orton record. Though possibly this alternation may some day be found somewhere in the British Isles, its development in three distinct New World regions is a noteworthy phenomenon. 13 Two further developments, one phonological and the other morphological, seem indigenous to the New World. For generations, observers of New York City speech have commented on the up-gliding diphthongal reflex of syllabic [a] — as [3ί, βϊ] in word, bird, heard and the like. Stigmatized by the schools and the butt of popular humor, this feature seems to have faded in the last generation; even less rare today is the homonymy of such pairs as coil and curl, characteristic of all Lowman's 1941 informants except the cultivated, and used by friends and relatives of mine of that level. This diphthong, including its extremely rounded variety, is so common in New Orleans, especially in the so-called "Irish Channel", that laymen sometimes attribute it to "Brooklyn influence". In less extreme forms it occurs in polite speech in the plantation South from South Carolina westwards. The physiological explanation is simple: abandoning lateral constriction while raising the tongue. But why should this operate in two parts of North America and nowhere in Britain? Could some kind of language contact be responsible? Equally indigenous to North America seems to be dove /dov/ as the preterite of dive. Though reported by Wright's English dialect dictionary, the only contextual example is from Longfellow's Hiawatha·, it was not recorded by Lowman, and not included in Orton's questionnaire. It is dominant in the American North, is well established in the Charleston area, has some lodgment on Chesapeake Bay, and seems

170

Raven / . McDavid,

Jr.

to be spreading among the younger and better educated (Atwood 1953: 9, Fig. 9). Even if some sporadic evidence should be found in some obscure British dialect, linguistic geography establishes dove as a cultivated American form. A closer look at the evidence, finally, may prevent snap judgments about the nature of linguistic changes. Callary (1975) and others suggest that in northern Illinois the raising of the vowel of bad and the fronting of the vowel of hot have spread from metropolitan Chicago. But there is ample evidence for this raising and this fronting in the rural speech of the Inland North — western New England and its derivatives — which was the primary source of northern Illinois speech (see Kurath-McDavid 1961: 1 0 3 - 1 0 4 , Maps 12, 13, 15). Even more amusing is the assertion of some educationists that working-class Chicagoans of Southern background pronounce rat and right alike. True, for many Southerners the vowel of right is monophthongal [a·], but the contrast with [ae·, ae] is signaled by tongue height instead of the fronting upglide. The confusion is not in Southern pronunciation but in the Midwestern perception, as is illustrated by the legend of the Southerner who, visiting a Michigan coed's apartment for a drink after the theater, asked for a piece of ice and was astounded by her enthusiastic amorous cooperation. Like other kinds of evidence, that from linguistic geography varies in quality. The designer of the questionnaire may not have included all the items one might like; the investigator may not have asked them in a way to elicit responses necessary for the most fruitful interpretation; the quality of phonetic transcription varies from field worker to field worker. Nevertheless, the records offer valuable insights to those interested in linguistic change, and should be consulted wherever they are available.

Notes 1. Members of the 1937 Linguistic Institute recall how dramatically Sapir followed his presentation of Grimm's statement with a demonstration that the Germanic consonant shift in Indo-European is paralleled by the Chipewyan consonant shift in Athapaskan. For Chipewyan spirants, as for proto-Germanic stops, the articulation of every phoneme is altered, but the structure of the ancestral consonant system is maintained (see Sapir 1931).

Linguistic geography and language change

171

2. Wenker's Sprachatlas des Deutschen Reiches exists in only two hand-prepared copies, in archives in Marburg and in Berlin. Wider distribution of the evidence was made possible by the re-edited version of Wrede et al. (1927 — 1956). 3. The investigation of German dialects continues today, though with new emphases. 4. The 44,251 completed questionnaires represent the work of about as many separate investigators, whose phonetic training and perceptions inevitably varied. It is not possible to identify informants (the completed questionnaire may sometimes represent a schoolmaster's synthesis — contrast Kurath et al. 1939, Chapter VI). The phonological thrust of the investigation was sometimes skewed by the discovery of lexical variants (e. g., Gaul and Ros alongside Pferd/Percthorse'), not fully accounted for by Mitzka's Wortatlas (1951 —1973). Nevertheless, the achievement is impressive. 5. These forms appear in the Northern Cursor Mundi (c. 1300) but not in the London dreams of Adam Davy (1307). It was the middle of the seventeenth century before they all became the norm in Standard English. The vector by which they were initially transmitted was probably the wool trade, in whose London guilds Northerners had attained important positions in fourteenth-century London. 6. The poverty of the unskilled masses has marched cheek to jowl with a shortage of the skilled, and consequent high wages. 7. Short-sighted Federal housing policies have resulted in the deterioration of the inner cities of many northern and western metropolitan areas into slums populated chiefly by blacks whose speech is basically southern American, little contaminated by contact with local white speech. Thus original regional differences have been transmuted into social differences, often more striking than regional differences among speakers of the same social class. 8. Cities were included in the network for Jaberg and Jud (1925 — 1940), often with more than one interview. But the informants, wherever found, represented local folk speech. 9. Even in phonology, there are often gaps in the coverage of Orton's questionnaire which make it difficult to develop usage profiles such as those Kurath developed for the American informants. For instance, there is inadequate representation of French origin, so that such common items as sauce and sausage are lacking from the inventory of the low back vowels. 10. This suggestion was confirmed by O'Cain (1972) for Charleston. 11. E. g., in the South Carolina Piedmont. As a speaker of such a variety, I experienced mild trauma in Michigan in 1937, in my first association with speakers who have the same structural differences as I do but have a contrast between high-front and mid-front where mine is between mid-front and low-front. Drinking beer can be misinterpreted. 12. O'Cain (1972) indicates that this homonymy is losing ground in Charleston, though persisting in smaller communities. 13. Chambers (1973) labels this phenomenon "Canadian rising". As pointed out, it is not strictly Canadian, though often so misidentified by those unfamiliar with the southern American parallels; nor is it "rising" but rather the preservation of an earlier and less open beginning, in certain positions, that is found in all positions in some dialects — an intermediate stage in the diphthongization of Middle English fij and /u/.

172

Raven I. McDavid, Jr.

References Atwood, E. Bagby 1953 A survey of verb forms in the eastern United States. Ann Arbor: University of Michigan Press. Blair, Walter—Raven I. McDavid, Jr. 1982 The mirth of a nation: America's great dialect humor. Minneapolis: University of Minnesota Press. Bloomfield, Leonard 1933 Language. New York: Henry Holt and Company. Callary, Robert 1975 "Phonological change and the development of an urban dialect in Illinois". Language in Society 4: 155 — 166. Chambers, J. ΚΙ 973 "Canadian rising", Canadian Journal of Linguistics 18: 113 — 135. Ellis, Alexander J. 1889 The existing phonology of English dialects. Vol. 5. On early English pronunciation. (Early English Text Society, Extra Series 56). London: Trübner. Ellis, Stanley (ed.) 1968 Studies in honor of Harold Orton (Leeds Studies in English 2). Leeds: School of English, University of Leeds. Fisher, John H. 1977 "Chancery and the emergence of standard written English in the fifteenth century", Speculum 52: 870-899. Gillieron, Jules — Edmond Edmont 1902 — 1910 Atlas linguistique de la France. 13 vols. Paris: Champion. Grimm, Jacob 1919 — 1937 Deutsche Grammatik. Göttingen: Dietrich'sche Verlagsbuchhandlung. Hall, Robert Α., Jr. 1936 "Linguistic theory in the Italian Renaissance", Language 12: 96 — 107. 1942 The Italian "questione della lingua": An interpretative essay. (University of North Carolina Studies in the Romance Language and Literature 4). Chapel Hill: University of North Carolina. Jaberg, Karl —Jakob Jud 1925 — 1940 Sprach- und Sachatlas Italiens und der Südschweiz. Zofingen: Ringier. Kurath, Hans, et al. 1977 Handbook of the linguistic geography of New England (2nd edition, revised by Audray R. Duckert). Published by arrangement with the American Council of Learned Societies, New York: AMS Press. (First published Providence: Brown University, 1939.) Kurath, Hans —Guy S. Lowman 1970 The dialect structure of Southern England: Phonological evidence (Publication of the American Dialect Society 54). University, ΑΙ.: University of Alabama Press. Kurath, Hans —Raven I. McDavid, Jr. 1961 The pronunciation of English in the Atlantic States. Ann Arbor: University of Michigan Press.

Linguistic geography and language change

173

Labov, William 1966 The social stratification of English in New York City. Washington: Center for Applied Linguistics. Levine, Lewis — Henry J. Crockett, Jr. 1967 "Speech variation in a Piedmont community: Postvocalic /-r/", in: Lieberson (ed.), 76 — 98. (Originally published in Sociological Inquiry 36, No. 2 [Spring 1966]; 204-226.) Lieberson, Stanley (ed.) 1967 Explorations in Sociolinguistics (International Journal of American Linguistics 33, No. 2/Part 2: Indiana University Research Center in Anthropology, Linguistics and Folklore, publication 44). McDavid, Raven I., Jr. 1948 "Postvocalic /-r/ in South Carolina: A social analysis", American Speech 23: 194-203. 1968 "Two studies of dialects of English", in: S. Ellis (ed.), 2 3 - 4 5 . 1981 "Low-back vowels in Providence: A note in structural dialectology", Journal of English Linguistics 15: 21 —29. Mitzka, Waither—Ludwig Erich Schmitt 1951 — 1973 Deutscher Wortatlas. 20 vols. Glessen: Schmitz. Moulton, William G. 1968 "Structural dialectology", Language 44: 451 —466. O'Cain, Raymond Κ. 1972 "A social dialect survey of Charleston, South Carolina", dissertation, University of Chicago. Orton, Harold —Eugen Dieth (eds.) 1962 — 1971 Survey of English dialects: Basic material. Introduction, 4 vols, (each in 3 parts). Leeds: E. J. Arnold. Orton, H a r o l d - N a t h a l i a Wright 1974 A word geography of England. London: Seminar Press. Orton, Harold — Stewart Sanderson—John Widdowson 1978 The linguistic atlas of England. London: Croom Helm. Pickford, Glenna R. 1956 "American linguistic geography: A sociological appraisal", Word 12: 211 — 233. Pop, Sever 1950 La dialectologie. Louvain: University of Louvain. Rice, Stuart A. (ed.) 1931 Methods in social science: A casebook. Chicago: University of Chicago Press. Sapir, Edward 1931 "The concept of phonetic law as tested in primitive languages by Leonard Bloomfield", in: Rice (ed.), 297-306. Schleicher, August 1861 Compendium der vergleichenden Grammatik der indogermanischen Sprachen. Weimar: H. Böhlau. Schmidt, Johannes 1872 Die Verwandschaftsverhältnisse der indogermanischen Sprachen. Weimar: H. Böhlau.

174

Raven I. McDavid, Jr.

Tolomei, Claudio 1535 II cesano. Vinegia: G. Giolito de Ferrari. Wrede, Ferdinand, et al. 1927 — 1956 Deutscher Sprachatlas. Marburg/Lahn: Elwert. Wright, Joseph 1898 — 1905 English dialect dictionary. 6 vols. London: Frowde. 1905 English dialect grammar. Oxford: Frowde.

Psycholinguistics: A research review Car lota S. Smith

The field of psycholinguistics deals with the psychological processes involved in language. Important subfields include language performance (encoding and decoding), the study of memory, speech perception and production, language development and acquisition. Other interdisciplinary fields such as neurolinguistics and reading will not be covered in this survey.

1. Language performance: words, sentences, texts Psycholinguists studying performance are interested in the processes and structures that underlie it — the steps from sound to meaning, and vice versa. In trying to get at these, researchers are forced to rely largely on indirect techniques. Typically a subject is given a task with a verbal stimulus and a direct behavioral response; the patterns of reaction or response times (RTs) are interpreted on the assumption that some time is taken up by linguistic processing. Differences in RTs lead to inferences about processing. Obviously experimental tasks and stimulae are themselves important variables in this kind of research. 1.1. Lexicon A number of questions in this field concern the lexicon, its structure and availability to the language user. Among the important questions are the following: How is the "mental lexicon" organized, and what is the internal structure of lexical entries? What are the processes invoked in finding a word in the lexicon? Is linguistic processing

176

Carlota S. Smith

relatively autonomous, with separate components for different types of processing? The mental lexicon appears to be organized according to the phonological form of words: Fay and Cutler (1977) cite patterns of spontaneous speech errors as evidence for this claim. Studies of rhyme detection suggest that both phonological and orthographic information is available in lexical representations (Donnenwerth-Nolan et al. 1981; Tanenhaus et al. 1980; Stanners — Forbach 1973). Morphemes rather than words are the units of the mental lexicon (Murrell — Morton 1974; Taft-Forster 1975; Stanners et al. 1979), although the role of derivational morphemes is unclear; apparently compound nouns have separate entries (Osgood — Hoosain 1974). Following the work of Putnam (1975) and Rosch (1973), it is hypothesized that lexical entries are organized in terms of prototypes rather than on a featural basis, as proposed by Collins and Quillian (1969). Word frequency affects the response time of subjects in several tasks. In lexical decision tasks, subjects are asked whether a visually presented string of letters is a word; RTs are faster to relatively frequent words (Forster —Bednall 1976; Taft 1979). Phoneme monitor tasks also show the effect of frequency. Here subjects listen to sentences, monitoring for a specified word-initial phoneme; it is assumed that latency for locating the phoneme increases with complexity of processing. Phoneme monitor RTs were faster to frequent than to relatively infrequent words (Foss 1969). In Rapid Serial Visual Presentation (RSVP) the words of a sentence are flashed very quickly on a screen and subjects are asked questions about them; frequency affects RTs in RSVP studies (Forster 1979). However, the work of Scarborough et al. (1977) suggests quite plausibly that recency and not frequency may be the factor that explains such results. There are several theories as to how lexical searches proceed. Morton (1970, 1979) proposes a word detector, or logogen model, in which a detector for a word becomes activated when it reaches threshold. A "spreading activation" model seems required to deal with the effects of context on lexical search (Collins — Loftus 1975); Becker and Killion (1977) suggest that the search is an active one. Evidence from different tasks shows that context can facilitate lexical search, which supports the active model (Meyer — Schvaneveldt 1976; Rumelhart 1977). Lexical decision RTs decrease when a target word is preceded by a related word (Loftus —Cole 1974; Meyer — Schvaneveldt 1971). Phoneme monitor studies with ambiguous words also indicate that context can

Psycholinguistics

177

affect search. When an ambiguous word preceded the target word RTs were relatively long, suggesting that all meanings of the word were accessed (Foss 1969); but in strongly biased contexts RTs suggest that only the relevant meaning of an ambiguous word is accessed (Swinney —Hakes 1976; for discussion of phoneme monitor work see Cutler—Norris 1979). Hoagabaum and Perfetti (1975) argue that the dominant meaning of an ambiguous word is ordered first in the mental lexicon, and therefore accessed first. Simpson (1981) shows that both context and dominance affect RTs in lexical decision tasks. However, Swinney (1979) and Tanenhaus et al. (1980) present evidence that all word meanings are accessed initially and that context affects the search process only at a second stage. The main current models for lexical searches are the following: (i) The logogen model, discussed above and updated in Jackson — Morton (1984). (ii) The search model, in which early stages of word recognition result in orthographic or phonological codes, which are compared to entries in the mental lexicon: the comparison is affected by frequency and by items that are semantically related to earlier words, cf. Forster—Davis (1984). (iii) The cohort model, developed by MarslenWilson and his colleagues, in which the candidates for recognition are the set of words sharing the properties of the signal, segment by segment (Marslen-Wilson 1984). (iv) The TRACE model uses a lowlevel trigger with quick access to information at various levels (McClelland — Elman 1986). For a summary of recent work in word recognition see Norris (1986). 1.2. Methodological questions It has been shown that experimental paradigms and different stimulus material produce different results. The moral may be that people respond differently to different situations; Foss and Blank (1980) argue, in this vein, that there are two modes of response to tasks such as those discussed here, namely prelexical and postlexical. They suggest that the mode used in a particular situation depends on contextual constraints, memory load, and clarity. Stress and other phonetic factors also affect responses (Cutler —Foss 1977; Cole et al. 1978). Some models of performance involve parallel processing with interactive components (Marslen-Wilson 1975); in others, processing takes place seriatim with autonomous components (Garrett 1975). Förster (1979) presents evidence for the autonomous model from RS VP tasks:

178

Carlota S. Smith

He did not find evidence for interaction of information, as the parallel processing model would predict. For example, the plausibility of a sentence did not appear to affect judgements involving syntax. Garrett (1981) argues that much evidence apparently favoring the interactive model can and should be accounted for within the more constrained autonomous model. An interactive model of language processing may be supported by some experiments with rhyme and category monitoring: subjects dealt with several types of information apparently without the discontinuity that a discrete, serial model would predict (Craik — Jacoby 1979; Lockhart et al. 1976). However, McClelland (1979) suggests a different interpretation: If the output of early stages of processing "cascades" to later stages before the early ones are completed, evidence for interaction of information does not necessarily indicate a common stage or process. Another technique for studying process involves shadowing: subjects repeat, or shadow, the words of a sentence immediately after hearing them. RT patterns vary depending on various factors; it is claimed that they show the effects of prior semantic and syntactic information and thus support an interactive model (Tyler — Marslen-Wilson 1977; Marslen-Wilson - Tyler 1980); see Norris (1982) for a competing interpretation. It has been demonstrated that prosodic structure affects comprehension. Cutler and Ladd (1983) survey recent models of prosody and its measurement. 1.3. Syntactic structure and processing The role of syntactic structure in language processing has been the focus of a good deal of study. Listeners are aware of the surface constituents of a sentence when there are no acoustic cues to constituent boundaries (Fodor et al. 1966, Fodor — Garrett 1967). It has been suggested that people do not standardly compute syntactic structure, relying instead on semantic processing strategies (Bever 1970; Rumelhart 1977); Holmes (1979) argues, however, that syntactic structure is always computed. See also the papers in Flores d'Arcais—Jarvella (1983). In the 1960s transformational generative grammar (TG) had an enormously stimulating effect on psycholinguistic work because it gave researchers specific linguistic structures with which to work (Miller 1962). Many experimental studies were done involving transformational rules and the structures postulated by TG, but by the late 1960s

Psycholinguistics

179

it was clear that a competence grammar such as that of Chomsky (1965) could not be related directly to psycholinguistic performance (Fodor — Garrett 1966; Watt 1970). Current theories in TG emphasize surface structure (rather than underlying structure and transformational derivations) and are explicitly designed to be relevant for language performance (Bresnan 1978). They are being implemented more or less directly in parsers and psycholinguistic models of performance (Marcus 1980; Kaplan 1973; Wanner - Maratsos 1978; F r a z i e r - R a y ner 1982). The notion of modularity plays an important role in theories of language processing; and rather generally in linguistics, psychology and other fields (Fodor 1983). A modular model has distinct subsystems that interact, rather than one large and intricate system. This allows for considerable flexibility. In studies of performance — language production and comprehension — the modular approach has led to a number of interesting hypotheses (Garfield 1987). According to the modularity thesis, the language processing system is made up of independent subsystems. In this framework the old question about whether or not language is simply an instance of general cognitive abilities is revived; for discussion see Tanenhaus et al. (1985), Cairns (1984), and Marslen-Wilson - Tyler (1980). In some models syntactic processing is independent. For instance, Frazier and Fodor (1978) argue that the initial step of parsing is exclusively syntactic. This means that parsing trees for sentences are built up without regard to semantic or pragmatic considerations, with parsing strategies such as High Attachment of constituents whenever possible. Frazier and Rayner (1982) present evidence favoring the hypothesis from ambiguous sentences, using recordings of eye-movement. There are competing accounts, however: Ford et al. (1982) argue that what look like structural preferences are really word based. Crain and Steedman (1985) and Tyler and Marslen-Wilson (1977) argue for an interaction hypothesis. Interaction hypotheses are not necessarily incompatible with modularity, but they provide the parser with access to semantic and pragmatic information from the earliest stages. Most studies deal with comprehension because the input can be controlled to at least some extent. However, Osgood and Bock (1977) present studies of production from a functionalist point of view, in which syntax is said to depend on combinations of cognitive and communicative factors. Bock (1982) suggests that there are two production modes, one primarily syntactic and the other primarily lexical.

180

Carlota S. Smith

Bever and Townsend (1979) take a functional approach in arguing that perceptual heuristics account for universal restrictions on subordinate clause word order. Grosjean et al. (1979) argue for a speech production mechanism that deals serially with units that correspond to phrase constituents. On the other hand, some evidence has been presented that the major unit for sentence planning corresponds to a deep or abstract level clause rather than surface clause (Ford — Holmes 1978). Speech errors are a source of evidence for production processes and constraints. Shattuck-Huffnagel (1979) argues for a single underlying production mechanism, with access to different levels of information, because the material erroneously inserted may share characteristics of different types with its replacement. Speech errors are also discussed by MacKay (1970); Fromkin (1973); Ellis (1979); Nooteboom and Cohen (1975); and Dell (1986). Levelt (1983) considers evidence from self-repairs as clues to speech production. 1.4. Text structure The structure of text and discourse is now an important topic of inquiry in psycholinguistics and related fields. The most interesting and fruitful approach has been that of the mental model. The model presents a structure which cannot be found "in" the discourse in any direct way, and which is intended to reflect the way the receiver understands the text. This is a considerable departure from rather traditional text-based approaches which looked for units of text structure that were larger than the sentence. Such models include the propositional content of a text as well as networks of coreference and other coherence relations in the text (Johnson-Laird 1983; Van Dijk — Kintsch 1983; Kamp 1981,1985). Discourse models must be cumulative to allow for changes, denials, etc., and although they may use sentence information they build up structures that are not sentential. Hobbs (1985) presents an artificial intelligence approach to discourse structure, especially cohesion. Anaphora is much studied in these and related frameworks (Hirst 1981; Webber 1981; Grosz-Sidner 1986; the topic is reviewed in Brown —Yule 1983). Anaphora is also studied experimentally (Garrod — Sanford 1977; Ehrlich — Rayner 1983). Important related fields of inquiry are inferencing and knowledge representation. Inference plays an important role in text understanding and the construction of models. Trabasso et al. (1984) deals with schemas and causal relations. H. Clark and his colleagues have dem-

Psycholinguistics

181

onstrated that inference takes place and that in some cases it takes time (Clark et al. 1983; Clark 1985). For a general discussion of recent approaches to text and discourse structure, see G. Brown —Yule (1983). To account for text understanding and the way people remember texts, the notion of schema has proved very useful. Researchers in psychology and in artificial intelligence have proposed that schemas, or macro-structures, organize the processing of texts, as well as the way knowledge is integrated and stored (Norman — Rumelhart 1975; Minsky 1975; Anderson 1976; Shank - Abelson 1977; Van D i j k Kintsch 1983). The idea is that people use high-level schemas for events and sequences that they encounter in the world; the schemas represent the skeletons of the situations. The schema or script of a visit to a restaurant is a famous example: knowing that the topic of a discourse involves a restaurant enables a person to understand references to waiters, menus, etc., with little preparation. It has been suggested that the same types of schemata appear in processing, memory, and other language activities. Research into people's use of schemas involves experiments with memory and retrieval, and computer simulation (Bower [ed.] 1975; Carpenter-Just 1975; B o b r o w - Collins 1975). Graesser et al. (1980) present evidence that people remember stories according to the predictions of a schema model. Spilich et al. (1979) show that previous knowledge affects the way people deal with new information. Vipone (1980) and Kieras (1982) investigate the power and limits of the schema approach, presenting evidence that it must be supplemented with less global notions. Most of this work deals with ex post facto rather than on-line processing.

2. The study of memory According to Nilsson (1979), current approaches to memory emphasize its continuity with other cognitive processes. This contrasts with the so-called cybernetic approach, in which memory is taken to involve the transfer of information from one compartment to another (Broadbent 1958). In the cognitive view there are interactions between memory and other processes: they are different aspects of the same system (Craik — Lockhart 1972; Craik — Tulving 1975). Models involving higher-level schemata are often invoked here: perception and memory, it is hypothesized, are organized by the same types of schemata (Ru-

182

Carlota S.

Smith

melhart 1977; Bobrow-Collins 1975; Lindsay - Norman 1977; Ε. E. Smith 1978). Memory structures are believed to consist of networks of related propositions, similar to those set up for story and knowledge organization. Nodes representing related knowledge are relatively close together in a network. There are some difficulties in applying this to memory experiments: it fails to account in a satisfactory way for certain kinds of relatedness (Rips et al. 1973; Holyoak —Glass 1975). The general view approach is close to procedural semantics, where memory is seen as active rather than passive (Miller —Johnson-Laird 1976). Two kinds of memory, the episodic and the semantic, were hypothesized by Tulving (1972). Episodic memory is essentially surface whereas semantic memory is featural. However, recent work does not support this clear dichotomy (Craik—Jakoby 1970; Ratcliff — McKoon 1981; Shoben et al. 1978). McCloskey and Santee (1981) found that subject-predicate relatedness affected RTs in a sentence verification task, suggesting that semantic information is featural quite generally; see also Glass —Holyoak (1975). Many studies of memory involve sentence verification; however, it is difficult to evaluate RT results based on this type of task. Forster and Olbrei (1973) argue that RTs show the time to evaluate and match, rather than comprehend or remember. Some other related studies involve the role of referential continuity in discourse (Garnham et al. 1982), and the effect of givennew information differences (Buschke — Schaier 1979). There is evidence for a two-stage process in the searching of semantic memory. At the first stage, the search asks whether there is any relevant information in memory; at the second the information, if any, is accessed. Answers to questions such as "Have you been to Paris?" are faster for distant, unfamiliar cities (Shanon 1979); in a lexical decision experiment of McCloskey and Bigler (1980), RTs were faster for rejective illegal strings ( g x g q ) than legal nonword strings ( g o v e ) . Other evidence for a two-stage process is given in Glucksberg — McCloskey 1981; Kolers-Palef 1976; Stanners - Forbach 1973; and Reder 1979. Memory for sentences can, apparently, take different forms. It has been shown that people tend to remember the meaning of a sentence rather than its form (Sachs 1967; Caplan 1972). Bransford and Franks (1971) and Bransford et al. (1972) found that people "recognized" sentences they had never seen before when the sentences were congruent semantically with what they had seen. Similarly, people compute and remember the presuppositions associated with a sentence and do

Psycholinguistics

183

not readily distinguish them in memory for the original sentence (Jarvella 1971; Clark - Haviland 1977; Clark - S h u n k 1980). But there is also evidence that people remember facts about surface structure: Baker (1978); Keenan et al. (1977); Bock (1977). The form in which a sentence is recalled can be affected by thematization and prompting (Perfetti — Goldman 1974; Anderson — Pickert 1978), and other contextual factors. People tend to answer questions in the same voice as the question (Wright 1972). Here as in other aspects of processing, apparent contradictions can be resolved with the recognition that people behave differently according to task and situation (see Glucksberg et al. 1973): The need for speed and/or accuracy can affect performance (Reed 1973); types of memory cues affect performance (Lockhart et al. 1976; Tulving 1974); and it is important to note that neither comprehension nor memory are all-or-nothing processes (Ratc l i f f - M c K o o n 1982).

3. Speech production and perception The essential question that has shaped research in this area concerns the relation between the speech system and the acoustical properties of speech. 3.1. Vocal tract hypothesis The question is a difficult one because no simple relation between acoustic signals, the organs of perception and production, and linguistic units has been discovered. In a series of influential papers, Fant has suggested that the organs of speech be regarded as physical systems which constrain the structure of the speech signal: the acoustic signal, then, is paramount (Fant 1960,1971). Speech articulation is considered as action toward the definite end of producing an acoustic pattern; in perception the speech organs extract information from the acoustic signal. This approach, sometimes called the vocal tract hypothesis, is contrasted by Studdert-Kennedy (1980) with an earlier view in which it was assumed that linguistic units are somehow represented in the accoustic signal and researchers attempted to explain the paradox of their absence by tailoring perceptual mechanisms for extracting them. The vocal tract hypothesis has all but superseded earlier theories that focused on motor commands as realizations of phonemic input

184

Carlota S. Smith

to the motor system (Liberman et al. 1967). The idea is that the phonemic input to the motor system consists of sequences of subphonemic commands, perhaps corresponding to distinctive features. Neutral mechanisms of production control are seen as a necessary mediation for perception. Studies modelling the vocal tract have shown that there are various ways the production system can realize a given phoneme, and that a single change in the articulatory program can have various acoustic effects (Lindblom 1962; Stevens — House 1972; MacNeilage — Ladefoged 1976). Plosives are studied in Summerfield — Haggard (1974) and Stevens — Klatt (1974); fricatives in Delattre (1968) and Massaro — Cohen (1966); affricates in Raphael — Dorman (1977). Statistical studies of covariation are summarized in Ladefoged — Harshman (1979). 3.2. Models of speech perception Models of speech perception have been developed consistent with the two hypotheses mentioned above. Some models are based on feature detector systems related to articulatory principles (Cooper 1975; Abbs —Sussman 1971). The notion of feature detectors for systems of this type is based on successful work in the field of vision, summarized in Lindsay —Norman (1977). Some difficulties with the motor hypothesis, and with the hypothesis that feature detectors underlie speech perception, are discussed in MacNeilage (1970). People perceive speech sounds as categorical rather than continuous: a segment is perceived as belonging to one category or another (Liberman 1970: Studdert-Kennedy 1976). Listeners tend to perceive distinctions that follow the phonemic categories of their language; but Pisoni et al. (1982) show that they can be trained to hear other distinctions as well. Some non-speech auditory stimuli are also perceived categorically (Cutting — Rosner 1974; Miller —Johnson-Laird 1976), as are some visual stimuli (Pastore et al. 1976). Speech signals are not all perceived categorically: Abramson (1979) presents evidence to this effect for the tones of Thai, for example. More generally, Fujisaki and Kawashima (n.d.) argue that the continuous and categorical modes of perception are not absolutely dichotomous in speech. See Repp (1983) for a summary of recent work on categorical perception. Much research focusing on the properties of acoustic signals has used artificial speech (Liberman et al. 1967; Fourcin 1979). Important

Psycholinguistics

185

cues to speech perception come from transitions; Shankweiler et al. (1977) found that vowels are identified more accurately in consonantal contexts than in steady state forms. The onset of voicing has been found to be a major cue for distinguishing types of plosives (Lisker — Abrahamson 1964, 1967). Recently attempts have been made to identify the invariant features of acoustic patterns that cue the perception of speech sounds (Stevens 1980); acoustic invariants can be identified when the articulatory organs are close to their target positions (Blumstein — Stevens 1979). Other sophisticated acoustical approaches to the invariance question include Searle et al. (1979) and Walley — Carrell (1983). The notion that speech sounds consist of distinctive feature bundles is due to Jakobson (1971); distinctive features are related to natural phonetic classes in Fant (1971). Stevens and Perkell (1977) claim that distinctive features are phonologic in nature and have quantal correlates in the domains of speech production, acoustics, and speech perception. Klatt (1979) considers the role of features in perception, and presents a model of lexical access. Prosodic information plays an important role in the perception of linguistic structure. Different degrees of duration can cue syntactic boundaries within a phrase (Cooper 1976; Lindblom 1978); Klatt (1976) summarizes factors affecting segmental duration. Klatt and Cooper (1975) argue that the durational structure of a sentence affects intelligibility and naturalness, especially in the areas of vowel duration, duration between stressed vowel onsets, and final phrase lengthening. Prosodic patterns can disambiguate sentences (Lehiste et al. 1975), but the relative importance of different aspects of prosodic patterning is poorly understood; timing is discussed in Huggins (1968). Work on the rhythmic structure of speech shows that it has a hierarchical rhythmic organization (Martin 1971, 1972). Speech perception is an active process in which people listen selectively and use their knowledge of language structure in constructing a linguistic representation. Listeners use contextual information so that they may hear material that does not actually occur. Warren and Obusek (1971) presented listeners with sentences from which a segment and its acoustic cues had been removed; the listeners claimed to have heard the missing segment. Foss and Blank (1980) show that people predict stress patterns on the basis of the early part of an utterance. In ongoing perception of connected speech, words may be accessed in the mind of the listener before phonemes in these words become

186

Carlota S. Smith

available as responses (Morton —Long 1976; Cole —Jakimik 1977). Fowler (1986) and Kelso et al. (1986) discuss speech processes in the light of event perception and action theory. See Jusczyk (1986) for a review of current speech perception research. Speech research is increasingly concerned with the interaction between different aspects of speech: acoustic, perceptual, physiological, neurological (Lindblom 1980).

4. Language development and language acquisition A fairly good picture of the course of first-language development has been established through naturalistic and maturational studies. This development is linked to cognitive development and the maturational stage of the child, and the relation between these factors has received a good deal of attention. Language acquisition, a closely related field, has been concerned with the mechanisms underlying the aquisition of language. 4.1.

Language development

4.1.1. Phonology Children arrive in the world equipped perceptually to deal with speech sounds: infants of one month are capable of categorical perception (Eimas et al. 1971; Eimas 1974). Work in this area, which involves special techniques, is summarized in Eimas — Tartter (1979). Studies of babbling from the earliest stages onwards both confirm and disconfirm the discontinuity hypothesis of Jakobson (1968). Patterns of babbling successively approximate the child's target language (Oller et al. 1976), contrary to the hypothesis, yet there is spectrograph^ evidence that children's production of sounds is quite different in speech and in babbling (Port — Preston 1972), which supports the claim of discontinuity. For a detailed review of studies of infant speech perception and auditory development, see Aslin et al. (1983) and Lindblom — Zetterstrom (1986). Language develops directly from early communication, rituals, and exchanges between young children and others (Bruner 1973/1974; Carter 1979). Work on the acquisition of pragmatics shows that there are distinct stages in this development (Bates 1976; Halliday 1975;

Psycholinguistics

187

Macnamara 1977). Piaget's work emphasizes that symbolic behavior and sensorimotor development precede language (Piaget 1926, 1972; see also Vygotsky 1962 for a somewhat different view of early language development). Cromer (1976) sums up the impressive evidence that cognitive development is a prerequisite for grammatical development. Studies of early communication emphasize the context and function of an utterance; a small number of communicative functions are expressed at the one-word stage (Greenfield — Smith 1976). The traditional view of early words as holophrases, which attributed sentential or propositional structure to the words, is no longer necessary with the tools of current speech-act theory (Dore 1975). Careful empirical studies of phonological development have shown that children begin to talk by acquiring a collection of words with idiosyncratic forms, rather than a rudimentary phonological system (Ferguson — Farwell 1975; Ingram 1977, 1979; see also Menyuk — Menn 1979). Only after they have learned a fairly large number of words do they begin to construct a phonological system (Kiparsky — Menn 1977). It has been argued that in syntax, too, children begin with individual cases of constructions and only later develop a rule system (Braine 1976). The development of phonology is reviewed and discussed in YeniKomshian et al. (1980); see also Schiefelbusch — Bricker (1981). Prosody and language development is discussed in a symposium in the Journal of Phonetics (Vol. 21 [1982]: 112-126). Furrow (1984) considers young children's use of prosody; Cutler and Swinney (1987) focus on developing prosodic systems and semantics for older children. 4.1.2. Early syntax Several longitudinal studies of the spontaneous speech of young children have produced some interesting generalizations about the early stages of child language; Ferguson and Slobin (1973) present several papers discussing this material. R. Brown (1973) shows that mean length of utterance (MLU) is a good indication of the linguistic development of children until their MLU is about 4.0; linguistic and chronological development do not correlate. Bringing together material from various studies, Brown found striking regularities in the way the children talked: the same types of notions and relations were expressed, and grammatical morphemes were acquired in the same order. The regularities are thought to reflect the dependence of early language

188

Carlota S. Smith

development on cognitive ability (Slobin 1973). The most fruitful analyses of the data of early word-combinations rely on rich, functional interpretation rather than syntactic analysis (Schlesinger 1974; Brown 1973). Children's comprehension ability may exceed production ability at the early stages, at least in part because of limitations on memory and faculties pertaining to the planning and production of utterances (Shipley et al. 1978; Olson 1973). There is some question as to whether abstract syntactic structure should be ascribed to children with an MLU of less than 2 (Bowerman 1973). One source of evidence for linguistic rules and structures is errors, or child-forms, that could only have been produced by the child's own grammars. Inflectional patterns such as Igoed thus provide invaluable insights. Acquisition of inflectional patterns is analyzed extensively in Brown (1973) and in some of the papers in Ferguson — Slobin (1973). There are three stages of learning: single instances, exceptionless (over-)generalizations, generalizations with exceptions. These three stages appear in syntax (cf. C. Chomsky 1969) and morphology (Bowerman 1974) as well. Klima and Bellugi (1967) use utterances that are ungrammatical for adults, such as When I can go? to construct a detailed account of the emergence of the auxiliary system of English. Their account suggests that children's emerging memory and computational abilities significantly affect language learning; see also Fay (1978). Errors in experimental studies using imitation provide clues as to how children structure linguistic input (Slobin —Welsh 1973; C. S. Smith 1970). 4.1.3. Later syntax Following the rudimentary stages, children appear to focus on surfacestructure configurations and grammatical-functional relations, as shown in studies of constructions such as the passive, relative clauses, and complement clauses (Sheldon 1974; deVilliers et al. 1979; Lederberg — Maratsos 1981; Hakuta 1981). At some stages in development children tend to rely on semantic strategies in comprehension; these strategies are to be distinguished from grammatical rules since they are essentially probabilistic. Extensive studies of conjunction show that patterns of development can be partially predicted from the structure of the language being learned (Lust 1977; Lust et al. 1980). More semantically oriented studies, including experimental studies of the development of anaphora, which focus on children's hypotheses as

Psycholinguistics

189

successive approximations to adult rules, are reported in Solan (1981); development of reference and deixis are studied in Maratsos (1976), Warden (1976), and Clark —Sengul (1978). There is a fairly extensive literature on the development of temporal reference (Cromer 1971; Antinucci-Miller 1976; Harner 1976, 1981; C. S. Smith 1980; Weist 1986); studies on complex adverbial sentences are summarized in Flores d'Arcais 1978. The development of word meanings proceeds with great speed: children learn approximately nine words a day until about the age of six (Carey 1978). Referential meanings develop gradually, some increasing in generality and others in specificity; over-generalization tends to be emphasized, probably because it is much easier to observe, but both occur (Anglin 1977; Ε. Clark 1973; Richards 1979). Referential words are usually learned at the functional level appropriate to the child's life: "dog" rather than "animal" or "dachshund", "flower" rather than "plant" or "petunia" (deVilliers-deVilliers 1978). Relational words have received a good deal of attention, partly because of their relation to cognitive development in the Piagetian framework (Donaldson-Wales 1970; Karmiloff-Smith 1977; Sinclair-de-Zwart 1967; Carey 1976). The development of word meaning is closely related to concept formation and other cognitive processes. (Gentner 1978; Levine — Carey 1982; G e l m a n - M a r k m a n 1986; Markman 1987). E.Clark (1981) discusses children's lexical and syntactic innovations. After the age of five, children are still learning. They do not yet control certain specialized syntactic rules (C. Chomsky 1969); according to Karmiloff-Smith (1979), they focus at this stage on complex categories in which linguistic and other factors interact. Metalinguistic awareness develops at this stage (Sinclair-de-Zwart 1971; Beilin 1975; Hakes 1980). Cromer (1976) argues that in spite of their considerable abilities, children over five are at an intermediate stage of language development. He notes that their linguistic abilities are not yet stable: complex tasks and distracting factors can affect their performance. A good deal of variation in children's abilities can be explained in terms of cognitive and computational load; only when the linguistic system is thoroughly mastered and integrated with others is linguistic performance consistent. The development of discourse competence has been the focus of recent research, as in Foster's 1982 study of topic development. Hickman (1982) looks at the way referents are introduced and maintained

190

Carlota S. Smith

in discourse; Stoel-Gammon and Hedberg (1984) and Karmiloff-Smith (1985) study the development of cohesion in children's narratives; and Preece (1987) shows that several narrative forms are produced conversationally by young children. Experimental studies using artificial language have been quite suggestive in showing that certain properties seem to be essential to the learning of an artificial language. By hypothesis, these properties are essential to natural languages as well. Moeser and Bregman (1982) show that semantic correlates to syntax facilitate learning. Morgan and Newport (1981) show that constituent groupings with semantic significance strongly facilitate learning. 4.2. Language acquisition Language-acquisition theory deals with questions of how humans acquire natural language. Its domain includes the mechanisms underlying first language acquisition, and the nature of language structure. Noam Chomsky has argued forcefully that humans acquire natural language because of their innate linguistic endowment (1965, 1975). Language learning cannot be explained as learning in the general sense, he claims, making the argument of "poverty of the stimulus". The linguistic input to young children underdetermines the grammars that they are learning. Rather, humans have a specific language faculty with specific content, and grammar is a matter of growth and triggering rather than of learning in the traditional sense. Evidence for this view of "grammar growth" is the rapidity of first-language acquisition, the relative absence of errors, and the fact that children are not instructed in their native language. Furthermore, and very tellingly, children acquire constraints for which they receive no direct evidence: this argues strongly for an innate language faculty. Detailed accounts of language learning from this strong nativist position are developed in Felix (1984) and Pinker (1984). Others take the position that general cognitive abilities are used in the construction of grammars by children. Some psychologists emphasize the importance of functional considerations and the social grounding of language development (Bates —MacWhinney 1982). Others argue that the essentials, such as grammatical categories, can be constructed using distributional and other problem-solving techniques, without invoking a special linguistic faculty (Moulton —Robinson 1981; Maratsos 1982; Bohannon 1984). Slobin (1973, 1985) emphasizes

Psycholinguistics

191

the role of strategies of learning that are keyed both to language and to what children know about the world. Bohannon and WarrenLeubecker (1985) present an account of the main approaches to these questions; see Wanner—Gleitman (1982) and Harris — Davies (1987) for discussions more sympathetic to the strong nativist position. Studies of the linguistic input to children are of central importance. Strong nativists argue that the input cannot explain the acquisition of language, whereas those with a more empiricist approach claim that it can. Research has established that speech directed to children tends to be short, well-formed, and distinctly pronounced, rather than illformed, mumbled, and complicated. Snow and Ferguson (1977) present a collection of research papers documenting this conclusion; but see also Pye (1986). Structural analysis of speech directed to children does not support claims that input can account for language acquisition (Newport et al. 1977; Gleitman et al. 1984). On the other hand, Furrow et al. (1979) argue that children's speech significantly mirrors the structural patterns of adult input. An important point made by learnability theorists involves the evidence that children receive about the language they are learning. In many accounts of learning, evidence about errors is a crucial factor, yet children do not receive direct negative evidence. Parents rarely correct grammatical errors directly (Brown — Hanlon 1970), and recent work in this area suggests strongly that indirect negative evidence is not available to children (Grimshaw 1986). The approach to learnability of Wexler and his colleagues has been influential (J. Wexler — Culicover 1980; K. Wexler 1982; see also White 1982). The topic of generalization and productivity is central to both language acquisition and linguistics. There is evidence that children change their grammars (Bowerman 1982), sometimes reorganizing and sometimes retreating from generalizations (deVilliers 1984; Randall 1984). The acquisition of dative structures is an interesting area for study because it involves limited generalization (White 1982; Pinker — Frost 1984); lexical productivity that is not entirely unlimited is discussed in C. S. Smith (1981). Linguists explicate the notion of a specific linguistic faculty through Universal Grammar, which represents the human linguistic endowment. Universal Grammar has a modular organization and has a rich structure. In this approach syntax is taken to be an autonomous subsystem, syntactic categories to be relatively abstract, and innate constraints on rules are expected. (Solan 1981; Jacubowicz 1984; C r a i n - N a k a y a m a 1987).

192

Carlota S. Smith

The various structures found in different languages are allowed for by a strong notion of parametrization. Among the topics most fruitfully studied are pronominal and anaphoric systems (Mazuka et al. 1986; K. Wexler — Manzini 1987; a collection of papers on this topic is presented in Lust [ed.] 1986), coordination (Lust —Chien 1984), and the pro-drop phenomenon (Hyams 1987, with particularly interesting comments from Lebeaux 1987). Cross-linguistic work is of great importance since the field attempts to arrive at generalizations about language. Some recent studies of languages other than English are MacWhinney 1975 (Hungarian), Berman 1981 (Hebrew), Snow et al. 1980 (Dutch), Johnston - Slobin 1979 (Italian, Serbo-Croatian, Turkish), Tanouye 1979 (Japanese), and Erbaugh 1978 (Chinese). Sketches of the acquisition of English, German, Hebrew, Japanese, Kaluli, Polish, French, Samoan, Turkish, and American Sign Language are presented in Slobin (1985). Note I would like to thank D o n Foss of the University of Texas for allowing me access to his recent discussion of psycholinguistics.

List of

abbreviations

CogPsych JChL JExP JExP: G JEXP: HLM JExp: HPP JPhon JPsyR JSPHRes JVLVB L&S Ling I Μ &C JAS PRCLD Ρ & Ρ Psycholog R PsychonSci QJExP

Cognitive Psychology Journal of Child Language Journal of Experimental Psychology — General — Human learning and memory — Human learning and perception Journal of Phonetics Journal of Psycholinguistic Research Journal of Speech and Hearing Research Journal of Verbal learning and Verbal Behavior Language and Speech Linguistic Inquiry Memory & Cognition Journal of the Acoustical Society of America Papers and Reports on Child Language Development Perception and Psychophysics Psychological Review Psychonomic Science Quarterly Journal of Experimental

Psychology

Psycholinguistics

193

References Abbs, James H. — Harvey M. Sussman 1971 "Neurophysiological feature detectors and speech perception: A discussion of theoretical implications", JSHR 14: 2 3 - 3 6 . Abrahamson, Adele 1977 Child language: An interdisciplinary guide to theory and research. Baltimore: University Park Press. Abramson, Arthur 1979 "Non-categorical perception of tone categories in Thai", in: Lindblom — Ohman (eds.), 127-134. Abramson, Arthur—Leigh Lisker 1970 "Discriminability along the voicing continuum: Cross-language tests", Sixth International Congress of Phonetic Sciences Prague: 595 — 673. Prague: Academia. Aksu, Ayhan 1978 "The acquisition of causal connectives in Turkish", PRCLD 15: 129 — 139. Alba, Joseph W. — Susan G. Alexander—Lynn Hasher —Karen Caniglia 1981 "The role of context in the encoding of information", JExP: HLM 7: 283-292. Anderson, John R. 1976 Language, memory and thought. Hillsdale, N. J.: Erlbaum. Anderson, Richard C. —Roberta D. Pickert 1978 "Recall of previously unrecallable information following a shift in perspective", JVLVB 17: 1 - 1 2 . Anglin, Jeremy 1977 Word, object, and conceptual development. New York: Norton. Antinucci, Francesco —R. Miller 1976 "How children talk about what happened", JChL 3: 167-170. Aslin, Richard N. —David B. Pisoni — Peter W. Jusczyk 1983 "Auditory development and speech perception in infancy", in: Μ. M. H a i t h - J . J. Campos (eds.), 573-689. Baker, Carl L. — John McCarthy (eds.) 1981 The logical problem of language acquisition. Cambridge, Mass.: MIT Press. Baker, Linda 1978 "Processing temporal relations in simple stories: Effect of input sequence", JVLVB 17: 559-572. Bates, Elizabeth 1976 Language and context: The acquisition of pragmatics. New York: Academic Press. Bates, Elizabeth — Brian MacWhinney 1982 "Functionalist approaches to grammar", in: Wanner—Gleitman (eds.), 173-218. Bates, Elizabeth —M. Masling — Walter Kintsch 1978 "Recognition memory for aspects of dialogue", JExP: HLM 4: 187-197.

194

Carlota S. Smith

Becker, Curtis Α. —Τ. H. Killion 1977 "Interaction of visual and cognitive effects in word recognition", JExP.HPP 7: 389-401. Beilin, Harry 1975 Studies in the cognitive basis of language development. New York: Academic Press. Berman, Ruth 1981 "Regularity vs. anomaly: The acquisition of Hebrew inflectional morphology", JChL 8: 265-282. Berwick, Robert C. — Amy Weinberg 1983 "The role of grammars in models of language use", Cognition 13: 1 —61. Bever, Thomas 1970 "The cognitive basis for linguistic structures", in: Hayes (ed.), 279 — 362. Bever, Thomas — David J. Townsend 1979 "Perceptual mechanisms and formal properties of main and subordinate clauses", in: Cooper—Walker (eds.), 159 — 226. Bloom, Lois (ed.) 1978 Readings in language development. New York: Riley. Blumstein, Sheila — Kenneth Stevens 1979 "Acoustic invariance in speech production", JAS 66: 1001 — 1017. Bobrow, David G. —Alan Collins (eds.) 1975 Representation and understanding: Studies in cognitive science. New York: Academic Press. Bock, J. Kathryn 1982 "Toward a cognitive psychology of syntax: information processing contributions to sentence formulation", Psycholog R 89: 1—47. Bohannon, John N. 1984 "Reciprocity in children's language environment", paper presented at the 9th Annual Conference on Language Development, Boston University, Massachusetts. Bohannon, John N. — Amye Warren-Leubecker 1985 "Theoretical approaches to language acquisition", in: J. B. Gleason (ed.), 173-217. Bouma, Herman —Don G. Bouwhis (eds.) 1984 Control of language processes. Vol. 10 of Attention and performance. Hillsdale, N.J.: Erlbaum. Bower, Gordon H. 1976 "Experiments on story understanding and recall", QJExP 28: 522 — 534. Bower, Gordon H. (ed.) 1975 The psychology of learning and motivation. Vol. IX. New York: Academic Press. Bowerman, Melissa 1973 "Structural relations in children's language: Syntactic or semantic?" in: Moore (ed.), 197-213. 1974 "Learning the structure of causative verbs: A study in the relation between cognitive, syntactic and semantic development", PRCLD 18: 142 — 178. 1982 "Reorganizational processes in lexical and syntactic development", in: Wanner-Gleitman (eds.), 319-345.

Psycholinguistics

195

Braine, Martin D. S. 1976 Children's first word combinations (Monograph of the Society of Research in Child Development 41). Chicago: Society of Research in Child Development. Bransford, John D.—James R. Barclay—Jeffrey J. Franks 1972 "Sentence memory: A constructive versus interpretive approach", CogPsych 3: 192-209. Bransford, John D.—Jeffrey J. Franks 1971 "The abstraction of linguistic ideas", CogPsych 2: 331—350. Bresnan, Joan 1978 "A realistic transformational grammar", in: Halle — Bresnan — Miller. Bresnan, Joan (ed.) 1982 The mental representation of grammatical relations. Cambridge, Mass.: MIT Press. Broadbent, Donald Eric 1958 Perception and communication. London: Pergamon. Brown, Gillian — George Yule 1983 Discourse analysis. Cambridge, England: Cambridge University Press. Brown, John (ed.) 1976 Recall and recognition. London: Wiley. Brown, Roger 1973 A first language. Cambridge, Mass.: Harvard University Press. Brown, Roger — Camille Hanlon 1970 "Derivational complexity and the order of acquisition in child speech", in: Hayes (ed.), 11—53. Bruner, Jerome 1973/1974 "From communication to language: A psychological perspective", Cognition 3: 225-287. 1975 "The ontogenesis of speech acts", JChL 2: 1 — 19. Buschke, Herman —Aron H. Schaier 1979 "Memory units, ideas, and propositions in semantic remembering", JVLVB 18: 549-563. Butterworth, Brian (ed.) 1981 Language production. Vol. I. New York: Academic Press. Cairns, Helen S. 1984 "Current issues in research in language comprehension", in: Naremore (ed.). Campbell, Robin N. —Philipp Τ. Smith (eds.) 1976 Recent advances in the psychology of language: Language development and mother-child interaction. New York: Plenum Press. Caplan, David 1972 "Clause boundaries and recognition latencies for words in sentences", P & P 12: 7 3 - 7 6 . Carey, Susan 1976 "Less may never mean more", in: Campbell — Smith (eds.), 109 — 132. 1978 "The child as word learner", in: H a l l e - B r e s n a n - M i l l e r (eds.), 2 6 4 - 2 9 3 .

196

Carlota S. Smith

Carlson, Roger D. —Björn Granstrom — Dennis H. Klatt 1979 "Some notes on the perception of temporal patterns in speech", Proceedings of 9th International Congress of Phonetic Sciences. Copenhagen: Institute of Phonetics, University of Copenhagen, 260 — 267. Carpenter, Patricia A. 1977 "Integrative processes in comprehension", in: LaBerge — Samuels (eds.), 217-242. Carpenter, Patricia Α. —Marcel A. Just 1975 "Sentence processing: A psycholinguistic processing model of verification", Psycholog R 82: 4 5 - 7 3 . Carter, Anne Lindsay 1979 "Pre-speech meaning relations: An outline of sensorimotor development", in: Fletcher—Garman (eds.), 71—92. Carterette, Edward C. — Morton P. Friedman (eds.) 1976 Handbook of perception. Vol. 7. New York: Academic Press. Chang, Frederik R. 1980 "Active memory processes in visual sentence comprehension: Clause effects and pronominal reference", M&C 8: 58 — 64. Chomsky, Carol 1969 The acquisition of syntax in children from 5 to 10. Cambridge, Mass.: MIT Press. Chomsky, Noam 1965 Aspects of the theory of syntax. Cambridge, Mass.: MIT Press. 1975 Reflections on language. New York: Pantheon Press. Clark, Eve 1973 "What's in a word? On the child's acquisition of semantics in his first language", in: Moore (ed.), 197 — 213. 1981 "Lexical innovations: How children learn to create new words", in: Deutsch (ed.), 299-328. Clark, E v e - C . J. Sengul 1978 "Strategies in the acquisition of deixis", JChL 5: 557-575. Clark, Herbert H. 1985 "Language use and language users", in: Lindzey—Aronson (eds.), 179-232. Clark, Herbert H. —Susan S. Haviland 1977 "Sentence comprehension and the given-new contract", in: Freedle (ed.), 1-40. Clark, Herbert H . - R o b e r t Schreuder-S. Buttrick 1983 "Common ground and the understanding of demonstrative reference", JVLVB V 22: 245-258. Clark, Herbert H . - D . H. Shunk 1980 "Polite responses to polite requests", Cognition 8: 111 — 143. Cohen, Antonie — Sibout G. Nooteboom (eds.) 1975 Structure and process in speech perception. Berlin: Springer-Verlag. Cole, Ronald A. — Jola Jakimik 1977 "Understanding speech: How words are heard", in: Underwood (ed.), 67-116.

Psycholinguistics

197

Cole, Ronald A. — Jola Jakimik — William E. Cooper 1978 "Perceptibility of phonetic features in fluent speech", JAS 64: 44 — 66. Collins, Anne M. —Elizabeth F. Loftus 1975 "A spreading-activation theory of semantic processing", Psychology 82: 402-428. Collins, A n n e M . — M. R. Quillian 1969 "Retrieval time for semantic memory", JVLVB 14: 2 1 5 - 2 3 9 . Cooper, William E. 1975 "Selective adaptation to speech", in: Restle et al. (eds.), 23 — 54. 1976 "The syntactic control of timing in speech production: A study of complement clauses", Journal of Phonetics 4, No. 2: 151 — 171. Cooper, William E. —Jeanne M. Paccia—Stephen E. Lapointe 1978 "Hierarchical coding in speech timing", CogPsych 10, No. 2: 154—177. Cooper, William E. — Edward C. T.Walker (eds.) 1979 Sentence processing: Psycholinguistic studies presented to Merrill Garrett. Hillsdale, N. J.: Erlbaum. Craik, Fergus I. M. — Larry L. Jacoby 1979 "Elaboration and distinctiveness in episodic memory", in: Nilsson (ed.), 145-166. Craik, Fergus I. M. —R. S. Lockhart 1972 "Levels of processing: A framework for memory research", JVLVB 11: 671-684. Craik, Fergus I. M. — Endel Tulving 1975 "Depth of processing and the retention of words in episodic memory", JEx Ρ: G 104: 2 6 8 - 2 9 4 . Crain, Stephen — Makoto Nakayama 1987 "Structure dependence in grammar formation", Language 63, No. 3: 522 — 543. Crain, Stephen —Mark Steedman 1985 "On not being led up the garden path: The use of context by the psychological syntax processor", in: Dowty — Karttunen — Zwicky (eds.), 320-358. Cromer, Richard 1971 "The development of the ability to decentre in time", British Journal of Psychology 62: 353 - 365. 1976 "Developmental strategies for learning language", in: Hamilton — Vernon (eds.), 305-358. Cutler, Anne 1976 "Beyond parsing and lexical lookup: An enriched description of auditory sentence comprehension", in: Wales —Walker (eds.), 133 — 149. Cutler, Anne — David Fay 1982 "One mental lexicon, phonologically arranged", Ling 113: 107 — 112. Cutler, Anne —Donald Foss 1977 "On the role of sentence stress in sentence parsing", Language & Speech 20: 1 - 1 0 . Cutler, Anne —D. Ladd 1983 Prosody: models and measurements. Berlin/New York: Springer-Verlag.

198

Carlota S. Smith

Cutler, Anne —Dennis Norris 1979 "Monitoring sentence comprehension", in: Cooper — Walker (eds.), 113-134. Cutler, Anne —David Swinney 1987 "Prosody and the development of comprehension", JChL 14: 145 — 167. Cutting, James E. — Barbara Rosner 1974 "Categories and boundaries in speech and music", P & P 16: 564 — 570. Darwin, Christopher J. 1976 "The perception of speech", in: Cartarette — Friedman (eds.), 175 — 226. Delattre, Pierre 1968 "From acoustic cues to distinctive features", Phonetica 18: 198 — 230. Dell, Gary S. 1986 "A spreading-activation theory of retrieval in sentence production", PsychReview 93: 283-321. Deutsch, Werner (ed.) 1981 The child's construction of language. New York: Academic Press. deVilliers, Jill 1984 "Syntactic overgeneralization", paper presented at the 9th Annual Conference on Language Development, Boston University, 1984. deVilliers, Jill — Peter deVilliers 1978 Language acquisition. Cambridge, Mass.: Harvard University Press. deVilliers, Jill —Helen Flusberg — Kenji Hakuta —M. Cohen 1979 "Children's comprehension of relative clauses", JPsyR 8: 499-518. Diehl, R. L.—Jeffrey L. Elman —Susan B. McCusker 1978 "Contrast effects on stop consonant identification", JExP : HPP 4: 5 9 9 609. Donaldson, Margaret—George Balfour 1968 "Less is more: A study of language comprehension in children", BJPsych 59: 461-471. Donaldson, Margaret — Roger Wales 1970 "On the acquisition of some relational terms", in: Hayes (ed.), 235 — 268. Donnenwerth-Nolan, Suzanne — Michael K.Tanenhaus —Mark S. Seidenberg 1981 "Multiple code activation in word recognition: Evidence from rhyme monitoring", JExP . HLM 7: 170-180. Dore, John 1974 "A pragmatic description of early language development", JPsyR 3: 343 — 350. 1975 "Holophrases, speech acts, and language universals", JChL 2: 21 —40. Dosher, Barbara A. 1982 "Effect of sentence size and network distance on retrieval speed", JExP.LMC 8: 173-207. Dowty, David — Lauri Karttunen — Arnold Zwicky (eds.) 1985 Natural language parsing. New York: Cambridge University Press. Dressier, Wolfgang U. (ed.) 1977 Trends in text-linguistics. New York and Berlin: de Gruyter. Ehrlich, Kate —Keith Rayner 1983 "Pronoun assignment and semantic integration during reading: Eye movements and immediacy of processing", JVLVB 22: 75 — 107.

Psycholinguistics

199

Eimas, Peter D. 1974 "Linguistic processing of speech by young infants", in: Schiefelbusch — Lloyd (eds.), 5 5 - 7 3 . Eimas, Peter D. —Joanne Miller (eds.) 1981 Perspectives on the study of speech. Hillsdale, N.J.: Erlbaum. Eimas, Peter D. — Einar R. Siqueland — Peter Jusczyk — Michael Vigorito 1971 "Speech perception in infants", Science 171: 303 — 306. Eimas, Peter D.—Vivien C. Tartter 1979 "On the development of speech perception: Mechanisms and analogies", in: H. W. R e e s - L . Lipsett (eds.), 155-193. Ellis, Andrew W. 1979 "Speech production and short-term memory", in: Morton — Marshall (eds.), 157-187. Erbaugh, Mary 1978 "Acquisition of temporal and aspectual distinctions in Mandarin", PRCLD15: 3 0 - 3 7 . Estes, William K. (ed.) 1978 Handbook of learning and cognitive processes. Vol. VI. Hillsdale, N.J.: Erlbaum. Fant, Gunnar 1960 Acoustic theory of speech production. The Hague: Mouton. 1968 "Analysis and synthesis of speech processes", in: Malmberg (ed.), 173-277. 1971 "Distinctive features and phonetic dimensions", in: Perren —Trim (eds.), 171-191. Fay, David 1978 "Transformations as mental operations", JChL 5: 143 — 152. Fay, David — Anne Cutler 1977 "Malaproprisms and the structure of the mental lexicon", Ling 18: 505 — 520. Felix, Sascha 1984 Two problems of language acquisition: On the interaction of Universal Grammar and language growth. Passau, W. Germany: University of Passau. Ferguson, Charles — Carol Farwell 1975 "Words and sounds in early language acquisition", Language 51: 419 — 430. Ferguson, Charles — Daniel Slobin (eds.) 1973 Studies of child language development. New York: Holt, Rinehart, Winston. Fischler, Ira 1981 "Context effects in word recognition", Cognition 10: 8 9 - 9 5 . Fletcher, Paul — Michael Garman (eds.) 1979 Language acquisition. Cambridge: Cambridge University Press. 1986 Language acquisition: Studies in first language development. Cambridge, England: Cambridge University Press. Flores d'Arcais, Giovanni B. 1978 "The acquisition of the subordinating constructions in children's language", in: Campbell-Smith (eds.), 349-393.

200

Carlota S. Smith

Flores d'Arcais, Giovanni B. — R. J. Jarvella (eds.) 1983 The processes of language understanding. New York: Wiley. Flores d'Arcais, Giovanni B.—William Levelt (eds.) 1970 Advances in psycholinguistics. New York: American Elsevier. Fodor, Jerry A. 1983 The modularity of mind: An essay on faculty psychology. Cambridge, Mass.: Bradford. Fodor, Jerry A. — Thomas Bever — Merrill F. Garrett 1966 "The active use of grammar in speech perception", P&P 1: 30 — 32. Fodor, Jerry A. — Merrill F. Garrett 1966 "Some reflections on competence and performance", in: Lyons —Wales (eds.), 135-179. 1967 "Some syntactic determinants of sentential complexity", P&P 1: 289 — 296. Fodor, Janet D. 1978 "Parsing strategies and constraints on transformations", Ling 19: 427 — All. Ford, Marilyn 1980 "Visual perception of rapidly presented word sequences of varying complexity", P&P 8: 215-221. Ford, Marilyn—Joan Bresnan — Ronald Kaplan 1982 "A competence-based theory of syntactic closure", in: Bresnan (ed.), 727-796. Ford, Marilyn — Virginia Holmes 1978 "Planning units and syntax in sentence production", Cognition 6: 35 — 53. Förster, Kenneth 1970 "Some effects of ambiguity on sentence comprehension", JVLVB 9: 699 — 706. 1979 "Levels of processing and the structure of the language processor", in: Cooper-Walker (eds.), 2 7 - 8 5 . Forster, Kenneth — Elizabeth S. Bednall 1976 "Terminating and exhaustive search in lexical access", M&C 4: 53 — 61. Forster, Kenneth — Carol Davis 1984 "Repetition priming and frequency attenuation in lexical access", Journal of Experimental Psychology: Learning, Memory and Cognition 10: 680 — 698. Forster, Kenneth —I. Olbrei 1973 "Semantic heuristics and syntactic analysis", Cognition 2: 319 — 347. Foss, Donald J. 1969 "Decision processes during sentence comprehension: Effects of lexical item difficulty and position upon decision times", JVLVB 8: 457—462. 1970 "Some effects of ambiguity upon sentence comprehension", JVLVB 9: 699-706. in press "Experimental psycholinguistics", Annual Review of Psychology. Foss, Donald J. — Michelle A. Blank 1980 "Identifying the speech codes", CogPsych 12: 1 - 3 1 . Foster, Susan 1982 "Learning to develop a topic", PRCLD 21: 6 3 - 7 0 .

Psycholinguistics

201

Fourcin, Adrian J. 1979 "Auditory patterning and vocal fold vibration", in: Lindblom—Ohman (eds.), 167-175. Fowler, Carol A. 1986 "An event approach to the study of speech perception from a direct-realist perspective", Journal of Phonetics 14, no. 1: 3 — 28. Frazier, Lyn—Janet D. Fodor 1978 "The sausage machine: A new two-stage parsing model", Cognition 6: 291-325. Frazier, Lyn — Keith Rayner 1982 "Making and correcting errors during sentence comprehension: Eye movements in the analysis of structurally ambiguous sentences", Cog Psych 14: 178-210. Freedle, Roy O. (ed.) 1977 Discourse processes and comprehension. Norwood, N.J.: Ablex. Fromkin, Victoria 1973 Speech errors as linguistic evidence. The Hague: Mouton. Fry, Dennis B. — Adele S. Abrahamson — Peter D. Eimas—Alvin M. Liberman 1962 "The identification and discrimination of synthetic vowels", L&S 5: 171 — 189. Fujisaki, Hiroya 1979 "On the modes and mechanisms of speech perception: Analysis and interpretation of categorical effects in discrimination", in: Lindblom — Ohman (eds.), 177-189. Fujisaki, Hiroya — Takashi Kawashima n. d. "Some experiments on speech perception and a model for the perceptual mechanism" (Rep. Engineering Res. Institute. Faculty of Engineering, U of Tokyo 28), 6 7 - 7 3 . Furrow, David 1984 "Young children's use of prosody", JChl 11: 203-214. Furrow, David — Katharine Nelson — Helen Benedict 1979 "Mothers' speech to children and syntactic development: Some simple relationships", JChL 6: 4 2 3 - 4 3 2 . Garfield, Jay L. (ed.) 1987 Modularity in knowledge representation and natural language understanding. Cambridge, Mass.: MIT Press. Garnham, Alan 1981 "Mental models as representations of texts", M&C 9: 560-565. Garnham, Alan — Jane Oakhill — Philip Johnson-Laird 1982 "Referential continuity", Cognition 11: 2 9 - 4 6 . Garrett, Merrill 1975 "The analysis of sentence production", in: G. H. Bower (ed.), 133 — 177. 1981 "Levels of processing in sentence production", in: Butterworth (ed.), 177-220. Garrod, Simon — Anthony Sanford 1977 "Interpreting anaphoric resolution", JVLVB 16: 7 7 - 9 0 . Gay, Thomas 1979 "Coarticulation in some consonant-vowel and consonant cluster vowel syllables", in: Lindblom — Ohman (eds.), 69 — 76.

202

Carlota S. Smith

Gelman, Rachel — Ellen Markman 1986 "Understanding natural kind terms: A developmental comparison", PRCLD Vol. 25: 4 1 - 4 8 . Gentner, Deirdre 1978 "On relational meaning: The acquisition of verb meaning", Child Development 49: 988-998. Glass, Arnold L. — Keith J. Holyoak 1975 "Alternative conceptions of semantic memory", Cognition 3: 313 — 339. Gleason, Jean B. (ed.) 1985 The development of language. Columbus, Ohio: Merrill. Gleitman, Lila —Elissa Newport — Henry Gleitman 1984 "The current status of the motherese hypothesis", JChL 11: 4 3 - 8 0 . Glucksberg, Sam —Michael E. McCloskey 1981 "Decisions about ignorance: Knowing that you don't know", JExP: HLM 7: 311 - 3 2 5 . Glucksberg, Sam — Tom Trabasso — Jerry Wald 1973 "Language and mental operations", CogPsych 5: 338 — 370. Goodman, George O.—James L. McClelland — Raymond W. Gibbs 1981 "The role of syntactic context in word recognition", M&C 9: 580-586. Graesser, Arthur — Sallie E. Gordon —Janet Sawyer 1979 "Recognition memory for typical and atypical actions in scripted activities", JVLVB 8: 319-332. Graesser, Arthur — George Mandler 1975 "Recognition memory for the meaning and surface structure of sentences", JExP.HLM 7: 238-248. Graesser, Arthur —Scott P. Robertson — Edward R. Lovelace — Daniel M. Swinehart 1980 "Answer to why questions expose the organization of story plot and predict recall of actions", JVLVB 19: 110-119. Greenfield, Patricia M.—J. H. Smith 1976 The structure of communication in early language development. New York: Academic Press. Grimshaw, Jane 1981 "Form, function, and the language acquisition device", in: B a k e r McCarthy (eds.), 165-182. 1986 "Linguistic mistakes: The role of negative evidence in language learning", paper presented at the 11th Annual Conference on Language Development, Boston University, 1986. Groenendijk, J. et al. (eds.) 1981 Formal models in the study of language. Amsterdam: North Holland. Grosjean, Francois — Lysiane Grosjean — Harlan Lane 1979 "The patterns of silence: Performance structures in sentence production", CogPsych 11: 5 8 - 8 1 . Grosz, Barbara — Candace Sidner 1986 "Attention, intention and the structure of discourse", Computational Linguistics 12: 175-204. Haith, Marshall M. —Joseph L. Campos (eds.) 1983 Infancy and developmental psychobiology. Vol. II of Handbook of child psychology. New York: Wiley.

Psycholinguistics

203

Hakes, David 1980 The development of metalinguistic abilities in young children. New York: Springer-Verlag. Hakuta, Kenji 1981 "Grammatical description vs configurational arrangement in language: The case of relative clauses in Japanese", Cognition 9: 97 — 236. Halle, Morris — Joan Bresnan — George Miller (eds.) 1978 Linguistic theory and psychological reality. Cambridge, Mass.: MIT Press. Halliday, Michael A. K. 1975 Learning how to mean. Amsterdam: Elsevier. Hamilton, Charles —McCay D. Vernon (eds.) 1976 The development of cognitive processes. New York: Academic Press. Harner, Lorraine 1976 "Children's comprehension of linguistic reference to past and future", JPsyR 5: 6 5 - 8 4 . 1981 "Children talk about the time and aspect of actions", Child Development 52: 4 9 8 - 5 0 6 . Harris, Harold J. —Mark Davies 1987 "Learning and triggering", First Language 7, No. 1: 3 — 39. Haviland, K. S. S . - H e r b e r t H. Clark 1974 "What's new? Acquiring new information as a process in comprehension", JVLVB 13: 512-521. Hayes, John R. (ed.) 1970 Cognition and the development of language. New York: Wiley. Hayes-Roth, Barbara — Frederick Hayes-Roth 1977 "The prominence of lexical information in memory: Representations of meaning", JVLVB 16: 119-136. Hickman, Maya 1982 "Creating referents in discourse: A developmental analysis of linguistic cohesion", in: Papers from the 16th Regional Meeting of the Chicago Linguistic Society. Chicago Linguistic Society: 192 — 203. Hirst, Graeme 1981 "Discourse oriented anaphora resolution in natural language understanding: A review", AmJComputLing 1: 85 — 98. Hogaboam, Thomas W. — Charles A. Perfetti 1975 "Lexical ambiguity and sentence comprehension", JVLVB 14: 265—274. Hobbs, Jerry 1985 On the coherence and structure of discourse (Center for the Study of Language and Information, Report No. 37). Stanford: Stanford University. Holmes, Virginia M. —R. Arwas — Merrill E. Garrett 1977 "Prior context and the perception of lexically ambiguous sentences", M&C 5: 103-110. Holyoak, Keith J. —Arnold L. Glass 1975 "The role of contradictions and counter-examples in the rejection of false sentences", JVLVB 14: 215-239. Horton, David L. — James J. Jenkins (eds.) 1971 The perception of language. Columbus, Ohio: Chas Merrill.

204

Carlota S. Smith

Huggins, Andrew W. 1968 "The perception of timing in natural speech. I: Compensation within the syllable", L&S 11: 1 - 1 1 . 1972 "On the perception of temporal phenomena in speech", JAS 51: 1279 — 1290. Hurtig, Richard 1978 "The validity of clausal processing strategies at the discourse level", Discourse Processes 1978: 195-202. Huxley, Renata —David E. Ingram (eds.) 1971 Language acquisition: Models and methods. New York: Academic Press. Hyams, Nina 1987 "Pro-drop as a parameter", in: Roeper — Williams (eds.), 1 —22. Ingram, David 1979 "Phonological patterns in the speech of young children", in: Fletcher — Garman (eds.) (1979). Jackson, A. — Jacqueline Morton 1984 "Facilitation of auditory word recognition", M & C 12: 568-574. Jacubowicz, Celia 1984 "Local Α-binding: A structural contraint in the development of the interpretation of referentially dependent expressions", Paper presented at the 9th Annual Conference on Language Development, Boston University, 1984. Jakobson, Roman 1968 Child language, aphasia, and phonological universals. The Hague: Mouton. 1971 "Observations sur le classement phonologique de consonnes", Proceedings of the 3rd Congress on Phonetic Science, Ghent, in: Roman Jakobson: Selected Writings, Vol. I. The Hague: Mouton, 1971, 272-279. Jarvella, Robert 1970 "Effects of syntax on running memory span for connected discourse", PsychonSci 19: 235-236. 1971 "Syntactic processing of connected speech", JVLVB 10: 409-416. Johnson-Laird, Philip N. 1983 Mental models: Toward a cognitive science of language, inference, and consciousness. Cambridge, Mass.: Harvard University Press. Johnston, James C. — Daniel Slobin 1979 "The development of locative expressions in English, Italian, Serbo-Croatian and Turkish", JChL 6: 529-546. Joshi, Aravind K. —Bonnie Webber—Ivan A. Sag (eds.) 1981 Elements of discourse understanding. Cambridge, England: Cambridge University Press. Jusczyk, Peter A. 1986 "A review of speech perception", in: Kaufman — Thomas — Boff (eds.), 27: 1-57. Just, Marcel Α. —Patricia Carpenter (eds.) 1976 Cognitive processes in comprehension. Hillsdale, N.J.: Erlbaum. Kamp, Johann 1981 "A theory of truth and semantic interpretation", in: Groenendijk et al. (eds.), 227-322.

Psycholinguistics 1985

205

"Context, thought, and communication", Proceedings of the Aristotelian Society 85: 2 3 9 - 2 6 1 . Kaplan, Ronald 1973 "A general syntactic processor", in: Rustin (ed.), 193 — 241. Karmiloff-Smith, Annette 1977 "More about the same: Children's understanding of post-articles", JChL 4: 377-394. 1979 "Language development after five", in: Fletcher—Garman (eds.), 455-474. 1985 "Language and cognitive processes from a developmental perspective", Language and Cognition Proceedings No. 1: 61 —83. Kaufman, Lloyd — James Thomas — Kenneth Boff (eds.) 1986 Handbook of perception and performance. New York: Wiley. Keenan, Janice M. — Brian MacWhinney — Deborah Mayhew 1977 "Pragmatics in memory: A study of natural conversation", JVLVB 16: 549-560. Kelso, J. A. Scott-Elliot L. Saltzman - Betty Tuller 1986 "The dynamical perspective on speech production: Data and theory", Journal of Phonetics 14, No. 1: 29 — 60. Kemper, Susan 1980 "Memory for the form and force of declaratives and interrogatives", M&C 8: 367-371. Kieras, David E. 1982 "Topicalization effects in cued recall of technical prose", M&C 9: 660 — 669. Kintsch, Walter 1976 "On recognizing linguistic stories", in: Just — Carpenter. 1977 The representation of meaning in memory. Hillsdale, N.J.: Erlbaum. Kiparsky, Paul —Lisa Menn 1977 "On the acquisition of phonology", in: Macnamara (ed.), 47 — 78. Klatt, Dennis 1976 "Linguistic uses of segmental duration in English: Acoustic and perceptual evidence", JAS 59: 1208-1236. 1979 "Speech perception: A model of acoustic-phonetic analysis and lexical access", JPhon 7: 279-312. Klatt, Dennis — William E. Cooper 1975 "Perception of segment duration in sentence contexts", in: Cohen —Nooteboom (eds.), 6 9 - 8 9 . Klima, Edward — Ursula Bellugi 1967 "Syntactic regularities in the speech of children", in: Lyons—Wales (eds.), 183-219. Kolers, Paul A . - S a n d r a R. Palef 1976 "Knowing not", M&C 4: 553-558. Kozimsky, Ely — Walter Kintsch — Lyle E. Bourne, Jr. 1982 "Decision making with texts: Informational analysis and schema acquisition', JExP . G 110: 363-380. LaBerge, David —S. Jay Samuels (eds.) 1977 Basic processes in reading: Perception and comprehension. Hillsdale, N. J.: Erlbaum.

206

Carlota S. Smith

Ladefoged, Peter — Richard Harshman 1979 "Formant frequencies and movement of the tongue", in: Lindblom — Ohman (eds.), 25 — 34. Lass, Norman J. (ed.) 1976 Contemporary issues in experimental phonetics. Springfield, IL: C. C. Thomas. 1983 Speech and language: Advances in basic research and practice. Vol. IX. New York: Academic Press. Lebeaux, David 1987 "Comments on pro-drop as a parameter", in: Roeper—Williams (eds.), 23-40. Lederberg, Amy R. — Michael Maratsos 1981 "Children's use of semantic analysis in the interpretation of missing subjects: Further evidence against the MDP", JPsyR 10: 8 9 - 1 1 0 . Lehiste, Ilse 1975 "The phonetic structure of paragraphs", in: Cohen — Nooteboom (eds.), 195-206. 1979 "The perception of sentence and paragraph boundaries", in: Lindblom — Ohman (eds.), 191-201. Lehiste, Ilse—J. P. Olive —L. A. Streeter 1975 "The role of duration in disambiguation of syntactically ambiguous sentences", JAS 60: 1199-1201. Levelt, William J. M. 1970 "A scaling approach to the study of syntactic relations", in: Flores d'Arcais-Levelt (eds.), 109-121. 1983 "Monitoring and self-repair in speech", Cognition 14: 41 — 104. Levine, Susan C. — Susan Carey 1982 "Up front: The acquisition of a concept and a word", JChL 9: 645 — 658. Liberman, Alvin M. 1970 "The grammars of language and speech", CogPsych 1: 301—323. Liberman, Alvin M. —Frank S. Cooper—K. S. Harris —Peter MacNeilage 1962 "A motor theory of speech perception", in: Proceedings of the Speech and Communication Seminar, Stockholm. Vol. II, Paper D3. Stockholm: Speech Transmission Laboratory, Royal Institute of Technology. Liberman, Alvin M. —Frank S. Cooper—David P. Shankweiler — Michael StuddertKennedy 1967 "Perception of the speech code", Psycholog R 74= 432-461. Lindblom, Björn Ε. 1962 "A spectrographic study of vowel reduction", JAS 35: 1773 - 1 7 8 1 . 1978 "Final lengthening in speech and music", Travaux de l'Institut de Phonetique de Lund 13: 33 — 56. Lund: Gleerup. 1980 "The goal of phonetics, its unification and application", Phonetica 37: 726-728. Lindblom, Björn Ε. —Michael Studdert-Kennedy 1967 "On the role of formant transitions in vowel recognition", JAS 35: 173 — 181. Lindblom, Björn Ε.— Sven Ohman (eds.) 1979 Frontiers of speech communication research. New York: Academic Press.

Psycholinguistics

207

Lindblom, Björn Ε. — Rolf Zetterstrom (eds.) 1986 Precursors of early speech. New York: Stockton Press. Lindsay, Peter H. —Donald A. Norman 1977 Human information processing. New York: Academic Press. Lindzey, Gardner — Elliot Aronson (eds.) 1985 Handbook of social psychology. Vol. II. (3rd edition). New York: Random House. Lisker, Leigh — Arthur Abramson 1964 "A cross-language study of voicing in initial stops: Acoustical measurements", Word 20, No. 3: 384-422. 1967 "Some effects of context on voice onset time in English stops", L&S 10: 1-28.

Lockhart, Robert S. — Fergus I. M. Craik —Larry L. Jacoby 1976 "Depth of processing in recognition and recall: Some aspects of a general memory system", in: J. Brown (ed.), 75 — 102. Loftus, Elizabeth — William R. Cole 1974 "Retrieving attribute and name from semantic memory", JExP 102: 1116-1122. Lust, Barbara 1977 "Conjunction reduction in child language", JChL 4: 2 5 7 - 2 8 7 . Lust, Barbara (ed.) 1986 Studies in the acquisition of anaphors. Dordrecht: Reidel. Lust, Barbara — Yu-Chin Chien 1984 "The structure of coordination in first language acquisition of Mandarin Chinese: Evidence for a universal", Cognition 17: 49 — 83. Lyons, John — Roger Wales (eds.) 1967 Psycholinguistics papers. Edinburgh: University Press. MacKay, David G. 1970 "Spoonerisms and the structure of errors in the serial ordering of speech", Neuropsychologia 8: 323 — 350. Macnamara, John "From sign to language", in: Macnamara (ed.), 11—35. 1977 Macnamara, John (ed.) 1977 Language learning and thought. New York: Academic Press. MacNeilage, Peter "Motor control of serial ordering of speech", PsychologR 7: 182 — 196. 1970 MacNeilage, Peter —Peter Ladefoged 1976 "The production of speech and language", in: Carterette — Friedman (eds.), 7 5 - 1 2 0 . MacWhinney, Brian 1975 "Rules, rote and analogy in morphological formations in Hungarian children", JChl 2: 6 5 - 7 8 . 1978 Processing a first language: the acquisition of morphophonology (Society for Research of Child Development Monograph 43). Chicago. Malmberg, Bertil (ed.) 1968 Manual of phonetics. Amsterdam, Holland: North Holland. Mandle, Heinz—Nancy L. Stein — Thomas Trabasso (eds.) 1984 Learning and comprehension of text. Hillsdale, N.J.: Erlbaum.

208

Carlota S. Smith

Mandler, George 1967 "Organization in memory", in: Spence — Spence (eds.), 327 — 372. Mandler, George —Nancy S. Johnson 1977 "Remembrance of things parsed: Story structure and recall", CogPsych 9: 111-151. Maratsos, Michael 1976 The use of definite and indefinite reference in young children: An experimental study in semantic acquisition. Cambridge: Cambridge University Press. 1982 "The child's construction of grammatical categories", in: Wanner — Gleitman (eds.), 240-266. Marcus, Mitchell 1980 A theory of syntactic recognition for natural languages. Cambridge, Mass.: MIT Press. Markman, Ellen 1987 "How children constrain the possible meanings of words", in: Neisser (ed.), 255-287. Marslen-Wilson, William 1975 "Sentence perception as an interactive parallel process", Science 189: 6-228.

1984

"Function and process in spoken word recognition", in: Bouma — Bouwhis (eds.), 125-150. Marslen-Wilson, William — Loraine Tyler 1980 "The temporal structure of spoken language understanding", Cognition 8: 1 - 7 1 . Marslen-Wilson, William —A. Welsh 1978 "Processing interactions and lexical access during word recognition in continuous speech", Cog Psych 10: 29 — 63. Martin, Jacqueline G. 1971 "Some acoustic and grammatical features of spontaneous speech", in: Horton—Jenkins (eds.), 47 — 68. 1972 "Rhythmic (hierarchical) vs. serial structure in speech and other behavior", PsychologR 79: 487-509. Massaro, Dominic W. — Marshall M. Cohen 1966 "The contribution of fundamental frequency and voice onset time to the [zi]-[si] distinction", JAS 60: 704-717. Mazuka, R. —Barbara Lust —T. Wakayama — William Snyder 1986 "Distinguishing effects of parameters in early syntax acquisition: A crosslinguistic study of Japanese and English", PRCLD 25: 73 — 82. McClelland, James L. 1979 "On the time relations of mental processes: An examination of systems of processes in cascade", PsychologR 86: 278 — 330. McClelland, James L. — Jeffrey Elman 1986 "The trace model of speech perception", Cognitive Psychology 18: 1 —86. McCloskey, Michael —Kurt Bigler 1980 "Focussed memory search in fact retrieval", M&C 4: 553 — 558. McCloskey, Michael —Sam Glucksberg 1979 "Decision process in verifying category membership statements: Implications for models of semantic memory", CogPsych 1: 1—37.

Psycholinguistics

209

McCloskey, Michael—Jeffrey Santee 1981 "Are semantic memory and episodic memory distinct systems?" JExP.HLM 7: 6 6 - 7 1 . Menyuk, Paula—Lise Menn 1979 "Early strategies for the perception and production of words", in: Fletcher-Garman (eds.) (1979), 4 9 - 7 0 . Meyer, David E. — Roger W. Schvaneveldt 1971 "Facilitation in recognizing pairs of words: Evidence of dependence between retrieval operations", JExP 90: 227—234. 1976 "Meaning, memory structure, and mental processes", Science 192: 27 — 33. Miller, George 1962 "Some psychological studies of grammar", American Psychologist 17: 748-762. Miller, George — Eugene Galanter — Karl Pribram 1960 Plans and the structure of behavior. New York: Holt, Rinehart, Winston. Miller, George — Philip Johnson-Laird 1976 Language and perception. Cambridge, Mass.: Harvard University Press. Miller, Joanne D.—Walter Kintsch 1980 "Readability and recall of short prose passages: A theoretical analysis", JExP. HLM 6: 335-354. Miller, Joanne D. - C h e t S. W i e r - Richard E. P a s t o r e - W . J. K e e l y - Robert J. Dooling 1976 "Discrimination and labelling of noise-buzz sequences with varying noiselead times: An example of categorical perception", JAS 60: 410—417. Minsky, Marvin 1965 "A framework for representing knowledge", in: Winston (ed.), 211—280. Moeser, Shannon D. —Albert S. Bregman 1972 "The role of reference in acquisition of a miniature artificial language", JVLVB 11: 759-765. Moore, Terence (ed.) 1973 Cognitive development and the acquisition of language. New York: Academic Press. Morgan, James L. — Elissa L. Newport 1981 "The role of constituent structure in the induction of an artificial language", JVLVB 10: 6 7 - 8 5 . Morton, John 1970 "A functional model for memory", in: Norman (ed.), 203 — 254. 1979 "Word recognition", in: Morton - Marshall (eds.), 107 - 1 5 6 . Morton, John —John Long 1976 "Effects of word transitional probability on phoneme identification", JVLVB 15: 4 3 - 5 1 . Morton, John—John C. Marshall 1979 Structures and processes. Vol. II of Psycholinguistics. Cambridge, Mass.: M.I.T Press. Moulton, Jane —George Robinson 1981 The organization of language. Cambridge: Cambridge University Press. Murreil, G. Α. —John Morton 1974 "Word recognition and morphemic structure", JExP 102: 9 6 3 - 9 6 8 .

210

Car lota S. Smith

Naremore, Rita C. (ed.) 1984 Recent advances in language sciences. San Diego, California: College Hill Press. Neisser, Ulrich (ed.) 1987 Concepts and conceptual development. Cambridge, England: Cambridge University Press. Newport, Elissa — Henry Gleitman — Lila Gleitman 1977 "Mother I'd rather do it myself: Some effects and noneffects of maternal speech style", in: Snow — Ferguson (eds.), 109—149. Nilsson, Lars-Göran 1979 "Functions of memory", in: Nilsson (ed.), 3 — 15. Nilsson, Lars-Göran (ed.) 1979 Perspectives on memory research. Hillsdale, N.J.: Erlbaum. Norman, Donald A. (ed.) 1970 Models of human memory. New York: Academic Press. Norman, Donald A. — David G. Rumelhart 1975 "Memory for knowledge", in: Norman — Rumelhart (eds.), 3 — 32. Norman, Donald A. — David G. Rumelhart (eds.) 1975 Explorations in cognition. LNR Research Group. San Francisco: Freeman. Nooteboom, Sibout G. —Antonie Cohen 1975 "Anticipation in speech production and its implications for perception", in: Cohen —Nooteboom (eds.), 124 — 145. Norris, Dennis 1982 "Autonomous processes in comprehension: A reply to Marslen-Wilson and Tyler", Cognition 11: 9 7 - 1 0 3 . 1986 "Word recognition: Context effects without priming", Cognition 22: 93 — 136. Oller, D. Kimbrough —Leslie A. Wieman —William J. Doyle —Carol Ross 1976 "Infant babbling and speech", JChL 3 : 1 - 1 1 . Olson, Gary 1973 "Developmental changes in memory and the acquisition of language", in: Moore (ed.), 145-157. Osgood, Charles 1980 Lectures on language performance. New York: Springer-Verlag. Osgood, Charles — Michael J. Bock 1977 "Salience and sentencing: Some production principles", in: Rosenberg (ed.), 8 9 - 1 4 0 . Osgood, Charles — Rumjahn Hoosain 1974 "Salience of the word as a unit in the perception of language", P & P 145: 168-192. Pastore, Richard E. — Carol Friedman —K. J. Baffuto — Elizabeth A. Fink 1976 "Categorical perception of both simple auditory and visual stimuli", paper presented at the meeting of the Acoustical Society of America, Washington, D.C., 1976. Perfetti, Charles A. — Susan R. Goldman 1974 "Discourse functions of thematization and topicalization", JPsyR 13: 70 — 79.

Psycholinguistics

211

Perren, G. Ε.—J. L. Μ. Trim (eds.) 1971 Selected papers from the Second International Congress of Applied Linguistics. Cambridge, England: Cambridge University Press. Piaget, Jean 1926 The language and thought of the child. New York: Harcourt Brace. 1972 The child's conception of the world. Totawa, N.J.: Littlefield, Adams. Pinker, Steven 1979 "Formal models of language learning", Cognition 7: 217 — 283. 1984 Language learning and language development. Cambridge, Mass.: Harvard University Press. Pinker, Steven—L. Frost 1984 "Constraining productivity: Children's passivization of verbs with different thematic roles", paper presented at the 9th Annual Conference on Child Language Development, Boston University, 1984. Pisoni, David B. — Richard N. Aslin—Alan J. Percy —Beth L. Hennessy 1982 "Some effects of laboratory training on identification and discrimination of voicing contrasts in stop consonants", JExP: HPP 8: 297 — 314. Port, Diane Kewley — Malcolm S. Preston 1972 "Early apical stop production: A voice onset time analysis", Haskins Lab Status Reports on Speech Research. SR 29/30: 125-149. Preece, A. 1987 "The range of narrative forms conversationally produced by young children", JChL 14: 353-373. Putnam, Hilary 1975 "Mind, language and reality", (Philosophical Papers, Vol. II). Cambridge: Cambridge University Press. Pye, Charles 1986 "Quiche Mayan speech to children", JChL 13, No. 1: 8 5 - 1 0 0 . Randall, Janet 1984 "Indirect positive evidence: A new route for retreat", paper presented at the 9th Annual Conference on Language Development, Boston University, 1984. Raphael, Lawrence J. —Michael F. Dorman 1977 "Perceptual equivalence of cues for the fricative-affricate contrast", JAS 6: 578 A. Ratcliff, Roger—Gail McKoon 1981 "Does activation really spread?" PsychologR 88: 4 5 4 - 4 6 2 . 1982 "Speed and accuracy in the processing of false statements about semantic information", JExP: LMC 8: 6 - 3 6 . Reber, Arthur S. —Rhianon Allen 1976 "Analogic and abstraction strategies in synthetic grammar learning: A functionalist approach", Cognition 6: 189 — 221. Reder, Lynne M. 1979 "The role of elaborations in memory for prose", CogPsych 11: 221—234. Reed, Adam V. 1973 "Speed-accuracy tradeoff in recognition memory", Science 181: 574—576. Rees, Η. W. — Laurence Lipsett (eds.) 1979 Advances in child development and behavior. Vol. 13. New York: Academic Press.

212

Carlota S. Smith

Repp, Bruno H. 1983 "Categorical perception: Issues, methods, findings", in: Lass (ed.), 243-335. Restle, Frank et al. (eds.) 1975 Cognitive theory. Vol. I. Hillsdale, N.J.: Erlbaum. Richards, Meredith 1979 "Sorting out what's in a word from what's not: Evaluating Clark's acquisition theory", JExP 27: 1—47. Rips, Lance J. —Edward J. Shoben—Edward E. Smith 1973 "Semantic distance and the verification of semantic relatedness", JVLVB 7: 2 9 - 3 2 1 . Roeper, Thomas —Edwin Williams (eds.) 1987 Parameters and acquisition theory. Dordrecht, Holland: Reidel. Rosch, Eleanor 1973 "On the internal structure of perceptual and semantic categories", in: Moore (ed.), 111-144. Rosch, Eleanor —B. Lloyd (eds.) 1978 Cognition and categorization. Hillsdale, N.J.: Erlbaum. Rosenberg, Sheldon (ed.) 1977 Sentence production. Hillsdale, N.J.: Erlbaum. Rubenstein, Herbert —M. Garfield—J. Millin 1970 "Homographic entries in the internal lexicon", JVLVB 9: 487-494. Rumelhart, David E. 1977 Introduction to human information processing. New York: Wiley. Rustin, Randall (ed.) 1973 Natural language processing. New York: Algorithmics Press. Sawashima, Masayuki —D. Steven Cooper (eds.) 1977 Dynamic aspects of speech production. Tokyo: University of Tokyo Press. Scarborough, Don L. —C. Cortese — H. S. Scarborough 1977 "Frequency and repetition effects in lexical memory", JExP.HPP 3: 1-17. Schiefelbusch, Richard—Diane Bricker (eds.) 1981 Early language: Acquisition and intervention. Baltimore: University Park Press. Schiefelbusch, Richard —Lyle Lloyd (eds.) 1974 Language perspectives. Baltimore: University Park Press. Schlesinger, Izchak M. 1974 "Relational concepts underlying language", in: Schiefelbusch — Lloyd (eds.), 129-151. 1977 "The role of cognitive development and linguistic input in language acquisition", JChL 4: 153-169. Searle, C. L.—J. Z. Jacobson —S. G. Rayment 1979 "Phoneme recognition based on human audition", JAS 65: 799-809. Shank, Roger — Raziel Abelson 1977 Scripts, plans, goals and understanding. New York: Halsted. Shankweiler, Donald—Winifred Strange — Robert Verbrugge 1977 "Speech and the problem of perceptual constancy", in: Shaw — Bransford (eds.), 315-345.

Psycholinguistics

213

Shanon, P. 1979 "Have you been to Paris?", Acta Psychologica 43: 313-328. Shattuck-Hufnagel, Stephanie 1979 "Speech errors as evidence for a serial-ordering mechanism in sentence production", in: Cooper—Walker (eds.), 295 — 342. Shattuck-Hufnagel, Stephanie — Dennis Klatt 1978 "The limited use of distinctive features and markedness in speech production: Evidence from speech errors", JVLVB 18: 41—56. Shaw, Robert—John Bransford (eds.) 1977 Perceiving, acting, and knowing. Hillsdale, N.J.: Erlbaum. Sheldon, Amy 1974 "The role of parallel function in the acquisition of relative clauses in English", JVLVB 13: 272-281. Shipley, Elizabeth — Carlota Smith — Lila Gleitman 1978 "A study in the acquisition of language: Free responses to commands", in: Bloom (ed.), 347-370. Shoben, Edward J. —Κ. T. Westcourt — Edward E. Smith 1978 "Sentence verification, sentence recognition and the semantic-episodic distinction", JExP HLM 4: 304-317. Simoes, Maria P. —Carol Stoel-Gammon 1979 "The acquisition of inflections in Portuguese: A study of the development of person markers on verbs", JChL 6: 53 — 68. Simpson, Greg B. 1981 "Meaning dominance and semantic context in processing of lexical ambiguity", JVLVB 11: 683-692. Sinclair-de-Zwart, Hermione 1967 Langage et operations: Sous-systemes linguistiques et operations concretes. Paris: Dunod. 1971 "Sensorimotor action patterns as a condition for the acquisition of syntax", in: Huxley —Ingram (eds.), 121 — 135. Slobin, Daniel 1966 "The acquisition of Russian as a native language", in: Smith —Miller (eds.). 1973 "Cognitive prerequisites for the acquisition of grammar", in: Ferguson — Slobin (eds.), 175-208. 1985 The cross-linguistic story of language acquisition. 2 vols. Hillsdale, N.J.: Erlbaum. Slobin, Daniel — Charles Welsh 1973 "Elicited imitation as a research tool in developmental psycholinguistics", in: Ferguson — Slobin (eds.), 485 — 495. Smith, Carlota S. 1970 "An experimental approach to children's linguistic competence", in: Hayes (ed.), 109-135. 1980 "The acquisition of time talk: Relations between child and adult grammar", JChL 7: 257-288. 1981 "Comments on the acquisition of productive morphology", in: B a k e r McCarthy (eds.), 151-163. Smith, Edward E. 1978 "Theories of semantic memory", in: Estes (ed.), 1 —56.

214

Carlota S. Smith

Smith, Frank — George Miller (eds.) 1966 The genesis of language. Cambridge, Mass.: M.I.T. Press. Snow, Catherine —N. Smith —Marion Hoefnagel-Hohle 1980 "The acquisition of some Dutch morphological rules", JChL 7: 539 — 554. Snow, Catherine — Charles Ferguson (eds.) 1977 Talking to children. Cambridge: Cambridge University Press. Solan, Larry 1981 "The acquisition of structural restrictions on anaphora", in: Tavakolian (ed.), 5 9 - 7 3 . Spence, Kenneth W. — Janet T. Spence (eds.) 1967 The psychology of learning and motivation. Vol. 1. New York: Academic Press. Spilich, G. J. —Gregg T. Vesonder—Harry H. Chiesi—James F. Voss 1979 "Text processing of domain-related information", JVLVB 18: 275-290. Stanners, Robert F. —Gary B. Forbach 1973 "Analysis of letter strings in word recognition", JExP: G 98: 31 —35. Stanners, Robert F.—James J. Neiser—W. P. Harnon —R. Hall 1979 "Memory representations for morphologically related words", JVLVB 18: 399-412. Stevens, Kenneth 1980 "Acoustic correlates of some phonetic categories", J ASA 68: 836 — 842. Stevens, Kenneth — Sheila Blumstein 1981 "The search for invariant acoustic correlates of phonetic features", in: Eimas —Miller (eds.), 1—38. Stevens, Kenneth — Arthur S. House 1972 "Speech perception", in: Tobias (ed.), 1 —62. Stevens, Kenneth — Dennis H. Klatt 1974 "The role of formant transitions in the voice-voiceless distinction for stops", JAS 55: 653-659. Stevens, Kenneth — Joseph S. Perkell 1977 "Speech physiology and phonetic features", in: Sawashima—Cooper (eds.), 323-341. Stoel-Gammon, Carol —N. L. Hedberg 1984 "A longitudinal study of cohesion in the narratives of young children", paper presented at the 3rd International Congress for the Study of Child Language, Austin, Texas, 1984. Studdert-Kennedy, Michael 1976 "Speech perception", in: Lass (ed.), 2 4 3 - 2 9 3 . 1980 "Speech perception", L & S 23: 45 - 66. Summerfield, Quentin A.— Mark P. Haggard 1974 "Perceptual processing of multiple cues and contexts: Effects of following vowel upon stop consonant voicing", JPhon 2: 279 — 295. Swinney, David A. 1979 "Lexical access during sentence comprehension: (Re)consideration of context effects", JVLVB 18: 645-659. Swinney, David Α. —David T. Hakes 1976 "Effects of prior context upon lexical access during sentence comprehension", JVLVB 15: 681-689.

Psycholinguistics

215

Taft, Marcus Α. 1979 "Recognition of affixed vowels and the word frequency effect", M & C : 263-272. Taft, Marcus A. — Kenneth Forster 1975 "Lexical storage and retrieval of prefixed words", JVLVB 14: 6 3 8 - 6 4 7 . Tanenhaus, Michael K. — Greg N. Carlson —Mark S. Seidenberg 1985 "Do listeners compute linguistic representations?" in: Dowty — Karttun e n - Z w i c k y (eds.), 359-408. Tanenhaus, Michael K. —Helen P. Flanigan — Mark S. Seidenberg 1980 "Orthographic and phonological priming in auditory and visual word recognition", M&C 8: 513-520. Tanenhaus, Michael K. —James M. Leiman —Mark S. Seidenberg 1979 "Evidence for multiple stages in the processing of ambiguous words in syntactic contexts", JVLVB 18: 4 2 7 - 4 4 0 . Tanouye, Ellen 1979 "The acquisition of verbs in Japanese children", PRCLD 17: 49 — 56. Tavakolian, Susan (ed.) 1981 Language acquisition and linguistic theory. Cambridge, Mass.: M.I.T. Press. Tobias, Jerry (ed.) 1972 Foundations of modern auditory theory. Vol. II. New York: Academic Press. Trabasso, Thomas — T. Secco — Paul van den Broek 1984 "Causal cohesion and story coherence", in: Mandle—Stein —Trabasso (eds.), 8 3 - 1 1 1 . Tulving, Endel 1972 "Episodic and semantic memory", in: Tulving — Donaldson (eds.), 381-403. 1974 "Cue-dependent forgetting", American Scientist 64: 74 — 82. Tulving, Endel — Margaret Donaldson (eds.) 1972 Organization of memory. New York: Academic Press. Tyler, Lorraine — William Marslen-Wilson 1977 "The on-line effect of semantic constraints on syntactic processing", JVLVB 16: 683 — 692. Underwood, Geoffrey (ed.) 1977 Strategies of information processing. New York: Academic Press. Umeda, Noriko 1975 "Vowel duration in American English", JAS 58: 434—445. 1977 "Consonant duration in American English", JAS 61: 846-858. van Dijk, Teun —Walter Kintsch 1977 "Cognitive psychology and discourse: Recalling and summarizing stories", in: Dressler (ed.), 61 —80. 1983 Strategies of discourse comprehension. Orlando, Florida: Academic Press. Vipond, Douglas 1980 "Micro and macroprocesses in text comprehension", JVLVB 19: 276 — 296. Vygotsky, Lev S. 1962 Thought and language. Trans. E. Hanfmann and G.Vakar. Cambridge, Mass.: M.I.T. Press.

216

Car lota S. Smith

Wales, Roger —Edward Walker (eds.) 1976 New approaches to language mechanisms. Amsterdam: North Holland. Walley, Amanda D.—Thomas D. Carrell 1983 "Onset spectra and formant transitions in the adult's and child's perception of place of articulation in stop consonants", JAS 73: 1011 — 1022. Wang, William S.-Y. 1976 "Language change", Ann NY Acad. Sei. 280: 61 - 7 2 . Wanner, Eric —Michael Maratsos 1978 "An ATN approach to comprehension", in: Halle — Bresnan — Miller (eds.), 119-161. Wanner, Eric —Lila Gleitman (eds.) 1982 Language acquisition: The state of the art. Cambridge, England: Cambridge University Press. Warden, David A. 1976 "The influence of context on children's use of identifying expressions and reference", British Journal of Psychology 67: 101 — 102. Warren, Robert E. —C. J. Obusek 1971 "Speech perception and phonemic restoration", P & P 9: 358 — 363. Watt, William C. 1970 "On two hypotheses concerning psycholinguistics", in: Hayes (ed.), 137-220. Webber, Bonnie L. 1981 "Discourse model synthesis: Preliminaries to reference", in: Joshi —Webb e r - S a g (eds.), 283-299. Webster, Ronald L. — Susan J. Schumacher —Β. B. Lubker 1968 "Changes in stuttering frequency as a function of various intervals of delayed auditory feedback", Journal of Abnormal Psychology 75: 45—49. Weist, Richard 1986 "Tense and aspect: Temporal systems in child language", in: Fletcher— Garman (eds.) (1986), 356-374. Wexler, Kenneth 1982 "A principle theory of language acquisition", in: Wanner—Gleitman (eds.), 288-317. Wexler, Kenneth — Peter Culicover 1980 Formal principles of language acquisition. Cambridge, Mass.: M.I.T Press. Wexler, Kenneth —Rita Manzini 1987 "Parameters and learnability in binding theory", in: Roeper—Williams (eds.), 4 1 - 7 6 . White, Lydia 1982 Grammatical theory and language acquisition. Dordrecht, Holland: Reidel. Winston, Patrick H. (ed.) 1965 The psychology of computer vision. New York: McGraw Hill. Wright, Patricia 1972 "Some observations about how people answer questions about sentences", JVLVB 11: 898-895. Yeni-Komshian, Grace H. —James F. Kavanaugh — Charles A. Ferguson (eds.) 1980 Child phonology. New York: Academic Press.

Lexicostatistics John A. Rea

One of the three most promising, and therewith most controversial developments in diachronic linguistics in the last half century was that of lexicostatistics, expounded and refined in a series of articles by the late Morris Swadesh. Lexicostatistic techniques compare selected vocabulary of related or presumably related languages, and have been used to postulate relationships among languages, to set up subgroupings among related languages, and to date the time of divergence of two or more related languages. It is this latter use, lexicostatistic dating, known also as glottochronology, that has received the greatest attention and led to the greatest controversy. Using lists of words chosen for their supposed universality and relative immunity to replacement by external influences, for which equivalents were sought in earlier and later forms in languages, Swadesh claimed to have established that there is a constant rate of attrition for items in the list, and derived a formula to express this — a formula slightly modified as the list itself underwent revisions. Thus in the instance of an earlier and a later stage of one language where the time separating the two is not known, the formula is claimed to permit estimation of that time with some precision. Much more important in the application is the case of two related languages, for which a modification of the formula could be used to "triangulate" the date at which they had begun to diverge. And by applying the technique to groups of related languages, one could not only determine the subgroupings of the language family in question, but could also establish dates at which their undifferentiated proto-languages were spoken. To calculate the "divergence time" of a pair of languages, the formula to be used is t = log C — log r, where t is the time depth to be

218

John A. Rea

determined (in millenia), C is the percentage of cognate items, and r is the retention rate (.805 in the earlier 200 word list, .86 in the later 100 word list). To claim to be able to pinpoint when two languages were one has been especially attractive to anthropologists, who hoped to have found in glottochronology a means of relating the linguistic "event" of divergence to historical and cultural data obtainable by other methods. This appeal is also testified to by the fact that a considerable number of articles dealing with lexicostatistics have appeared in anthropological rather than linguistic journals. If one could determine the date when two languages or dialects began to diverge, one might infer when the peoples who spoke them separated, and perhaps might further attempt to locate the original speakers geographically (see, especially, Dyen [1965]). Such a desire to relate linguistic knowledge to movements of people is, of course, not new, as witness various essays on the Urheimat of the "Indo-Europeans" and attempts to surmise facets of their culture from vocabulary data — although not using quantifiable approaches. However, there quickly began to appear strong criticisms, and equally strong defenses, of various aspects of lexicostatistics. Several, including Chretien (1962), Gleason (1965), Merwe (1966), and Sankoff (1973) have attacked, defended, or sought to modify the mathematics involved, which was first examined by Lees (1953). Others, such as Coseriu (1965) and Hoijer (1956), outlined difficulties encountered in finding ready equivalents to the supposedly universal, non-cultural vocabulary items; and nearly all practitioners have found it necessary to modify Swadesh's lists, commonly by simple omission of difficult items, as Healy (1969: 122 — 123) finds Dyen doing. That this is not a mathematically innocent, self-compensating expedient is clear, since Swadesh (1955), Gleason (1959), and others have found that items in the lists are not equally resistant to replacement. Gudschinsky (1956 a) questioned whether there could in fact be such a universal, non-cultural list; and the data in Rea (1973) show clearly that those who helped prepare the original lists in Romance had severe difficulties on that score. More recently Guiter (1984) has used a revised 100-word list that includes 15 items which do not appear in either of Swadesh's lists, and has applied a revised formula in a series of articles dealing primarily with Indo-European, but also encompassing several nonIndo-European languages including Basque.

Lexicostatistics

219

A more telling problem arose as the formulae were applied to pairs and groups of languages where the linguist could check results against known dates. Extreme cases, such as Neo-Melanesian, examined by Hall (1959), and Icelandic, examined by Bergsland and Vogt (1962), showed that whatever the average rate of vocabulary might be, individual cases could vary from this so considerably that the method seemed scarcely the "precision instrument for dating linguistic splits" Swadesh (1955) had hoped for. Although proponents argued that in some cases identifiable factors such as contact, tradition, tabu, etc., (Hymes 1960: 8 — 9) might skew the results, such variation clearly meant that the results might be suspect in instances where we lacked knowledge to factor out possible similar disturbances. Further, these undesirable results also infected the data in Romance (Coseriu 1965; Cross 1964; Rea 1958), Greek (Diebold 1964), and Germanic (Arndt 1959) — the very languages on which the formulae of lexicostatistics were based! Besides its use for determining divergence times purporting to be actual dates, lexicostatistics has found much favor among those wishing to demonstrate relationship among particular languages, and to construct Stammbäume showing linguistic subgroupings. Foremost among those using the method to such ends has been Isidore Dyen of Yale, whose meticulous classifications of Austronesian languages show again the practical aim of relating linguistic data to cultural data, and especially to migrations of peoples. Swadesh had early (1951) pointed out the utility of lexicostatistics for establishing linguistic relationships, exuberantly suggesting (1954) that with as few as one percent cognates one might establish "mesophyla", and with four or so percent "microphyla". Given that borrowing and other mutual accidents are hard to rule out in groups where diachronic phonology has not been firmly established, and given the fact that chance resemblances may fall within these ranges, as Bender (1969) has shown, such small percentages of apparent cognates have not seemed safe to many. Lexicostatistical materials have also been used to posit relationships, subgroupings, and divergence times in a number of studies of Amerindian languages, although here too there has been dissent, including Pierce (1965), Campbell (1977), and Longacre (1960), the latter being strongly critical of some of Swadesh's classifications. More recently, several studies of African language classification have made use of lexicostatistical comparisons, frequently with word lists modified to accommodate to conditions of that situation. One notes Nurse and

220

John A. Rea

Philippson (1980), as well as Ladefoged (1972). But again there has to date been no fine-meshed study based on dozens of dialects in wellestablished families such as Romance, similar to Dyen's work in Austronesian, to determine with precision how well lexicostatistical approaches to subgroupings compare with accepted classifications, although there are indications that the former may be considerably distorted (Rea 1984). Although it is likely that lexicostatistics will continue to be used by some along with other tools of diachronic linguistics, its shortcomings in those areas where verification has been possible, along with its emphasis on vocabulary (which linguists place below phonology and morphology in diachronic study), and its concern with shared retentions of such vocabulary (by contrast with traditional attention to shared innovations) have contributed to its waning interest to linguists. One can scarcely skim the surface of the considerable bibliography in a short space. Gudschinsky's (1956 b) introduction is still perhaps the best place to begin, followed by the nearly issue-length study by Hymes (1960) in Current Anthropology and the discussion accompanying it. The discussion following Bergsland and Vogt (1962) contains both criticisms and defenses, while the Dyen (1973) volume includes a variety of more recent material. But the fullest and one of the most severely critical of the studies is the monograph by Tischler (1973). The lengthy bibliography it includes, references in items cited above, and subsequent listings in standard annual linguistic bibliographies will lead to a plethora of further sources.

References Arndt, Walter W. 1959

"The performance of glottochronology in Germanic", Language 35: 110 — 192. Bender, Marvin 1969 "Chance CVC correspondences in unrelated languages", Language 45: 519-531. Bergsland, Knut — Hans Vogt 1962 "On the validity of glottochronology", Current Anthropology 3: 115 — 153. Campbell, Lyle 1977 Quichean linguistic prehistory (University of California Publications in Linguistics 81). Berkeley: University of California. Chretien, C. Douglas 1962 "The mathematical models of glottochronology", Language 32: 11—37.

Lexicostatistics

221

Coseriu, Eugenio 1965 "Critique de la glottochronologie appliquee aux langues romanes", ACILR X: 8 7 - 9 6 . Cross, Ephraim 1964 "Lexicostatistics has not yet attained the status of a science", PICL 9: 480-489. Diebold, A. Richard 1964 "A control case for glottochronology", An A 66: 987 — 1006. Dyen, Isidore 1965 A lexicostatistic classification of the Austronesian languages {IJ AL Memoir 19). Bloomington: Indiana University. Dyen, Isidore (ed.) 1973 Lexicostatistics in genetic linguistics. The Hague: Mouton. Gleason, Η. Α., Jr. 1959 "Counting and calculating for historical reconstruction", Anthropological Linguistics 2: 22 — 32. Gudschinsky, Sarah 1956 a "Three disturbing questions concerning lexicostatistics", IJAL 22: 212 — 213. 1956 b "The ABC's of lexicostatistics (glottochronology)", Word 12: 175-210. Guiter, Henri 1984 "Datation de divergences romanes", RLiR 48: 2 6 9 - 2 7 9 . Hall, Robert Α., Jr. 1959 "Neo-Melanesian and glottochronology", IJAL 25: 2 6 5 - 2 6 7 . Healy, Alan 1969 Review of Dyen 1965. Linguistics 52: 111-125. Hoijer, Harry 1956 "Lexicostatistics: A critique", Language 32: 49 — 60. Hymes, Dell H. 1960 "Lexicostatistics so far", Current Anthropology 1: 3 —44 and 340—345. Ladefoged, Peter (ed.) 1972 Language in Uganda. London: Oxford University Press. Lees, Robert B. 1953 "The basis of glottochronology", Language 29: 113 - 1 2 7 . Longacre, Robert E. 1960 Review of Mapas de clasificaciön lingüistica de Mexico y las Americas, by Morris Swadesh. Language 36: 397—410. Merwe, Nikolas J. van der 1966 "New mathematics for glottochronology", Current Anthropology 7: 485 — 500. Nurse, Derek —Gerard Philippson 1980 "The Bantu languages of East Africa: A lexicostatistical survey", in: Edgar C. Polome —C. P. Hill (eds.), 2 6 - 6 7 . Pierce, Joe E. 1965 "Hanis and Miluk: Dialects or unrelated languages", IJAL 31: 323 — 325. Polome, Edgar C . - C . P. Hill (eds.) 1980 Language in Tanzania. London: Oxford University Press.

222

John A. Rea

Rea, John A. 1958 "Concerning the validity of lexicostatistics", I J AL 24: 145-150. 1973 "The Romance data for glottochronology", in: Sebeok (ed.), 355 — 367. 1984 "Lexicostatistic skewing in northwest Sardinia". Paper presented at 1984 meeting of Linguistic Society of America, 27 — 30 December 1984, Baltimore. Sankoff, David 1973 "Mathematical developments in lexicostatistic theory", in: Sebeok (ed.), 93-113. Sebeok, Thomas A. (ed.) 1973 Current trends in linguistics, 11. The Hague: Mouton. Swadesh, Morris 1951 "Diffusional cumulation and archaic residue as historical explanations", Southwest Journal of Anthropology 7: 1—21. 1952 "Lexicostatistic dating of prehistoric ethnic contacts", Proceedings of the American Philosophical Society 96: 452—463. 1954 "Perspectives and problems of Amerindian comparative linguistics", Word 10: 306-332. 1955 "Towards greater accuracy in lexicostatistic dating", IJAL 21: 121 — 137. Tischler, Johann 1973 Glottochronologie und Lexicostatistik (Innsbrucker Beiträge zur Sprachwissenschaft 11). Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck.

Theoretical Models of Change

The Neogrammarian hypothesis Kurt R. Jankoxvsky

The advancement of empirical language studies during the first decades of the nineteenth century was coupled with, if not stringently dependent upon, the steadily increasing experience of regularity in sound transitions from preceding to succeeding language stages. The notion that "sound laws suffer no exception" began to surface by the middle of the century and was formulated as a binding operational procedure for the first time in 1876 by August Leskien. What later on came to be perceived by many contemporary linguists as an outright provocation hardly appeared as an unreasonably rigid methodological assertion in Leskien's rather casual formulation: "If exceptions are understood to mean those cases where an expected sound change did not occur due to certain recognizable causes, ... there can of course be no objection to the statement that the sound laws do not operate without exception" (Leskien 1876: XXXIV; see also pp. XXVIII and l). 1 Some of the subsequent restatements, notably those by Hermann Osthoff, Karl Brugmann, and Hermann Paul, were indeed somewhat provocative; not so much because they implied a demand for greater consistency in the search for, and the application of, laws, but because their authors combined the proclamation of the regularity principle with a fairly general condemnation of previous practices and the claim to have launched, by insisting upon a more rigorous, comprehensive methodological approach, a truly new era in scientific language analysis, vastly superior to the procedure of their contemporaries and predecessors alike. The earliest use of the term sound law ("Lautgesetz"; see Wechssler 1900: 399) in the work of Franz Bopp (see, e. g., Bopp 1825: 195) and Rasmus Rask (see, e.g., Rask 1818) indicated no more than the

224

Kurt R.

Jankowsky

discovery of a general tendency of regularity in language development. The same holds true for the findings of Jacob Grimm. The regularity phenomenon inherent in the formulation of his sound shifts ("Lautverschiebungen") of 1822 does not preclude the expected, and experienced, existence of exceptions. "The sound shift occurs in the majority of cases, but never involves every single instance" (Grimm 1822: 590). Over the years, the awareness grew among linguists that their methodology might come extremely close in exactitude to that of the natural sciences. When through the achievements of Carl Lottner (1862), Hermann Grassmann (1863), and Karl Verner (his article was published in 1877, but had been discussed with friends much earlier; see Jankowsky 1972: 117) the exceptions to Grimm's "sound law" had been reduced to a negligible residue, the resulting euphoria which enwrapped especially, but not exclusively, the Neogrammarians was a very understandable reaction. On the basis of this experience, the Neogrammarians felt justified in formulating methodological demands which took for granted what by less daring, more conventional standards would have required a great deal of further empirical research, i. e., that sound changes proceed with all-pervading regularity. Where exceptions are observed, they are believed to be due either to another sound law not yet detected or to the intervening force of analogy. No other possibility exists (see, e.g., Delbrück 1880: 60). The initially implied equation of sound laws and the physical laws of nature was apparently an overstatement. A more cautious wording would have avoided much of the heat in the ensuing unproductive controversy which did unfold and continue to grow in spite of immediate modifications. These were forthcoming in quick succession and almost spontaneously — that is, only in part as a response to criticism. They were coupled with much-needed elaborations on both the nature of sound laws and the characteristic features of analogy as well as on how the two concepts interact (see, e.g., Delbrück 1880: 101-128). The endeavor to provide comprehensive explanations of the theoretical basis and at the same time to supplement and illustrate theory by a great variety of possible applications is not exactly the coordinated effort of a homogeneous group. The Neogrammarians, even though their loose association is held together for many years by an unmistakable, strong esprit de corps, pursue their individual objectives as individual scholars. They share to a great extent many basic assumptions, but each of them places his emphasis on different priorities and

The Neogrammarian

hypothesis

225

defends controversial items rather independently against friend and adversary alike. In our definition, the Neogrammarian group, as now seems widely accepted, includes the following eight members (cf. Jankowsky 1972: 127): a) The four Indo-Europeanists: Karl Brugmann (1849 —1919) Berthold Delbrück (1842-1922) August Leskien (1840-1916) Karl Osthoff (1847-1909) b) The four Germanists: Wilhelm Braune (1850-1926) Friedrich Kluge (1856-1926) Hermann Paul (1846-1921) Eduard Sievers (1850-1932) Even several years before the appearance of his Declination im Slavisch-Litauischen und Germanischen in 1876, Leskien had begun propagating his ideas on sound law and analogy, especially in his academic lectures at the University of Leipzig (see Paul 1877: 316, 322; Masing 1883: 22). The methodological leader of the group, Hermann Paul, provides a listing of important books and articles of the earlier years of the "movement" which elaborate on and, more significantly, illustrate the crucial concepts of the Neogrammarian linguistic methodology (see Paul 1877: 321). Apart from Leskien's Declination, they include works of Braune (1876), Sievers (1876 a), Osthoff (1875, 1876), and Brugmann (1876 a, 1876 b). Sievers' Lautphysiologie of 1876, which saw five editions until 1901 and is essential as a source book to this very day, gave the Neogrammarians the much-needed solid foundation for their principal goal of exploring the true nature of the sounds and the factors involved in their historical change. Two years later, in the introduction to the first volume of their Morphologische Untersuchungen, Osthoff and Brugmann furnished a comprehensive statement of the general objectives which the group intended to pursue. This 18-page "manifesto" is by far the most provocative of the Neogrammarian publications (see Osthoff-Brugmann 1878: I I I - X X ) . Osthoffs monograph Das physiologische und psychologische Moment in der sprachlichen Formenbildung (1879) broadened the theoretical basis by providing a thorough investigation of the psychological component in sound formation as well as sound transition and of its interrelation with the physical aspect of language.

226

Kurt R.

Jankowsky

The movement entered into its climactic phase with the publication of Hermann Paul's Prinzipien der Sprachgeschichte (1880) which was admiringly labeled by a contemporary fellow-linguist as "epochmaking in regard to the study of every aspect of language life it touched" (Karsten 1894: 314). Such an evaluation, substantively endorsed even by many linguists of that time from outside the Neogrammarian fold, has hardly lost its truth value to this very day. Paul's Prinzipienlehre, defined by him as "the most suitable means for us to gain insight into the causal interconnection of events" (Paul 1920 b: 15), is employed in his Prinzipien der Sprachwissenschaft to exemplify in one particular discipline the resulting scientific approach "which mediates between the Gesetzeswissenschaften and the Geschichtswissenschaften" (Paul 1920 b: 15). Language as a product of human culture requires for Paul a historical treatment (Paul 1920 a: 1). Detecting the conditions of its development constitutes "the basis for the methodology to be pursued in the investigation of each factor" (1920 a: 3). The factors involved are both physical and psychological: The psychological factor is the most essential factor in all cultural activities ... But the psychological aspect is nevertheless not the only factor; there is no culture on a purely psychological foundation ... Whenever we enter the field of historical development, we have to deal, apart from psychological, with physical forces (1920 a: 6).

How do these two sets of forces combine in the life of language? "It is a fact of fundamental importance which we must never lose sight of that all purely psychological interrelation ["Wechselwirkung"] occurs only within the individual. All interaction among individuals is brought about only indirectly, mediated by physical means" (1920 a: 12). Accordingly, Paul considers it "as one main task for the doctrine of principles of the science of culture ["Prinzipienlehre der Kulturwissenschaft"] to state the general conditions under which the psychological and physical factors, pursuant to their own specific laws, combine to achieve a common purpose" (1920 a: 7). Paul's views on the historical as the only truly scientific approach to language analysis (see Paul 1920 a: 20), on the need to search for exceptionless laws ("Science is served only by cogent law, not by arbitrariness" [1877: 318]), on the supplementary nature of the physical and the psychological aspects of language, on the individual as the originator ("Actual language does exist only in the individual" [1877: 325]) and the initiator of all language change ("The changes in language occur in [the speech of] the individual" [1920 a: 34]), and lastly on the

The Neogrammarian

hypothesis

227

need to penetrate to the causes of all language-related phenomena (see Paul 1920 b: 13) are paralleled by some (especially by Brugmann and Osthoff) and largely shared by all of the Neogrammarians. The most consistent, and probably also the most moderate and balanced, discussion of the Neogrammarian tenets by a group member is provided by Berthold Delbrück, the foremost linguistic historiographer among them, in a variety of publications, including Einleitung in das Sprachstudium (1880), Grundfragen der Sprachforschung (1901), and "Das Wesen der Lautgesetze" (1902). The programmatic manifesto of 1878 (see Osthoff — Brugmann 1878: III —XX) is summarized by Franz Misteli in an extensive review of the first volume of the Morphologische Untersuchungen (Misteli 1880). The main points of this summary are the following: It is four demands that this recent movement [ = the Neogrammarians] places before linguistics: 1) The sound laws have to be regarded as operating without exceptions. There are no two [types of] sound changes, the one according to laws and the other sporadic, the one regular and the other irregular ... 2) What does not fit the forms dictated by the sound laws, has to be taken and explained as analogy formation ... 3) More recent languages ... have to be accorded greater importance for the study of older language stages than they have had before, especially if one is convinced that analogy was as much at work in the latter as it is in the former ... Demand 1) presupposes and requires a much greater precision in the treatment of phonetic questions than was customary so far; the more recent languages offer the most valuable material for it. 4) At the same time, far greater restrictions are [to be] placed upon the reduction of the forms of the individual languages to IE proto-forms; those forms should, instead, first be examined within their specific language groups in accordance with the first two demands, and only the result thus obtained should be utilized for [the investigation of] IE language conditions (Misteli 1880: 366).

It is obvious from the above that sound law and analogy formation are the dominating factors in the four demands. What had been a clouded issue in the discussion at the time, and may not be entirely settled even today, is the question as to what extent the two concepts and the methodology underlying their application can be called an innovation that was single-handedly achieved by the Neogrammarians. Naturally, we must not expect to find a plausible answer in the manifesto of 1878. On the other hand, there is good reason to take the following words of Brugmann, written seven years later, at their face value and interpret them, since they have not been contradicted, as characterizing also the position of his friends: "I for one have always considered these newer views as the organic and consistent develop-

228

Kurt R. Jankowsky

ment of the older endeavors, and this opinion has grown on me from year to year" (Brugmann 1885: 125). In 1887, G. I. Ascoli cited the above passage and remarked: "After this statement ... it should by no means be unreasonable to conclude that from now on any further discussion would be superfluous" (Ascoli 1887: 104). He is even more specific in another instance within the same context: "When checking into the entire controversy thoroughly and comprehensively without bias it seems indeed almost impossible that it could ever exist" (1887: 171). Gustaf Ε. Karsten, a few years after Ascoli, added a new dimension to the evaluation of the dispute by focusing on the larger framework and recognizing the achievements of all parties concerned: In a general way it was a movement from casualty to causality. To say that any one man started this movement a few years before or during the decade from 1870 to 1880 would be unfair. There is an uninterrupted course of development from Schleicher to Brugmann, from Whitney to Paul. On the other hand, however, it would be equally unfair not to grant that during the time mentioned a vague impression, a sub-conscious motive, evolved into the clear statement and the strict observance of a principle (Karsten 1894: 312).

The consequences of "the clear statement and the strict observance" of the principle in question become most apparent when tracing back the main stages in the development of the two concepts involved. Of greatest importance in this "line of heritage" is, first of all, Rudolf von Raumer. In contrast to Jacob Grimm, he insisted on a strict separation of sound and letter and comprehensive study of sound physiology. Directing the attention of linguists away from nebulous concepts like "Sprachgeist" and towards the individual as the only one responsible for bringing about all changes in language (see von Raumer 1858: 374 and Jankowsky 1972: 88-91), he set in motion tendencies that eventually led to the thorough investigation of both mental processes occurring in the speech production and the phonetic properties of the sound inventory (see Esper 1973: 23 and Benware 1974: 79 — 80). The Neogrammarians have repeatedly acknowledged von Raumer's important contribution for the development of their own position (see, e.g., Delbrück 1919: 114-115 and Paul 1901: 119). Rudolf von Raumer is aware of analogy as a force in the process of sound change, but he does not see any reason for treating it extensively (see von Raumer 1858: 376). Hence, he does not see yet, as his followers do, the complementary status of sound law and analogy. His comments, however, on Georg Curtius' differentiation of

The Neogrammarian

hypothesis

229

regular and irregular changes do point in this general direction (see von Raumer 1861: 435, note). For August Friedrich Pott the term "Lautgesetz" had already become a household word. But still he is far from positing that it operates without exceptions. He established that, even in the case of the "mere letter", language is not ruled "by the lawlessness of unrestrained arbitrariness, but by reasonable freedom, i. e., restriction via specific laws, founded in the nature of the sounds" (see Pott, quoted in Jankowsky 1972: 87). He employed analogy fairly often, and indeed somewhat complementary to sound law. However, analogy formations were interpreted by him as misformations and used to explain anomalous phenomena, i.e., forms which deviate from the anticipated regular development (see Delbrück 1919: 182). While Georg Curtius achieved significant advances concerning the regularity principle, his attitude towards analogy, even though he considered it a "concept of fundamental importance" (Curtius 1886: 52), was ambiguous at best. Delbrück's assessment is harsh, though to the point: "Curtius did not come up with a comprehensive treatment on ... analogy; instead, he did, as other linguists have done, occasionally make use of analogy as a principle of explanation" (Delbrück 1880: 104). In spite of his contention that "in the life of sounds, firm laws can be identified with the greatest certainty, laws which assert themselves with the consequence of forces of nature" (Curtius 1866: 80), he differentiated between regular or constant and irregular or sporadic changes (see Curtius 1866 passim), a viewpoint which he never modified. "We will be able to find laws and rules, even if not quite without exceptions and deviations" (1866: 79) seems to have been his ever-recurring observation. On the other hand, the extent of his progress as against his predecessors is duly acknowledged, for instance by Delbrück: "One must consider ... that his endeavor was very particularly directed towards identifying a stricter order in the world of sounds" (Delbrück 1880: 101). When Curtius refers to analogy formations as "in principle possible everywhere, but nowhere occurring of necessity" (Curtius 1885: 39), he came close to the Neogrammarian position (see Osthoff 1879: 22). By his discussion of analogy, however, in terms of a "secondary force ... the misinformation and aberration as against the healthy forma-

230

Kurt R. Jankowsky

tion" (Curtius 1885: 44), he provoked strong criticism on the part of the Neogrammarians (see Brugmann 1885: 78). According to the contention of numerous contemporaries of the Neogrammarians, August Schleicher professed to, and practiced, an attitude towards the sound law and analogy principles which was in no way different from that of the Neogrammarians. Thus, Johannes Schmidt, staunch supporter as well as objective critic of Schleicher,2 repeatedly asserted that "it was Schleicher who was the first to teach that all changes which the IE words have undergone from the oldest times to the present day were caused by two factors, by sound laws operating without exceptions and by interfering ... analogies" (Schmidt 1887: 304-305). That Schleicher placed emphasis on the exceptionless operation of sound laws and on the importance of analogy was specifically recognized by the Neogrammarians, for example by Brugmann (1885: 31) and Delbrück ("... Schleicher established a large number of carefully considered and well-founded sound laws, designed to serve as a guideline for every linguist, he has undoubtedly brought about an extraordinary achievement by his activity of probing and classifying" [1919: 101]). Nevertheless, the Neogrammarians, in spite of Schmidt's contention to the contrary, did not find in the work of Schleicher conclusive evidence "that he admitted of no sound laws other than those operating without exception" (Delbrück 1919: 102). Brugmann, furthermore, aligned Schleicher with Curtius as linguists who "proceeded in many respects from assumptions that were incompatible with the very nature of language history" (Brugmann 1885: 31). The manifesto statement of the Morphologische Untersuchungen (Osthoff — Brugmann 1878) contains several passages (e.g., 1878: IX) which are directed against Schleicher, even though his name is not mentioned (see Schmidt 1887: 303 — 304). Brugmann also takes issue with Schleicher as the main representative of the view that "the languages organized and gained perfection in pre-historical time to degenerate and decay in historical time" (Brugmann 1885: 32), with direct bearing on Schleicher's pronouncement: "The history of language consists ... of two totally different parts: 1) The history of the development of language, prehistorical period; 2) the history of the decay of language, historical period" (Schleicher 1850: 13). The Neogrammarians felt in many respects much closer to Wilhelm Scherer because of his "more consistent adherence to the sound laws" (see Paul 1877: 321) and "the more extended use of the so-called false

The Neogrammarian

hypothesis

231

analogy for the explanation of forms" (Paul 1879: 308). Scherer, although severely criticized for his excessive speculations and "most glaring contradictions" (Paul 1879: 311), is amply praised on account of his "achievement to have provided an effective stimulus for tackling the question of how transformations and innovations in a language are brought about" (Osthoff — Brugmann 1878: XI). In matters of sound law and analogy, Scherer is claimed by the Neogrammarians as a trend-setting model. Especially in the case of Leskien, "the seed spread by Scherer fell on fertile ground" (1878: XII). For Leskien's methodological principles of 1876, arrived at "by reflecting on the concept of 'sound law' and 'exception to the law' more thoroughly than has ever been done before (1878: XII), statements by Scherer like the following, which Leskien undoubtedly knew, must have had that "seeding effect" to which Osthoff and Brugmann refer: "Sound changes ... proceed according to firm laws that do not suffer any interference other than by other laws" (Scherer 1875: 107). Scherer, whom Erwin Esper rightly calls "the man who may be considered the last German link in the chain leading from Grimm to the neogrammarians" (Esper 1973: 22), must also be credited with having removed "a stumbling-block", namely "Schleicher's notion that linguistic change — 'decay' — had been confined to the historical period" (1973: 22). The well-known passage in Scherer to which Esper implicitly refers has become a cornerstone in the theoretical framework of the Neogrammarians: "One can hardly close one's eyes any longer to the insight that the differentiation between development and decay or ... between nature and history of language is an error. I for my part have encountered everywhere only development, only history" (Scherer 1868: X). There cannot be any doubt that William D. Whitney played a very important, if not a decisive role, not only in the formation of the general attitude of the Neogrammarians toward language ("In a certain sense, we all belong to his pupils" [Delbrück 1897: 84 — 85]), but also, more specifically, with regard to the sound law and analogy concepts and their close interrelation (see, e.g., Paul 1901: 126). At the time when the Neogrammarians were about to launch their "movement", the two major works of Whitney became available in German, i. e., Language and the study of language (Whitney 1867), translated by Julius Jolly in 1874, and The life and growth of language (Whitney 1875), translated in 1876 by Leskien. That these translations amounted to an endorsement of what Whitney had to say about language is an

232

Kurt R. Jankowsky

obvious fact, corroborated to boot by the numerous references of individual Neogrammarians to Whitney's writings. To comment on them or even to merely list them would fill some pages. Perhaps the most revealing reference is a passage written by Brugmann in 1885. He cited Whitney's evaluation of the work done by German linguists. As to detailed research, Brugmann wrote (paraphrasing Whitney): "We are far ahead of all other nations, and we do have indisputably the leadership, it is true, in all matters regarding this aspect" (Brugmann 1885: 39). But on the other hand — and now I quote directly from Whitney: "There is among them [i.e., the German linguists] ... such discordance on points of fundamental importance, such uncertainty of view, such carelessness of consistency, that a German science of language cannot be said to have an existence" (Whitney 1875: 318 — 319). Brugmann accepted this attack as being directed against those pre-1875 German linguists to whom he, too, found himself in opposition, and derived from it a challenge for post-1875 German linguists, most particularly for him and his Neogrammarian associates, asserting that Whitney's demand "to bring about an agreement in the questions of principles so that an organic coordination ... of all works of detail ... may become possible, has already been met today to no small extent" (Brugmann 1885: 3 9 - 4 0 ) . Brugmann's claim (filed for himself and for his group) — that a substantial advancement was achieved for linguistic theory during the decade following the publication of Whitney's Life and growth of language — focused undoubtedly for the main part on the sound law and analogy principles. The ingredients were supplied, it is true, by a great many distinguished linguists of the immediate past, but it was the Neogrammarians who fitted the parts together and consolidated them into a complex framework which has served ever since, with only minor modifications over the years, as a widely applied and reliable working hypothesis for the historical linguist. The resulting expansion and refinement involved, above all, the structured interrelation of the two concepts: "There is most frequently a causal relationship between the emergence and incorporation of analogy formations and the phonetic change" (Brugmann 1885: 81). Analogy undergoes a thorough reappraisal: "Analogy is today no longer the utterly destructive driving force with no purpose other than to interfere with the straight course of the phonetic development of languages, b u t . . . an eminently important factor in the process of both learning and speaking the language". Accordingly, a new purpose is

The Neogrammarian

hypothesis

233

discovered in the operation of analogy, namely, "on the one hand, to produce, in reaction to the form-destructing operation of the sound laws, a firmer association among what belongs together and, on the other hand, to neutralize the phonetic merger of forms with distinctive functions" (Brugmann 1885: 8 5 - 8 6 ) . Paul's formulation is equally forceful: "Each disorganization is followed by a reorganization. The stronger the force of the attack by the sound laws against the groups, the greater the activity of new creation [by analogy]" (Paul 1920 a: 198). Similar passages are found in the writings of other Neogrammarians (see, e. g., Osthoff 1879:1; Delbrück 1902: 293). It is important to note that the idea of close sound law-analogy interrelation was clearly pronounced already in Leskien's 1876 publication (1876: 2). While Leskien proved for the first time conclusively the complementarily important status of analogy as a counterpart to the sound law concept, a detailed investigation of how analogy functions was supplied by other Neogrammarians, for instance by Brugmann (see Brugmann 1876a: 317 — 320) and especially by Paul (e.g., Paul 1877: 3 2 8 - 3 3 1 and, most importantly, Paul 1920a: 106-120). Very shortly after these Neogrammarian treatments of the subject, the first responses by other linguists appeared. From the many publications that could be cited I refer to only a few which I consider most representative of both the thoroughness and the generally favorable attitude found in the majority of the studies. Masing (1883) mainly gives a restatement of what by then must have been known as "the Neogrammarian position". Wheeler (1887) deals exhaustively with this topic in a booklet that has remained extremely valuable reading material to this very day. And, lastly, Karsten (1894), like Wheeler, though on a less ambitious scale, ventures probing evaluations and interpretations, along with constructive criticism, on the basis of accurate presentations of pertinent Neogrammarian pronouncements. All three authors, like numerous others not mentioned here, have also treated extensively the relationship of analogy to sound law. Paul's "proportional groups", which he uses as a model to describe the operation of analogy, have emerged as the most appropriate device for dealing with the phenomena involved. That further refinement was needed and not all instances of change caused by analogy could be explained via the proportional equation (Paul 1920 a: 117: "animus : animi = mensa : x") did not go unnoticed even at the time of the first reactions to the Neogrammarian propositions and has continued to

234

Kurt R.

Jankowsky

remain a favorite topic of linguistic discussion to this very day (see, e.g., Esper 1973: 40). It is not surprising at all that modern linguists utilized Paul's overall treatment of analogy (see especially Paul 1920 a: 106 — 120) for a comparison with Chomsky's competence and performance terminology (Best 1973: 27 — 30). Another inveterate subject matter, then and now, is the question as to what proof can be established for the correctness of the regularity principle. Is an inductive or at least a deductive substantiation possible? The first to address this problem comprehensively was Osthoff. He realizes the impossibility of establishing "that all sound laws in languages detected so far ... are observed by us in their exceptionless operation" (Osthoff 1879: 6) and, hence, has to replace provability by probability. Delbrück argues along similar lines: "Inductively the exceptionless operation of the sound laws is not provable" (Delbrück 1880: 115), adding the plausible comment: "Not the empirical sound laws, but the sound laws per se operate without exception" (Delbrück 1902: 294). The deductive proof advanced by Osthoff as well as by Delbrück is unconvincing at best. In spite of Masing's reasoning for the regularity principle being deductively "provable and proven" (Masing 1883: 34), I side with Karsten (1894: 314), who contends that "neither the inductive nor the deductive method yielded conclusive arguments". The entire question, so it seems to me, boils down to being a moot point, since as a working hypothesis the regularity principle has been in use for over a century now and is relied upon by Neogrammarians and their followers as well as by their opponents. Where it is employed consistently, the results have been not only extremely useful, but also crucially important for the investigations of the historical linguist at large. Its value as a working hypothesis is fully recognized by Johannes Schmidt, who very often took issue with the theoretical position of the Neogrammarians, but of whom Wheeler correctly said that in "his worthy practice, we find no essential point of difference from that of the neogrammarians" (Wheeler 1887: 42). Hugo Schuchardt, the most formidable critic of the "new movement", likewise directs his pungent attacks (see Schuchardt 1885) against the formulation of the regularity principle only. "Otherwise, however, I am not aware of any principal opposition to the Neogrammarians" (Schuchardt in 1886, quoted in Wheeler 1887: 42). A final word on how the Neogrammarians fare generally in the linguistic discussion today. There is no doubt that things have changed significantly during the last ten or fifteen years. I would hope that this

The Neogrammarian

hypothesis

235

change is the result of a far-reaching change in reading habits: Neogrammarian writings are read again. Hopefully, Piero Meriggi's complaint of 1966 is unnecessary today: "These precipitous critiques are generally so completely unfounded that I feel compelled to draw the conclusion: whoever writes about the Neogrammarians in this way, has never read them" (Meriggi 1966: 3). One of the benefits of a more extensive reading of the Neogrammarians and about them will dissolve without fail the widely held opinion, first invented by Schuchardt (1885: 1 — 3), that "Neogrammarianism" is equatable with the sound law and analogy principles. The realization that this opinion is false — that, instead, the scope of Neogrammarian thought and practice by far transcends the limits of sound law and analogy — will be a decisive step in the direction of a better understanding of what role the two concepts play in the life and growth of language. Recent publications on the subject elaborate with increasing frequency on the relevancy of the Neogrammarian work for current linguistic studies. "It is indicative of the quality of work then achieved that already at their time the investigation had advanced to asking questions which to this very day have remained valid ... One even has to recognize that the answers which have been attempted since then often were pursued along avenues first explored by the Neogrammarians". These words by Karl-Heinz Best (1973: 44), including the pertinent substantiation, can easily be paralleled almost ad libitum by quoting from existing literature of the last decade. Paul, more than any other Neogrammarians, figures prominently in these accounts. "His points of departure in many instances constitute anticipations of the linguistic issues of the twentieth century" (Cherubim 1973: 312). That Konrad Koerner can furnish us with ample and conclusive proof to support his "claim that Paul's Prinzipien had ... a profound influence on the development of Saussure's linguistic thought" (Koerner 1972: 301) fits the general picture. Only a few years ago two important monographs appeared, one celebrating the one hundredth anniversary of Pauls and Braunes Beiträge (see Fromm — Ganz — Reis 1978), the other commemorating a century-long Neogrammarian influence on the shaping of linguistic thought (see Transactions 1978). Both volumes contain an abundance of fruitful discussions on many aspects — including the sound law and analogy principles — of Neogrammarian theory and practice, and must convince even the skeptic that no labor is lost when a renewed effort is made to better understand the linguistic currents of today by

236

Kurt R. Jankowsky

delving into that important portion of our linguistic past which has been indelibly marked by the immensely influential activity of the Neogrammarians. "No reputable historical linguist has brought the charge that the 'working hypothesis' of the regularity of sound laws does not work " (Pulgram 1955: 61). We can continue to work on its refinement, but it is undoubtedly here to stay.

Notes 1. Translations from German to English, here and elsewhere in the paper, are mine. 2. See, e. g., Schleicher's "family tree theory" and Schmidt's critical evaluation of it (Schmidt 1872) which led to his (supplementary) counterconcept of the "wave theory".

References Ascoli, Graziadio Isaia 1887 Sprachwissenschaftliche Briefe. Trans. Bruno Güterbock. Leipzig: S. Hirzel. Benware, Wilbur A. 1974 The study of Indo-European vocalism in the 19 th century. Amsterdam: John Benjamins. Best, Karl-Heinz 1973 Probleme der Analogieforschung. Munich: Max Hueber. Bopp, Franz 1825 "Vergleichende Zergliederung des Sanskrits und der mit ihm verwandten Sprachen II (Reflexiv)", Abhandlungen der Berliner Akademie, phil.-hist. Klasse (Reprinted in: Kleine Schriften zur vergleichenden Sprachwissenschaft. Leipzig: Zentralantiquariat DDR, 1972: 1 — 32). Braune, Wilhelm 1876 "Über die Quantität der ahd. Endsilben", Beiträge zur Geschichte der deutschen Sprache und Literatur 2: 125 — 167. Brugmann, Karl 1876 a "Nasalis sonans in der idg. Grundsprache", Studien zur griechischen und lateinischen Grammatik 9: 285 — 338. 1876 b "Zur Geschichte der stammabstufenden Declination", Studien zur griechischen und lateinischen Grammatik 9: 363 — 406. 1885 Zum heutigen Stand der Sprachwissenschaft. Strassburg: K. J. Trübner. 1900 "Zu dem 'Vorwort' zu Band 1 der Morphologischen Untersuchungen von Osthoff und Brugmann", Indogermanische Forschungen 11: 131 — 132.

The Neogrammarian hypothesis

237

Cherubim, Dieter 1973 "Hermann Paul und die moderne Linguistik. Zur Studienausgabe von H.Pauls Prinzipien der Sprachgeschichte", Zeitschrift für Dialektologie und Linguistik 40: 310 — 322. Christie, William, Jr. (ed.) 1976 Current progress in historical linguistics. Amsterdam: North-Holland. Curtius, Georg 1866 Grundzüge der griechischen Etymologie. Leipzig: B. G. Teubner. 1885 Zur Kritik der neuesten Sprachforschung. Leipzig: S. Hirzel. 1886 "Bemerkungen über die Tragweite der Lautgesetze, insbesondere im Griechischen und Lateinischen", in: Windisch (ed.), 50 — 94. (First published 1870.) Delbrück, Berthold 1880 Einleitung in das Sprachstudium. Ein Beitrag zu Geschichte und Methodik der vergleichenden Sprachforschung. Leipzig: Breitkopf und Härtel. 1897 "Letters from foreign scholars concerning Professor Whitney", Journal of the American Oriental Society 19: 83 — 85. 1901 Grundfragen der Sprachforschung. Strassburg: Κ. J. Trübner. 1902 "Das Wesen der Lautgesetze", Annalen der Naturphilosophie 1: 277 — 308. Esper, Erwin A. 1973 Analogy and association in linguistics and psychology. Athens: Univ. of Georgia Press. Fromm, Hans —Peter Ganz—Marga Reis (eds.) 1978 Beiträge zur Geschichte der deutschen Sprache und Literatur 100. Grassmann, Hermann 1863 "Über die Aspiration und ihr gleichzeitiges Vorhandensein im Auslaute der Wurzeln", Zeitschrift fur vergleichende Sprachforschung 12: 81 — 138. Grimm, Jakob 1822 Deutsche Grammatik. Vol. 1. (2nd edition). Göttingen: Diederichsche Buchhandlung. (First published 1819; quoted from 1822 ed.) Jankowsky, Kurt R. 1972 The Neogrammarians. A re-evaluation of their place in the development of linguistic science. The Hague: Mouton. 1976 "The psychological component in the work of the early Neogrammarians and its foundations", in: Christie (ed.), 267 — 284. 1979 "Typological studies in the 19th century and the Neogrammarian sound law principle", Forum Linguisticum 4: 159 — 173. 1981 Review of Dieter Cherubim, Grammatische Kategorien. Das Verhältnis von 'traditioneller' und 'moderner' Sprachwissenschaft (Tübingen: Max Niemeyer, 1975), Zeitschrift für Dialektologie und Linguistik 48: 68 — 69. 1983 Review article on Beiträge zur Geschichte der deutschen Sprache und Literatur. Vol. 100 (1978), Forum Linguisticum 7: 2 3 9 - 2 5 9 . Karsten, Gustaf E. 1894 "The psychological basis of phonetic law and analogy", Publications of the Modern Language Association of America 9: 312 — 341. Koerner, E. F. Konrad 1972 "Hermann Paul and synchronic linguistics", Lingua 29: 274—307.

238

Kurt R. Jankowsky

Leskien, August 1876 Die Declination im Slavisch-Litauischen und Germanischen. Leipzig: S. Hirzel. Lottner, Carl 1862 "Ausnahmen der ersten Lautverschiebung", Zeitschrift für vergleichende Sprachforschung 11: 161—205. Masing, Ferdinand 1883 Lautgesetz und Analogie in der Methode der vergleichenden Sprachwissenschaft. St. Petersburg: Kaiserliche Akademie der Wissenschaften. Meriggi, Piero 1966 "Die Junggrammatiker und die heutige Sprachwissenschaft", Die Sprache 12: 1 - 1 5 . Misteli, Franz 1880 "Lautgesetz und Analogie. Methodologisch-psychologische Abhandlung", (On the occasion of Osthoff — Brugmann, Morphologische Untersuchungen, vol. 1 [1878]). Zeitschrift für Völkerpsychologien·. 3 6 5 - 4 7 5 and 1 2 : 1 - 2 7 . Osthoff, Hermann 1875 Forschungen im Gebiete der idg. nominalen Stammbildung. Jena: H. Costenoble. 1876 "Zur Frage des Ursprungs der germanischen n-Deklination". Beiträge zur Geschichte der deutschen Sprache und Literatur 3: 1—89. 1879 Das physiologische und psychologische Moment in der sprachlichen Formenbildung. Berlin: C. Habel. Osthoff, Hermann —Karl Brugmann 1878 Morphologische Untersuchungen auf dem Gebiete der idg. Sprachen. Vol. 1. Leipzig: S. Hirzel. Paul, Hermann 1877 "Die Vocale der Flexions- und Ableitungssilben in den ältesten germanischen Dialecten", Beiträge zur Geschichte der deutschen Sprache und Literatur 4: 315-475. 1879 Untersuchungen über den germanischen Vocalismus. Halle: Max Niemeyer. 1901 Grundriss der germanischen Philologie. Vol. 1. (2nd edition). Strassburg: Karl J. Trübner. (First published 1891.) 1920 a Prinzipien der Sprachgeschichte. (5th edition). Halle: Max Niemeyer. (First published 1880.) 1920 b Aufgabe und Methode der Geschichtswissenschaften. Berlin: Walter de Gruyter. Pulgram, Ernst 1955 "Neogrammarians and soundlaws", Orbis 4: 61 —65. Rask, Rasmus 1818 Undersegelse om det gamle Nordiske eller Islandske Sprogs Oprindelse. Köbenhavn: Gyldendal. Raumer, Rudolf von 1837 Die Aspiration und die Lautverschiebung. Leipzig: F. A. Brockhaus. 1858 "Die sprachgeschichtliche Umwandlung und die naturgeschichtliche Bestimmung der Laute", Zeitschrift für die österreichischen Gymnasien 9: 353-373. (Reprinted in Raumer 1863: 368-393.)

The Neogrammarian

hypothesis

239

1861

"Die geschichtliche Entwicklung der Laute", Zeitschrift für die österreichischen Gymnasien 12: 2 6 7 - 2 9 8 . (Reprinted in Raumer 1863: 4 0 5 - 4 4 3 . ) 1863 Gesammelte sprachwissenschaftliche Schriften. Frankfurt & Erlangen: Heyder & Zimmer. Scherer, Wilhelm 1868 Zur Geschichte der deutschen Sprache. Berlin: Weidmann. 1875 Review of Sprachwissenschaft, trans. Julius Jolly (München: Th. Ackermann, 1874 [Translation of Whitney 1867]), Preussische Jahrbücher 35: 106-111. Schleicher, August 1850 Die Sprachen Europas in systematischer Ubersicht. Bonn: Η. B. König. Schmidt, Johannes 1869 "Nachruf August Schleicher", Zeitschrift für vergleichende Sprachwissenschaft 18: 3 1 5 - 3 2 0 . 1872 Die Verwandtschaftsverhältnisse der idg. Sprachen. Weimar: Hermann Böhlau. 1887 "Schleichers Auffassung der Lautgesetze", Zeitschrift für vergleichende Sprachforschung 28: 303 — 312. Schuchardt, Hugo 1885 Über die Lautgesetze. Gegen die Junggrammatiker. Berlin: Robert Oppenheim. Sievers, Eduard 1876 a "Kleine Beiträge zur deutschen Grammatik III. Die starke Adjectivdeclination", Beiträge zur Geschichte der deutschen Sprache und Literatur 2: 98-124. 1876 b Grundzüge der Lautphysiologie zur Einführung in das Studium der Lautlehre der idg. Sprachen, n. p. (2nd edition 1881 as Grundzüge der Phonetik zur Einführung in das Studium der Lautlehre der idg. Sprachen. Leipzig: Breitkopf & Härtel.) Verner, Karl 1877 "Eine Ausnahme der ersten Lautverschiebung", Zeitschrift für vergleichende Sprachforschung 23: 97 — 130. Wechssler, Eduard 1900 "Gibt es Lautgesetze?", in: Festgabe für Hermann Suchier. Halle: Niemeyer, 3 4 9 - 3 5 8 . Wheeler, Benjamin Ide 1887 Analogy and the scope of its application in language. Ithaca, Ν. Y.: Cornell Univ. Press. Whitney, William D. 1867 Language and the study of language. New York: Scribner's. 1875 The life and growth of language. London: Henry S. King & Co. Wilbur, Terence H. (ed.) 1977 The Lautgesetz-controversy: A documentation ((1885 — 86). Amsterdam: John Benjamins. Windisch, Ernst (ed.) 1886 Georg Curtius. Kleine Schriften. Leipzig: S. Hirzel. *** Transactions of the Philological Society. Commemorative Volume: The Neogrammarians. Oxford: Basil Blackwell.

A structural view of sound-change Archibald A. Hill

A generation ago I wrote an article called "Phonetic and phonemic change", that had at least the virtue of being a relatively early attempt at structural description of sound-change. The article seems now somewhat naive, and it is blind to such areas as sentence phonemics. It also assumes not only the validity of the phoneme, but also the reality of phonological drift over a period of time, both of which are matters that have since been often challenged. Ten years later, in 1946, Henry Hoenigswald restated some of the essential points in a far more elegant and formal fashion, yet my first attempt still seems to me relevant to current linguistic theorizing. Thus Martin Joos in his collection of 1957 remarked of my essay that though I used a transcription system which I would later have abandoned, at least few of my points would have been lost (Joos 1957: 84). I can most easily describe the matters which seem to me still relevant as a series of conflicting formulations. They are, in order of importance, as follows: First, is the concept of the phoneme still valid, or is it now invalid? Is it now still necessary, or is it unnecessary? Second, are sound-changes regular, or do they admit of sporadic and random exceptions? Third, do sound-changes come about by shifts in the phonemes (and their components) or do they occur word by word? Fourth, if, as it apparently does, the data presents evidence for both lexical and phonemic diffusion of change, is this final conflict such as to force abandonment of one view or the other about the nature of sound-change, or are the views capable of being reconciled? I shall try to present a reasonably brief view of the structuralist position on all these points.

242

Archibald

Α.

Hill

Far and away the most important of the points of difference that I have mentioned is the question of the validity of the phoneme. In the 1930s it was impossible to foresee that the currently dominant transformationalist-generative school of linguists would often say that the phoneme was neither necessary, motivated, nor valid. It is not my purpose to discuss the reasons for these attacks on the phoneme; instead, I shall merely outline my reasons for still making use of it. They are, first of all, reasons of economy in historical statement. Thus the phoneme functions much as do the "sound types" in the Neogrammarian formulations of Rask and Grimm, such as that Latin ρ t k correspond to Germanik / θ χ. The cases where there is an apparent exception to such regularity are usually explainable in terms of variation in the environment, as when difference in word-stress accounts for the appearance of an unexpected medial consonant in English father in contrast with Latin pater, the difference being explained by the so-called Verner's law. This convenience of statement, furthermore, avoids the complications of morphophonemics, as when the variant developments of the first vowels of nation and national are accounted for by setting up a special morphophonemic entity which varies between the vowels of hate and hat, though the vowel of nation is identical with that of the first syllables of native and nativist, which does not vary as does the vowel of nation. But more important than this simplicity of historical statement is consideration of the function of the phoneme in the whole surface structure of language. One of Bloomfield's influential statements about language was that language items are all divided into "sames" and "differents". The "differents" serve as identifiers of forms, cutting continua into discrete entities. The "sames" do not thus identify and segment. Thus a possible definition of the phoneme is that it is a minimal differentiator of forms larger than itself, and in which it is contained. A greatly simplified statement of the position of the phoneme in the complete analysis of the surface structure of language is that phonemes identify words and word elements as same or different; words and word elements identify sentences as same or different; sentences, finally, identify utterances and discourses as same or different. There is, however, a difficulty in saying that different words identify different sentences. If two words are phonologically identical but different in meaning, then it is their meanings which identify sentences;

A structural view of sound-change

243

and the result would seem to be circular, in that meaning gives difference and difference gives meaning. But as I said in 1970, the meaning of such words, indeed of all words, is conditioned by the discourse context. The identification of word-meaning is then not circular, and closely resembles the analysis of identifiers at all levels. If we can, as I think we must, accept the presence of identifiers at all levels, it further seems to me necessary to describe them in nature and function at all levels of surface structure, before final semantic, grammatical, and syntactic description of deep structure can be completed. One reason for my opinion here is acceptance of the adage ascribed to St. Thomas Aquinas — there is nothing in intellect which was not first in sense. Another reason for defense of the phoneme is the same as that used by Sturtevant in 1947 for defense of the doctrine of regularity in sound-change. I can employ his defense of regularity, since, as I have said above, regular sound-changes are most easily and simply stated in terms of phonemics. Sturtevant pointed out that phonemic entities, the items used in historical linguistics, have led to the enormous volume, exactitude, and consistency of comparative historical linguistics. If the results of comparative linguistics are valid, they justify the phoneme. If these results can be shown to be invalid, the phoneme collapses. Needless to say, I do not think it will. There are two other points which should be mentioned in connection with phonemes. One is that there is some empirical evidence for their reality in the occurrence of spoonerisms, where phonemes are shifted about from slot to slot, very much as if a child were shifting alphabet blocks. And finally, I should mention that I find Sanford Schane's generally excellent article of 1971 on the phoneme largely irrelevant to my purpose. Schane is assigning the phoneme a modest place in transformational-generative analysis of surface structure. It is the modesty of the place he assigns it that is responsible for the amusing condescension of the statement with which his article ends: As generativists, if we acknowledged him [i. e., the phoneme], then it was as an illegitimate child. Perhaps we can now recognize the little bastard for what he is (Schane 1971: 521).

The second major disagreement mentioned at the start of this article is the dispute, still not thoroughly settled, as to whether sound-changes are regular, or are irregular with sporadic and unpredictable exceptions. The Neogrammarian position was that sound-changes were

244

Archibald Α. Hill

always and without exception regular, and that apparent exceptions were the result of such causes as borrowing, or analogy. A typical example of borrowing is skirt from Northern Middle English (ultimately Old Norse), beside native shirt. A special type of borrowing is spelling pronunciation, as the American pronunciation of coney, in Coney Island, with the vowel of stone instead of the older and still British pronunciation with the vowel of stun. A typical example of analogy is the pronunciation of covert with the first vowel made identical with overt, instead of the older vowel, that of cover. As for regularity, once again it is proper to quote Sturtevant, since he is particularly revealing on that subject. Speaking of the days when students of language still spoke of sporadic changes, he says: It was possible to connect any two words of any two languages by merely assuming a sufficient number of sound-changes. Voltaire's sneering description of etymology as the science in which consonants count for little and vowels for nothing continued to be deserved for many years after the death of its author (Sturtevant 1917: 72 — 74).

Going on to the days after Leskien had set up the pronouncement of Ausnahmslosigkeit, Sturtevant continues: The matter of regularity was vigorously debated for many years, and the outcome ... is a compromise. Few scholars now deny the possibility of exceptions to the phonetic laws, but in practice all reputable linguists assume that these laws are regular, and all refrain from advancing etymologies which violate them (1917: 72 — 74).

A very striking example of the way in which acceptance of Ausnahmslosigkeit led to advance in etymological study lies in the words eleven and twelve. No less a scholar than the great Franz Bopp equated Latin undecim with Germanic forms like the Gothic ain-lif, thus equating the second element with the form which appears in the numbers thirteen, fourteen, etc. And even the equally great August Schleicher agreed, saying that a change from d to / was a sporadic change not unknown elsewhere in the world. It was Rasmus Rask who recognized the improbability of the Bopp — Schleicher derivation and referred the Germanic forms, as Pedersen points out, to a formation "one left [after ten]". A belief in regularity of sound-change is still the best defense against the amateur etymologist, like one I knew who railed against the derivation of the weakly stressed pronoun in a sentence like "watch 'em go", from the Old English pronoun hem rather than from the modern strong form (from Old Norse) them. My acquaintance used to cite this derivation as a typical example of philological over-

A structural view of sound-change

245

complication and absurdity. Loss of h under weak stress is regular, while loss of the voiced sound which begins them is not. It is, however, necessary to point out that regularity is not accepted even as a heuristic working principle by all linguists. Among them is Roger Lass, who in 1980 cited the American pronunciation of ass 'fundament' from Old English ears and British arse, as having been produced by a "sporadic and relatively rare" change (1980: 77). Such a statement would seem to imply that he thinks no further explanation is necessary.1 The third question with which I started, it will be remembered, was whether dissemination of sound-change occurs word by word, or phoneme by phoneme. Obviously, such changes as analogies, borrowings, and the like occur word by word, and there are others as well, such as haplologies. As the late Bernard Bloch used to remark, the adjective haplogical is itself an example of the change, since the etymology would lead us to expect *haplological. There can be no doubt that some changes proceed word by word. Further, I think that some true sound-changes, that is changes in pronunciation due to phonetic conditioning, also proceed word by word. Thus Labov in 1981 cites the distribution of variant vowels in a series of words which had an earlier short open ο: Every dialect and subdialect of American English shows a different distribution. My own speech shows tense [i.e., with the vowel of law] moth, wroth, cloth, but lax [i.e., with the vowel of hot ] Goth, tense strong, long, song, but lax ping-pong, thong; tense moral and coral, but lax sorrel and tomorrow (Labov 1981: 284).

In Labov's list, as exemplification of individual variation, my own speech differs from his as follows: The /o/ of law occurs in thong where Labov has /a/, and the vowel /a/ occurs where he has /o/ in moral and coral. In addition I have the vowel of stone in wroth, where he has /o/. It would seem to me that such a distribution is nowadays much more easily explained as variant sound-change, diffused word by word rather than phoneme by phoneme. But Labov goes on to demonstrate that there are also sound-changes which are gradual, unconscious, and are diffused phoneme by phoneme. His study of living language and its changes lead him to the following conclusion: In English, the progressive devoicing of final consonants seems to be a regular, Neogrammarian shift. A more dramatic lenition of postvocalic stops is to be found in Liverpool, where voiceless stops become affricates and fricatives — beginning with /k/, and proceeding to jtj and perhaps /p/ among younger children. My own

246

Archibald Α. Hill

explorations of this process show that it is both gradual and regular. Such a replay of Grimm's Law reminds us that it is no accident that the Neogrammarians' most brilliant successes were scored on the original leniting shifts of Grimm's Law. There is every reason to think that this was a gradual process, a phonetic output rule: as paradigmatic an example of Neogrammarian change as we might look for (Labov 1981: 302).

I think that every linguist must be grateful for Labov's conclusive and brilliantly evidential study. There are, furthermore, some important implications from his results which should be stated here. Thus if all sound-change were diffused word by word, the structure of language would be much more chaotic than it is. All forms would be individually variable like the list of /o/ and /a/ forms cited by Labov. In fact, were that the only type of change, I find myself wondering whether there would be any systematic and patterned distribution of sounds as identifiers at all. It seems to me also that the change of distinctive features from ancestral /p t k/ with voiceless closure to Germanic /f θ χ/ with voiceless constriction instead, represents the type of sound-change which most clearly demonstrates the patterned quality of sounds as identifiers. The pattern is equally symmetrical before and after the change, and this quality in sounds is what makes possible an analysis which is in structural terms, and even the description of highly individual variation like those in the pronunciation of thong and coral in terms of phonemes arranged on a symmetrical scheme of combinations of distinctive features. In short, a final definition of a structuralist attitude to sound-change is that no soundchange can be adequately described without taking into account its effect on the pattern of phonemes.

Note 1. Lass is apparently unaware of my study of 1940, "Early loss of r before dentals", PMLA 55: 308 — 359. In that study I tried to show that loss of r before such sounds as 5 was actually widespread, though usually suppressed by regression. As is characteristic of regressive changes, the restoration of r extends to forms which did not originally have it, as in Standard English parsnip from earlier pasnip. It is then characteristic of forms subject to regression that they show a less than perfectly regular distribution.

A structural view of sound-change

247

References Hill, Archibald A. 1936

"Phonetic and phonemic change", Language 12: 15 — 22. (Cf. Joos 1957 below). 1940 "Early loss of r before dentals", PMLA 55: 3 0 8 - 3 5 9 . 1970 "Laymen, lexicographers, and linguists", Language 46: 245 — 258. Hoenigswald, Henry M. 1946 "Sound change and linguistic structure", Language 22: 138 — 141. (Cf. Joos 1957 below). Joos, Martin 1957 Readings in linguistics. Vol. I. Chicago: University of Chicago Press. (Joos reprints Hill 1936: 8 1 - 8 4 , and Hoenigswald 1946: 1 3 9 - 1 4 3 [both with comments]). Labov, William 1981 "Resolving the Neogrammarian controversy", Language 57: 267 — 308. Lass, Roger 1980 On explaining language change (Cambridge Studies in Linguistics 27). Cambridge: Cambridge University Press. Pedersen, Holger 1931 Linguistic science in the nineteenth century: Methods and results (trans. V. W. Spargo). Cambridge, Mass.: Harvard University Press. [1962] [Reprinted as: The discovery of language: Linguistic science in the nineteenth century. Bloomington: Indiana University Press]. Schane, Sanford 1971 "The phoneme revisited", Language 47: 503-521. Sturtevant, Edgar H. 1917 Linguistic change: An introduction to the historical study of language. Chicago: University of Chicago Press. (Reprinted with an introduction by Eric Hamp 1961). 1947 An introduction to linguistic science. New Haven: Yale University Press. (Reprinted 1960).

The transformational-generative model Robert D. King

It is not really possible to speak of a single model of change in transformational-generative grammar (TG) at the present time, simply because there is not a single theory of TG that commands universal acceptance. Or to put it in a different way that is slightly more accurate, there is a TG model of linguistic change, but that model has so many variations in the details that one would do well to proceed as though there were different models. The several different models of change that could be posited stand in a one-to-one relationship with the several different variants of TG that have followers in the linguistic discipline. The fact is that today, unlike in 1968, there is not a single monolithic form of TG that everyone accepts as correct. Some linguists who are TG grammarians believe that rules are ordered, others do not; some believe that rules are ordered but by universal principles of language (intrinsically) and not by statements particular to the specific grammar of a language (extrinsically); some TG linguists believe that there can be a great distance between underlying phonemic and surface phonetic representations (abstract phonology), others do not (concrete phonology). In a sense, then, every variant of synchronic TG will have its own model of change, or at least its own typology of possible changes. In the version of TG that permits extrinsic ordering of rules, one possible kind of change will be rule reordering; but rule reordering is not a possible change in a theory that has no ordered rules. Similarly, other versions of TG will delimit the possible set of changes depending on the constraints on the synchronic organization of the grammar. However, there is an advantage in the exposition of all this diversity so characteristic of TG in the 1980s rather than acting as if it did not

250

Robert D. King

exist — as if we were still situated in the linguistic world of 1965 — 1969 when such disagreements as TG linguists had, and they were few, revolved around details on the periphery of the theory and did not go to its heart.

1. Typologies of change In this discussion I will confine myself to phonological (morphophonemic) change and leave open the question of whether any of this applies to syntactic change. Theories of syntax are even more numerous and less tightly constrained than theories of phonology, so my comments above concerning the multiplicity of typologies of phonological change (within TG) apply with even greater force to models of syntactic change. The organization of phonology in TG circa 1968 went roughly as follows: The phonological component of a grammar consists of a set of rules, some of which may be ordered with respect to others, i.e., some of which may be required to precede others, in order of application. Underlying forms, whose exact feature specification is determined by the evidence of morphophonemic alternations and by economy, are acted on by the phonological rules in succession, ultimately producing the surface phonetic representation. To cite one example from the paradigm TG description of a language from this period (Chomsky — Halle 1968), the grammar of English contains (among many others) three rules ordered in the sequence given: Velar Softening, which changes k to s before i; Trisyllabic Laxing, which laxes a tense vowel followed by two syllables; and Vowel Shift, which changes tense i to [ai]. The underlying form of, for example, criticize is /krltik+Tz/, with various crucial parts of the underlying representation supported by morphophonemic alternations like critic/criticism, criticjcrisis, and criticize I criticism. The derivation of criticize will then go as follows: Underlying representation: Velar Softening: Trisyllabic Laxing: Vowel Shift: (Other rules): Surface phonetic representation:

/krltik +Tz/ krltis+Iz kritis+Tz kritis + aiz [krltasaiz]

The transformational-generative

model

251

A phonology, in this theory, then consists of an underlying form /F/ which is acted on by the (partially) ordered rules R , , R 2 , ..., R n to produce the phonetic output [P], In this theory the set of possible changes is as follows (for further details and examples see King 1969: 39 — 63): Transformational-generative grammar circa 1968 A. Primary Change 1. Rule Addition (at the end of the grammar) 2. Rule Insertion (in the middle of the grammar) B. Simplification 1. Rule Change a. Rule Loss b. Rule Reordering c. Rule Simplification 2. Lexical Change (Relexicalization) This enumerates the ways in which a grammar of a language may change, a classification of the ways in which two grammars from two successive periods of a language may differ. What is being claimed, therefore, is that if two chronologically sequential grammars of a language are compared, they will differ by one of the kinds of change listed under A and B: a rule will be added or lost, two rules will reorder, and so on. Relexicalization means that the underlying representation in the lexicon changes: if a language has /pad/ = [pad] and a rule is added changing d to t without qualification, and if the rule produces no morphophonemic alternations, then the form /pad/ in the lexicon will be replaced by /pat/. Relexicalization is of course commonplace and not controversial. That a change in underlying representation is a possible change is implicit in every theory that posits a level of representation more abstract than the phonetic. How much of this typology is still valid, given what we have learned about phonological change since the 1960s? Rule Addition is still solid, though we now know, thanks to the pioneering work of William Labov (e. g., 1970, 1972), how complicated it really is in real time. It appears that a rule is innovated with various kinds of conditioning, among them phonological, morphological, and social: Certain phonetic environments favor or impede the operation of the rule, likewise certain morphological environments; and the rule is more apt to be used in

252

Robert D. King

one social situation rather than another. As time goes on one of these factors will tend to become dominant and crowd out the others, yielding in the end rules that are either "purely" phonological, or "purely" morphological, or "purely" stylistic. In other words, Rule Addition is a possible kind of change (indeed, probably the most common). However, the process by which a rule is added is immensely complicated, as Labov has richly demonstrated, and goes far beyond simply tacking on a rule at the end of a grammar. Rule Insertion, the placing of an innovated rule in the middle of a grammar so that it operates on a level of representation more abstract than the surface phonetic, attracted a great deal of discussion in the late 1960s. This was partly because it provided dramatic confirmation of abstract levels of representation, partly because it offered a neat solution to certain long-standing historical puzzles, notably Lachmann's "Law" in Latin. Rule Insertion is now regarded as bogus; it does not occur (see King 1973; Hogg 1976). New rules, when they occur, are placed at the end of the rules in the earlier grammar, not in the middle of them. It is useful to differentiate Primary Change (Type A change, i.e., Rule Addition) from changes like Rule Loss and Rule Reordering. The latter seem typically to be motivated by the tendency toward greater simplicity, in various senses. Simplification is characteristic of child language, so we assume that Type Β changes (Simplification) arise in the acquisition of language by a new generation of speakers. Rule Loss is still secure as a possible change. It now appears clear, however, that rules do not simply get lost from the grammar of a language; they must first become hard to learn. Specifically, it now seems correct that a necessary, though not sufficient, condition for Rule Loss is that the rule first become opaque. The notion of rule opacity is due to Paul Kiparsky (1971) and is defined in this way: A rule A —> Β / C D i s said to be opaque if i) A occurs in the environment C D, or if ii) Β occurs in environments other than C D. The importance of branch i) of opacity in Rule Loss is established; that of branch ii) is not certain. (For an in-depth study of one particular instance of Rule Loss, see King 1980.) Rule Reordering is highly problematic for various reasons, chief among them the fact that the status of rule ordering in synchronic grammars is itself problematic. One of the leading variants of TG, Natural Generative Phonology, denies that phonological rules are ordered at all (Vennemann 1974; Hooper 1976). Other variants require

The transformational-generative

model

253

rules to be intrinsically ordered, i.e., sequenced by universal principles (Koutsoudas —Sanders —Noll 1974). One such principle, for example, is Proper Inclusion Precedence, whereby the less general rule must precede the more general rule in case the structural description of one properly includes the other. If such a principle is in fact a language universal, then of course it would be impossible for the two rules to reorder against the universal principle. Rule Reordering is a possible change only in a theory that has extrinsic ordering, i.e., explicit statements that Rule A must precede Rule B. The ordering-no ordering controversy raged throughout the 1970s. On balance I believe that most rules, if they are ordered at all, are ordered by general principles (intrinsically), and that at most a small residue of rules is ordered extrinsically. But I do not think the argument can be regarded as closed (King 1976). If provisionally we do assume that rules are extrinsically ordered, then certain generalizations can be made that have to do with the motivation for reordering. Initially, the motivation advanced for reordering was a functional one — one based on efficiency (Kiparsky 1968). The idea was that rules tend to shift into an order that maximizes their utilization: bleeding orders are replaced by non-bleeding orders; non-feeding orders are replaced by feeding orders. It now appears that this principle is wrong; the correct principle that motivates rule reordering is the "paradigm principle": Rules tend to shift into the order that increases paradigm regularity (King 1973). In other words, if two rules apply in the order A —B, and if the order B —A will produce more regular paradigms, then reordering from A —Β to B —A will tend to occur. There are no well-proven cases in which rules have reordered from a non-feeding order to a feeding order. Rule Simplification is still valid. More precisely, one should speak of the "simplification of the structural description of a rule". If something happens before [p k], then we may predict that eventually the same something will happen before [p t k] — the natural class [p t k] is simpler by a feature than [p k]. Based on this discussion we may posit the following taxonomy of change for present-day TG: Transformational-generative grammar circa 1982 A. Primary Change 1. Rule Addition (at the end of the grammar)

254

Robert D. King

B. Simplification 1. Rule Change a. Rule Loss (provided the rule is opaque) b. Rule Reordering (?) c. Rule Simplification 2. Lexical Change (Relexicalization)

2. Grammar transmission We assume that the adult speakers of a language community have a grammar underlying their speech output. The grammar of an adult cannot change radically, but it can be changed by rule addition. The child, developing his language ability by hearing what is spoken around him and by the desire to understand and be understood, constructs a grammar (underlying forms, rules) that will enable him to speak as others do. Various factors come into play in the "Language Acquisition Device" (LAD) of the child: universal constraints on underlying forms and the form of rules, the desire to simplify, the functional necessity of producing speech that is not sharply discontinuous with the existing speech of the community, and so on. Through rule addition and simplification language changes over time, as follows:

Generation 1:

Grammar 1

Generation 2:

LAD

1. Language Universals 2. Evaluation Metric 3. Functional Constraints

Grammar 2

Speech Output 1

The transformational-generative

model

255

This model of language transmission, and of the relationship between adult and child grammars, has remained essentially unchanged since the 1960s. It is, of course, like any model an over-simplification. It says too little about the social matrix of change (Weinreich — Labov —Herzog 1968). It says too little about acoustic-auditory factors in the initiation and transmission of change (Ohala 1974). It does not take proper account of the importance of a kind of change called "abductive" (Andersen 1973). And it underestimates the amount of change that takes place in adult speech (Labov 1972). But to say that the model does too little in one regard or another is a rather cheap criticism to make, since the same criticisms can be made of any other model of linguistic change. The model is a skeleton that provides the minimal framework for describing and trying to explain change, and for this the TG model has withstood the test of time.

References Andersen, Henning 1973 "Abductive and deductive change", Language 49: 765 — 793. Bach, Emmon —Robert T. Harms (eds.) 1968 Universals in linguistic theory. New York: Holt, Rinehart & Winston. Bruck, Anthony —Robert Fox —Michael W. LaGaly (eds.) 1974 Papers from the Parasession on Natural Phonology. Chicago Linguistic Society Annual Meeting, 18 April 1974. Chicago: Chicago Linguistic Society. Chomsky, Noam —Morris Halle 1968 The sound pattern of English. New York: Harper & Row. Dingwall, William O. (ed.) 1971 A survey of linguistic science. College Park: University of Maryland Press. Herzog, Marvin I. et al. (eds.) 1980 The field of Yiddish. Philadelphia: Institute for the Study of Human Issues. Hogg, Richard M. 1976 "The status of rule reordering", Journal of Linguistics 12: 103 — 123. Hooper, Joan B. 1976 An introduction to natural generative phonology. New York: Academic Press. King, Robert D. 1969 Historical linguistics and generative grammar. Englewood Cliffs: Prentice Hall. 1973 "Rule insertion", Language 49: 5 5 1 - 5 7 8 . 1976 "In defense of extrinsic ordering", in: Koutsoudas (ed.), 76 — 96.

256 1980

Robert D. King

"The history of final devoicing in Yiddish", in: Herzog et al. (eds.), 371 — 430. Kiparsky, Paul 1968 "Linguistic universale and linguistic change", in: Bach —Harms (eds.), 170-202. 1971 "Historical linguistics", in: Dingwall (ed.), 577-649. Koutsoudas, Andreas (ed.) 1976 The application and ordering of grammatical rules. The Hague: Mouton. Koutsoudas, Andreas — Gerald Sanders — Craig Noll 1974 "On the application of phonological rules", Language 50: 1—28. Labov, William 1970 "The study of language in its social context", Studium Generale 23: 30 — 87. 1972 "The internal evolution of linguistic rules", in: Stockwell — Macauley (eds.), 101-171. Lehmann, Winfred P. — Yakov Malkiel (eds.) 1968 Directions for historical linguistics. Austin: University of Texas Press. Ohala, John J. 1974 "Phonetic explanation in phonology", in: Bruck — Fox — LaGaly (eds.), 251-274. Stockwell, Robert P . - R o n a l d K. S. Macauley (eds.) 1972 Linguistic change and generative theory. Bloomington: Indiana University Press. Vennemann, Theo 1974 "Words and syllables in natural generative grammar", in: Bruck —Fox — LaGaly (eds.), 346-374. Weinreich, Uriel —William Labov —Marvin I. Herzog 1968 "Empirical foundations for a theory of language change", in: Lehmann — Malkiel (eds.), 9 7 - 1 9 5 .

Other Approaches

Dialect geography Peter

Trudgill

As an academic study linguistic geography can be said to have begun towards the end of the last century. 1 In 1876 Georg Wenker began the work that was eventually to lead to the compilation of the Deutscher Sprachatlas (Wrede 1926 — 1956; see also Mitzka 1952). Wenker sent out questionnaires by post to every school district in the German Empire (about 40,000 in all) with the request that teachers "translate" the 40 questionnaire sentences into the local dialect. Wenker received over 44,000 replies, and his technique proved to be a very successful way of obtaining information as to which vocabulary items were used in different parts of the German-speaking area. The method, however, was not very successful in obtaining material on pronunciation, since normal orthography is not adequate for representing fine shades of differences in sound. Subsequent studies, including later work on German, therefore employed the "direct method". This is generally recognized to have been developed in connection with the compilation of the Atlas linguistique de la France (Gillieron — Edmont 1902 — 1910) — a work in thirteen volumes. Field-work for this atlas was begun in 1897 and was carried out entirely by Edmond Edmont under the direction of Jules Gillieron who had earlier (1880) published a small linguistic atlas of an area of French-speaking Switzerland. Gillieron bought Edmont a bicycle and sent him out, with a previously developed questionnaire, to record responses in 639 different French communities. The questionnaire was much larger than the one used in Germany, containing eventually about 1900 items, but the amount of data obtained was much less than that obtained by Wenker, since the geographical coverage was much less dense. Publication could therefore follow much more quickly. Attempts were made, however, to ensure

258

Peter

Trudgill

that the coverage was adequate by selecting areas for study according to a fixed geometric pattern. (Similar methods are being employed in a more recent survey of Sweden — see Dahlstedt 1972.) The "direct method" itself involved the field-worker in interviewing informants in person and noting down their actual pronunciation in a specially devised phonetic script. Both the postal and direct methods continue to be employed, with various refinements, to the present day, supplemented in the last 30 years or so by the availability of the portable tape-recorder. Once data has been gathered, by either method, it is possible to display the linguistic material, obtained on maps and, if resources permit, to combine the maps into atlases. (Mapping problems involved in this kind of activity are discussed in Moulton 1972, which is also a very useful survey of European work in this field as a whole.) European atlases in this tradition are presented or discussed in the following works: Romanian (Pu§cariu et al. 1938 — ; Pu§cariu 1938); Swiss German (Hotzenköcherle et al. 1962 — ; Hotzenköcherle 1962); Italian (Jaberg — Jud 1928-1940; Franceschi 1967); Dutch (Kloeke 1 9 3 9 - ; Blancquaert et al. 1930-); Catalan (Griera 1 9 2 4 - ) ; Spanish (Diego 1946); Swedish (Benson 1 9 6 5 - ; Götlind - Landtmanson 1940-1950); Norwegian (Hoff 1967); Danish (Bennike - Kristensen 1898-1912); Irish (Wagner 1958); Welsh (Thomas 1973); Ukrainian (Zylko et al. 1948); Lithuanian (Senkus 1956); Bulgarian (Stojkov 1957); Hungarian (Bärczi 1957); Polish (Nitsch-Urbahczyk 1957-1960). Later works are also under way for French (Seguy 1954 — ; Nauton 1957—1963; Gardette 1 9 5 0 - ; Haust et al. 1953 — ) and German (Klein - Schmitt 1961 - , 1 9 6 5 - ; Schmitt 1 9 6 3 - , 1965-). For English we have dialect surveys of the United States and Canada (see Kurath et al. 1939 — 1943; K u r a t h - M c D a v i d 1961; K u r a t h - B l o c h 1939); England (Orton et al. 1 9 6 2 - , 1978; O r t o n - W r i g h t 1974; Kolb 1966; and see Wakelin 1973); and Scotland (see Mcintosh 1952; Catford 1957; M a t h e r Speitel 1975-). Work on linguistic geography elsewhere has produced the following: Afrikaans (Louw 1959 — ); South American Spanish (Florez 1961; Navarro 1948; Cardenas 1958); and Brazilian Portuguese (Rossi 1963). Word atlases, dealing only with the regional distribution of vocabulary items, have also been particularly popular (Kretschmer 1918; Lindqvist 1947; Davis 1948; Kurath 1949; Mitzka - Schmitt 1 9 5 1 - ; Reed 1957; Wood 1961; Hankey 1960); studies of grammatical variation (Atwood 1953) less so.

Dialect geography

259

In the work of Wenker and those who followed him it emerged that, in most cases, words and pronunciations were not randomly distributed geographically, but were confined to particular regions. This was not particularly surprising since it had been recognized for a long time that dialect areas existed. However, something also emerged which many linguists did find somewhat startling. When the geographical distribution of a particular linguistic feature had been plotted on a map, it was possible to draw lines on the map separating areas which had the feature from those which did not. 2 When maps for individual features were combined into overall displays incorporating a number of dialect differences, it was found that very few of the isoglosses actually coincided. Instead, lines drawn for individual features crisscrossed the maps (and each other) in an apparently haphazard fashion. This was initially a rather distressing revelation. Linguists (or many of them) had previously assumed that dialects were discrete entities, that speakers were either speakers of a particular named dialect or not, and that dialect boundaries actually existed. Now it became clear that this was far from being the case. Dialect boundaries as such could not be found, and one initial reaction was to suggest that there were "no such things as dialects" — "dialect" was thought to be perhaps a meaningless concept (cf. Gauchat 1904). The reasons for this reaction are not difficult to understand. In Germany, for instance, it was a fact well-known to linguists that dialect speakers in the north of the country said hüs 'house', müs 'mouse', etc. with a long, pure vowel, whereas speakers in the south said haus, maus, etc. with a diphthong. The task of linguistic geography, it had been assumed, was simply to find out where the boundary between the two types lay. It turned out, however, that if isoglosses were drawn for each ü/au word separately, few of them followed exactly the same route. There were not, that is, simply hüs dialects and haus dialects: there were hüs, müs dialects; hüs, maus dialects; haus, maus dialects, and so on. When other words, vowels and consonants were taken into consideration the position became very complex indeed. Eventually dialectologists recovered from this shock. First, it began to be recognized that, in most cases, patterns could be detected in dialect maps, and dialectologists were able to point to types of phenomena that tended to recur from map to map. For example, zones which were crossed by relatively few isoglosses were picked out and labeled "focal areas". Focal areas tended, in turn, to be surrounded either by "transition areas", which were crossed by many isoglosses

260

Peter Trudgill

and did not belong clearly to any focal area, or by "bundles of isoglosses" — isoglosses, it was true, did not often coincide, but they did frequently run close together and in the same general direction. For further discussion see Wagner (1927) and Bach (1950). It therefore began to appear that dialects, while obviously no longer to be regarded as discrete entities, remained valid as relative units. There were some (focal) areas that could be labeled unambiguously, say "Central Bavarian" or "Southern Bavarian" dialect areas (Kranzmayer 1956), and other (transition) areas that were something intermediate. 3 This is typical of much linguistic data. Language is a dynamic phenomenon and is continually subject to change. Continua, gradata and variability are frequently encountered, and most differences are of the more-or-less rather than the either-or variety. Dialect maps also led to the development of an interest in why particular isoglosses happened to be located at particular places, and in some cases explanations could be advanced (Jaberg 1936; Bach 1950; Moser 1954). For example, it was noted from the configuration of certain isoglosses that linguistic forms had obviously spread outwards as innovations from particular centers. These were generally either urban centers, or major lines of communication such as the Danube. Kranzmayer (1956) showed that, in many respects, the Central Bavarian dialect of German (including Munich, Vienna and the Danube Valley) was innovating, while North Bavarian (including the Regensburg and Nuremberg areas) and South Bavarian (southern Austria) were more conservative. Central Bavarian, for instance, has lost I in words like Salz 'salt' and Geld 'money', while the other dialects have retained it. Thus the area around the Danube has become a focal area as the result of the outward diffusion of linguistic innovations. It could also be shown that the spreading of new words or pronunciations took the form of wedges driven into the territory of older forms, and where two wedges joined up isolated "islands" might be left behind. These more conservative zones were termed "relic areas" and tended to be located in isolated areas like mountain valleys or on the distant periphery of language areas. Transition zones, on the other hand, resulted from the fact that different innovations traveled similar but not identical distances in different directions. This differential location of isoglosses could often be accounted for in terms of the chronology of their origin, together with changes in communications networks at different periods of history. (An important study on the spread of linguistic forms in Dutch is Kloeke 1927).

Dialect geography

261

It was also apparent that linguistic innovations tended to spread further along major rivers than they did over more difficult terrain, and that bundles of isoglosses sometimes coincided with political boundaries, past and present, or with physical barriers. A major study by Frings (1956), originally published in 1922, deals with both these factors in a treatment of the dialects of the German Rhineland. A bundle of isoglosses runs across Germany from west to east, including lines for northern Aws/southern haus 'house'; N.maken/S.machen 'make'; N.datjS.das 'that'; N.dorp/S.dorf 'village'. However, when the bundle reaches the Rhineland the isoglosses separate out into what has become known as the "Rhenish fan". Frings was able to show that the area bounded by the maken/machen line north of Cologne and the dorpjdorf, hüsjhaus line south of Bonn was coextensive with the old diocese of Cologne. Similarly, the area to the south of the dorpldorf line and to the north of the datjdas line, north of Mainz, was coextensive with the former diocese of Trier. These had been relatively stable ecclesiastical administrative areas from the Middle Ages until the end of the eighteenth century, and Frings suggests that linguistic innovations spreading from the culturally dominant area of southern Germany travelled along the Rhine to urban centers (such as Cologne and Trier) and then outwards from these centers to the edges of the administrative areas which they dominated. Because of phenomena such as these, linguistic maps and atlases compiled from dialect surveys can be used as research tools in historical linguistics and, to a certain extent, history. (Notable examples of work of this type are Jaberg and Jud (1928), and Kloeke (1950), who was able to show which dialects of Dutch had been dominant in the formation of Afrikaans.) In the words of Bottiglioni (1954), "Just as a geologist moves from the morphological aspect of the ground to discover the sedimentary processes that have determined it, so the linguist needs a faithful representation of the linguistic area to reconstruct its history". Maps can demonstrate the probable direction a linguistic change has taken, and help to shed light on problems such as the relative age of two current forms. In fact, a whole school of linguistics grew up around factors such as these. Bartoli's (1925) "Neolinguistics" — a name later modified to "Linguistica Spaziale" (1945) — is based, among other things, on five principles or "areal norms" which can be used in the study of historical linguistics. Three of these "norms" are genuinely geographical (see Bonfante 1947):

262

Peter

Trudgill

1) If, of two linguistic forms, one is found in isolated areas and the other in areas more accessible for communication, then the former is the older. 2) If, of two linguistic forms, one is found in peripheral areas and the other in central areas, then the former is the older. 3) If, of two forms, one is used over a larger area than the other, then that is the older. "Linguistica Spaziale" became discredited because, as a method of historical linguistics, it proved to be of doubtful value, since many exceptions, especially to principle 3), were readily found. Principle 1), too, can easily be contradicted, with a British example. The dialects of the rural north-east of Scotland (particularly Buchan) have an / at the beginning of words such as what and which which is known (largely from the spelling) to be newer than the hw found elsewhere in Scotland, including the industrialized Lowlands (as well as in other parts of the English-speaking world). Similarly, principle 2) can be "disproved": Scottish accents and accents in the south of England (peripheral areas) have a vocalic distinction put Iputt, could/cud. Accents in the North and Midlands of England (the central area) have, on the other hand, no distinction. However, we know from other sources (older descriptions and texts) that it is Scotland and the south of England which are innovating. Earlier stages of the English language had no distinction between put and putt. The "Neolinguists" attempted to treat these principles as "laws", and their lack of flexibility in the handling of these norms proved to be their downfall. However, it cannot be denied that the principles have some validity if they are regarded as guidelines rather than as "laws". If, for example, we consider different accents of English as it is spoken in Britain, we find that these accents fall, in one respect, into two separate groups: Those which actually pronounce the r in words like cart and car (where the r comes before a consonant or at the end of a word), and those which do not. If we confine our attention to old-fashioned rural dialect speech, it turns out that there are three main separate "r-pronouncing" areas in England — areas where pa and par are not pronounced the same. These three areas are, broadly speaking, the south-west, part of Lancashire and the northwest, and part of the north-east — stretching up into Scotland. There is, on the other hand, only one continuous non-r-pronouncing area. This suggests very strongly that it is the loss of r which is the innovation — cah and caht are newer pronunciations than car and cart. We can

Dialect geography

263

assume that this is the case since it is most unlikely that an identical innovation would have started in three widely separated areas at once. We can therefore assume that "r-dropping" began somewhere in the east of the country and has since spread north and west, creating wedge-like patterns which have divided the original r-pronouncing region into three. Good overviews of work in linguistic geography are provided in Pop (1950-1951), Kurath (1972), and Petyt (1980). Dialectology of the traditional, linguistic geography type revealed that, allowing for factors such as migration and ethnic group differentiation, geographically contiguous dialects tend to be more similar than those which are spatially widely separated. Other studies have shown that this is equally true of geographically contiguous languages. Languages which are spoken in the same general geographical region tend to have features in common even if they are not closely related (from a historical point of view) to each other, and even if they are totally mutually incomprehensible. Isoglosses can be drawn for many grammatical and phonetic phenomena which bear no relation to language frontiers. In Europe, for example, front rounded vowels of the type found in French peu, German böse are used in Finnish, Swedish, Norwegian, Danish, German, Dutch, French, and northern dialects of Italian. It is true that, with the exception of Finnish, these languages are all related. However, we can point out that French, for example, is much more closely related to Spanish, which does not have such vowels, than to German, which does. English, which is a close relation of Dutch and German, does not have front rounded vowels either. Similarly, the uvular r pronunciation is found in southern Norway; southern Sweden; Denmark; parts of Germany, Holland, Belgium, Luxembourg and Switzerland; most of France; and part of northwestern Italy. This pronunciation is thought to have started in Paris, although this is not certain, and to have spread to the other areas in the last 250 years or so (see Trudgill 1974). Tongue-tip trilled or flapped r, on the other hand, is known to be an older pronunciation (i. e., it was formerly used in French, German and Danish) and is geographically peripheral in Europe. It is found in northern Scandinavia, Scotland, and part of northern England, in the north; Portugal, Spain, Italy and Greece, in the south; and Austria, Hungary, and the Slavonicspeaking countries in the east.

264

Peter

Trudgill

In many cases these areal features are due to a linguistic innovation, such as uvular r, originating in one dialect and then, as a result of diffusion processes, spreading to neighboring dialects, regardless of language barriers. Where this kind of diffusion takes place on a large scale, it can give rise to the development of "linguistic areas" (German Sprachbund). These are geographical regions where languages, whatever their historical relationships, have a large number of characteristics in common. The subject of linguistic areas was first treated, in connection with the study of American Indian languages, by Boas (1940, first published in 1917) and Sapir (1921). Areal studies of these languages have been felt to be valuable in an analysis of their relationship to each other and their history, and in the study of migrations. Subsequent studies have included Voegelin (1945) and Darnell — Sherzer (1971). But perhaps more influential in the development of the study of linguistic areas have been the writings of Trubetzkoy (1931) and Jakobson (1931, 1938), using data drawn from European languages. Later general writings on this subject include Bonfante — Sebeok (1944); Becker (1948); Dyan (1956); Wolff (1959); and Diebold (1960). The best known linguistic area in Europe is undoubtedly the Balkans. The languages spoken in this area, particularly Romanian, Bulgarian, Macedonian, Albanian, and Greek, are not in most cases closely related, but nonetheless show striking resemblances to one another in many respects. Golab (1959) has shown that there are a number of grammatical similarities between Macedonian and certain varieties of Romanian, and Kazazis (1965, 1967) has made a similar study of syntactic similarities between Albanian, Bulgarian, Greek and Romanian. Other treatments of the Balkans as a linguistic area are Sandfeld (1932); Pavlovic (1958); Birnbaum (1966); and Afendras (1968, 1970). Perhaps the best known and most striking example of linguistic similarity in the Balkans is the fact that four of the languages (Albanian, Romanian, Bulgarian, and Macedonian) all have a post-posed definite article — i. e., the form corresponding to English the is placed after the noun rather than before it, for example: Albanian Romanian

ujk lup 'wolf

ujku lupul 'the wolf

This feature does not occur in related languages which happen not to be spoken in the Balkans. For example, Italian (which is related to

Dialect geography

265

Romanian) and Russian (which is related to Macedonian and Bulgarian) do not have such forms for the definite article. Traditionally, linguistic geography has been geographical only in the sense that it has been concerned with the spatial distribution of linguistic phenomena. More recently, however, some linguists have been attempting to improve work in this field through the study and adoption of certain of the concepts and techniques developed by social geographers. The work that has so far been carried out, although there is as yet little of it, suggests that in fact geographical practice and theory may be of considerable value to linguistics. It appears, for example, that geographical cartographic techniques have much to offer. Hitherto, linguists have normally simply drawn isoglosses on maps separating areas which have a particular feature from those which do not. In actual fact, maps of this type often represent a gross distortion of the linguistic facts: They do not demonstrate what percentage of speakers in these areas actually use a particular feature, or how often they use it; they do not show what age, sex, or social class these speakers are; and they tell us nothing about situational or stylistic factors involved. They are inaccurate because they represent as a discrete break what is in fact a continuum, and as static what is in fact a dynamic situation; differences in language are, as stated above, usually a matter of more-or-less rather than all-or-nothing. We therefore require some way of making dialect maps quantitative. It has been argued (Trudgill 1974; Chambers — Trudgill 1980) that in order to obtain more realistic and useful descriptions of geographical variation in language, linguists should begin to attempt to construct maps rather like those that appear in Hägerstrand (1952). Maps such as these can not only demonstrate density of usage of particular linguistic forms, but also tell us more about the direction of linguistic change, and in much greater detail. As Hägerstrand has said: "When studying changes we cannot draw boundary lines and observe their displacements. Instead we must ascertain the spatial diffusion of ratios". This point can be illustrated with reference to linguistic research carried out in southern Norway (see Trudgill 1974). Here we have been concerned with plotting the diffusion of particular variants of vowel sounds outwards from an urban area into the surrounding countryside. The lines drawn on the maps produced by this study are not isoglosses in the usual sense. Rather they are contours interpolated between sampling points, in accordance with average numerical indices of vowel quality recorded by speakers (cf. Labov 1966) in the manner described

266

Peter Trudgill

by Robinson and Sale (1969, Chapter 7) for height and other contours, and by Hägerstrand (1952). Maps of this type provide a useful manner of portraying linguistically gradient phenomena like vowel sounds, and of handling quantitative data such as percentage of r-usage. They also give a much more accurate picture of the diffusion processes involved in linguistic change than normal isogloss maps, and for that reason are more useful in searching for explanations as to why linguistic changes follow certain routes and not others (see also Trudgill 1983).

Notes 1. Much of the material in this essay is derived from Trudgill (1975). 2. Strictly speaking, lines of this type are known as "isoglosses" on word maps and "isophones" on maps dealing with pronunciation. Frequently, however, "isogloss" is used in both cases. On some problems concerning the drawing of isoglosses, with reference to the Scots-English border area, see Speitel (1969). 3. "Dialect" also came to be a rather more pragmatic term: The fact that linguists used the term "Central Bavarian Dialect" did not mean that they could not also use other terms like "North Central Bavarian Dialect" if they wished to for some particular purpose.

References Afendras, E. 1968

"The Balkans as a linguistic area: A study in phonological convergence". Unpublished Ph. D. dissertation: John Hopkins University. 1970 "Quantitative distinctive feature typologies and a demonstration of areal convergence" (review article), Applied Linguistics 9: 49 — 81. Atwood, E. Bagby 1953 A survey of verb forms in the Eastern United States. Ann Arbor, Michigan: University of Michigan. Bach, Adolf 1950 Deutsche Mundartforschung. Heidelberg: Winter. Bärczi, G. 1957 "Les travaux de l'atlas linguistique de la Hongrie", Acta Linguistica Academiae Scientarum Hungaricae : 1—52. Bärtoli, Matteo 1925 Introduzione alia neolinguistica. Genova: L. S. Olschki. 1945 Saggi di linguistica spaziale. Torino: Bond. Becker, Henrik 1948 Der Sprachbund. Berlin: Humboldt & Biichere. Bennike, V. — Μ. Kristensen 1898 — 1912 Kort over de danske folkemaal med forklaringer. Kebenhavn: Gyldendal.

Dialect geography

267

Benson, S. 1965 Südschwedischer Sprachatlas. Lund: Gleerup. Birnbaum, Henrik 1966 "On typology, affinity and Balkan linguistics", Zbornik za filologiju i linguistiku 9: 17 — 30. Blanquaert, E. et al. 1930— Reeks Nederlandse Dialectatlassen. Antwerpen: De Sikkel. Board, C. et al. (eds.) 1975 Progress in geography. London: Edward Arnold. Boas, Franz 1919 "Introduction", International Journal of American Linguistics 1: 1—8. (Reprinted in Race, Language and Culture, New York: MacMillian, 1940: 199-210). Bonfante, Giuliano 1947 "The neolinguistic position", Language 23: 344 — 375. Bonfante, Giuliano —Thomas A. Sebeok 1944 "Linguistics and the age and area hypothesis", American Anthropologist 46: 382-386. Bottiglioni, G. 1954 "Linguistic geography: Achievements, methods and orientations", Word 10: 375-387. Cardenas, D. N. 1958 "The geographic distribution of the assibilated r, rr in Spanish America", Orbis 7: 407-414. Catford, Ian 1957 "The linguistic survey of Scotland", Orbis 6: 105-121. Chambers, J. —Peter J. Trudgill 1980 Dialectology. London: C.U.P. Dahlstedt, Karl-Hampus 1972 Svensk dialektgeografi, dess metoder och uppgifter. Umeä: Umeä University. Darnell, R. —Joel Sherzer 1971 "Areal linguistic studies in North America: A historical perspective", International Journal of American Linguistics 37: 20 — 28. Davis, A. L. 1948 "A word atlas of the Great Lakes region", unpublished dissertation, University of Michigan. Diebold, A. 1960 "Determining the centers of dispersal of language groups", International Journal of American Linguistics 26: 1 — 10. Diego, V. G. de 1946 Manual de dialectologia. Madrid: Cultura Hispanica. Dyen, Isidore 1956 "Language distribution and migration theory", Language 32: 611 —626. Flörez, L. 1963 "El espanol en Colombia y su atlas linguistico", Thesaurus 18: 268 — 356. Franceschi, T. 1967 "Achievement des recherches dialectales pour FAtlante Linguistico Italiano", Zeitschrift für Mundartforschung Beihefte, Neue Folge 3: 232 — 238.

268

Peter Trudgill

Frings, Theodor 1956 Sprache und Geschichte. Halle: Niemeyer. Gardette, P. 1950— Atlas Unguistique et ethnographique du Lyonnais. Lyon: Institut de Linguistique Romane des Facultes Catholiques de Lyon. Gauchat, L. 1904 "Gibt es Mundartgrenzen?", Archivum für das Studium der neueren Sprachen 111: 365-403. Gillieron, Jules 1880 Petit atlas phonetique du Valais roman. Paris: H. Champion. Gillieron, Jules — Edmond Edmont 1902 —191 OAtlas linguistique de la France. Paris: E. Champion. Golab, S. 1959 "Some Arumanian — Macedonian isogrammatisms and the social background of their development", Word 15: 415—435. Götlind, J. —S. Lantmanson 1940 — 1950 Västergötlands folkmal. Uppsala: Lundequist. Griera, A. 1924— Atlas linguistic de Catalunya. Barcelona: Institut d'Estudis Catalans. Hägerstrand, Τ. 1952 The propagation of innovation waves (Lund Studies in Geography, Series B: Human Geography, No. 4). Lund: Gleerup. Hankey, C. 1960 A Colorado word geography (Publications of the American Dialect Society 34). University, Ala.: University of Alabama Press. Haust, J., et al. 1953 Atlas linguistique de la Wallonie. Liege: Vaillant-Carmanne. Hoff, I. 1967 "Historisch-phonematische Mundartkarten", Zeitschrift für Mundartforschung Beihefte, Neue Folge, 3: 373 — 387. Hotzenköcherle, Rudolf 1962 Einführung in den Sprachatlas der deutschen Schweiz. Bern: Francke. Hotzenköcherle, Rudolf, et al. 1962— Sprachatlas der deutschen Schweiz. Bern: Francke. Jaberg, Karl 1936 Aspects geographiques du language. Paris: E. Droz. 1928 — 1940 Sprach und Sachatlas Italiens und der Südschweiz. Zofingen: Ringier. Jaberg, Karl —Jakob Jud 1928 Der Sprachatlas als Forschungsinstrument. Halle. Jakobson, Roman 1931 "Über die phonologischen Sprachbünde", in: Selected writings. Vol. I. The Hague: Mouton. 1938 "Sur la theorie des affinites phonologiques entre les langues", in: Selected writings. Vol. I. (Also in Troubetzkoy, 1949). Appendix. Kazazis, K. 1965 "Some Balkan constructions corresponding to Western European infinitives", Unpublished dissertation, Indiana University.

Dialect geography 1967

269

"On a generative grammar of the Balkan languages", Foundations of Language 3: 117 — 123. Klein, Kurt —Ludwig Erich Schmitt 1961— Siebenbürgisch — deutscher Sprachatlas. Marburg: Elwert. 1965— Tirolischer Sprachatlas. Marburg: Elwert. Kloeke, G. 1927 De Hollandsache Expansie. 's Gravenhage: Nijhoff. 1939— Taalatlas van Noord- en Zuid-Nederland. Leiden: Brill. 1950 Herkomst en groei van het Afrikaans. Leiden: Leiden University Press. Kolb, E. 1966 Phonological atlas of the northern region: The six northern counties, north Lincolnshire and the Isle of Man. Bern: Francke. Kranzmayer, Eberhard 1956 Historische Lautgeographie des gesamtbairischen Dialektraumes. Graz: Böhlau. Kretschmer, Paul 1918 Wortgeographie der deutschen Umgangssprache. Göttingen: Vandenhoeck & Ruprecht. Kurath, Hans 1949 A word geography of the eastern United States. Ann Arbor, Michigan: University of Michigan. 1972 Studies in area linguistics. Bloomington: Indiana University Press. Kurath, Hans, et al. 1939 Handbook of the linguistic geography of New England. Providence: Brown University. Kurath, Hans —Raven I. McDavid 1961 The pronunciation of English in the Atlantic states (Studies in American English 3). Ann Arbor, Michigan: University of Michigan. Kurath, Hans, et al. 1939 — 1943 Linguistic atlas of New England. Providence: Brown University. Labov, William 1966 The social stratification of English in New York City. Washington: Center for Applied Linguistics. Lindqvist, N. 1947 Sydväst-Sverige i spräkgeografis belysning. Lund: Blom. Louw, S. 1959— Afrikaanse Taalatlas. Pretoria: Pretoria University Press. Mather, J . - H . Speitel 1975— The linguistic atlas of Scotland. London: Croom Helm. Mcintosh, Angus 1952 An introduction to a survey of Scottish dialects. Edinburgh: Nelson. Mitzka, Walther 1952 Handbuch zum deutschen Sprachatlas. Marburg: Elwert. Mitzka, Waither — Ludwig Erich Schmitt 1951— Deutscher Wortatlas. Glessen: Schmitz. Moser, Hugo 1954 "Sprachgrenzen und ihre Ursachen", Zeitschrift fur Mundartforschung 22: 87-111.

270

Peter Trudgill

Moulton, William G. 1972 "Geographical linguistics", in: Sebeok (ed.), 196-222. Nauton, P. 1957—1963 Atlas linguistique et ethnographique du Massif Central. Paris: Centre National de la Recherche Scientifique. Navarro, T. 1948 El espanol en Puerto Rico: contribucion a la geograßa lingiiistica hispanoamericana. Rio Piedras: Editorial Universitas Universidad del Puerto Rico. Nitsch, K. —S. Urbariczyk 1957— 1960 Maly atlas gwar polskich. Wroclaw: Ossolineum. Orton, H. 1978 The linguistic atlas of England. London: Croom Helm. Orton, H., et al. 1962— Survey of English dialects. Leeds: Arnold. Orton, H . - N . Wright 1974 A word geography of England. London: Seminar. Pavlovic, M. 1958 "Les processus balkanistiques et la perspective du parier serbe de Prizren", Orbis 7: 4 6 - 5 3 . Petyt, Κ. M. 1980 The study of dialect. London: Deutsch. Pop, Sever 1950—1951 La dialectologie. 2 vols. Louvain/Gembloux: Duculot. Pu§cariu, Sergiu 1938 "Der rumänische Sprachatlas", Archivum für vergleichende Phonetik 2: 107-117. Pu§cariu, Sergiu, et al. 1938— Atlasul linguistic roman. Cluj-Sibiu: Rumänische Bibliothek. Reed, C. E. 1957 "Word geography of the Pacific Northwest", Orbis 6: 8 6 - 9 3 . Robinson, A. — R. Sale 1969 Elements of cartography. New York: Wiley. Rossi, Ν. 1963 Atlas previo dos Falares Baianos. Rio de Janeiro: Institute National do Livro. Sandfeld, Kristian 1932 Linguistique balkanique. Paris: E. Champion. Sapir, Edward 1921 Language. New York: Harcourt & Brace. Schmitt, Ludwig Erich 1963— Luxemburgischer Sprachatlas. Marburg: Elwert. 1965— Schlesischer Sprachatlas. Marburg: Elwert. Sebeok, Thomas A. (ed.) 1972 Current trends in linguistics 9, Part 1. The Hague: Mouton. Seguy, J. 1954— Atlas linguistique et ethnographique de la Gascogne. Toulouse: Institut d'etudes meridionales de la faculte des lettres.

Dialect geography 1973 Senkus, J. 1956

271

"Les atlas linguistiques de la France par regions", Langue Frangaise 18: 65-90. Lietuviu kalbos atlaso medziagos rinkimo programa. Lithuanian Language and Literature.

Vilnius: Institute of

Speitel, Η. H. 1969 "An areal typology of isoglosses: Isoglosses near the Scottish-English border", Zeitschrift für Dialektologie und Linguistik 1: 50 — 66. Stojkov, S. 1957 "Bülgarski dialekten atlas", Ezik i literatura 12: 2 8 5 - 2 9 1 . Thomas, A. R. 1973 The linguistic geography of Wales. Cardiff: University of Wales Press. Trubetzkoy, N. S. 1931 "Phonologie et geographie linguistique", TCLP IV: 2 2 8 - 2 3 4 (reprinted as Appendix 3 in his Principes de phonologie, 343 — 350. Paris: C. Klincksieck, 1949). Trudgill, Peter J. 1974 "Linguistic change and diffusion: Description and explanation in sociolinguistic dialect geography", Language in Society 3: 215 — 246. 1975 "Linguistic geography and geographical linguistics", in: Board et al. (eds.). 1983 On dialect: Social and geographical perspectives. Oxford: Blackwell. Voegelin, Charles F. 1945 "The influence of area on American linguistics", Word 1: 54 — 58. Wagner, K. 1927 Deutsche Sprachlandschaften (Deutsche Dialektgeographie 23). Marburg: Elwert. Wagner, Heinrich 1958 — Linguistic atlas and survey of Irish dialects. Dublin: Institute for Advanced Studies. Wakelin, Martyn 1973 English dialects. London: Athlone. Wolff, H. 1959 "Subsystem typologies and area linguistics", Anthropological Linguistics 1, No. 7: 1 - 8 8 . Wood, G. 1961 Word distribution in the interior South (Publications of the American Dialect Society 35). University, Ala.: University of Alabama Press. Wrede, Ferdinand 1926 — 1956 Deutscher Sprachatlas. Marburg: Elwert. 1963 Kleine Schriften. Marburg: Elwert. Zylko, F., et al. 1948 Prohrama dlja zbyrannja materialiv do dialektolohicnoho atlasa ukrajins'koji movy. Kiev: Institute of Linguistics.

Social stratification of language John Baugh

1. Introduction One of the major themes that is found throughout Labov's work considers the historical implications that can be drawn from synchronic linguistic research, bridging the synchronic/diachronic distinction (see Harms, this volume). I will review some of the major trends in Labov's quantitative studies, along with tentative suggestions for future analyses. I recognize that there is considerable disagreement among linguists regarding historical procedures; however, we will not review these debates again here. With future goals in mind we shall consider the strengths and limitations of quantitative studies in linguistic variation, along with their potential diachronic utility. A glimpse of Labov's position appears in the opening remarks of his Presidential Address to the Linguistic Society of America: If such a thing as an advanced study of human behavior exists, linguistics may claim to hold that position — perhaps for no other reason than its ability to state precisely the issues being argued ... The orientation toward linguistic research that I would like to demonstrate here begins with a somewhat different perspective. It is motivated by a considerable respect for the intelligence of our predecessors, and for the evidence that led them to their conclusions (1981: 2 6 7 - 2 6 8 ) .

Labov's research, leading to this position, grew out of an interdisciplinary background, including training as a professional chemist, and that "scientific" background seems to have influenced his work. Throughout his linguistic career Labov has attempted to build upon time-tested principles in linguistics as well as the sister (social) sciences. While this paper concentrates on Labov's research, it would be wrong to imply that he is somehow solely or primarily responsible for these

274

John Baugh

research trends. As he states, his studies have utilized the previous work of others. Lehmann (1981) outlines the broader relationship more completely in his article titled "Historical linguistics and sociolinguistics". Labov's earliest historical insights can be found in his 1963 paper on "The social motivation of a sound change". Without going into great detail here, this research examined vocalic variation among natives of Martha's Vineyard, and found that personal loyalties and aspirations played a major role in linguistic production. For our purpose here, the interdisciplinary foundations of the Martha's Vineyard research set the stage for his work that followed, and, in particular, his continued objective to overcome the perceived distinction between synchronic and diachronic research.

2. Theoretical foundations The early work in American dialectology and the principles of sound change embodied in Grimm's Law and Verner's Law were instrumental in Labov's early formulations. More recent observations, like those of Chomsky and Halle (1968), have influenced his research as well. The major work that followed Martha's Vineyard was Labov's classic study of The social stratification of English in New York City (1966). Again, an interdisciplinary perspective was employed. A great deal has already been said on many occasions regarding the relative feasibility of various interdisciplinary approaches, and I have little to add to this discussion. From a theoretical point of view, the goals of sound interdisciplinary research have always been shared by social scientists, and it is this positive outlook that I would prefer to stress. Although the New York City research was not initiated as a historical project, it has several implications, many of which were of direct concern to minority populations. Educators began to call on linguists, and Labov in particular, to shed some additional light on the linguistic problems of minority speakers. One result of this effort was the collaborative work by Labov, Cohen, Robins, and Lewis (1968), A study of the non-standard English of Negro and Puerto Rican speakers in New York City. It was slightly prior to that time, during the latter part of 1967, that Weinreich, Labov, and Herzog produced a paper that was specifically concerned with historical issues, namely, "Empirical foundations for a theory of language change". This paper contains

Social stratification

of language

275

some of the major theoretical statements regarding the utility of the study of linguistic variables. Weinreich states his position as follows: We will, finally, suggest that a model of language which accommodates the facts of variable usage and its social and stylistic determinants not only leads to more adequate descriptions of linguistic competence, but also naturally yields a theory of language change that bypasses the fruitless paradoxes with which historical linguistics has been struggling for over half a century ... We will argue that the generative model for the description of language as a homogeneous object is itself needlessly unrealistic and represents a backward step from structural theories capable of accommodating the facts of orderly heterogeneity (1968: 99 — 100).

As indicated previously, the debate over the best historical procedures is far from being resolved, and the available literature suggests that several productive approaches have been developed through both inductive and deductive procedures. In the best of all circumstances we should maintain flexible theories which are capable of accommodating research advances in linguistics and, perhaps, other fields. At this point it would be helpful to review the formal notation that Labov developed to handle quantitative studies of variation.

3. Analytic advances Labov (1969) introduced variable rules in his analysis of copula variation in Black English. Since that time the concept has been refined and developed in other studies (Cedergren — Sankoff 1974), and has been most successful with analyses of phonology and morphophonemics. In order to fully appreciate the necessity of variable rules, we must first recognize Labov's desire to complement the rule-writing scheme employed by Chomsky and Halle (1968). As most readers know, their original formulations accounted for two cases, categorical rules and optional rules: A. Categorical rules Β. Optional rules

X —> [Υ] / A X —• (Υ) / A

Β Β

Labov introduced the following variable rule to account for instances of rank-ordered variation in Black English copula variation, and the notation has been adopted and refined by others (see Cedergren - Sankoff 1974):

276

John Bough

C. Variable rules D. Detailed variable rules

X —• / A

/ [fea A ] \ / [fea Β] \ J X -> / ( [fea C] ) (

Β

[fea P] Tea Q] )

[fea V]

For a more thorough discussion of the relationship between synchronic rules and their diachronic implications, see Harms in this volume. As far as Labov's research is concerned, the goal has been to provide a formal notation that has the capability of accounting for instances of systematic variation in languages. Moreover, as a heuristic, variable rules have changed considerably since their original formulation. While variable rules are not obligated to examine the relationship between social and linguistic forces on language, they have that potential, and it is about the power of the notation that some linguists have expressed great concern. These arguments are well documented in the works of Bickerton (1971), Wolfson (1976), Hupsek (1986), Kay and McDaniel (1979), and Romaine (1981). If a general criticism can be found it lies in the concern for how the various social and linguistic factors are selected, and, once selected, how they parse with the conceptual (i. e., cognitive) points of reference among individuals within a speech community. Most of these criticisms have been addressed elsewhere (Labov 1972 and Sankoff — Labov 1979). For the purpose at hand, however, we must appreciate that the selection of factors, both linguistic and social, requires extreme care. Previous studies have utilized various combinations of factors to determine which specific factor groupings provide the best statistical fit, but statistical evidence alone is not adequate to our task. Elsewhere (Baugh 1980) I have discussed this problem at greater length; as new findings are developed we will be better equipped to select adequate factors with greater reliability. In the meantime there is great reconstructive potential for synchronic analyses.

4. Historical insights: Black English copula variation One of the clearest examples of how synchronic studies can provide historical insights can be found in the area of Black English copula variation. Readers should keep in mind that the original version of

Social stratification

1.00

of language

111

Jamaica

.75 . «-s . ^ \

.50 **

^

Ν

y

Ν X

.25

•

0.00 NP

Loc.

Adj

1.00

Vb + ing

gon(na)

Vb + ing

gon(na)

Gullah

.75 .50 .25 0.00 NP

Adj

Loc.

Harlem (The Cobras)

1.000 •

.750

/ /

•/

.500

~

/

~ ~

/

.250

/ /

t/

/ 0.000 f

NP

Loc.

_ Adj

Vb -f ing

gon(na)

det. # NP

Grammatical hierarchy based on % of deletion for Jamaican and Gullah Jamaican

%

Adj. gonna NP _V Loc.

66 32 22 17 17

Figure 1.

Gullah gonna Adj. V Loc. NP

% 88 52 52 22 11

Copula absence in three Black English communities.

the variable rule was developed to examine this problem. With the advantage of new computer-aided calculations, the original Harlem data from Labov's New York City research were compared to similar patterns of copula absence in Gullah and Jamaican Creole English (see Baugh 1980). The early positions were perceived as being in opposition, as Fasold observed:

278

John Baugh

Proponents of the Creole history of Vernacular Black English often disagree with linguists who have studied the dialect synchronically over the degree to which decreolization has progressed ... Accepting Labov's analysis of the modern dialect is not tantamount to a denial of the Creole origin hypothesis, but simply to recognize that Vernacular Black English has reached a late post-Creole stage (1976: 79).

My own research in this area reinforces Fasold's position; the data that are illustrated in Figure 1 (above) compare copula absence in three Black English communities. Again, recalling the oral traditions of Black America, along with the paucity of reliable linguistic documentation, it was evident that new diachronic procedures would be needed to resolve the opposing historical hypotheses. Figure 1 reveals dimensions of this diachronic potential. In this instance the purely linguistic factors suggest strong parallels. The environments listed at the bottom of each graph represent the grammatical features that immediately follow the verb "to be", or, more specifically, where the verb "to be" is absent in these Black English dialects. Through a careful analysis of the synchronic variation in three distant locations, that is, including geographic and diachronic distance, we can see at a glance that the patterns are quite similar. This similarity in turn indicates a common ancestry. Returning once again, to the broader historical theme of this paper, the preceding evidence has been restricted to a comparison of purely linguistic factors. Comparable social data are not available for all three communities; nevertheless, the pattern illustrated in Figure 1 is clear. This kind of comparative analysis, between one group and/or region and others, can be applied to any spoken (or signed) language. In other words, the division between oral and written cultures is diminished when synchronic data are the bases of comparison. Moreover, with the improved technical quality of tape-recorded speech, today's linguist is in an excellent position to provide historical data for tomorrow, that is, data which are far more precise than the written documents that we now use for reconstruction.

5. Conclusions and future prospects In the light of the heuristic nature of variable rules, along with the interdisciplinary foundation of Labov's work, it would be difficult to draw strong conclusions at this stage of development. It would be more productive to look toward the future, and to determine the

Social stratification of language

279

potential utility of quantitative procedures. First, readers must appreciate that variable rules may or may not be comparable to the generative rules developed by Chomsky (again, see Harms in this volume). They nevertheless adopt a similar notation; this is not a limitation. Rousseau and Sankoff (1978) have developed programs specifically for the purpose of processing large quantities of social and linguistic data. Moreover their programs provide for a series of calculations that will reveal the combination of factors that yields the most reliable statistical fit. Sankoff and Labov (1979) outline the technical dimensions of this orientation, and readers are directed to their discussion for a more complete account of the potential and limitations of these procedures. In the meantime we can proceed with cautious optimism, because we have made significant discoveries through quantitative techniques.

References Baugh, John 1980 "A re-examination of the Black English copula", in: Labov (ed.). Bickerton, Derek 1971 "Inherent variability and variable rules", Foundations of Language 7:457 — 492. Cedergren, H. C. — David Sankoff 1974 "Variable rules: Performance as a statistical reflection of competence", Language 50: 333 — 355. Chomsky, Noam —Morris Halle 1968 The sound pattern of English. New York: Harper and Row. Fasold, Ralph 1976 "One hundred years from syntax to phonology", in: Steever—Walker — Mufwene (eds.), 7 9 - 8 7 . Harms, Robert T. 1990 "Synchronic rules and diachronic 'laws'", this volume. Hupsek, M. 1986 "Linguistic variation, context, and meaning: A case of 'ing/in' variation in North American workers' speech", Language in Society 15: 149 — 164. Kay, Paul —C. K. McDaniel 1979 "On the logic of variable rules", Language in Society 8: 151 — 188. Labov, William 1963 "The social motivation of a sound change", Word 19: 273 — 309. 1966 The social stratification of English in New York City. Washington, D. C.: Center for Applied Linguistics. 1969 "Contraction, deletion, and inherent variability of the English copula", Language 45: 715 — 762.

280

John Baugh

1972 Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press. 1981 "Resolving the Neogrammarian controversy", Language 57: 267 — 308. Labov, William —Paul Cohen — Clarence Robins —John Lewis 1968 A study of the non-standard English of Negro and Puerto Rican speakers in New York City. Washington, D. C.: USOE Final Report, Research Project 3, 288. Labov, William (ed.) 1980 Locating language in time and space. New York: Academic Press. Lehmann, Winfred P. 1981 "Historical linguistics and sociolinguistics", International Journal of the Sociology of Language 31: 11—27. Lehmann, Winfred P. —Yakov Malkiel (eds.) 1968 Directions for historical linguistics: A symposium. Austin: University of Texas Press. Romaine, Suzanne 1981 "The status of variable rules in sociolinguistic theory", Journal of Linguistics 17: 9 3 - 1 1 9 . Rousseau, Pascale — David Sankoff 1978 "Advances in variable rule methodology", in: Sankoff (ed.). Sankoff, David—William Labov 1979 "On the use of variable rules", Language in Society 8: 189-223. Sankoff, David (ed.) 1978 Linguistic variation: Models and methods. New York: Academic Press. Steever, S. —Carol Walker—Salikoko Mufwene (eds.) 1976 Papers from the Parasession on Diachronic Syntax. Chicago Linguistic Society 12th Regional Meeting, 22 April 1976. Chicago: Chicago Linguistic Society. Weinreich, Uriel—William Labov —Marvin I. Herzog 1968 "Empirical foundations for a theory of language change", in: Lehmann — Malkiel (eds.), 9 5 - 1 8 8 . Wolfson, Nessa 1976 "Speech events and natural speech", Language in Society 5: 189 — 209.

Contact and interference Franklin

Southworth

Introduction The field of language contact and interference was critically reviewed by Uriel Weinreich in his brilliant and seminal book, Languages in contact (Weinreich 1974 [originally published 1953]). Though many studies of linguistic interference have appeared since that time, relatively few of these have addressed, in any significant way, any of the important theoretical issues raised by Weinreich. For this reason, the following discussion takes primarily the form of a review and updating of Weinreich's work. It is assumed that materials published prior to 1953 have been adequately covered by Weinreich, whose bibliography contains 658 entries and is accompanied by a subject index. Of works published since that time, the focus is on those which bring data to bear on important issues, and no attempt will be made to cover all the primary treatments of interference between two languages. We are concerned here with change resulting from language contact, i. e., the effects on a language of the transfer of elements or features (phonic, grammatical, or lexical) from another language with which it is in contact. Weinreich's definition of language contact ("... languages will be said to be IN CONTACT if they are used alternately by the same persons" [1974: 1]) is probably used implicitly by most workers in this field, though contact may also be said to take place even for individuals with very low levels of competence in the secondary language (see, e. g., Diebold 1961). From the point of view of change, the focus in studies of contact is generally on sizable groups of speakers rather than on individuals. Thus, for example, studies of linguistic interference in the Balkans range from investigations of area-wide

282

Franklin

Southworth

typological similarities (for example, Havranek 1933) or features shared among several languages (for example, Schaller 1975: 134 — 190; Sandfeld 1930: 163-209), to studies of the effects of dominant languages on the speech of small immigrant groups (for example, Coteanu 1957; Recatas 1934).1 Weinreich's definition of interference ("... the rearrangement of patterns that result from the introduction of foreign elements into the more highly structured domains of language, such as the bulk of the phonemic system, a large part of the morphology and syntax, and some areas of the vocabulary ..." [1974: 1]) is probably also implicit in most work on the subject — especially given its vagueness with regard to what is excluded. Concerning the included domains, Weinreich says: "It would be an oversimplification to speak here of borrowing, or mere additions to an inventory" (1974: 1). In studies of interference among Balkan languages, it is usual to make a sharp distinction between lexical and structural borrowing, and in discussing the former to provide separate discussions of nouns, verbs, function words, and affixes: Sandfeld (1930), for example, singles out adverbs, prepositions, and conjunctions of Turkish origin for a separate discussion. Students of linguistic interference in South Asia (the term "convergence" is the usual one here) consider lexical borrowing to be a quite separate phenomenon from convergence proper — though the borrowing of affixes, function words, and basic verbs is often considered along with grammatical convergence (see, for example, Emeneau 1971). Similarly Sherzer (1976), dealing with areal linguistic studies in North America, considers only shared structural features (under the headings of phonology and morphology).

Phonic interference Weinreich's division of phonic interference into four types (underdifferentiation of phonemes, over-differentiation of phonemes, reinterpretation of distinctions, and phone substitution [1974: 18 — 19]) is well-known and has been widely used by others. Redefinitions of similar types of changes in terms of generative phonology have been undertaken (see, for example, King 1969: 39 — 63), but perhaps not with specific reference to interference. Thus, for example, a case reported by Emeneau (1956) involving the diffusion of a sub-phonemic distinction between alveolar and palatal affricates across major lan-

Contact and interference

283

guage boundaries could perhaps be logically classified according to Weinreich's scheme as a case of (conditioned) phone substitution, whereas in a generative phonological framework it would be considered as a case of the addition of a rule (by transfer) — in this case, a rule which converts alveolar affricates to palatals before non-low front vowels. Some of the more important discussions of phonic interference have little direct bearing on the understanding of change, since they are presented in terms of typological or area-wide similarities without any attempt to understand them dynamically. This is largely true, for example, for Havränek (1933) and Schaller (1975: 134-190) for the Balkan languages, it is true for Sandfeld (1930) (see, for example, 102 — 104), for Ramanujan and Masica (1969) for South Asian languages, and largely true for the materials presented in Sherzer (1976), though here in a number of cases it has been possible to make historical inferences. By contrast, most of the work on phonic interference in South Asia has had a strong diachronic bias, and has been moreover specifically concerned with the question of substratum influence, particularly the hypothesis of Dravidian (or other) influence on early Indo-Aryan. The literature on the origin of the retroflex-dental contrast in Indo-Aryan provides examples of the types of argumentation used: for example, see Kuiper (1967) for a statement of the problem and references to earlier work; Southworth (1974) for a discussion of the relevance of frequency and geographical distribution and Hock (1975) for an attempt to explain the phenomenon in terms of purely internal change. Regarding the predictability of phonic interference, Weinreich refers to the work of Reed, Lado, and Shen (1948) in predicting interference in various languages by English-speaking learners. He raises the problem of the different renditions of English /Θ/ and /δ/ by French speakers (who generally identify them with their /s/ and jxj respectively) and by Russian speakers (who tend to reproduce them with their jtj and / d/), without proposing any definite solution, though he makes the statement that "... the possibilities of structural explanation are most promising" (1974: 20). Further on (1974: 25) he quotes with approval Roman Jakobson's statement that a language "accepts foreign structural elements only when they correspond to its tendencies of development" (Jakobson 1938: 54), but does not tell us how to apply this notion without circularity, or what types of evidence might be mustered to test it.

284

Franklin

Southworth

In discussing the phonic treatment of transferred morphemes, Weinreich (1974: 26 — 28) mentions a number of factors which are presumed to determine the extent to which a form is integrated or adapted to the phonological patterns of the borrowing language: level of the individual speaker's competence in the source language; attitudes toward the source language; the extent to which borrowing is influenced by the written forms of words; and also folk-etymological or analogical factors (cf. for example American Italian /broko'lino/ 'Brooklyn'). Kelkar (1957) has noted that in order to explain the phonic transformations of English words in "Marathi English", it is not sufficient to know the level of the speaker's competence in English, since local expectations also play a role — i. e., the speaker whose pronunciation departs too much from the local norm in the direction of the native English or American norm will be criticized as a showoff, or simply as aberrant. Holden (1976) discusses the differential rates of assimilation of various features in borrowings from French into Russian. Sa'id (1967: 39 — 57) treats the phonology of borrowings into modern standard Arabic. Newton (1963) discusses the interaction of phonological shape, gender, and other factors in the assimilation of loanwords. Weinreich raises a point to which little attention has been paid, either before or since the appearance of Languages in contact, namely, that different constraints apply to the phonic treatment of words depending on whether speakers are performing in a secondary language or are introducing words from their secondary language into their primary language: ... a bilingual's connected utterance in language S [i.e., a secondary language], even if imperfect, must nevertheless approximate the phonemics of S sufficiently to be intelligible to its unilingual hearers. On the contrary, the use of a word borrowed from S in a Ρ [i. e., primary language] — utterance is not inhibited by the need to conform to an extraneous phonemic norm; the mechanisms of interference therefore affect individual loanwords with particular force (Weinreich 1974: 28).

Grammatical interference Weinreich notes that "the problem of grammatical interference [is] currently among the most debated questions of general linguistics", and illustrates the problem by juxtaposing quotes from Meillet and Sapir on the one hand, and Hugo Schuchardt on the other, which take

Contact and interference

285

completely opposing views regarding the extent to which grammatical interference is possible (Weinreich 1974: 29; on this point see also Emeneau 1956; Sherzer 1976: 1—6; Fought 1982). Weinreich appears to believe that this apparently fundamental disagreement among reputed and responsible scholars is "due to the lack of agreement between them on fundamental terms and concepts" (1974: 29). In order to avoid falling into the same trap himself, Weinreich proposes first to make an absolute difference between morphemes (segmental or suprasegmental) and grammatical relations (including order, agreement, dependence and similar relations, and modulations of stress and pitch); secondly, to introduce two relative distinctions, namely that of more or less obligatory categories, and that of greater of less "syntagmatic boundedness of morphemes used to express categories" (1974: 29 — 30). This approach, then, recognizes the following types of grammatical interference: 1) the use of a morpheme from language A (presumably with its grammatical relations as defined in A, mutatis mutandis); 2) the application of a grammatical relation of language A to a Bmorpheme in Β speech (what might nowadays be called the transfer of a rule, rule constraint, or the like); 3) grammatical calquing, i. e., the change in the functions of a B-morpheme (in Β speech) on the model of A. Weinreich suggests that "it may be possible to range the morpheme classes of a language in a continuous series from the most structurally integrated inflectional ending, through ... prepositions, articles ... to full-fledged words ... and on to independent adverbs and ... unintegrated interjections" (1974: 35). He criticizes Haugen (1950: 224) for proposing such a "scale of adoptability" without sufficiently emphasizing its hypothetical nature. A similar criticism could be made of much work on linguistic interference in the Balkans and in South Asia; in fact, in many such cases, the assumption of the existence of such a scale is implicit, not explicit as in Haugen's case. Weinreich gives a rather lengthy discussion of the questions of the transferability of "highly bound morphemes", making use of the data available to him, which was almost exclusively drawn from European languages (1974: 31—37). Apart from the transfer of affixes which is achieved secondarily, i.e., by resegmentation after borrowing (e.g., in the case of English -ize or -ation), Weinreich finds that there are only a few reliably reported cases of the transfer of bound morphemes such as inflectional affixes, and that in all the known cases the transfer has taken place between highly congruent systems (1974: 32 — 33). Fur-

286

Franklin

Southworth

thermore, this seems to be a highly ephemeral phenomenon, which rarely if at all gets beyond informal colloquial speech to find a place in standard or formal usage (1974: 33). Cases reported since Weinreich wrote on the subject probably support his statement for the most part, but not completely. Emeneau (1962 b, Chapter 4) points, for example, to a suffix in Brahui (a Dravidian language) apparently transferred from Makran Baluchi (an Iranian language) which fills a function not found in other Dravidian languages; and notes the pronominal suffixes found among Indo-Aryan and Iranian languages in a contiguous region including Pakistan, Afghanistan, and eastern Iran (Emeneau 1965). Both these cases would seem to contradict Weinreich's requirement regarding congruence of systems. The phenomena treated by a number of writers under the heading of "structural convergence" among South Asian languages typically involve the transfer of what Weinreich called "grammatical relations", independently of lexical borrowing. (Lexical borrowing is also frequent among these languages, but sometimes predominantly in the opposite direction from that of structural convergence.) For a discussion of this, see Emeneau (1954, 1956, 1962 a, 1971, 1974); Pandit (1972); Masica (1976); Southworth - Apte (1974); Shapiro - Schiffman (1981); and Nadkarni (1975). On grammatical interference in the Balkans, the most important earlier work is Sandfeld (1930: 163 — 210, with many references to previous work), which ascribes many of the common features to Greek origin. Schaller (1975) is a valuable recent summary, though it provides little new data. Schaller (1977) is an extremely useful bibliography of Balkan linguistics, topically organized and including a combined subject and author index. Sherzer (1976) provides an overview of studies of grammatical convergence among the native languages of North America, with a substantial bibliography (over 500 titles). Hymes (1971: 65 — 90) discusses the transfer of items from various components of a language, and the relationships between such transfers and the processes of creolization and pidginization. Rocencvejg, in his brief monograph (1976), suggests a distinction between "direct interference" (the wholesale transfer of idiosyncratic properties from one language to another) and "indirect interference" (the substitution of shared properties for idiosyncratic properties).

Contact and interference

287

Lexical interference Weinreich's discussion of lexical interference distinguishes a) simple words (Weinreich 1974: 47 — 50), treated under the headings of outright transfer and extensions of native words on the basis of foreign models, and b) compound words and phrases (1974: 50 — 53). Under b) the following well-known types are distinguished: 1) transfer of analyzed compounds (Florida Spanish objetores concientes 'conscientious objectors'); 2) reproduction in terms of equivalent native words, including loan translations proper (American Portuguese estar direito 'to be right'), loan renditions (Vaterland, Halbinsel, Wolkenkratzer), and loan creations (Yiddish mitkind 'sibling'); 3) hybrid compounds (Florida Spanish home plato 'home plate'). (See Haugen 1956: 55 — 59 for a parallel discussion.) Casagrande, in his detailed study of Comanche linguistic acculturation (1954—1955), distinguishes between "primary" and "secondary" linguistic accommodation of new cultural items. The former includes extension of old words and new coinages (Weinreich's "loan creations") — these being the mechanisms requiring the least knowledge of the other language (in this case English, and secondarily Spanish) — and the latter includes outright transfer or loan translation which require greater bilingual competence. In the case of Comanche the mechanisms of primary accommodation account for over 90% of the acculturation vocabulary. Heath (1978, 1981) presents a case involving very high rates of lexical diffusion (up to 78% in some semantic areas), though the necessary historical work has not yet been done which would determine the sources of items, and thus it is not clear what part of these percentages may represent original cognates. Reviewing other cases of intensive lexical diffusion (Scotton — Okeju 1973; Newman 1974; Haugen 1953), Heath concludes that the Australian cases studied by him show the highest levels of lexical diffusion reported in the literature.

Factors determining interference Weinreich's discussion of factors affecting interference provides a 2by-2 grid, opposing "stimuli" (i. e., predisposing factors) to "resistance factors" on the one hand, and structural (i.e., internal linguistic) factors to non-structural (i.e., social or psychological) factors on the

288

Franklin

Southworth

other (1974: 2 2 - 2 5 , 4 3 - 4 4 , 5 6 - 6 1 , 6 3 - 6 7 ; especially the table on 64 — 65). Apart from those general factors favoring interference (structural differences, positive attitudes toward the source language, individual propensity for speech mixture) or opposing it (maintenance of systemic stability and intelligibility, mother-tongue loyalty, puristic attitudes), Weinreich lists specific factors in each of the four categories for each of the main types of interference mentioned. He also treats the "total amount of interference" as a separate category. This scheme suggests the need for testing the significance of each of these factors relative to the others — a research need which has hardly been touched up to the present day, with the result that we still have nothing but opinions and impressions to work with in this field. A few studies, however, may be mentioned. Heath (1978) points out certain structural factors which affect the borrowability of lexical items: e. g., a class of verbs which frequently occurs "naked" (without affixes) in the source language is subject to more frequent borrowing than another class which does not so occur. 2 Rayfield (1970) puts forward the view that structural differences between the borrowing and source languages are relatively unimportant, alongside cultural and individual factors. Scotton and Okeju (1973) claim that the type of cultural contact is the primary variable in determining the amount and type of borrowing. A developing literature on structural constraints in code-switching may ultimately make important contributions to the question: see, for example, Lipski (1977); McClure (1977); Timm (1977); Valdes Fallis (1976); Zentella (n.d.); Southworth (1980). Among the non-structural factors mentioned by Weinreich as either favoring or disfavoring interference, attitudinal factors (social value of the source or recipient language, language loyalty, puristic attitudes, intolerance of recognizable loanwords) are prominent. A few recent studies contain conclusions relating to the presumed importance of these or other factors in promoting or inhibiting interference, but in the absence of cross-cultural or cross-linguistic studies, none of these can be considered conclusive. Heath mentions the following among the factors he believes to be responsible for the high levels of lexical diffusion in Arnhem Land (northern Australia): frequent intermarriage with high levels of intrafamilial bilingualism, small size of groups, and absence of prestige differentiation among the languages involved (1981: 363). Newman (1974) presents a case in which intermarriage seems not to be a factor,

Contact and interference

289

since the group in question (the Bella Coola of the northwest coast of North America) reportedly intermarried frequently with two different neighboring groups, but borrowed vocabulary extensively only from one of them; Newman concludes that attitudes toward the cultures of the two groups were the determining factor. The cases of interference reported from South Asia would suggest that intermarriage cannot be regarded as a necessary explanation in any general sense, since many of the cases of extreme interference reported there (including both grammatical convergence and high levels of lexical borrowing) occur in contexts involving little intermarriage (see, for example, the case of Saurashtri in Pandit 1972). Sherzer and Bauman (1972) suggest that lexical borrowing may occur in any type of contact situation, and in fact some types of lexical similarities may result from shared ecological features without implying any direct contact at all. Structural diffusion, however, requires very special conditions of contact; from the perspective of native American studies, they suggest that this type of interference takes place only under conditions of "intimate face-to-face contact between whole communities" (1972: 144). In Sherzer's survey of linguistic areas in native North America, he emphasizes repeatedly that spatial contact is not sufficient to lead to this type of linguistic interference: I stress that it is not mere contact in space which leads to diffusion and therefore to the development of linguistic areas. Rather, internal linguistic factors together with certain socio-cultural conditions (more precisely, communicative conditions: type of social organization; density of population; degree of trade, intermarriage, and multilingualism; attitudes toward language; etc.) can encourage or inhibit the diffusion of linguistic traits (1976: 255; italics in original).

While many might agree with this statement, it is clear that neither "degree of multilingualism" nor "attitudes toward language" can be viewed as unanalyzable primes. Thus, when Sherzer (1976: 244) invokes language attitudes as the primary factor responsible for the failure of the Southwest culture area to develop into a linguistic area (in spite of the presence of predisposing factors such as high density of population, long period of contact, and relative cultural homogeneity), we must ask what other social factors lie behind these attitudes.

Value of contemporary studies One may briefly mention the potential contribution which can be made by detailed sociolinguistic studies to the testing of hypotheses regarding the factors affecting linguistic interference. Payne (1980), in a detailed

290

Franklin Southworth

study of the acquisition of Philadelphia phonological patterns by children from other areas, indicates that both age and structural differences have to be taken into account in attempting to predict the extent to which the new pattern will be acquired. Baugh (1980) reviews the evidence for the two theories of the origin of copula deletion in the black English vernacular (BEV), and concludes that "the synchronic status of BEV copula has been influenced by West Indian Creoles, as well as by SE [Standard English]" (1980: 104). Klein (1980) discusses the circumstances under which Spanish-English bilinguals in New York interpret the simple present and continuous present tenses according to English norms. Blom and Gumperz (1972) present an analysis of the transfer of features from Bokmäl (a variety of standard Norwegian) into a local dialect, with a detailed discussion of the social conditions under which this takes place. Gumperz and Wilson (1971) discuss grammatical interference among three local languages in a border region of India, indicating the social factors and cultural attitudes which lie behind this behavior.

Notes 1. See Ramanujan — Masica (1969) and Southworth (1974) for comparable studies in South Asia. 2. See Lindenfeld (1971) and Pfaff (1979) for additional discussions of structural impediments to interference.

References Baugh, John 1980 "A reexamination of the Black English copula", in: Labov (ed.), 83 — 106. Blom, Jan-Petter — John J. Gumperz 1972 "Social meaning in linguistic structures: code-switching in Norway", in: Gumperz — Hymes (eds.), 407 — 434. Casagrande, Joseph B. 1954; 1955 "Comanche linguistic acculturation", IJAL 20: 1 4 0 - 1 5 1 , 2 1 7 - 2 3 7 ; 21: 8-25. Coteanu, Jon 1957 "A propos des langues mixtes (sur l'istro-roumain)", in: Melanges linguistiques publies ά l'occasion du VIHe Congres International des Linguistes a Oslo, du 5 au 9 Aoüt 1957. Bucarest: Editions de l'Academie Populaire Roumaine, 1 2 9 - 1 4 8 .

Contact and interference

291

Diebold, A. Richard 1961 "Incipient bilingualism", Language 37: 96 — 112. Dil, Anwar S. (ed.) 1980 Language and linguistic area: Essays by Murray B. Emeneau. Selected and introduced by Anwar S. Dil. Stanford California: Stanford University Press. Emeneau, Murray B. 1954 "Linguistic prehistory of India", Proceedings of the American Philosophical Society 98: 2 8 2 - 2 9 2 (Reprinted in Dil 1980: 85-104). 1956 "India as a linguistic area", Language 32: 3 — 16 (Reprinted in Dil 1980: 105-125). 1962 a "Bilingualism and structural borrowing", Proceedings of the American Philosophical Society 106: 4 3 0 - 4 4 2 (Reprinted in: Anwar S. Dil (ed.), 38-65). 1962 b Brahui and Dravidian comparative grammar (University of California publications in linguistics, Vol. 27). Berkeley and Los Angeles: University of California Press. (Chapter 4, "Iranian and Indo-Aryan influence on Brahui", 4 7 - 6 1 , reprinted in revised form in Dil 1980: 333-349.) 1965 India and historical grammar. (Annamalai University, Department of Linguistics Publication No. 5) Annamalainagar, India: Annamalai University (pp. 1—24 reprinted as "Diffusion and evolution in comparative linguistics" in Dil 1980: 66 — 84; pp. 25 — 75 reprinted as "India and linguistic areas" in Dil 1980: 126-166). 1971 "Dravidian and Indo-Aryan: The Indian linguistic area", in: Andree F. Sjoberg (ed.), 3 3 - 6 8 (reprinted in Dil 1980: 167-196). 1974 "The Indian linguistic area revisited", in: Southworth — Apte 1974: 92 — 134 (Reprinted in Dil 1980: 197-249). Fought, John 1982 "The reinvention of Hugo Schuchardt" (review article), Language in Society 11: 4 1 9 - 4 3 6 . Gumperz, John J. —Robert Wilson 1971 "Convergence and creolization: A case from the Indo-Aryan/Dravidian border", in: Hymes (ed.), 151-168. Haugen, Einar 1950 "The analysis of linguistic borrowing", Language 26: 210 — 231. 1953 The Norwegian language in America. 2 vols. Philadelphia: University of Pennsylvania Press. 1956 Bilingualism in the Americas: A bibliography and research guide (Publications of the American Dialect Society 26). University, Ala.: University of Alabama Press. Havranek, Bohuslav 1933 "Zur phonologischen Geographie: das Vokalsystem des balkanischen Sprachbundes", Archives neerlandaises de phonetique experimentale 8/9: 119-125. Heath, Jeffrey 1978 Linguistic diffusion in Arnhem Land. Canberra: AUS. 1981 "A case of intensive lexical diffusion", Language 57: 335 — 367.

292

Franklin Southworth

Hock, Hans Heinrich 1975 Substratum influence on (Rig Vedic) Sanskrit? (University of Illinois Studies in the Linguistic Sciences No. 2). Urbana: Dept. of Linguistics, University of Illinois. Holden, Cyril 1976 "Assimilation rates of borrowing and phonological productivity", Language 52: 131-147. Huttar, George L. 1965 "Sources of creole semantic structures", Language 51: 684—695. Hymes, Dell H. (ed.) 1971 Pidginization and creolization of languages. Cambridge: Cambridge University Press. Jakobson, Roman 1938 "Sur la theorie des affinites phonologiques des langues", in: Actes of the 4th International Congress of Linguists, 27 Aug. —1 Sept. 1936. Copenhague: Eimar Munksgaard: 48 — 59. Kelkar, Ashok R. 1957 "Marathi English: A study in foreign accent", Word 13: 268-282. Keller, Gary V. et al. (eds.) 1976 Bilingualism in the bicentennial year and beyond. Jamaica New York: Department of Foreign Languages, York College, CUNY. King, Robert D. 1969 Historical linguistics and generative grammar. Englewood Cliffs, N.J.: Prentice-Hall, Inc. Klein, Flora 1980 "A quantitative study of syntactic and pragmatic indications of change in the Spanish of bilinguals in the U.S.", in: Labov (ed.), 69 — 82. Kuiper, F. B. J. 1967 "The genesis of a linguistic area", Indo-Iranian Journal 10: 81 — 102 (Reprinted in Southworth-Apte 1974: 135-153). Labov, William (ed.) 1980 Locating language in time and space. New York: Academic Press. Lindenfeld, Jaqueline 1971 "Semantic categorization as a deterrent to grammatical borrowing: A Yaqui example", IJAL 37: 6 - 1 4 . Lipski, John M. 1977 "Code-switching and the problem of bilingual competence", The Fourth Lacus Forum Columbia, S. C.: Hornbeam Press, 263 — 277. Masica, Colin P. 1976 Defining a linguistic area: South Asia. Chicago: Chicago University Press. McClure, Erica 1977 "Aspects of code-switching in the discourse of Mexican-American children", in: Muriel Saville-Troike (ed.), 9 3 - 1 1 5 . Nadkarni, Mangesh V. 1975 "Bilingualism and syntactic change in Konkani", Language 51: 672 — 683.

Contact and interference

293

Newman, Stanley 1974 "Linguistic retention and diffusion in Bella Coola", Language in Society 3: 201-214. Newton, Brian 1963 "Grammatical integration of Italian and Turkish substantives into modern Greek", Word 19: 2 0 - 3 0 . Pandit, P. B. 1972 India as a sociolinguistic area. (Gune Memorial Lectures 1968). Poona: University of Poona. Payne, Arvilla C. 1980 "Factors controlling the acquisition of the Philadelphia dialect by out-ofstate children", in: Labov (ed.), 143-178. Pfaff, Carol 1979 "Constraints on language-mixing", Language 55: 291 —318. Ramanujan, A. K. —Colin P. Masica 1969 "Toward a phonological typology of the Indian linguistic area", in: Thomas A. Sebeok (ed.), 543-577. Rayfield, Joan R. 1970 The languages of a bilingual community (Janua linguarum, series practica, 77). The Hague: Mouton (reviewed by J. Fishman, Language 48 (1972): 969-975). Recatas, B. 1957 L'etat actuel du bilinguisme chez les macedo-roumains du Pinde et le role de la femme dans le langage. Paris: Libraire E. Droz. Reed, David W . - Robert L a d o - Y a o Shen 1948 "The importance of the native language in foreign language learning", Language Learning 1: 17—23. Rocencvejg, V. Ju. 1976 Linguistic interference and convergent change (Janua linguarum, series maior, 99). The Hague: Mouton (English version of the Russian Jazykovye kontakty, lingvisticeskaja problematika, 1972). (Reviewed by Michael S. Flier, in Language 54 (1978): 236). Sa'id, Majed F. 1967 Lexical innovation through borrowing in modern standard Arabic (Princeton Near Eastern Papers 6). Princeton, N. J.: Program in Near Eastern Studies, Princeton University (reviewed by P. Abboud in Language 47 (1979): 229 — 232). Sandfeld, Kristian 1930 Linguistique balkanique: problemes et resultats. Paris: Campion (author's translation of his Balkanfilologien, which appeared in 1926 in Danish). Saville-Troike, Muriel (ed.) 1977 Georgetown University round table on languages and linguistics. Washington, D. C.: Georgetown University Press. Schaller, Helmut Wilhelm 1975 Die Balkansprachen. Heidelberg: Carl Winter. 1977 Bibliographie zur Balkanphilologie. Heidelberg: Carl Winter Universitätsverlag.

294

Franklin Southworth

Scotton, Carol Myers—John Okeju 1973 "Neighbors and lexical borrowings", Language 49: 871—879. Sebeok, Thomas Α. (ed.) 1969 Linguistics in South Asia (Current trends in linguistics 5). The Hague: Mouton. Shapiro, Michael C. —Harold Schiffman 1981 Language and society in South Asia. New Delhi: Motilal Banarssidas. Sherzer, Joel 1976 An areal-typological study of American Indian languages north of Mexico. Amsterdam: North-Holland Publishing Company. Sherzer, Joel —Richard Bauman 1972 "Areal studies and culture history: language as a key to the historical study of culture contact", Southwestern Journal of Anthropology 28: 131 — 152. Sjoberg, Andree F. (ed.) 1971 Symposium on Dravidian civilization. Austin, New York: Jenkins Publishing Company, Pemberton Press. Southworth, Franklin C. 1974 "Linguistic stratigraphy of North India", in: Southworth — Apte (eds.), 201-223. 1980 "Indian bilingualism: some educational and linguistic implications", in: V. Teller —S. J. White (eds.), 121-146. Southworth, Franklin C. —Mahadeo L. Apte (eds.) 1974 "Contact and convergence in South Asian languages", International Journal of Dravidian Linguistics 3: 1 - 1 9 7 (in Part 1), 2 0 1 - 2 4 5 (in Part 2). Teller, V . - S . J. White (eds.) 1980 Studies in child language and multilingualism. New York: The New York Academy of Sciences. Timm, L. A. 1977 "Code-switching in war and peace", The Fourth Lacus Forum: 236 — 249. Columbia S. C.: The Hornbeam Press. Valdes Fallis, Guadalupe 1976 "Social interactions and code-switching patterns: A case study in Spanish/ English alternation", in: Gary V. Keller et al. (eds.), 53 — 85. Weinreich, Uriel 1974 Languages in contact: Findings and problems. The Hague: Mouton (originally published as Number 1 in the series Publications of the Linguistic Circle of New York, New York, 1953). Zawawi, Sharifa M. 1979 Loan words and their effect on the classification of Swahili nominals. Leiden: Brill. Zentella, Ana Celia n. d. Code switching and interactions among Puerto Rican children (Sociolinguistic Working Papers Number 50). Austin, Texas: Southwest Educational Development Laboratory.

III. Types of Language Change

Phonological Change

Phonetic, phonemic, and phonotactic change Elmer H. Ant onsen

The structural approach to phonological change focuses on the analysis of changes between chronologically disparate phonological systems, each of which is to be analyzed synchronically in terms of distinctive vs. nondistinctive units, i. e., phonemes vs. allophonic variants, and on the relationship among the phonemes in each system (Hockett 1958).* The intent is to uncover significant changes in the phonological systems of different chronological stages and to make it possible on purely linguistic grounds to determine when a given language has undergone sufficient structural change to warrant the assumption that it is no longer the "same" language. This determination also enables investigators to define various stages on the basis of purely linguistic criteria and brings greater precision to the use and understanding of terms referring to prehistoric stages (Lehmann 1961; Makaev 1961; Antonsen 1965). The comparison of systems is also intended to provide insight into the mechanisms and causes of sound change (Martinet 1955; Moulton 1961). Diachronic phonology is faced with all the problems of synchronic phonological analysis, plus the necessity of recovering the primary data to be analyzed (Penzl 1971). Historical phonologists are therefore directly involved in the theoretical issues confronting all phonologists (Greenberg 1979). The basic tenets of structural phonology are described in Hockett (1958) and Lyons (1968).

* For references, see Elmer H. Antonsen, "Structuralist interpretation", this volume.

298

Elmer Η.

Antonsen

1. Phonemic and phonotactic change On the basis of the comparison of the synchronic systems of distinct chronological stages of various languages, different types of phonemic changes have been catalogued, whereby an older phonemic system undergoes restructuring. Some investigators (Hoenigswald 1960; Moulton 1967) have identified up to ten different types of restructuring, but Penzl (1971) believes these can be reduced to just four principal ones: 1) phoneme-shift, e.g., early O H G j\)j th becomes /d/ d with no coalescence with other phonemes; 2) phoneme-merger, either complete or partial, e. g., PIE jo, a, a/ become P G /a/ in all instances, while P G /a/ in root syllables before /-i, -T, -j/ becomes O H G /e/; 3) phonemesplit, e.g., PIE jsj becomes PG jsj or jzj by Verner's Law; and 4) phoneme-loss, e. g., the loss of M E jkj in Ε knot, knight. According to Penzl, all other types distinguished by other scholars can be subsumed under these four, e. g., PG /sk/ becomes Ε or Ger. jsj as a subtype of phoneme-shift (or phoneme merger?), while pre-OHG j ö j undergoes phoneme-split to become O H G juoj. While it may seem inconsequential to debate a matter such as whether Old High German diphthongization of long mid vowels represents a phoneme-shift (noncoalescence with other phonemes) or a phoneme-split (into two phonemic units), there are deeper theoretical implications, since the relationships within the systems have changed radically, as seen from Table 1. (adapted from Moulton 1961; Antonsen 1972). Table 1.

Changes within pre-Old High German and Old High German

pre-OHG

/!/

/ü/

βr

βr

Μ

OHG

/!/ 2 /u/ /Ü/ 2 Μ /δ/ /Ö/2 /ä/ Μ

The long mid vowels of pre-Old High German (with superscript 1 ) do not correspond to those of Old High German (superscript 2 ), since the former have undergone diphthongization to O H G /ia/ and /üö, uo/, while O H G jej and jö, 5/ result from a conditioned monophthongization of pre-OHG /ai/ and /au/. We are clearly dealing with radically different systems in which the original distinctive features have undergone significant shifts. A somewhat different situation applies to the Germanic consonant shift, traditionally described as PIE /p t k, b d g, b h d h g h / becoming

Phonetic, phonemic, and phonotactic

change

299

pre-Gmc. /f J> h, ρ t k, b d g/. The only restriction on this stage is the rule that in a sequence of pre-Germanic obstruents, the first must be a fricative and the second a stop, so that PIE /sp, st, sk, pt, kt/ etc. become pre-Gmc. /sp, st, sk, ft, ht/ etc., that is, the stops in second position do not shift. Because there is no merger (aside from these clusters) of the original three series, Lehmann (1953, 1961) sees the same phonemic structure in both systems. He believes that Germanic does not become a distinct "language" until after Verner's Law has caused the merger of pre-Gmc. /f J) h/ with pre-Gmc. jb d g/ in certain environments and the Germanic accent shift has removed the original conditioning factor (absence of stress on the preceding vowel).

2. Phonetic change The structuralist preoccupation with determining phonemic contrasts usually has not led to the neglect of subphonemic, or phonetic, change. In the case of reconstructed languages, it is sometimes impossible to determine phonetic values for certain contrasts, e. g., PIE /b h d h g h /. Although the contrast is reconstructable, the available evidence does not permit us to draw any definite conclusions concerning the distinctive features which define that contrast. In more favorable circumstances, however, it is possible to establish the distinctive features of reconstructed phonemes, as well as the (nondistinctive) features of their allophonic variants. In the case of Proto-Germanic, for example, we can reconstruct the vowel system with considerable certainty (van Coetsem 1970; Cercignani 1979). Most investigators agree in establishing contrasts among vowels usually designated as i, e, a, u, i, e1, e2, ä, δ, ü on the basis of Germanic comparative evidence, but internal reconstruction and comparison with Proto-Indo-European reduce the series from ten to just eight phonemes, since ä always derives from /a + n/ before /h/, and /e2/ is clearly a secondary development (van Coetsem 1970). The contrasts among the various phonemes, and therefore the distinctive features (Jakobson —Fant —Halle 1951), can be determined on the basis of their development into the daughter languages (Antonsen 1972). Thus, a, e, δ in post-root syllables cause a lowering of root-syllable /i, u/ to /e, ο/ (so called α-umlaut, never conditioned by the mid vowel /e/), so that these three vowels must have displayed the feature [ + low], and therefore e and δ may be represented by /äe/ and /a/, respectively

300

Elmer Η. Antonsen

(confirmed by the treatment of Lat. /δ/ and /ä/ in early loans as /ü/ and /ä/, respectively). The vowel /a/ must have been central, because it subsequently developed both front and back allophones in root syllables, while /i, e, äe/ and /u, ü, ä/ developed back and front allophones, respectively, which indicates that the feature distinguishing these latter two series was spread vs. rounded. The system of oppositions can therefore be stated in terms of the distinctive features in Table 2 (from Antonsen 1972): Table 2. System of oppositions Μ

Ν

Μ

/Τ/

Μ

—

-

—

—

+

+

+

-

—

—

—

—

+

—

—

—

+

/a/ Long/tense Low Rounded High

/a/

+ + —

+ + +

+

The results of this feature analysis produce a skewed system for early Pro to-Germanic, which may also be represented as in Table 3: Table 3. Early Proto-Germanic vowels

High Mid

Low

Spread

Rounded

Spread

/i/ /e/

/U/

/Τ/

/a/

/ae/

Rounded μ

jaj

The skewed nature of the system is also confirmed by the vocalic symbols found in the Germanic runic alphabet and their acrophonic names (Antonsen 1975). In the case of a language without modern descendants, such as Gothic, the problems of establishing phonological contrasts (as opposed to graphic ones) and their distinctive features are more complex, but by no means insuperable (as might be inferred from Marchand 1970). The comparative method, internal reconstruction, and the evidence of loanwords permit us to reconstruct the phonological system (Antonsen 1972), although we cannot determine which of the oppositions, front/back or spread/rounded, and long/short or tense/lax, were distinctive. One or the other of each set of distinctions was certainly present. Arguments advanced to show that Gothic "disproves" the presence of umlaut allophones in Proto-Germanic are therefore unfounded; Gothic simply provides no evidence concerning

Phonetic, phonemic, and phonotactic change

301

the question, since allophones are not designated. There is no theoretical basis for denying the possible existence of front allophones of Gothic non-front phonemes or of back allophones of non-back ones. Morphological restructuring, such as PG */gastiz/ > Go. gasts does not call for the phonemicization of such allophones, since the change in structure is not a regular phonological one, but a morphological, analogical one (Antonsen 1969). Distinguishing between phonological and morphological restructuring is crucial to the reconstruction of diachronic phonological systems. Dealing with phonological change from allophonic variation to phonemic contrast, Penzl (1965) believes with good reason that all phonological change is traceable to allophonic variation. The question of when and how the phonemicization took place, however, has posed vexing problems. In the case of Germanic umlaut, the reduction or loss of the conditioning factor has been seen as the prime motivator, e. g., OHG mahti > MHG mähte (or earlier in pre-OHG */warmijan-/ > OHG warmen /wärmen/). It has been demonstrated, however, that general restructuring of the system can result in the phonemicization of allophones without the loss of the conditioning factor, as in the case of PG /u/ = [o] before low vowels (Kurylowicz 1952). From Proto-Germanic to Northwest Germanic, the long/tense vowel system illustrated in Table 3. was restructured through shifts of PG /äe/ > /ä/, of /ä/ > /ö/, and through the appearance of a new /e/ (so-called /e2/). The features of this new five-vowel long/tense system (except for length/tenseness itself) were shared in a one-to-one correlation with the vowels of the short/lax system, provided that [o] functioned as the short/lax correlate of long/tense /5/, rather than merely as an allophone of /u/. For that reason, we find NwG horna 'horn' attested with a new phoneme in spite of the retained /-a/, and this same phoneme has even been morphologically introduced into environments where it was originally impossible, e.g., holtijaz 'son of Holt(a-gastiz)' and dohtriz 'daughters' (Antonsen 1975). Similarly, restructuring through the East Scandinavian monophthongization of NwG /ai, au, iu/ > ESc. /äe, q, 0, y/ resulted in the phonemicization of ESc. /ae, q, 0, y/, former umlaut allophones of NwG /a, o, u/ (Antonsen 1978); in Old English, the monophthongization of NwG /iu, eo, au/ > OE /m, Ä, 5/ produced the phonemes /ui, a , a/ from former allophones of NwG /i, e, a/ (Antonsen 1967); and finally, in Gothic, the monophthongization of PG /ai, au/ > Go.

302

Elmer Η.

Antonsen

jab, a/ was responsible for the phonemic status of Go. /ae, ä/, formerly allophones which arose before pre-Go. /r, h/ (D'Alquen 1974). Phonemicization through system-balancing of this type produced the curious situation (for structuralists) whereby the former allophones could be designated by either the old or the new phoneme-designation, e. g., NwG kurne = /korne/ 'corn', but NwG horna = /horna/ 'horn'; OE medu or meodu = /mAdu/ 'mead'. Such spelling variations were the subject of considerable controversy, involving the supposed designation of allophones (for Old English, see Stockwell — Barritt 1951; for Old Norse, Benediktsson 1963). Far from indicating allophonic writing habits (or even dialectal phonological differences), such spellings are indicative of phonemic indeterminacy (Haugen 1969), just as in the case of M H G tac ~ tag. In discussing the same principle to explain Notker's "Anlautgesetz" (law of initial position) in Old High German, which Penzl (1955) had reluctantly considered an example of allophonic spelling, Moulton (1979) pointed out that the phonemic indeterminacy indicated by Notker's spelling variations, e.g., ter bruoder 'the brother (nominative singular)' vs. tes pruoder 'the brother's (genitive singular)', corresponds to a taxonomic phonemic analysis. Had Notker written according to systematic phonemic principles, he would have designated a single underlying form, and we would know nothing about his "Anlautgesetz".

Evidence Robert

Kyes

Prior stages in the evolution of the phonology of a language must be deduced indirectly, either through inspection of extant artifacts or through observation of a modern descendant. In his Principien der Sprachgeschichte (1886: 320), Hermann Paul saw writing as the only medium through which past linguistic circumstances can be reconstructed, and recommended careful investigation of the complex and flexible relationship between writing and speech. Ε. H. Sturtevant, in Linguistic change (1917: 10), relegated writing to a subordinate position, of interest to the linguist only because of its utility in language restoration. Structuralists of the 1930s, 1940s, and 1950s did not question the importance of written evidence for historical linguistics, but few sought to clarify their assumptions or explain their procedures for drawing conclusions from the evidence. Generativists have generally ignored the matter altogether. Robert D. King devotes an entire chapter of his Historical linguistics and generative grammar (1969) to theories of scribal practice, but concludes that the enterprise is so fraught with difficulties and uncertainties as to be "not a satisfying field for the linguist" (1969: 213). A few scholars, however, acknowledge principled orthographic analysis to be a prerequisite to phonological analysis, e.g., Charles Hockett (1959: 578); W.Sidney Allen (1965: vi, and 1968: vi); lipo Tapani Piirainen (1968: 20); Herbert Penzl (1971: 5, 19, 3 1 - 3 2 , 34, and 1971: 305); James M. Anderson (1973: 16); Charles V. J. Russ (1976: 65); and with strong reservations William J. Jones (1979: 256). Paul (1886: 321) stated that the purpose of an alphabet is to record phonological differences that occur within a dialect, but only those differences which have functional value. The notion that orthographic

304

Robert Kyes

contrasts imply phonological contrasts became a guiding principle somewhat later; one of the first to apply it to a problem in historical linguistics was W. Freeman Twaddell, in his "A note on Old High German umlaut" (1938), accounting thereby for medieval German scribes' failure to record the difference between their back and front rounded vowels. In numerous publications Herbert Penzl has shown how the cautious and qualified observance of this principle, as a working hypothesis, can aid in the discovery of early phonologies and phonological changes; we cite here only "The phonemic split of Germanic k in Old English" (1947), "Umlaut and secondary umlaut in Old High German" (1949), "The evidence for phonemic changes" (1957), Lautsystem (1971: 13 — 43), "Scribal practice" (1971), and Vom Urgermanischen zum Neuhochdeutschen (1975: 2 2 - 2 6 ) . The goals of historical phonology, according to Penzl, are to reconstruct earlier phonological systems, and to identify not only the phonemic oppositions that operate within them, but also allophonic alternations and phonetic features. The first assumption is that scribes tend to distinguish phonemes in their spellings, but not allophones. The procedure is one of multidimensional comparison, comparing spellings of the same or similar morphemes in the same text (homographic analysis), in other contemporary texts (heterographic comparison), and in the same dialect at different times (diagraphic comparison). The appearance of a new orthographic alternation that can be correlated with conditioning factors suggests a phonemic split; examples of this are treated by Twaddell (1938), Penzl (1947, 1949), Kurath (1956), and Allen (1965: 23 [on Lat. -gn- in lignum vs. lego]). Coalescence of symbols implies phonemic merger. Reverse spellings point to merger or loss; see Gillis Kristensson (1979: 304-307). Piirainen (1968) bases a phonological analysis of an early seventeenth-century German text on a computer-assisted comparison with Middle High German. One exhaustive comparative study is Ursula Schulze's Studien zur Orthographie und Lautung der Dentalspiranten s und ζ im späten 13. und frühen 14. Jahrhundert (1967). Jones (1979) wants to keep principles to a minimum, regarding them as "theoretical or practical preconceptions" that could distort "highly malleable" historical evidence (1979: 256). Possibilities for such distortion are indeed numerous and unpredictable. Paul (1886: 232 — 335) showed that the match between letter and sound may be disturbed by etymological or analogical spellings. All writing systems are borrowed

Evidence

305

by one people from another, and with them go many of the donor's scribal conventions, which may have little relevance for the borrower's phonemic system. Furthermore, once a writing system gains acceptance, there is pressure to maintain it. Paul (1886: 327 — 330) noted that orthography neither causes nor prevents phonological change, but ignores it. Stockwell and Barritt (1961: 76) argue that the weight of tradition even in medieval times must have been enormous, and doubt the likelihood of early grapheme-phoneme equivalence. While etymological spellings, imported conventions, and indigenous traditions represent conscious attempts to fashion a writing system whose relationship to the users' phonemic system is not isomorphic, one must also reckon with unconscious phenomena like a scribe's inattention, misinterpretation, or incompetence; see King (1969: 205 — 206). All of these factors conspire to render the relationship between orthography and pronunciation opaque, and frustrate our attempts to identify and describe phonological changes.

References Allen, W. Sidney 1965

Vox Latina: A guide to the pronunciation of Classical Latin. Cambridge: Cambridge University Press. 1968 Vox Graeca: A guide to the pronunciation of Classical Greek. Cambridge: Cambridge University Press. Anderson, James M. 1973 Structural aspects of language change (Longman Linguistics Library No. 13). London: Longman. Hockett, Charles F. 1959 "The stressed syllables of Old English", Language 35: 5 7 5 - 5 9 7 . Jones, William J. 1979 "Graphemic evidence for the diphthongization of jej and /of: The case re-opened", Neophilologus 63: 250—259. King, Robert D. 1969 Historical linguistics and generative grammar. Englewood Cliffs: PrenticeHall. Kristensson, Gillis 1979 "On the evidence for phonemic change", Neuphilologische Mitteilungen 80: 3 0 4 - 3 0 7 . Kurath, Hans 1956 "The loss of long consonants and the rise of voiced fricatives in Middle English", Language 32: 435—445. Paul, Hermann 1886 Principien der Sprachgeschichte (2nd edition). Halle: Niemeyer.

306

Robert Kyes

Penzl, Herbert 1947 "The phonemic split of Germanic k in Old English", Language 23: 34 — 42. 1949 "Umlaut and secondary umlaut in Old High German", Language 25: 223-240. 1957 "The evidence for phonemic changes", in: Pulgram (ed.), 193 — 208. 1971 Lautsystem und Lautwandel in den althochdeutschen Dialekten. München: Hueber. 1971 "Scribal practice, phonological change, and biuniqueness", The German Quarterly 44: 305-310. 1975 Vom Urgermanischen zum Neuhochdeutschen: Eine historische Phonologie (Grundlagen der Germanistik 16). Berlin: Schmidt. Piirainen, lipo Tapani 1968 Graphematische Untersuchungen zum Frühneuhochdeutschen (Studia Linguistica Germanica 1). Berlin: de Gruyter. Pulgram, Ernst (ed.) 1957 Studies presented to Joshua Whatmough on his sixtieth birthday. The Hague: Mouton. Russ, Charles V. J. 1976 "The data of historical linguistics: Sources for the reconstruction of pronunciation from written records", York Papers in Linguistics 6: 65 — 73. Schulze, Ursula 1967 Studien zur Orthographie und Lautung der Dentalspiranten s und ζ im späten 13. und frühen 14. Jahrhundert (Germanistische Forschungen, Neue Folge 19). Tübingen: Niemeyer. Stockwell, Robert P . - C . Westbrook Barritt 1961 "Scribal practice: Some assumptions", Language 37: 75 — 82. Sturtevant, Ε. H. 1917 Linguistic change: An introduction to the historical study of language. Chicago: University of Chicago Press (reprinted 1961). Twaddell, W. Freeman 1938 "A note on Old High German umlaut", Monatshefte fur deutschen Unterricht 30: 177-181.

Structuralist interpretation Elmer H. Ant onsen

Structuralists have seen the causes of phonological change in both psychological and physiological factors connected with the speech act. Martinet (1955) and most other structuralists consider the development of umlaut allophones in Germanic root syllables to be the result of an anticipation of the vowel in following syllables, which is psychological in origin but physiological in its production. Martinet and also Dal (1967) believe that the phonemicization of these allophones is not so much dependent upon the reduction or loss of the conditioning factors as on the psychological shifting of the function of distinguishing among the given morphological and lexical units from the nonroot to the root syllables, so that the conditioning factors, the nonroot vowels, become redundant in function and therefore eventually lose their distinctiveness. Considering the numerous nondistinctive variants assignable to any given phoneme in a particular system, Hockett (1958) believes that the ultimate cause of sound change is to be found in the inadequacies of a speaker's ability to perceive and to articulate precisely the same sounds, so that he actually aims at an idealized target, the given phoneme, without consistently hitting the "bull's eye". Further, Hockett believes that the configuration of the oral cavity, with greater physical space in the front than in the back, may be responsible for the relatively greater instability of back-spread phonemes compared to front-rounded ones, since there is less room for maximal phonological space (Hockett 1959; see now also Moulton 1987). The concept of phonological space is one which Martinet (1955), in particular, has made use of in trying to explain intertwining developments in phonological systems, whereby a shift in one element sets off

308

Elmer Η. Antonsen

a shift in neighboring ones in a so-called "push- and drag-chain" relationship. Examples are the Modern English and the Norwegian and Swedish great vowel shifts. Directly connected with this concept is that of the "hole in the pattern", which results when one or more phonemes in a system shift to leave a void, as illustrated by the ProtoGermanic short/lax vowel system in Table 3. (see p. 300), which is skewed by the absence of a distinct reflex of PIE joj. The "horror vacui" of phonological systems (or a tendency toward symmetry) eventually fills this void with a new phoneme derived from an allophonic variant of PG /u/. One might perhaps argue whether the change in the pre-Old High German long/tense vowel system described in conjunction with Table 1. (see p. 298) represents the development and subsequent filling of "holes in the pattern" or the working of the "push- and drag-chain" concept (Moulton 1961). In any case, with this change, the Old High German system shows that one series has left the original system (to become diphthongs) and been replaced by a new series (from older diphthongs). Whether there ever actually was a genuine "hole in the pattern" is not clear. (Critiques of these concepts are found in Galton 1977, Jeffers 1974, and Collinge 1970.) One of the problems created by the "push- and drag-chain" concept is the fact that if it holds, we should expect no "holes in the pattern" to develop. In other words, why do some phonological changes result in shifts of greater or lesser extent, while others result in mergers or other developments leaving phonological voids? Martinet (1952, 1955) seeks to answer this question on the basis of the functional loads of the contrast involved, maintaining that distinctions with lower functional loads are more susceptible to loss than those with higher ones. It lies in the nature of the structuralist approach to language change, with its assumption of subphonemic, allophonic variation (whether conditioned or in free variation) as a prior condition for phonemic change, that phonological change is considered to be gradual, whereas instantaneous change is a reflex of morphological restructuring or of external influence (Penzl 1965). References Anderson, John M. —Charles Jones (eds.) 1974 Theory and description in phonology. Vol. II of Historical Linguistics. Proceedings of the First International Conference on Historical Linguistics, 2 — 7 September 1973, Edinburgh. Amsterdam: North-Holland.

Structuralist interpretation

309

Antonsen, Elmer H. 1965 "On defining stages in prehistoric Germanic", Language 41: 19 — 36. 1967 "On the origin of the Old English digraph spellings", Studies in Linguistics 19: 5 - 1 7 . 1969 "Zur Umlautfeindlichkeit des Oberdeutschen", Zeitschrift für Dialektologie und Linguistik 36: 201 —208. 1972 "The Proto-Germanic syllables (vowels)", in: van Coetsem — Kufner (eds.), 117-140. 1975 A concise grammar of the older Runic inscriptions (Sprachstrukturen A 3). Tübingen: Max Niemeyer. 1978 "On the notion of 'archaicizing' inscriptions", in: Weinstock (ed.), 282 — 288.

Benediktsson, Hreinn 1963 "Some aspects of Nordic umlaut and breaking", Language 39: 409 — 431. Cardona, George — Norman H. Zide (eds.) 1987 Festschrift for Henry Hoenigswald: On the occasion of his seventieth birthday. Tübingen: Gunter Narr. Cercignani, Fausto 1979 "Proto-Germanic */i/ and */e/ revisited", Journal of English and Germanic Philology 78: 7 2 - 8 2 . Collinge, Ν. E. 1970 Collecteana Linguistica: Essays in general and genetic linguistics (Janua Linguarum, Series Minor 21). The Hague: Mouton. Dal, Ingerid 1967 "Über den /-Umlaut im Deutschen", Neuphilologische Mitteilungen 68: 47-64. D'Alquen, Richard J. Ε. 1974 Gothic ai and au: A possible solution (Janua Linguarum, Series Practica 151). The Hague: Mouton. Dressier, Wolfgang U. —Oskar Pfeiffer (eds.) 1977 Phonologica 1976. Akten der dritten Internationalen Phonologie-Tagung Wien, 1.—4. September 1976 (Innsbrucker Beiträge zur Sprachwissenschaft 19). Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck. Galton, Herbert 1977 "A different view of economy in phonological change", in: Dressier — Pfeiffer (eds.), 201 - 2 0 4 . Greenberg, Joseph H. 1979 "Rethinking linguistics diachronically", Language 55: 275 — 290. Haugen, Einar 1969 "Phonemic indeterminacy and Scandinavian umlaut", Folia Linguistica 3: 107-118. Hockett, Charles F. 1958 A course in modern linguistics. New York: MacMillan. 1959 "The stressed syllables of Old English", Language 35: 575-597. Hoenigswald, Henry M. 1960 Language change and linguistic reconstruction. Chicago: U. of Chicago Press.

310

Elmer Η. Antonsen

Jakobson, Roman —Gunnar Fant —Morris Halle 1951 Preliminaries to speech analysis. Cambridge: MIT Press. Jarceva, Viktorija N. (ed.) 1961 Voprosy germanskogo jazykoznanija: Materialy vtoroj naucnoj sessii po voprosam germanskogo jazykoznanija. Moskva: Akademija Nauk. Jeffers, Robert J. 1974 "On the notion 'explanation' in historical linguistics", in: Anderson — Jones (eds.), 231-256. Kurytowicz, J. 1952 "The Germanic vowel system", Biuletyn Polskiego Towarzystwa Jgzykoznawczego 11: 50 — 54. Lehmann, Winfred P. 1953 "The conservatism of Germanic phonology", Journal of English and Germanic Philology 52: 140-152. 1961 "A definition of Proto-Germanic: A study in the chronological delimitation of languages", Language 37: 67 — 74. Lyons, John 1968 Introduction to theoretical linguistics. Cambridge: Cambridge University Press. Makaev, Enver A. 1961 "Ponjatie obscegermanskogo jazyka", in: Jarceva (ed.), 44—67. Marchand, James W. 1970 "Gotisch", in: Schmitt (ed.), 9 4 - 1 2 2 . Martinet, Andre 1952 "Function, structure, and sound change", Word 8: 1 —32. 1955 Economie des changements phonetiques: Traite de phonologie diachronique. Berne: A. Francke. Moulton, William G. 1961 "Zur Geschichte des deutschen Vokalsystems", Beiträge zur Geschichte der deutschen Sprache und Literatur (Tübingen) 83: 1 —35. 1967 "Types of phonemic changes", in: To honor Roman Jakobson: Essays on the occasion of his seventieth birthday, 11 October 1966. Vol. Ill, 1393 — 1407 (Janua Linguarum, Series Maior 31 — 33). The Hague: Mouton. 1979 "Notker's 'Anlautgesetz' in: R a u c h - C a r r (eds.), 241 - 2 5 2 . 1987 "On vowel length in Gothic", in: C a r d o n a - Z i d e (eds.), 281-291. Nöth, Winfried 1974 "Perspektiven der diachronen Linguistik", Lingua 33: 199 — 233. Penzl, Herbert 1955 "Zur Erklärung von Notkers Anlautgesetz", Zeitschrift für deutsches Altertum 86: 196-210. 1965 "Phonem, Allophon und Sprachlaut in der historischen Sprachwissenschaft", in: Zwirner-Bethge (eds.), 4 5 8 - 4 6 2 . 1971 Lautwandel und Lautsystem in den althochdeutschen Dialekten. München: Max Hueber. Rauch, Irmengard —Gerald F. Carr (eds.) 1979 Linguistic method: Essays in honor of Herbert Penzl (Janua Linguarum, Series Maior 79). The Hague: Mouton.

Structuralist interpretation

311

Russ, Charles 1978 "Recent developments in historical phonology", New German Studies 6: 175-194. Schmitt, Ludwig Erich (ed.) 1970 Sprachgeschichte. Vol. I of Kurzer Grundriß der germanischen Philologie bis 1500. Berlin: de Gruyter. Stockwell, Robert P . - C . Westbrook Barritt 1951 Some Old English graphemic-phonemic correspondences — ae, ea, and a (Studies in Linguistics, Occasional Papers 4). Washington: Georgetown University. van Coetsem, Frans 1970 "Zur Entwicklung der germanischen Grundsprache", in: Schmitt (ed.), 1-93. van Coetsem, Frans — Herbert Kufner (eds.) 1972 Toward a grammar of Proto-Germanic. Tübingen: Max Niemeyer. Weinstock, John (ed.) 1978 The Nordic languages. Vol. III. Proceedings of the third International Conference of Nordic and General Linguistics, 5 — 9 April 1976, Austin. Austin: University of Texas. Zwirner, Eberhard — Wolfgang Bethge (eds.) 1965 Proceedings of the Fifth International Congress of Phonetic Sciences, 16 — 22 August 1964, Münster. Basel: S. Karger.

Synchronic rules and diachronic "laws": The Saussurean dichotomy reaffirmed Robert T. Harms

Rules which relate one stage of a linguistic form to another are common to both generative grammar and historical linguistics. Unfortunately, this similarity between the synchronic rules of generative grammar and the "laws" of language change (often called "diachronic rules") has at times led to misunderstanding. Unlike the rule-like statements of sound change regularities between given (arbitrarily chosen) chronological stages of a language, the rules of generative grammar are strictly synchronic. They may be said to reflect the processes by which a speaker derives a correct pronunciation for sequences of morphemes, within a grammatical system containing lexical representations of morphemes which are not identical with their phonetic realizations in an utterance; for example, /dis + ζ/ 'dish' + plural (lexical representation of the two morphemes in 'dishes') requires a rule of schwa insertion and a rule for the correct value of lax /i/ to yield the phonetic form [disaz]. A second confusing similarity between generative phonology and historical linguistics stems from the often striking resemblance between the abstract lexical representations and the hypothetical reconstructed forms of earlier stages; for example, modern English /wld/ 'wide' (as for Chomsky and Halle 1968), with a series of rules providing a synchronic "vowel shift": wld > wlyd > weyd > wäeyd > wäyd. But the justification for such forms and rules is purely synchronic. The speaker of modern English must account for the numerous morpheme variants which preserve the effects of the English Vowel Shift (e. g., wide — width, bite — bit, etc.). One way would be to treat such pairs as suppletive in nature, i. e., by listing each morpheme separately (with

314

Robert Τ. Harms

no rules); alternatively, the phonetic variants could be derived by rules from a single representation (as in Chomsky — Halle 1968). Whichever decision is correct for the speaker of modern English, the facts of historical change cannot enter into the language-acquisition process, and are therefore irrelevant. In a sense, the language-acquisition process may be seen as analogous in certain respects to the historical method of internal reconstruction, but that process is guided solely by values leading the learner to an optimal synchronic system with no concern for historical truth. From a somewhat different perspective, Stockwell (1972) argues in effect that although the learner indeed has no access to the history of his language, knowledge of dialects and history might well lead the linguist to prefer a synchronic analysis which correctly reflects historical change as a more plausible reflection of native competence. To further illustrate this distinction we might note that the Germanic sound shifts known as Grimm's Law were never a synchronic rule of any stage of Germanic (although it is not difficult to state these as a single rule under the conventions of generative phonology), since they did not lead to morpheme variants (cf. King 1969: 39). Contrast, however, the extreme position of Lightner (1975: 618) that sets of forms such as paternal — Jupiter — father necessitate a synchronic rule mirroring Grimm's Law. On the other hand, Verner's Law, by virtue of the numerous morpheme alternations it produces, had synchronic rule counterparts in every Germanic language except Gothic — although by Proto-Germanic times it applied only to those morphemes with conditioned variants, and an abstract stress position need not have been one of the terms of the synchronic rule. Finally, economy and simplicity are frequently invoked in evaluating both synchronic rules and diachronic statements of change. The simplicity metric of synchronic linguistics, which is viewed as directing the learner toward the forms and rules of a grammar, must be clearly separated from the notions of economy, plausibility, and elegance used to support given hypotheses of historical change. By contrast, Andersen's claim (1973: 780) that "adaptive change is a more complex phenomenon than evolutive change" represents an attempt to link synchronic and diachronic notions of economy. Generative phonology has contributed little to the understanding of how to evaluate the economy of given historical claims. Especially misguided were a number of early statements that the best historical account relating two stages of a language is that which posits the fewest changes consistent

The Saussurean dichotomy reaffirmed

315

with constraints on synchronic grammars (as in Postal 1968: 265), which in turn led to arguments for the initiation of fairly abstract rules with grammatical conditioning — such as the loss of final nasals except in the first singular in Estonian (Kiparsky 1971 b: 19, even in the face of contradictory evidence and a more convincing explanation in the source which he cites, i.e., Kettunen 1929).

1. Rules and forms in language change The study of phonological rules is at the same time the study of morphemes and their lexical forms (also known as "underlying", "systematic phonemic"). Together the rules and forms constitute interdependent components of a grammar. In general, the stronger the requirement that morpheme variants be united by a common underlying form, the greater the phonetic difference (degree of abstractness) between lexical and surface representations, and the greater the need for rules and other formal devices to relate the two levels of representation. In the general study of language change, rules and lexical forms must always be considered jointly. The generative historian must establish the relevant synchronic portions of each language system being compared. Only a handful of generative histories, and these dealing with limited sets of rules supported by historical record, are worthy of note: Halle —Keyser 1971 (English stress), Wolfe 1972 (English vowel shift), and Entenman 1977 (French vowel nasalization).

2. The form of rules The standard form of a segmental phonological rule is: F —• C / A Β i. e., the class of segments F (focal class; F is never a sequence of segments) in the environment ( = '/') following A and preceding Β are changed in those feature values specified in C. (Note that C does not replace F.) If F is zero ('0'), C represents an epenthetic segment; if C, then F is deleted. The absence of an environment indicates an unconditioned process, such as a phonetic detail rule or a rule of absolute neutralization (synchronic total merger — generally regarded as questionable following Kiparsky 1968).

316

Robert Τ. Harms

Two distinct aspects of a rule must be recognized: 1) the structural description (SD) — those conditions which determine the rule's operation (i. e., the sequence AFB); and 2) the structural change (SC) — its phonetic effect (i.e., C). An SD —SC transformational format handles processes in which either F or C is viewed as more than one segment, e.g., French vowel nasalization (cf. Schane 1973: 68): SD: V

C [ +nasal]

The transformational format is generally employed to express coalescence {oy > ε), decomposition (a > oa), and segmental split (η > ntj s; this type of change differs from true epenthesis, in which the inserted segment may not be viewed as an extension of A or B). The synchronic permutation of segments (metathesis) by transformational rules (as in Keyser 1975: 404) does not appear to be justified (cf. Webb 1976). The economy (simplicity) of a rule is a function of the complexity of its SD — a measure of the generality of F, A, and B. The SC, on the other hand, must be judged in terms of phonetic plausibility relative to a given conditioning environment or the typological naturalness (markedness) of the output segment. Despite numerous attempts the SC has not been successfully incorporated into the notion of phonological simplicity. The most notable such attempt was the theory of linking proposed by Chomsky and Halle (1968: 400 — 435). Although their argument included apparent support from language change (palatalization in Slavic, English umlaut) — and the theory was immediately applied to the diachronic study of sibilants in Spanish (Harris 1969) and Russian vowel reduction (Lightner 1968) — it has been precisely the counterevidence from sound change that has shown such theories to be ill-founded (cf. Bach —Harms 1972). Morpheme structure rules (Halle 1959) and conditions (Stanley 1967), phonological "conspiracies" (Kisseberth 1970), and surfacephonetic constraints (Shibatani 1973) are best interpreted as reflections of a single aspect of phonological organization: the interaction of segmental and nonsegmental domains of control (morpheme, word, syllable). Among these various approaches, only Shibatani's surface constraints, by virtue of recognizing the independent role of prosody (i. e., the syllable), may be shown to have any diachronic significance (cf. especially 102—104). A single change in prosodic structure may often call forth several segmental adjustments. Early Slavic, for ex-

The Saussurean dichotomy reaffirmed

317

ample, shifted to an exclusively open-syllable type, but numerous independent segmental rules were introduced within a fairly short period in conjunction with the new constraint (among others: nasalization, monophthongization, epenthesis, and diverse adjustments in consonant clusters). In recent years, recognition of the important role of prosodic structures has led to several new specialized formal systems: autosegmental phonology (Goldsmith 1976) — a generative revitalization of the long components of Harris (1947); metrical phonology (Liberman — Prince 1977) — phrase and word stress assignment by means of binary branching "prosodic" trees, and its extension to the constituent structure of segments within a syllable (Kiparsky 1979). The application of these models to problems of sound change may be found in Prince 1980 (Finnic gradation), Ingria 1980 (Grassman's Law in Greek, compensatory lengthening), and de Chene —Anderson 1979 (532); but the contribution of these efforts is notational at best. Among the numerous other formal rule types proposed in the generative literature, two deserve mention here: "via rules" and "variable rules". "Via rules" have been proposed (Vennemann 1972) to establish nonproductive correspondences among similar-yet-suppletive morpheme variants. For example, reflexes of the English Vowel Shift are treated as follows: lexical /wayd/ 'wide' is related to /wid + Θ/ 'width' via an /ay/ —> /i/ directional association statement (as opposed to deriving one from the other or both from underlying /wid/ as in the standard theory). In this view virtually all of the segmental rules of Chomsky and Halle (1968) would be via associations. Borrowed forms, such as Latin vocabulary in Spanish, may also lead to via references. The variable rule of William Labov (1969: 737 — 742) gave formal recognition to quantifiable systematic variation in the application of an optional rule. For example, a simplified version of auxiliary contraction is given (1969: 739) as: 9 —> (0) / [ß pro] # #

+T

Co # # [aVb]

i. e., the optional deletion of the auxiliary ( # # [ + T]Co # # ) reduced vowel is favored by a following verb ([a Vb]) and a preceding pronoun ([β pro]). Greek letters are used here to rank (a "most favorable", β "next most favorable", etc.) and quantify social as well as linguistic factors influencing the operation of the rule. The variable rule, origi-

318

Robert Τ. Harms

nally hailed for bridging the Saussurean dichotomy, is currently viewed as an important heuristic device rather than as a rule of synchronic grammar.

3. The life cycle of a rule The initiation of sound change — i.e., the introduction of a phonological rule — in generative studies has largely been treated (or ignored) in accord with the Postal (1968: 283) tradition that, apart from universal constraints on grammars, languages could change without limit and in unpredictable ways. The converse view, that "synchronic rules should be so constrained that they cannot violate possible types of historical change", was invoked by Stockwell (1972) in arguing against "exchange rules" (rules in which the focal class and the output class are switched; e. g., North Lapp i > e and e > i in unstressed syllables). Bach and Harms (1972: 18) propose "strong naturalness constraints on the initiation of phonetic rules ... [which constraints] are essentially diachronic". The subsequent modification of these rules under formal and social forces may be seen as the basis for the "unnatural" ("crazy") rules of synchronic grammars. The apparent initiation of rules conditioned by nonphonetic (i.e., syntactic, morphological, and lexical) factors can generally be traced to 1) a misguided application of diachronic economy (as discussed above), 2) the absence of documentation of the crucial intervening stages, and 3) the failure to consider the mediating role of prosody in associating grammatical structures and segmental rules. For example, the flapping of intervocalic /t/ in American English is generally said to be conditioned by adjacent segments and boundaries (cf. Kahn 1976), but the contrasting word-initial /t/'s of by tomorrow (flapped) and buy tomatoes (unflapped) reveal a contrast in prosodic foot structure (i. e., only foot-internal /t/ flaps) — a level of prosody largely ignored in descriptive studies (including the many sociolinguistic attempts to document the details of change in progress). In yet another case we find that early Baltic Finnic had a rule deleting final vowels from the endings of finite verbs (cf. Harms 1964: 110 — 112), but once the SOV sentence structure of Finnic — no longer rigid in Baltic Finnic — is considered, the original conditioning role of utterance-final prosody is clear. "Lexical diffusion" (Wang 1969), by which rules are claimed to be arbitrarily tied to specific morphemes, does not represent

The Saussurean dichotomy reaffirmed

319

a rule-governed process at all, but rather a mode of lexical restructuring (cf. Catford 1974: 22, where the process is labeled "substitution"). We find, for example, that the Welsh evidence for diffusion presented by Chen and Wang (1975: 257) reveals a typical case of dialect mixture in an intermediate zone (cf. Sommerfelt 1962: 74 — 75, which stands in sharp contrast with Chen and Wang's use of that data to bolster their viewpoint). A rule with a high degree of naturalness presents a different sort of problem: Does the rule represent grammatical control or is it the consequence of universal constraints on speech production? For example, the [s] variant of the English plural, commonly held to be a rule of assimilation, may be seen instead to result from general (and inviolable) constraints on voicing control, and is thus not a rule of grammar — a "nonrule" (Harms 1978). The "persistent rules" of Chafe (1968), a source of apparent ordering difficulties, tend to reflect nonrule constraints of this kind. Indeed, nonrules, by virtue of their subtle interaction with substantive aspects of prosody (timing, intensity, etc.), may well constitute one of the major sources of rule initiation in situations of sociolinguistic influence. The nonrule output of one dialect can serve as a model for the introduction of a phonological rule in a second dialect, one with slightly different prosody (cf. Harms 1976). Along similar lines, recent work by Ohala (1974 a, 1974 b) and Catford (1974) suggests an explanation of rule initiation in the physical nature of articulation and perception with reference to a sociolinguistic context. Many rules, especially those which cause morpheme alternations, persist and evolve in transmission from one generation or social group to another. Phonetic rules conditioned by «-ary feature values (for example, spirantization of Spanish voiced stops unless preceded by a phonetically homorganic consonant — i.e., 'η coronal' but not 'coronal') are susceptible to restructuring on a binary basis (as prosodic word-final lengthening in English has led to the phonological "tensing" of nonlow vowels). The structural description of a rule tends to be extended to a more general focal class (for example, diphthongization of mid vowels extended to nonhigh vowels) or environment (for example, palatalization before [i] extended to front vowels). Because of the essential bond between generality and rule economy, this sort of SD extension is frequently called "rule simplification", and at times inappropriately applied to the output (SC) as well. Extension of the SC, on the other hand, is difficult — if not impossible — to characterize

320

Robert Τ. Harms

in terms of rule formalism. Output "hypercorrection" (following Labov 1965: 112) may lead to exaggeration (for example, breaking of [ä] to [oa] has the initial component raised to [u]; affrication of [ky, t5] to [c]), but the workings of contrastive function, areal accommodation, and typological simplification (markedness) can also be demonstrated. The final stages in the evolution of a rule are marked by increased lexical conditioning, typically manifested in the form of diacritic features, minor rules (cf. Kiparsky 1971 a: 598), morphologized rules (cf. Holden 1972). Contrary to common lore, the morphologization of a rule need not entail loss of productivity. For example, Russian velar softening (a reflex of the Slavic first palatalization) remains automatic with diminutive -ka formations: for example, monolog 'monologue' + ka yields monolozka (Holden 1972: 118). One final twist in the alteration of a rule is "rule inversion", by which F —• C / A Β is restructured as C —> F in an environment approximating the complement of A B; as when certain stem vowels of a paradigm are lost via apocope and syncope, the remaining vowels may be reinterpreted as the result of epenthesis (cf. Vennemann 1972).

References Andersen, Henning 1973 "Abductive and deductive change", Language 49: 765 — 793. Anderson, John M. — Charles Jones (eds.) 1974 Historical linguistics. Vol. II. Amsterdam: North-Holland. Bach, Emmon — Robert T. Harms 1972 "How do languages get crazy rules?" in: Stockwell — Macaulay (eds.), 1-21. Bruck, Anthony —Robert Fox —Michael LaGaly (eds.) 1974 Papers from the Parasession on Natural Phonology, 18 April 1974. Chicago: Chicago Linguistic Society. Catford, Ian C. 1974 "Natural sound changes: Some questions of directionality in diachronic phonetics", in: Bruck — Fox — LaGaly (eds.), 21 — 29. Chafe, Wallace L. 1968 "The ordering of phonological rules", UAL 34: 115 — 136. Chen, Chin-Chuan- William S.-Y. Wang 1975 "Sound change: Actuation and implementation", Language 51: 255 — 281. Chomsky, Noam —Morris Halle 1968 The sound pattern of English. New York: Harper & Row. de Chene, B. — Stephen R. Anderson 1979 "Compensatory lengthening", Language 55: 505 — 535.

The Sanssurean dichotomy

reaffirmed

321

Dingwall, William Orr (ed.) 1971 A survey of linguistic science. College Park: University of Maryland, Linguistics Program. Entenman, George 1977 The development of nasal vowels (Texas Linguistic Forum 7). Austin: University of Texas at Austin, Department of Linguistics. Goldsmith, John A. 1976 Autosegmental phonology. Bloomington: Indiana University Linguistics Club. Gribble, Charles E. (ed.) 1968 Studies presented to Professor Roman Jakobson by his students. Columbus: Slavica Publishers. Halle, Morris 1959 The sound pattern of Russian. The Hague: Mouton. Halle, Morris —Samuel Jay Keyser 1971 English stress. New York: Harper & Row. Harms, Robert T. 1964 Review of The structure and development of the Finnish language, by Lauri Hakulinen. Word 20: 105-114. 1976 "The segmentalization of Finnish nonrules", in: Harms — Karttunen (eds.), 73-88. 1978 "Some nonrules of English", in: Jazayery — Polome — Winter (eds.), 39 — 51. Harms, Robert T. — Frances Karttunen (eds.) 1976 Papers from the Transatlantic Finnish Conference, 10 April 1976, Austin. (Texas Linguistic Forum 5) Austin: University of Texas, Department of Linguistics. Harris, James W. 1969 "Sound change in Spanish and the theory of markedness", Language 45: 538-552. Harris, Zellig 1947 Methods in structural linguistics. Chicago: University of Chicago Press. Holden, Cyril T. 1972 "Loan words and phonological systems", unpublished dissertation, University of Texas at Austin. Ingria, Robert 1980 "Compensatory lengthening as a metrical phenomenon", Linguistic Inquiry 11: 465-495. Jazayery, Μ. Α. —Edgar C. Polome—Werner Winter 1978 Linguistic and literary studies in honor of Archibald A. Hill. Vol. II. The Hague: Mouton. Kahn, David 1976 Syllable-based generalizations in English phonology. Bloomington: Indiana University Linguistics Club. Kettunen, Lauri 1929 Eestin kielen äännehistoria. Helsinki: Suomalaisen Kirjallisuuden Seura. Keyser, Samuel Jay 1975 "Metathesis and Old English phonology", Linguistic Inquiry 6: 377—411.

322

Robert Τ. Harms

King, Robert 1969 Historical linguistics and generative grammar. Englewood Cliffs: PrenticeHall. Kiparsky, Paul 1968 How abstract is phonology? Bloomington: Indiana University Linguistics Club. 1971 a "Historical linguistics", in: Dingwall (ed.), 576-649. 1971 b Phonological change. Bloomington: Indiana University Linguistics Club. 1979 "Metrical structure assignment is cyclic", Linguistic Inquiry 10: 421 —441. Kisseberth, Charles W. 1970 "On the functional unity of phonological rules", Linguistic Inquiry 1: 291-306. Kreidler, Charles W. (ed.) 1965 Georgetown University monograph series on language and linguistics 18. Washington, D. C.: Georgetown University Press. Labov, William 1965 "On the mechanism of linguistic change", in: Kreidler (ed.), 91 — 114. 1969 "Contraction, deletion, and inherent variability of the English copula", Language 45: 715-762. Liberman, Mark —Alan Prince 1977 "On stress and linguistic rhythm", Linguistic Inquiry 8: 249 — 336. Lightner, Theodore M. 1968 "An analysis of akan'e and ikan'e in Modern Russian using the notion of markedness", in: Gribble (ed.), 188-200. 1975 "The role of derivational morphology in generative grammar", Language 51: 617-638. Ohala, John J. 1974a "Experimental historical phonology", in: Anderson — Jones (eds.), 353 — 389. 1974 b "Phonetic explanation in phonology", in: Bruck — Fox — LaGaly (eds.), 251-275. Postal, Paul 1968 Aspects of phonological theory. New York: Harper & Row. Prince, Alan S. 1980 "A metrical theory for Estonian quantity", Linguistic Inquiry 11: 511 — 562. Schane, Sanford A. 1973 Generative phonology. Englewood Cliffs: Prentice-Hall. Shibatani, Masayoshi 1973 "The role of surface phonetic constraints in generative phonology", Language 49: 8 7 - 1 0 6 . Smith, Μ. E. (ed.) 1972 Studies in honor of George L. Trager. The Hague: Mouton. Sommerfeit, Alf 1962 Diachronic and synchronic aspects of language. The Hague: Mouton. Stanley, Richard 1967 "Redundancy rules in phonology", Language 43: 393 — 436.

The Saussurean dichotomy reaffirmed

323

Stockwell, Robert P. 1972 "Problems in the interpretation of the Great English Vowel Shift", in: Smith (ed.), 344-362. Stockwell, Robert P . - R o n a l d K. S. Macaulay (eds.) 1972 Linguistic change and generative theory. Bloomington: Indiana University Press. Vennemann, Theo 1972 "Rule inversion", Lingua 29 (1972): 2 0 9 - 2 4 2 . Wang, William S.-Y. 1969 "Competing sound changes as a cause of residue", Language 45: 9 — 25. Webb, Charlotte 1976 "The status of metathesis as a phonological process", in: Sixth California Linguistic Association Conference Proceedings, San Diego, 77—89. San Diego: San Diego University Press. Wolfe, Patricia M. 1972 Linguistic change and the Great Vowel Shift in English. Berkeley: University of California Press.

Morphophonology* Frans van Coetsem and Susan McCormick

1. Introduction The purpose of this chapter is to examine the diachronic aspect of morphophonology, a term which implies a link between the areas of phonology and morphology. 1 Before addressing the historical perspective, we will briefly discuss the notion of morphophonology in general, as its status and definition are points of debate. The treatment of morphophonology in the literature ranges from the establishment of a separate morphophonological component, to an analysis of such phenomena as integral parts of a comprehensive phonology/morphophonology/morphology component. In between are descriptions that view the status of morphophonology in various ways. Some see it as a definable interface between phonology and morphology, but do not accord it status as an autonomous language component. Others admit a vaguely defined area between phonology and morphology, but do not posit separate status. Still other analyses disallow an independent morphophonological description by strictly assigning morphophonological study to either phonology, morphology, or both. A broad definition of morphophonology includes the study of the phonological shape and structure of the morphemes of a language, specifically the study of the phonological differences, i.e., the alternations that occur within morphemes. A more limited definition of morphophonology includes alternations only insofar as they have a morphological significance. In both cases, alternation is the key notion * We would like to thank Linda R. Waugh for some useful comments.

326

Frans van Coetsem and Susan

McCormick

in morphophonology and serves as the basis for further distinctions, such as the status of an alternation with morphological function as either the primary (relevant) or concomitant (redundant) marker of the morphological distinction in question. The differing viewpoints on how morphophonology should be defined has also led to disagreement among linguists as to which types of alternations are to be included within the morphophonological domain and which belong to phonology and morphology proper. In spite of the divergent opinions concerning the validity and definition of a separate morphophonological section in linguistic description, the occurrence of alternations that operate within morphemes is a linguistic fact that should be described and accounted for. In the next few pages, we will isolate some representative examples of alternations, and look at the various major approaches to synchronic morphophonological analysis. We will then discuss the diachronic perspective of morphophonology, in particular how the synchronic alternation types relate to specific diachronic developments. For a survey (with bibliography) of the general and representative views on morphophonology, see Kilbury (1976) and Dressier (1985: especially 1 — 8).

2. Alternation An alternation is composed of alternants and can be schematized as an operation of the type: A ~ Β

C...)

The alternation occurs within morphemes. When speaking of morphemes in this context, especially pertinent are the notions of "code frequency" (either an open or a closed list) and "message frequency". Alternations also occur in sets. It is useful to distinguish some representative examples of alternations, as they are found in the literature: 1) A vocalic alternation delineates tense, for example, in English give ~ gave (ablaut), or number in foot ~ feet and in German Mutter 'mother' ~ Mütter 'mothers' (umlaut). This type of alternation performs a morphological function and is non-automatic; it contains morphological information and is said to have a morphological "environment" or to be morphologically "conditioned". 2 In the example give ~ gave the environment is present ~ past; in foot ~ feet it is

Morphophonology

327

singular ~ plural. The alternation is also the primary or relevant marker of the morphological distinction. In some cases, the alternation is productive, as in sing ~ sang (cf. substandard bring ~ brang for bring ~ brought), while in others it is not, for example, oo ~ ee in foot ~ feet.3 Competing patterns are not uncommon and one morphological function may be marked in more than one way. In a case like this there is a concomitant (redundant) marker of the morphological distinction beside the primary (relevant) marker. It may not always be clear which of the markers is the primary one. In German Hand 'hand' ~ Hände 'hands', for instance, the a ~ ä alternation and the plural marker -e concurrently signal the singular-plural distinction. It should also be noted that tonal oppositions within the same morphemes may perform morphological functions. 2) A second type of alternation is morphologically conditioned and non-automatic, for example, / ~ ν in wife ~ wives. Here, however, the alternation serves only as a concomitant or redundant marker of the morphological distinction, and is not productive. 4 English oo ~ ee in foot ~ feets (beside entirely regularized foots) is another example, occurring during the acquisition process of English, as well as in substandard English. 3) In some cases an alternation performs a morphological function, but has a phonological conditioning which is restricted in different ways and is non-automatic. Compare [ϊ] ~ [ε] ~ [ε] (or jij ~ /ε/ ~ /ε/) in creep ~ crept ~ crept, keep ~ kept ~ kept, dream ~ dreamt ~ dreamt, next to dreamed ~ dreamed. Such verbs behave very much like ablaut verbs and are normally grouped with them. The alternation [ϊ] ~ [ε] also occurs in derivational patterns like deep ~ depth, and in loans such as serene ~ serenity. Similar alternations are [ai] ~ [i], e.g., wide ~ width, divine ~ divinity, and [ei] ~ [ae], e.g., shade ~ shadow, profane ~ profanity. English [k] ~ [s] in electric ~ electricity, opaque ~ opacity, occurring only in loans, also belongs to this category. 4) A fourth group consists of phonologically conditioned alternations that perform no morphological function, but occur within certain welldefined morphological categories. For example, the plural in English shows an assimilation that depends on the voicing value of a final consonant, e.g., backs ~ bags [z]. Although the s ~ ζ alternation does not occur in the specified phonological environment throughout the language (cf. sins [sinz] and since [sins]), it is entirely predictable within the morphological category in question. In this sense it can be

328

Frans van Coetsem and Susan

McCormick

viewed as automatic. 5 The indefinite article in English is similarly restricted, e. g., a day vs. an apple. 5) Finally, there is the type of variation or alternation that is fully automatic, does not perform a morphological function, and is in no way morphologically constrained. An alternation like this is seen in the word-final devoicing in German Tag [k] ~ Tages [g]. This type of alternation exemplifies well the Prague School archiphoneme, where the opposition between two phonemes is neutralized in a specific phonological context. Some further remarks concerning alternations can be made: i) An alternant may be zero, for example, 0 ~ η in Old High German fähan 'to catch' ~ fieng 'caught', past. ii) In cases like bring ~ brought, think ~ thought, the alternating parts comprise more than one phonemic unit. An extreme example is suppletion or a suppletive alternation, seen in cases like go ~ went. iii) An alternation pattern of two alternants appears to be common and basic. Alternations of more than two alternants are, however, not uncommon, as ablaut verbs in Germanic and the Germanic languages show. In Middle High German, for example, a verb like nemen 'to take' displays no less than six alternants, namely i ~ e ~ a ~ x ~ B/C D becomes opaque when, as the result of the action of other rules, phonetic A becomes possible in the environment C D. Schindler (1974: 5) writes that according to the testimony of most of the individual Indo-European languages a final *-n after a long vowel is lost. In Greek, however, *-n after a long vowel becomes possible again as a result of two rules introduced later, namely 1) *-m > -n (i.e., in word-final position) and 2) the loss of *-t in the sequence *-nt (e. g., pherön 'carrying' < *pherönt). As a result of the loss of the rule requiring the elimination of the final -n after a long vowel, final -n is restored in the nominative singular of the «-stem nouns. For example, after the stem-

358

Philip Baldi and William R.

Schmalstieg

final *-n had been lost in a Pre-Greek form *khiö a new nom. sing. khiön got its stem-final *-n a second time on the basis of the stem of oblique cases such as the gen. sing, khion-os. The attempt to introduce formal methods, such as the application of ordered rules, is a noble effort in the continuing human hope of introducing a strict methodology into the discipline of historical linguistics. In fact, however, counterexamples, inconsistencies and inattention to empirical evidence render all known formalisms imperfect. As Feyerabend (1978: 55) writes: "... no single theory ever agrees with all the known facts in its domain." It was the abandonment of the American structuralists' strict methodological requirement of beginning linguistic analysis on the phonemic level which allowed the 'generative revolution' to take place. A good case could now be made for the abandonment of most contemporary methodologies in that they limit the scope of their investigation to such a degree that the results are usually at least somewhat distorted.

4. Word-formation Stein (1977: 220) writes that word-formation "traditionally distinguishes between derivation and compounding; in derivation bound morphemes are used ... Compounds, on the other hand, always consist of at least two free morphemes." Derivational morphemes occur typically as infixes, prefixes and suffixes, while compounds typically consist of etymologically free morphemes, e. g., blackbird (from black and bird). The process of derivation results in a new word with its own paradigm (e.g., verb inflate and noun inflation), and frequently is accompanied by meaning differences not predictable from the sum of the two parts. For example, the English word demonstrable means 'able to be demonstrated' while the word danceable (as in danceable music) does not mean 'able to be danced' but rather 'music to which one can dance'. Derivational morphology must be distinguished from inflectional morphology, which defines the grammatical relationship of a word to other words in a syntagm. Inflectional morphology takes place within paradigms, and never alters the meaning of the root, for example, Lat. fllius 'son', fllii 'of the son', flliö 'to the son', laudö Ί praise', laudäs 'you praise', laudäbämus 'we were praising', etc.

Morphological

change

359

It is customary to classify compounds as either "endocentric" or "exocentric". An endocentric compound is one which has the same function as one of its members, for example, the compound blackbird functions in a sentence just like bird. An exocentric construction, on the other hand, is one in which the construction does not belong to the same form or function class of either of its constituents, for example, John ran is an exocentric construction because it is neither a nominal (like John) or a verbal (like ran). In compounding, exocentric constructions are quite unusual, though such English examples as devilmay-care (e. g., your devil-may-care attitude disturbs me) show an adjectival or modifying function which is not found in any of the constituent parts of the compound. There are borderline cases, for example, redcap is not a kind of cap, but it does function as a noun just like the word cap. According to Marchand (1969: 10): "The coining of new words proceeds by way of combining linguistic elements on the basis of a determinant/determinatum relationship called syntagma. When two or more words are combined into a morphological unit on the basis just stated, we speak of a compound." There is a natural human tendency to see both likeness and difference. Thus in the word steamboat the similarity with others of its kind is expressed by boat, the difference by the word steam. From the semantic viewpoint the determinatum {boat) represents the element whose range of applicability is limited by the determinant (steam). In traditional terminology such a compound would be an endocentric compound. Marchand excludes the so-called exocentric compounds from consideration and calls them derivatives. Among the early Indo-European languages, compounding is most systematically represented by Sanskrit. Compounds in Sanskrit are declinable stems made up of two or more declinable stems. Generally speaking, compounds are nominal alternatives for phrase-level expressions, for example, knob of the door = doorknob, or, in Skt. asvasya ghosas 'horse's neigh' can also be expressed as asvaghosas. There are five principal types of compound in Sanskrit, the first three based on the relation of the elements of the compound to each other, the last two defined by the relation of the whole compound to other elements in the utterance (cf. Whitney 1964). The main groups are as follows (Sanskrit examples are quoted in their uncombined forms): 1) dvandva, 'two-and-two', copulative or coordinate compounds. In this type, no member is subordinate to any other, e.g., pänipäda-

360

Philip Baldi and William R.

Schmalstieg

'hands and feet' ( = päni ca padäu ca); ahinakula — 'snake and mongoose' ( = ahis ca nakulas ca); miträvaruna — 'Mitra and Varuna' (= mitras ca varunas ca). 2) tatpurusa, or dependent compounds. In this class one member is in some kind of dependent case relationship to the head, modifying it or qualifying it in some way. Thus nagaragamana- = nagaram gamanas 'going to the city', is an accusative tatpurusa where the relationship is one of object {nagaram 'to the city'); hastakrta- — hastena krtas 'made by hand', where there is an instrumental relationship. 3) karmadhäraya, or descriptive compounds. In this class the first member modifies the second and is either an adjective, or a noun used adjectivally modifying a noun, or an adverb modifying an adjective. Some examples are nilotpala- = nilam utpalam 'blue lotus' (adjective + noun); räjadanta- = dantas räjä asti 'the tooth is king' = 'kingtooth' (molar). 4) bahuvrihi, or possessive compounds. The bahuwihi compounds are analogous to the English compounds of the type redcap and Bluebeard. These compounds refer neither to caps nor to beards, but rather to those who possess or are characterized by them. The Sanskrit word bahuvrihi characterizes the class. In composition it is bahus 'much' and vrihis 'rice', but it refers neither to bahus nor vrThis. Instead, it refers to one who possesses or is characterized by 'much rice'. Another example is: süryatejas- 'someone characterized by the brightness of the sun' ( = süryasya tejas yasya sas). 5) avyayibhäva, or indeclinables. These compounds are formed by modifying a declinable stem by means of an indeclinable (an avyaya), and the resultant compound occurs only in the accusative singular neuter and functions as an adverb. Some examples are: yathäkämam 'at will' ( = yathä 'as' + käma- 'desire'; yathäsamkhyam 'respectively' (= yathä + samkhya- 'enumeration'). Hoenigswald (1977: 9 — 12) writes that the trend towards endocentricity is new in Indo-European. Originally most compounds with truly non-deverbative second elements are either "Rektionskomposita" transformed from phrases in which this second element appears as governed, e. g., Gk. eperetmos '[seated] at the oars' or else they are of the possessive compound type which we have seen in Sanskrit as the bahuvrihi type. Compounds could then change their meaning from exocentric to endocentric, thus räja-putra- 'having kings for sons' acquires a meaning 'the king's son' with a shift of stress to räja-puträ-. The Greek compound prothuron (< pro 'before' and thurön 'doors')

Morphological

change

361

originally meant only 'area in front of the house' and only later came to refer specifically to a door in front of another door. Similarly in earlier forms of the language a suffixed word was exocentric in that it performed a different function from its root word. Thus the agent suffix -ter- and -tor- could be attached from verbs, e. g., Latin actor 'agent' (from ag-ö Ί lead; do'). Later forms of the Indo-European languages allow what one might call endocentric sufflxation, for example, Hündchen is a diminutive dog and whitish is a degree of white, i.e., both words perform the same function as the root from which they are derived, Hund 'dog' and white, respectively. Lipka (1977: 155 — 156) writes that in the process of lexicalization the syntagmatic character of complex lexemes is lost. Thus in the pronunciation of the English word forecastle considerable phonological change has taken place. Fewer phonological changes are still possible in the creation of compounds, for example English clergyman, policeman, etc., in which only the reduction of the vowel in [-man] is noticed. But the old form can be reconstituted if the derivation is still evident and doublets can occur, for example, waistcoat can be pronounced [weskat] or [weisksut]. An important element in the lexicalization of compounds is the loss of motivation, or semantic recoverability of the compound parts. This may occur as the result of extra-linguistic or linguistic changes. An example of the former is furnished by the word blackboard which is now used to refer to the common green board. An example of the latter is furnished by mincemeat in which the second element harks back to a time when meat meant any kind of human food. A further step in the loss of motivation is the passage to an idiom. For example, to understand cannot be comprehended from its elements under- and -stand. Suffixes may be formed from old compounds when the second element of the compound is lost as an independent word (Henzen 1965: 109). Thus gött-lich 'godlike' derives from an old possessive compound (bahuvrihi) the second element of which derives from Germanic *-likaz 'body'. Thus etymologically göttlich meant something like 'having God's body'. When *likaz was lost as a separate word, the element -lich became a suffix. It could then spread to other words, thus Gott 'God' is to göttlich as Glück 'luck, fortune' is to glück-lich 'lucky, fortunate' (1965: 110). A fairly common phenomenon is the concatenation of several suffixes to produce a new suffix, for example, the -keit of German Finsterkeit 'darkness' derives (from -ic plus -heit, (1965: 111).

362

Philip Baldi and William R. Schmalstieg

Note 1. Since the purpose of this chapter is a brief review of the recent literature on analogy, we have omitted the vast history with references to Esper 1973 and Anttila — Brewer 1977.

References Anderson, John M. — Charles Jones (eds.) 1974a Syntax, morphology, internal and comparative reconstruction. Vol. I of Historical linguistics. Proceedings of the First International Conference on Historical Linguistics, 2 — 7 Sept. 1973, Edinburgh (North-Holland Linguistics Series, 12a). Amsterdam/New York: North-Holland/Elsevier. 1974b Theory and description in phonology. Vol. 2 of Historical linguistics. Proceedings of the first International Conference on Historical Linguistics, 2 — 7 Sept. 1973, Edinburgh (North-Holland Linguistics Series, 12). Amsterdam/New York: North-Holland/Elsevier. Anttila, Raimo 1972 An introduction to historical and comparative linguistics. New York: MacMillan. 1977 Analogy (Trends in Linguistics: State-of-the-Art Reports). The Hague: Mouton. Anttila, Raimo — Warren A. Brewer 1977 Analogy: A basic bibliography (Amsterdam Studies in the Theory and History of Linguistic Science, Series V, Vol. 1). Amsterdam: John Benjamins. Baldi, Philip 1979 "Typology and the Indo-European prepositions", Indogermanische Forschungen 84: 49 —61. Best, Karl-Heinz 1973 Probleme der Analogieforschung. München: Hueber. Brekle, Herbert E. — Dieter Kastovsky (eds.) 1977 Perspektiven der Wortbildungsforschung. Bonn: Bouvier Verlag Herbert Grundmann. Brogyanyi, B. (ed.) 1979 Festschrift for Oswald Szemerenyi on the occasion of his 65th birthday. Amsterdam: John Benjamins. Buck, C. D. 1933 Comparative grammar of Greek and Latin. Chicago: University of Chicago Press. Dingwall, William Orr (ed.) 1971 A survey of linguistic science. College Park, Maryland: Linguistics Program, University of Maryland. Esper, Erwin Allen 1973 Analogy and association in linguistics and psychology. Athens: University of Georgia Press.

Morphological change

363

Fairbanks, Gordon H. 1977 "Case inflections in Indo-European", JIES 5: 101 - 1 3 1 . Feyerabend, Paul 1978 Against method. London. Henzen, Walter 1965 Deutsche Wortbildung. Tübingen: Max Niemeyer. Hoenigswald, Henry 1977 "Diminutives and tatpurusas: The Indo-European trend toward endocentricity", JIES 5: 9 - 1 3 . Jespersen, Otto 1921 Language: Its nature, development and origin. New York: Norton. King, Robert D. 1969 Historical linguistics and generative grammar. Englewood Cliffs, New Jersey: Prentice-Hall. Kiparsky, Paul 1971 "Historical linguistics", in: William Orr Dingwall (ed.), 576-649. 1974 "Remarks on analogical change", in: John M. Anderson — Charles Jones (eds.) (1974b), 257-275. 1980 "Concluding statement", in: Elizabeth Closs Traugott et al. (eds.), 409-417. Kurylowicz, Jerzy 1958 "Ogolne tendencje zmian analogicznych", Biuletyn polskiego towarzystwa jgzykoznawczego 17: 207 — 219. 1960a "La nature des proces dits 'analogiques'", Acta Linguistica 5 (1949), 5: 15 — 37. Reprinted in Esquisses linguistiques. Wroclaw and Krakow: Polska Akademia Nauk 1960, 6 6 - 8 6 . 1960b "Odpowiedz jezykoznawstwa", Biuletyn polskiego towarzystwa jgzykoznawczego 19, 203 — 210. Lakoff, Robin 1969 "Another look at drift", in: R. Stockwell - R. Macaulay (eds.), 172-198. Lehmann, Winfred P. 1973 Historical linguistics: An introduction (2nd edition). New York: Holt. Li, C. N. (ed.) 1974 Word-order and word-order change. Austin: University of Texas Press. Lipka, Leonhard 1977 "Lexikalisierung, Idiomatisierung und Hypostasierung als Probleme einer synchronischen Wortbildungslehre", in: Herbert Ε. Brekle — Dieter Kastovsky (eds.), 155 — 163. Manczak, W. 1958 "Tendances generates des changements analogiques", Lingua 7: 298 — 325, 387-420. 1960 "Odpowiedz Prof. J. Kurylowiczowi", Biuletyn polskiego towarzystwa j§zykoznawczego 19: 191—201. 1979 "L'apophonie e/o en grec", in: B. Brogyanyi (ed.), 529 — 535. Marchand, Hans 1969 The categories and types of present-day English word-formation (2nd edition). München: Beck.

364

Philip Baldi and William R. Schmalstieg

Sapir, Edward 1921 Language. New York: Harcourt, Brace, Jovanovich. Saussure, Ferdinand de 1959 Course in general linguistics, trans. Wade Baskin. Edited by Charles Bally and Albert Sechehaye. New York: Philosophical Library. Schindler, Jochem 1974 "Fragen zum paradigmatischen Ausgleich", Die Sprache 20: 1 —9. Schmalstieg, William R. 1980 Indo-European linguistics: A new synthesis. University Park/London: Penn State University Press. Stein, Gabriele 1977 "The place of word-formation in linguistic description", in: Herbert E. Brekle - Dieter Kastovsky (eds.), 219-235. Stockwell, Robert P. — R. Macaulay (eds.) 1969 Linguistic change and generative theory. Bloomington: Indiana University Press. Thomason, S. G. 1980 Review of Analogy by R. Anttila. Language 56: 418-424. Traugott, Elizabeth C. et al. (eds.) 1980 Papers from the Fourth International Conference on Historical Linguistics, 26 — 30 March 1979, Stanford. Amsterdam: John Benjamins. Venneman, Theo 1974a "Topics, subjects and word-order: From SXV to SVX via TVX", in: John M. Anderson — Charles Jones (eds.) (1974a), 339 — 376. 1974b "An explanation of drift", in: C. N. Li (ed.), 269-305. Villar, F. 1981 Dativo y locativo en el singular de la flexion nominal indoeuropea. Salamanca: Ediciones Universidad de Salamanca. Whitney, William Dwight 1964 Sanskrit grammar. Cambridge, Mass.: Harvard University Press.

Syntactic change Win/red P. Lehmann

1. Introduction The study of syntactic change has a long background on which current and future study must be based. A brief review cannot survey this in any detail. Moreover, surveys with bibliography are available (Delbrück 1893: 1 - 8 8 ; Porzig 1924; Wackernagel 1926: 1 - 6 9 ; Havers 1931: xi — xviii, 209 — 270). These surveys deal largely with the IndoEuropean language family, but include the view that all such study must concern itself with the theoretical problems of universal syntax (Porzig 1924: 126). And Havers specifically discusses non-Indo-European languages, as does Koschmieder (1965). These works remain important, even though we find outmoded approaches and theoretical inadequacies in them. They include well-argued treatments of fundamental issues accompanied by pertinent data and bibliography. Among these issues are the following: the scope of syntax; definition of the sentence and its constituents; the processes of syntax; phenomena like deixis, anaphora, gender, person, and so on. These issues continue to be topics of discussion in subsequent works, as examination of recent publications discloses. Few surveys have been made for these in more recent works (cf. Sebeok 1973; Birnbaum 1978). The brief space allotted this essay allows little more than a sketch of the current activity and some problems, and no treatment at all of work on syntactic change in specific languages, however much one would like to include studies by scholars like Karl-Horst Schmidt and Antonio Tovar. The proceedings of conferences also illustrate attention to diachronic syntax (Lehmann - Malkiel 1968, 1982; Li 1975, 1976, 1977; Steever et al. 1976). Further, international conferences on his-

366

Winfred P. Lehmann

torical linguistics have devoted considerable attention to syntax (Anderson - Jones 1974; Christie 1976; Mäher et al. 1981; Traugott et al. 1980; Ahlquist 1981/1982; Fisiak 1980, 1984; Ramat 1980). Points on which agreement has been reached and generally maintained are the following. The sentence, rather than the sound (phoneme) or word is the fundamental unit of language (Brugmann 1904: 281, 623 ff.). Some scholars, to be sure, propose the text as the fundamental unit (Junker 1924: 20); in the second edition of the Grundriß Brugmann suggests a similar view when he prefers Äußerung '(complete) utterance' to Satz 'sentence' (1906:1). Syntactic studies, however, generally focus on sentences and their constituents, leaving larger structures to different undertakings, in the past referred to as stylistics and now as discourse analysis or text linguistics. Sentences are regarded as strings of syntactic elements. Produced in time, these have a horizontal dimension. To express different meanings in such strings, their elements must be selected from competing entities, such as the possible nouns that might fill the position of an object. These two dimensions lead to the images popular a generation ago for representing the syntagmatic and paradigmatic dimensions of language. As the basic dimension, the syntagmatic is more fundamental. An accurate understanding of language then takes arrangement as its fundamental process, selection as secondary to it. Morphology is simply a means to delimit further the competing entities available in selection; in a language with eight inflections for case, the competing entities at some position in the string are more highly delimited formally than in a language with two such inflections. Treatment of morphology should not therefore direct the syntactic component of a grammar; it was the development of historical and general linguistics that directed earlier linguists to assign it that role. Such a role for morphology is no longer maintained in most recent work on syntactic change. In dealing with syntax a small number of processes have been recognized. Bloomfield lists four: order or arrangement, selection, modulation or intonation, modification or sandhi (1933: 163 — 164). Brugmann cites the first three of these (1925: 1); Paul's listing (1920: 123 — 124) is less compact than Brugmann's and Bloomfield's. Various periods have concentrated on one of these processes. Nineteenthcentury historical linguistics virtually confined its attention to syntactic change involving selection. Change in arrangement has received considerable attention in the last decades. Change in modulation has

Syntactic

change

367

attracted little concern, neither has change in sandhi rules. The concentration of any one period or any one linguist is determined in part by views on the relative importance of these processes in the formation of sentences. Different periods also use different terminology for the same processes or constructions. One example is ellipsis in coordinated constructions; in older works this is referred to as "economy" (Sparsamkeit, Paul 1920: 314 — 315), in more recent as "gapping". As another example, nominal subject and object clauses of past grammars are now widely referred to as complements. Eventually a consensus or a partial consensus on terminology may be reached. Historical phonologists no longer deal with consonant variations as "grades" or Stufen, though the terms have been maintained for vowel variations, as in apophony or ablaut. Attention to change in terminology is important for relating current study with that of the past, and especially for preventing failure to make use of earlier findings. It is also important to observe a major difference between current attention to syntactic change and attention to it in the past, especially with reference to protolanguages. Earlier scholars made no attempt to reconstruct syntactic patterns. Brugmann, with his usual clarity, begins the preface to the second edition of his Grundriß with the statement that historical linguistics still cannot present the facts of language history in the context and chronological sequence in which they occurred. He states further that not even a relative chronology can be proposed (1897: ix); and he repeats his position in the preface to his volume on morphology (1906: vii). Similarly, Meillet begins his preface stating that his book has a very limited objective — to indicate briefly the agreements between the various Indo-European languages (1937: vii). Brockelmann follows the same approach for Semitic (Brockelmann 1913). Current historical syntactic studies, on the other hand, do propose chronological relationships, attempting to follow "the form that history must assume" (Brugmann 1897: ix; see, for example, Lehmann 1974). Students using older "historical" grammars and essays must note the difference.

2. The sites of syntactic change Syntactic change may affect any of the four processes: arrangement, selection, intonation, or sandhi. Moreover, phonological patterns signal these processes, and accordingly phonological change may bring

368

Winfred P. Lehmann

about syntactic change. Syntactic change may therefore be related to what I here call sites. It should be noted that the four processes apply to morphology as well as to syntax if the two are distinguished. Morphological units are constructed in specific patterns of arrangement, as well as by the three other processes. In this essay inflectional morphology is treated as an integral part of syntax, in accordance with the aim of examining change in languages of every structure, from isolating to polysynthetic, and of accounting for equivalent markers, such as prepositions, suffixes or postpositions for case. This treatment does not imply that morphological change, or phonological change, should not be examined independently as well. Semantic change apparently has little, if any effect on syntactic change; rather, the reverse is true. For example, French pas, originally 'foot, step', underwent semantic change when ne preceding it was reduced and lost. Moreover, when auxiliaries are introduced (such as forms of the root *wes- 'live', represented still in was, were; or *bhew'grow', represented in be, been; or Japanese eru 'get', represented in the potential suffix -e-), the change in their function is brought about by syntactic innovation rather than restructuring of meaning in these verbs. This is not to deny the complexity of meaning in many verbs, as pointed out by Gulstad (1974: 123). Surveying treatments of syntactic change, we confine ourselves here to five sites. It may be assumed that a specific syntactic change is centered in any of the five sites, and that scholarly concentration on one as opposed to another reflects in part scientific caution. For example, only since Schmidt (1926) have linguists concerned themselves with the effect of change in arrangement on further syntactic change. The concern has been highly prominent in recent years, but still has not been as thoroughly treated as has change in selection or phonological change affecting syntactic patterns. For much concern on change in arrangement has been directed at "word-order change" rather than change of order among syntactic elements. Yet syntax, as Brugmann long ago noted, is not simply the study of arrangement of words, but rather of affixes and other morphemes, or formatives as Brugmann labeled them. Change in arrangement therefore cannot confine itself to change in the order of words (see Hetzron 1980). Such limited syntactic change, however, has been the focus of much study (Canale 1976 and his references; Friedrich 1975). See also the limited scope of the first conference at Santa Barbara (Li 1975). Some of the partici-

Syntactic change

369

pants did, however, deal with change also in affixes and syntactic categories (Steele 1975; Lehmann 1975b). The continuing emphasis on words reflects the nineteenth-century concentration on selection. The concern with selection led nineteenth-century linguists and their followers to emphasize change of morphological elements. If the verbal system of a language was modified, as by reduction of morphological categories like an aorist and a perfect with development of compound forms, it was the loss of selection entities rather than modification in patterns of arrangement to which the change was ascribed. The same applies to the loss of nominal inflections and the introduction of adpositions (Schmidt 1926: 492), as well as many other selection changes, such as "loss of gender". These changes were only secondarily related to change in arrangement, which might accord with syntactic universale referring to general categories of meaning. Schwyzer cites the differences between Meillet and Sechehaye on this matter, himself following Meillet in concentrating on form (1950: 7 — 11). Yet at the same time he pointed out Brugmann's late and experimental proposal to deal with universal psychological categories (1918). Only recently has this proposal been generally adopted, in the widespread concern for what today are called universale. Phonological change is taken by some linguists to be the primary site of syntactic change. This view is presented forcefully by Vennemann (1975; see also Hamp 1976, 1979). Vennemann's eighth generalization states the basis for it: "Every morphological system is destroyed in time by phonological change" (1975: 293). According to this view other processes need to be adopted upon the "destruction" of the selection markers. Yet Vennemann recognizes that morphology may be "replenished". The Greek ί-aorist, replenished after loss of intervocalic -s-, is an example. Even linguists singling out phonology do not therefore consider it the sole site of syntactic change in language. In their treatments of syntactic change some works are largely devoted to support of particular theoretical approaches rather than explanation of data, e. g., the stratificational (Christie 1974), the transformational (Kiparsky 1968; Lakoff 1968; Traugott 1968,1969; Lightfoot 1979), the anti-transformational (Anttila 1972; see also Makkai — Makkai 1976). Differing approaches lead to varying terminology. For example, the widespread use of tree-diagrams in transformational grammar led to the introduction of metaphorical terms such as raising, left-branching, right-branching and so on. And the emphasis on cognitive psychology brought about much publication on the cognitive

370

Winfred P. Lehmann

processes involved in linguistic activities, including linguistic change, as well as further terminology (H. Andersen 1973, among others). Such differences in focus are surveyed elsewhere. This essay reviews recent treatments of syntactic change which treat by any theory alteration of constructions with reference to the sites and sources of change.

3. Use of a framework The necessity of carrying out syntactic studies in terms of a framework has been implicit in approaches to language for some time, as in statements by Brugmann (1904: viii) and in Meillet's clipped characterization of language as a system ou tout se tient. Yet neither Brugmann nor Meillet drew the obvious conclusions, following instead the course of historical linguistics which dealt first with the more accessible elements of language. Recent studies emphasize the importance of treating syntactic change in terms of a framework. Various frameworks are proposed; besides distinguishing between underlying and surface structure, a distinction between system and norm — as well as parole — developed by Coseriu has been taken over by some scholars (Coseriu 1962; H. Andersen 1973). Syntactic constructions are assumed to be interrelated (Smith 1978), so that the change of one affects others (Lehmann 1978: 37 — 42, 400 — 421; Hawkins 1979). For example, the presence of prepositions as opposed to postpositions in a language is related to the position of finite verbs (V) with regard to their objects (O); VO languages typically have prepositions, OV languages have postpositions. Syntactic constructions associated in this way are identified in several publications (Greenberg 1966; Lehmann 1978). Greenberg labeled languages with numerals. Most linguists now take their labels from the central syntactic construction, the clause. They set up two large classes, VO and OV (Bartsch — Vennemann 1982: 33). Each has three potential sub-types: SVO, VSO, VOS; SOV, OSV, OVS. It has been estimated that 40% of languages are SVO, about 10% VSO, about 50% SOV, with only minute numbers of the other types. In accordance with the axiom that languages are always in flux, few languages show consistency of type. Inconsistencies reflect ongoing syntactic change. When languages cannot be identified for a given type, they are called ambivalent (Lehmann 1972).

Syntactic

change

371

Study of consistent languages has identified syntactic patterns associated with basic clause types. Greenberg's synchronic presentation treats interrelationships as implicational universals. From a historical point of view such interrelated patterns are found in languages because changes in syntactic constructions are brought about by changes in key constructions of the syntactic framework (Lehmann 1973a, 1974; Vincent 1980; Jeffers 1976b who is sceptical). An early treatment of such interrelated changes is that of P. W. Schmidt (1926: esp. 488 — 491). According to him, the position of accusative objects with regard to their verb is changed when the position of the genitive with regard to its head noun changes; similarly the position of adjectives to the head of their noun phrase. Schmidt's proposal suffered from his view that the position of genitives was determined by external or non-linguistic considerations: the type of culture. Original (Ur-) cultures have preposed genitives. Change to postposed position came through language mixture during the migrations resulting from social and economic changes brought about by matriarchical agriculture. This view may well have hindered further consideration of interrelated syntactic change at the time. Schmidt also discussed a syntactic change which he considers "purely internal", the shift from preposed to postposed genitives in the Romance languages (1926: 491—494). This change he relates to the loss of the Latin case suffixes which gave rise to prepositional constructions. Current theory on syntactic change gives precedence to patterning of principal clauses, and thus to remodeling of morphology. Other observations of Schmidt's are touched on below, such as the slowness of syntactic change. This is illustrated by the length of time required for the change of arrangement of the Romance genitive, and also of the genitive in English (Fries 1940; Chung 1977). Most linguists dealing with syntactic change maintain the same position with regard to extent of time. The use of a framework has led some scholars to propose a cycle in language change. Vennemann, for example, states that "languages develop back and forth, or cyclically, among language types" (1974a: 371). Yet cyclical change has been severely criticized; Ebert provides a critique, with references (1976: vii —xvii). Below we deal further with proposed causes of syntactic change. Here we have been primarily concerned to illustrate the central importance ascribed to a framework in recent treatment of syntactic change — by which it is assumed that such change is carried out primarily in relation to other syntactic or linguistic change.

372

Winfred P. Lehmann

4. Syntactic change involving arrangement Accurate understanding of language has led to the current concern with patterns of arrangement, their interrelationships, and their changes (Lehmann 1969 etc.; Aitchison 1979; Parker 1980). In this concern, sentences are examined for their essential elements: verb and object, with subject as a further constituent. Modifiers and function words supplement these, as do formatives. The many studies of change from OV to VO order or the reverse illustrate recognition of the central importance of arrangement in syntax. It is agreed that linguistic change takes place in underlying rather than surface forms (Lightfoot 1976; Lehmann 1976). This recognition is expressed variously. In structuralist terminology, change at the phonological level was said to be phonemic, that is, in underlying phonology, rather than phonetic, that is, at the surface (Hoenigswald 1960). An alternate term for phonemic was structural, another emic. Transformational terminology refers to reanalysis and restructuring (Traugott 1974). Identification of the time of completion of change was the goal of frequently cited essays, such as Twaddell's (1938) and that of Weinreich, Labov, and Herzog (1968). Both essays deal with phonological change, which provides more palpable evidence than does syntactic change. In view of the difficulties involved in identifying "underlying form" and the changing views of individuals, not to speak of schools, the details of syntactic change are often murky. Yet few linguists reconstruct surface forms; Watkins in his later publications is among the most prominent (1976). When syntactic change is carried out, the earliest locus is assumed to be in principal clauses, rather than in subordinate clauses. Moreover, declarative clauses rather than interrogative or negative are identified as the initial sequences affected. Subordinate clauses are held to be conservative. German is cited as an example; see also Schmidt on the Latin development of prepositions — postpositions are maintained longest in interrogative and negative clauses of Latin (1926: 492). Many studies have been carried out on change in relative clauses (Antinucci - Gebert 1979; Chr. Lehmann 1982; Romaine 1981; for accurate insights Justus 1976 is highly important). Hierarchization of change is also assumed for syntactic constructions. Clause order is taken as most influential (Lehmann 1971,1973a). In the same way, modified constructions as opposed to government

Syntactic change

373

constructions are affected last. Adjective order is among the last affected. Order in attributive compounds is even more conservative (Lehmann 1969; Tischler 1979). Vennemann associates many of these constructions in a pattern, operator: operand (Bartsch — Vennemann 1982: 35, 179, and bibliography), in which the verb is the keystone. These generalizations are inoperative when language contact interferes with the natural development of a language. Inadequate work has been done on this important problem. Sandfeld's (1930) pioneering study in areal linguistics concentrated on individual characteristics of the Balkan languages. We need investigations of innovations in such characteristics as related to the overall structure in the several Balkan languages, as well as in all languages affected by contact. Such studies may emerge from the important project on the linguistic geography of Europe (Weijnen — Kruijsen 1979); investigations of characteristic syntactic patterns are planned in this project. Another area studied, though largely for synchronic purposes, is South Asia (Emeneau 1956); some investigations have, however, touched on the diachronic consequences of the contacts between the Dravidian and the Indo-Aryan languages (Ratanajoti 1975). Other language families must be investigated for insights into change in arrangement. Studies are available for Chinese (Li - Thompson 1974a, 1974b, 1980; Tai 1976); for NigerCongo (Givon 1975; Hyman 1975); for Proto-Salish (Ingram 1975; see Lehmann 1975a); for Semitic (Givon 1979); for Uto-Aztecan (Jacobs 1975); and for Yuman (Langdon 1977). All such studies of languages attested only recently must be examined with great care, because in the absence of historical information we cannot determine whether aberrant syntactic features are the result of external influences.

5. Change in selection: morphosyntactic change As noted above, morphosyntactic change was given great prominence in nineteenth-century historical linguistics. Theoretical argument wavered between primary attention to sentences and to morphology. As classic examples we may cite Brugmann's Grundrisse. The first edition of 1886 — 1900 neatly separated phonology, morphology, and syntax with Delbrück producing the three volumes on syntax (1893 — 1900). The second edition, however, adopted the view that morphology should embrace discussion of the use of forms, i. e., their syntactic role, as well as their structure (1897 — 1916). Without Delbriick's participation

374

Winfred P. Lehmann

this edition led only through morphosyntax, leaving the impression that syntax was of secondary interest. According to Anttila, "there is no independent syntactic change at all" (1972: 256). Brugmann did produce a volume on syntax; published posthumously, it dealt only with the simple sentence (1925). Brugmann's modified view in the second edition was applied in the important historical grammars of Schwyzer (1939 — 1953) and Szantyr (1965), as well as in many subsequent works, even general introductions (Lehmann 1962). These introductions reflect the structuralist position, which concentrated on phonology and morphology, while the approach to syntax was being hammered out; see Sturtevant-Hahn (1951), for which the proposed syntactic volume never appeared. Harris's important article (1946) pointed to a principled treatment of syntax (cf. Joos's attendant note, 1957: 153). Yet the article did not shift attention away from phonology and morphology until later, as the further articles in Joos's anthology indicate. Similarly, Koschmieder's succession of articles had little effect (1965). The introductory handbooks continued the nineteenth-century emphasis except for Historical linguistics (Lehmann, 1973b: 173 —189, 200 — 202; see Traugott's comments on the first edition, 1969). The shift came about as attention to universale and typology in global language study brought attention to language of all structures, not merely the highly inflected early Indo-European dialects. General typology suggests that incidence of inflection is a function of language type. Those types with nominal elements of the sentence not separated by the verbal element require means for distinguishing subject and object. The distinguishing feature may consist of particles, but inflection is a common device. On the other hand, SVO languages have little need for inflections, especially if their order is consistent and applied in subordinate as well as principal clauses. As has often been stated, however, morphology is conservative (Meillet 1967: 111; Givon 1971). English, the Scandinavian languages, and the Romance languages provide excellent examples of the loss of inflection as languages shift from SOV to SVO order. Among problems of interest in morphosyntax is the sequence of loss of inflections, and the site of retentions or innovations. Firstly, is there a sequence of loss for markers of government (see Delbrück 1907)? And when do markers for agreement, as in gender classes, arise (Brugmann 1897; Fodor 1959; Shields 1979)? How are subclasses of nominal elements affected in these losses? Pronouns, for example,

Syntactic change

375

retain government markers longer than do nouns. Modifiers of nouns, e. g., adjectives, give up government markers before agreement markers; French, for example, retains adjectival markers for the plural; English shows such retentions only in demonstratives {this: these), not in descriptive adjectives. Secondly, how long are government markers retained? Chinese has none. In English the distinction remaining in pronouns is tenuous, as demonstrated by the widespread uncertainty in use of I:me. Thirdly, does the loss of inflection affect word class inventory? When do new word classes arise, such as articles (Blazer 1980)? And fourthly, what morphological distinctions are maintained in connecting sentences within a text? For example, verb forms indicating subordination had important uses in Classical Greek and Latin to connect sentences, as noted by the term consecutio temporum. These and other problems are being widely studied (Eliason 1980; H. Andersen 1973; Lüdtke 1980). Many were investigated in earlier historical linguistics, but with attention to specific languages or language families (Meillet 1938, 1948 and various articles; Benveniste 1971: 113 —192; Kurylowicz 1964). It is always important to remember, however, that many Indo-Europeanists viewed their work in the context of universal grammar, even if they were not constantly explicit about it, as many current linguistic studies are. Further studies illustrate widespread concern with the source and changes of syntactic entities (S. Anderson 1980; Comrie 1980; Givon 1971; Greenberg 1980; Haiman 1974; Hale 1970; Jeffers - Zwicky 1980; Justus 1976; Meillet 1967: 111; Prosdocimi 1978).

6. Intonation change Few studies are available of syntactic change affecting intonation (cf., however, Dahlstedt 1901, and especially Hoenigswald's important article, 1980). It is also difficult to investigate, because of inadequate information in records of the past. The necessary information may be inferred when accent is indicated (Ananthanarayana 1970, 1979). Metrics are highly important in providing that information (Allen 1973). Yet even when descriptions of older patterns are provided, as in texts of Classical Greek, there may be no subsequent work on examining change in intonation. The area very much needs attention.

376

Winfred P. Lehmann

7. Sandhi change Change in sandhi rules is also poorly studied. Individual grammars, as of the Middle Indie dialects, point out differences between the thoroughly described sandhi rules for Sanskrit and those for Middle Indie dialects such as Pali and the Prakrits. Details of change are pointed out in historical Romance linguistics, as of the treatment of final nasals in Latin and their loss subsequently. But as with intonation, there has been little attempt to arrive at general principles (Gauthiot 1913).

8. Sources of change One of the major contributions of recent historical theory concerns the sources of change. In earlier theory it was often assumed that change was initiated by some specific cause, possibly external to language. Prokosch, for example, proposed that the Germanic consonant shift was initiated by the vigorous leaders of the Germanic Völkerwanderung (1939: 52 — 57). Others, such as those cited by Prokosch, proposed causes like change in location, in climate, and so on. Among reasons for change then would be decisions by the speakers, cultural movements, and migration to new homes. A modified position is represented in Bloomfield's Language with its Chapters 21, "Types of phonetic change" and 22, "Fluctuation in the frequency of forms". The view presented in Chapter 22 was widely studied in dialect geography and sociolinguistics. The recent contribution has come with regard to a proposed initial cause. Linguists have progressively adopted the position that languages are not homogeneous structures (Meillet 1967: 133 — 138). An initial proposal by Fries and Pike (1949) required some time to be accepted; but the increasing study of varying forms of language has led to widespread departure from the assumption of a stable, ideal standard, whether comparable to a Saussurean langue or a Chomskyan ideal language. Patterns in functional, geographical, and social dialects as well as idiolects vary from the normal pattern of a language. Moreover, speakers adapt their patterns in accordance with the other languages they are used to (Meillet 1967: 133; Lehmann 1973b: 201; Nadkarni 1975). Further, functional investigations have led to wider understand-

Syntactic change

377

ing of marking. A language may be typically SVO; but for reasons of emphasis, humor and the like, OVS or VSO order might be used, as illustrated by Wolcott Gibbs' spoof of time style: "Backwards ran sentences till reeled the mind". If such patterns come to predominate, the unmarked might be reduced until they are lost; as aberrant patterns in subvarieties of language might be. Another potential source of innovation is tension between grammatical and functional patterns (Vennemann 1974a). Functional requirements may conflict with those of grammar, so that the topic or theme may be placed first, even when not the subject in SVO languages or SOV. Languages have been identified as topic-prominent or subjectprominent (Li — Thompson 1976). Using this dichotomy it has been proposed that OV languages are characteristically topic-prominent, VO languages subject-prominent (Lehmann 1975b). If topics come to be identified with subjects, there may be a change to SVO structure. In short, there is no need to search for mysterious external causes to trigger a shift. The bases for change are often present in a language. They may also be provided through language contact (Meillet 1967: 9 8 - 9 9 ; Lehmann 1974: 242-251; L i - T h o m p s o n 1974: esp. 207-213). The process of change was subcategorized in the influential article of Weinreich, Labov, and Herzog (1968). Change there is examined as a series of problems: The Actuation Problem; the Transition Problem; and the Embedding Problem. Embedding must be examined "in the linguistic structure ... and in the social structure" (1968: 183 — 188). Applied to syntactic change this approach seeks to determine how and when a change begins. Some changes have been ascribed in great part to interaction between languages, such as the fundamental shift of the early Indo-European dialects from OV to SVO structure, and the shift of Sinhalese from SVO to OV (Lehmann 1974; Ratanajoti 1975). Specific constructions have also been sorted out, such as the use of sequences after the clause rather than within it. Such sequences are often datives, used to express purpose, also through nominal elements of verbs. Delbrück identified the pattern as a tag or sequel (,Schleppe); Givon saw in it a departure from OV towards VO structure (1975). Complex problems in transition concern the sequence of constructions when syntactic shift takes place. Some studies suggest that shift of clause order, such as VO to OV, is followed by a shift of government constructions, e.g., prepositions to postpositions, and then of modi-

378

Winfred P. Lehmann

fication constructions, such as positions of relative clauses, genitives, and the like. Others suggest a less systematic process. Embedding, whether in the linguistic or the social structure, is viewed as a long process at the syntactic level. Fries identified a duration of a millennium in "embedding" of postposed replacing preposed position of the genitive in English (1940). Schmidt's earlier proposal on the shift of the genitive in the Romance languages also proposed a long period, without supplying chronological details (1926; see also Chung 1977). In this way recent study of the sources of change and its implementation has avoided simplistic answers. Yet much work on reliable sources is needed. Of the three possible sources for reliable examples, only the Indo-European has been examined in any detail.

9. Credence in proposals of change Linguists have seen their task in accordance with Whitney's statement: to assemble, arrange, and explain the whole body of linguistic phenomena (1892: 6; similarly Havers 1931: 1 and others). For historical linguistics the first task is fundamental. It is also difficult, especially for examination of the earliest periods of attestation, and unfortunately often incompletely discharged. When the task of assembling reliable data, also referred to as achieving observational adequacy, is neglected, the subsequent tasks of description and explanation cannot be achieved (Havers 1931: esp. 4). Historical linguists must recognize that observational adequacy may be impossible because of the inherent limitations of available data. Gothic, for example, is available only in a slavish translation of a Greek text that has not been discovered. Any generalization about Gothic, especially its syntax, must accordingly be performed with great care. Historical linguists have noted that only those syntactic sequences differing from the standard Greek of the fourth century A. D. can be used in making statements about historical Gothic and Germanic syntax, and in drawing generalizations about syntactic change. Even such sequences require wariness, as do most texts available for historical purposes. Many are literary. Further, many are produced in accordance with metrical conventions. All such texts may have an abnormally high incidence of marked patterns, so high that the regular, unmarked pattern may be difficult to detect. A complexity of potential

Syntactic change

379

pitfalls has been usefully indicated in compact form by Nagy (1974; see also Hoenigswald 1980, and Havers 1931). In view of these pitfalls a historical syntactician develops procedures which may fail to occur to general linguists. One such procedure is identification of syntactic constructions which do not lend themselves to marking. The comparative is one example; adpositions are another. On the other hand, adjective position and even position of the verb with regard to its object are open to variation. Generalizations based on them must accordingly be carefully scrutinized. A purely statistical approach is highly misleading unless it sorts out marked patterns. (Readers may understand why references are omitted at this point.) It may also be clear why reductionists have proposed attempted evasions of any historical explanation. Two loci for such evasions have attracted many followers: social and psychological. Saussure selected the social; transformational theories have turned to the psychological. Following Saussure's approach, historical study was carried out on parole, not langue. In the transformational view, competence embraces an ideal language. Competence of successive generations involves abduction of differing grammars. Historical linguists may study the relations between them. Work using reductionist approaches has not led to notable advances in historical study, including that of syntax. Many of the most important historical linguists have had a firm basis in attention to the aspect of language which is now referred to as sociolinguistic. Brugmann's manifesto urges study of the contemporary language and dialects (Lehmann [ed.] 1967:197 — 209). Meillet's unsurpassed booklet (1967) calls attention to a wide array of language varieties which subsequent sociolinguistic study has not yet encompassed. Historical linguistics finds important data in language variation, especially that of earlier periods. The great contributions of nineteenth-century historical linguistics arose from treatment of variation in phonological data. Jacob Grimm pointed out such variation when he assembled the Germanic data and propounded his widely known rules of correspondence (Lehmann [ed.] 1967: 46 — 60). Subsequent linguists concentrated on those variations, commonly known as exceptions to Grimm's Law. The method is comparable to the treatment of residues in the physical sciences. By this method they provided explanations like the impressive formulations known as Grassmann's Law and Verner's Law (translations available in Lehmann 1967). The understanding which has been

380

Winfred P. Lehmann

reached concerning the data in phonology may be examined in Kortlandt's remarkable article on the history of the Baltic accent phenomena (1974). Historical syntax is at a stage of development similar to that of midnineteenth-century phonology. Much of the data has been assembled and described. These descriptions, like Grimm's of the Germanic consonant correspondences, include statements of the exceptions. Until recently there has been no means to deal with the exceptions, no means to provide explanations, as Hirt so poignantly stated (cf. Lehmann 1974: 4). The means is now available. Rather than through general statements about details of that means (e.g., McCone 1979; Campbell—Mithun 1980), progress in understanding the data and in expanding syntactic theory will come from attention to the residues. Such attention in the most thoroughly described language families will provide the methods for dealing with the syntax of other languages.

References Ahlquist, Anders (ed.) 1981/1982 Papers from the Fifth International Conference on Historical Linguistics. Galway, 1981. Amsterdam: Benjamins. Aitchison, Jean 1979 "The order of word order change", TPS: 4 2 - 6 5 . Allen, William S. 1973 Accent and rhythm. Prosodic features of Latin and Greek: A study in theory and reconstruction. Cambridge: Cambridge University Press. Althaus, Hans Peter — H. Henne — Η. Ε. Wiegand (eds.) 1980 Lexikon der germanistischen Linguistik (2nd edition). Tübingen: Niemeyer. Ananthanarayana, Halli S. 1970 "Intonation contours in Vedic: A hypothesis", Vishveshvaranand Indological Journal 8: 1 —19. 1979 A syntactic study of Old Indo-Aryan. Hyderabad: Osmania University. Andersen, Henning 1973 "Abductive and deductive change", Language 49: 765 — 793. 1980 "Morphological change: Towards a typology", in: Fisiak (ed.), 1—50. Anderson, John M.— C. Jones (eds.) 1974 Historical linguistics: Proceedings of the First International Conference on Historical Linguistics. 2 — 7 Sept. 1973, Edinburgh. 2 vols. Amsterdam: North-Holland. Anderson, Stephen R. 1980 "On the development of morphology from syntax", in: Fisiak (ed.), 51-69.

Syntactic change

381

Antinucci, F. — A. Durand — L. Gebert 1979 "Relative clause structure, relative clause perception, and the change from SOV to SVO". Cognition 7: 145-176. Anttila, Raimo 1972 An introduction to historical and comparative linguistics. New York: Macmillan. Austerlitz, Robert (ed.) 1975 The scope of American linguistics. Lisse: de Ridder. Bartsch, Renate — Theo Vennemann 1982 Grundzüge der Sprachtheorie. Tübingen: Niemeyer. Benveniste, Emile 1971 Problems in general linguistics [Transl. Mary E. Meek], Coral Gables: University of Miami Press (first published 1966). Birnbaum, Henrik 1978 Linguistic reconstruction: Its potentials and limitations in new perspective (JIES Monograph 2). Washington: Institute for the Study of Man. Blazer, Deering 1980 "Old French articles and word order change", Folia Linguistica Historica 1: 295-304. Bloomfield, Leonard 1933 Language. New York: Holt. Brettschneider, Gunter — Christian Lehmann (eds.) 1980 Wege zur Universalienforschung. Tübingen: Narr. Brockelmann, Carl 1913 Grundriss der vergleichenden Grammatik der semitischen Sprachen, Vol. II. Syntax. Berlin: Reuther & Reichard (Reprinted 1961, Hildesheim: Olm). Brogyanyi, Bela (ed.) 1979 Festschrift for Oswald Szemerenyi. Amsterdam: Benjamins. Brugmann, Karl 1897 The nature and origin of the noun genders in the Indo-European languages [Transl. Ε. Y. Robbins], New York. 1897 and 1906 Vergleichende Laut-, Stammbildungs- und Flexionslehre der indogermanischen Sprachen (= Grundriss der vergleichenden Grammatik der indogermanischen Sprachen). 2 vols. Strassburg: Trübner. (Originally published as Vol. I & II of Karl Brugmann — Berthold Delbrück, Grundriss der vergleichenden Grammatik der indogermanischen Sprachen. 5 vols. Strassburg: Trübner, 1886-1900). 1904 Kurze vergleichende Grammatik der indogermanischen Sprachen. Strassburg: Trübner. 1918 Verschiedenheiten der Satzgestaltung nach Maßgabe der seelischen Grundfunktionen in den idg. Sprachen (Berichte der Sächsischen Akademie der Wissenschaften phil.-hist. Klasse, 70.6). Berlin: Akademie-Verlag. Bynon, Theodora 1977 Historical linguistics. Cambridge: Cambridge University Press. Campbell, Lyle — Marianne Mithun 1980 "The priorities and pitfalls of syntactic reconstructions", Folia Linguistica Historica 1: 19 — 40.

382

Winfred P. Lehmann

Christie, William M. Jr. 1974 "A stratificational view of linguistic change", in: Anderson — Jones (eds.), 65-99. Christie, William M., Jr. (ed.) 1976 Current progress in historical linguistics. Proceedings of the Second International Conference on Historical Linguistics, 12—16 January 1976, Tucson, Arizona (North-Holland Linguistic Series 31). Amsterdam: NorthHolland. Chung, Sandra 1977 "On the gradual nature of syntactic change", in: Li (ed.), 7 — 55. Comrie, Bernhard 1980 "Morphology and word order reconstruction: Problems and prospects", in: Fisiak (ed.), 8 3 - 9 6 . Coseriu, Eugenio 1962 Teoria del linguaje y lingiiistica general. Madrid: Gredos (German translation: Sprachtheorie und allgemeine Sprachwissenschaft. München: Fink, 1975). Dahlstedt, A. 1901 Rhythm and word-order in Anglo-Saxon and Semi-Saxon with special reference to their development in Modern English. Lund: Möller. Delbrück, Berthold 1893 — 1900 Vergleichende Syntax der indogermanischen Sprachen. Vol. I —III. Straßburg: Trübner. 1907 Synkretismus. Straßburg: Trübner. Ebert, Robert P. 1976 "Introduction", in: Steever et al. (eds.), vii—xviii. Eliason, Stig 1980 "Case, word order and coding in a historical linguistic perspective", in: Fisiak (ed.), 127-139. Emeneau, Murray B. 1956 "India as a linguistic area", Language 32:3 — 16. Fisiak, Jacek (ed.) 1980 Historical morphology. The Hague: Mouton. 1984 Historical syntax. Papers presented for the Third International Conference on Historical Syntax, 31 March —3 April 1981, Blazejewko, Poland. Berlin: Mouton. Fodor, Istvän 1959 "The origin of grammatical gender I", Lingua 8: 1 —41. Friedrich, Paul 1975 Proto-Indo-European syntax: The order of meaningful elements (JIES Monograph 1). Butte, Montana College of Mineral Science and Technology. Fries, Charles C. 1940 "On the development of the structural use of word-order in Modern English", Language 16: 199-208. Fries, Charles C. — Kenneth L. Pike 1949 "Coexistent phonemic systems", Language 25: 29 — 50.

Syntactic change

383

Gauthiot, Robert 1913 La fin de mot en indo-europeen. Paris: Geuthner. Givon, Talmy 1971 "Historical syntax and synchronic morphology: An archaeologist's field trip", in: Papers from the 7th Regional Meeting of the Chicago Linguistic Society. 1 6 - 1 8 April 1971, Chicago: 394-415. Chicago: Chicago Linguistic Society. 1975 "Serial verbs and syntactic change: Niger-Congo", in: Li (ed.), 47 — 112. 1979 On understanding grammar. New York: Academic Press. Greenberg, Joseph H. 1966 "Some universals of grammar with particular reference to the order of meaningful elements", in: Greenberg (ed.), 73 — 113. 1979 "Rethinking linguistics diachronically", Language 55: 275 — 290. 1980 "Circumfixes and typological change", in: Traugott et al. (eds.), 233 — 241. Greenberg, Joseph H. (ed.) 1966 Universals of language. (2nd edition) Cambridge, Mass. MIT Press. (First published 1963.) Greenberg, Joseph H. — Charles A. Ferguson — Edith A. Moravcsik (eds.) 1978 Universals of human language. Vol. I —IV. Stanford: Stanford University Press. Gulstad, Daniel E. 1974 "Reconstruction in syntax", in: Anderson — Jones (eds.) (Vol. I): 117-158. Haiman, John 1974 Targets and syntactic change. The Hague: Mouton. Hale, Kenneth 1970 "The passive and ergative in language change: The Australian case", in: Wurm - Laycock (eds.), 757-781. Hamp, Eric P. 1976 "Why syntax needs phonology", in: Steever et al. (eds.), 348 — 364. 1979 "On syllabic reduction and syntax", in: Brogyanyi (ed.), 343 — 350. Harris, Zellig S. 1946 "From morpheme to utterance", Language 22: 161 — 183. (Reprinted in Joos [ed.] 124-153.) Havers, Wilhelm 1931 Handbuch der erklärenden Syntax. Heidelberg: Winter. Hawkins, John A. 1979 "Implicational universals as predictors of word order change", Language 55: 618-648. Heilmann, Luigi (ed.) 1974 Proceedings of the IIth International Congress of Linguists, 28 Aug —2 Sept 1972, Bologna-Florence. Bologna: il Mulino. Hetzron, Robert 1980 "On word order and morpheme order", in: Brettschneider — Lehmann (eds.), 175-179. Hoenigswald, Henry M. 1960 Language change and linguistic reconstruction. Chicago, University of Chicago Press.

384 1980

Winfred P. Lehmann

"Notes on reconstruction, word-order, and stress", in: Ramat (ed.), 69-87. Hyman, Larry M. 1975 "On the change from SOV to SVO: Evidence from Niger-Congo", in: Li (ed.) (1975), 113-147. Ingram, David 1975 "A note on word order in Proto Salish", IJAL 41: 165-168. Jacobs, Roderick 1975 Syntactic change: A Cupan (Uto-Aztecan) case study. Berkeley and Los Angeles: University of California Press. Jeffers, Robert J. 1976a "Syntactic change and syntactic reconstruction", in: Christie (ed.), 1—10. 1976b "Typological shift and change in complex sentence structure", in: Steever et al. (eds.), 136-149. Jeffers, Robert J. — Arnold M. Zwicky 1980 "The evolution of clitics", in: Traugott et al. (eds.), 221 - 2 3 1 . Joos, Martin (ed.) 1957 Readings in linguistics. Washington: ACLS. Joseph, Brian 1980 "Linguistic universals and syntactic change", Language 56: 345 — 370. Juilland, Alphonse (ed.) 1976 Linguistic studies offered to Joseph Greenberg. Saratoga, CA: Anima Libri. Junker, Heinrich F. Κ. 1924 "Die indogermanische und die allgemeine Sprachwissenschaft", in: Stand und Aufgaben der Sprachwissenschaft. Festschrift für Wilhelm Streitberg, 1—64. Heidelberg: Winter. Justus, Carol 1976 "Relativization and topicalization in Hittite", in: Li (ed.), 215 — 245. Kiparsky, Paul 1968 "Tense and mood in Indoeuropean syntax", Foundations of Language 4: 30-57. Kortlandt, Frederik Η. Η. 1974 "On the history of Baltic accentuation", in: Anderson — Jones (eds.) (Vol. II), 295-309. Koschmieder, Erwin 1965 Beiträge zur allgemeinen Syntax. Heidelberg: Winter. Kurylowicz, Jerzy 1964 The inflectional categories of Indo-European. Heidelberg: Winter. Labov, William 1982 "Building on empirical foundations", in: Lehmann — Malkiel (eds.), 17-92. Lakoff, Robin 1968 Abstract syntax and Latin complementation. Cambridge, Mass.: MIT Press. Langacker, Ronald W. 1977 "Syntactic reanalysis", in: Li (ed.), 5 1 - 1 3 9 . Langdon, Margaret 1977 "Syntactic change and SOV structure: The Yuman case", in: Li (ed.) 295-290.

Syntactic change

385

Lehmann, Christian 1982 Der Relativsatz. Tübingen: Narr. Lehmann, Winfred P. 1958 "On earlier stages of the Indo-European nominal inflection", Language 34: 179-202. 1969 "Proto-Indo-European compounds in relation to other Proto-Indo-European syntactic patterns", Acta Linguistica Hafniensia 12: 1—20. 1971 "On the rise of SOV patterns in New High German", in: Schweisthal (ed.), 1 9 - 2 4 . 1972 "Contemporary linguistics and Indo-European studies", PMLA 87: 976-993. 1973a "A structural principle of language and its implications", Language 49: 47-66. 1973b Historical linguistics. An introduction (2nd edition). New York: Holt, Rinehart and Winston. (First published 1962) 1974 Proto-Indo-European Syntax. Austin: University of Texas Press. 1975a "The challenge of history", in: Austerlitz (ed.), 41 - 58. 1975b "From topic to subject in Indo-European", in: Li (ed.), 445—456. 1976 "Surface and underlying structure in typological study", in: Juilland (ed.), 517-526. 1979 "Internal reconstruction and historical syntax", Studies in Language 3: 65-89. Lehmann, Winfred P. (ed.) 1967 A reader in nineteenth-century historical Indo-European linguistics. Bloomington: Indiana University Press. 1978 Syntactic typology. Austin: University of Texas Press. Lehmann, Winfred P. — Yakov Malkiel (eds.) 1968 Directions for historical linguistics. Austin: University of Texas Press. 1982 Perspectives on historical linguistics. Amsterdam: Benjamins. Li, Charles N. (ed.) 1975 Word order and word order change. Austin: University of Texas Press. 1976 Subject and topic. New York: Academic Press. 1977 Mechanisms of syntactic change. Austin: University of Texas Press. Li, Charles N. — Sandra A. Thompson 1974a "Historical change of word order: A case study in Chinese and its implications", in: Anderson — Jones (eds.) 199—217. 1974b "An explanation of word order change: SVO —• SOV", Foundations of Language 12: 2 0 1 - 2 1 4 . 1976 "Subject and topic: A new typology of language", in: Li (ed.) (1976), 457-489. 1980 "Synchrony and diachrony: The Mandarin comparative", Folia Linguistica Historica 1: 231 —250. Lightfoot, David W. 1976 "The base component as a locus of syntactic change", in: Christie (ed.), 17-32. 1977 "Syntactic change and the autonomy thesis", Journal of Linguistics 13: 191-216. 1979 Principles of diachronic syntax. Cambridge: Cambridge University Press.

386

Winfred P. Lehmann

Lüdtke, Helmut 1980 "The place of morphology in a universal cybernetic theory of language change", in: Fisiak (ed.), 273-281. Mäher, J. Peter — Allan R. Bomhard — Konrad Koerner (eds.) 1981 Papers from the Third International Conference on Historical Linguistics, 22 — 26 Aug 1977, Hamburg. Amsterdam: Benjamins. Makkai, Adam — V. B. Makkai 1976 "The nature of linguistic change and modern linguistic theories", in: Christie (ed.), 235-259. McCone, Kim P. 1979 "The diachronic possibilities of the Indo-European 'amplified' sentence: A case history from Anatolian", in: Brogyanyi (ed.), 467 — 487. Meillet, Antoine 1916 "Sur la methode ä employer en syntaxe", BSL 20: 133-137. 1938 and 1948 Linguistique historique et linguistique generale. 2 vols. Paris: Champion/ Klincksieck. 1937 Introduction a I'etude comparative des languages indo-europeennes. Paris: Hachette. (Reprinted 1964, University, Ala.: University of Alabama Press). 1967 The comparative method in historical linguistics [Transl. Gordon B. Ford, Jr.]. Paris: Champion. Nadkarni, M. C. 1975 "Bilingualism and syntactic change in Konkani", Language 51: 672 — 683. Nagy, Gregory 1974 Review of The language of Hesiod in its traditional context, by G. P. Edwards (1971), The Canadian Journal of Linguistics 25: 219 — 224. Parker, Frank 1980 "Typology and word order change", Linguistics 18: 269 — 288. Paul, Hermann 1920 Prinzipien der Sprachgeschichte. Halle: Niemeyer. Porzig, Walter 1924 "Aufgaben der indogermanischen Syntax", in: Stand und Aufgaben der Sprachwissenschaft. Festschrift für Wilhelm Streitberg, 126 — 151. Heidelberg: Winter. Prokosch, Eduard 1939 A comparative Germanic grammar. Philadelphia: University of Pennsylvania Press. Prosdocimi, Aldo L. 1978 "Diacronia; ricostruzione. Genera proxima e differentia specifica", LeSt 13: 335-371. Ramat, Paolo 1976 "Linguistic reconstruction and typology", JIES 4: 189 — 206. Ramat, Paolo (ed.) 1980 Linguistic reconstruction and Indo-European syntax. Amsterdam: Benjamins. Ratanajoti, Hundirapola 1975 "The syntactic structure of Sinhalese and its relation to that of the other Indo-Aryan dialects", Dissertation, University of Texas at Austin.

Syntactic change

387

Romaine, Suzanne 1981 "Syntactic complexity, relativization, and stylistic levels in Middle Scots", Folia Linguistica Historica 2: 71 —97. Sandfeld, Kristian 1930 Linguistique balkanique. Paris: Champion. Schmidt, P. Wilhelm 1926 Die Sprachfamilien und Sprachenkreise der Erde. Heidelberg: Winter. Schweisthal, K. G. (ed.) 1971 Grammatik, Kybernetik, Kommunikation. Bonn: Dümmler. Schwyzer, Eduard 1939 — 1953 Griechische Grammatik. 3 vols. München: Beck. Sebeok, Thomas A. (ed.) 1973 Diachronic, areal, and typological linguistics. Vol. XI of Current Trends in Linguistics. The Hague: Mouton. Shields, Kenneth 1979 "A theory of gender change", Glossa 13: 2 7 - 3 8 . Smith, Donald L. 1978 "Mirror images in Japanese and English", Language 54: 78 — 122. Steele, Susan 1975 "On some factors that affect and effect word order", in: Li (ed.) (1975), 199-268. 1977 "Clisis and diachrony", in: Li (ed.) (1977), 539-579. Steever, Sanford B. - C. A. Walker - S. S. Mufwende (eds.) 1976 Papers from the Parasession on Diachronic Syntax. Chicago Linguistic Society 12th Regional Meeting, 22 April 1976. Chicago: Chicago Linguistic Society. Sturtevant, Edgar Η. — E. Adelaide Hahn 1951 A comparative grammar of the Hittite language. New Haven: Yale University Press. Szantyr, Anton 1965 Lateinische Syntax und Stilistik. Vol. II of Lateinische Grammatik. Manu Leumann (ed.). München: Beck. Tai, James H.-Y. 1976 "On the change from SVO to SOV in Chinese", in: Steever et al. (eds.), 291-304. Timberlake, Alan 1977 "Reanalysis and actualization in syntactic change", in: Li (ed.) (1977), 141-177. Tischler, Johann 1979 "Zu den syntaktischen Grundlagen der Nominalkomposition: Die Reihenfolge der Teilgliedcr", in: Brogyanyi (ed.), 853 — 868. Traugott, Elizabeth Closs 1968 "Diachronic syntax and generative grammar", Language 41: 402—415. 1969 "Towards a grammar of syntactic change", Lingua 23: 1 —27. 1974 "On the notion of 'restructuring' in historical linguistics", in: Heilmann (ed.), 921-928. 1980 "Meaning change in the development of grammatical markers", Language Sciences 2: 44—61.

388

Winfred P. Lehmann

Traugott, Elizabeth — R. Labrum — S. Shepherd (eds.) 1980 Papers from the 4th International Conference on Historical Linguistics. 26—30 March 1979, Stanford. Amsterdam: Benjamins. Twaddell, W. Freeman 1938 "A note on Old High German umlaut", Monatshefte fur deutschen Unterricht 30: 177-181. Vennemann, Theo 1974a "Topics, subjects and word order: From SXV to SVX via TVX", in: Anderson — Jones (eds.), I: 339 — 379. 1974b "Analogy in generative grammar: The origin of word order", in: Heilmann (ed.), 7 9 - 8 3 . 1975 "An explanation of drift", in: Li (ed.), 264 — 305. 1976 "Categorial grammar and the order of meaningful elements", in: Juilland (ed.), 6 1 5 - 3 4 . Vincent, Nigel 1980 "Iconic and symbolic aspects of syntax: Prospects for reconstruction", in: Ramat (ed.), 4 1 - 6 8 . Wackernagel, Jacob 1926 — 1928 Vorlesungen über Syntax 1—2 (2nd edition.) Basel: Birkhäuser. Watkins, Calvert 1976 "Towards Proto-Indo-European syntax: Problems and pseudo-problems", in: Steever et al. (eds.), 305-326. Weinreich, Uriel — William Labov — Marvin I. Herzog 1968 "Empirical foundations for a theory of language change", in: Lehmann - M a l k i e l (eds.), 9 5 - 1 9 5 . Weijnen, A. — J. Kruijsen 1979 Atlas Linguarum Europae. Second questionnaire. Assen: van Gorcum. Whitney, William Dwight 1892 Language and the study of language (5th edition). New York: Scribner. Wurm, S. A. - D. C. Laycock (eds.) 1970 Pacific linguistic studies in honor of Arthur Capell (Pacific Linguistics, Series C, No. 13). Canberra: Australian National University.

Lexical Change

Onomasiological change: Sachen-change reflected by Wörter Ladislav

Zgusta

Whereas semantics studies the problems of meaning, with linguistic phenomena being the starting point of the investigation, onomasiology proceeds from the "things" to the expressions that denote them. This statement immediately shows one of the strong limitations of onomasiology: It is restricted to the study of lexical phenomena (sentences, texts, speech acts, etc., are not studied), and it usually focuses on expressions with a denotative or representational function, preferably of fully or at least semi-terminological character. Nor does it ever reach the philosophical or psychological depth of interpretation frequently attained in semantic studies. It can be conceived as a rather applied and sometimes immediately practical branch of linguistics which investigates naming within coherent segments of reality; for instance, agricultural terminology, the names of diseases, words relating to sport activities, etc. In a more theoretical understanding, it can be perceived as based either on the Saussurean idea that a (lexical) unit's 'valeur' is determined by its relations to the adjacent (lexical) units, or on philosophical and/or scientific considerations of the relation between 'things' and their taxonomies, and words. Considered from this point of view, an "onomasiological lexical change" is caused by changes in the denoted things, which can belong to any of the following three categories: 1) new things come into existence or are discovered; 2) things cease to exist and/or fall into oblivion; 3) things change.

390

Ladislav

Zgusta

1. New things come into existence or are discovered When a new thing comes into existence, is invented, discovered or taken cognizance of, usually a new expression denoting it comes into existence as well. Such an addition to the lexicon clearly must be considered lexical change. The most frequent sources of such new expressions are the following: a) An already existing word is endowed with a new meaning, frequently in addition to its older senses. For instance, with the arrival of various mechanical gadgets and a widespread use of metallic screws, the word nut developed a new sense, that of the thing which fits around and secures a screw. Any of the changes of lexical meaning known in semantics — narrowing of the sense, metaphor, metonymy, etc. — can take place in this situation. b) A word that had existed in the language but became obsolete or was lost is revived and endowed with the new sense. For instance, turnpike road, usually shortened to turnpike, was an obsolete word denoting something archaic by the late nineteenth century; however, it was revived with the construction of express highways in the United States, some of which are financed by toll. c) A new word comes into existence by any word-formational process the respective language has (most frequently it is derivation and composition) or a collocation is lexically stabilized to yield the necessary multiword lexical unit (or composite lexeme): for instance, derivations like Amer. French calumet 'Indian pipe' (from Picardian calumeau, French chalumeau 'straw'); compounds like railway, multiword lexical units like express train, French pomme de terre. Such collocations tend to be of descriptive character. d) The newly coined descriptive multiword lexical unit is frequently abbreviated, namely 1) by ellipsis (e.g., turnpike road > turnpike', French chemin de fer metropolitain, the official designation of the Paris subway > m0tropolitain; Czech Röntgenovy paprsky 'R's rays' = 'Xrays' > röntgen, rentgen 'X-rays'); 2) by truncation (e. g., preparatory school > prep in informal language, which involves both ellipsis and truncation; confidence man 'professional swindler', first attested by the Oxford English Dictionary in 1884, soon truncated into con man; the best example is, however, French metropolitain > metro 'subway', because there is no doubt as to the onomasiological novelty of the thing in the late nineteenth century 1 ; or 3) by acronymy (e.g., radio

Onomasiological

change

391

detecting and ranging > radar). Some of the portmanteau and other similar words also belong here. e) A word is borrowed, usually from the language of the people through whom acquaintance with the thing was made. For instance, potato, tomahawk come from languages of the American Indians. f) The necessary word is sometimes not borrowed but loan-translated (or calqued); that is, its morpheme-to-morpheme translation is endowed with the overall lexical meaning it has in the source language. For instance, Fr. pomme de terre 'apple of the soil' = 'potato' > Germ. Erdäpfel 'soil-apple' = 'potato'. These loan-translations frequently are only partial, combined with some indigenous word-formational material, or the source language serves as a model only. For instance, Engl, railway > Fr. chemin de fer 'way of iron' = 'railway' (rail is the model for fer) > Germ. Eisenbahn 'iron way' = 'railway' (exact calquing) > Czech zeleznice 'railway' (zelezo 'iron' + -ni-ce, indigenous derivational suffixes). Sometimes the source language is the model for the development of a new sense: for instance, Greek ptosis 'fall; case (gram.)' > Lat. casus 'fall' and later 'case (gram.)'. g) Proper names are a frequent source of new expressions. Typically, we have cases of naming from the place of origin (e.g., china, denim < de Nimes, jeans < Genoese), from the inventor {Röntgen's rays — see above [d]), or for purely honorific purposes (polonium for Poland; volt for Count Volta). h) The new expression is coined on the basis of existing linguistic material, but without regard for the normal derivational processes. For instance, the Czech playwright Capek created the word robot around 1920; it is inspired by Czech robota 'heavy, compulsory work', but cannot be regularly derived from it. (From Czech, it went into English and other languages as a borrowing.) Gas was invented by van Helmont in the seventeenth century, its inspiration being Greek chaos. i) It seems to happen only infrequently that a new word is coined quite freely, without a graspable source. The brand name Kodak, which by now approaches the status of a general noun (at least in some languages into which it has been borrowed), is said to be such a free creation. What onomatopoeic, sound-symbolic, or synesthetic stimuli might have influenced such a coinage remains mostly unknown. Although the coining of new words and expressions can be observed, given sufficient data, in any language in any space of time, there are

392

Ladislav

Zgusta

two situations in which new coinages are extremely frequent. The first of them obtains when a cultural community comes into a sudden and extensive contact with another, radically different culture. The study of the resulting acculturation processes in the lexicon is one of the important tasks of anthropological linguistics. The second situation obtains when some areas of human activity, particularly scientific enquiry and technology, expand rapidly, with many discoveries and inventions being made constantly. This latter situation entails the increased importance of terminology. This word can be used either in reference to a set or sets of defined technical terms, or to their study. The former form an important membership class in the lexicon of a language; the latter is an important branch of applied linguistics. It has some specificities of its own: It is so strongly normative that some states create authoritative bodies endowed with the exclusive right to coin and define technical terms whose use is obligatory in that state; and it is concerned not only with terms already existing, but also with normative rules of how to coin new terms in the future. Scientific terminology has specificities of its own. For instance, in many European languages, Greek and Latin morphemes are frequently used for terminological purposes. The two situations where new coinages are frequent often overlap, typically in states which were emancipated from the political rule of other states and which try to introduce their own language as the medium of administration, culture, and all the other registers. In such cases, another branch of applied linguistics is involved, namely, language planning. Considered from this point of view, the respective goal — namely, the modernization of language, in general, and more specifically, the introduction of the necessary new terms — can be conceived of as a change planned for the future. A branch of applied linguistics adjacent to terminology is the study of brand names and their coinage. Brand names are sometimes closer to proper names (e. g., the detergent Ajax) and less often to a general noun (xerox: Greek xeros 'dry'), so that their study tends to belong instead to onomastics. However, the coinage of brand names for artifacts like medications has onomasiological aspects. It may go without saying that the "new things" referred to above can be immaterial, i.e., abstract. Such "abstract new things" again can be received from another culture. For instance, with the spread of Christianity, many words and expressions referring to notions like

Onomasiological

change

393

transubstantiation, misericordia, eucharistia were either borrowed or loan-translated into many other languages. However, "new abstract things" or notions also do develop within a culture. Typically, they have to do with new conceptualizations, i. e., with the realization that some phenomena and their underlying principles belong together, form a whole. For instance, while the differences between the literary and the colloquial variants of Greek, Arabic, and other languages had been known since times immemorial, it was Charles Ferguson who conceptualized them as a whole, calling them "diglossia". Or they have to do with a conceptualization of existing phenomena combined with new ones: thus was "cybernetics" established and named by N. Wiener. However, the terminological coinage may come late after the conceptualization. For instance, the problem of why an almighty and supremely loving God should allow the existence of evil has been known to Christian theologians since antiquity, with extensive discussions in such important and widely-read authors as St. Augustine and St. Thomas Aquinas, but the terminological coinage came only with Leibniz's French treatise on theodicee (borrowed in English as theodicy), which word he coined from Greek theos 'god' and dike 'justice' = 'justification of God'. Naturally, it is impossible to prove whether the conceptualization came only with the word or whether it was reached before it; however, the discussions of the problem are as coherent and as exhaustive before the coinage as afterwards, so that it would seem that at least the unified relation of the phenomena, their forming a whole, was grasped. Sometimes, very concrete "things" have to wait for a unified term as well. For instance, the type of highway constructed for the exclusive use of motor-cars, which is called It. autostrada, Germ. Autobahn, Fr. autoroute, Brit. Engl, motorway, has no unified term in American English: it is called freeway, expressway, limited access highway, or interstate (highway). Whether the lack of a unique, precise term testifies to the lack of a unified concept is a moot point. One more remark. A new conceptualization and terminological coinage does not necessarily have to have serious sources, or motives: In 1754, when H. Walpole conceptualized as a unified property the ability of the three fabular Princes of Ceylon ( = obsol. Serendip) to keep making lucky and unexpected discoveries by accident and called it serendipity, he was hardly engaging in scientific enquiry. Whether, however, the artificial coinage of a word like absquatulate 'to get away'

394

Ladislav

Zgusta

(from abscond and/or 'go and squat somewhere else') also belongs here, is a moot question: On the one hand, the word refers to no "new thing"; on the other hand, one could perhaps conceive of the desire to express oneself jocularly (for which purpose the word seems to have been coined) as an onomasiological force of its own.

2. Things cease to exist The opposite phenomenon consists of the fact that things, concrete or abstract, cease to exist and the respective words become obsolete. In antiquity, waxed tablets were a frequently-used material for writing; in Old English, they were called aestel. However, the use of these tablets and consequently also of the word was later discontinued. Many other names of garments, dishes, weapons, but also of institutions, beliefs, superstitions are lost when culture changes. Such lexical loss has, however, various degrees. Nobody will organize an old-style auto-da-fe in our days, and so the words denoting the garments of the condemned, sambenito and samarra, originally borrowed from Spanish, are lost from today's language, both in Spanish and in English; however, they occur in texts which, in contrast to Old English, can be read by a person without special linguistic training, and so they still belong to the passive vocabulary of some speakers of Modern English. There is also the following difference: a knowledgeable person who talks about auto-da-fe to other experts will not hesitate to use the words sambenito and samarra when referring to the garments, thereby reviving them for the purpose of discourse in that register. On the other hand no expert on Old English realia will use aestel for reference; he will speak about the "wax tablets" or use some other descriptive phrase, and if he mentions aestel at all, then it is only to inform his hearers as to what the old word was, not for direct reference in his discourse. This example shows that the "loss" of the thing is not a sufficient condition for the total loss of the word; the intelligibility of the language of the respective texts as a whole seems to be another decisive factor. But in each of these cases, one must realize that, at least theoretically, any old word can be revived by the process illustrated above in lb). Also, it may happen that an author of a, say, historical novel will use such forgotten words in his text to give it the proper "atmosphere"; we know several such cases from the works of Sir Walter Scott. Such

Onomasiological change

395

a word is then present at least in the passive lexicons of the readers of those texts. One more remark: sometimes only one of the senses of a polysemous word is lost. For instance, with the disappearance of the steam locomotive, one of the senses of tender, namely, 'railroad car attached to the locomotive and designed to carry fuel' is getting obsolete and may eventually be lost.

3. Change in the denoted thing The third type of onomasiological change is caused by a change in the denoted thing. For instance, the word car referred during the entire nineteenth century either exclusively or prevalently to a vehicle normally driven by animals, but in the twentieth century its nearly exclusive reference is a motor car. The horn originally was a musical instrument made from animal horns, but today it is made from metallic alloys only. This type of change is frequently called substitution. Many linguists hesitate to accept it as an onomasiological change. Indeed, if we take the meaning of the word car to be 'vehicle for transportation of persons, animals, and goods', then no change of the meaning occurred at all. Similarly, if horn is 'a sort of trumpet with a certain color of sound', nothing happened. Note that in both cases the broader definition captures the unchanging function, or purpose, of the thing, whereas the narrower definitions seem to deal with accidentals. This is why we, without compunction, sail on ships propelled by motors. Still, there is, at least historically, a slight difference, because horn is derivationally tied to that part of the animal's body, whereas car lacks any derivational motivation. Yet derivational motivation usually gets lost (as in to sail), so that one can suppose that the substitution involved in horn also took place after speakers had lost awareness of the derivation. Such a loss of derivational motivation is itself a phenomenon sometimes related to the substitution. The smallest particle of matter was called ätomos by the Greek philosophers, the morphemes of which suggest the meaning 'indivisible', and this term was borrowed into modern physics. The atom was split at an age when most scientists knew Greek, but the term remained, and today not even a Greek scholar feels a sense of contradiction in the collocation subatomic particles.

396

Ladislav

Zgusta

Sometimes, however, the motivation is not lost, and if there is a change in the thing, the word changes as well. An example from our days: To save some expense with typesetting, the numbered notes that accompany the text of a paper, the footnotes of more affluent times, are frequently set at the end of the article, chapter, or book. In some editions, they pass on as footnotes (loss of derivational motivation), but in some they already are called endnotes (a new word, derivationally motivated; e.g., ANS Bulletin No. 64). Most examples mentioned in the preceding survey are drawn from English and other modern languages. However, it would be easy to find examples for la), b), c), e), f)> 2), and 3) in many other languages — if not all — the frequency of the single types depending both on cultural developments and on the structure of the single languages. It might prove difficult to find frequent examples for Id), g), and h) in old languages and in some modern ones; but, for instance, the technical terminology of ancient native Sanskrit linguistics abounds with such coinages.

Note 1 Interestingly, both forms were borrowed into Russian. In everyday language, spoken and written, 'subway' is metro', in more official texts it is metropoliten (with the non-palatalizing e, here spelt as e). Metro has been borrowed into a number of other languages; the full version of the French multiword lexical unit has never been borrowed anywhere else.

References a)

Onomasiology

de Gorog, Ralph 1981 "The application of onomasiology to synonymy, word formation, and etymology", Word 32: 9 9 - 1 0 8 . Heger, Klaus 1964 "Die methodologischen Voraussetzungen von Onomasiologie und begrifflicher Gliederung". Zeitschrift für romanische Philologie 80: 486 — 516. Quadri, Bruno 1952 Aufgaben und Methoden der onomasiologischen Forschung (Romania Helvetica 37). Bern: Francke. Wiegand, Herbert Ernst 1970 "Synchronische Onomasiologie und Semasiologie", Germanistische Linguistik 3: 243 — 267.

Onomasiological change

397

b) Onomasiology within lexicology Reichmann, Oskar 1976 Germanistische Lexikologie. Stuttgart: J. B. Metzler. c) Acculturation, changes of lexical units Zgusta, Ladislav 1971 Manual of lexicography. The Hague/Paris: Mouton. d)

Terminology.

Felber, Helmut 1985 Manual of terminology. Paris: UNESCO. Rey, Alain 1979 La terminologie: noms et notions (Que sais-je, 1730). Paris: Presses Universitaires de France. Riggs, Fred W. (ed.) 1981 The COCTA Conference: Proceedings of the Conference on Conceptual and Terminological Analysis in the Social Sciences. Frankfurt am Main. Indeks Verlag. η. d. Interconcept report: A new paradigm for solving terminology problems of the social sciences (Reports and Papers in the Social Sciences 47). n. p.: UNESCO. n. d. Help for social scientists: A new kind of reference process (Reports and Papers in the Social Sciences 57). n. p.: UNESCO. Sager, Juan — D. Dungworth — P. F. McDonald 1980 English special languages. Wiesbaden: Oscar Brandstetter. Wüster, Ε. 1974 "Die allgemeine Terminologielehre — ein Grenzgebiet zwischen Sprachwissenschaft, Logik, Ontologie, Informatik und den Sachwissenschaften", Linguistics 119: 61 — 106. e) Language planning Rubin, Joan — Björn Jernudd 1971 Can language be planned? Sociolinguistic theory and practice for developing nations. Honolulu: University of Hawaii Press. Rubin, Joan — Björn Η. Jernudd — Jyotirindra Das Gupta — Joshua A. Fishman — Charles A. Ferguson (eds.) 1977 Language planning processes (Contributions to the Sociology of Language Vol. 21). The Hague/Paris/New York: Mouton. Rubin, Joan — Roger Shuy (eds.) 1973 Language planning: Current issues and research. Washington: Georgetown University Press. f ) Historical processes Algeo, John 1977 (1980) "Blends, a structural and systemic view", American Speech 52: 47 — 64. 1978 (1980) "The taxonomy of word making", Word 29: 122-131. (63 different types of word making are discerned.) 1980 "Where do all the new words come from?" American Speech 55: 264 — 277.

398

Ladislav Zgusta

Fisiak, Jacek (ed.) 1985 Historical semantics — Historical word-formation. Berlin/New York: Mouton. Gusmani, Roberto 1984 "A proposito della motivazione linguistica", Incontri linguistici 9: 11 —23. g) Semantic processes Coseriu, Eugenio 1981 "Les procedes semantiques dans la formation des mots", Cahiers Ferdinand de Saussure 35: 3 — 16. Stern, Gustaf 1964 Meaning and change of meaning, with special reference to the English language. Bloomington: Indiana University Press. (First published Göteborg: Elanders, 1931.) h) Borrowing and loan translation Gusmani, Roberto 1981 Saggi sull' interferenza linguistica. Firenze: Olschki. i) Archaic and obsolete words Bax, Marcel 1983 "Die lebendige Dimension toter Sprachen. Zur pragmatischen Analyse von Sprachgebrauch in historischen Kontexten", Zeitschrift fur Germanistische Linguistik 11: 1—21. j) General information Numerous analyses of concrete cases and some theoretical discussions can be found in all volumes of the journal Wörter und Sachen.

Semantic change* John Alge ο

The first studies of semantics were historical. The term "semantic", although used as early as 1665 in reference to predicting the weather from signs, was popularized in English through the 1900 translation of Michel Breal's Semantics: Studies in the science of meaning. Breal had proposed the term semantique as early as 1883 and the English equivalent had turned up by 1893 (Read 1948; Burchfield 1972-86). Although reacting against the Neogrammarian preoccupation with phonetic changes (which he superciliously referred to as "physiological grammar"), Breal followed the Neogrammarian model in looking for "laws" that govern semantic change in language. Breal's laws were, however, universal principles, not descriptions of a change in a particular language at a particular time. Thus they were more analogous to natural laws than to Neogrammarian linguistic ones. They were primarily a taxonomic device for ordering the chaos of meaning changes. Historical studies of semantic change reached an apogee with the 1931 publication of Gustaf Stern's Meaning and change of meaning. Earlier historical studies had been principally concerned with classifying the kinds of semantic change and especially with looking for psychological and cultural reasons for change (Breal's laws). Stern continued that trend. By examining words that have shifted their meanings, he arrived at an a posteriori, sevenfold classification of semantic change: 1) Substitution, sense change with a nonlinguistic cause — a change in the referent or in our knowledge of it or in our attitude toward it. * I am grateful to two of my colleagues, Jared Klein and Salikoko Mufwene, who read and made suggestions for "Semantic change" and "Borrowing".

400

John Algeo

For example, ship changed its meaning when its referent changed from a vessel propelled by wind power to one propelled by steam. 2) Analogy, of which one type is popular etymology. For example, fruition originally meant 'enjoyment' but because it sounds like fruit, it came to mean 'bearing fruit, completion' (Rex 1969). 3) Shortening. When the blue devils 'melancholy' was shortened to the blues, the color term acquired a new meaning, or at least had one aspect of its meaning reinforced (Morris 1976). 4) Nomination, the intentional naming of a referent with a term not previously used for it. Murray Gell-Mann's use of quark for an elementary subatomic particle is an example of this class, as are most instances of deliberate metaphor, euphemism, and other figures of speech. 5) Transfer, the use of a word for a new referent that somehow is similar to older referents of the word. Spontaneous metaphors, such as synesthesia — a sharp taste, a dark sound, a sweet sight — are examples. 6) Permutation, a meaning change resulting from a shift in focus. For example, want shifted from 'lack (something desirable)' to 'desire'. Shifts from the concrete to the abstract or vice versa and some forms of metonymy belong here. 7) Adequation, a shift in which characteristics of the referent are considered central to the meaning of the word. This class includes generalization, restriction, amelioration, and pejoration. The decline of interest in diachronic semantics that followed Stern's work can be related to a general lack of interest in historical matters on the part of structuralists and to the disinclination of American structuralists to concern themselves with any except differential meaning. Leonard Bloomfleld (1933), however, summarized earlier classifications of semantic change under a number of categories that have been widely used in later works. Changes related to scope are narrowing {meat 'food' to 'edible flesh') and widening {bird 'young birdling' to 'feathered vertebrate'). Changes resulting from figures of speech are metaphor {bitter 'biting' to 'harsh of taste'), metonymy {cheek 'jaw' to 'side of face'), synecdoche {silver 'kind of metal' to 'table utensils'), hyperbole {starve 'die of hunger' to 'be hungry'), and litotes or understatement {lust 'pleasure' to 'violent desire'). Changes in the emotional association of words are degeneration {knave 'servant boy' to 'rogue') and elevation {knight 'boy' to 'man of military rank below baron').

Semantic change

401

The most assiduous of recent historical semanticists has been Stephen Ullmann, whose classification has an a priori and therefore neater basis for division than did Stern's or Bloomfield's. Ullmann (1951, 1962) has sorted semantic change into the following types: I. Semantic change due to linguistic conservation (when the referent of the word changes, but the word is retained) II. Semantic change due to linguistic innovation A. Transfer of names through 1. Similarity of senses (metaphor) 2. Contiguity of senses (metonymy) B. Transfer of senses through 1. Similarity of names (popular etymology) 2. Contiguity of names (ellipsis) Without an independently motivated theory of semantics, all taxonomies of semantic change seem somewhat forced. Ullmann (1972) has written a useful history of semantic taxonomies in Western Europe, which shared the decline in historical interest, though not the neglect of semantics, that was typical of American structuralism. However, recently a new interest in historical semantics has been discernable (Lehmann 1976), and even Breal's laws have been given a fresh lease by the renewed attention to language universale (Ullmann 1966). In addition to the intrinsic interest of semantic universale as indications of the structure of the human mind, they are important for giving what Leonard Bloomfield (1933: 430) called "some measure of probability by which we can judge of etymologic comparisons." In historical reconstruction, the sound laws of the Neogrammarians provide a basis for deciding whether two forms are cognates, but in addition the forms must be related semantically. They must be the same in meaning or have meanings one of which can plausibly be a historical development of the other or both of which can plausibly be derived from some common earlier meaning. Universale of meaning and of meaning change (even though only statistical probabilities) are necessary for reliable linguistic reconstruction. The difficulty of arriving at convincing universale, however, is notorious. For one thing, it is important that the data to be described be kept distinct from the language of the describer, as Emile Benveniste has convincingly shown (1954). According to John Lyons (1977: 620), the three most important developments in diachronic semantics since the 1920s are these:

402

John Algeo

1) "The application of the principles of structuralism in tracing the history of particular semantic fields" (cf. Lyons 1977: 250 — 261). Jost Trier had proposed that the meaning of a term can be dealt with only in comparison with the meanings of other terms belonging to the same lexical field, illustrating his point with a study of terms for 'intellect'. The same point can be seen in the color spectrum: What red means depends partly on the meaning of orange, purple, pink, and brown; individual colors have to be defined with respect to the whole range of colors in the spectrum. Brent Berlin and Paul Kay (1969) have studied color systems with respect to universale and implicational hierarchies. Since the cardinal doctrine of structuralism is that a language is "un systeme ού tout se dent", each item in the system (such as the meaning of a particular word) has value and can be defined only as it relates to other items in the system. When English replaced its native earn 'mother's brother' and fcedera 'father's brother' with the single term uncle, the change was part of a restructuring of the lexical field of family relationship (Lehmann 1976: 8). Structurally what is important is not the change of meaning in an isolated word, but changes in a whole field. 2) "The implementation of the principle that the history of the vocabulary of a language cannot be studied independently of the social, economic, and cultural history of the people." Although during the twentieth century much emphasis has been placed on the autonomy of linguistics (with phonetics and semantics sometimes viewed as peripheral disciplines, if not downright unlinguistic ones), it seems clear that the meanings of words cannot be studied apart from the nonlinguistic contexts in which words are used. The insistence on treating semantic and cultural change together is sometimes known as the words-and-things (Wörter und Sachen) movement. 3) "The realization that diachronic and dialectal variation are ultimately inseparable." Lyons could have made an even stronger statement. It is not merely dialect variation but synchronic variation of any kind that is inseparable from historical change. It is true that diachronic semantics, like all the rest of historical linguistics, presupposes an adequate synchronic description of the historical stages to be studied. It is necessary to distinguish the two approaches to language study; failure to do so results in the "etymological fallacy", by which the historically-minded prescriptivist may suppose that, because disheveled once meant 'with disarranged hair'

Semantic change

403

(OF chevel 'hair'), a bald man cannot be so described, or that, because disinterested once meant 'impartial', it may not now mean 'unconcerned'. (Never mind that 'unconcerned' is what the word first meant in English — the view of historical prescriptivists seldom focuses beyond their immediate concern.) When two forms are the same in pronunciation but different in meaning, we have to decide whether they are the same word with a range of meanings (polysemy) or whether they are different words that happen to be pronounced alike (homonymy). Both alternatives are examples of ambiguity — polysemy as the result of semantic change and homonymy often as the result of phonological change. The history that produces the two sorts of ambiguity is, however, often at variance with, and irrelevant to, the perception of present-day speakers. For example, Lyons (1981: 46) points out that most persons today would probably think of sole 'bottom of the foot' and sole 'a kind of fish' as different words because of their sharp divergence in sense; yet historically they are from the same etymon (Latin solea 'sandal', the fish being named for its shape) and are thus instances of polysemy rather than of homonymy. On the other hand, shock (of grain) and shock (of hair) are probably usually thought of as the same word — the latter a metaphoric use of the former; however, the words have different histories, being homonyms that have semantically converged in the popular imagination. In these cases diachronic and synchronic facts conflict and must therefore be carefully distinguished. On the other hand, it is also true that a too rigid separation of synchronic and diachronic studies is misguided. In any etat de langue, a multitude of variants — pronunciations, words, grammatical constructions — will be competing with one another. It is unrealistic to regard any language as the competence of an idealized speaker-hearer whose usage admits no alternative forms. A language without variants is the dream of a logician or other such visionary. In real language we constantly face a choice among competing forms, which are actually changes that seem static only because they have been recorded by the camera of synchronic description. If nature abhors a vacuum, language abhors a crowd. When two words develop similar forms, one of the words usually disappears, with a consequent change in the meaning of some other word that takes over the functions of the one crowded out. A frequently cited example of this process is the development in Gascony of the Latin cattus 'cat' and gallus 'cock'. The two forms by regular sound change

404

John Algeo

fell together as gat, creating an awkward homonymy. As a result, gat was retained only for the sense 'cat' and a variety of new words were introduced for the sense 'cock'. The influence of "homonymic clash" on semantics has been the subject of studies by Edna Williams (1944) and John Orr (1962). We might also talk of "polysemic clash". When a word like gay adds to its older meaning 'joyous' the newer sense 'homosexual', the older meaning is threatened. Younger speakers now seldom use the word in its earlier meaning. In such homonymy and polysemy, synchronic use and history overlap. Most people are also aware of language variation among the generations, thinking of some forms as "old-fashioned" and of others as "new-fangled". That awareness, which Lyons has called "a kind of diachrony-in-synchrony" (1977: 621), fudges the distinction between the two approaches to language for it means that speakers usually have a sense that their language is changing and are even aware of the direction of change. Thus a single etat de langue includes historical development within it. To sever descriptive from historical linguistics completely is to produce an incomplete description. Many recent introductory books on semantics include no treatment of historical change, or at most only a casual reference. Yet because of the dependence of diachronic study on synchronic description, some of those introductions provide useful background information for the student of historical semantics. George L. Dillon's Introduction to contemporary linguistic semantics (1977) and Janet Dean Fodor's Semantics: Theories of meaning in generative grammar (1977) are two such works. In ignoring diachrony they faithfully reflect the prejudice of much contemporary American semantic study, a prejudice that has been reinforced by a shift of interest during the last twenty years away from lexical semantics toward propositional semantics, thus emphasizing grammar at the expense of lexis. Joseph Voyles (1973) and Paul Werth (1974) have attempted to account for semantic change within new theories of grammar. However, they seem mainly to have restated traditional schemata in a generative mode, rather than to have arrived at an independent explanation based on new theoretical assumptions. Two other works that also give very useful surveys of synchronic semantics but touch upon historical matters lightly are Frank R. Palmer's Semantics: A new outline (1981), which grew out of a series of lectures he gave at the Linguistic Institute at Buffalo, New York, in the summer of 1971, and Geoffrey Leech's Semantics (1981), which

Semantic change

405

aims both at giving a general orientation to the subject and at developing a coherent theory of semantics. A useful survey of the field of semantics that does pay a good deal of attention to historical matters is R. A. Waldron's Sense and sense development (1979). The second half of Waldron's work is devoted to diachrony, including a useful summary of important earlier studies and a survey of types of semantic change. The most comprehensive treatment of semantics with some attention to historical matters is John Lyons's two-volume study, Semantics (1977). Short introductions to historical semantics can be found in a number of surveys of linguistics or of the history of particular languages (for example, Bloomfield 1933; Anttila 1972; Arlotto 1972; L. Palmer 1972; Lehmann 1973; Pyles - Algeo 1982). Historical semantics has affinities with other disciplines, especially with literary studies. James Bradstreet Greenough and George Lyman Kittredge's Words and their ways in English speech (1901), though dated, shows what can be achieved when a great classicist and a great scholar of English literature pool their talents to observe the lexicon of their native language. James Copley's Shift of meaning (1961) discusses 250 words, tracing their changes in meaning from Elizabethan times to the present day. The book is useful to the student of earlier literature, perhaps as much as an admonition to be conscious of changing meanings as for the information it gives about the 250 words it treats. C. S. Lewis's Studies in words (1967) looks in detail at a number of words whose changes in meaning mark different intellectual periods in the history of English society. Susie I. Tucker's Enthusiasm (1972) is a similar, but even more detailed examination of the history of a single word. Philip Howard's New words for old (1977) consists of 43 short, popular essays on changes in the use of words — some essays devoted to a single vogue word, others to a kind of change, and others to a lexical field. It is directed to the general reader. Similarly, The private lives of English words (1984) by Louis Heller, Alexander Humez, and Malcah Dror describes notable developments in the meanings of over 400 words. These works use sound scholarship to capture the attention of the layperson with an interest in word histories. They thus establish a much-needed link between the scholar and the general public.

406

John Algeo

References Anderson, John M. — Charles Jones (eds.) 1974 Syntax, morphology, internal and comparative reconstruction. Vol. I of Historical linguistics. Proceedings of the First International Conference on Historical Linguistics, 2—7 September 1973, Edinburgh (North-Holland Linguistics Series 12a). Amsterdam: North-Holland. Anttila, Raimo 1972 An introduction to historical and comparative linguistics (esp. 133 — 153: "Semantic change")· New York: Macmillan. Arlotto, Anthony 1972 Introduction to historical linguistics (esp. 165 — 183: "Semantic change"). Boston: Houghton Mifflin. Benveniste, Emile 1971 Problems in general linguistics (esp. 249 — 264: "Semantic problems in reconstruction"). Trans. Mary Elizabeth Meek. Coral Gables, Fla.: University of Miami Press. (First published in French in Word 10 [1954]: 251-264.) Berlin, Brent — Paul Kay 1969 Basic color terms: Their universality and evolution. Berkeley: University of California Press. Bloomfield, Leonard 1933 Language (esp. 425 — 443: "Semantic change"). New York: Holt. Breal, Michel 1900 Semantics: Studies in the science of meaning. Trans. Mrs. Henry Cust. London: Heinemann. (First published as Essai de semantique. Paris: Hachette, 1897.) Burchfield, R. W. (ed.) 1972 — 1986 A supplement to the Oxford English Dictionary. 4 vols. Oxford: Clarendon. Copley, James 1961 Shift of meaning. London: Oxford University Press. Dillon, George L. 1977 Introduction to contemporary linguistic semantics. Englewood Cliffs, N.J.: Prentice-Hall. Fodor, Janet Dean 1977 Semantics: Theories of meaning in generative grammar. New York: Crowell. Greenberg, Joseph H. (ed.) 1966 Universals of language (2nd edition). Cambridge, Mass.: M.I.T. Press. Greenough, James Bradstreet — George Lyman Kittredge 1901 Words and their ways in English speech. New York: Macmillan. Heller, Louis — Alexander Humez — Malcah Dror 1984 The private lives of English words. Foreword by Geoffrey Leech. London: Routledge. Howard, Philip 1977 New words for old. New York: Oxford University Press.

Semantic change

407

Leech, Geoffrey 1981 Semantics (2nd edition). Harmondsworth, Middlesex: Penguin. Lehmann, Winfred P. 1973 Historical linguistics: An introduction (esp. 205 — 235: "Semantic change and changes in the lexicon") (2nd edition). New York: Holt, Rinehart and Winston. 1976 "Diachronic semantics: 1976", in: Rameh (ed.), 1 - 1 3 . Lewis, C. S. 1967 Studies in words (2nd edition). Cambridge: Cambridge University Press. Lyons, John 1977 Semantics. 2 vols. Cambridge: Cambridge University Press. 1981 Language, meaning and context. London: Fontana. Morris, Kenneth M. 1976 "Blue as a marker of intensification". American Speech 51: 35—44. Orr, John 1962 Three studies on homonymics. Edinburgh: Edinburgh University Press. Palmer, Frank R. 1981 Semantics: A new outline (2nd edition). Cambridge: Cambridge University Press. Palmer, Leonard R. 1972 Descriptive and comparative linguistics: A critical introduction (esp. 300 — 340: "Etymology and change of meaning"). New York: Crane, Russak. Pyles, Thomas — John Algeo 1982 The origins and development of the English language (esp. 238 — 259: "Words and meanings") (3rd edition). New York: Harcourt Brace Jovanovich. Rameh, Clea (ed.) 1976 Semantics: Theory and application. Georgetown University Round Table on Languages and Linguistics 1976. Washington, D.C.: Georgetown University Press. Read, Allen Walker 1948 "An account of the word 'Semantics'", Word 4: 7 8 - 9 7 . Rex, Richard 1969 "Five hundred years of fruition", American Speech 44: 179 — 190. Sebeok, Thomas A. (ed.) 1972 Linguistics in Western Europe. Vol. IX of Current trends in linguistics. The Hague: Mouton. Stern, Gustaf η. d. Meaning and change of meaning, with special reference to the English language. Bloomington: Indiana University Press. (First published Gothenburg: Elanders Boktryckeri Aktiebolag, 1931). Tucker, Susie I. 1972 Enthusiasm: A study in semantic change. Cambridge: Cambridge University Press. Ullmann, Stephen 1957 The principles of semantics (2nd edition). New York: Philosophical Library. (First published Glasgow: Jackson, 1951.)

408

John Algeo

1962 Semantics: An introduction to the science of meaning. Oxford: Blackwell. 1966 "Semantic universale", in: Greenberg (ed.), 217 — 262. 1972 "Semantics", in: Sebeok (ed.), 343-394. Voyles, Joseph B. 1973 "Accounting for semantic change", Lingua 31: 95 — 124. Waldron, R. A. 1979 Sense and sense development (2nd edition). London: Deutsch. Werth, Paul Ν. 1974 "Accounting for semantic change in current linguistic theory", in: Anderson — Jones (eds.), 377 — 415. Williams, Edna Rees 1970 The conflict of homonyms in English. Hamden, Conn.: Archon. (First published New Haven: Yale University Press, 1944.)

Borrowing John Algeo

Borrowing, even more than semantic change, is correlated with cultural innovation, whether resulting from casual tourist contacts, military conquest, trade relations, or the most intimate exchanges in a bilingual community. The best theoretical study of the subject is still Einar Haugen's "The analysis of linguistic borrowing", in which he defines borrowing as "the attempted reproduction in one language of patterns previously found in another" (1950: 212). Borrowing involves either or both of two processes: substitution and importation. When English speakers borrowed the Yiddish word we spell chutzpah, they imported the morpheme. Some speakers also imported the initial voiceless velar fricative [x], whereas others substituted a native English [h] for it. Like sounds, morphemes can be substituted as well as imported; English imported the French son et lumiere, but it also substituted native morphemes to produce the competing form sound and light. These two processes are fundamental to a classification of the types of borrowing. Three types of borrowings have been generally recognized, which Haugen has called "loanwords", "loanshifts", and "loanblends": 1) Loanwords result from morphemic importation without morphemic substitution, for example, salsa 'a type of Caribbean dance music' imported from Spanish or gulag 'forced labor camp for political dissidents' imported from Russian. 2) Loanshifts result from substitution of native morphemes in a foreign pattern, without the importation of any foreign morpheme. According to The second Barnhart dictionary of new English (Barnhart — Steinmetz — Barnhart 1980), reeducation in the sense 'political indoctri-

410

John Algeo

nation' is a loanshift from Vietnamese hoc tap, which also appears, although much less frequently, as a loanword in English. Loanshifts are of two kinds: a. Loan translations are substitutions motivated by a similarity of meaning between the foreign model and the native substitute. According to The second Barnhart dictionary, sliding time is a loan translation from German Gleitzeit 'flexible time, variable working hours for the convenience of the employee', with the two English words as semantic equivalents of the two parts of the German compound. When the substitution involves a partial overlapping of the semantic ranges of the foreign model and the native substitute, the form is called a caique. Anttila (1972: 140) cites the example of American Finnish harja 'brush (for cleaning)', which has taken on the additional meaning 'paintbrush' through the influence of English brush. b. Semantic loans are substitutions motivated by a similarity of form between the foreign model and the native substitute. Haugen cites Portuguese grosseria 'rude remark' as acquiring under American influence the sense 'grocery'. When the foreign model and the native substitute are as semantically remote from each other as in this example, one is tempted to think of the type as a "loan mistranslation". 3) Loanblends are hybrid forms, part importation and part substitution. English coffee klatch is a loanblend from German Kaffeeklatsch, literally 'coffee gossip', with the first morpheme substituted and the second imported. In addition to Haugen's major study of borrowing, Uriel Weinreich's Languages in contact (1953) is an important theoretical treatment of the subject. Especially relevant is the section on "Lexical interference" (1953: 47 — 62). The way in which lexical borrowing influences the phonology of the reproducing language has been treated by Charles C. Fries and Kenneth L. Pike in their study of "Coexistent phonemic systems" (1949). Anna Granville Hatcher's Modern English wordformation and neo-Latin (1951) treats the influence of the word formation patterns of one language on those of others, specifically neoLatin copulative compounds on the modern vernaculars English, French, German, and Italian. Good short surveys of borrowing have been made as parts of introductory textbooks on general or historical linguistics by Leonard Bloomfield (1933), Charles F. Hockett (1958), Raimo Anttila (1972), Anthony Arlotto (1972), Winfred Lehmann (1973), and Theodora Bynon (1977).

Borrowing

411

The most complete study of the history of lexical borrowing in English is Mary S. Serjeantson's History of foreign words in English (1935); it is now badly out of date since it does not cover either borrowings made during the last fifty years or newer evidence about older borrowing. Serjeantson's work urgently needs to be updated. The introduction to A. J. Bliss's Dictionary offoreign words and phrases in current English (1966) is useful. A short article by Garland Cannon and Beatrice Mendez Egle, "New borrowings in English" (1979), is the most up-to-date study now available, although Cannon has underway a fuller treatment of the subject. Chapter 12 in Thomas Pyles and John Algeo's Origins and development of the English language (1982) is a brief overview of the history of English borrowing. The influence of individual foreign languages on the English word stock has been much studied. Steven M. Benjamin and Luanne von Schneidemesser (1979) have made a convenient bibliography of studies of German loanwords. Sol Steinmetz (1981) has considered the influence of Yiddish and Hebrew on the speech of the Jewish community of the United States. John Geipel (1971) has treated the extensive and important Scandinavian influence on earlier English. Garland Cannon (1981) has demonstrated that Japanese borrowings are more numerous, older, and more pervasive than one might have thought. G. Subba Rao (1954) wrote a model study of Indian loanwords in English, recently supplemented by the glossary of R. E. Hawkins (1984). Lorenzo Dow Turner (1949) did another model study of African loanwords. A useful adjunct to the study of the history of borrowing in English is A chronological English dictionary (1970) by Thomas Finkenstaedt, Ernst Leisi, and Dieter Wolff. Based on the Shorter Oxford English dictionary, it lists some 80000 words in the order of their earliest attested use. Although English is a great borrower, it is also a lender. Hans Galinsky (1980) has treated recent borrowings by German from American English. Stanley Aleong and Michel Cretien (1981) have studied the specialized vocabulary of Citizens' Band radio lingo in Canadian French, with a dominant influence from the United States. Lawrence M. Davis (1979) has examined the puristic reaction in Israel to English "contamination" of modern Hebrew. Robert W. Hellstrom (1976) has described an American dialect of Finnish with marked English influence — a speechway he dubs "Finglish". Donald L. Smith (1974) has reported an unusual influence of English on Japanese, by which loanwords from English are combined with one another in constructions

412

John Alge ο

English has never known; for example, saradu doggu, literally 'salad dog', is a sandwich consisting of salad greens in a hotdog bun. Perhaps the most unusual example of borrowing from English is that reported by Robbin Battison (1978), who describes the influence of English via finger spelling on American Sign Language. The studies mentioned here are illustrative of the variety and scope of work on borrowing. They are very far from an exhaustive list of such studies, which would require a volume many times the size of this entire book. They do, however, include some important works on the subject and a selection of useful and informative treatments of particular aspects.

References Aleong, Stanley — Michel Cretien 1981 "Can Smokey the Bear speak French? Adapting CB lingo in Canadian French", American Speech 56: 260 — 268. Anttila, Raimo 1972 An introduction to historical and comparative linguistics (esp. 154—178: "External change: Borrowing"). New York: Macmillan. Arlotto, Anthony 1972 Introduction to historical linguistics (esp. 184—195: "Borrowing"). Boston: Houghton, Mifflin. Barnhart, Clarence L. — Sol Steinmetz — Robert K. Barnhart 1980 The second Barnhart dictionary of new English. Bronxville, N.Y.: Barnhart. Battison, Robbin 1978 Lexical borrowing in American Sign Language. Silver Spring, Md.: Linstok. Benjamin, Steven M. — Luanne von Schneidemesser 1979 "German loanwords in American English: A bibliography of studies, 1872-1978", American Speech 54: 210-215. Bliss, Alan Joseph 1966 A dictionary of foreign words and phrases in current English. New York: Dutton. Bloomfield, Leonard 1933 Language (esp. 444 — 495: "Cultural borrowing", "Intimate borrowing", and "Dialect borrowing"). New York: Holt. Bynon, Theodora 1977 Historical linguistics (esp. 217 — 239: "Lexical borrowing"). Cambridge: Cambridge University Press. Cannon, Garland 1981 "Japanese borrowings in English", American Speech 56: 190 — 206. Cannon, Garland — Beatrice Mendez Egle 1979 "New borrowings in English", American Speech 54: 23 — 37.

Borrowing

413

Davis, Lawrence M. 1979 "The perils of purism: The anti-English purists in Israel", American Speech 54: 175-184. Finkenstaedt, Thomas — Ernst Leisi — Dieter Wolff 1970 A chronological English dictionary. Heidelberg: Winter. Fries, Charles C. — Kenneth L. Pike 1949 "Coexistent phonemic systems", Language 25: 29 — 50. Galinsky, Hans 1980 "American neologisms in German", American Speech 55: 243 — 263. Geipel, John 1971 The Viking legacy: The Scandinavian influence on the English and Gaelic languages. Newton Abbot, Devon: David and Charles. Hatcher, Anna Granville 1951 Modern English word-formation and neo-Latin. Baltimore: Johns Hopkins Press. Haugen, Einar 1950 "The analysis of linguistic borrowing", Language 26: 210 — 231. Hawkins, R. E. 1984 Common Indian words in English. Delhi: Oxford University Press. Hellstrom, Robert W. 1976 "Finglish", American Speech 51: 8 5 - 9 3 . Hockett, Charles F. 1958 A course in modern linguistics (esp. 402—424: "The conditions for borrowing", "Kinds of loans", "Adaptation and impact"). New York: Macmillan. Lehmann, Winfred P. 1973 Historical linguistics: An introduction (esp. 205 — 235: "Semantic change and changes in the lexicon") (2nd edition). New York: Holt, Rinehart and Winston. Pyles, Thomas — John Algeo 1982 The origins and development of the English language (esp. 292 — 317: "Foreign elements in the English word stock") (3rd edition). New York: Harcourt Brace Jovanovich. Rao, G. Subba 1954 Indian words in English: A study in Indo-British cultural and linguistic relations. Oxford: Clarendon Press. Serjeantson, Mary S. 1935 A history of foreign words in English. London: Routledge and Kegan Paul. Smith, Donald L. 1974 "Ribbing Ingrish — Innovative borrowing in Japanese", American Speech 49: 185-196. Steinmetz, Sol 1981 "Jewish English in the United States", American Speech 56: 3 — 16. Turner, Lorenzo Dow 1974 Africanisms in the Gullah dialect. Ann Arbor: University of Michigan Press. (First published Chicago: University of Chicago Press, 1949.) Weinreich, Uriel 1963 Languages in contact: Findings and problems. The Hague: Mouton. (First published New York: Linguistic Circle of New York, 1953.)

Etymology Edgar C. ΡοΙοπιέ

The purpose of etymology is to trace back the lexical component of a language to its origins and to describe the basic meaning, history, and development of the terms involved. It is therefore closely associated with historico-comparative linguistics, as it has to apply the rules of diachronic phonological and morphological change to determine the original form or reconstruct the prototype of each lemma in the language from which it originated — whether it be derived from a prehistoric protolanguage or some dialect of an earlier language, or borrowed from an outside source. The term "etymology" — borrowed from the Old French into Middle English under the form ethimologie — derives ultimately from the Greek ετυμολογία which designated the attempts of the Greek philosophers to discover the "true sense" of the words "according to their origin" (Grk. το ετνμον), an endeavor which led to the two opposed schools of thought illustrated by Plato's Cratylus. While Cratylus defends the claim advanced by Heraclitus that there is a natural link (Grk. φύσει) between name and object, Plato sides with Democritus who considers the attribution of names as a matter of "arbitrary determination" (Grk. θέσει), but he fails to draw the proper consequences from this view on account of his own concept of "truth" (Ducrot - Todorov 1983: 131). In the Middle Ages, purposeful etymological investigation was thwarted by theological concerns which fostered the belief in the monogenesis of language and the derivation of all tongues from Biblical Hebrew. Apart from some enlightened forerunners in the seventeenth and eighteenth centuries, etymological research only started on a genuinely scientific basis with the development of linguistics in the

416

Edgar C. Polome

nineteenth century. Thus, in spite of many shortcomings in his methodology and the organization of his materials, August Friedrich Pott (in his Etymologische Forschungen auf dem Gebiete der indogermanischen Sprachen [1833 — 1836; 2nd edition in ten vols., 1859 — 1876]) showed a reasonable understanding of sound change. However, as the early nineteenth-century comparativists were strongly interested in reconstructing the original culture of the people into whose linguistic past they probed, their etymological work was closely linked with linguistic paleontology, and they already based their search for the Proto-Indo-European homeland and their assumptions about the dispersal of the Indo-European dialects on the study of the vocabulary, e. g., Adolphe Pictet, in Les origines indo-europeennes ou les Aryas primitifs, essai de paleontologie linguistique (1859 — 1863). The publication of the Grundzüge der griechischen Etymologie of Georg Curtius in 1858 marks a turning point in etymological research: in an effort to correlate comparative linguistics with Greek philology, he differentiated between "regular" and "sporadic" sound change, also eliminating worthless etymologies through "discriminating criticism" (Pedersen 1962: 90). Equipped with August Schleicher's confident reconstruction of Indo-European, August Fick then published his Wörterbuch der indogermanischen Grundsprache in ihrem Bestände vor der Völkertrennung, which became a Vergleichendes Wörterbuch der indogermanischen Sprachen in the following editions (2nd edition 1870-1871; 3rd edition 1874-1876). With the thorough transformation of the picture of the Indo-European protolanguage after 1876 with the work of Brugmann and the other "Neogrammarians", Fick's dictionary had to be completely recast in its fourth edition, in which the second volume — Urkeltischer Sprachschatz (1894) — was written by Whitley Stokes and edited by Adalbert Bezzenberger, whereas the third — Wortschatz der Germanischen Spracheinheit (1909) — was compiled by the Norwegian linguist Alf Torp, assisted by Hjalmar Falk. In the meantime, the foundations for constructive work in the Romance languages had also been laid by Friedrich Diez with his Etymologisches Wörterbuch der romanischen Sprachen (1853). The time was now ripe for the efflorescence of an impressive array of etymological dictionaries of various Indo-European languages and for the growing development of etymological research in other language families. To mention but a few in the six decades between 1880 and 1940 in the Indo-European field, let us list the following by language group:

Etymology

417

a) Indo-Aryan: C. C. Uhlenbeck, Kurzgefaßtes etymologisches Wörterbuch der altindischen Sprache (Amsterdam 1898 — 1899). Only the introduction and the first part (a-jü) have appeared of Ernst and Julius Leumann's Etymologisches Wörterbuch der Sanskrit-Sprache (Leipzig 1907), and similarly, little has appeared of Walter Wiist's Vergleichendes und etymologisches Wörterbuch des Alt-Indoarischen {Altindischen) (Heidelberg 1935 — ). Ralph L. Turner, A comparative and etymological dictionary of the Nepali language (London 1931). Paul Horn, Grundriß der neupersischen Etymologie (Straßburg 1893), to which Heinrich Hübschmanns Persische Studien (Straßburg 1895) serves as a supplement. Heinrich Hübschmann, Etymologie und Lautlehre der ossetischen Sprache (Straßburg 1887). Georg Morgenstierne, An etymological vocabulary of Pashto (Oslo 1927). b) Greek: Prellwitz, Etymologisches Wörterbuch der griechischen Sprache (Göttingen 1892; 2nd edition 1905). Emile Boisacq, Dictionnaire etymologique de la langue grecque (Paris/ Heidelberg 1909-1916). c) Italic: Alois Walde, Lateinisches etymologisches Wörterbuch (Heidelberg 1906; 2nd edition 1910; a third edition, thoroughly revised by J. B. Hofmann, appeared in two volumes in 1938 — 1954). Brief etymological data were given by Rudolf Thurneysen in the monumental Thesaurus linguae latinae from 1901 on (its publication was momentarily interrupted by World War I in 1915). Fr. Muller, Altitalisches Wörterbuch (Göttingen 1926). Alfred Ernout and Antoine Meillet, Dictionnaire etymologique de la langue latine. Histoire des mots. (Paris 1932; 2nd edition 1939; 3rd edition 1951; 4th edition by J. Andre 1979). d) Romance: Wilhelm Meyer-Lübke, Romanisches etymologisches Wörterbuch (Heidelberg 1911-1920; 3rd edition [thoroughly revised] 1930-1935). For the individual Romance languages, let us briefly refer, among others, to the dictionaries of Roque Barcia (1880 — 1883) and E. de

418

Edgar C. Polome

Echegaray (1887-1889) for Spanish; of F. A. Coelho (1890) and A. Nascentes (1932) for Portuguese; of F. Zambaldi and O. Pianigiani (1907) for Italian; of S. Puscariu (1905) for Rumanian; of W. von Wartburg (1928-), Ε. Gamillscheg (1929), Ο. Bloch and W. von Wartburg (1932), A. Dauzat (1938) for French; etc. e) Celtic: Alexander Macbain, An etymological dictionary of the Gaelic language (Inverness 1896; 7th edition, revised by C. MacPharlain, 1911). Victor Henry, Lexique etymologique des termes les plus usuels du breton moderne (Rennes, 1900). e) Germanic: C. C. Uhlenbeck, Kurzgefaßtes etymologisches Wörterbuch der gotischen Sprache (Amsterdam 1896; 2nd edition 1900). Sigmund Feist, Vergleichendes [etymologisches] Wörterbuch der gotischen Sprache (Halle 1909; 2nd edition Halle 1923; 3rd edition Leiden 1939). Hjalmar Falk and Alf Torp, Norwegisch-dänisches etymologisches Wörterbuch (Heidelberg 1910 — 1911) — originally published as Etymologisk Ordbog over det norske og det danske sprog (Kristiania 1903-1906). Alf Torp, Nynorsk etymologisk ordbok (Oslo 1919). Elof Hellquist, Svensk etymologisk ordbok (Lund 1922; 2nd edition 1939; 3rd edition 1948). Jakob Jakobsen, An etymological dictionary of the Norn language in Shetland (edited by A. Horsböl, London/Copenhagen 1928 — 1932; an early version was edited by M. Mikkelsen as Etymologisk ordbog over det norrone sprog pä Shetland, Copenhagen, 1921). Walter W. Skeat, An etymological dictionary of the English language (Oxford 1879-1882; revised edition 1909). Ernest Weekley, An etymological dictionary of modern English (London 1912). Ferdinand Holthausen, Etymologisches Wörterbuch der englischen Sprache (Leipzig 1917; 2nd edition 1927; 3rd edition Göttingen 1949). Ferdinand Holthausen, Altenglisches etymologisches Wörterbuch (Heidelberg 1934; 2nd edition 1963). Friedrich Kluge, Etymologisches Wörterbuch der deutschen Sprache (Strasbourg 1883 — with 7 additional editions between 1884 and 1915; 9th edition Berlin — Leipzig 1921; following editions [from the 11th, 1934, on] revised by A. Goetze, later [17th edition 1951] by W. Mitzka

Etymology

419

and A. Schirmer, and [from the 18th, 1960, till the 21st, 1975] by W. Mitzka alone). Johannes Franck, Etymologisch woordenboek der Nederlandsche taal (The Hague 1892; 2nd edition, revised by N. van Wijk, 1912-1929; Supplement by C. B. van Haeringen, 1936). Jozef Vercoullie, Beknopt etymologisch woordenboek der Nederlandsche taal (Ghent 1890; 2nd edition 1898; 3rd edition 1925). f) Slavic: Franz Miklosich, Etymologisches Wörterbuch der slawischen Sprachen (Vienna 1886). Erich Bernerker, Slawisches etymologisches Wörterbuch (incomplete: A - M o r - ; Heidelberg 1908-1913). Antonin Preobrazenskij, Etimologiceskij slovar' russkogo jazyka (incomplete: A - Su-; Moskva 1910-1918). Aleksander Brückner, Slownik etymologiczny jezyka polskiego (Krakow 1926-1927; 2nd edition 1957). Josef Holub and Frantisek Kopecny, Strucny slovnik etymologicky jazyka cesko-slovenskiho (2nd edition Praha 1937; 3rd edition published as Etymologicky slovnik jazyka ceskeho, Praha 1952). g) Baltic: Reinhold Trautmann, Baltisch-slawisches Wörterbuch (Göttingen 1923). Erich Berneker, Die preußische Sprache. Texte, Grammatik, etymologisches Wörterbuch (Straßburg 1896). Reinhold Trautmann, Die altpreußischen Sprachdenkmäler. Einleitung, Texte, Grammatik, Wörterbuch (Göttingen 1910). There are also concise etymological notes in the four-volume Lettischdeutsches Wörterbuch (Latviesu valodas värdmca) of J. Mühlenbach and J. Endzelin (Riga 1923-1932). h) Albanian: Gustav Meyer, Etymologisches Wörterbuch der albanischen Sprache (Straßburg 1891). i) Armenian: Heinrich Hübschmann, Armenische Studien. 1.1: Grundzüge der armenischen Etymologie (Leipzig 1883). Although not to the same extent, similar work was being done in such families as Finno-Ugric, as the well-known Magyar etymologiai

420

Edgar C. Polome

szotär {Lexicon critico-etymologicum linguae Hungaricae) of Zoltan Gombocz and Jänos Melich (Budapest 1914—) may show. Reviewing sample entries of a number of these dictionaries, including also the extensive historical dictionaries like the Oxford New English dictionary on historical principles, M. de Vries and L. A. te Winkel's Woordenboek der Nederlandsche Taal, and the Ordbok over Svenska Spräket published by the Swedish Academy, A.S.C. Ross (1958) tentatively applies the coding system of symbolic logic to define the relation between cognate linguistic items, but soon decides that the most efficient way to approach the study of etymology is to provide the student with a detailed introduction into the diachronic development of English phonology within the dialect of Germanic from which it derives, and to examine further how Germanic correlates within Indo-European with the other branches of that family both in the phonology and, to a certain extent, in the morphology. After surveying the way entries have been structured in a set of dictionaries, Yakov Malkiel (1976: 54—65) establishes a kind of typology of the presentation of the evidence: a) listing the cognates without explanation, as F. Holthausen does in his dictionaries (his method has hardly changed from his 1934 Old English etymological dictionary to his 1948 Vergleichendes und etymologisches Wörterbuch des Altwestnordischen —Altnorwegisch-isländischen, einschließlich der Lehn- und Fremdwörter sowie der Eigennamen [Göttingen: Vandenhoeck & Ruprecht]); b) supplying the etymon with a minimal amount of cultural information and/or word history, like the French etymological dictionary of O. Bloch and W. von Wartburg (Paris: Presses Universitaires de France, 1950; 2nd edition); c) choosing a definite etymology, but mentioning divergent views in a decreasing order of acceptability, as A. Walde and J. B. Hofmann do in the third edition of the Latin etymological dictionary (Heidelberg: Carl Winter, 1938 — 1954), and in most cases, Sigmund Feist in the third edition of his Gothic etymological dictionary (Leiden: E. J. Brill, 1939) does as well; d) advocating one specific solution, while listing, discussing, and also refuting explanations which the author deems less plausible — an approach typically used by Jan de Vries in both his Altnordisches etymologisches Wörterbuch (Leiden: Ε. J. Brill, 1961) and his Nederlands etymologisch woordenboek (Leiden: E. J. Brill, 1971);

Etymology

421

e) picking out a preferred explanation while simply providing bibliography on divergent opinions; f) definitely stressing one's favorite explanation, but giving a fairly detailed presentation of the etymological problem with a more or less exhaustive bibliographical review of the proposed hypotheses — a method used by the new Etymologisches Wörterbuch des Althochdeutschen of A. Lloyd and Ο. Springer (vol. 1: A — bezziro; Göttingen: Vandenhoeck & Ruprecht, 1988). Obviously, these approaches can overlap, and they can occur in various combinations within one dictionary. The delicate part is the choice of the etymology that is presented as the most plausible, since the reasons why others are rejected or doubted can always appear controversial and may often depend on the position of the author versus a definite theoretical issue in the diachronic development of the language or the reconstruction of the protoform. Thus, if one accepts, with Elmar Seebold (1980), that PIE *ghw- becomes *b- initially in Germanic, ON bani 'death, murderer' can be derived from the IE root *ghwen- 'hit, kill', and the phrase orms bani 'serpent's killer', designating E>6rr in the 22nd stanza of Hymiskvida, becomes a neat parallel of the Vedic epithet of Indra, Vrtrahän- 'killer of the ophidian Vrtra- (cf. Watkins 1987: 288 — 293). If, on the contrary, one sticks with the more generally accepted view that initial PIE *ghw- yields either *g- or *vv(cf. Polome 1987), ON bani, OE, OHG bana 'killer' are to be derived from a root *bhen- 'hit, injure', which also appears in Avestan banta'ill' — illness being ascribed to the action of evil spirits striking the victim in Zoroastrian belief (cf. for further discussion, Lloyd/Springer 1988: [cols.] 460 — 462). As the Iranian term stands isolated (with a verbal form banaysn 'they sicken'), its relationship with the Germanic terms is, however, often considered dubious (cf. Lehmann 1986: 61). The above example illustrates the plight of the etymologist who has to coordinate the phonological and semantic changes conditioning the reconstruction of the "etymon" with the cultural data and the linguistic context. Since the assumed root *bhen- does not meet Meillet's rule of thumb that one needs cognates in at least three non-contiguous IndoEuropean dialects in order to postulate the existence of the relevant Indo-European protoform, it rests exclusively on a Germanic-Iranian correspondence (a postulated relation with Welsh bon-clust 'box on the ears' appears to be improbable; cf. Pokorny 1959: 126). Isoglosses of this type between distant Indo-European dialects can, however, be

422

Edgar C. Polome

the reflexes of characteristic archaisms, as are the religious terms found in Indo-Iranian and in Italic and Celtic, e.g., OInd. sraddhä-, Αν. zrazdä-, Lat. credere, Olr. cretim 'believe, have faith in' < IE *kreddhe- 'put trust in', or in Anatolian and Italic, or Germanic, e. g., Lat. sacer 'sacred, hallowed': Hitt. saklai- 'custom, rite', or ON pulr 'cultic speaker' (G Kultredner): Hitt. talliia- 'call solemnly upon the deity' (Polome 1975). However, in such cases, the relevant correspondences belong to a set within the archaic technical terminology of IndoEuropean — a type of vocabulary that has been carefully collected and studied in detail by T. Gamkrelidze and V. Ivanov (1984). But to assess the validity of the comparison of ON bani with Avest. banta-, it is important to examine the other cases of co-occurrence of lexical items in Germanic and Iranian: the Indo-European etymological dictionary of J. Pokorny (1959) does not list any entry where the correspondence is limited to those two languages, but has seven entries (including the root *bhen-) in which a third language is involved. They do not at all constitute a semantically coherent set, but their further analysis may throw some additional light on the problems the etymologist has to cope with: a) under IE *rebh~, Pokorny (1959: 853) lumps together modern Persian raftan 'go', Ossetic räväg 'quick', Middle Irish reb 'sleight of hand, play' ( < Celtic *reba), and Germanic *reb- 'move turbulently', the only semantic link between these terms being apparently a "rapid movement" — but, as J. Vendryes (1974: R - l l ) points out, the meaning of Pers. raftan is too divergent, and the only valid connection is between Celt. *rebä and the German dialectal verb reben 'be fidgety' (MHG reben 'dream, be perturbed'). b) similarly, Avestan kasu- 'small', German hager 'lean', and Lithuanian nukaseti 'be completely weakened' serve as a base to reconstruct an IE *kak- 'grow thin' (Pokorny 1959: 521-522), but again the Iranian term has the more general sense of 'small, limited, inferior' (Bartholomae 1961: cols. 460 — 461), whereas the Lithuanian verb apparently belongs to a set of "popular" words such as the Lithuanian simplex kaseti 'get thin, wane, wither' and Russian cäxnut' 'waste away' (Fraenkel 1962: 227 — 228). This however, implies that the -xin the Slavic terms would either reflect *-ks- or be "derogatory" like the palatalization in Lithuanian kioka 'weakly person, dragging himself around' (Fraenkel 1962: 255) — a view which Slavicists like Vasmer (1958: 306) and Trubacev (1987: 338) only share insofar as German hager is left out of the comparison.

Etymology

423

These two cases illustrate the necessity of closely examining the position of the terms compared within their narrower semantic field in closely cognate languages before going out on a limb to connect them with words whose phonological shape would allow a direct link, but whose meaning is much too loosely related. The careful analysis of the semantic component of the etymology is a basic requirement: one of the major objections made against compendia like Pokorny's Indo-European etymological dictionary, and its predecessor, the twovolume Vergleichendes Wörterbuch der indogermanischen Sprachen of Alois Walde, edited by Julius Pokorny (with a 3rd volume of indices compiled by Konstantin Reichhardt [Berlin: Walter de Gruyter, 1927 — 1932]), is their reconstruction of numerous homonymous roots with rather general meanings, e.g., IE *ger-x 'pull together, gather', *ger-2 'call out hoarsely' (onomatopoeic), *ger-3 'turn, wind', *ger-4 'be awake, awaken' (Pokorny 1959: 382 — 390). This, in turn, accounts for the abundance of allegedly synonymous roots: at least nine are listed as meaning "swell", and the same number appears with the meaning "cut", while no less than fifteen are supposed to mean "turn", and there are even more examples of "hit", or "bend", or the like. Admitted that reconstructed roots reflect a common denominator of the forms attested in the various dialects, making them semantically so vague is quite implausible. What is required is an investigation of the various fields in which the basic meaning branches off, with specific data on the socio-cultural or material motivation for the substantial changes of meaning involved. A neat example of such an inquiry is provided by Jacqueline Manessy-Guitton (1988) when she describes the interrelation and the progressive development of the very divergent meanings of the Greek terms derived from IE root *H2eig- designating a forward movement. A wild goat is a rapidly moving, jumping animal — hence, its Greek name: αιξ\ similarly, the titmouse is a perpetually active bird, so forward that the Greek would use the phrase aiyi&ov τοληρότερος 'more cheeky than a titmouse' — hence, again, the derivation of its name from *aiy-. From the concept of "moving forward", one arrives easily at the notion of "attacking" or "projecting" to which words like αίγίζω 'attack violently, rend asunder', αιγανέη 'hunting spear, javelin', or αίγλη 'light' (as a projection of luminous particles). A secondary development from "darting forward" is "tall and slim" (cf. French elance), or "steep" — hence, the name of the black poplar, αίγειρος, and the Homeric adjective αΐγίλιψ 'steep', which could be based on a neuter *αιγιλ- (but the derivation remains obscure).

424

Edgar C. Polome

Pursuing the examination of the lexical correspondences between Germanic, Iranian, and a third Indo-European language, we can point out two that appear to simply reflect old terms for specific feelings or actions resulting from an emotional drive or impulse. a) IE *kormo- 'pain, agony, disgrace' (Pokorny 1959: 615), in Avest. fsaroma- '(sense of) shame', Middle Pers. sarm 'shame', OHG har(a)m 'pain, grief, insult, misfortune', ON harmr 'grief, sorrow', OE hearm 'harm, grief, affliction, calumny, insult', OCS sramj> (< *sormu) 'shame'. b) IE *ieug- 'arouse, be stirred up" (Pokorny 1959: 512), in OPers. yaud- 'move restlessly, get excited, become restless', Goth, jiuka (plur.) 'quarrels', OE geocor 'full of hardship', M H G jouchen 'chase, drive', Tocharian AB yuk- 'vanquish', yutk- (< *yuktk-) 'be concerned about'. Sometimes closer study shows that the root is also attested in additional Indo-European languages, for example in the case of the old heteroclitic neuter *aier- ~ aien- (Pokorny 1959: 12), where Armenian ayg 'daybreak' can be added to Avestan ayard, gen. ayan 'day', Greek ήρι < *αιερ-ι (loc.) 'early', Gothic air, ON är 'early', OHG er 'previously, before', etc. (cf. Lehmann 1986: 18). Similarly, Latin texere 'weave' could be semantically better connected with Ossetic taxun 'weave', Armenian t'ek'em 'turn, plait, wind', OHG täht 'wick', ON jyättr 'thread, wire, wick' than with the terms referring to carpentry and working with an axe: Vedic tasti (3rd plur. täksati] 'hews, shapes, does carpentry work', Avest. tasaiti 'does carpentry work, cuts', Greek τέκτων 'carpenter', Olr. tal(< *tokslo-) 'axe', OHG dehsa(la) 'doubleaxe', ON pexla 'axe', Lith. tasyti 'hew', OCS tesati 'hew', which show a palatal -k-. Ultimately, if the "satem"-palatalization is a dialectal development in Indo-European, both groups could be related to the same root *tek~, enlarged or not by -s-, whose original meaning might be preserved in Anatolian by Hittite taks- 'join together', as this verb could apply either to the threads in weaving or the woodwork in carpentry. While the rules of phonological development and morphological derivation have to be most strictly applied, one must always keep in mind that, in etymology, each word constitutes a specific case, and that its history has to be carefully retraced. If one looks up the etymology of the term for "bed" in one of the standard dictionaries, e.g., Kluge (1975: 71), Germanic *badja-, which yielded Gothic badi, Old English bedd, Old Frisian, Old Saxon bed, Old High German betti

Etymology

425

'bed', one is usually referred to an Indo-European root *bhect- 'dig, pierce' which is abundantly documented by Hittite beda- 'pierce, prick', Latin fodere 'dig', fossa 'ditch', Welsh bedd, Cornish bedh, Breton bez 'grave', Lithuanian bedü, besti 'stick, bore, dig', Latvian b§du, best 'dig, bury', bedre 'pit', Old Prussian boadis 'prick', Old Slavic bodd, bosti 'stick' (cf. Lehmann 1986: 55). It would properly be, as Klein (1966: 158) translates it: "a couch dug into the ground", the key to the semantic development being supplied by meanings like 'lair of animals, nest' in Old Swedish basdhil and 'small plot in a garden' in the Old High German gloss bedde. agellus in horto, 'flowerbed' in Old English bedd (wyrtbedd 'garden-bed'). However, Jan de Vries (1961: 29; 1971: 34) considers this interpretation highly improbable, pointing out that "the sleeping place of man would definitely never be called after the lair of an animal; such a primitive stage at which people might sleep in holes dug into the soil already belonged to the past in Indo-European times ... The co-occurrence of the meanings "bed" and "grave" is ascribable to the religious views that consider sleep and death as closely interrelated." Therefore, de Vries proposes another etymology, linking Germanic *badja- with *bada- 'bath': as the latter derives from IE *bhe~, found, e.g., in O H G bäen 'bake bread' (cf., with -g-enlargement: Greek φώγω 'roast', OE bacan 'bake', etc.), its original meaning would be "warm bath", and the bed would be "the place where one keeps warm". Although Vittore Pisani apparently shares such a view (in Paideia, vol. 13, 1958: 193), one cannot help finding this hypothesis rather ad hoc. What do we know about the sleeping habits of the ancient Germanic people? As H. Hinz (1976: 316) points out, the oldest wooden bed extant in the Germanic world dates back to the sixth century A. D.; in early times, they may originally have slept on the floor, but archeological finds indicate that earthen benches along the walls of prehistoric dwellings already served to sit as well as to sleep. There is no mention anywhere of "dug out couches", and it might be remembered that Tacitus, when he wants to stress the mir a feritas ('astonishing wildness') of the Finns and contrasts them with the Germanic people, explicitly refers to the Finns' sleeping on the ground, with their lack of a permanent home, as traits of their barbaric savage state (Germania, chapter 46). What the early Germanic *badja- characterizes is accordingly, as P. Buchholz (1976: 319) rightly specified, "whatever softer bedding one sleeps on", in particular the bolster that supports the head: Old Norse bedr 'bolster, feather-bed', as well as the early Finnish

426

Edgar C. Polome

loan from Germanic patja 'bolster', point in that direction, and etymologically, they can be derived without further ado from IE *bhedh'bend, press'. When stable wooden structures were introduced, they were simply named metonymically after the soft support people were wont to lie on. Several scholars have presented their views on etymology in recent years, and much of what they have said pertains to the methodology and history of lexicography and semantic analysis, but their presentation has usually been illustrated by excellent examples clarifying their views. Besides the already mentioned works of A. S. C. Ross and Yakov Malkiel, the following definitely deserve particular attention: Vittore Pisani, L'etimologia. Storia — Questioni — Metodo (Torino: Rosenberg & Sellier, 1947; 2nd edition Brescia: Paideia, 1967) — translated into German: Die Etymologie. Geschichte — Fragen — Methoden (Internationale Bibliothek für allgemeine Linguistik; München 1975). Pierre Guiraud, L'etymologie ("Que Sais-je?"; Paris: Presses Universitaires de France, 1964). Elmar Seebold, Etymologie. Eine Einführung am Beispiel der deutschen Sprache (München: C. Beck, 1981). Helmut Birkhan, Etymologie des Deutschen (Langs Germanistische Lehrbuch Sammlung, Vol. 15; Bern/Frankfurt/New York: Peter Lang, 1985). There are also two important collections of articles on the problem of etymology and etymological dictionaries: Etymologie, edited by Rüdiger Schmitt (volume 373 in the series "Wege der Forschung"; Darmstadt: Wissenschaftliche Buchgesellschaft, 1977). Das etymologische Wörterbuch. Fragen der Konzeption und Gestaltung, edited by Alfred Bammesberger (Regensburg: Friedrich Pustet, 1983). Moreover, since 1963, the Institute for the Russian Language of the Academy of Sciences of the U.S.S.R. publishes annually a volume of articles entitled Etimologija. Although the recent revival of lexicography has been concentrated more on descriptive than on historical dictionaries, etymology has also enjoyed a renewal of interest since World War II. The requirements for a methodologically sound approach to etymological lemmata have probably been spelled out in the clearest way by Jürgen Untermann in the description of his projected Osco-Umbrian dictionary in the just quoted volume edited by A. Bammesberger (1938: 295 — 312, esp. 301):

Etymology

427

1) Documentary evidence: For languages with limited documentation, all the attestations of the term have to be given in the proper context, with the necessary discussion of the inflectional forms and the critical apparatus relevant to readings, variants, emendations or any other alterations of the preserved text; in languages with abundant material, the coverage must be such that it provides a complete documentation on all the possible occurrences of the lemma in its various forms or indicates the specific inflectional pattern to which it belongs, i.e., if a verb is a regular member of a definite class so that all its forms are readily predictable, the latter need not be listed, but all irregularities have to be mentioned in context. 2) Meaning: Preference should always be given to "contextual meaning", but other sources may be used, e.g., for older languages, the more easily understandable text in the original language from which a document was apparently translated; bilingual inscriptions; old glossaries and other lexicographical compilations; quotes in ancient authors; etc. Etymological interpretation should only be used as a last resource. 3) Semantic field: It may prove useful to properly situate the term within its semantic field, to distinguish it from its alleged synonyms and to pinpoint its particular usage in contrast or alternation with other terms — thus, for example, the colloquial terms for "dog" mutt, pooch, and cur, will be distinguished by the features: [crossbred] and [friendly], according to the matrix: pooch mutt cur

[ +friendly] [ +crossbred] [ +crossbred] [ — friendly]

4) Derivations and compounds: The whole word family of the base term is examined and included — this is particularly illustrated by the French etymological dictionaries of Latin and Greek, namely A. Ernout and A. Meillet's already quoted Dictionnaire etymologique de la langue latine (Paris: Klincksieck, 4th edition by J. Andre, 1979) and Pierre Chantraine's Dictionnaire itymologique de la langue grecque. Histoire des mots (Paris: Klincksieck, 1968 — 1980; 4 volumes in 5 parts, completed by J. Taillardat, O. Masson and J.-L. Perpillou, and edited by M. Lejeune). 5) Etymology (proper): Cognates of the term within the language family to which it belongs if it derives from the "protolanguage". Reconstruction of the protoform and possibly of intermediate stages

428

Edgar C.

Polome

of development. Full examination of the semantic development from the original meaning to the attested sense(s). If there are conflicting etymologies, it is necessary to clearly justify the choice of a specific solution on the basis of the evidence — linguistic and extra-linguistic — and to thoroughly discuss the rejected etymologies critically. It is not sufficient to simply mention "unacceptable" or "incorrect", as is often the usage in some German etymological dictionaries: explanations like those supplied by Lehmann (1986) in his Gothic etymological dictionary or Lloyd and Springer (1988) in their Old High German etymological dictionary are required. 6) Word history: This does not only cover the diachronic development of the term, but also the motivation for calling the item by the name given to it (what the Germans call Benennungsmotiv) — a process that E. Seebold has illustrated in his book (quoted above) on etymology (1970: 230 — 238) with regard to the "earwig", and in his contribution to Bammesberger's volume on the etymological dictionary (1983: 261—276) by discussing a set of terms like German Hahn, Lunge, and Ähre. Word history is especially crucial when dealing with non-inherited words: since the earliest times cultural items have been designated by borrowed terms which have often become so integrated into the language that they are hardly identifiable. A classic example is the case of the words for "hemp". It seems now fairly well established that the plant (Cannabis sativa) originated in central Asia, but spread wide and far at a very early date. Its wild forms were already valuable for economic purposes, as the fibers obtained from the bast of the stems could easily be used to produce coarse textiles, while the seeds provided animal feed. It was cultivated in China as early as the middle of the third millennium B. C., and hemp fabrics appear in Anatolia about the eighth century B. C. Between 700 and 300 B.C., the Sarmatians and Scythians became familar with it in the south-western part of Russian Asia. Its use as a drug, known in India since 1000 B.C., did not reach the Mediterranean before post-classical times (Zohary — Hopf 1988: 119 — 121). As hemp spread as a fiber-crop from Greece to Italy and Sicily about 100 B.C., it is not astonishing that Latin borrowed its name — cannabis — from the Greek κάνναβις, whose origin remains disputed. Herodotus (IV. 74 — 75) describes the plant as alien (the plant only became familiar in Greece in Hellenistic times), and he specifies that the Scythians used its seeds to purify themselves

Etymology

429

in ritual baths (hemp oil served to make soap), and that they became intoxicated with its narcotic drug, whereas the Thracians wove fabrics resembling linen with its fibers. While Low German, Middle Dutch, and dialectal modern Dutch kennep reflect a direct borrowing from Latin cannabis (cf. Old WestFlemish canaps > W. Flem. kampst, kemp(st); de Vries 1971: 251), just as in the Celtic words for "hemp" (cf. Ir. cnäib, etc.), the Germanic terms mostly show an initial h- (cf. ON hampr, OE hcenep, OHG hanaf, etc.) which points to borrowing from a different source before the consonant shift. This source could be either Scythian, as for Russian Church Slavic konoplja (from which, in turn, Lith. kanäpes is presumably derived), or Thracian, as for Albanian kanep and perhaps also Old Prussian knapios (cf. Toporov 1984: 93 — 98). The ultimate origin of the name of the plant, which contains the sequence *kVnVb/p-, remains obscure (there is already a Sumerian kunibu 'hemp'). Civilization introduces new products in definite linguistic areas, and with each of them comes the term designating it, but it is sometimes difficult to retrace the road that it has taken to get there: Thus, Gothic calls 'oil' alew, a neuter -α-stem which implies an underlying form *oleuom in the language from which it was borrowed; such a transitional form has been assumed for the Latin loan *oleiuom from Greek έλai(f)ov 'oil' around the middle of the second century B.C. The problem is to determine how this form got into Germanic. The idea of a Celtic intermediary has been suggested, but that implies that the borrowing must predate the Germanic lowering of *o to a, and that the Goths would have gotten the term from an eastern Celtic tribe with which they traded in Bohemia, Moravia or Silesia, and which would have Celtic *olevom < OLat. *oleivom. But there is no trace of early oil trade in the Germanic territory, and no evidence of a continental Celtic *olevom (the insular terms are late borrowings from British Latin). It could, however, be assumed that the term might have entered Germanic, together with the word for "lamp" — Gothic lukarn, from provincial Latin *lücerna 'oil-lamp' — via the Raetic population of the southern Alps, with which the Cimbri came into close contact at the end of the second century B.C. In Raetic, Latin *oleuom would have appeared as alewa- (Polome 1985: 310); unfortunately, in view of the scarcity of the documentation on the Raetic language, there is no actual attestation of the term (cf. Lehmann 1986: 27). As there is no archeological evidence to corroborate any assumption made about the transmission of such an important Mediterranean product as "oil"

430

Edgar C. Polome

to the early Germanic world, all our hypotheses about the background of Gothic alew must of necessity remain somewhat speculative. Although many problems remain unsolved — and some are likely not to be resolved for lack of adequate documentary evidence — the considerable progress made in the field of etymological research can easily be measured by comparing the recent etymological dictionaries with their nineteenth or early twentieth-century predecessors. Limiting ourselves again to the field of Indo-European and to the major works, some of which have already been quoted, we might make the following comments: a) Proto-Indo-European: There is no updated comparative etymological dictionary of Indo-European. Pokorny's 1959 compendium is essentially a shortened, revised version of Walde's Vergleichendes Wörterbuch der indogermanischen Sprachen, which he edited in the early 1930s, though some material has been added, especially in Celtic and the "recently discovered" languages (Hittite and Tocharian; Mycenian Greek is not taken into consideration). A semantically organized dictionary covering the essentials of the Indo-European vocabulary is provided by Carl Darling Buck's Dictionary of selected synonyms in the principal Indo-European languages (Chicago: Chicago University Press, 1949). Moreover, Gamkrelidze and Ivanov's magnum opus (1984) now offers a temporary substitute for an updated semasiological lexicon of the protolanguage, in which the cultural background is discussed in considerable detail in correlation with the views of the authors on the level of civilization and original homeland of the Indo-Europeans. Recently completed was also the Indo-European comparative dictionary compiled by Stuart A. Mann (Hamburg: H. Buske, 1984 — 1987) — a rich, but fairly uncritical collection of Indo-European roots, presented without further discussion or relevant bibliography. b) Indo-Iranian: There is a complete Concise etymological Sanskrit dictionary in four volumes by Manfred Mayrhofer (Heidelberg: Carl Winter, 1956 — 1980), which is now being replaced by a completely new Etymologisches Wörterbuch des Altindoarischen (Heidelberg: Carl Winter, 1986 — ), whose first part concentrates on the older language. For later Indo-Aryan, the Comparative dictionary of Indo-Aryan languages (London: Oxford University Press, 1966) by R. L. Turner, with its Indexes (1969) and Phonetic analyses (1971) is available. A new Avestan dictionary has been in preparation for a long time (cf. Bernfried Schlerath, Awesta-Wörterbuch. Vorarbeiten: 1. Index

Etymology

431

locorum zur Sekundärliteratur des Awesta. 2. Konkordanz [Wiesbaden: Ludwig Reichert, 1968]). About the requirements for such a work and about the more comprehensive Iranian dictionary planned by W. P. Schmidt after the pattern of Turner's work on Indo-Aryan, see B. Schlerath's article on "Stand und Aufgaben der iranischen Lexikographie" (in Bammesberger 1983: 229-242). c) Anatolian: While the Hethitisches Wörterbuch of Johannes Friedrich (Heidelberg: Winter, 1952-1954; with three supplements - 1: 1957, 2: 1961, 3: 1966) gave only sparse etymological comments, and Heinz Kronasser's Etymologie der hethitischen Sprache (Wiesbaden: Otto Harrassowitz, 1962 — 1968; indexes by Erich Neu, 1988) focuses on a) graphemics and phonology, and b) word-formation, two etymological dictionaries of Hittite are presently under publication: the first volume of Johann Tischler's Hethitisches etymologisches Glossar (Innsbrucker Beiträge zur Sprachwissenschaft, vol. XX: 1; Innsbruck: Sprachwissenschaftliches Institut der Universität), providing a well-documented critical survey of the etymological literature on the Hittite terms beginning with A to K, was completed in 1983, whereas one volume containing parts I and II of Jaan Puhvel's Hittite etymological dictionary (in the series Trends in Linguistics, Documentation, Vol. 1; Berlin/ New York/Amsterdam: Mouton), covering the words beginning with A, E, and I, was published in 1984. Puhvel's work is more philologically oriented and more personal and eclectic in its presentation of the etymologies than Tischler's. d) Greek: There are two recent Greek etymological dictionaries: one by the Swedish Indo-Europeanist Hjalmar Frisk — Griechisches etymologisches Wörterbuch (3 volumes; Heidelberg: Carl Winter, 1960 — 1972); the other by the French Hellenist Pierre Chantraine — Dictionnaire etymologique de la langue grecque. Histoire des mots (4 volumes in 5 parts, completed by Jean Taillardat, Oliver Masson and Jean-Louis Perpillou, edited by Michel Lejeune; Paris: Klincksieck, 1968 — 1980). To the parts covering A to Κ in these works, Guy Jucquois and Bernard Devlamminck added a useful supplement in 1977: Complements aux dictionnaires etymologiques du grec ancien (Louvain: Peeters). e) Italic: Except for the new editions of A. Walde's Lateinisches etymologisches Wörterbuch by J. Β. Hofmann (3rd edition completed in 1956; Heidelberg: Carl Winter) and A. Ernout and A. Meillet's Die-

432

Edgar C. Polome

tionnaire etymologique de la langue latine (4th edition by J. Andre; Paris: Klincksieck, 1979), no important new etymological work has appeared in the field of Italic, but Jürgen Untermann is preparing an etymological dictionary of Osco-Umbrian, as indicated above. f) Celtic: The Lexique etymologique de l'irlandais ancien, begun by J. Vendryes under the joint sponsorship of the Dublin Institute for Advanced Studies and the Centre National de la Recherche Scientifique in Paris, progresses very slowly: since 1959, only 6 fascicles covering the letters A, B, C, Μ, N, O, P, R, S, T, U have appeared, although the work is continued by E. Bachellery and P.-Y. Lambert (cf. their note in Bammesberger 1983: 17 — 24). In the meantime, other works have appeared: Leon Fleuriot, Dictionnaire des gloses en vieux-breton (Paris: Klincksieck, 1964); Enrico Campanile, Profilo etimologico del cornico antico (Pisa: Pacini, 1974). A comprehensive Dictionnaire etymologique du breton ancien, moyen et moderne. Origine et histoire des mots (Rennes: Ogam-Celticum, 1973 — ) was begun by Christian Guyonvarc'h, but its publication seems to have been temporarily interrupted after the first 480 pages (covering about half of "A") had appeared. g) Germanic: Although there is no recent etymological dictionary of Germanic as such (one is in preparation by E. C. Polome), part of the material is covered by Elmar Seebold's Vergleichendes und etymologisches Wörterbuch der germanischen starken Verben (The Hague: Mouton, 1970). In the individual languages, a number of important new dictionaries have been published in the post-World War II decades, namely: W. P. Lehmann's Gothic etymological dictionary (1986), which completely revises and updates Sigmund Feist's Vergleichendes Wörterbuch der gotischen Sprache (3rd edition; Leiden: Ε. J. Brill 1939) and incorporates the useful additions of Bernard Devlamminck and Guy Jucquois' Complements aux dictionnaires etymologiques du gotique. 1. A — F (Louvain: Peeters, 1977). The three works that have appeared in the field of Old Norse are very different in scope and approach: F. Holthausen's Vergleichendes etymologisches Wörterbuch des Altwestnordischen (Göttingen: Vandenhoeck & Ruprecht, 1948) shows, as already indicated above, the usual terseness characterizing the work of its author, whereas the Isländisches etymologisches Wörterbuch (Berne: Francke, 1956) of Johannesson is narrowly dependent on Pokorny's (1959) etymologies. On the contrary,

Etymology

433

Jan de Vries endeavors to take an original approach in his Altnordisches etymologisches Wörterbuch (Leiden: Ε. J. Brill, 1961; 2nd edition 1962), but relies too heavily on the disputable views of Jost Trier on early Germanic and Indo-European society in correlation with word history. As far as English is concerned, The Oxford dictionary of English etymology (Oxford: Clarendon Press, 1966) by C. T. Onions, with the assistance of W. G. S. Friedrichsen and R. W. Burchfield, is essentially expanding and updating the valuable etymological notes of the great Oxford English dictionary, and the Concise Oxford dictionary of English etymology (Oxford: Clarendon Press, 1986) is merely a revised abridgment of the same, edited by T. F. Hoad. The two-volume Comprehensive etymological dictionary of the English language (Amsterdam/London/New York: Elsevier, 1966 — 1967) by Ernest Klein is a useful, but basically uncritical compilation of currently accepted etymologies. Interesting and quite original is the effort of Calvert Watkins to regroup under their "roots" the English terms listed in the American heritage dictionary of the English language (Boston 1969) in a Dictionary of Indo-European roots (Boston: Houghton Mifflin, 1985). For Old English, Alfred Bammesberger, who is preparing an etymological dictionary to replace Holthausen's Altenglisches etymologisches Wörterbuch (Heidelberg: C. Winter, 1934; 2nd edition 1963), has published a couple of important preliminary studies: Beiträge zu einem etymologischen Wörterbuch des Altenglischen (Heidelberg: C. Winter, 1979) — corrections and additions to Holthausen's work — and English etymology (Heidelberg: Carl Winter, 1984) — a survey of phonology and word-formation. For Dutch, Jan de Vries compiled a Nederlands etymologisch woordenboek (Leiden: Brill, 1971) on the model of his Old Norse dictionary (cf. Polome in Bammesberger 1983: 217 — 228). A most useful complement to Dutch etymological dictionaries is provided by S. P. E. Boshoff and G. S. Nienaber, Afrikaanse Etimologiee (Pretoria: Suid-Afrikaanse Akademie vir Wetenskap en Kuns, 1967). The Etymological glossary to the Old Saxon Heliand by Samuel Berr (Bern and Frankfurt: H. Lang, 1971) looks more like a data-base for an etymological study of the lexicon of the Old Saxon poetic language as presented by Edward H. Sehrt in his Vollständiges Wörterbuch zum Heliand und zur altsächsischen Genesis (Göttingen: Vandenhoeck & Ruprecht, 1925; 2nd edition 1966), as it merely lists forms and meanings, with abbreviated bibliographical references, without any comment whatsoever.

434

Edgar C. Polome

For German, until recently, only the revised reprints of Kluge's etymological dictionary (latest: 1975) were available besides concise popular works like Ernst Wasserzieher's Woher? Ableitendes Wörterbuch der deutschen Sprache (Bonn: Ferd. Dümmler, 1974; 18th edition, revised by W. Betz) — published in the G.D.R. as Kleines etymologisches Wörterbuch der deutschen Sprache (Leipzig: Bibliographisches Institut, 1979) — or Duden Etymologie: Herkunftswörterbuch der deutschen Sprache, edited by G. Drosdowski, P. Grebe et al. (Mannheim/Wien/Zürich: Bibliographisches Institut, 1963). In spite of the fact that it was prepared for a linguistically more sophisticated readership, the Concise German etymological dictionary of M. O'C. Walshe (London: Routledge & Kegan Paul, 1951) remains fairly limited in scope and coverage, although it also includes Middle High German terms no longer extant in the modern language. In 1986 the first fascicle of the comprehensive Deutsches etymologisches Wörterbuch of Rolf Hiersche (Heidelberg: Carl Winter) appeared, which deals with the word history as well as the etymological problems in greater depth than any work hitherto in the field of Germanic studies. Complemented by the excellent Etymologisches Wörterbuch des Althochdeutschen of A. Lloyd and Ο. Springer (Göttingen: Vandenhoeck & Ruprecht), whose first volume appeared in 1988, it will provide German with some of the best modern etymological coverage. h) Baltic: Two major works have appeared: Ernst Fraenkel's Litauisches etymologisches Wörterbuch (Heidelberg: Carl Winter, 1962 — 1965) and V. Ν. Toporov's Prusskij jazyk: Slovar' (Moskva: Nauka, 1975-), up to the letter L. i) Slavic: While the publication of the monumental Lexicon linguae palaeoslovenicae (Slovnik jazyko staroslovenskeho; Praha 1966 — ) progresses slowly, only one of the three comparative dictionaries of Slavonic appears to move steadily ahead: Fourteen volumes have appeared of the Etimologiceskij slovar' slavjanskix jazykov — praslavjanskij leksiceskij fond, edited by Oleg N. Trubacev (Moskva: Nauka, 1974 — ), but the Vergleichendes Wörterbuch der slawischen Sprachen of Linda Sadnik and Rudolf Aitzetmüller (Wiesbaden: O. Harrasowitz) has come to a stop after the first volume (A/B), which took twelve years (1963 — 1975) to produce. Also very slow is the publication of the Polish Slownik praslowianski, commissioned by the Linguistic Committee of the Academy of Sciences (Wroclaw: Ossolineum, 1974 — [latest volume: V, 1984, "D"]).

Etymology

435

In the individual Slavic languages, Vasmer's Russisches etymologisches Wörterbuch (Heidelberg: C. Winter, 1953 — 1958) and its Russian translation, edited by Ο. N. Trubacev (2nd edition; Moscow 1986 — 1987), is being complemented rather than replaced by the Etimologiceskij slovar' russkogo jazyka published by the Moscow Lomonosov University Philosophical Faculty since 1963, as the latter's presentation and discussion of non-Slavic cognates appears to be much more limited. In Polish, on the contrary, Aleksander Brückner's Slownik etymologiczny jgzyka polskiego (Krakow 1927; reprint: Warszawa 1957) is definitely being superseded by Franciszek Slawski's Slownik etymologiczny j§zyka polskiego, whose publication in Cracow (since 1952) seems to have been temporarily interrupted in 1982, after volume 5 (up to the end of "L") came out. There is a detailed etymological dictionary of Serbo-Croatian in three volumes by Petar Skok: Etimologijski rjecnik hrvatskoga Hi srpskoga jezika (Zagreb: Jugoslovenska akademija umetnosti, 1971 —1973; additional index volume, 1974), edited by M. Deanovic and L. Jonke. For Czech, Vaclav Machek, Etymologicky slovnik jazyka ceskeho (Praha: Academia, 1968) offers more substantial and personally elaborated etymological comments than the older work with the same title by Josef Holub and Frantisek Kopecny (Praha: Statni nakladatelstvi, 1952). For Bulgarian, the concise Etimologiceski i pravolisenü recnikü na bülgarskija knizovenü ezikü of St. Mladenov (Sofia, 1941) is being replaced by an elaborate Bülgarski etimologicen recnik under the redaction of V. Georgiev, I. Gülübov, J. Zaimov and S. Ilcev (Sofia: Bülgarska Akademija Nauk, 1971 —; 3 volumes published hitherto). j) Albanian: Only a selection of 250 "core" terms is examined by Martin E. Huld in his Basic Albanian etymologies (Columbus: Slavica, 1984). k) Armenian: R. Solta's Die Stellung des Armenischen im Kreise der indogermanischen Sprachen (Vienna: Mechitharists, 1960) is an examination of the Indo-European component of the Armenian lexicon, which is also the focus of the new etymological dictionary compiled by John A. C. Greppin (the first fascicle covering the letter "A" appeared in 1983 in Volume 141 of Bazmavep — Revue d'etudes armeniennes (Venice: Saint-Lazare), 235 — 322). 1) Tocharian: The Lexique etymologique des dialectes tokhariens of Albert Joris van Windekens, published in the Bibliotheque du Musion

436

Edgar C. Polome

(Vol. 11; Louvain 1941) has been replaced by Volume I of his Le tokharien confronte avec les autres langues indo-europeennes. La phonetique et le vocabulaire (Louvain: Centre International de Diabetologie Generale, 1976). An Etymologicum Tocharicum, thoroughly different in its conception, is being prepared for publication by Werner Winter (cf. his remarks in Bammesberger 1983: 318 — 325). In fields like the Romance languages etymological research also made considerable progress after World War II, and here, too, a number of important new etymological dictionaries were published, while work continued on older, more ambitious projects and popular dictionaries were revised and reprinted. Restricting ourselves to just a few of those, let us mention the outstanding Diccionario critico etimologico de la lengua castellana of Joan Corominas (4 volumes; Madrid: Gredos/Bern: Francke, 1954 — 1957), of which an abridged form appeared in 1961 (Madrid: Gredos) and quickly went through a second (1967) and third (1974) revised edition; the whole work was then thoroughly recast in 5 volumes and published posthumously by Jose A. Pascual under the title Diccionario critico etimologico castellano e hispänico (Madrid: Gredos 1980 — 1983). There are several recent etymological dictionaries available for Italian, e.g., Giacomo Devoto's Avviamento alia etimologia italiana. Dizionario etimologico (Firenze: F. LeMonnier, 1966); Carlo Battisti and Giovanni Alessio's Dizionario etimologico Italiano (Firenze: G. Barbera, 1950 — 1957; reprinted 1975); Manlio Cortelazzo and Paolo Zolli's Dizionario etimologico della lingua italiana (5 volumes; Bologna: Zanichelli, 1979 — 1985; reprinted 1984 — 1988), but the only work providing a coverage fully corresponding to the methodological requirements of modern etymological research is Max Pfister, Lessico etimologico Italiano (Wiesbaden; Ludwig Reichert, 1979 — ), of which two volumes of the twenty planned are completed and a third is in progress. In the field of French, while the Altfranzösisches Wörterbuch of A. Tobler and E. Lommatzsch is moving towards completion, a Dictionnaire etymologique de l'ancien frangais is being prepared under the editorship of Kurt Baldinger (Quebec: Les Presses de l'Universite Laval, 1971—; cf. about this work and Baldinger's Dictionnaire onomasiologique de l'ancien occitan and Dictionnaire onomasiologique de l'ancien gascon, his contribution in Bammesberger 1983: 25 — 47). Meanwhile, work progresses on the monumental etymological dictionary of modern French initiated by Walter von Wartburg in the 1920s:

Etymology

437

Französisches etymologisches Wörterbuch. Eine Darstellung des galloromanischen Wortschatzes (vol. 1 [Bonn/Leipzig: F. Klopp, 1928]; later vols. [Basel: Helbing & Lichtenhahn]). In 1969 Ernst Gamillscheg published a completely reworked second edition of his Etymologisches Wörterbuch der französischen Sprache (Heidelberg: Carl Winter), while the often reprinted Dictionnaire etymologique de la langue frangaise of Albert Dauzat (Paris: Larousse; 10th edition, 1949) was replaced in 1964 by the Nouveau dictionnaire etymologique et historique (Paris: Larousse), edited by Jean Dubois and Henri Mitterand. The new edition of H. Titkin's Rumänisch-deutsches Wörterbuch by Paul Miron (Wiesbaden: Otto Harrassowitz, 1986 — 1989; three volumes) contains updated etymological notes, onomastic data and comments on linguistic geography, as well as references to the first documented attestation of each term. Worth mentioning are also a number of important dictionaries dealing with the languages or dialects of specific areas, e. g., Antoni M. Alcover and Francesch Β. de Moll, Diccionari catalä-valenciäbalear: inventari lexical y etimolögich de la lengua que parlen Catalunya espanyola ... (ten volumes; Palma de Mallorca 1930 — 1962; 2nd edition, vols. 1—2, 1962 — 1968); Max Leopold Wagner, Dizionario etimologico sardo (3 vols.; Heidelberg: Carl Winter, 1957 — 1964); etc. An exhaustive bibliography on etymology would also include the numerous monographs on specific topics such as semantic fields in definite languages or lexical items based on particular derivational patterns, to say nothing of the rich documentation on onomastics, from ancient hydronymy to theophoric place-names. They are listed in the relevant etymological dictionaries as well as in the current linguistic bibliographies.

References Bammesberger, Alfred (ed.) 1983 Das etymologische Wörterbuch. Fragen der Konzeption und Gestaltung (Eichstätter Beiträge. Abteil. Sprache und Literatur, Bd. 8). Regensburg: Friedrich Pustet. Bartholomae, Christian 1961 Altiranisches Wörterbuch. Berlin: Walter de Gruyter (Reprint of 1904 edition). Beck, Heinrich — Hans Jankuhn — Kurt Ranke — Reinhold Wenskus (eds.) 1976 Reallexikon der Germanischen Altertumskunde. Vol. 2. Berlin/New York: Walter de Gruyter.

438

Edgar C. Polome

Buchholz, Peter (see Hinz — Buchholz) — Jan de Vries 1961 Altnordisches etymologisches Wörterbuch. Leiden: E. J. Brill. 1971 Nederlands etymologisch woordenboek. Leiden: E. J. Brill. Cardona, George — Norman H. Zide (eds.) 1987 Festschrift für Henry Hoenigswald on the occasion of his seventieth birthday. Tübingen: Gunter Narr. Dahlstedt, Karl-Hampus (ed.) 1975 The Nordic languages and modern linguistics. Vol. II. Umeä, Sweden: Kunglige Skytteanska Samfundets Handlingar. Ducrot, Oswald — Tzvetan Todorov 1983 Encyclopedic dictionary of sciences of language. Trans, by Catherine Porter. Baltimore/London: The Johns Hopkins University Press. (Originally published as Dictionnaire encyclopedique des sciences du langage. Paris: Editions du Seuil, 1972,) Fraenkel, Ernst 1962 — 1965 Litauisches etymologisches Wörterbuch. Vol. 1. A — privekiüoti (1962); Vol. 2. privyketi —zvolgai (1965). Nachträge, Wortregister, Berichtigungen, Nachwort. Heidelberg: Carl Winter/Göttingen: Vandenhoeck & Ruprecht. Gamkrelidze, Thomas V. — Vjaceslav V. Ivanov 1984 Indoevropejskij jazyk i indoevropejcy. Rekonstrukcija i istoriko-tipologiceskij analiz prajazyka i protokultury. Tbilisi: Izdatel'stvo Tbilisskogo Universiteta (English translation by Johanna Nichols forthcoming). Hinz, Η. — Peter Buchholz 1976 "Bett und Bettzeug. 1. Archäologie. 2. Schriftliche Überlieferung", in: Beck - Jankuhn - Ranke - Wenskus (eds.), 316-320. Jazayery, Mohammed Ali — Werner Winter (eds.) 1988 Languages and cultures: Studies in honor of Edgar C. Polome (Trends in Linguistics: Studies and Monographs 36). Berlin: Mouton de Gruyter. Klein, Ernest 1966 —1967 Λ comprehensive etymological dictionary of the English language. Vol. 1. A - Κ (1966); Vol. 2. L - Z (1967). Amsterdam/London/New York: Elsevier Publishing Company. Kluge, Friedrich 1975 Etymologisches Wörterbuch der deutschen Sprache (21 st edition, reprint of 20th [1967] edition, revised by Walther Mitzka). Berlin/New York: Walter de Gruyter. Lehmann, Winfred P. 1986 A Gothic etymological dictionary, based on the third edition of " Vergleichendes Wörterbuch der Gotischen Sprache" by Sigmund Feist (Bibliography prepared by Helen-Jo J. Hewitt). Leiden: E. J. Brill. Lloyd, Albert L. — Otto Springer 1988 Etymologisches Wörterbuch des Althochdeutschen. Vol. 1. Göttingen: Vandenhoeck & Ruprecht. Malkiel, Yakov 1976 Etymological dictionaries: A tentative typology. Chicago/London: University of Chicago Press.

Etymology

439

Manessy-Guitton, Jacqueline 1988 "Les champs semantiques de la racine *H2eig- en grec", in: Jazayery — Winter (eds.), 4 1 9 - 4 3 7 . Mayrhofer, Manfred — Martin Peters — Oskar E. Pfeiffer (eds.) 1980 Lautgeschichte und Etymologie. Akten der VI. Fachtagung der Indogermanischen Gesellschaft, Wien, 24—29 September 1978. Wiesbaden: Ludwig Reichert. Ölberg, Hermann — Gernot Schmidt (eds.) 1985 Sprachwissenschaftliche Forschungen. Festschrift fur Johann Knobloch (Innsbrucker Beiträge zur Kulturwissenschaft 23). Innsbruck: Institut für Sprachwissenschaft der Universität. Pedersen, Holger 1962 The discovery of language: Linguistic science in the nineteenth century. Trans. John Webster Spargo. Bloomington: Indiana University Press. Pokorny, Julius 1959 Indogermanisches etymologisches Wörterbuch. Vol. I. Bern/München: Francke. Polome, Edgar C. 1975 "Old Norse religious terminology in Indo-European perspective", in: Dahlstedt (ed.), 126 — 136. (Reprinted in Language, Society, and Paleoculture: Essays by Edgar C. Polome. Anwar S. Dil (ed.), Stanford: Stanford University Press, 1982: 285-295). 1985 "Two etymological notes: 1. Gothic alew. 2. The Germanic divine name Nehalennia", in: Ölberg - Schmidt (eds.), 309-323. 1987 "Initial PIE in Germanic", in: Cardona - Zide (eds.), 3 0 3 - 3 1 3 . Ross, Alan S. C. 1958 Etymology, with special reference to English (The Language Library). London: Andre Deutsch. Seebold, Elmar 1980 "Etymologie und Lautgesetz", in: Mayrhofer — Peters — Pfeiffer (eds.), 431-484. Toporov, V. N. 1975Prusskij jazyk: slovar'. (4 vols, available). Vol. 1. Α - D (1975); Vol. 2. E - H (1979); Vol. 3. I - K (1980); Vol. 4. K - L (1984). Moskva: Nauka. Trubacev, Ο. N. (see Vasmer — Trubacev). Vasmer, Max 1953-1958 Russisches etymologisches Wörterbuch. Vol. 1. Α - Κ (1953); Vol. 2. L - S s u d a (1955); Vol. 3. S t a - J a (1958). Heidelberg: Carl Winter. Vasmer, Max — Ο. N. Trubacev 1986-1987 Etimologiceskij slovar' russkogo jazyka. Vol. 1. A-D (1986); Vol. 2. Ε-Muz (1986); Vol. 3. Muza-Sjat (1987); Vol. 4. T-Jascur (1987). (2nd. edition = Russian expanded and revised translation of Max Vasmer's Russisches etymologisches Wörterbuch, 1953 — 1958). Moskva: Progress. Vendryes, J. 1959— Lexique etymologique de l'irlandais ancien. (continued since 1978 by E. Bachellery and P.-Y. Lambert; six volumes available: A [1959], Β [1981], C [1987], Μ Ν Ο Ρ [1960], R S [1974], Τ U [1978]). Dublin: Institute for Advanced Studies/Paris: Centre National de la Recherche Scientifique.

440

Edgar C. Polome

Watkins, Calvert 1987 "How to kill a dragon in Indo-European", in: Watkins (ed.), 270-299. Watkins, Calvert (ed.) 1987 Studies in memory of Warren Cowgill (1929 — 1985): Papers from the Fourth East Coast Indo-European Conference, Cornell University, 6 — 9 June 1985. (Studies in Indo-European Language and Culture, NS Vol. 3) Berlin/New York: Walter de Gruyter. Zohary, Daniel — Maria Hopf 1988 Domestication ofplants in the Old World: The origin and spread of cultivated plants in West Asia, Europe, and the Nile Valley. Oxford: Clarendon Press.

Change of Languages

Language families and subgroupings, tree model and wave theory, and reconstruction of protolanguages Henry M.

Hoenigswald

1) The motif of linking languages by lines of ancestry and descent ("Italian is descended from Vulgar Latin": "Middle English is an older stage of Modern English") is ancient, familiar, and well embedded in intellectual folklore. Equally familiar is a further development: the classification and subclassification of "related" languages into families and subfamilies ("English and German are more closely related than are English and Latin") with the implication that the common "ancestor" of a family is also the ancestor of the (sub)ancestor of a subfamily, and that the subancestor represents an intermediate stage between its ancestor and its descendants. This mode of presentation reflects a traditional dependence on a particular graph-theoretical device which is employed in many fields of knowledge to depict either systematic relationships of some sort, or the sequence of events (often irreversible) in time (Platnick — Cameron 1977): the unstemmed, rooted tree (Hoenigswald 1973a, 1975,1987). In such a tree, the vertices stand for languages, observed, recorded, or inferred; the edges and paths for lines of descent; and their direction away from the root point for the passage of time (Stewart 1976). 2) It is an open question to what extent the reliance on the tree has been a factor in the selection of the data and in the formation of the concepts with which we endeavor to interpret language historically; metaphors and formalisms are notoriously tyrannical. However obvious the applicability of the tree schema to some aspects of language may be, it is also true that the cladistic (tree-oriented) point of view

442

Henry Μ.

Hoenigswald

was not invented for linguistics and that the earlier or concurrent function which it had in other disciplines was greatly relevant. Some of these interdisciplinary influences, which are part of general intellectual history, were acknowledged explicitly, as when August Schleicher proclaimed himself a Darwinist of sorts (Hoenigswald 1963; Koerner 1978). Other influences — probably far more genuine — existed within the personalities of the practitioners in the form of well-assimilated modes of thinking, too deeply ingrained even to be specifically discussed, as when the same Schleicher transferred the principle of the exclusively shared copying error from manuscript work to linguistics, where it surfaced as the principle of shared innovation. 3) In stemmatics (manuscript study), genealogy, and evolutionary biology the tree is a convenience employed for the purpose of graphically representing certain important relationships in time. There is no general pretense that it is capable, in and of itself, of representing all possible relationships. For instance, stemmata (trees showing the manuscript tradition of a text) cannot do justice to cases in which a scribe conflates several manuscript sources to produce his version: proper trees have no loops. Genealogical trees are essentially either parthenogenetic or restricted to the male line; if information about the other parent is incorporated the symbolism will have to transcend the tree schema. This is so obvious that it hardly ever leads to confusion — the metaphor of the tree is no more insidious in these fields than it is in graph theory itself. One reason for this is that the vertices and edges stand on the whole for well-circumscribed entities with clear physical boundaries: manuscripts, organisms (human individuals, etc.), species; copying activities, acts of procreation, mutations. Not so in language. A vertex is occupied by "a language", and an edge or path symbolizes "a line of descent" — entities of which it can precisely not be said that they are "given". Rather, they are created by definition, and there is always a suspicion that there may be circularity, and that the very relations which the tree conveys enter into the definitions of its vertices (Hoenigswald 1977). Language family trees are thus not only limited by the nature of the tree model; they are precarious to begin with. 4) This does not mean that language family trees are not useful devices so long as the impossible — a complete depiction of full history — is not asked of them. This lesson was, however, learned only slowly and painfully as linguistics came to be increasingly formalized over the last two centuries. The important work was naturally done in connection with specific technical problems, in Indo-European, Romance, and

Language families and subgroupings

443

occasionally elsewhere. Porzig (1954, Chapter 1) provides informative guidance through this history. After a period of more or less intuitive use of the tree model, certain difficulties began to be discovered, notably by H. Ebel (1852) and A. Pictet (1859): it turned out that affinities among related languages do not necessarily occur in mutually exclusive ways (e. g., between languages A Β C D against languages Ε F, or between A Β against only C D) but also, for different features, in overlapping fashion (the preceding, but also A B C against D Ε or the like). This state of affairs cannot be formulated in terms of successive bifurcations or multifurcations. Not that the overlapping is necessarily altogether random, certainly not in the Indo-European material then studied (other language families such as Finno-Ugric present a different picture). Affinities of the telling kind (see below) exist between neighboring groups of speakers, thus producing a chain on the geographical map. J. Schmidt (1872) made the problem famous by advocating the so-called "wave theory" to replace the tree; but his final dictum to the effect that "metaphors [Bilder] have but little value in scientific [linguistic] discourse" should never be forgotten. In 1884 came K. Brugmann's work in which we have the most influential (cf. Dyen 1953, 1978), though not the first, statement of the requirement that account be taken not of shared properties (which could, after all, be retained properties) but of shared innovations — a requirement, as we have seen, 3), with its own historical interest, doctrinal as well as practical. It has sometimes been held (Watkins 1966; Markey 1976) that significant retentions should be accorded the same standing. We must, however, remember that retentions and innovations are not independent phenomena but converses. An innovation is a non-retention, and while shared retentions are compatible with a subgrouping, innovations are indicative of one. 5) Thus the lines were drawn, but the consequences varied. Porzig's chronicle (1954) may be interpreted as showing that those practitioners who did not simply give one particular sharing an arbitrary preference over others, thus clinging to the tree schema at all costs, became cautious and refused to recognize the existence of subgroupings within Indo-European upward from the obvious subfamilies like Slavic, Germanic, and so forth. Others developed important refinements. In the most famous instance, it was found that a sharing, namely that of the distinction between e, a, and ο in Armenian, Greek, Italic, and Celtic, was only a retention, and that it was Indo-Iranian with its uniform a that had innovated. Or it was found (if not always remembered in a

444

Henry Μ.

Hoenigswald

polemic) that superficially identical innovations in two descendant languages were not identically ordered and did not, therefore, constitute a sharing. Both in Osco-Umbrian (but not in the otherwise "close" Latin) and in Aeolic Greek (but not in most other forms of Greek) *kw merges with *p into ρ before a front vowel, yet in the sequence *-VkwetV-, *-kw- was apparently still intact when *-e- (in this position) was lost in Osco-Umbrian, since the outcome visible in the texts is -kt- and not -pi-. In other words, a proper application of the "comparative" method and of internal reconstruction can sometimes remove an apparent overlapping. (That the innovations discussed here and elsewhere are in such large numbers sound changes rather than morphemic-semantic changes does not mean that the latter are less important when they can be identified. It is only a tribute to the fact that sound changes, insofar as they merge contrasts, can be recognized as innovations ipso facto [Hoenigswald 1977].) 6) It is not always the case, however, that there is a tree to be recovered. Scholars are fairly well agreed (Bloomfield 1933; Bynon 1977) that only "clear cleavage" of speech communities can be expected to result in non-overlapping innovation. Conversely, they agree that incidence of true overlapping could be taken as an indication that there was no clear break so that innovations could have spread over an area in which some other innovations had already taken place (Goetze 1941). This is a situation in which the line of descent is not well defined and where the celebrated distinction between straight-line material and dialect borrowing is weak. It was also pointed out that the presence of sharp language boundaries in historical times does not in itself prove a clear prehistoric cleavage since it can result from the disappearance of transitional dialects between two centers. Related languages may, then, have well-defined boundaries and still exhibit the contradictory overlappings which frustrate the setting up of subancestries. 7) Even where there was sufficiently clear cleavage, innovations may be duplicated in separate events. If such duplication is not merely accidental it may be seen as related to either a universal or to a typological factor. Some changes may be "natural" or "trivial". Others may be semi-predictable from the state of affairs as it existed before the division of the community, as when an existing allophonic diversification makes it probable for a given phonemic entity to undergo real ("phonemic") sound change along the fissures already in phonetic existence. As for typological factors, it may be true that they are best

Language families

and subgroupings

445

discussed as setting target structures for linguistic areas in which not only unrelated but also related languages abut, and not as entailing specific changes out of the several which might be said to serve the same target. It is indeed in cases where the specific machinery differs that we can most easily study the contrast between a genealogical and a typological classification. 1 Nevertheless, the chances for a duplication in machinery as well are considerable (Dyen 1978). 8) Textbook treatments of these problems are often couched in the old terms of the family tree and the wave model, with an insistence that these do not contradict each other, and with an implicit realization that they are graphic representations of a simplified state of affairs. However, writers are often reluctant to admit that the simplification is forced on the observer by no other factor than his or her subservience to the model. The challenge in this insidious situation is to analyze the simplifications and to think of other formalisms — not necessarily geometrical ones — to gain heuristic advantages. Among the more explicit contributions are Pulgram (1953) and Bynon (1977). A geometric symbolism of a new kind was proposed by Southworth (1964; taken up in Anttila 1972). More fundamental questions were raised by Swadesh's glottochronological proposal (see Hymes 1971, with a complete bibliography of Swadesh's work since ca. 1950) and, especially, through its more recent development by Dyen (1975) and his associates. In some sense, statistical properties (e. g., vocabulary sharing under sound change) of related languages can be expressed in tree form in such a way as to allow a chronological interpretation which converges with the interpretation of the conventional tree that results from the comparative method (Hoenigswald 1973b, 1987). In addition, the conventional tree itself, with the overlappings that leave some trees in a problematic state, may be subjected to certain statistical considerations (Dyen 1978). 9) For reasons hinted at above, 5), the reconstruction of protolanguages and subancestors by such instruments as the comparative method is primarily concerned with the distribution of phonological contrast and homonymy among the morphs. On morphemic matters — inflection, derivation, syntax, semantics — algorithms of a comparable kind are not really available. Our uniformitarian beliefs oblige us, however, to interpret the immediate results of our formal procedures beyond the boundaries of mere non-contradictoriness, under the guidance of our knowledge, such as it is, of universale and typologies. The best understanding of the extent to which reconstructions are

446

Henry Μ.

Hoenigswald

"only formal" or, on the other hand, "concrete" (Lehmann — Zgusta 1979) could be obtained from an analysis of the rare instances in which something very close to the protolanguage is accessible independently, as with the Romance languages and Latin. It is at any rate impossible to maintain that our reconstructions are nothing but formulas for observed correspondences, as if correspondences had ever been collected without the thought of concrete antecedents. 10) The record abounds with substantive work on the classification and subclassification of languages in many parts of the world. The reference works on the languages of the world (Meillet — Cohen 1952 and later) are, to be sure, usually not explicit enough on points of methodological interest. For certain much-discussed problems in IndoEuropean we refer once again to Porzig (1954). Here the subclassification controversy concerning the Dardic languages of Gilgit and Kashmir is omitted, presumably as taking place inside a recognized subdivision, Indo-Iranian. Similarly, the Ingvaeonic question, socalled, is a classification problem inside the Germanic group (Markey 1976). The decipherment of the Linear Β script gave rise to an elaborate reexamination of the ancient Greek dialects (Risch 1955; Cowgill 1966). Ancient Indo-European Dialects (Birnbaum — Puhvel 1966, including Beeler, Birnbaum, Collinder, Cowgill, Emeneau, Hamp, Lane, Lehmann, Polome, Puhvel, Senn, Watkins, Winter, all 1966) was the result of a conference in which the Indo-European picture was comprehensively surveyed. Activity in other fields has become more and more lively; as a sample we mention Greenberg's restatement of African and "Semito-Hamitic" languages (1963, etc.) and the first steps that have been taken by Gelb (1977) and others (Cagni 1981) to fit the discoveries at Ebla in Syria to the known Semitic evidence.

Note 1 In Great Lakes Algonquian the same typological target — a t:n rather than a t : 0 : n subsystem — was attained in some languages by merging *θ with *t and in others by merging *θ with *n.

Language families and subgroupings

447

References Abraham, Werner (ed.) 1975 Vt Videam: Contributions to an understanding of linguistics. For Pieter Verbürg on the occasion of his seventieth birthday. Lisse: The Peter de Ridder Press. Ageeva, R. A. 1966 "O klassifikacii afrikanskix jazykov Dz. Grinberga (Obzor)", Jazyki Afrikv. 281-297. Ambrosini, R. 1976 Introduzione alia linguistica storica. Pisa: Libraria Editrice Athenaeum. Anttila, Raimo 1972 An introduction to historical and comparative linguistics. New York/London: Macmillan. Beeler, Madison S. 1966 "The interrelationships within Italic", in: Birnbaum — Puhvel (eds.), 51-58. Bender, M. L. 1976 "Genetic classification of languages: Genotype vs. phenotype", Language Sciences 43: 4 — 6. Birnbaum, Henrik 1966 "The dialects of common Slavic", in: Birnbaum — Puhvel (eds.), 153-197. 1978 "Linguistic reconstruction", in: Dressler — Meid (eds.), 475—477. Birnbaum, Henrik — Jaan Puhvel (eds.) 1966 Ancient Indo-European dialects. Berkeley/Los Angeles: University of California Press. Bloomfield, Leonard 1933 Language. New York: Henry Holt. Bonfante, Giuliano 1931 I dialetti indoeuropei (Annali del Reale Istituto Orientale di Napoli Vol. 4, No. 9). Napoli (Reprinted Brescia Paideia, 1976). Bopp, Franz 1853 Über die Sprache der alten Preussen in ihren verwandtschaftlichen Beziehungen. Berlin: F. Dümmler. Brogyanyi, Bela (ed.) 1979 Studies in diachronic, synchronic, and typological linguistics. Festschrift for Oswald Szemerenyi on the occasion of his sixty-fifth birthday (Amsterdam Studies in the Theory and History of Linguistic Science IV. Current Issues in Linguistic Theory). Amsterdam: John Benjamins. Brugmann, Karl 1884 "Zur Frage nach den Verwandtschaftsverhältnissen der indogermanischen Sprachen", Internationale Zeitschrift für allgemeine Sprachwissenschaft 1: 226-256. Bynon, Theodora 1977 Historical linguistics. Cambridge: Cambridge University Press.

448

Henry Μ. Hoenigswald

Cagni, Luigi (ed.) 1981 La lingua di Ebla. Atti del Convegno Internazionale, Napoli, 21—23 Aprile 1980 (Istituto Universitario Orientale. Seminario di Studi Asiatici, Series minor, 14). Napoli: Istituto Universitario Orientale. Cameron, Η. Don 1987 "The upside-down cladogram: Problems in manuscript affiliation", in: Hoenigswald — Wiener (eds.), 227 — 242. Christie, William M. (ed.) 1976 Current progress in historical linguistics (North-Holland Linguistics Series 31). Amsterdam: North-Holland. Cole, Roger W. (ed.) 1977 Current issues in linguistic theory. Bloomington/London: Indiana University Press. Collinder, Björn 1966 "Distant linguistic affinity", in: Birnbaum - Puhvel (eds.), 199-200. Cowan, William (ed.) 1982 Eleventh Algonquian Conference. Ottawa: Carleton University. Cowgill, Warren C. 1966 "Ancient Greek dialectology in the light of Mycenaean", in: Birnbaum — Puhvel (eds.), 7 7 - 9 6 . Delbrück, Berthold 1869 Review of Wörterbuch der indogermanischen Grundsprache, by A. Fick. Zeitschrift für vergleichende Sprachforschung 18: 73 — 80. 1881 Introduction to the study of language. A critical survey of the history and methods of comparative philology of Indo-European languages. Transl. of Einleitung in das Sprachstudium by Eva Channing. (Leipzig: Breitkopf und Härtel, 1880). Leipzig: Breitkopf und Härtel. (New edition with a foreword and selected bibliography by E. F. K. Koerner [Amsterdam Studies in the Theory and History of Linguistic Science I. Amsterdam Classics in Linguistics 8] Amsterdam: John Benjamins Β. V., 1974.) Devine, A. M. 1971 "Ausnahmslosigkeit und Stammbaumtheorie", Lingua 26: 348 — 369. Devoto, Giacomo 1962 Origini indeuropee. Firenze: Sansoni. Dressier, Wolfgang — Wolfgang Meid (eds.) 1978 Proceedings of the 12th International Congress of Linguists, 28 Aug—2 Sept 1977, Vienna. Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck. Durante, Marcello 1977 "Tipicitä della ricostruzione linguistica", in: Simone — Vignuzzi (eds.), 9-17. Dyen, Isidore 1953 Review of Malgache et Manjaan by Otto Ch. Dahl, Language 29: 577 — 590 (reprinted in his Linguistic Subgrouping and Lexicostatistics. The Hague/ Paris: Mouton, 1975, 2 9 - 4 9 ) . 1973a "The impact of lexicostatistics on comparative linguistics", in: Dyen (ed.), 75-84.

Language families and subgroupings 1978

449

"Subgrouping and reconstruction", in: Jazayeri — Polome — Winter (eds.), 3 3 - 5 2 . Dyen, Isidore (ed.) 1973b Lexicostatistics in genetic linguistics. Proceedings of the Yale Conference. Yale University, April 3—4, 1971 (Janua Linguarum: Studia Memoriae Nicolai van Wijk Dedicata; Series major, 62). The Hague/Paris: Mouton. Ebel, Η. 1852 "Celtisch, Griechisch, Lateinisch", Beiträge zur vergleichenden Sprachforschung 1: 4 2 9 - 4 3 7 . Elmendorf, William 1973 "Lexical determination of subgrouping", in: Dyen (ed.): 108 — 117. Emeneau, Murray B. 1965 India and historical grammar (Annamalai University Department of Linguistics Publication No. 5). Annamalainagar, India: Annamalai University. (pp. 1—24 reprinted as "Diffusion and evolution in comparative linguistics" 66 — 84, and pp. 25 — 75 reprinted as "India and linguistic areas 126 — 66)" in: Language and linguistic area: Essays by Murray Emeneau. Ed. Anwar S. Dil. Stanford: Stanford University Press. 1980. 1966 "The dialects of Old Indo-Aryan", in: Birnbaum — Puhvel (eds.), 123-138. *Ezard, Bryan 1977 "Lexicostatistical and other methods of classifying languages", in: Proceedings of the S. 1. L. Consultants' Seminar, 49 — 63. Ukarumpa: Summer Institute of Linguistics. Fairbanks, Gordon H. 1969 "Language split", Glossa 3: 4 9 - 6 6 . Fick, August 1873 Die ehemalige Spracheinheit der Indogermanen Europas. Göttingen: Vandenhoeck und Ruprecht. *Fodor, Istvän 1966 The problems in the classification of the African languages: Methodological and theoretical conclusions concerning the classification system of Joseph H. Greenberg (Studies in Developing Countries). Budapest: Center for Afro-Asian Research of the Hungarian Academy of Sciences. Gamkrelidze, Τ. V. 1978 "Language typology and linguistic reconstruction", in: Dressler — Meid (eds.), 4 8 0 - 4 8 2 . Gelb, Ignace J. 1977 Thoughts about Ebla: A preliminary evaluation (Syro-Mesopotamian Studies 1/1). Malibu: Undena Publications. *Gercenberg, L. G. 1978 "K voprosu ο klassifikacii indoevropejskich jazykov", Izvestija Akademii Nauk SSSR 37: 387-392. *Giginejsvili, Β. ΚΙ 972 "Comparative reconstruction and the problem of the parent-language", Voprosy jazykoznanija 4: 40 — 52. Goddard, Ives 1982 "Eastern Algonquian as a genetic subgrouping", in: Cowan (ed.), 143-158.

450

Henry Μ. Hoenigswald

Goetze, Albrecht 1941 Review of Development of the Canaanite dialects by Ζ. S. Harris, Language 16: 167-170. Göschel, J. — Pavle Ivic — K. Kehr (eds.) 1977 Dialekt und Dialektologie: Ergebnisse des Internationalen Symposiums "Zur Theorie des Dialekts", Marburg/Lahn, 5 — 10 September 1977 (Zeitschrift für Dialektologie und Linguistik Beiheft 26). Wiesbaden: Steiner. Greenberg, Joseph H. 1955 Studies in African linguistic classification. New Haven: Compass Publication Company. 1963 The languages of Africa (International Journal of American Linguistics, Vol. XXIX. 1, Part 2 = Indiana University Research Center in Anthropology, Folklore, and Linguistics, Publication 21). Bloomington: Indiana University Press. 1987 Language in the Americas. Stanford: Stanford University Press. Haas, Mary R. 1969 The prehistory of languages. The Hague: Mouton. Hall, Robert Α., Jr. 1950 "The reconstruction of Proto-Romance", Language 26: 6 — 27 (Reprinted in Readings in Linguistics, ed. M. Joos. Vol. I. Washington: American Council of Learned Societies, 1957, 303-314). 1964 Introductory linguistics. Philadelphia/New York: Chilton Books. Η amp, Eric P. 1966 "The position of Albanian", in: Birnbaum — Puhvel (eds.), 97 — 122. 1976 "On some principles of lexical-phonological comparison", in: Christie (ed.), 203-213. Haudry, Jean 1979 "Une illusion de la reconstruction", BSL 74: 175-189. Hetzron, Robert 1976 "Two principles of genetic reconstruction", Lingua 38: 39 — 108. Hockett, Charles F. 1958 A course in modern linguistics. New York: The Macmillan Company. 1981 "The phonological history of Menominee", Anthropological Linguistics 23: 5 1 - 8 7 . Hodge, Carleton Τ. 1974 "A set of postulates for comparative linguistics", Lacus Forum 1: 209 — 216. Hoenigswald, Henry M. 1963 "On the history of the comparative method", Anthropological Linguistics 5: 1 - 1 1 . 1973a Studies in formal historical linguistics (Formal Linguistics Series, Vol. 3). Dordrecht/Boston; D. Reidel Publishing Company. 1973b "The comparative method", in: Sebeok et al. (eds.), 51 - 6 2 . 1975 "Schleicher's tree and its trunk", in: Abraham (ed.), 157-160. 1977 "Intentions, assumptions, and contradictions in historical inguistics", in: Cole (ed.), 168-194. 1987 "Language family trees, topological and metrical", in: Hoenigswald — Wiener (eds.), 257-267.

Language families and subgroupings

451

Hoenigswald, Henry M. — Linda F. Wiener (eds.) 1987 Biological metaphor and cladistic classification. Philadelphia: University of Pennsylvania Press. Hymes, Dell H. 1971 "Bibliography of Morris Swadesh", in: Sherzer (ed.), 324-336. Jazayery, Mohammed Ali — Edgar Polome — Werner Winter (eds.) 1978 Linguistic and literary studies in honor of Archibald A. Hill. Vol. II. The Hague: Mouton. Koerner, E.F. Konrad 1978 "Toward a historiography of linguistics: 19th and 20th century paradigms", in: Toward a historiography of linguistics: Selected essays (Amsterdam Studies in the Theory and History of Linguistic Science, III. Studies in the History of Linguistics, Volume 19). Amsterdam: John Benjamins Β. V. Kretschmer, Paul 1896 Einleitung in die Geschichte der griechischen Sprache. Göttingen: Vandenhoeck und Ruprecht. Kroeber, Alfred L. — C. D. Chretien 1937 "Quantitative classification of Indo-European languages", Language 13: 83-103. 1939 "The statistical technique and Hittite", Language 15: 6 9 - 7 1 . Lane, George S. 1966 "On the interrelation of the Tocharian dialects", in: Birnbaum — Puhvel (eds.), 213-234. Lehmann, Winfred P. 1966 "On the grouping of the Germanic languages", in: Birnbaum — Puhvel (eds.), 1 3 - 2 8 . 1973 Historical linguistics (2nd edition). New York: Holt, Rinehart and Winston, Inc. Lehmann, Winfred P. — Ladislav Zgusta 1979 "Schleicher's tale after a century", in: Brogyanyi (ed.), 455 — 466. Leskien, August 1876 Die Deklination im Slavisch-Litauischen und Germanischen. Leipzig: Hirzel. Lord, Robert 1966 Teach yourself comparative linguistics. London: The English Universities Press Ltd. Lottner, E. 1858 "Über die Stellung der Italer innerhalb des europäischen Stammes", Zeitschrift für vergleichende Sprachforschung 7: 18—49; 162 — 193. *Makaev, Enver A. 1977 Obscaja teorija sravitel'nogo jazykoznanija. Moskva: Nauka. Markey, Thomas L. 1976 Germanic dialect grouping and the position of Ingvaeonic (Innsbrucker Beiträge zur Sprachwissenschaft 15). Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck. *Mazzola, Michael Lee 1978 "The Romance stammbaum in the South", Semasia 5: 23 — 35.

452

Henry Μ. Hoenigswald

Meillet, Antoine 1922 Les dialectes indo-europeens (2nd edition). Paris: Librairie ancienne Edouard Champion, Editeur. 1925 La methode comparative en linguistique historique (Instituttet for sammenlignende kulturforskning, Serie A: Forelesninger II). Oslo etc.: H. Aschehoug & Co. (W. Nygaard), etc. 1958 "Note sur une difficulte generale de la grammaire comparee", in: Linguistique historique et linguistique generale by Antoine Meillet, 36—43. Paris: Librairie Honore Champion. Meillet, Antoine — Marcel Cohen (eds.) 1952 Les langues du monde (New edition). Paris: Centre national de la recherche scientifique. Michelena, Luis 1963 "Lenguas y protolenguas", Acta Salmanticensia, Filosofia y Letras Vol. 17, no. 2. Monteil, V. 1965 "La classification des langues de l'Afrique", Bulletin de l'Institut Frangais d'Afrique Noire (Β) 27: 155-168. Pictet, Α. 1859 Les origines indoeuropennes. Paris: Cherbuliez. Pisani, Vittore 1940 Geolinguistica e indeuropeo (Atti della Reale Accademia Nazionale dei Lincei. Memorie, Serie VI, Classe di scienze morali, storiche, e filologiche, Anno cccxxxvi, vol. ix, fasc. ii, pp. 111 — 378). Roma. Platnick, Norman I. — H. Don Cameron 1977 "Ciadistic methods in textual, linguistic, and phylogenetic analysis", Systematic Zoology 26: 380 — 385. Polome, Edgar C. 1966 "The position of Illyrian and Venetic", in: Birnbaum — Puhvel (eds.), 59-76. Porzig, Walter 1954 Die Gliederung des indogermanischen Sprachgebiets. Heidelberg: Carl Winter. Priestly, Tom 1973 "Subgrouping in comparative reconstruction", Anthropological Linguistics 15: 299-323. Puhvel, Jaan 1966 "Dialectal aspects of the Anatolian branch of Indo-European", in: Birnbaum - Puhvel, 235-247. Pulgram, Ernst 1953 "Family tree, wave theory, and dialectology", Orbis 2: 67 — 72. Risch, Ernst 1955 "Die Gliederung der griechischen Dialekte in neuer Sicht", Museum Helve ticum 12: 61 —76 (reprinted in The Language and Background of Homer. Ed. G. S. Kirk, Cambridge/New York: Heffer/Barnes and Noble, 1964, 90-105). Sankoff, David 1973 "Parallels between genetics and lexicostatistics", in: Dyen (ed.), 90 — 105.

Language families and subgroupings

453

Schleicher, August 1861 —1862 Compendium der vergleichenden Grammatik der indogermanischen Sprachen. Weimar: Böhlau. Schmidt, Johannes 1872 Die Verwantschaftsverhältnisse der indogermanischen Sprachen. Weimar: Böhlau. Senn, Alfred 1966 "The relationships of Baltic and Slavic", in: Birnbaum — Puhvel (eds.), 139-152. Shevoroshkin, Vitalij V. — Thomas L. Markey (eds.) 1986 Typology, relationship, and time: A collection of papers on language change and relationship by Soviet linguists. Ann Arbor: Karoma. Simone, R. — U. Vignuzzi (eds.) 1977 Problemi della ricostruzione linguistica. Atti del Congresso Internazionale di Studi, Pavia, 1 — 2 ottobre 1975 (Publicazione della Societä di Linguistica Italiana 11). Roma: Bulzoni. Southworth, Franklin 1964 "Family tree diagrams", Language 40: 557 — 565. Stang, Christian S. 1973 Lexikalische Sonderübereinstimmungen zwischen dem Slavischen, Baltischen und Germanischen (Skrifter utgitt av det Norske Videnskaps-Akademi i Oslo, II. Hist.-filos. Klasse, N.S. 11). Oslo. Stewart, Ann Harlemann 1976 Graphic representation of models in linguistic theory. Bloomington/London: Indiana University Press. Strunk, Klaus 1977 "Heterogene Entsprechungen zwischen indogermanischen Sprachen", Zeitschrift für vergleichende Sprachforschung 90: 11 — 36. Sturtevant, Edgar H. 1939 "The pronoun *so *sa *tod and the Indo-Hittite hypothesis", Language 15: 1 1 - 1 9 . Swadesh, Morris 1971 The origin and diversification of language. Joel Sherzer (ed.), Chicago/New York: Aldine/Atherton. Tovar, Antonio 1977 "Die Bildung der indogermanischen Dialekte", in: Göschel — Ivic — Kehr (eds.), 143-154. Walde, Alois 1917 Über älteste sprachliche Beziehungen zwischen den Kelten und Italikern. Innsbruck: Wagner'sche k.k. Universitätsbuchdruckerei, R. Kiesel. Wang, William, S. Y. 1987 "Representing language relationships", in: Hoenigswald — Wiener (eds.), 243-256. Watkins, Calvert W. 1966 "Italo-Celtic revisited", in: Birnbaum - Puhvel (eds.), 2 9 - 5 0 . Wells, Rulon S. 1987 "The life and growth of language: Metaphors in biology and linguistics", in: Hoenigswald — Wiener (eds.), 39 — 80.

454

Henry Μ. Hoenigswald

Wiener, Linda F. 1987 "Phonetic and genetic classification", in: Hoenigswald — Wiener (eds.), 217-226. Winter, Werner 1966 "Traces of early dialectal diversity in Old Armenian", in: Birnbaum — Puhvel (eds.), 201-212. Wyatt, William 1970 "The prehistory of Greek dialects", Transactions of the American Philological Association 101: 557 — 632. Zeuss, Johann Kaspar 1837 Die Deutschen und die Nachbarstämme. München: Lentner.

The development of standard language (koine) and dialect: Language split and dialect merger Thomas L. Markey

In any science, any true science, there are always issues and non-issues, but the former consistently receive less attention than the latter. Nonissues continue to fascinate, while real issues have terminal interest. This is because issues are soluble, soluble in the several senses so brilliantly illustrated by Peter B. Medawar in his The art of the soluble, while non-issues are not. This accounts for the fact that the non-issues are perennially with us, while issues are not. The Flat Earth Society has persisted, while Joseph Priestley's theory of dephlogisticated air has not. Quite simply, Priestley failed to distinguish a problem from a counter-example. Non-issues feed off problems (and real or contrived insolubles), while issues (and theories) are either supported or refuted by counter-examples. Real issues are subject to proof and/or disproof, while non-issues are not. Moreover, the terms and propositions of issues are clearly defined and definable, while those of non-issues are not. Now, not to be mistaken, both issues and non-issues are subject to analysis, description, and even interpretation, but the interpretation that solves an issue merely reinterprets a non-issue. Witness generative semantics versus Extended Standard Theory in the transformational debate over recoverability. Issues, when solved, are consigned to the handbooks of science (as laws or formulae), while the non-issues continue to be chronicled in state-of-the-art reports or occupy countless articles and monographs as an interminable debate. The point of this lengthy prolegomenon is that the very topics encompassed by the title of this report are, by and large, non-issues, albeit crucial non-issues. I will explain. In the course of my explanation I will provide, naturally enough, a state-of-the-art report. 1

456

Thomas L. Markey

But a coda to this inaugural assessment is perhaps in order. Much, it seems to me, in contemporary linguistics is comprised of non-issues, and rather confused non-issues at that: the confusion of local with global in discourse analysis (e. g., focus/topic vs. ostension), the construal of an apodictic pragmatics as a natural conditioning of communication (i.e., confusion of network with field), and so on. The centrality of non-issues in linguistics may be due to the fact that linguistics is such a comparatively young science. Consider the discrediting of alchemy as chemistry refined its experimentation. But the objection of youth is vacuous if one considers Aristotle a linguist. The crux of the matter lies in the fact that linguistics has scarcely begun to ask real theoretical questions, questions about the thinking about thinking, about a principled contrast between representation and presentation. For now, it seems, what is physically available for the linguist is logically inaccessible (consider, in computers, the contrast between what is logically deleted and what is physically retained). The suggestion here — and the critical yardstick for the survey that follows — is that, once we have changed our frame of reference, then what is physically available will be logically accessible. Issues will thereby become distinct from non-issues and what have been designated as non-issues may well join the ranks of the soluble. We know that dialects merge, that languages split, or rather glide apart in time, and that standard languages may be born after long and careful cultivation (e. g., German in the course of the eighteenth century) or evolve rapidly as the result of precipitous socio-economic pressures (e.g., Modern Hebrew under the aegis of Eliezer Perelman a.k.a. Ben-Yehuda), but what has defied definition all along is "dialect". Traditionally, "dialect" was defined as a limited lect, a somehow socially or geographically (or both) restricted lect, a subvariety of some standard language, either hypothetical (as in the case of Afrikaans) or real (as in the case of French). 2 But is American English a dialect of British English? Is Friulan a dialect of Northern Italian or RhetoRomance? Such questions end in qualitative and quantitative absurdities, for precisely which or how many features, and from which linguistic level or levels, are required to determine linguistic proximity to or distance from some norm? Is Occitan merely a dialect of French, or an independent subspecies of Franco-Provencal? 3 What, in summary, is the nature of variation? What triggers it, and how can it be

Language split and dialect merger

457

measured? These were the basic questions posed by the sociolinguists, and principally by William Labov. 4 For nearly two decades now, Labovians have attempted to define split, merger, and variation in strictly quantitative terms. But, as I have pointed out elsewhere,5 their "glottometry" has failed to make the necessary distinctions between speech community (e. g., the Balkans) and language community (e. g., West Frisian), nature and norm, language and function, or to allow for non-regular variation or discontinuous (though non-catastrophic) change, or, in some principled way, to consider the fact that not all stigmatized features are regressive. More basically, however, societal pressure and prestige is not, as the Labovians have rather simplistically contended, everywhere correlated or correctable with linguistic norms for which there may well be no such thing as a random sample. Despite these (and other) objections — and I am not alone in leveling these charges — the Labovian brand of sociolinguistics has dominated the field for the past decade and enjoyed enormous and unquestioned influence in the investigation of dialect merger and the microcosmic effects of language split. Nevertheless, this school has failed to provide anything other than rather trivial quantitative definitions of the notions of dialect and standard language. Statistics are, as Pfaff (1973) has electrifyingly shown, merely indices of the superficial mechanisms of change and change processes, and they are inadequate indices at that. While they describe, they do not explain change. They do not come to grips with the inner, lingual dynamics of what is at stake in the course of split and merger or the construction of standards. The sociolinguistic fact that, say, the greatest percentage of people in Jonestown who use sheeps are red-headed males with an annual income in excess of $ 30,000.00 says nothing about the linguistic motivation behind sheep : sheep —> sheep : sheeps. Such sociolinguistics is mere social and linguistic taxonomy. It may predict new data, but not new phases. It is static, not dynamic, and is thus incommensurate with something as clearly dynamic as change and variation. In more refined scientific terms, this approach is based on the concept of the amplitude of transitional probabilities, and this is but a threshold, a stage one, to a more penetrating analysis. Fundamentally — the bottom line — the Labovian approach is a study of the entropies of microphenomena that may or may not (probability) promote dialect merger or split. Yet, if one were forced to categorize the Labovians philosophically, then one would surely characterize them as quantificational positivists.

458

Thomas L. Markey

In diametric opposition to the Labovians, who posit an essentially social basis for convergence/divergence, C.-J. Bailey and his former colleagues, Willi Mayerthaler (currently at the University of Klagenfurt) and Peter Mühlhäusler (currently at Oxford University), at the Technical University of Berlin, have attempted to probe the ideational motivations behind split, merger, and variation in time within the framework of what has been termed "developmental linguistics". 6 This is essentially a deterministic view of change. It is based on the concept of linguistic relativity, a notion that is reflected in the works of such distinguished predecessors as Wilhelm von Humboldt (typological relativity), Johannes Schmidt (the spatial relativity of the Wellentheorie), and Hugo Schuchardt (the relativity of Wörter und Sachen). Developmentalism involves a non-absolutist, gradualistic conception of markedness theory (more vs. less marked :: unmarked and not receptive to marking vs. unmarked and receptive to marking). The markedness of items is coreferenced to their environments (before and after), thereby providing a parametric for change. For example, more marked in an unmarked environment becomes less marked in a marked environment: m/ (-m)—>m/ ( + m). This is considered a natural change, where natural is defined as an inherent property and not as a trivial economy measure or a metric of simplicity. Natural, too, is the persistence of unmarked forms upon neutralization of a marked/ unmarked (more/less marked) contrast. Clearly, the goal of developmentalism is to predict new phases, not new data. Further notions of developmentalism are constructional iconicity (when a formal asymmetry refers to a deeper semantic asymmetry, e.g., the positive [big] versus the comparative [bigger]); processual and substantive universals; implicational hierarchies ranged along temporal continua: evaluative scaling in terms of naturalness; and feature stripping and loading. By and large, dialects (and languages) converge by stripping and diverge by loading. (Cf. creolization and decreolization.) Consider, for example, the loss of nominal gender in Modern English (as against Old English) as the dialects of Anglo-Saxon merged versus the gender loading (pronouns, nouns, and even verbs) of the Semitic languages as they split from some larger suprafamily of languages. The determinism inherent in developmentalism denies the possibility of randomness in dialect/language merger and split; there may well be con-fusion, but not unprincipled mixture. Take the case of sheep : sheep —> sheep : sheeps cited above. Rather than statistically identify a probable "social" cause for this "analogical" (in traditional terms) extension,

Language split and dialect

merger

459

the developmentalist would infer a strictly ideational basis for this transformation. Identical singular/plural marking is abnatural, for it is not constructionally iconic: The singular is usually, if not universally, less complexly marked than the plural; hence, the plural is "naturally" structurally greater than the singular, thereby denoting its structural or con-structural iconicity. Iconically encoded rules are natural, while non-iconically encoded rules are not. The shift to sheeps merely restores an inherent natural order and, in this sense, simplifies the grammar. While the color of hair, annual income level, relative social status, and degree of education of those who say sheeps may be anthropologically interesting, this detail is immaterial to linguistic insight. Un- or abnatural changes and variants are more marked than natural changes or variants. This essentially deterministic view of change clearly vitiates any abductive approach to merger and split and invalidates proportional models of analogy. 7 But what — a question rarely asked in linguistics — are the noetic, the knowledge-deriving and -building, bases of sociolinguistics and developmentalism in their searches for answers to the "non-issues" of merger/split and standardization? If Labovianism may be characterized as quantificational positivism, then the developmental approach may be represented as naturalistic determinism or, perhaps better, critical naturalism. For the developmentalist, natural changes are not merely quantitative. Moreover, there is an obvious tendency to reject categories of substance in favor of quality, relation, activity, and process as central, virtually metaphysical, notions. Here it should be recalled that the bulk of current mainstream theoretical inquiry has so far eschewed dynamic-panchronic processes in favor of static-synchronic descriptions (sold as interpretations) of individual languages. Labov and company have concentrated on probabilities, some of which are endorsed by empirical fact. Chomsky has concentrated on an essentially processual phenomenology of the moment, the strictly linear synchronic transformation, rule governance, and binding — "Time" and "Variation" having been excluded from the generative Weltanschauung by adherence to strict uniformity of the moment. Developmentalists deny the efficacy of the synchronic/diachronic dichotomy, and there is some textual evidence that, even for Saussure, this dichotomization was simply a pretheoretical prejudice. 8 Developmentalists are — once again in the more refined terms of the science of the quantum — concerned with the superpo-

460

Thomas L. Markey

sition of basic states (mlin) and the interference of effects, rather than (ä la Labov) the effects of interference. Labovian and developmentalist approaches meet in the realm of creolistics. In recent years, the proper assessment of creolization, that is, true, stratified creolization such as Negerhollands or Haitian French as opposed to fusion creolization such as English (Anglo-Saxon + Norman French), Yiddish (Alemannian + Hebrew-Aramaic), or Romontsch (Alemannian + Rheto-Romance + unidentified substratal residue), has been one of the most exciting and productive arenas for investigating split/merger and standardization. 9 Pidgins and Creoles do not result from some random mixture (see Mühlhäusler 1980). They reveal a principled reductionist reconstitution of language that reflects Haeckel's Law: ontogeny recapitulates phylogeny. By now it is abundantly clear (or should be) that Creoles do not result from mindless monogenetic and gradualistic replacement by relexification from prestigious superstrates (English, French, Portuguese, etc.) of the lexical contents of an unformated jargon. There is no conclusive evidence that some one lingua franca sequentially took on the garb of whatever prestige language it contacted, thereby giving rise to Portuguese Creole, English Creole, then French Creole, and so on. Nor are Creoles forged by maximally Europeanizing reflexification and minimal grammatical restructuring of some nebulous (West African, Polynesian, Arabic, etc.) basilectal input. Creoles are discontinuous phenomena. Yet, despite vast geographical separation and differences in inputs, they confront us with a basic uniformity, a fundamental universality of construction. For example, all true Creoles mark noun plurals by suffixation of the third plural anaphor, e.g., Papiamentu buki-nan 'bookthey' = '(the) books'. Such uniformity in morpho-syntactic strategy has prompted Derek Bickerton to assert that creolization evinces an innate, prewired and preprogrammed, basis for human language — a bioprogram that effectually maps the development of language in the species (see Bickerton 1981)10. Creolization is simultaneously macromerger and macrosplit. Moreover, Creoles can standardize by decreolization (e.g., Papiamentu) as they are transposed in time along continua from basilectal to acrolectal varieties, from varieties that are more to less creole-like, in the direction of some norm (e. g., Standard French).11 Perhaps more than anything else, creole studies have forced linguists to reconsider two basic topics, namely, typology and the continuity/

Language split and dialect

merger

461

discontinuity dilemma, matters that directly impinge upon merger and split. As far as typology is concerned, the wisest course is to admit that we are currently faced with a host of unanswered questions. To what extent do typological differences/similarities and incongruities/congruities impede or further split/merger? To what extent does typology constrain the limits of standardization? What criteria (syntactic, morpho-syntactic, or strictly morphological) take precedence in defining typologies, and which grammatical sectors (word order, deixis, diathesis) are more/less susceptible to restructuring (split/merger) upon contact? Even more exasperating is the question of whether or not (though it seems to be the case) certain typological strategies result in a superpositioning of split and merger themselves in what I have termed "compartmentalization", the reserved relation of a process or marker from one input for the lexicon from that input in a fusion situation (see Markey 1981: 9 —14).12 Note, for example, that in Cappadocian Greek, vowel harmony, a typological characteristic of Turkish but not of Greek, is restricted to Turkish loans. Consider, further, Anglo-Saxon leaves (plural) vs. Anglo-Norman foliage (collective), that is, the compartmentalization of count- vs. mass-based typologies, and note the co-occurrence restrictions on -age (e. g., plumage, postage) in English. For my money, some of the most exciting recent work in typology, work that has provided a vital background for answering the above questions, is that by Talmy Givon, initially in his On understanding grammar (1979) and subsequently in his Syntax: A functional-typological approach (1983). The semantics of split imply discontinuity and those of merger continuity, but of the two, discontinuity is by far the more subtle and conceptually complicated notion. For the physicist, the character of discontinuity is captured by indeterminancy relations and Bohr's Paradox. For the neo-evolutionists, such as Stephen Jay Gould, the results of discontinuity are termed "punctuated equilibrium", an obviation of the "missing links" of Darwinian gradualism. Discontinuity invalidates the sterile homologies of a static and contingent structuralism. Discontinuity invalidates the probative value of dependency relations and linearity for describing the enfolding of processes in time. Anyone can appropriately conjure up the image of discontinuity by focusing on the alternate perspectives of a Necker cube. But more appropriately here, note the discontinuity in the flip-flop from plural [ +count, — mass] to collective [ — count, + mass] throughout the history of the

462

Thomas L. Markey

Indo-European languages. Recall the compartmentalization of leaves/ foliage cited above. Discontinuities are splits and mergers without transitional phases. In some very enlightening recent work, Hansjakob Seiler, in conjunction with the project on universals at the University of Cologne, has pursued the problem of discontinuous change (though this is not his primary intention). He has foreshadowed an iconic mapping of discontinuity (e.g., see Seiler 1981).13 In effect, he has suggested that discontinuity may well be perceived only by an iconic linearity of disruption, by flip-flops in time (*=•) as a punctuated and punctuating gradualism. Note, for example, his definition of the apposite polarities in a morphosyntactic continuum as being respectively less marked vs. more marked and more grammaticalized vs. less grammaticalized (g/g):


m

m

Phase 1

Phase Ν

g

transitional phases

g

where A is the flip of Β (and conversely) in much the same manner as the circles on a Necker cube are discontinuous flips. In possession strategies, for example, simple juxtaposition (John house) would be Phase 1, while possession-marking on the verb would be Phase Ν {John has a house), and 1 —»Ν is the flip of Ν —> 1, by definition without transitional phases. Conceptual patterning such as flips, Necker cubes, and the like would, for now, seem to be one of the best ways of coming to grips with the discontinuity inherent in split and merger. Indeed, such typological mapping, in the mathematical sense of "typology", is reminiscent of the diagrammatics employed by Rene Thom to illustrate his catastrophe theory, our best morphology to date for comprehending discontinuous change. 14 But linguists, physicists, and evolutionists are not the only professionals perplexed by discontinuity. In the realm of literary criticism, deconstruction, particularly as advanced by Jacques Derrida, is really just an attempt to come to terms with discontinuity. Nature, it seems, periodically deconstructs

Language split and dialect merger

463

norms by discontinuous effects, and this is but another of the "effects" of interference. Finally, then, there is the "non-issue" of standardization, perhaps the most "non" of all the topics reviewed here. Interest in standardization has usually grown out of a fascination with bilingualism. Recall, for example, that the effluence of Einar Haugen's pioneering work on bilingualism was his inquiry into the riksmäljbokmällnynorsk controversy, that is, his investigation of the concessional evolution of a Standard Norwegian (see Haugen 1953, 1966).15 But awareness of bilingualism or diglossia is, in reality, an awareness of differences and similarities between tokens in reference to a common type. Closing the gap between tokens and types and elevating some one token to the level of general acceptance is all that standardization is really about. This assessment also accounts for why most discussion of "the standard" for a particular language is so tediously atomistic. Continuing with Norwegian, note Andre Bjerke's discussion of syd vs. sor 'south', where the former is considered Dano-Norwegian, the latter native (Bjerke 1962). So we have Sydpolen and Sorpolen and no norm, no standard, no logical editorial rule, but total chaos. Bjerke rejects SerAfrika as a violation of his particular Sprachgefühl, but accepts SorNorge 'South Norway', a native environment for a native sor. But why? Obviously because of markedness reversal. Standardization proceeds, by and large, by unmarking — nativizing — marked variants by placing them in unmarked, native environments. Standardization is the linguistic response to an identity crisis, and this is precisely what Otto Jespersen was talking about back in 1925 in his Mankind, nation and individual from a linguistic point of view, a work that became Haugen's bible. But the imposition of prescriptive rules for crystallizing a standard is, of course, not an inherently linguistic, but an extralinguistic, a sociolinguistic, matter. The extra-linguistic aspects of standardization take the form of language-specific histories. 16 There are, however, some very general and typical events in the histories of standardization. They have been admirably documented for German by Eric A. Blackall (1959). The sequential events are: vindication (acquisition of self-respect as a prestige lect) —• stabilization (accompanied by increasing explicitness) —• standardization of literary media —> recognition of stratification (e. g., Schriftsprache — Umgangssprache — Mundart). Standards may also be — and in fact usually are — forged on the several anvils of (1) political, (2) philosophical, or (3) purely intellectual expediency. An

464

Thomas L. Markey

example of (1) is Modern Hebrew; of (2) the increasing refinement of German in the nineteenth century at the hands of an idealism even more insistent in its bravado than the idealism of a Josiah Royce; and of (3) West Frisian, albeit still frequently regarded as a lingua rustica in comparison with Standard Dutch (see Markey 1981: 202 — 206). But to document individual histories is not my purpose here, which is rather to outline general principles and events. Another general principle/event is enforcement, either by an academy or some other recognized authority, or by nature. The natural case may be illustrated by an analogue. It is like students cutting across a campus by a favorite shortcut who finally see their convention officially condoned by an administration which paves the path, thereby leaving the students free to find another path. This brings me to my final point on standardization: to recognize it as a partner of theory if the process of nativization were viewed in terms of the various trends I have outlined here, namely, sociolinguistics, developmentalism, functional typology, and creolistics, as well as the inquiry into continuity versus discontinuity, all of which are but ways of analyzing the "non-issues" of language split and dialect merger.

Notes 1. The literature, even the relevant literature, on the forging of koines or lingua francas, language split, and dialect merger is so vast that merely to reference it would more than occupy the total space allotted. It is a sign of this embarras de recherche that standardization alone has been accorded its own journal, Language Problems and Language Planning, edited by Richard E. Wood. The varieties of English (and the split and merger of its dialects, as well as, tangentially, standardization) have also recently (1980) been given a journal, English World-Wide, edited by Manfred Görlach. Language as a communicative device (as opposed to the performative mechanism of Chomsky) has also very recently (1981) received its journal, Language and Communication, edited by Roy Harris. In this report, I can only critically summarize the high points of what seem to me to be leading and provocative (the two are not synonymous) trends in the sociology of language that subsumes standardization, split, and merger. 2. French, with its rigidity of norms, its well-nigh specious precision at the hands of an academy which adjudicates innovation, and the preciosite in its social stratification of usage, is probably the parade example of an exceedingly well-groomed standardized language. Afrikaans, on the other hand, insecure in its claim to any status, has gone so far as to label its birth a miracle and to mark it with a public monument. Its more conservative supporters violently disclaim any possibility of its creole origins, but see Markey (1981). For yet other languages, such as Osage,

Language split and dialect merger

3. 4.

5. 6. 7.

8.

9. 10.

465

the question of standardization is purely hypothetical, at best a very moot question, and perhaps even smacks of insult. On this matter, an interesting exercise in the definition of dialect as opposed to language, see Pierre Bee (1969). See, for example (and I can only offer a sampling here), Labov (1963, 1966, 1972a, 1974); Labov - Yaeger - Steiner (1972); Weinreich - Labov - Herzog (1968); Trudgill (1971, 1979); Chambers - Trudgill (1980); Hudson (1980); Milroy (1980). The best introduction to Labovian sociolinguistics, a recapitulation of much of his earlier work, is Labov (1972b). See Markey — Tyma (1978); Markey (1981). For further criticism see Bailey (1977, 1980). See Markey (1981); Bailey (1981, 1982). The proposition that there is a primarily abductive basis for change innovation is one of Henning Andersen's major claims to theoretical insight. The abductive/ deductive complementarity is, of course, the brainchild of Charles Sanders Pierce, the founder of pragmatism (ca. 1878). Logically, definition of abduction as a nonevident minor premise permits only probable conclusions, but, essentially, it is the relational "grammatology" of the two terms that concerns Andersen (see Andersen 1973). Unlike Pierce, who applied this set to semantics, Andersen and his followers apply it to linguistic change. In his quest for philosophical analogues to linguistic phenomena, Andersen has fallen into the subtle trap laid by incommensurability: he has employed a stative relation to explain a process. Note, for example, how the validity of this approach has been severely questioned in C.-J. Bailey (1980), where he contends that Andersen "... makes the fatal assumption of static frameworks that different [abductive] outputs imply different [deductive] underlying structures, assuming that speakers who merge bad, bared, and beard must utilize adaptive rules to distinguish them in dealing with them in the speech of others and in producing them in their own formal styles. This view of the matter commands little psychological credibility" (1980: 164). This and similar objections also serve to discredit Raimo Anttila's rather counter-intuitive views of change as Gej/a/i-shuffling (see Anttila 1975). For further controversion of the abductive/deductive view and proportional analogy, see Willi Mayerthaler (1980a). On the implications of naturalness and markedness theory for linguistic change, see also Willi Mayerthaler (1977, 1980b [revised and translated 1983], and 1982). The evidence for this from the cahiers is not strikingly clear, but this contention may be inferred from a close-reading analysis of Saussure's writings prior to his teaching in the Cours. At any rate, I must point out that only religions, not sciences, accept the second-hand reports of disciples as gospel. The literature on creolistics has grown enormously in just the past two or three years, but for a succinct introduction to the major issues and trends, see Hill (1979). Bickerton's thesis is controversial, of course, not only because it counters nativistic rationalism with empirical fact, but also because incredulous opponents see a failure to make allowances for the sociological context of creolization (plantation economies), for the fact that some non-additive processes in Creoles (e. g., the surpasscomparative in English-based Creoles, the reflexive marked by terms for body-parts in French-based Creoles) appear as traces of substratal strategies. Be all this as it may, Bickerton's thesis is revolutionary and has attracted more mass-media attention

466

11.

12. 13.

14.

15. 16.

Thomas L. Markey than any work in linguistics in decades. Note the reviews by Begley (1982); Bruner - Feldman (1982); Morrison (1982); and Anshen (1982). Note, for example, that Papiamentu, the indigenous Creole of the ABC islands (Aruba, Curasao, Bonaire) of the Netherlands Antilles, has progressed far down the path of decreolization. Unlike so many Creoles, it has even been vindicated qua language. Although still not officially sanctioned as a language for the classroom, there is an extensive religious and secular literature, particularly since the 1940s, and prescriptive grammars by natives. See Reinecke et al. (1975:147 — 209). Contrast the Papiamentu situation with that of Tahitian, not a Creole, which never appears in written form except in missionary translations of the Bible, a result of the restrictive colonialism of metropolitan France. For an intriguing discussion of the anthropological dimensions of compartmentalization, see Moore (1981). In the exposition that follows, I have been somewhat irresponsible, for I have given a "visual" interpretation of Seller's work and thus pushed his views into a dimension he may not have intended. Nevertheless, with apologies to Seiler, I think the "visualization" of discontinuous processes in time is an extremely important means for coming to grips with change and change processes. Thorn's theory, like Bickerton's, is highly controversial and has been hailed as fraudulent or as a brilliant breakthrough with little middle ground. However, the truly important thing about Thom is that he has asked a very meaningful question and set about answering it in a highly informative way. The classic presentation is Thom (1972). For guides, see Saunders (1980); Woodcock - Davis (1978). For a survey of current trends in bilingualism, see Paradis (1978). The classic description of these matters is that by Uriel Weinreich (1953).

References Andersen, Henning 1973 "Abductive and deductive change", Language 49: 567 — 593. Anshen, Frank 1982 Review of Bickerton 1981. "The Week in Review", The New York Times 22 August 1982: 22EY. Anttila, Raimo 1975 "Generalization, abduction and language", in: Koerner (ed.), 263 — 296. Bailey, Charles J. 1977 "Variation and linguistic analysis", Papiere zur Linguistik 12: 5 — 56. 1980 "Old and new views on language history and language relationships", in: Lüdtke (ed.), 139-181. 1981 "Developmental linguistics", Folia Linguistica 15: 29 — 37. 1982 On the Yin and Yang of nature of language. Ann Arbor: Karoma. Bee, Pierre 1967 La langue occitane (Que sais-je?, No. 1059). Paris: Presses Universitaires de France.

Language split and dialect merger

467

Begley, Sharon 1982 "Fossils of language", Review of Bickerton 1981. Newsweek 15 March 1982: 80. Bickerton, Derek 1981 Roots of language. Ann Arbor: Karoma. Bjerke, Andre 1962 Hva er godt riksmäl? Spersmäl og svar. Oslo: Riksmälsforbundet. Blackall, Eric A. 1959 The emergence of German as a literary language. Cambridge: Cambridge University Press. Bruner, Jerome — Carol Feldman 1982 Review of Bickerton 1981. The New York Review of Books 24 June 1982: 34-36. Castille, G. P. - Gilbert Kushner (eds.) 1981 Persistent peoples: Cultural enclaves in perspective. Tucson: University of Arizona Press. Chambers, J. K. — Peter J. Trudgill 1980 Dialectology (Cambridge Textbooks in Linguistics). Cambridge: Cambridge University Press. Givon, Talmy 1979 On understanding Grammar. New York: Academic Press. 1983 Syntax: A functional typological approach. Ann Arbor: Karoma. Haugen, Einar 1953 The Norwegian language in America: A study in bilingual behavior. Philadelphia: University of Pennsylvania Press (2nd edition Bloomington: Indiana University Press, 1969). 1966 Language conflict and language planning: The case of Modern Norwegian. Cambridge, Mass.: Harvard University Press. Heilmann, Luigi (ed.) 1974 Proceedings of the 11th International Congress of Linguists. 28 August/ September 1972, Bologna —Florence. Bologna: il Mulino. Hill, Kenneth C. (ed.) 1979 The genesis of language. Ann Arbor: Karoma. Hudson, R. A. 1980 Sociolinguistics. Cambridge: Cambridge University Press. Koerner, E. F. Konrad (ed.) 1975 The transformational generative paradigm and linguistic theory. Amsterdam: John Benjamins. Labov, William 1963 "The social motivation of a sound change", Word 19: 273 — 309. 1966 The social stratification of English in New York City. Washington, D.C.: Center for Applied Linguistics. 1972a Language in the inner city: Studies in the Black English vernacular. Philadelphia: University of Pennsylvania Press. 1972b Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press. 1974 "On the use of the present to explain the past", in: Heilmann (ed.), 825-851.

468

Thomas L. Markey

Labov, William — Malcah Yaeger — Richard Steiner 1972 A quantitative study of sound change in progress (Report on National Science Foundation Contract GS-3287). Philadelphia: United States Regional Survey. Lehmann, Winfred P. — Yakov Malkiel (eds.) 1968 Directions for historical linguistics. Austin: University of Texas Press. Lüdtke, Helmut (ed.) 1980 Kommunikationstheoretische Grundlagen des Sprachwandels. Berlin/New York: de Gruyter. Markey, Thomas L. 1981 Frisian (Trends in Linguistics: State of the Arts Reports 13). The Hague: Mouton. 1981 "Diffusion, fusion and creolization: A field guide to developmental linguistics", Papiere zur Linguistik 24: 3 — 37. 1982 "Afrikaans: Creole or non-creole?", Zeitschrift für Dialektologie und Linguistik 49: 169-207. Markey, Thomas L. — Stephen Tyma (eds.) 1973 "Introduction", in: The analytic and the descriptive, Vol. II of Studies in European linguistic theory, xvi —xix. (Giessener Beiträge zur Sprachwissenschaft Bd. 13) Grossen-Linden: Hoffmann-Verlag. Mayerthaler, Willi 1977 Studien zur theoretischen und zur französischen Morphologie (Linguistische Arbeiten 40). Tübingen: Niemeyer. 1980a "Aspekte der Analogietheorie", in: Lüdtke (ed.), 8 0 - 1 3 0 . 1980b Morphologische Natürlichkeit. Wiesbaden: Athenaion. (Revised and translated, Ann Arbor: Karoma, 1983.) 1982 "Markedness and historical linguistics", paper presented at the 2nd Annual Symposium on History, Linguistics, and Philology, 15 — 17 April 1982, University of Michigan. Medawar, Peter Brian 1967 The art of the soluble. London: Methuen. Milroy, Lesley 1980 Language and social networks (Language in Society No. 2). Oxford: Blackwell. Moore, Janet R. 1981 "Persistence with change: A property of sociocultural dynamics", in: Castille — Kushner (eds.), 23 — 36. Morrison, Philip 1982 Review of Bickerton 1981. "Books", Scientific American 247, No. 2: 26-33. Mühlhäusler, Peter 1980 "Warum sind Pidginsprachen keine gemischten Sprachen?", in: Ureland (ed.), 139-159. Paradis, Michel 1978 Aspects of bilingualism. Columbia, S. C.: Hornbeam Press. Pfaff, Carol 1973 "A sociolinguistic study of Black children in Los Angeles", dissertation University of California, Los Angeles.

Language split and dialect merger

469

Reinecke, John E. et al. 1975 A bibliography of pidgin and Creole languages (Oceanic Linguistcs Special Publication 14). Honolulu: The University Press of Hawaii. Saunders, P. T. 1980 An introduction to catastrophe theory. Cambridge: Cambridge University Press. Seiler, Hansjakob 1981 POSSESSION as an operational dimension of language (Arbeiten des Kölner Universalienprojekts 42). Köln: Institut für Sprachwissenschaft. Thom, Rene 1972 Stabilite structurelle et morphogenese. Reading, Mass.: Benjamin. Trudgill, Peter J. 1971 "The social differentiation of English in Norwich", Dissertation Edinburgh University. Trudgill, Peter J. (ed.) 1979 Sociolinguistic patterns in British English. London: Edward Arnold. Ureland, P. Sture (ed.) 1980 Sprachvariation und Sprachwandel. Tübingen: Niemeyer. Weinreich, Uriel 1953 Languages in contact: Findings and problems. New York: Linguistic Circle of New York. Weinreich, Uriel — William Labov — Marvin Herzog 1968 "Empirical foundations for a theory of language change", in: Lehmann - Malkiel (eds.), 9 7 - 1 9 5 . Woodcock, Alexander — Monte Davis 1978 Catastrophe theory. New York: E. P. Dutton.

Contact linguistics: Research on linguistic areas, strata, and interference in Europe P. Sture

Ureland

Introduction Contacts between peoples and nations have always been crucial for human progress. Any research into the effects of linguistic and cultural interaction in the present or the past opens up a rich field of study which is developing more and more into an area of its own: "contact linguistics". Since language contacts are also cultural contacts, the traditional borderline between linguistics (in the narrower sense) and other disciplines (sociology, cultural anthropology, history, psychology, etc.) is felt as an artificial barrier which hampers our efforts a) to describe the geographical spread and historical dimension of language, b) to explain its social use, and c) to describe and explain language change. An adequate treatment of the topics language contact and language change must take account of the results of research in the following areas: a) the language blending and substratum hypotheses; b) the linguistic area hypotheses of the Balkan linguistic type of areal studies ("Sprachbund"); and, in particular, those of the Prague phonologists ("phonologische Sprachbünde"); c) the interference hypotheses of the American structuralists; d) the morphological and syntactic typology of American ethnologists and ethnolinguists; e) the hypotheses on pidginization and creolization of the creolists; f) research on language contact in Europe since Weinreich 1953.

472

P. Sture

Ureland

Unfortunately, all these areas of research cannot be treated in a short article. I will therefore restrict myself to discussing points a) and b) from a historiographic viewpoint and to sketching an outline of f). (For a detailed discussion of c), d), and e), see Ureland 1978c.)

1. A short historiography of contact linguistics Contact linguistics or convergence linguistics has not always been recognized as a subdiscipline worthy of independent study in the history of linguistics. Other more categorial attempts at theory formation of a systematic nature (Neogrammarian, structuralist, generative or communicational-theoretical) have dominated and continue to dominate the field. It is true that lasting contributions have been made in the study of language change during the past century, thanks to such categorial theory constructions as those suggested by the historical comparativists or Neogrammarians. But this was done at the cost of gaining an insight into the more complex problems of linguistic variation and linguistic diversity caused by language contact and language change (consider Schuchardt's position in the nineteenthcentury linguistic establishment). The preoccupation with abstractions at the expense of authentic research (field work on bilingualism or bilectalism) resulted in one-dimensional laws ("Lautgesetze") or rules (generative rule charts) which were not always in agreement with the range of language variation, the social parameters, or the geographical distribution of language. This preoccupation with theory at a safe distance from social, geographical, and historical facts is a characteristic not only of our predecessors but also of the majority of contemporary historical linguists. Few linguists today present really new data or insights into the mechanism of linguistic change, but rather quote or excerpt from the old handbooks. True historical research which involves the time-consuming reading of old manuscripts and traditional philological work is seldom carried out. The evolutional-Darwinian view of language change is still amply represented in the standard works of historical linguistics, although the new perspective on language change resulting from research on contact between languages should merit equal space and importance in all descriptions of language change. How is it that this state of affairs still prevails today? In order to answer this question we have to examine the history of contact linguistics.

Contact linguistics

473

2. Language contact hypotheses 2.1. Language blending and stratalinguistic research It is an enigma in the history of linguistics how the contact aspect of language change was so slow in being accepted. Almost at the same time as the proclamation of the infallibility of the sound laws by Brugmann and Osthoff in 1878 and the subsequent debate on the onedimensional nature of these laws (cf. in particular Schuchardt's 1885 sharp criticism), Miklosich and Schuchardt were carrying out extensive investigations on the blending or convergence of the continental European languages. 1 Schuchardt's interest in pidgins and Creoles should also be mentioned as an example of his openness to new fields of ethnic studies and his capacity to break the bonds of Europeancentered research (cf. Schuchardt 1882 — 83 Kreolische Studien /— VI). Schuchardt is usually hailed as the father of Creole studies, although he was by no means the first European in the past century to devote himself to pidgins and Creoles (cf., e. g., the long list of other European "creolists" in Le Page 1980: 117-118). However strong the dogmatic genealogical-biological view of language change may have been in the Neogrammarian camp, other approaches to language change asserted themselves (cf. the chapter on "Language blending" in the second edition of Paul 1886: 390 — 403 [not in the first edition of 1880]; Windisch 1897; and Schuchardt's 1917 review of Saussure 1916), in particular among the less dogmatic of the dialectologists, who were the most representative critics of categorial sound-law-thinking beside Schuchardt. In his view the rise and evolution of the Romance languages from Vulgar Latin were due to different ethnic composition (substrate: Celts, Iberians, Thracians, etc.) in the provinces conquered by the Romans. The underlying strata were thought to have caused changes in the Latin spoken by the Romans in the various parts of the Roman Empire (cf. Ascoli 1881 —1882). This type of thinking in terms of geological layers spread to other fields as well: those of Celtic, Germanic, Slavic, and IndoEuropean. New ingenious terms for different types of historical contacts were coined which described the direction of language contact and how these contacts had caused language change (e.g., von Wartburg's 1932 "superstratum" and Valkhoffs 1932 "adstratum"). These approaches to language contact research which aimed to describe and explain language change were based on the stratalinguis-

474

P. Sture

Ureland

tic-genealogical model and they reached a climax shortly before the outbreak of the Second World War in 1939. On the eve of the outbreak of the war, Brondal, Cohen, Gamillscheg, Pisani, Valkhoff, von Wartburg, Whatmough, and others prepared papers for the 5th International Congress of Linguists which was to be held in Brussels between August 28 and September 2, 1939. Because of the invasion of Poland on September 1, 1939, these papers were never presented or published. We are fortunate to have them at our disposal in a volume edited by Kontzi (1982d: 72 — 101), where they are presented under the same title as that of the congress section planned for Brussels: "Substrat, Superstrat, und Adstrat". These papers all reflect the great fascination which the stratalinguistic view exerted upon this pre-war generation of linguists. What is striking in reading these papers is the expert knowledge of past language contact, the insights into the historical background of the languages involved, the presentation of the social and ethnic background of vanished speech communities, together with a rich stream of pertinent linguistic facts which are all used to support the hypotheses presented. The scholars mentioned here had a thorough training in their own fields before they started generalizing on the basis of their observations of language change — a characteristic which is not found today among those linguists who propose sweeping hypotheses and among those who have a flair for evolving theories but who do not want to do the empirical work necessary to construct convincing theories on language change. As a result of the catastrophic effects of the German Nazis' racism and of the Second World War itself, there was little interest in this kind of ethnic-historical approach to language change after the war in Germany, Scandinavia, and other parts of Europe. Thus scholars turned to other goals and methods in describing languages and language change. In Romance studies, however, the ethnic approach lingered on, unaffected by Nazi-German racial distortions. This fact is well illustrated by the articles collected in Kontzi (1982d). The articles contained in this volume show a development from historically-oriented stratalinguistic approaches to structurally-oriented ones (cf., for instance, the critical articles against one-sided substratomania from a Romanistic and general linguistic point of view by Nielsen (1952), Malmberg (1961), Francescato (1969), Tovar (1951), Baldinger (1963), or Delattre (1969-70).

Contact linguistics

475

2.2. Linguistic area research The hypothesis of linguistic change as a result of language contact gained considerable support from the discovery of typological similarities between the very different languages of the Balkans (cf. Miklosich 1862, 1869, 1870-1871, 1872, 1884-1890; and in particular Sandfeld [1930], who suggested that a "Balkan linguistic area" existed). The importance of including the areal linguistic dimension in describing the languages of Europe was also stressed by a number of researchers, in particular phonologists during the 1930s. The hypothesis of the "linguistic area" in the sense of Trubetzkoy (1931a) or Jakobson (1931) grew out of phonological comparisons between the Central and East European languages. The phonological correlations between related and unrelated languages worked out by these two scholars were to be used as the basis of the phonological universale proposed later on in the 1950s. Sandfeld also included morphological and syntactic characteristics in order to support his hypothesis of a "Balkan linguistic area". Thanks to the research carried out by Trubetzkoy and Jakobson, synchronic linguistics was to gain a status equal to that of historical linguistics and a structural (phonological) typology was created, which allowed typological comparisons of a purely synchronic nature, without consideration of the genealogical or socio-historical reasons underlying the similarities which had been found between languages spoken on the European continent. The leading scholars in Germanic, Romance, and English studies did not consider the problem of establishing such a typology of linguistic areas to be central to their research. This problem was given greater emphasis in linguistic disciplines which were involved more directly with language contacts between IndoEuropean and non-Indo-European languages on the fringes of Europe: that is, in Slavic studies (cf. Trubetzkoy 1931b, 1939a, 1939b; Jakobson 1931, 1938; Skalicka 1935; and Isacenko 1934), in Finno-Ugric studies (cf., e.g., Lewy 1942: 83 and 1959-1960), and in Celtic studies (cf. Pokorny 1927-1930; Lewy 1959-1960; Wagner 1959). In recent years this interest in linguistic areas and language contacts as potential factors underlying language change has gone through a renaissance, also in studies dealing with the periphery of Europe: Slavic-FinnoUgric (cf. Veenker 1967, 1982, 1985; Batori 1979, 1982; Panzer 1982, 1984), North-Germanic-Baltic-Slavic (cf. Ureland 1977, 1978b, 1979a; Lehiste 1978; Panzer 1982, 1984; Romance-Celtic (Rohr 1963, 1969, 1982); Romance-Arabic (Steiger 1948-49; Rizzitano 1965; Giese 1966;

476

P. Sture

Ureland

Rohlfs 1972; Pellegrini 1972; Kontzi 1980, 1982a, and 1982b). For all this research the impact of the structural comparative phonology of the Prague school was necessary for an overall view of the languages of Europe, and equally necessary was the contact aspect between IndoEuropean and non-Indo-European languages found in the stratalinguistic approach. The two factors contributed towards an areal typology of European and Asian languages. Jakobson, Skalica, and Isacenko advanced the hypothesis of a "Euro-Asian linguistic area" in East Europe. They were stimulated by Trubetzkoy's structural phonology, which was based on phonological and prosodic correspondences between East-Slavic, Volga-Finnish, and Tatarian languages in Russia. Jakobson also spoke of a "Baltic linguistic area", which not only encompassed Baltic and Baltic Finnish languages (Latvian, Lithuanian, Livian, Estonian, and Votian), but also North Germanic (Swedish, Norwegian, and Danish) and West Slavic languages (Kashubian) (cf. a more recent critical approach to this idea in Lehiste 1978). Although the latter linguistic area has not been so generally accepted, it has been applied to explain certain correspondences found between these languages in the verb phrase (enclitical modalities, prosody, etc.) (cf. Pisani 1974; Ureland 1977, 1980; Lehiste 1978, etc.). The most famous linguistic area is the one mentioned above which was proposed by Sandfeld and which he called the "Balkan linguistic area" after studying Miklosich 1862, 1869, 1872, and 1884-1890. Thus an areal typology has been established not only for East and Sout-East Europe, but also for West and South-West Europe. Pokorny (1927 — 30) laid the foundation of such a linguistic area by using the substratum approach; Lewy (1959 — 60) calls this area the "Atlantic linguistic area" which includes Basque, Spanish, French, English, and Irish. Wagner (1959) goes a step further and also includes Berber and other North-African languages in an "Euro-African linguistic area". According to Wagner, this overall territory includes a smaller "British linguistic area", which covers British Celtic (Welsh, Cornish, and Breton) and Goidelic Celtic (Irish and Scots Gaelic) but also English. All this research on linguistic areas in Europe is a means of reconstructing mutual contacts which have led to common linguistic changes or adaptations in the phonology, morphology, and syntactic systems of the languages which are still visible today. The linguistic area approach offers us linguistic perspective from which we gain descriptive adequacy and explanatory power. Even though it may be true that the

Contact

linguistics

477

term "linguistic area" has become empty as a result of excessive use (see, e.g., Decsy [1973], who speaks of a "Viking", a "Peipus" or a "Rokytno linguistic area"), the terms "Balkan", "Atlantic", "British", and "Baltic linguistic areas" have definitely been accepted because of the great correspondences between the languages in the east, west and north of Europe. Nevertheless, it seems necessary to define the concept with an explicit number of structural features, and when these are present one is entitled to claim that a given number of languages constitute a linguistic area. However such an explicit catalogue of criteria does not exist.

3. The rise of ethnolinguistics in Europe Although interest in the ethnic approach had been reduced by the events of the Second World War in Europe, it was very much alive in America, where the conditions were much more favorable than in Europe. The long tradition of ethnolinguistics since the days of Boas, Sapir, and Whorf supported the continuity of ethnical studies in America, which were to bear fruit in the later works of Greenberg, Weinreich, and Labov as well as in the long series of Creole and pidgin studies. The lack of genuine historical research on an ethnic basis in Europe explains the great impact of synchronic structuralist and generative linguistics in all European countries after the war. The vacuum left by the decline of the German type of historical linguistics (historische Sprachwissenschaft) and the rejection of contact studies on an ethnic-historical basis made younger linguists in Europe receptive to new theoretical approaches during the 1950s and 1960s. In spite of the fact that little ethnic and historical research was carried out during these decades, strong forces are now being mobilized in opposition to the one-dimensional type of systemic linguistics which disregards linguistic variation and historical change. In the mid-1970s, the European linguists had recovered from the anti-ethnic outlook of the 1950s and found inspiration in the new research on bilingualism and sociolinguistics as presented by American linguists working on emigrant languages in the USA (e.g., Haugen 1953), language contact areas in Europe (cf. Weinreich 1953; Sommerfelt 1962; Jud 1920; Gartner 1910; Planta 1931, etc.). The publications of Labov, Fishman, Ferguson, and Gumperz on socially and ethnically diversified speech communities

478

P. Sture

Ureland

were of great importance for the awakening of the dormant interest in the ethnic approach in Europe. The renewal of interest in an ethnically-based type of linguistics made it clear to most European linguists that the bilingual individual constituted an important focus of research for describing linguistic change. A number of phonological, morphological, syntactic, and lexical innovations were shown to result from interference and transference processes which took place in the speech of bilinguals. Linguists did not focus their attention on the entire speech community and abstract systems but rather on the bilingual or multilingual speakers. In the heyday of Neogrammarianism and structural and generative grammar, the fact was forgotten or ignored that man is a multilingual creature, who can learn and master several languages with varying degrees of perfection. Bilingualism and diglossia are potential capacities which every human being possesses at birth. He can develop these abilities to a near perfect degree, if the ethnic and social circumstances force him to do so. One can ask oneself whether there is any individual who has recourse only to one standardized code in the sense of the structuralists or to only one system of rules in the sense of the generative grammarians. Every human being lives with a multitude of linguistic varieties around him, which he learns to understand and often speak within a comparatively short period of time. If he were unable to learn or understand these varieties he would also be unable to meet the complex requirements of the speech acts. Furthermore, in European areas of multilingualism or in areas with a strongly developed type of diglossia of dialects and a standard language (e.g., Schleswig, Grisons in Switzerland, Alsace, or East Belgium), the inadequacy of systematic linguistics was still more obvious. Here the intricate network of domains, levels of style, and registers had to be investigated before general claims as to linguistic development and change could be formulated. The multidimensionality of linguistic speech acts in these areas contrasts sharply with the onedimensional rule systems as elaborated within the framework of generative grammar. In order to describe linguistic change, the multidimensional language gift of man must be taken into consideration. Convergence linguistics or, as we prefer to call it in Europe, "contact linguistics", is one road to a better understanding of language change and consequently of the nature of language.

Contact

linguistics

479

4. European studies in multilingualism and contact linguistics Because of the over-emphasis on linguistic universals and the neglect of the enormous linguistic differences between the languages of the world, the essential factors causing language change could not be established. However, this statement is only partly true in view of certain activities in Europe which date back to about 1970. A true interest in and deep involvement with ethnic perspectives of the European languages can be observed in a long series of publications (including Festschriften and symposium proceedings) which have not yet attracted widespread attention. First we should mention Kloss (1969) Grundfragen der Ethnopolitik im 20. Jahrhundert or the new edition in 1978 of his Entwicklung neuerer germanischer Kultursprachen seit 1800, which originally came out in 1952 and which was hardly accepted then because of its ethnically-based approach to the Germanic languages (cf. Kloss 1981: 3). Secondly, two handbooks on linguistic minorities in Europe by Straka (1970) and Stephens (1978) are signs of a new awakening of interest in ethnicity in Europe. Thirdly, four important Festschriften have been published which also indicate a growing interest in ethnic problems: the Festschrift for Betz, Sprachliche Interferenz (1981) edited by Kolb and Lauffer; the two Festschriften for Wandruszka, Interlinguistica (1971), edited by Bausch and Gauger, and Europäische Mehrsprachigkeit (1981), edited by Pöckl; and in addition a very interesting book published in connection with the fiftieth anniversary of Hugo Schuchardt's death in 1977 by Lichem —Simon (1980). In these four Festschriften, which all have an almost programmatic character, the dogma of homogeneity in systemic linguistics is attacked. As the titles indicated, heterogeneity and polysystematism are the focus of interest: interference, interlinguistics (i. e., contact linguistics), polylingualism, and last but not least the name of Hugo Schuchardt, who was rediscovered in the 1970s by linguists interested in bilingual and creole studies (see, e.g., the translations of some of his works edited by Markey [1979] and edited by Gilbert [1980]). In describing linguistic change, it is important to refer to Schuchardt because for him language was not a homogenous system, but rather a multitude of systems and subsystems, in which diachrony and synchrony played an equal role and in which the process of blending (Mischung) was a major cause of change in the history of a

480

P. Sture

Ureland

language. "Blending permeates every linguistic development; it occurs between close dialects, between related and even between quite unrelated languages" (transl. of a quotation in Denison 1980: 8).

5. Centers of research on contact linguistics in Europe Finally, a number of symposia on language contact in Europe have been organized which indicate the trend back towards ethnically-based linguistics and a new interest in historical and social change in the languages in Europe. Two centers of contact linguistic research have played a leading role here: the Research Center on Multilingualism in Brussels and the Linguistic Circle of Mannheim in West Germany. In a joint effort to further contact linguistic studies in Europe, eight symposia have been held so far, which have resulted in the publication of a great number of articles on the contacts between European languages. Many of these publications are already on the market (cf. Ureland 1978c; 1979b, 1980b, 1981, 1982b, and 1985; and Neide 1980, 1983a, b, c, d, and 1985; Neide - Ureland - Clarkson 1986; Ureland - Clarkson 1984). The main concern has been to establish a forum for discussion, where all the different data from contact-linguistic research can be collected, not only for one European language in contact with another, but for a great number of European languages in contact with each other. The articles published in the volumes just mentioned are characterized by an almost positivistic delight in presenting the magnitude of variants and linguistic changes caused by language contact. The linguistic and ethnic-cultural variables underlying the interference and transference processes are extensively dealt with, even though it has not yet been possible to establish a typology of language contacts. More attention has been given to linguistic change with reference to the bilingual individual. Whereas the Brussels symposia (cf. Neide 1980, 1983a, b, c, d) presented an overall socio- and ethnolinguistic spectrum of all sorts of contacts between languages in contact and/or conflict, the Mannheim publications have focused more on specific areas and formal linguistic problems of contact, whereby the presentation of linguistic data plays an indispensable role: The North Sea area (1978), Dialects and standard languages in multilingual areas of Europe (1979), The problem of variation and language change (1980), The situation of the linguistic and cultural minorities in Europe (1981),

Contact linguistics

481

The achievement of stratalinguistic research and Creole studies (1982), The genesis of languages and peoples (1985), Language contact in the Hanseatic League (1987) (in the framework of The Linguistic Circle of Mannheim, LAMA). Two symposia on "Language Contact and Conflict" were organized by the Research Centre on Multilingualism in Brussels (cf. Neide 1980 and 1983a, b, c). At the 13th International Congress of Linguists held in Tokyo in 1982, two working groups were conjointly organized by the two research centers (cf. Ureland 1983a: 1277 — 1286), the papers of which were also conjointly published (cf. Neide — Ureland — Clarkson 1986). Between them the two centers of contact-linguistic research have published hundreds of articles on various aspects and areas of languages in contact. More articles will appear from this rich field of research when the proceedings of the 1988 symposia are published ("Language Contact in the British Isles" organized by the Mannheim Linguistic Circle, and "The Third Contact and Conflict Symposium", organized by the Brussels Research Centre).

6. Research on migrant workers, language death, linguistic minorities, and linguistic boundaries in Europe 6.1. Migrant workers The ethnically-based research on languages and language change in Europe is not limited to the two centers of research on contact linguistics. Besides the two series of publications emanating from the conference activities in Mannheim and Brussels, more volumes can be mentioned here which are the result of a similar interest: Molony — Zobl - Stölting 1977; Meisel 1977; Meid - Heller 1981; Werner 1980; Caudmont 1982; Messner — Wandruszka 1984; Oksaar 1984; Denison — Sornig — Gadler — Grassegger 1986. The activities during the past five years have been on an ever-increasing scale, so that this type of research will probably last long into the 1990s. In addition there are a number of research undertakings which are examining the change in the subvarieties of the standard languages, and which will be of importance directly or indirectly for the general theory of language change. A series of articles (e.g., Clyne 1968, 1980; Ferguson 1977, etc.), books (e.g., Clyne 1975; Klein - Dittmar 1979, etc.), and research projects on the language of the migrant workers (e.g., the

482

P. Sture

Ureland

Heidelberger Forschungsprojekt Pidgin-Deutsch 1975; Meisel 1975; and Dittmar — Haberland 1978), and dissertations on the same topic (e.g., Orlovic — Schwarzwald 1978) are an expression of a deep involvement with ethnic, social, and ecological aspects of the languages in Europe, whereby the causes and effects of language change play a central role. 6.2. Language death In addition to these investigations on the migrant workers and their languages, some other aspects of language contact should be mentioned which have been the topics of collections of articles and conferences in the past few years. First, the phenomenon of language death has lately been a topic which has inspired a number of articles collected in Dressier — Wodak-Leodolter 1977b. Language death in a number of European countries is dealt with in Denison 1977 (northern Italy), Dressier - Wodak-Leodolter 1977a (Brittany), and Dorian 1977 (East Sutherland), etc. Other contributions which deal with the linguistic changes in areas of language death are Dorian 1973, 1978, 1981 (East Sutherland), Walker 1978, Larsen 1984 (North Frisland), Greene 1981 (Ireland, Scotland, Wales and the Faroes), Ambrose — Williams 1983 (Wales), and Fennell 1981, 1985 (Ireland). 6.3. Linguistic minorities The following language contact conferences should be mentioned: The working group on "Language Contact in Scandinavia" held at the Fourth International Conference of Nordic and Modern Linguistics in Oslo (June 1980) (cf. Ureland - Clarkson 1984); the First International Conference on Minority Language held in Glasgow (September 1980) (cf. Haugen — McClure — Thomson 1981); the conference held in Meran, Italy (October 1980) on "Linguistic Problems and European Unity"; the Linguistic Minorities Project at the University of London; the investigation of the Slovenian language in Kärnten, Austria (cf. Denison 1981); the Second Symposium on Languages and Cooperation in Urbino, Italy (September 1981); the symposium on "International Languages and Vernaculars" and the sections 13 and 14, "Bilingualism and Immigrants' Language Problems" and "Language Planning and Policy", at the 6th International Congress of Applied Linguistics held in Lund, August 1981 (cf. Sigurd — Svartvik

Contact linguistics

483

1981). A conference on "The Romance Languages in the Eastern Alps" was held in Salzburg (October 1982) (cf. Messner — Wandruszka 1984); a comparison between European linguistic minorities was the topic of a conference held in Bayreuth in 1983 (cf. Hinderling 1986). In Neuchätel a symposium on bilingualism was held in 1984 (cf. Liidi 1986) and in Graz a conference on the important concept of maternal language (Muttersprache) (cf. Denison et al. 1986). The founding of new linguistic journals during the past ten years in Europe such as The Journal of Multilingual and Multicultural Development, Clevedon (Great Britain), Multilingua, Berlin, and Sociolinguistica, Tübingen, reflects the increasing need for publication possibilities for this new and expanding field of linguistics. In 1983 and 1986 two more international conferences on minorities took place in Turku, Finland, and Galway, Ireland (cf. Moide — Sharp 1984 and Mac Eoin - Ahlqvist - 0 hAodha 1987a and 1987b). 6.4. Language contact in historical linguistics Also in the field of historical linguistics a number of conferences and symposia have been held which focus on the contacts between languages in the past: "The Achievement of Stratalinguistic Research and Creole Studies" (cf. Ureland 1982b), "The Genesis of Languages and Peoples in Europe" (cf. Ureland 1985), "The First Symposium on (Middle) Low German in Scandinavia" (cf. Schöndorf — Westergaard 1987), "Language Contact in the Hanseatic League" (cf. Ureland 1987b). A second symposion on (Middle) Low German in Scandinavia was held in Copenhagen in May 1987 (cf. Hyldgaard-Jensen, in press). A symposium on "Language Contact in the British Isles" is planned for 1988 on the Isle of Man, to be organized by the Linguistic Circle of Mannheim (LAMA) in its series of symposia "Language Contact in Europe" (cf. Ureland — Broderick 1990). 6.5. Linguistic boundaries and linguistic islands In this context the entire range of research on the languages on the Germanic-Romance, Germanic-Slavic, and West-Germanic-North Germanic language boundaries should be mentioned: Belgium: Vanacker 1967, 1973, 1977; Verdoodt 1973; Neide 1978, 1979; Taeldeman 1978, 1982; van de Craen 1980; Fischer 1980; Persoons - Versele 1980; Persoons 1981; Trim 1981; Quix 1981; Beardsmore 1983; Deprez

484

P. Sture

Ureland

— Persoons 1983; Deprez — Persoons — Struelens — Wynants 1983; Luxembourg: Hoffmann 1979, Knowles 1980; Lorraine: Cadiot 1980; Alsace: Becker-Dombrowski 1981; Hartweg 1980, 1983, 1986, Ladin 1982, 1983a, 1983b; French Switzerland: Kolde 1980, 1982; The Grisons {Switzerland): Billigmeier 1979; Diekmann 1979, 1981, 1982a, 1982b, 1986; Kristol 1980; Decurtins 1981, 1985; Darms 1985; Cathomas 1977, 1982; Camartin 1982; Goebl 1982a, 1982b, 1986; Ureland 1982a, 1983b; Viletta 1983; Northern Italy: Egger 1977, 1979, 1981, 1983; Riedmann 1972, 1973, 1979; Pernstich 1981; Langer 1983; Craffonara 1981; Born 1983; Höglinger 1980; Kramer 1981; Francescato 1981a, 1981b; Denison 1977,1979,1980,1981b; Hornung 1980,1983; Ureland 1982a; Meid - Heller 1979; Meid 1985a, 1985b; 2 Roumania: Rein 1979, 1980; Steinke 1979. Language contacts on the German-Danish linguistic boundary in the north has been the topic in Hyldgaard-Jensen 1980; S0ndergaard 1980b, 1981, 1984; North Frisland: Walker 1978, 1980; Wilts 1978. Case studies of language contacts in the British Isles have also been carried out by Ureland (1978a); Roider 1980, 1981; Broderick (1985) (Insular Celtic); Ireland: Mac Eoin (1982); Scotland: Strauss (1982). Interesting contact situations in the present and the past in France are described in Lemarchand-Unger 1981 (Brittany); Provence: Diekmann 1981; North-West France: Rohr 1982; Spain: Diaz Lopez 1980; Tovar 1978; Winkelmann 1983, 1985; Sardinia: Rindler Schjerve 1979, 1983; South Italy: Kattenbusch 1981; Rohr 1981; Italy: Radtke 1982, 1985; Holtus 1985; Schmitt-Brandt 1985 (Latin). (For an overall historical and stratalinguistic treatment of the rise of the Romance languages see Kontzi 1978, 1982d; Hubschmid 1982; and Schmitt 1982.) The contact processes which have given rise to the Maltese language has been treated in Kontzi 1980, 1982a, 1982b, 1982c, 1983. Turning north again to Germany and the Netherlands, German and Dutch have been treated from historical, sociolinguistic and ethnolinguistic points of view in Polome 1985 (Proto-Germanic); Goossens 1985 (Dutch); Pabst 1980; Stellmacher 1980 (Low German), Munske 1982, 1984, 1986 (Latin and French as superstrata); Bellmann 1982 (Slavic-German); Jelsma — van der Plank — Gorter 1983, Sjölin 1980 (Low German-Frisian-Dutch). East European language contacts have so far not been the focus of the symposia. However, a few articles must be mentioned here, which have treated eastern language contacts: Yugoslavia: Veiter 1983; Sojat

Contact

linguistics

485

1981; Toporisic 1981; Neweklowski 1985 (Slovenian); Auburger 1980; Reiter 1985 (Macedonian); Albania: Breu 1985; Greece: Kramer 1980 (Aromunian); Moldavia: Werner 1980b; Soviet Union and Finland; Bätori 1979, 1980, 1982; Veenker 1967, 1982, 1985 (Finno-Ugrian and Slavic); Panzer 1982 (East and West Slavic); Panzer 1984 (North Germanic and East Slavic); The Balkans: Steinke 1982 and Szabo 1985.3 6.6. Language contact in Scandinavia In the Scandinavian field a number of books and publications can be mentioned which have appeared and which indicate that there is a basic reorientation from monolingual and ethnocentric Nordic preoccupation in Scandinavia towards bilingual and multiethnic perspectives of all languages spoken in Scandinavia. This reorientation stems from Finno-Ugric studies on the linguistic situation of the Sami and Finnish minorities in the north (cf. especially Hansegärd 1967, 1968, 1971, 1975, 1978, 1979, 1985, 1990).4 North Germanic and FinnoUgric contacts have been treated in the symposia edited by Baudou and Dahlstedt (1980) and by Nylund Torstensson and Wande (1979), and likewise in a Festschrift (cf. Dahlstedt — Hanson — SahlmanKarlsson 1982). Only a selected list of publications for the profuse literature on language contact in Northern Scandinavia and Finland can be presented here: Allardt — Starck 1981 (Finland); Broch — Jahr 1981, 1984, Jahr 1982 (Russo-Norwegian); Collinder 1980, Dahlstedt 1967, 1982, Helander 1984, Korhonen 1982, 1984, Nesheim 1979, Sköld 1979, Johansson 1977, 1985, SOU ( = Statens offentliga utredningar) (Sami-North-Germanic-Finnish); Loman 1972, 1974, 1977a, 1977b (Swedish-Finnish); Lägreid 1984, Panzer 1984, Söderlind 1984 (North Germanic-Old Russian); Larsen 1984, Sondergaard 1984 (Danish-Low German-North Frisian) 5 ; Munske 1984 (French-Scandinavian); Ureland 1984 (Swedish-English); Ureland 1980a (North Germanic-East Slavic-Continental Germanic-Latin). The problem of Scandinavia as a cultural and linguistic unit has also been treated by a number of scholars: Elert 1981; Haugen 1980, 1981; Sigurd 1977. A special volume on Scandinavian language contacts has been published (cf. Ureland — Clarkson 1984). Symposia on the problem of bilingualism, in particular that of the recent immigrants into Scandinavia, have been held: "Grannspräk och

486

P. Sture

Ureland

minoritetsspräk" 1975, Skutnabb-Kangas 1977, Stedje — Trampe 1979, Ejerhed — Henrysson 1981 (minority symposia). Periodicals and research projects on ethnic groups and their linguistic situation in Scandinavia have also been started, e.g. Schwarz 1966, 1976; Swedish as a Target Language (SUM) 1985; Allwood — Strömqvist — Yoionmaa 1981 — 1986; Scandinavian Working Papers on Bilingualism 1982 — 1985; Finno-Ugric Language Contacts in Sweden (FUSKIS) 1984. An intense debate on the theory and methods of teaching and learning Scandinavian languages as a second language has been going on since the 1970s, e.g. Skutnabb-Kangas 1981, 1984 (English translation); Hyltenstam 1985; Hyltenstam — Maandi 1986; and Boyd 1985. A summary of the debate on semilingualism (cf. Hansegärd 1968, 1975, 1985, and 1990) and Scandinavian studies on bilingualism up to 1983 is found in Bratt Paulston 1983 and Skutnabb-Kangas 1984: 248-263. See also Hansegard's 1975, 1977a and 1977b reply to criticism against his semilingual hypothesis, lately also in Hansegärd 1985 and 1990.

Conclusion The field of contact linguistics is expanding so much that it is impossible here to enumerate all the valuable contributions which have appeared during the past few years. For the general theory of language change, the methods applied in research on living languages in contact are indispensable in investigating the consequences of past language contact. By evaluating the results gained in modern language contact studies it is possible to avoid the unconstrained hypotheses of stratomania. It should, however, be remembered that all contact studies are of importance for describing linguistic change, because it is just as relevant to know the present in order to be able to describe the past as it is to know the past in order to be able to describe the present.

Notes 1. Cf., e.g., Miklosich 1862 (Slavic-Roumanian), 1869 (Slavic-Greek), 1 8 7 0 - 7 1 (Albanian-Slavic and Albanian-Romance); 1884—90 (Turkish-East European languages), 1872 (Slavic-Hungarian); and Schuchardt 1884 (Slavic-German and SlavicItalian.

Contact linguistics

487

2. On the publications on Rhaetoromanic dialects and subdialects see the comprehensive bibliography by Iliescu and Siller-Runggaldier (1985). 3. For further references on continental language contacts (especially German in contact with other languages) see the extensive bibliography published in Jakob (1987). 4. A summary of the entire research on minorities in Northern Scandinavia is presented in Ureland (1987a). 5. A short bibliography on Danish-German bilingualism is presented in Sondergaard (1980b).

References Allardt, Erik — Christian Starck 1981 Spräkgränser och samhällsstruktur. Finlandssvenskarna i ett jämförande perspektiv. Lund: Almqvist and Wiksell. Allwood, Jens — Sven Strömqvist — Kaarlo Voionmaa 1982 — 1986 "The ecology of adult language acquisition — Second language acquisition by adult immigrants": A research project of the European Science Foundation Centre of the Swedish Group: Department of Linguistics, University of Gothenburg. Nordic Linguistic Bulletin 6, No. 1: 17 — 38. Althaus, Hans Peter — Helmut Henne — Herbert Ernst Wiegand (eds.) 1980 Lexikon der germanistischen Linguistik. Tübingen: Niemeyer. Ambrose, John — Collin Η. Williams 1981 "On the spatial definition of minority: Scale as an influence on the geolinguistic analysis of Welsh", in: Haugen — McClure — Thomson (eds.), 5 3 - 7 1 . Ascoli, G. 1881 —1886 "Lettere glottologiche", first letter in Rivista di filologia e d'Istruzione Classica Vol. X (1881 —82); second and third letters in Archivio glottologico Italiano Vol. X (1886). (Translated into German in Kontzi (ed.); 2 9 - 5 3 . ) Auburger, Leopold 1980 "Sprachvariation in der Entwicklung der makedonischen Standardsprache (mit besonderer Berücksichtigung der Entwicklung der Akzentstruktur)", in: Ureland (ed.), 1 - 3 8 . Axelsson, M. — Ä. Viberg (eds.) 1985 Inlärning av species och lexikal struktur. (Svenska som malspräk SSMReports No. 2, September 1985). Stockholm: Institute of Linguistics. Axelsson, Μ. — Ä. Viberg et al. (eds.) 1985 Spräkutveckling och undervisningsmodeller. (Svenska som malspräk SSMReport No. 1, March 1985) Stockholm: Institute of Linguistics. Baldinger, Kurt 1963 "Besprechung von B. Malmberg, 1961", Zeitschrift für romanische Philologie 79: 276-278. Batori, Istvän 1979 "Russen und Finnougrier. Zweisprachigkeit und sprachliche Interferenz", in: Ureland (ed.) (1979), 1 - 2 6 .

488 1980

P. Sture Ureland

"Russen und Finnougrier. Zweisprachigkeit und Sprachwechsel", in: Neide (ed.), 389-397. 1982 "Versuch einer Typologie des Sprachkontaktes anhand der finnougrischen Sprachen und ihrer Kontakte", in: Ureland (ed.) (1982), 355-369. Baudou, E. — Karl Hampus Dahlstedt (eds.) 1980 Nordskandinaviens historia i tvärvetenskaplig belysning (Acta Universitatis Umenis. Umea Studies in the Humanities 24). Umeä: Almqvist and Wikseil. Bausch, Karl-Richard — Hans Martin Gauger 1971 Interlinguistica. Festschrift zum 60. Geburtstag von Mario Wandruszka. Tübingen: Niemeyer. Beardsmore, Hugo Baetens 1983 "The 'supreme-language' hypothesis applied to Brussels", in: Neide (ed.) (1983a), 1 5 - 3 9 . Becker-Dombrowski, Carola 1981 "Zur Situation der deutschen Sprache im Elsaß", in: Ureland (ed.) (1981), 149-179. Bellmann, Günter 1982 "Vorschläge zur Integrationstypologie auf der Grundlage des slawischdeutschen Sprachkontaktes", in: Ureland (ed.) (1982), 265-276. Billigmeier, Robert H. 1979 A crisis in Swiss pluralism. The Hague: Mouton. Born, Joachim 1983 "Domäne und Attitüde in den ladinischen Dolomitentälern", in: Neide (ed.) (1983b), 259-272. Boyd, Sally 1985 Language survival. A study of language contact, language shift and language choice in Sweden. Göteborg: University of Gothenburg. Bratt Paulson, Christiana 1983 Forskning och debatt om tväspräkighet. Stockholm: Norstedt. Breu, Walter 1985 "Das Albanische als National- und Minderheitensprache", in: Ureland (ed.) (1985), 4 1 5 - 4 3 6 . Broch, Ingvild — Ernst Häkon Jahr 1981 Russenorsk — Et pidginspräk i Norge (Tromsc-Studier i Spräkvitenskap III). Oslo: Novus. 1984 "Russenorsk: A new look at the Russo-Norwegian pidgin in northern Norway", in: Ureland — Clarkson (eds.), 21—65. Broderick, George 1985 "The development of Insular Celtic", in: Ureland (1985) 153-180. Brandal, Viggo 1939 Preprint of a paper for the 5th Congress of Linguists in Brussels in 1939, in: Kontzi (ed.), 100-101. Brugmann, Karl — Hermann Osthoff 1878 — 1890 Morphologische Untersuchungen auf dem Gebiet der indogermanischen Sprachen. Vols. I —IV. Leipzig: S. Hirzel. Cadiot, Pierre 1980 "Situation linguistique de la Moselle germanophone: un triangle glossique", in: Neide (ed.) (1980), 325-334.

Contact linguistics

489

Camartin, Iso 1982 "Integration und Assimilation von Anderssprachigen (dargestellt an der Sprachsituation in Graubünden)", in: Ureland (ed.) (1982), 107-118. Cathomas, Bernard 1977 Erkundigungen zur Zweisprachigkeit der Rätoromanen. Bern/Frankfurt a. M.: Lang. 1982 "Rätoromanische Spracherhaltung — Konzepte — Maßnahmen — Wirkungen", in: Ureland (ed.) (1982), 119-129. Caudmont, Jean (ed.) 1982 Sprachen im Kontakt — Langues en contact. Tübingen: Narr. Clyne, Michael 1968 "Zum Pidgin-Deutsch der Gastarbeiter", Zeitschrift für Mundartforschung 35: 130-139. 1975 Forschungsbericht Sprachkontakt. Kronberg/Ts.: Scriptor. 1980 "Sprachkontakt/Mehrsprachigkeit", in: Althaus — Henne — Wiegand (eds.), 641-646. Cohen, Marcel 1939 Preprint of a paper for the 5th Congress of Linguists in Brussels in 1939, in: Kontzi (ed.), 100. Collinder, Björn 1980 "Det samiska spräket och samernas äldre historia", in: Baudou — Dahlstedt (eds.), 193-206. Craen, Pete van de 1980 "Frenchification processes in Brussels: A model for verbal strategies", in: Neide (ed.) (1980), 399-406. Craffonara, Lois 1981 "Die kulturelle und politische Situation der Sellaladiner. (Frühjahr 1981)", in: Ureland (ed.) (1981), 8 1 - 1 0 9 . Dahlstedt, Karl-Hampus 1967 "Some observations on Scandinavian-Lappish place names in Swedish Lapland", Lapps and Norsemen in Olden Times (Institutet for sammenlignende Kulturforskning, Serie A: XXVI), Oslo. 1982 "Ord for 'Renko' i nordsvenska dialekter", in: Dahlstedt et al. (eds.), 21-66.

Dahlstedt, Karl-Hampus — Äke Hansson — Siiri Sahlman-Karlsson (eds.) 1982 Spräkhistoria och spräkkontakt i Finland och Νord-Skandinavien. Umeä: Skytteanska samfundet. Darms, Georges 1985 "Aspekte der Entstehung einer neuen Schriftsprache", in: Ureland (ed.) (1985), 377-390. Decsy, Gyula 1973 Die linguistische Struktur Europas. Wiesbaden: Harrassowitz. Decurtins, Alexi 1981 "Zum deutschen Sprachgut im Bündnerromanischen — Sprachkontakt in diachronischer Sicht", in: Ureland (ed.) (1981), 111-147. 1985 "Die Bestrebungen zur schriftsprachlichen Vereinheitlichung der bünderromanischen Idiome — Zur Vorgeschichte des 'Rumansch Grischun'", in: Ureland (ed.) (1985), 349-376.

490

P. Sture Ureland

Delattre, Pierre 1969-1970 "La theorie celtique et les substrats", Romance Philology 23: 480-491. (Also in Kontzi (ed.), 320-335). Denison, Norman 1977 "Language death or language suicide", Linguistics 191: 13 — 22. (Also in Dressier — Wodak-Leodolter (eds.), 13 — 22). 1979 "Zur Triglossie in der Zahre", in: Ureland (ed.) (1979), 2 7 - 3 7 . 1980 "Heterogenität und Kompetenz", in: Wandruszka — Dressier (eds.), 6-25. 1981a "A linguistic ecology for Europe", Grazer Linguistische Studien" 15: 81-95. 1981b "Conservation and adaptation in a plurilingual context", in: Meid — Heller (eds.), 3 3 - 5 2 . Denison, Norman — K. Sornig — H. P. Gadler — H. Grassegger 1986 Muttersprache (η) (Grazer Linguistische Studien 27). Deprez, Kas — Yves Persoons 1983 "Het effect van de onderzoeker op attitudinele uitspraken van brusselse scholieren T.A.V. de vlaamse gemeenschap in Brüssel", in: Neide (ed.) (1983c), 5 7 - 7 7 . Deprez, Kas — Yves Persoons — Marina Streulens — Armel Wynants 1983 "L'autre 'Liberte du pere de famille'. Les enfants non-neerlandophones dans l'enseignement neerlandophone ä Bruxelles", in: Neide (ed.) (1983c), 90-102. Diaz Lopez, Cesar 1980 "Diglossia and social cleavage: The case of Galicia", in: Neide (ed.) (1980), 225-235. Diekmann, Ε. 1979 "Minderheitenprobleme in der Rätoromania", in: Ureland (ed.), 39 — 54. 1981 "Zur Situation des Okzitanischen als sprachliche und kulturelle Minderheit in Frankreich", in: Ureland (ed.) (1981), 181-199. 1982a "Sprachkontakte in Graubünden", in: Caudmont (ed.), 259-282. 1982b "Soziolinguistische Aspekte deutsch-rätoromanischer Interferenzbeziehungen in Graubünden", in: Ureland (ed.) (1982), 131-154. 1986 "Zur Sprachvariation, Sprachnormierung und Sprachpflege im Bündnerromanischen", in: Holtus — Ringger (eds.), 537 — 556. Dittmar, Norbert — H. Haberland 1978 Papers from the First Scandinavian-German Symposium on the Language of Immigrant Workers and their Children, Roskilde. Roskilde: Universitetets Center ( = Rolig Papir 12). Dorian, Nancy C. 1973 "Grammatical change in a dying dialect", Language 49: 413 — 438. 1977 "The problem of the semi-speaker in language death", in: Dressier — Wodak-Leodolter (eds.), 2 3 - 3 2 . 1978 "The fate of morphological complexity in language death: Evidence from East Sutherland Gaelic", Language 54: 590-609. 1981 Language death. The life cycle of a Scottish Gaelic Dialect. Philadelphia: The University of Philadelphia Press.

Contact linguistics

491

Drachmann, Gaberell (ed.) 1977 Salzburger Beiträge zur Linguistik. Tübingen: Narr. Dressler, Wolfgang U. — Ruth Wodak-Leodolter 1977 "Language preservation and language death in Brittany", in: Dressier — Wodak-Leodolter (eds.), 3 3 - 4 4 (cited as 1977a). Dressier, Wolfgang U. — Ruth Wodak-Leodolter (eds.) 1977 "Language death", International Journal of the Sociology of Language 12 (cited as 1977b). Ebneter, Theodor 1980 "Diasystem vs. Kontakt: der Ausdruck der Zukunft im Deutschen, Rätoromanischen und Norditalienischen", in: Werner (ed.), 43 — 59. Egger, Kurt 1977 Zweisprachigkeit in Südtirol. Bozen: Athesia. 1979 "Morphologische und syntaktische Interferenzen an der deutsch-italienischen Sprachgrenze in Südtirol", in: Ureland (1979), 55 — 104. 1981 Dialekt und Hochsprache in der Schule. Beiträge zum Deutschunterricht in Südtirol. Bozen: Athesia. 1983 "Mehrsprachigkeit in Südtirol", in: Neide (ed.) (1983d), 147-162. Ejerhed, Eva — J. Henrysson (eds.) 1981 Tväspräkighet. Papers from the Third Nordic Conference on Bilingualism, 4—5 June 1980, Umeä (Acta Universitatis Umensis/Umeä Studies in Humanities 36). Umeä: Umeä University/Stockholm: Almqvist and Wiksell. Eiert, Claes C. 1981 Internordisk spräkförstäelse. Papers given at a symposium in Rungstedtgaard, 24—26 March 1980, Copenhagen (Acta Universitatis Umensis/ Umeä Studies in the Humanities 33). Umeä: Umeä University. Fennell, Desmond 1981 "Can a shrinking linguistic minority be saved? Lesson from the Irish experiment", in: Haugen — McClure — Thomson (eds.), 32 — 39. 1985 "The relationship of the new Irish nation to Gaelic since the 19th century", in: Ureland (ed.) (1985), 247-272. Ferguson, Charles A. 1977 "Simplified register, broken language and Gastarbeiterdeutsch", in: Molony - Zobl - Stölting (eds.), 2 5 - 3 9 . Fischer, Georg 1980 "Untersuchungen zur Spracheinstellung in der Areler Gegend", in: Neide (ed.) (1980), 131-139. Fishman, Joshua A. 1968 Readings in the sociology of language. The Hague: Mouton. Francescato, G. 1969 "Sostrato contatto linguistico e apprendimento della lingua materna", Archivo glottologico italiano 54: 10—28. (Also in Kontzi (ed.), 193 — 211). 1981a "Friulian", in: Ureland (ed.) (1981), 139-148. 1981b II Bilingue Isolate. Bergamo: Minerva Italica. Gamillscheg, Ernst 1939a Preprint of a Paper for the 5th Congress of Linguists in Brussels in 1939, in: Kontzi (ed.), 8 3 - 8 4 .

492 1939b

P. Sture Ureland

"Zur Geschichte der germanischen Lehnwörter im Italienischen", in: Zeitschrift für Volkskunde 1939: 8 9 - 1 2 0 . (Also in Kontzi (ed.), 336-366). Gärding, E. — G. Bruce — R. Bannert (eds.) 1978 Nordic prosody. Lund: Department of Linguistics. Gartner, Th. 1910 Handbuch der rätoromanischen Sprache und Literatur. Halle: Niemeyer. Gendron, J.-D. - Peter H. Neide 1986 Plurilingua VI. Plurilinguisme en Europe et au Canada. Perspectives de Recherche. Bonn: Dümmler. Giese, Wilhelm 1966 "Der Maureneinfall in Europa und seine kulturellen Auswirkungen", Zeitschrift für Romanische Philologie 82: 1 — 8. Gilbert, Glenn (ed.) 1980 Pidgin and Creole languages. Selected essays by Hugo Schuchardt. Cambridge: Cambridge University Press. Goebl, Hans 1982a "Kulturgeschichtliche Bedingtheiten von Kontaktlinguistik — Bemerkungen zum gegenwärtigen Stand der 'Questione Ladina'", in: Ureland (ed.) (1982), 155-169. 1982b Review of Deutsch und Italienisch in Südtirol, by Johannes Kramer (Reihe Siegen/Beiträge zur Literatur- und Sprachwissenschaft, Band 23. Heidelberg: Winter 1981.) Ladinia VI: 223-249. 1986 "Problems and perspectives of contact linguistics from a Romance scholar's point of view", in: Neide — Ureland — Clarkson (eds.), 125 — 150. Goossens, Jan 1985 "Herauslösung und Herausbildung des Niederländischen", in: Ureland (ed.) (1985), 287-304. Grannspräk 1975 Grannspräk och minoritetsspräk i Norden (Stockholm: Norstedt). Greene, David 1981 "The Atlantic group: Neo-Celtic and Faroese", in: Haugen — McClure — Thomson (eds.), 1—9. Gumperz, John J. 1972 Directions in sociolinguistics. The ethnography of communication. New York: Holt, Rinehart & Winston. Hagström, Björn 1984 "Language contact in the Faroes", in: Ureland — Clarkson (eds.), 171-189. Hansegärd, Nils Erik 1967 Recent Finnish loanwords in Jukkasjärvi Lappish (Acta Universitatis Upsaliensis. Studia Uralica et Altaica Upsaliensia 3). Uppsala: Almqvist & Wiksell. 1968 Tväspräkighet eller halvspräkighet. Stockholm: Aldus/Bonniers. 1971 Finnish loanwords in Jukkasjärvi Lappish classified according to their meaning. Supplement to Recent Finnish Loanwords in Jukkasjärvi Lappish by the same author (Acta Universitatis Upsaliensis 6). Uppsala: Almqvist & Wiksell. 1975 "Tväspräkighet eller halvspräkighet?" Invandrare och minoriteter 3: 7 — 13.

Contact linguistics 1977a

493

"Paneldiskussion om dubbel halvspräkighet", in: Skutnabb-Kangas (ed.), 212-227. 1977b "Loman och halvspräkigheten", Invandrare och minoriteter (ed. by D. Schwarz) 4: 3 6 - 5 1 . 1978 The transition of the Jukkasjärvi Lapps from nomadism to settled life and farming (Studia ethnographica upsaliensia XXXIX). Uppsala: Almqvist & Wiksell. 1979 "Scandinavian as a second language in the educational policy among Lapps and Finns", in: Nylund Torstensson — Wande (eds.), 33—43. Spraken i det Norrbottens flnska omrädet. Arbetsrapport 3: 1988. Luleä: 1985 Hogskolan i Luleä. 1990 Den norrbottensfinska spräkfrägan. En äterblick pä haivspräkighersdebatten. Uppsala Multiethnic Papers 19. Uppsala: Centre for Multiethnic Research. Hartweg, Frederic 1980 "Le dialecte alsacien: domaines d'utilisation", in: Neide (ed.) (1980), 75-82. 1983 "L'erosion lexicale en dialecte alsacien", in: Neide (ed.) (1983), 281 - 2 9 1 . 1986 "Dialektliteratur und lexikalische Entwicklungstendenzen: Zwei Untersuchungen zum Elsässischen", in: Gendron — Neide (eds.), 45 — 53. Hattori, Shirö — K. Inoue 1983 Proceedings of the XIHth International Congress of Linguists, 27 August — 1 September 1982, Tokyo. Tokyo. Haugen, Einar 1953 The Norwegian language in America. Vols. I —II. Philadelphia: University of Pennsylvania Press. 1980 "Language problems and language planning: the Scandinavian model", in: Neide (ed.) (1980), 151-157. 1981 "Language fragmentation in Scandinavia: Revolt of the minorities", in: Haugen — McClure — Thomson (eds.), 100 — 119. Haugen, Einar — J. D. McClure — D. S. Thomson (eds.) 1981 Minority languages today. Edinburgh: University Press. Heiander, Elina 1984 Om trespräkighet. Umeä: Almqvist and Wiksell. Hinderling, Robert 1986 Europäische Sprachminderheiten im Vergleich. Stuttgart: Steiner. Höglinger, Elisabeth 1980 "Interferenzen des Ladinischen und Italienischen in das Deutsche von Grödner Schülern. Beitrag zur Problematik des Deutschunterrichts in den Schulen der ladinischen Ortschaften Südtirols", Ladinia VI: 287 — 302. Hoffmann, F. 1979 Sprachen in Luxemburg. Wiesbaden: Steiner. Holtus, Günter 1985 "Glottogenese in Italien. Zur Entstehung und zu den Anfängen der italienischen Sprache", in: Ureland (ed.) (1985), 133-151. Holtus, Günter — Kurt Ringger (eds.) 1986 Raetia antiqua et moderna. Festschrift W. Theodor Elwert zum 80. Geburtstag. Tübingen: Niemeyer.

494

P. Sture Ureland

Hornung, Maria 1980 "Sprachkontakt und Sprachkonflikt in den von Österreich aus besiedelten deutschen Sprachinseln in östlichen Oberitalien", in: Neide (ed.) (1980), 91-95. 1983 "Zum Einfluß der Staatssprache auf die Gestaltung von Lehnwörtern in mehrsprachigen Bereichen der alten österreichisch-ungarischen Monarchie und in ihren Folgestaaten", in: Neide (ed.) (1983c), 307-317. Hubschmid, Johannes 1982 "Vorindogermanische Substratwörter in den romanischen Sprachen — Methodische Überlegungen und Forschungsergebnisse", in: Ureland (ed.) (1982), 2 7 - 3 8 . Hyldgaard-Jensen, Karl 1980 "Die Begegnung des Dänischen und des Deutschen in Schleswig", in: Neide (ed.) (1980), 237-241. in press Der niederdeutsche Einfluß auf die skandinavischen Sprachen. Proceedings of the 2nd Symposium on (Middle) Low German in Scandinavian, 18 — 20 May 1987, Copenhagen. Hyltenstam, Kenneth 1985 Svenska i invandrarperspektiv. Lund: Liber. Hyltenstam, Kenneth — Katrin Maandi 1986 Nordens spräk som mälspräk. Forskning och undervisning. Stockholm: Institute of Linguistics. Iliescu, Maria — Heidi Siller-Runggaldier 1985 Rätoromanische Bibliographie. Innsbruck: Institut für Romanistik, University of Innsbruck. Isecenko, A. 1934 "Der eurasische Sprachbund", Orient und Occident 17: 30 — 34. Jahr, Ernst Häkon 1982 "Language contact in northern Norway", in: Ureland (ed.) (1982), 307-320. Jakob, Gerhard 1987 Deutschsprachige Gruppen am Rande und außerhalb des geschlossenen deutschen Sprachgebiets — Eine bibliographische Zusammensetzung zum Thema 'Sprache' aus der Zeit nach 1945. Mannheim: Institut für deutsche Sprache. Jakobson, Roman 1931 "Über die phonologischen Sprachbünde", TCLP 4: 234-240. 1938 "Sur la theorie des affinities phonologiques des langues" Actes du IV Congres International de Linguistes. K0benhavn: E. Munksgaard, 1938: 48 — 58. (See Roman Jakobson, Selected Writings I. Phonological Studies. The Hague: Mouton, 1962: 234-46.) Jelsma, G. — R van der Plank — D. Gorter 1983 "A social map of bilingual Friesland", in: Neide (ed.) (1983c), 181-192. Johansson, Henning 1977 Samerna och sameundervisningen in Sverige. Diss. Umeä: Pedagogiska Institutionen. 1985 Elevernas eller skolans kultur och samhälle? Sammanfattning av ett projekt i norra Norrbotten. Umea: Pedagogiska Institutionen.

Contact linguistics Jud,Jakob 1920

495

"Zur Geschichte der bündnerromanischen Kirchensprache", in: Jahresbericht der Historisch-Antiquarischen Gesellschaft von Graubünden. Chur. Kattenbusch, Dieter 1981 "Die provenzalische Minderheit in Süditalien", in: Ureland (ed.) (1981), 41-51. Keim, Inken 1984 Das Deutsch türkischer Gastarbeiter. Tübingen: Narr. Keskitalo, A. J. 1981 "The status of the Sami language", in: Haugen — McClure — Thomson (eds.), 151-162. Klein, Wolfgang — Norbert Dittmar 1979 Developing grammars: The acquisition of German syntax by foreign workers. Berlin/Heidelberg: Springer. Kloss, Heinz 1969 Grundfragen der Ethnopolitik im 20. Jahrh. Wien: Braumüller. 1978 Die Entwicklung neuer germanischer Kultursprachen seit 1800. (2nd edition). Düsseldorf: Schwann. (Originally published as Die Entwicklung neuer germanischer Kultursprachen von 1800 bis 1950. München: Paul, 1952.) 1981 "Unorthodoxe Betrachtungen über Volksgruppen und Volksgruppensprachen in Europa", in: Ureland (ed.) (1981), 1 — 16. Knowles, Jane 1980 "Multilingualism in Luxembourg", in: Neide (ed.), 355 — 361. Kolb, Herbert — Hartmut Lauffer (eds.) 1981 Sprachliche Interferenz. Festschrift W. Betz. Tübingen: Niemeyer. Kolde, Gottfried 1980 "Vergleichende Untersuchungen des Sprachverhaltens und der Spracheinstellungen von Jugendlichen in zwei gemischt-sprachigen Schweizer Städten", in: Neide (ed.) (1980), 243-253. 1982 "Überlegungen zur vergleichenden Sprachkontaktforschung (Am Beispiel der Schweizer Städte Biel und Freiburg)", in: Caudmont (ed.), 59 — 95. Kontzi, Reinhold 1980 "Das Maltesische am Schnittpunkt von Abendland und Morgenland", in: Neide (ed.) (1980), 105-111. 1982a "Das Zusammentreffen der arabischen Welt mit der romanischen und seine sprachliche Folgen", in: Kontzi (ed.), 387 — 449. 1982b "Maltesisch: Sprachgeschichtliche und areallinguistische Aspekte", in: Ureland (ed.), 6 3 - 8 7 . 1982c "Italienisches im Maltesischen gezeigt an verschiedenen Textsorten", in: Caudmont (ed.), 321 - 3 3 7 . 1983 "Methodische Probleme des Maltesischen: Faktoren, die in zweisprachiger Situation die Erhaltung einer kleinen Sprache hemmen oder fördern", in: Neide (ed.) (1983b), 348-358. Kontzi, Reinhold (ed.) 1978 Zur Entstehung der romanischen Sprachen (Wege der Forschung Bd. 162). Darmstadt: Wissenschaftliche Buchgesellschaft.

496 1982

P. Sture Ureland

Substrate und Superstrate in den romanischen Sprachen (Wege der Forschung Bd. 475). Darmstadt: Wissenschaftliche Buchgesellschaft, (cited as 1982d) Korhonen, Olavi 1982 Samisk-finska bättermer och ortnamnselement och deras slaviska bakgrund. Umeä: Umea Universitet. 1984 "Language contact as reflected in the migration of the Finns, the Saamis and the Vikings", in: Ureland — Clarkson (eds.), 67 — 96. Kramer, Johannes 1980 "De tweetaligheid der Aromoenen in Griekenland", in: Neide (ed.) (1980), 255-261. 1981 Deutsch und Italienisch in Südtirol (Reihe Siegener Beiträge zur Literaturund Sprachwissenschaft Band 23). Heidelberg: Winter. Kristol, Andres, M. 1980 "400 Jahre Mehrsprachigkeit in Bivio", in: Ureland (ed.) (1980), 9 5 - 1 1 9 . Ladin, Wolfgang 1982 Der elsässische Dialekt — museumsreif? Strasbourg: Saide. 1983a "Sozio- und psycholinguistische Aspekte der Polyglossie im Department Untereisass — zur Problematik von Spracherzählungen in Sprachgrenzgebieten", in: Neide (ed.) (1983b), 3 8 - 4 8 . 1983b "Mehrsprachigkeit in Ostfrankreich", in: Neide (ed.) (1983d), 9 5 - 1 0 6 . Lägreid, Annelies 1984 "The Scandinavians in Russia: On extra-linguistic factors in the language contact between North Germanic and East Slavic Peoples in the Early Middle Ages", in: Ureland - Clarkson (eds.), 9 7 - 1 1 0 . Langer, Alexander 1983 "Chancen und Hindernisse für Zweisprachigkeit in Südtirol", in: Neide (ed.) (1983c), 171-180. Larsen, Nils-Erik 1984 "Statistical investigations of language death in a North Frisian community", in: Ureland - Clarkson (eds.), 191-220. Lehiste, Ilse 1978 "Polytonicity in the area surrounding the Baltic Sea", in: Gärding — Bruce - Bannert (eds.), 237-247. Lemarchand-Unger, Brigitte 1981 "Die Verwendung der bretonischen Sprache unter soziolinguistischen Aspekten", in: Ureland (ed.) (1981), 201-218. Le Page, Robert B. 1980 "Hugo Schuchardt's Creole studies and the problem of linguistic continua", in: Lichem — Simon (eds.), 114 — 145. Lewy, Ernst 1942 Der Bau der europäischen Sprachen. (Proceedings of the Royal Irish Academy Vol. XLVIII, Section C, No. 2) Dublin: Royal Irish Academy. 1959 — 1960 "In Atlantischer Landschaft. Spuren unerklärter Sprachbeziehungen im westeuropäischen Raum", Sprachforum 3: 219 — 221.

Contact linguistics

497

Lichem, Klaus — H. J. Simon (eds.) 1980 Hugo Schuchardt. Vorträge und Aufsätze des Schuchardt-Symposiums 1977 in Graz. Heft 10. Österreichische Akademie der Wissenschaften. Phil.-hist. Klasse. Sitzungsberichte 373 Band. Wien: Österreichische Akademie der Wissenschaften. Loman, Bengt 1972 Spräk och samhälle. Spräksociologiska problem. Lund: Gleerups. 1974 Spräk och samhälle 2. Spräket i Tornedalen. Lund: Gleerups. 1977a "Den klasslösa syntaxen", in: Loman (ed.) 1977b, 5 — 39. 1977b Spräk och samhälle 3. Social variation i talspräkets syntax. Lund: Liber. 1977c "Svenska spräkets framtid i Finland", in: Sigurd (ed.), 58 — 68. Lüdi, Georges (ed.) 1986 Devenir bilingue — parier bilingue. Actes du 2e Colloque sur le bilinguisme, 20 — 24 September 1984, Universite de Neuchatel, Neuchatel. Tübingen: Niemeyer. Mac Eoin, Gearoid 1982 "Linguistic contacts in Ireland", in: Ureland (ed.) (1982), 2 2 7 - 2 3 5 . Mac Eoin, Gearoid — Anders Ahlqvist — Donnacha 0 hAodha (eds.) 1987a Third International Conference on Minority Languages: General Papers. 20 — 26 June 1986, University College, Galway, Ireland. In: Journal of Multilingual and Multicultural Development Vol. 8, No. 1 & 2. 1987b Third International Conference on Minority Languages: Celtic Papers. 20 — 26 June 1986, University College, Galway, Ireland (Multilingual Matters 32). Clevedon, Avon: Multilingual Matters. Malmberg, Bertil 1961 "Linguistique iberique et ibero-romane", Studia linguistica 15: 57 — 113. (Also in Kontzi (ed.), 128-185.) Markey, Thomas L. (ed.) 1979 The ethnography of variation. Selected writings on pidgins and Creoles. By H. Schuchardt. Transl. T. L. Markey. Ann Arbor: Karoma. Meid, Wolfgang 1985a Der erste zimbrische Katechismus, christlike unt korze Dottrina. Die zimbrische Version aus dem Jahre 1602 der Dottrina Christiana Breve des Kardinals Bellarmin in kritischer Ausgabe (Innsbrucker Beiträge zur Sprachwissenschaft 47). Innsbruck: Institut für Sprachwissenschaft der Universität. 1985b Der zweite zimbrische Katechismus. Dar klöane Catechismo vor des Beloseland. Die zimbrische Version aus dem Jahre 1813 und 1842 des Piccolo Catechismo ad Uso del Regno d'Italia von 1807 in kritischer Ausgabe (Innsbrucker Beiträge zur Sprachwissenschaft 48). Innsbruck: Institut für Sprachwissenschaft der Universität. Meid, Wolfgang — Karin Heller 1979 Italienische Interferenzen in der lautlichen Struktur des Zimbrischen (Österreichische Akademie der Wissenschaften. Philosophisch-Hist. Klasse, Sitzungsberichte, Band 353). Wien: Verlag der österreichischen Akademie der Wissenschaften. 1981 Sprachkontakt als Ursache von Veränderungen der Sprach- und Bewußtseinsstruktur (Innsbrucker Beiträge zur Sprachwissenschaft 34). Innsbruck: Institut für Sprachwissenschaft der Universität.

498

P. Sture Ureland

Meisel, Jürgen 1975 "Der Erwerb des Deutschen durch ausländische Arbeiter", Linguistische Berichte 38: 5 9 - 6 9 . Meisel, Jürgen (ed.) 1977 Pidgin — Creoles: Languages in contact. Tübingen: Gunter Narr Verlag. Messner, Dieter — Mario Wandruszka (eds.) 1984 Das Romanische in den Ostalpen. Wien: Verlag der österreichischen Akademie der Wissenschaften. Miklosich, Franz 1862 "Die slawischen Elemente im Rumänischen", Denkschriften der Wiener Akademie der Wissenschaften phil.-hist. Klasse 12, 1—70. 1869 "Die slawischen Elemente im Neugriechischen", Sitzungsberichte der Wiener Akademie der Wissenschaften philos.-hist. Klasse 63: 529 — 566. 1870 — 1871 "Albanische Forschungen (1. Die slavischen Elemente im Albanischen. 2. Die romanischen Elemente im Albanischen und 3. Die Form entlehnter Verba im Albanischen)", Denkschriften der Wiener Akademie der Wissenschaften, philos.-hist. Klasse 19: 337-374, and 20: 1 - 8 8 , 315-323. 1872 "Die slawischen Elemente im Magyarischen", Denkschriften der Wiener Akademie der Wissenschaften philos.-histor. Klasse 21: 1 —74. 1884 — 1890 "Die türkischen Elemente in den südost- und osteuropäischen Sprachen (Griechisch, Albanisch, Rumänisch, Bulgarisch, Serbisch, etc.)", Denkschriften der Wiener Akademie der Wissenschaften, philos.-hist. Klasse 34, 35, 37, & 38; 1884: 289-338; 1885: 105-192; 1889: 1 - 8 8 ; 1890: 1 - 1 9 4 . Moide, Bertil — Derrick Sharp (eds.) 1984 Second International Conference on Minority Languages, 6 — 12 June 1983, Turku/Äbo, Finland. In: Journal of Multilingual and Multicultural Development Vol. 5, No. 3 & 4. Molony, C. - H. Zobl - W. Stölting (eds.) 1977 Deutsch im Kontakt mit anderen Sprachen. Kronberg/Ts.: Skriptor. Munske, Horst 1982 "Die Rolle des Lateins als Superstratum im Deutschen und in anderen germanischen Sprachen", in: Ureland (ed.) (1982), 237—263. 1984 "French transferences with nasal vowels in the graphematics and phonology of the Germanic languages", in: Ureland — Clarkson (eds.), 231-280. 1986 "What are mixed languages?" in: Neide — Ureland — Clarkson (eds.), 81-95. Neide, Peter 1978 "Sprachkonflikt und Sprachwechsel in Brüssel", in: Ureland (ed.) (1978), 19-41. 1979 Volkssprache und Kultursprache. Wiesbaden: Steiner. Neide, Peter (ed.) 1980 Sprachkontakt und Sprachkonflikt (Zeitschrift für Dialektologie und Linguistik, Heft 32). Wiesbaden: Steiner. 1983a Plurilingua I. Theorie, Methoden und Modelle der Kontaktlinguistik. II. Bonn: Dümmler. 1983b Plurilingua II. Theorie und Methoden der Kontaktlinguistik. Bonn: Dümmler.

Contact linguistics 1983c 1983d 1985 Neide, Peter 1986

499

Plurilingua III. Vergleichbarkeit von Sprachkontakten. Bonn: Dümmler. Plurilingua IV. Mehrsprachigkeit und Multilingualismus. Bonn. Dümmler. Plurilingua V. Methoden der Kontaktlinguistik. Bonn: Dümmler. — P. Sture Ureland — Iain Clarkson (eds.) Language Contact in Europe. Proceedings of the Working Groups 12 and 13 at the XHIth International Congress of Linguists. 29 August —4 September 1982, Tokyo. Tübingen: Niemeyer.

Nesheim, A. 1979 "Cultural contact of the Lapps with their neighbours", in: Nylund Torstensson — Wande (eds.), 95 — 104. Neweklowsky, Gerhard 1985 "Das Werden der slowenischen Schriftsprache", in: Ureland (ed.) (1985), 391-402. Nielsen, Nils Age 1952 "La Theorie des substrats et la linguistique structurale", Acta linguistica 7: 1 - 7 . (Also in Kontzi (ed.), 186-192.) Nylund Torstensson, Eivor — Erling Wande 1979 Cultural contacts in the Arctic area. Symposium Proceedings of the 500Year Jubilee of the University of Uppsala, 6 — 8 June 1977. In: FennoUgrica Suecana 2. Uppsala: Uppsala Universitet. Oksaar, Els 1984 Spracherwerb — Sprachkontakt — Sprachkonflikt. Berlin: de Gruyter. Orlovic-Schwarzwald, Μ. 1978 Zum Gastarbeiterdeutsch jugoslawischer Arbeiter im Rhein-Main-Gebiet (Mainzer Studien zur Sprach- und Volksforschung 2) Wiesbaden: Steiner. Pabst, Klaus 1980 "Das preussische Geschäftssprachengesetz von 1876 — Sprachwechsel nationaler Minderheiten als Mittel politischer Integration", in: Neide (ed.) (1980), 191-200. Panzer, Baidur 1982 "Genetische und stratalinguistische Faktoren in der Entstehung der ostslavischen Schriftsprachen", in: Ureland (ed.) (1982), 321—337. 1984 "Parallel internal development or interference? Post-positive article and possessive perfect in North Germanic and Northern Russian", in: Ureland - Clarkson (eds.), 111-132. Paul, Hermann 1886 Principien der Sprachgeschichte (2nd edition). Halle: Niemeyer. Pellegrini, G. B. 1972 Gli arabismi nelle lingue neolatine, con speciale riguardo all' Italia. Brescia: Paideia: 635-666. Pernstich, Karin 1981 "Deutsch-italienische Interferenzen in der Südtiroler Presse", dissertation, University of Innsbruck. Persoons, Yves — Mireille Versele 1980 "Taalverschuiving in Malmedy", in: Neide (ed.) (1980), 4 1 3 - 4 2 0 . 1981 "Neubelgien - Malmedy", in: Ureland (ed.) (1981), 251 - 2 6 4 .

500

P. Sture Ureland

Pisani, Vittore 1939 Preprint of a paper for the 5th Congress of Linguists in Brussels in 1939, in: Kontzi (ed.), 78. 1974 Indogermanen und Europa. München: Fink. Planta, Robert von 1931 "Über Ortsnamen, Sprach- und Landesgeschichte Graubündens", Revue de Linguistique Romane 7: 80—100. Pöckl, Wolfgang (ed.) 1981 Europäische Mehrsprachigkeit. Festschrift zum 70. Geburtstag von Mario Wandruszka. Tübingen: Niemeyer. Pokorny, Julius 1927 — 1930 "Das nicht-indogermanische Substrat im Irischen", Zeitschrift für celtische Philologie XVI (1927): 9 3 - 1 4 4 , 231 - 2 6 6 , 363-394; XVIII (1930), 233-248. Polome, Edgar 1985 "Methodological aspects of glotto- and ethnogenesis of the Germanic people", in: Ureland (1985), 4 5 - 7 0 . Poulsen, J. H. 1981 "The Faroese language situation", in: Haugen — McClure — Thomson (eds.), 144-151. Quix, Marie-Paule 1981 "Altbelgien-Nord", in: Ureland (ed.) (1981), 224-235. Radtke, Edgar 1982 "Regional-italienisch im Meridione — Zur Interferenzanalyse Dialekt — Hochsprache", in: Ureland (ed.) (1982), 8 9 - 1 0 5 . 1985 "Ein 'neues' Italienisch — Zur Wertung der jüngsten Sprachentwicklung", in: Ureland (ed.) (1985), 273-285. Rein, Kurt 1979 "Neuere Entwicklungstendenzen der deutschen Sprache in Rumänien", in: Ureland (ed.) (1979), 125-147. 1980 "Diglossie und Bilingualismus bei den Deutschen Rumäniens", in: Neide (ed.) (1980), 264-269. Reiter, Norbert 1985 "Die Sprachlichkeit des Makedonischen", in: Ureland (ed.) (1985), 402-413. Riedmann, Gerhard 1972 Die Besonderheiten der deutschen Schriftsprache in Südtirol (Duden-Beiträge 39). Mannheim: Duden. 1973 "Wandlungen in der Stadtsprache", in: Bozen — Stadt im Umbruch, 371-381, Bozen. 1979 "Bemerkungen zur deutschen Gegenwartssprache in Südtirol", in: Ureland (ed.) (1979), 149-181. Rindler-Schjerwe, Rosita 1980 "Zur aktuellen Konfliktsituation des Sardischen als Minoritätensprache^', in: Neide (ed.) (1980), 201-208. 1983 "Zum Phänomen der Interferenz im Prozeß des Sprachwechsels — Ein Beitrag zur empirischen Sprachkontaktforschung", in: Neide (ed.) (1983a), 81-91.

Contact linguistics

501

Rizzitano, Umberto 1965 Gli Arabi in Italia. In: L'occidente e I'islam nell'alto medioevo, 2 — 8 april 1964. Vol. I. Settimane di studio del centro italiano di studi sull'alto medioevo XII. Spoleto: Presso la Sede de Centro. Rohlfs, Gerhard 1972 Historische Sprachschichten im modernen Sizilien. München: Bayerische Akademie. Rohr, Rupprecht 1963 Das Schicksal der betonten lateinischen Vokale in der Provincia Lugdunensis Tertia. Berlin: Duncker & Humboldt. 1969 "Lat. anlautendes c und g in der Normandie", Zeitschrift für französische Sprache und Literatur LXXXIX, Heft 3: 2 5 2 - 2 7 7 . 1981 "Sprachliche Minderheiten in Europa am Beispiel des Kalabro-Albanischen", in: Ureland (ed.) (1981), 5 3 - 5 8 . 1982 "Nordgermanische und bretonische Nachwirkungen in der französischsprachigen Bretagne", in: Ureland (1982), 2 1 5 - 2 2 6 . Roider, Ulrike 1980 "Der Sprachwechsel bei den keltischen Völkern", in: Neide (ed.) (1980), 421-424. 1981 "Zweisprachigkeit und grammatische Inter- und Transferenz im Keltischen der Britischen Inseln", in: Ureland (1981), 2 8 5 - 2 9 5 . Sandfeld, Kristian 1930 Linguistique balkanique. Problemes et resultats. Paris: Champion. Saussure, Ferdinand de 1916 Cours de linguistique generale. Paris: Payot. Schmitt, Christian 1982 "Die Ausbildung der romanischen Sprachen. Zur Bedeutung von Varietät und Stratum für die Sprachgenese", in: Ureland (ed.) (1982), 39 — 61. Schmitt-Brandt, Robert 1985 "Entstehung der lateinischen Standardsprache", in: Ureland (ed.) (1985), 121-132. Schöndorf, Kurt Erich — Kai-Erik Westergaard 1987 Niederdeutsch in Skandinavien. Akten des 1. nordischen Symposiums 'Niederdeutsch in Skandinavien' in Oslo 27.2. —1.3.1985 (Beihefte zur Zeitschrift für Deutsche Philologie). Berlin: Schmidt. Schuchardt, Hugo 1882 — 1883 Kreolische Studien, Vol. I— V. Sitzungsberichte der Wiener Akademie der Wissenschaften, Vols. 101, 102, 103, 105. (For a detailed bibliography, see Spitzer (ed.), 2 2 - 2 3 . ) 1884 Slawo-deutsches und Slawo-italienisches. Graz: Leuschner & Lubensky. 1917 Review of de Saussure 1916, Literaturblatt für germanische Philologie XXXVIII: 1 - 9 . 1928 "Über die Lautgesetze — gegen die Junggrammatiker", in: Spitzer (ed.), 51 — 107. (Originally published Berlin: Oppenheim, 1885.) Schwarz, David (ed.) 1966 Swedish minorities. Stockholm: Aldus and Bonniers. 1976 Invandrare och minoriteter. Stockholm.

502

P. Sture Ureland

Sigurd, Bengt (ed.) 1977 De nordiska spräkens framtid. Lund: Berlings. Sigurd, Bengt — Jan Svartvik (eds.) 1981 AILA '81, Proceedings. Vol. I. Sections and Workshops (Association Internationale de Linguistique Appliquee). Lund: Wallin & Dalholm. Sjölin, Bo 1980 "On the modalities of alternation in intragroup bilingual speech", in: Neide (ed.) (1980), 271-281. Skalicka, V. 1935 "Notes sur la declination des langues eurasiatiques", Archiv Orientälni 7: 351-354. Sköld, Tryggve 1979 "The earliest linguistic contacts between Lapps and Scandinavians", in: Nylund Torstensson — Wande (eds.), 105 — 116. Skutnabb-Kangas, Tove 1977 Papers from the First Nordic Conference on Bilingualism, 29 — 30 September 1976. Meddelanden frän Institutionen for Nordisk Filologi. Helsinki: Helsingfors Universitet. 1981 Tväspräkighet. Lund: Liber. 1984 Bilingualism or not: The education of minorities. Transl. Lars Malmberg and David Crane (Multilingual Matters 7). Clevedon, Avon: Multilingual Matters. (Translation of Tväspräkighet 1981.) Söderlind, Stefan 1984 "The realm of the Rus': A contribution to the problem of the rise of the East Slavic Kingdom", in: Ureland — Clarkson (eds.), 133 — 170. Sommerfeit, Alf 1962 Diachronic and synchronic aspects of language. The Hague: Mouton. SOU 1975 Statens offentliga utredningar: Nos. 99—100. Samerna i Sverige. Stöd at spräk och kultur (Stockholm: Utbildningsdepartementet). Sondergaard, Bent 1980a "Vom Sprachenkampf zur sprachlichen Koexistenz im deutsch-dänischen Grenzraum", in: Neide (ed.) (1980), 297-305. 1980b Die Anfänge der bilingualen Forschung im deutsch-dänischen Grenzgebiet — Eine Bibliographie. Köbenhavn: Danmarks Paedagogiske Bibliotek. 1981 "The fight for survival", in: Ureland (ed.) (1981), 297-306. 1984 "Language contact in the German-Danish border region: The problem of interference", in: Ureland — Clarkson (eds.), 221 — 229. Sojat, Antun 1981 "Resultate des Kontakts der kroatischen Standardsprache mit der Urbanen Umgangssprache Zagrebs", in: Ureland (ed.) (1981), 59 — 68. Spitzer, Leo (ed.) 1928 Hugo Schuchardt Brevier. Ein Vademecum der allgemeinen Sprachwissenschaft (2nd edition). Halle/Saale: Max Niemeyer. (Reprinted Darmstadt: Wissenschaftl. Buchgesellschaft, 1978.) Stedje, Astrid — Peter af Trampe 1979 Tväspräkighet (Föredrag vid det andra tväspräkighetssymposiet, 18 — 19 maj 1978). Stockholm: Akademilitteratur.

Contact linguistics

503

Steiger, Α. 1948 — 1949 "Aufmarschstraßen des morgenländischen Sprachgutes", Vox Romanica 10: 4 1 - 6 2 . Steinke, Klaus 1979 "Die sprachliche Situation der deutschen Minderheit in Rumänien", in: Ureland (ed.) (1979), 183-203. 1982 "Probleme der diachronen Sprachkontaktforschung. Am Beispiel der Balkan-Sprachen", in: Ureland (ed.) (1982), 3 3 9 - 3 5 4 . Stellmacher, Dieter 1980 "Mehrsprachigkeit der Niederdeutschen — ein theoretisches oder praktisches Problem?", in: Neide (ed.) (1980), 3 8 3 - 3 9 5 . Stephens, Mike 1978 Linguistic minorities in Western Europe. Dyfed: Gomer. Straka, Manfred (ed.) 1970 Handbuch der europäischen Volksgruppen. Wien: Braumüller. Strauss, D. 1981 "Aspects of German as a minority language in Western Europe", in: Haugen - McClure - Thomson (eds.), 189-200. 1982 "Schottland — einsprachig oder dreisprachig?", in: Ureland (1982), 297-306. SUM Svenska som malsprak (Stockholm: University, Institute of Linguistics). Szabö, Adam T. 1985 "Sprache als identitätstiftender Faktor in Rumänien", in: Ureland (ed.) (1985), 235-245. Taeldeman, Johan 1978 "Französische-flämische Sprachinterferenz in Flandern", in: Ureland (ed.) (1978), 4 3 - 6 6 . 1982 " 'Ingwäonismen' in Flandern", in: Ureland (ed.) (1982), 227-296. Toporisic, Joze 1981 "Slowenisch-deutsche Sprachkontakte", in: Ureland (ed.) (1981), 6 9 - 7 9 . Tovar, Antonio 1951 "La sonorisation et la chute des intervocaliques. Phenomene latin occidental", Revue des Etudes Latines 29: 102 — 120. (Revised by the author in Kontzi (ed.), 251-273.) 1978 "Typologische Perspektiven des Baskischen", in: Ureland (ed.) (1978), 67-81. Trim, Richard 1981 "Central Old Belgium", in: Ureland (ed.) (1981), 237-250. Trubetzkoy, N. S. 1931a "Die phonologischen Systeme", TCLP 4 (1931), 9 6 - 1 1 6 . 1931b "Die Konsonantensysteme der ostkaukasischen Sprachen", Caucasia 8: 1-52. 1939a "Gedanken über das Indogermanenproblem", Acta Linguistica I: 81 —89. 1939b "Zur phonologischen Geographie der Welt", in: Preprints to the Proceedings of the 5th Congress of Linguists to be held in Brussels in 1939. Ureland, P. Sture 1977 "Some comparative aspects of pronominal cliticalization in the Baltic language area", in: Drachmann (ed.), 301 —319.

504 1978a

P. Sture Ureland

"Die Bedeutung des Sprachkontakts in der Entwicklung der Nordseesprachen", in: Ureland (ed.) (1978), 8 1 - 1 2 1 . 1978b "Typological, diachronic, and areal linguistic perspectives of North Germanic syntax", in: Weinstock (ed.), 116 — 141. 1979 "Prehistoric bilingualism and pidginization as forces of linguistic change", The Journal of Indo-European Studies 7: 77 — 104 (cited as 1979a). 1980 "Language contact in Scandinavia as an impetus to language change", in: Neide (ed.) (1980), 4 4 1 - 4 5 1 (cited as 1980a). 1982 "Typological, areal linguistic and statistical aspects of Raeto-Romanic reflexives", in: Ureland (ed.) (1982), 171 - 2 0 0 (cited as 1982a). 1983a "Presentation of the working groups 12 and 13 at the Xlllth International Congress of Linguists in Tokyo, 1983", in: Hattori — Inoue (eds.), 1277-1286. 1983b "Language contact in the Alps — Report on a research project", in: Neide (ed.) (1983), 251-257. 1984 "The influence of American English on American Swedish — A case study on the nature of interference", in: Ureland — Clarkson (eds.), 281-324. 1987 "Language contact research in Northern Scandinavia", in: Mac Eoin — Ahlqvist - 0 hAodha (eds.) (1987a), 4 3 - 7 3 . (cited as 1987a). Ureland, P. Sture (ed.) 1978 Sprachkontakt im Nordseegebiet. Akten des 1. Symposiums über Sprachkontakte in Europa. Tübingen: Niemeyer (cited as 1978c). 1979 Dialekte und Standardsprachen in mehrsprachigen Gebieten Europas. Akten des 2. Symposiums über Sprachkontakt in Europa. Tübingen: Niemeyer, (cited as 1979b). 1980a "Language contact in Scandinavia as an impetus to language change", in: Neide (ed.), 441-451. 1980b Sprachvariation und Sprachwandel. Probleme der Inter- und Intralinguistik. Akten des 3. Symposiums über Sprachkontakt in Europa. Tübingen: Niemeyer (cited as Ureland 1980b). 1981 Kulturelle und sprachliche Minderheiten in Europa. Aspekte der europäischen Ethnolinguistik und -politik. Akten des 4. Symposiums über Sprachkontakt in Europa. Tübingen: Niemeyer. 1982 Die Leistung der Strataforschung und der Kreolistik — Typologische Aspekte der Sprachkontakte. Akten des 5. Symposiums über Sprachkontakt in Europa. Tübingen: Niemeyer, (cited as Ureland 1982b). 1985 Entstehung von Sprachen und Völkern. Glotto- und ethnogenetische Aspekte europäischer Sprachen. Akten des 6. Symposiums über Sprachkontakt in Europa, Mannheim 1984. Tübingen: Niemeyer. 1987 Sprachkontakt in der Hanse. Akten des 7. Symposiums über Sprachkontakt in Europa, Lübeck 1986. Tübingen: Niemeyer (cited as 1987b). Ureland, P. Sture — Iain Clarkson (eds.) 1984 Scandinavian language contacts. Proceedings of the Workgroup on Language Contact in Scandinavia held at the 4th International Conference of Nordic and Modern Linguistics, 23 — 27 June 1980, Oslo. Cambridge: Cambridge University Press. Ureland, P. Sture — George Broderick (eds.) 1990 Language Contact in the British Isles. Proceedings of the 8th International Symposium on Language Contact in Europe, September 18 — 24, 1988, Douglas, Isle of Man. Tübingen: Niemeyer.

Contact linguistics

505

Valkhoff, Marius 1932 Latijn, Romaans, Roemeens. Amersfoort. 1939 Reprint of a paper for the 5th Congress of Linguists in Brussels in 1939, in: Kontzi (ed.), 80. Vanacker, V. F. 1967 "Syntaktische Daten aus französisch-flämischen Tonbandaufnahmen", in: Verhandlungen des Internationalen Dialektologenkongresses II, 844 — 855. Wiesbaden: Steiner. 1973 "Een onnederlandse konstruktie in zuidwestlijke dialekten", in: Album Willem Ρέε, 367 — 377. Tongeren: Michiels. 1977 "Syntaktische overeenkomsten tussen Frans-Vlaamse en Westvlaamse dialekten", in: De Franse Nederlanden. Les Pays-Bas Frangais, 206 — 217. Jaarboek. Reekems: Stichting "Ons Erfdeel". Veenker, Wolfgang 1967 Die Frage des finnougrischen Substrats in der russischen Sprache. Bloomington: Indiana University/The Hague: Mouton. 1982 "Russisch-finnougrische Strata Wirkungen", in: Ureland (ed.) (1982), 371-390. 1985 "Ethnogenese und weitere Entwicklung der mordvinischen Sprachen", in: Ureland (ed.) (1985), 317-348. Veiter, Theodor 1983 "Mehrsprachigkeit in Jugoslawien", in: Neide (ed.) (1983c), 155-170. Verdoodt, Albert 1973 Les problemes des groupes linguistiques en Belgique. Louvain: Institut de Linguistique. Viberg, Ä. (ed.) 1985 Andraspräksinlärning (Svenska som mälspräk (SSM-Reports) No. 3, August 1985). Stockholm: Institute of Linguistics. Viletta, Rudolf 1983 "Untersuchungen zur Mehrsprachigkeit in der Schweiz unter besonderer Berücksichtigung des Rätoromanischen", in: Neide (ed.) (1983d), 107-146. Wagner, Heinrich 1959 Das Verbum in den Sprachen der Britischen Inseln. Tübingen: Niemeyer. Walker, Alastair G. H. 1978 "Nordfriesisch — eine sterbende Sprache", in: Ureland (ed.) (1978), 129-148. 1980 "Some factors concerning the decline of the North Frisian tongue", in: Neide (ed.) (1980), 4 5 3 - 4 6 0 . Wartburg, Walter von 1932 "Die Ursache des Auseinanderfallens der Galloromania in zwei Sprachgebiete: Französisch und Provenzalisch" (Vortrag gehalten in der Sitzung der Phil. hist. Klasse der Sächs. Akademie der Wiss. zu Leipzig vom 18. Mai 1932), Forschungen und Fortschritte 8. Jg, Nr. 21, 20. Juni 1932: 268-269. 1939 Preprint of a paper for the 5th Congress of Linguists in Brussels 1939, in: Kontzi (ed.), 8 7 - 8 9 .

506

P. Sture Ureland

Weinreich, Uriel 1953 Languages in contact. The Hague: Mouton. Weinstock, John (ed.) 1978 The Nordic languages and modern linguistics. Austin, Texas: University Press. Werner, Reinhold 1980 "Der slawisch-griechisch-rumänische Sprachwechsel in walachischen und moldauischen Inschriften 1600-1800", in: Neide (ed.) (1980), 4 6 1 - 4 6 6 (cited as Werner 1980b). Werner, Reinhold (ed.) 1980 Sprachkontakte. Tübingen: Narr (cited as 1980a). Whatmough, Joshua 1939 Preprint of a paper for the 5th Congress of Linguists in Brussels 1939, in: Kontzi (ed.), 7 7 - 7 8 . Windisch, Ernst 1897 "Zur Theorie der Mischsprachen und Lehnwörter", Berichte über die Verhandlungen der sächsischen Gesellschaft der Wissenschaften zu Leipzig, philos.-hist. Klasse, 49: 101 — 126. Leipzig. Winkelmann, Otto 1983 "Sprachgrenzen und Mehrsprachigkeit in den Zentralpyrenäen", in: Neide (ed.) (1983c), 193-203. 1985 "Vom Dialekt zur Nationalsprache — Die Entwicklung des Kastilischen während der Reconquista", in: Ureland (ed.) (1985), 193 — 208. 1975a

Grannspräk och minoritetsspräk i Norden. Nordisk t spräkseminarium 4—6 April 1975. Hanaholmens kulturcentrum. Esbo, Finland. Ed. Nordiska Rädet. Stockholm: Norstedt. 1975b Heidelberger Forschungsprojekt Pidgin-Deutsch. Sprache und Kommunikation ausländischer Arbeiter. Kronberg/Ts.: Scriptor. 1975c "Samerna i Sverige", in: Statens offentliga utredningar. Betänkande av sameutredningen. SOU 1975. Vols. 99 and 100 (Appendices). Stockholm: Göteborgs Offsettryckeri AB. 1978 "Dänisch, Nordfriesisch, Hoch- und Niederdeutsch in Schleswig-Holstein", in: Ureland (ed.) (1978), 149-166. 1980 Flerkulturell kompetens. Uppsatser fran ett symposium i Göteborg. Papers in Anthropological Linguistics 7. Göteborg: Department of Linguistics. 1982 — 1985 Scandinavian Working Papers on Bilingualism. Eds. Kenneth Hyltenstamm and Äke Viberg. No. 1 — 5. Stockholm: University of Stockholm, Institute of Linguistics. 1984 Finsk-ugrisk spräkkontakt i Sverige (FUSKIS) I. Finsk dialektutveckling i en svensk industristad (FIDUS). Uppsala: Publications of the Finno-Ugric Institute at the University of Uppsala. Multilingua, Journal of Interlanguage Communication. Ed. J. C. Sager. Berlin: Mouton — de Gruyter. Statens offentliga utredningar 1975 (SOU). See Samerna i Sverige.

Creolization and language change Ian Hancock

Creole studies, though becoming increasingly central to issues in both theoretical and sociolinguistics, and despite the wealth of literature it has generated, remains unclear in its definitions and its terminology. The present essay seeks to contribute to the clarification of the field with an overview of the main events in its development, a listing of classifications of languages which have been considered to be pidgins and Creoles, and a categorizing of the principal scholars according to their theories of pidginization and creolization. The material presented here incorporates some of the unpublished work of the late John E. Reinecke,* the twentieth-century father of Creole studies. The formalized study of pidginized and creolized languages dates from the last quarter of the nineteenth century, although discussions of such languages first began to appear in the mid-1600s. At the time of its publication in 1975, the Bibliography of pidgin and Creole languages was able to list several thousand titles, and in the decade since then hundreds more, including some 35 full-length books, have appeared. It is anomalous, then, that specialists in pidginization and creolization phenomena are still far from achieving a consensus as to what actually constitutes such languages, or what the processes are which bring them into existence. It is still popularly reiterated by non-specialists that a Creole is a pidgin language which has acquired native speakers, and that "a pidgin is a mixture of two languages" (Smitherman 1980: 32). Such definitions turn up with some frequency in dictionaries, and sometimes in textbooks of general linguistics. However, it is beginning to be argued by * With the permission of Aiko Reinecke.

508

Ian Hancock

some scholars that a Creole may not ever need to have been a pidgin, and that pidgins are neither mixtures of languages nor the result of only two languages coming into contact. Pidginized and creolized languages have been referred to in literature at least since the Middle Ages, though not as discrete linguistic systems. Usually, their lexical similarities to the corresponding metropolitan language have led to their being regarded as imperfectly-learned versions of the latter. Earlier this century the Encyclopedia Britannica was able to define Pidgin English as "an unruly bastard jargon, filled with nursery imbecilities, vulgarities and corruptions"; the late Mario Pei, whose popularizations of linguistics were widely read thirty years ago, believed that pidgins were "a marker of man's inefficiency and erroneous thinking" (Pei 1960: 258). Even early specialists such as Sayer, whose 117-page book Pidgin English (1944) was the first fulllength work on this topic, believed he was dealing with a "quaint and crude macaronic jargon." Attitudes such as these survive today in the academic world. There are Creole specialists who will confirm that their colleagues do not regard their particular field as a serious or even a legitimate one; this is especially true in Europe, where conservative linguistic attitudes perhaps change more slowly. But when such critics are challenged, by those creolists, it becomes evident that no effort has been made on their part to understand what a pidgin or a Creole language, or the process of pidginization, actually consists of. The point has already been made that creolists themselves are not in agreement as to what a pidgin or a Creole language, or the process of pidginization, is. It might be argued that the vague nature of Creole studies itself contributes to the attitudes which exist towards it. Is it the domain of the acquisitionist, or the sociolinguist, or the theoretician? Creolists include among their number historians, formalists, comparativists, and applied linguists. Yet while some linguists in other disciplines may question the validity of Creole studies, others are finding in it the key to language change and the understanding of human communication. Creole studies have grown out of a gradual coming-together of ideas, over the years, by individual scholars in these various areas, and it remains a broad and ill-defined field. It was not until 1959 that the first conference devoted to the subject was held, which brought different scholars, hitherto working for the most part independently,

Creolization

and language

change

509

together on the campus of the (then) University College of the West Indies at Mona, in Jamaica. Many types of languages have been dealt with under the heading of pidgin or Creole. Schuchardt, Coelho and other nineteenth-century scholars dealt with several kinds of contact situation yielding such languages, although most of their categories are not regarded today as representative of either pidgins or Creoles. The categories set up by those earlier writers were enumerated and discussed in Reinecke (1937: 55 — 83), and are listed here. The connection with processes of language acquisition has been acknowledged from the very beginning, but still remains unresolved. 1) Plantation Creole. The linguistic outcome of plantation societies, in which large numbers of speakers of many different languages live separately from, and subservient to, a much smaller dominant group which shares only one language. These have typically been situations of enslavement, and have yielded the best-known among the Creole languages, e. g. Haitian, Sranan. 2) Settlers' Creole. This results from the imposition of the speech of a European minority upon a non-European majority, and that majority's modified acquisition of it. Examples include Afrikaans and Angola Portuguese. 3) Trade jargons. The native language of the dominant group is maintained in drastically-reduced form, and used by that group for interaction with other populations. Examples include the Lingua Franca, Chinook Jargon, and the Inuit Trade Jargon. 4) Artificial languages. Intentionally-created languages having no historical development, and usually having entirely regular and maximally efficient structures. Examples are Esperanto, Volapük and Nordlinn. 5) Settlers' mixed jargons. The native languages of immigrant groups modified by being transplanted into a new linguistic environment: Volga German, Pennsylvania German, American Finnish. 6) Colonial dialects. American or Australian English; Canadian French; Brazilian Portuguese; Mexican Spanish. 7) Foreigners' mixed speech. Conventionalized varieties of host languages acquired by immigrants in their new countries: German-influenced English represented by the Hans Breitmann Ballads or the Katzenjammer Kids; "Yinglish" (Yiddish-influenced English); "Spanglish" (Spanish-influenced English). 8) Dying minor languages. Languages in a state of attrition, in which speakers appear to lose rule competence in the reverse order that they

510

Ian

Hancock

are acquired by LI learners. Examples include Yurok, Western Yiddish, Dalmatian. 9) Babu languages. The result of the imposition of colonial languages through the printed medium without spoken-language models, yielding unnaturally-acquired and unidiomatic non-native varieties of those languages: Filipino Bamboo English; Indian "Chichi" English. 10) Lingua francas. Koineized varieties of existing languages used for communication over large areas, typified by a lack of social registers or intrinsic cultural identity (expression of its own folklore, proverbs, etc.). Commercial Dyula, varieties of KiSwahili, English, Hausa, and Hindi. We dismiss most of the above categories today because it is now generally accepted that pidgin and Creole languages involve not only reduction, but restructuring. A century and a half ago, Wilhelm von Humboldt explored the notion of inner and outer linguistic form, discussed for creolized languages in particular by Dell Hymes (1971: 65 — 87). While all of the above are typified by reduction, i.e., the loss, to a greater or lesser extent, of surface rules and narrower lexical distinctions (and compensationally by the expansion of syntactic function and semantic range of individual lexical items), perhaps only the first involves actual restructuring, the yielding of "new" grammars. The term "restructuring" suggests a) the linguistic outgrowth from one specific prior language, and b) that the existing structure of that language is then reshaped. While this has been argued for by a number of creolists in various ways (listed below), more recent Creole theory claims that rather than the grammar of any given Creole deriving from its already-existing metropolitan (i.e., non-creolized, lexically-related) congener, it emerges from the brain, where a linguistic blue-print exists as part of our genetic makeup. If this is the case, "restructuring" is not the process involved. According to this argument, known as the "Bioprogram Hypothesis" (Bickerton 1981), without any single model to serve as a target, learners in these multilingual situations unconsciously produce their innate grammar. While this can happen wherever human beings are involved, given the right social circumstances, it may be regarded as a universal biological process. Since there is no universal lexicon, such languages must acquire their morpheme stock from specific, external sources, although they are then subject to universal processes of lexical generation (Hancock 1980).

Creolization and language change

511

The fact that pidgins and Creoles share so many of their characteristics, despite their widespread occurrence around the world, was noted very early on, and was the principal reason for the development of interest in them. The bioprogram hypothesis attempts to explain these similarities by universal processes, but there have been many other theories, some of which continue to have adherents. Advocates of the main ones are listed here and the specific sources cross-referenced to the bibliography, though the names included are far from exhaustive. These are not mutually exclusive hypotheses, and it is evident that sometimes quite different processes resulted in creolization. As Smith et al. (1987) state: "The most important moral that can be drawn ... is that the development of each creole must be examined individually." A table of the chronological development of the field is also appended below, and includes a list of most of the principal works which have been published. This is discussed further for the earlier period in Hancock (1987a). The main hypotheses and their

advocates

1) Pidginization and creolization result from the suddenness of the social contact: Hesseling, Tagliavini. 2) The processes are the same as those which operate upon metropolitan languages, though greatly accelerated: Van Name, Meyer, Wackernagel, Goodman. 3) The processes are not the same as stated in 2: Jespersen. 4) They result from the contact of speakers from "advanced" cultures with speakers of "primitive" cultures: Clough, Poyen-Bellisle, Lenz, Adler. 5) The socially-dominant speakers intentionally simplify their speech for the subordinate population: Chevillard, Jespersen, Schuchardt, Schultze. Parallels with baby-talk have been made by Schultze, Adam, and Egger, and see also Ferguson. 6) Belief that the learners are too simple-minded to grasp the intricacies of European linguistic systems: Bertrand-Bocande, Schultze. 7) A master-slave relationship is necessary: Schuchardt, Tagliavini. 8) Trade contact is necessary: Schuchardt, Sayer. 9) Some European languages are more easily (or are already partially) creolized than others, and lend themselves to further reduction: Schuchardt, Van Name. 10) Some of the creolization process is independent of external influence: Jespersen, Schuchardt, Hesseling, Coelho, Lenz, Van Name, Bickerton, Hancock, Baker — Corne.

512

Ian

Hancock

11) The process of creolization is reversed in continuing sustained contact with the lexically-related metropolitan congener; the concept of "decreolization": Van Name, Jespersen, Bloomfield, most modern creolists. 12) Degrees of creolization or decreolization can result in a continuum of registers ranging from broadly-creolized to highly metropolitanized: Van Name, Schuchardt, Jespersen, Lenz, DeCamp. 13) Differences among Creoles of shared lexical base are due to different componential make-up in the formation of each, more than to processes of decreolization: Hancock (1986). 14) Pidgins and Creoles are European lexicons on non-European grammatical frameworks: Bertrand-Bocande, Adam, Comhaire-Sylvain, Herskovits — Herskovits. 15) Their structures are generated by universal, genetic processes: Tagliavini, Hjelmslev, Molony, Bickerton, Byrne. 16) They have a non-European substrate: Thomas. 17) They have no non-European substrate: Coelho. Current arguments relating to this and to 15 and 16 above are found in Muysken — Smith. 18) They share similar structures because they originate from one earlier pidgin which has undergone replacement of its vocabulary, but not its structure, in different places in contact with different colonizing powers (the "Relexification Hypothesis"): Thomas, Whinnom, Stewart, Taylor, Thompson. 19) They developed in domestic, i.e., socially-equable, household, situations rather than in master-slave situations: Lichtveld, Valkhoff, Tonkin, Hancock. 20) All European-related pidgins and Creoles share a common ultimate origin (the "Monogenetic Hypothesis"): Bos, and proponents of the relexification hypothesis, 18 above. 21) All pidgins and Creoles sharing a common lexical base in a specific geographical locus (e. g., the Atlantic or the South-West Pacific) share a common ultimate origin: Hancock, Cassidy. 22) Each pidgin and Creole developed separately and independently of any others (the "Polygenetic Hypothesis"): Meyer, Coelho, Tagliavini, Hall, some proponents of the universale hypothesis. 23) Pidginization is the result of learners' errors becoming stabilized or "frozen": Coelho, Lenz. For discussions for and against this, see Schumann (1978) and Andersen (1983).

Creolization and language change

513

24) Very many languages must exist in the formative situation for pidginization to occur: Hesseling, Tagliavini, Whinnom, Bickerton, Hancock. 25) Speakers of the lexifier language must leave the situation in order for creolization to progress: Whinnom. 26) Pidginization can result from only two languages in contact: Gonzales, Smitherman. 27) A nautical element is significant in the development and spread of pidgins: Faine, Reinecke, Tagliavini, Hancock. 28) A nautical element is not significant: Goodman, Bickerton. 29) Creolization can occur without prior pidginization: Valdman (1977). Appendix: A selective chronology of Creole studies 1353 The first-known text in Lingua Franca, from Djerba, in Tunisia (Grion). 1455 First recorded instance of Afro-Portuguese, in verse, 14 years after the arrival of first Africans in Portugal (Teyssier). 1520 Humorous poem in Lingua Franca by Encina, in the Jerusalem tour guides (Harvey et al.). 1612 Haedo discusses the Lingua Franca of Algiers. 1640 Bouton refers to a "mixed jargon" in use between Caribs and Europeans in Martinique. 1655 Pelleprat suggests that Antilles Creole French was the result of reinforcement of learners' errors by the Europeans. 1659 Chevillard suggests that Antilles Creole French was the result of the Europeans' deliberately modifying their speech. 1671 Moliere incorporates Lingua Franca text in his Le Bourgeois Gentilhomme, and uses the word Sabir in print for the first time, an early French equivalent for "pidgin". 1718 Herlein includes a short dialogue in Sranan in the second edition of his work. 1735 Atkins records first use in English of the word Crceoles to describe the inhabitants of Jamaica. 1751 First glossary of Macao Lusoasian Creole published in China (see R. W. Thompson 1959). 1770 Magens writes the first systematic grammar of a Creole language: Virgin Islands Creole Dutch. 1774 Moravian Brethren publish the first whole text in a Creole: a psalm book in Virgin Islands Creole Dutch. First pedagogical materials appear in the same year for the same language. 1778 Van Dyk publishes the first study of Sranan, Moravian Brethren publish a catechism in the same language, and Schumann's dictionary of Saramaccan is completed.

514 1780 1781 1783 1802 1811 1818 1820 1825 1829 1830

1836 1842 1844 1846 1849 1852 1856 1859 1863 1868 1869 1870 1872 1875 1876 1877

Ian Hancock

The first lexicon of Pasar Malay and of Java Lusoasian Creole, the Nieuwe Woordenschat is published by Lodewyk Domenicus. Further scriptural translations made into Sranan and Creole Dutch. C. L. Schumann completes the third (and only dated) edition of his dictionary of Sranan. Ducoeurjoly describes Haitian Creole French. Berrenger completes the second (and only dated) edition of his grammar of Ceylon Creole Portuguese. Wesleyan Missionaries publish scriptures in the same language. Chrestien publishes the first work containing Mauritian Creole French. First-known publication in Papiamentu, the Declaracion corticu di catecismo. Moravians translate the New Testament into Sranan; first published catechism in Mauritian Creole French. Greenfield's Defence appears; the first full-length publication dealing with the status of Creole languages, and first to compare two Creoles. A lexicon and phrase book in the Lingua Franca is produced for French troops in Algeria. First discussion in print of China Coast Pidgin English, by Williams. First grammar of a French Creole (for Martinique) published by Goux. St. Matthew translated into Papiamentu; van der Vegt's manual of Sranan published. Marbot publishes his version of La Fontaine's fables in Martinique Creole French. Hale's work on Chinook Jargon appears. Hery publishes his stories in Reunion Creole French. A grammar of Dutch, in Papiamentu, is written by Putman. Cerepanov writes about the Russian-Chinese contact language of Kjachta. Bonneau, a French journalist, provides a brief description of Haitian Creole French, remarking on its "African" grammar. First use of the word pigeon (i. e., "pidgin") in print, in connection with China Coast Pidgin English. Teza's grammatical outline of Papiamentu. Russell's description of Jamaican, the first to describe a Caribbean Creole. Thomas publishes his grammar of Trinidad Creole French. First comparative study of several Creoles by Van Name. Saint-Quentin publishes his grammar of Cayenne Creole French; Teza publishes on Ceylon Lusoasian Creole. First Papiamentu dictionary, by Van Ewijk. Leland's Pidgin-English sing-song in China Coast Pidgin, included extended discussion of pidginized languages. Turiault finishes his grammar of Mauritian Creole French.

Creolization and language change

515

1879 Egger compares pidginization and hypocorism (baby-talk). 1880 Mercier publishes on Louisiana Creole French. 1880 — 1886 Coelho's work during this period stimulates extensive interest in varieties of Creole Portuguese by various writers. 1881 — 1920 Schuchardt produces a series of important monographs, and also stimulates interest in the field. Regarded as the father of nineteenthcentury creole studies. 1882 Hahn is the first to link creolization to Afrikaans. 1883 Gullah first used in American literature by Joel Chandler Harris. Lucien Adam's comparative study of Creole French dialects. 1884 Steere remarks on Zanzibar pidginized KiSwahili. Harrison's study of Black English published. 1885 Parepou's Atipa is the first novel to appear entirely in a Creole. 1889 Grade publishes a number of articles on West African Pidgin English. 1891 Outline of Annobon Creole Portuguese by Vila. 1894 Poyen-Bellisle completes first doctoral dissertation on a creole-related topic (the structure of Antilles Creole French). 1905 Hesseling lays foundation for the monogenetic hypothesis in his study of Creole Dutch. 1908 First linguistic account of Gullah by Bennett. 1909 Stefansson's essay on the Inuit trade jargon of Herschel Island. 1910 Links between manual signing and pidginization discussed by Long. 1922 Jespersen deals with several pidgins in his language textbook. 1931 Tagliavini summarizes the field of Creole studies to that date. 1933 Bloomfield provides definitions for "pidginization", "creolization", and "decreolization" which are taken as basic in the field for the following half-century. 1937 Reinecke's doctoral dissertation brings together all previous work in the area and presents much new data on hitherto unstudied pidgins and Creoles, a work so far unsurpassed for its completeness. This, and his Bibliography of pidgin and creole languages (1975) earn him the title

1939 1940 1942 1951 1956 1957 1959 1960 1961

of father of twentieth-century creole studies. Hjelmslev presents his "optimum grammar" hypothesis. First thesis on Krio presented by Sawyerr. Robert A. Hall, Jr. begins writing on pidgin and creole languages, and helps draw academic attention to the field. Robert B. Le Page begins linguistic survey of the Caribbean. Whinnom publishes his study of Philippine Creole Spanish. Chomsky's Syntactic structures appears, and helps influence later studies of pidginization phenomena. First creole conference held in Jamaica. The Le Page/DeCamp study of Jamaican Creole is published. Proceedings of the 1959 conference published.

516 1964 1965 1966

1967 1968 1969

1971 1972

1973

1974

1975

1976

1977

Ian

Hancock

Goodman's comparative study of Creole French dialects published. Whinnom publishes first explicit statement of the relexification hypothesis. Hall's Pidgin and Creole languages appears, the first volume wholly devoted to the field. Bailey's transformational description of Jamaican published. Valkhoff s work on Afrikaans and Creole Portuguese restates Hesseling's earlier hypothesis. Cassidy and Le Page publish the first edition of the Dictionary of Jamaican English. Mona Conference on Pidginization and Creolization brings over fifty specialists together. Le Page/Tabouret-Keller survey of multilingualism in Belize begins. Hancock publishes first comparative study of anglophone Atlantic Creoles; Nagara publishes on Japanese Pidgin English. Proceedings of the 1968 conference published, edited by Dell Hymes. Georgetown Round Table on Pidgins and Creoles, with sixty participants. Kay and Sankoff present a paper entitled "A language universals approach to pidgins and Creoles", published in the proceedings (1974, below). Appearance of Dillard's Black English stimulates interest in Creole connections of Afro-American dialects for the rest of the decade. A UNESCO-sponsored Conference on Pidgins and Creoles is held in Trinidad in August. The Carrier pidgin founded as a newsletter for creolists; Heine's PidginSprachen im Bantu-Bereich is the first full-length study of a non-IndoEuropean-related group of pidgins. Publication of Mühlhäusler, Pidginization and simplification of language·, DeCamp and Hancock (eds.), Pidgins and Creoles: Current trends and prospects', and Todd, Pidgins and Creoles. International Conference on Pidgins and Creoles held in Honolulu, with over 100 specialists participating. Appearance of Reinecke et al., A bibliography of pidgin and Creole languages, containing, in addition to the bibliographical entries, introductory essays on the different groups of pidgins and Creoles it deals with. Colloquium on Romance Languages in Contact with Non-Romance Languages held at Wuppertal, largely on Creoles. The first conference of the Society for Caribbean Linguistics, entitled "New Directions in Creole Studies", is held in Georgetown, Guyana. Plans to publish the journal Etudes Creoles made by AUPELF. Dillard edits special issue of the International Journal of the Sociology of Language entitled Sociohistorical factors in the formation of the Creoles. First and only two issues of the Journal of Creole Studies published; Amsterdam Creole Studies and Espace Creole also appear, as well as Valdman's Pidgin and creole linguistics·, Meisel's Langues en contact:

Creolization

and language change

517

Pidgins, Creoles', and Taylor's Languages of the West Indies. The Twelfth International Congress of Linguists in Vienna includes a workshop on pidgin and Creole languages. 1978 Second Society for Caribbean Linguistics conference held in Barbados in July. 1979 Conference on Theoretical Orientations in Creole Studies held in St. Thomas, Virgin Islands. Publication of Hancock (ed.), Readings in creole studies', Markey's translation of selected papers by Schuchardt; the translation by Markey and Roberge of selected papers by Hesseling; and Hill (ed.), The genesis of language. 1980 Publication of papers from the 1975 Honolulu conference (Day 1980). Appearance of Sankoff, The social life of language', Gilbert's translation of selected papers by Schuchardt; Alleyne's Comparative Afro-American; and Valdman and Highfield (eds.), Theoretical orientations in creole studies. Third biennial conference of the Society for Caribbean Linguistics held in Aruba in September. 1981 Publication of Highfield and Valdman (eds.), Historicity and variation in creole studies', Muysken (ed.), Generative studies on Creole languages', and the appearance of Bickerton's Roots of language, in which he refines his bioprogram hypothesis which revolutionizes the field. 1982 Fourth biennial conference of the Society for Caribbean Linguistics is held in Paramaribo, Suriname, and dedicated to the memory of the late John E. Reinecke. 1983 Publication of Boretzky's Kreolsprachen, Substrate und Sprachwandel; Carrington (ed.), Studies in Caribbean language; and Woolford and Washabaugh (eds.), The social context of creolization. Fourth International Creole Studies Colloquium is held in Lafayette, Louisiana, in May, and the York Creole Conference, entitled "Urban Pidgins and Creoles", at the University of York in September. 1984 Fifth Biennial Conference of the Society for Caribbean Linguistics held in Mona, Jamaica, August 29th — September 1st. Publication of Todd, Modern Englishes: Pidgins and Creoles. 1985 International workshop on Creoles held in Amsterdam in April. Publication of Le Page and Tabouret-Keller, Acts of identity; Hancock (ed.), Diversity and development in English-related Creoles; and Dalphinis' African and Caribbean languages. 1986 Publication of Muysken and Smith (eds.), Substrata vs. universals in creole genesis; papers from the 1985 Amsterdam Creole Workshop; and Mühlhäusler, Pidgin and Creole linguistics. First issue of the new Journal of Pidgin and Creole Languages appears. A Creole workshop held at the University of York in April, and the Sixth Biennial Conference of the Society for Caribbean Linguistics takes place at Mona, in Jamaica, in August.

518 1987

Ian Hancock Publication of Gilbert (ed.), Pidgin and Creole languages: Essays in memory of John E. Reinecke, and Smith, The genesis of the creole languages of Suriname. A symposium on Pidgin and Creole Languages held at the University of Dulsberg, in March.

References Acton, Thomas — M. Dalphinis (eds.) 1987 Language, Blacks and Gypsies. London: Karia Press. Adam, Lucien 1883 Les idiomes negro-aryen et maleo-aryen; essai d'hybridologie linguistique. Paris: Maisonneuve & Cie. Adler, Max 1977 Pidgins, Creoles and lingua francos. A sociolinguistic study. Hamburg: Helmut Buske. Alleyne, Mervyn 1980 Comparative Afro-American. Ann Arbor. Karoma. Ammon, Ulrich — Norbert Dittmar — Klaus Mattheier (eds.) 1987 Sociolinguistics: An international handbook of the science of language and society. Berlin: de Gruyter. Andersen, Roger (ed.) 1983 Pidginization and creolization as language acquisition. Rowley & London: Newbury House Inc. Ardener, Erwin (ed.) 1971 Social anthropology and language. London: Tavistock. Atkins, John 1735 A voyage to Guinea, Brazil and the West Indies. London. (The term "Craeoles" appears on page 244.) Bailey, Beryl Loftman 1966 Jamaican Creole syntax. Cambridge University Press. Baker, Philip — Chris Corne 1982 Isle de France Creole: Affinities and origins. Ann Arbor: Karoma. Bennett, John 1908 "Gullah: A Negro patois", South Atlantic Quarterly 7: 332-347. Berrenger, John 1811 A grammatical arrangement on the method of learning the corrupted Portuguese as spoken in India. Colombo: The Government Press. Bertrand-Bocande, E. 1849 "Notes sur la Guinee Portugaise ou Senegambie Meridionale", Bulletin de la Societe de Geographie, 3rd series, 12: 57 — 93. Bickerton, Derek 1981 Roots of language. Ann Arbor: Karoma. 1984 "The language bioprogram hypothesis", Behavioral and Brain Sciences 7: 173-221. Bloomfield, Leonard 1933 Language. London: George Allen & Unwin.

Creolization and language change

519

Bonneau, Alexandre 1856 "Les noirs, les jaunes, et la litterature frangaise en Haiti", Revue Contemporaine 29: 107-155. Boretzky, Norbert 1983 Kreolsprachen, Substrate und Sprachwandel. Wiesbaden: Harrassowitz. Bos, Alphonse 1880 "Note sur le Creole que Ton parle ä l'lle Maurice, ancienne lie de France", Romania 9: 571 - 578. Bouton, Jacques 1640 Relation de l'establissement des frangois depuis l'an de 1635 en Vile de Martinique. Paris: Cramoisy. Brown, Arthur (ed.) 1980 The Ann Arbor Black English case. Gainsville: Publication of the John Dewey Society. Byrne, Frank 1983 "Pidgin and Creole languages: How, who, where, when and why", UDO Papers on Language 1: 23—48. Carrington, Lawrence (ed.) 1983 Studies in Caribbean language. St. Augustine: Society for Caribbean Linguistics Publication. Cassidy, Frederic 1964 "Towards the recovery of early English-African pidgin", Colloque sur le Multilingualisme, Brazzaville. Brazzaville: 267 — 277. Cassidy, Frederic — Robert B. Le Page 1967 Dictionary of Jamaican English. Cambridge: Cambridge University Press. Cerepanov, S. I. 1853 "Kjaxtinskoe kitajskoe narecie russkago jazyka", Izvestija Vtorago Otdelenija Imperatorskoj Akademii Νauk: 370 — 373. Chevillard, Andre 1659 Les desseins de son eminence de Richelieu pour l'Amerique. Rennes: J. Durand. Chomsky, Noam 1957 Syntactic structures. The Hague: Mouton. Chrestien, Francois 1820 Les essais d'un bobre africain. Published by the author. 1820 (2nd edition St. Louis: Deroullede et Cie, 1831). Clough, James C. 1876 On the existence of mixed languages. London: Longmans, Green & Co. Coelho, Francisco A. 1880 — 1886 "Os dialectos romanicos ou neo-latinos na Africa, Asia e America", Boletim da Sociedade de Geographia de Lisboa 2: 129 — 196, 3: 451—478, 6: 705-755. Comhaire-Sylvain, Suzanne 1936 Le Creole haitien. Port-au-Prince: published by the author. Dalphinis, Morgan 1985 Caribbean and African languages. London: Karia Press. Day, Richard R. (ed.) 1980 Issues in English Creoles: papers from the 1975 Hawaii conference. Heidelberg: Julius Groos Verlag.

520

Ian Hancock

DeCamp, David 1971 "Towards a generative analysis of a post-creole speech continuum", in: Hymes (ed.), 349-370. DeCamp, David — Ian Hancock (eds.) 1974 Pidgins and Creoles: Current trends and prospects. Washington: Georgetown University Press. Dillard, Joe 1972 Black English. New York: Random House. Dillard, Joe (ed.) 1976 Sociohistorical factors in the formation of the Creoles = International Journal of the Sociology of Language No. 7. Domineus, Lodewyk 1780 Nieuwe woordenschat uyt het Nederduitsch in het Maleish en Portugeesch. Batavia: n. p. Ducoeurjoly, S. J. 1802 Manuel des habitans de Saint-Domingue, suivi du premier vocabulaire franςais-creole. 2 vols. Paris: Lenoir. Egger, Emile 1879 "Observations et reflexions sur la developpement de l'intelligence et du langage chez les enfants", in: Compte rendu de l'Academie des Sciences Morales et Politiques. (Reprinted 1887, Paris: Picard.) Faine, Jules 1936 Philologie creole: etudes historiques et etymologiques sur la langue Creole d'Haiti. Port-au-Prince: Imprimerie de l'Etat. Ferguson, Charles A. 1971 "Absence of copula and the notion of simplicity: A study of normal speech, baby talk, foreigner talk and pidgins", in: Hymes (ed.), 141 —150. Gilbert, Glenn (ed.) 1980 Pidgin and creole languages: Selected essays by Hugo Schuchardt. Cambridge: Cambridge University Press. Gilbert, Glenn (ed.) 1987 Pidgin and creole languages: Essays in memory of John E. Reinecke. Honolulu: University of Hawaii Press. Gonzales, R. J. 1967 "Pachuco, the birth of a Creole language", Arizona Quarterly 33: 343-356. Goodman, Morris 1964 A comparative study of the Creole French dialects. The Hague: Mouton. Goux, M. 1842 Catechisme en langue Creole. Paris: de Η. Vrayet de Surcy. Grade, Paul 1889 "Bemerkungen über das Negerenglisch an der Westküste von Afrika", Archiv 83: 261 -272. Greenfield, William 1830 A defence of the Surinam Negro English version of the New Testament. London: Samuel Bagster. Grion, Giusto 1891 "Farmacopea e Lingua Franca del dugento", Archivo Glottologico Italiano 12: 1 8 1 - 1 8 6 .

Creolization and language change

521

Haedo, Diego de 1927 Topogräfla e historia general de Argel (1612). 3 vols. Madrid: Soc. de Bibliofilos Espanoles. Hahn, Theophilus 1882 On the science of language and its study, with special reference to South Africa. Capetown: Michaelis. Hale, Horatio 1890 An international idiom: A manual of the Oregon trade language, or "Chinook Jargon". London: Whittaker & Co. Hall, Robert Α., Jr. 1942 "Two Melanesian Pidgin texts, with commentary", Studies in Linguistics 1 (6): 1 - 4 . 1955 Hands off Pidgin English! Sydney: Pacific Publications. 1966 Pidgin and Creole languages. Ithaca: Cornell University Press. Hancock, Ian 1969 "A provisional comparison of the English-related Atlantic Creoles", African Language Review 8: 7—72. 1980 "Lexical expansion in Creole languages", in: Valdman — Highfield (eds.), 63-88. 1986 "The domestic hypothesis, diffusion and componentiality: An account of Atlantic anglophone Creole origins", in: Muysken — Smith (eds.), 71-102. 1987a "History of research on pidgins and Creoles", in: Ammon — Dittmar — Mattheier (eds.), 446—456. 1987b "A preliminary classification of the anglophone Atlantic Creoles", in: Gilbert (ed.), 264-333. Hancock, Ian (ed.) 1979 Readings in Creole studies. Ghent: Story-Scientia. 1985 Diversity and development in English-related Creoles. Ann Arbor: Karoma Publishers. Harris, Joel Chandler 1883 Night with Uncle Remus. Boston: Osgood. Harrison, James 1884 "Negro English", Anglia 7: 232-279. Harvey, L. — R. Jones — Keith Whinnom 1967 "Lingua Franca in a villancico by Encina", Revue de Litterature Comparee 41: 573-579. Heine, Bernd 1973 Pidgin-Sprachen im Bantu-Bereich. Berlin: Reimer. Helmig van der Vegt, Α. 1844 Proeve eener handleiding om het Neger-Engelsch. Amsterdam: Van Kampen. Herlein, J. D. 1718 Beschryvinge van de volks-plantinge Zuriname. Leeuwarden. Herskovits, Melville — Frances Herskovits 1936 Suriname folk-lore. Columbia University Contributions to Anthropology 27. New York: Columbia University Press.

522

Ian Hancock

Hery, Louis-Emile 1849

Esquisses africaines:

Fables Creoles et explorations

dans l'interieur de l'ile

Bourbon. Reunion: Vital Delval. Hesseling, Dirk C. 1905 Het Negerhollands der Deense Antillen. Leiden: Sijthoff. Highfield, Arnold — Albert Valdman (eds.) 1981 Historicity and variation in Creole studies. Ann Arbor: Karoma Publishers. Hill, Kenneth C. (ed.) 1979 The genesis of language. Ann Arbor: Karoma Publishers. Hjelmslev, Louis 1939 "Etudes sur la notion de parente linguistique: Relations de parente des langues Creoles", Revue des Etudes Indo-Europeennes 2: 271 —286. Humboldt, Wilhelm von 1836 Ueber die Verschiedenheit des menschlichen Sprachbaues und ihren Einfluss auf die geistige Entwicklung des Menschengeschlechts. Berlin: Kgl. Akademie der Wissenschaften. Hymes, Dell (ed.) 1971 Pidginization and creolization of languages. Cambridge: Cambridge University Press. Jespersen, Otto 1922 Language: Its nature, development and origin. London: George Allen & Unwin, Ltd. Leland, Charles G. 1876 Pidgin-English sing-song. London: Kegan Paul, Trench & Trübner. Lenz, Rodolfo 1928 El Papiamento, la lengua criolla de Curazao. Santiago: Balcells y Cia. Le Page, Robert B. — David DeCamp 1960 Jamaican Creole. London: Macmillan & Co. Le Page, Robert B. — Andree Tabouret-Keller 1985 Acts of identity: Creole-based approaches to language and ethnicity. Cambridge: Cambridge University Press. Lichtveld, Lou 1954 "Enerlei creools?" West-Indische Gids 35: 5 9 - 7 1 . Long, J. Schuyler 1910 The sign language: A manual of signs. Washington: Gibson Bros. Magens, Joachim M. 1770 Grammatica over det Creolske sprog. Kjebenhavn: Gerhard Giese Salikath. Marbot, Frangoise-Achille 1846 Le bambous: Fables de La Fontaine. Fort-de-France: Ruelle & Arnaud. Markey, Thomas L. 1979

The ethnography

of variation: Selected

writings on pidgins and Creoles ( =

translations of selected works by Schuchardt). Ann Arbor: Karoma Publishers. Markey, Thomas — Paul Roberge (eds.) 1979

On the origin and formation

of Creoles: A miscellany

of articles

translation of Hesseling). Ann Arbor: Karoma Publishers. Meisel, Jürgen (ed.) 1977 Langues en contact: Pidgins — Creoles. Tübingen: Gunther Narr.

(=

a

Creolization and language change

523

Mercier, Alfred 1880 "Etude sur la language Creole en Louisiane", Comptes Rendues de la Athenee Louisanais 5: 378 — 383. Meyer, Paul 1872 Review of Thomas 1869, Revue Critique NS 6: 156-158. Moliere, Jean Baptiste 1671 Le bourgeois gentilhomme. Paris: Le Monnier. Molony, Carol 1970 Processes of Philippine Creole Spanish lexical change. Privately-circulated monograph. Stanford. Mühlhäusler, Peter 1974 Pidginization and simplification of language. Canberra: The Australian National University. 1986 Pidgin and Creole linguistics. Oxford: Basil Blackwell. Muysken, Pieter (ed.) 1981 Generative studies on Creole languages. Dordrecht: Foris Publications. Muysken, Pieter — Norval Smith (eds.) 1986 Substrata vs. universals in Creole genesis. Amsterdam/Philadelphia: Benjamins. Nagara, Susumu 1969 Japanese Pidgin English in Hawaii: A bilingual description. Honolulu: Hawaii University Press. Parepou, Alfred ( = Auguste de Saint-Quentin) 1885 Atipa, roman guyanais. Paris: Auguste Ghio. Pei, Mario 1960 The story of language. New York: Mentor Books. Pelleprat, Pierre 1655 Relation des missions de Pierre Pelleprat de la Compagnie de Jesus dans les lies et dans la terre ferme de l'Amerique meridionale. Paris: Cramoisy. Poyen-Bellisle, Rene de 1894 Les sons et les formes du Creole dans les Antilles. Baltimore: J. Murphy & Co. Putman, Jacobus Johannes 1849 Proeve eener Hollandsche spraakkunst. Curasao: Santa Rosa. Reinecke, John 1937 "Marginal languages: A sociological survey of the Creole languages and trade jargons", Dissertation Yale University. 1938 "Trade jargons and Creole dialects as marginal languages", Social Forces 17: 107-118. Reinecke, John et al. 1975 A bibliography of Pidgin and Creole Languages. Honolulu: Hawaii University Press. Russell, Thomas 1868 The etymology of Jamaican grammar. Kingston: DeCordova, McDougall & Co. Saint-Quentin, Auguste de 1872 "Etude sur la grammaire Creole", in: Introduction ά l'histoire de Cayenne 101 — 169. Ed. Alfred de Saint-Quentin. Antibes: J. Marchand.

524

Ian Hancock

Sankoff, Gillian 1980 The social life of language. Philadelphia: Pennsylvania University Press. Sawyerr, Harry A. E. 1940 "The Sierra Leone patois: A study of its growth and structure, with special reference to the teaching of English in Sierra Leone", M. Ed. thesis: Durham University. Sayer, Edgar S. 1944 Pidgin English. Toronto: E. Sayer. Schuchardt, Hugo 1882-1891 Kreolische Studien Vols. I - I X . Vienna: G. Gerold & Sohn. Schultze, Ernst 1933 "Sklaven- und Dienersprachen", Sociologus 9: 377-418. Schumann, C. L. 1783 Neger-Englisches Wörter-Buch. Paramaribo: The Moravian Archives. Schumann, John H. 1978 The pidginization process: A model for second language acquisition. Rowley & London: Newbury House, Inc. Smith, Norval 1986 The genesis of the creole languages of Suriname. Amsterdam: The University Press. Smith, Norval — Ian Robertson — Kay Williamson 1987 "The Ijo element in Berbice Dutch", Language in Society 16: 49 — 90. Smitherman, Geneva 1980 "Talkin' and testifyin' on Ann Arbor's Green Road", in: Brown (ed.), 26-51. Steere, Edward 1884 A handbook of the Swahili language, as spoken at Zanzibar. London: Society for Promoting Christian Knowledge. Stefänsson, Vilhjälmur 1909 "The Eskimo trade jargon of Herschel Island", American Anthropologist 11: 217-232. Stewart, William A. 1963 "Relexification as a factor in the evolution of Creole languages", paper presented before the Linguistic Society of America, December 1963 meeting, Chicago. Tagliavini, Carlo 1931 "Creole, lingue", Enciclopedia Italiana. Vol. XI. Milan: Instituto Giovanni Treccani, 1931, pp. 8 3 3 - 3 5 . Taylor, Douglas (ed.) 1963 "The origin of the West Indian Creole languages: Evidence from grammatical categories", American Anthropologist 65: 800—814. 1977 Languages of the West Indies. Baltimore: Johns Hopkins University Press. Teyssier, Paul 1959 La langue de Gil Vicente. Paris: Klincksieck. Teza, Emilio 1863 "II dialetto curassese", Politecnico 11: 342-352. Thomas, John J. 1869 The theory and practice of Creole grammar. Port of Spain: The Chronicle Publishing Office.

Creolization and language change

525

1870

"Creole philology", Trübner's American and Oriental Literary Record, 1st December 1870: 5 7 - 5 8 . Thompson, R. Wallace 1959 "Two synchronic cross-sections in the Portuguese dialect of Macao", Orbis 8: 2 9 - 5 3 . 1961 "A note on some possible affinities between the Creole dialects of the old world and those of the new", Creole Languages Studies 2: 107 — 113. Todd, Loreto 1974 Pidgins and Creoles. London: Routledge & Kegan Paul. 1984

Modern Englishes: Pidgins and Creoles. O x f o r d : Blackwell.

Tonkin, Elizabeth 1971 "Some coastal pidgins of West Africa", in: Ardener (ed.), 129—155. Turiault, Jean 1877 Etude sur le Creole de la Martinique. Brest: Imprimerie Lefournier. Valdman, Albert 1977 "Creolisation sans pidgin", in: Meisel (ed.), 105 — 136. Valdman, Albert (ed.) 1977 Pidgin and Creole linguistics. Bloomington/London: Indiana University Press. Valdman, Albert — Arnold Highfield (eds.) 1980 Theoretical orientations in Creole studies. New York: Academic Press. Valkhoff, Marius 1966 Studies in Portuguese and in Creole. Johannesburg: Witwatersrand University Press. Van Dyk, Pieter 1778 Nieuwe en nooit bevoorens geziene onderwijzinge in het Bastert Engels, of Neeger Engels. Amsterdam: Van Egmont. Van Ewijk, Petrus 1875 Nederlandsch-Papiamentsch-Spaansch woordenboekje. Wilemstad. Van Name, Addison 1869 — 70 "Contributions to Creole grammar", Transactions of the American Philological Association 1: 123 — 167. Vila, Istrio 1891 Elementos de la gramatica Ambii ο de Annobon. Madrid: A. Perez Dubrull. Wackernagel, Jakob 1904 "Sprachtausch und Sprachmischung", Nachrichten von der Königlichen Gesellschaft der Wissenschaften zu Göttingen 1904: 26—43. Whinnom, Keith 1956 Spanish contact vernaculars in the Philippine Islands. Hong Kong: The University Press. 1965 "The origin of the European-based pidgins and Creoles", Orbis 14: 509-527. Williams, S. Wells 1836 "Jargon spoken at Canton", Chinese Repository 4: 428 — 435. Woolford, Ellen — W. Washabaugh 1983 The social context of creolization. Ann Arbor: Karoma Publishers.

Bi- and multilingualism: Code-switching, interference and hybrids Nicole

Domingue

Changes occurring in the linguistic choices of speakers do not, for obvious reasons, take place instantaneously. Rather, they result from situations where contact between two or more languages produces wide-spread bilingualism. 1 The phenomenon of bilingualism can be researched from two points of view, the psycholinguistic point of view focusing on the bilingual individual, and the sociolinguistic approach which takes as its object of study the bilingual society. Since this volume is devoted to language change and since language change, though initiated by individuals, occurs only through the medium of society, the societal aspect of bilingualism alone will be discussed. Two main concerns prevail in the research on societal bilingualism: 1) an attempt at a systematic description which gives rise to the elaboration of a model, and 2) a study of the effects of the bilingual situation on the languages in contact.

1. Description of bilingualism The description of bilingualism requires first that the society in which it occurs be circumscribed and the concept of "speech community" is most often used to do so. As speech communities are defined in terms of shared knowledge of linguistic usage (Gumperz 1968; Hymes 1974), their size, social and political extension, and memberships can vary greatly. The second descriptive problem is the choice of criteria used: The number of speakers for each language, the types of languages, their

528

Nicole

Domingue

functions, the attitudes of speakers, have all been considered in descriptions of bilingual situations, either alone or in combination. For instance, Stewart (1968) proposes a classification based on the three following criteria: types of languages, percentage of users (not necessarily native speakers), and functions. Admittedly, this model is purely descriptive and is constructed as a background for language-planning activities. Ferguson (1970) finds it also useful to discuss the multilingual context of the teaching of Arabic in Ethiopia. Most studies in societal bilingualism are not, however, specifically directed toward solving practical problems, and their main emphasis is to explain the roles of the various languages at hand and the attitudes of the speakers toward them in terms of contextual factors. This functional approach is taken by Ferguson (1959) to define, at least partly, the concept of "diglossia". The definition of this term becomes purely functional in Fishman (1967), emphasizing the universality of functional differentiation in the linguistic behavior of human groups. Bilingual communities are viewed as instances of multi-style societies (Hymes 1974: 30), and most researchers subsume under one term — "code" — all speech varieties, be they languages, dialects, styles, or registers, belonging to the verbal repertoire of a speech community (Hymes 1974). The main concern of studies in societal bilingualism is the discovery of the interactional factors of the context which regulate the uses of the codes in a community. The most salient of these factors appear to be those pertaining to the situation, the participants, and the topic. Hymes (1974: 54 — 62) describes them in detail, and expands them into sixteen components of speech. Halliday (1973) prefers to consider as relevant a distinction between field of discourse, mode of discourse, and style of discourse. While a consideration of these factors is invaluable when one wants to describe particular speech events, researchers engaged in the description of bilingual communities must go one step further and try to classify the contextual factors in a systematic fashion. Such an approach is exemplified in Rubin (1962), where a series of binary oppositions represents the verbal behavior of Paraguayan society. Similar systematizations of relevant features can be found, for instance, in Tanner (1967) and Sankoff (1971). These classifications, however, remained impressionistic until Fishman (1972) attempted to explain the phenomenon of choice within a verbal repertoire by bringing the description to a higher level of abstraction. He argued (1972: 18) that the description of a bilingual society should include "a means

Bi- and multilingualism

529

of examining and relating speakers' individual, momentary choices to relatively stable patterns of choice that exist in their multilingual setting as a whole." In his view the concept of "domain" which characterizes sets of situations, themselves defined in terms of interlocutors, topics, and places, provides the possibility to establish such patterns. The concept of domain, as it is constructed, may include different cultural realities but it plays the same role across cultures, thus permitting comparisons (Fishman 1970: 33). The distinction between stable and unstable bilingualism is also made possible thanks to the use of this concept. The bilingual situation is seen as an equilibrium between forces, which sociolinguists have also tried to explain by historical analyses. The questions are as follows: What led to the bilingual situation? Is the situation changing? And if so, how? General factors affecting societies are convincingly put forward to explain the origins and the evolution of bilingualism. Political changes, economic conditions, social and demographic movements (Kloss 1967; Inglehard — Woodward 1967; Brosnahan 1973; Lieberson 1981) all contribute to the maintenance of, and shifting in, the bilingual situation. Recently, however, social psychologists have shown that, often, these factors are not directly responsible for change occurring in bilingual societies, but are channeled through the speakers, whose attitudes are taken to be the immediate factor regulating a bilingual situation (see, for instance, Lambert 1967; Giles — St. Clair 1979). Other properly socio-psychological factors (such as language loyalty, ethnicity, and personality, as they are shown to bear on the question of language choice) also need to be taken into consideration (Fishman 1964; Giles 1977).

2. The languages in contact Bilingual communities, though not necessarily composed exclusively of bilingual individuals, are communities whose speakers share similar sets of values in reference to the codes used. From the point of view of the bilingual speaker, switching between codes in response to accepted community values is part of his linguistic capabilities and reflects changes which occur in the context of the speech event. As early as 1953, however, Weinreich notes that it is necessary to make a distinction between individuals "who have control of their switching, holding it close to the ideal pattern, and those who have difficulty in

530

Nicole Domingue

maintaining or switching codes as required" (1968: 73). Those latter individuals are heard to alternate codes even when the context remains constant and Gumperz (1970: 136) insists that this type of switching must be studied as it "serves definite and clearly understandable communicative ends". Consequently, a distinction between "situational switching" and "metaphorical switching" is established: situational switching, also called "code changing" or simply "code switching", is governed by overt social rules which are rather easily detected by careful observation and involve shifts at obvious structural points, "resulting in sentences which are sequentially L t or L 2 " (Pfaff 1979: 298). Metaphorical switching, often called "code-mixing", seems to be mostly the reflection of stylistic and/or psychological factors. It "entails transferring linguistic units from one code into another. Such a transfer ... results in developing a new ... code of linguistic interaction" (Kachru 1978: 108). The majority of research on code-mixing is geared to answering the following two questions: 1) What produces code-mixing? and 2) What are the linguistic rules which govern it? In reference to 1), it has been found that mixing of L, expressions in L 2 often establishes ethnic identity ties or shows a tendency toward greater informality. Gumperz (1970) suggests also that the degree of emotional involvement with the topic may produce mixing. The use of a mixed code can be deliberately chosen when there is a conflict of purpose: Pfaff (1979) reports the case of a politician whose formal speech needed to be expressed in English but whose goal for ethnic identity could best be transmitted in Spanish; he decided to use both and to mix. Mixing can also be caused by "trigger words" (Clyne 1980) (that is, loanwords, proper names, quotes, set phrases) or by difficulty in finding the right word. As for question 2), a definitive answer has yet to be found and the search for constraints on the possible occurrence of code-mixing has been the subject of most recent research in the field (for instance, Pfaff 1979; Lipski 1978; Kachru 1978). It is now generally accepted that similarity of the surface structure of utterances increases the likelihood of a mix (Sridhar — Sridhar 1980). Thus, the study of mixed codes has the potential to make apparent differential degrees of closeness of syntactic relations within the sentence. Despite the fact that code-mixing produces an additional code available to the members of a bilingual community, it must be described in terms of the two codes which are mixed (Pfaff 1979: 314) since these two codes remain unaltered in the mixes. This fact distinguishes cru-

Bi-and multilingualism

531

daily between code-mixing and interference, since, according to Weinreich (1968: 1), interference phenomena are "instances of deviation from the norms of either language which occur in the speech of bilinguals as a result of their familiarity with more than one language." Interference, then, introduces changes in the languages in contact. At this point it is necessary to follow Weinreich in his distinction between two phases of interference: interference in speech, which occurs at the performance level in bilingual speech, and interference in language, which results from the establishment of patterns in the competence of the speakers and does not depend on bilingualism. Mackey (1970) argues, however, that the distinction is not a clear-cut one. He shows that interference in language — which he calls "integration" — is not necessarily absent from the competence of bilinguals and he attempts to determine when an instance of interference in speech becomes integrated, that is, becomes a borrowing. Borrowings at all levels of linguistic description have been identified, and the field of immigrant language studies has been particularly rich in this respect (Haugen 1973), as well as in noting the effects of borrowing on the spread of English (Kachru 1982). Interference, when it is extreme, may also result in the formation of mixed languages, sometimes called "hybrids" (Whinnom 1971). According to Whinnom, hybrids differ from pidgins in that they result from secondary hybridization — a mixture of the two languages of a bilingual society — rather than tertiary hybridization, which depends on a multilingual situation to produce a relatively stable pidgin. The process of pidginization is not yet completely understood and it is too early to decide whether the distinction between secondary and tertiary hybridization is useful at the functional level of analysis.

Note 1 Unless otherwise indicated, the term "bilingualism" will be used throughout this chapter as a cover term to describe both bilingual and multilingual situations.

References Alatis, James E. (ed.) 1970 Bilingualism and language contact: Anthropological, linguistic, psychological, and sociological aspects. (Monograph Series on Languages and Linguistics 23) Washington, D.C.: Georgetown University Press.

532 1978

Nicole Domingue

International dimensions of bilingual education. Washington, D.C.: Georgetown University Press. Bailey, Richard W. — Jay L. Robinson (eds.) 1973 Varieties of present-day English. New York: Macmillan. Brosnahan, L. F. 1973 "Some historical cases of language imposition", in: Bailey — Robinson (eds.), 40 - 54. Clyne, Michael G. 1980 "Triggering and language processing", Canadian Journal of Psychology 34, No. 4: 400 - 406. Di Sciullo, A.-M. - Muysken, P. - Singh, R. 1986 "Government and code-mixing", Journal of Linguistics 22: 1 —24. Ferguson, Charles A. 1959 "Diglossia", Word 25: 325 - 340. 1970 "The role of Arabic in Ethiopia: A sociolinguistic perspective", in: Alatis (ed.), 355-370. Fishman, Joshua A. 1964 Language loyalty in the United States. New York: Yeshiva University. 1967 "Bilingualism with and without diglossia; diglossia with and without bilingualism", The Journal of Social Issues 23, No. 2: 29 — 38. 1970 Sociolinguistics: A brief introduction. Rowley, Mass.: Newbury House. 1972 "The relationship between micro- and macro-sociolinguistics in the study of who speaks what language to whom and when", in: Pride — Holmes (eds.), 1 5 - 3 2 . Fishman, Joshua A. (ed.) 1968 Readings in the sociology of language. The Hague: Mouton. 1978 Advances in the study of societal multilingualism. The Hague: Mouton. Giles, Howard (ed.) 1977 Language, ethnicity and intergroup relations. New York: Academic Press. Giles, Howard - Robert N. St. Clair (eds.) 1979 Language and social psychology. Baltimore: University Park Press. Gumperz, John J. 1968 "The speech community", in: International encyclopedia of the social sciences. New York: Macmillan, 381—386. 1970 "Verbal strategies in multilingual communication", in: Alatis (ed.) (1970), 129-147. Halliday, Μ. A. K. — Angus Mcintosh — Peter Strevens 1973 "The users and uses of language", in: Bailey — Robinson (eds.), 9 — 37. Haugen, Einar 1973 "Bilingualism, language contact, and immigrant languages in the United States: A research report 1956-1970", in: Sebeok (ed.), 505-591. Hymes, Dell 1974 Foundations in sociolinguistics. An ethnographic approach. Philadelphia: University of Pennsylvania Press. Hymes, Dell (ed.) 1971 Pidginization and creolization of languages. Cambridge: Cambridge University Press.

Bi- and multilingualism

533

Inglehart, R. F. — M. Woodward 1967 "Language conflicts and political community", Comparative Studies in Society and History 10: 2 7 - 4 0 . Kachru, Braj B. 1978 "Code-mixing as a communicative strategy in India", in: Alatis (ed.) (1978), 107-124. Kachru, Braj B. (ed.) 1982 The other tongue: English across cultures. Urbana: University of Illinois Press. Kloss, Heinz 1967 "Bilingualism and nationalism", The Journal of Social Issues 23, No. 2: 39-47. Lambert, Wallace E. 1967 "A social psychology of bilingualism", The Journal of Social Issues 23, No. 2: 9 1 - 1 0 9 . Lieberson, Stanley 1981 Language diversity and language contact. Edited by Anwar S. Dil. Stanford: Stanford University Press. Lipski, John M. 1978 "Code-switching and the problem of bilingual competence", in: Paradis (ed.), 250-264. Mackey, William F. 1970 "Interference, integration and the synchronic fallacy", in: Alatis (ed.) (1970), 195-228. Paradis, M. (ed.) 1978 Aspects of bilingualism. Columbia, S. C.: Hornbeam Press. Pfaff, Carol W. 1979 "Constraints on language mixing", Language 55, No. 2: 291 — 318. Poplack, S. - D. Sankoff 1984 "Borrowing: The synchrony of integration", Linguistics 22, No. 1: 99-135. Pride, J. B. — Janet Holmes (eds.) 1972 Sociolinguistics. Harmondsworth: Penguin. Rubin, Joan 1962 "Bilingualism in Paraguay", Anthropological Linguistics A, No. 1: 52 — 58. Sankoff, Gillian 1971 "Language use in multilingual societies: Some alternative approaches", in: Pride — Holmes (eds.), 33 — 51. Sebeok, Thomas A. (ed.) 1973 Linguistics in North America. Vol. X of Current trends in linguistics. The Hague: Mouton. Sridhar, S. N. - Kamal K. Sridhar 1980 "The syntax and psycholinguistics of bilingual code-mixing", Canadian Journal of Psychology 34, No. 4: 407 — 416. Stewart, William A. 1968 "A sociolinguistic typology for describing national multilingualism", in: Fishman (ed.) (1968), 531-545.

534

Nicole Domingue

Tanner, Nancy 1967 "Speech and society among the Indonesian elite: A case study of a multilingual community", Anthropological Linguistics 9, No. 3: 15 — 39. Weinreich, Uriel 1968 Languages in contact: Findings and problems. The Hague: Mouton. (Originally published 1953, New York: Linguistic Circle of New York.) Whinnom, Keith 1971 "Linguistic hybridization and the 'special case' of pidgins and Creoles", in: Hymes (ed.), 9 1 - 1 1 6 .

Subject Index

Abduction 353, 465 Ablaut 326, 327, 328, 329, 330, 332, 333, 334, 335, 336, 338, 339, 340, 342 Accent 262 — shift 4 Acculturation — vocabulary 287 Acoustic 183, 184, 185, 186 Acquisition — of dative 191 — of first language 186, 190 Acrolect 29, 460 Acronymy 390 Actuation Problem 377 Adequation 400 Affixes 285 — inflectional 285 Agglutination 3, 44, 56, 62 Agreement 13 Allomorphy 357 Allophone 99 Allophonic — representation 98 Alphabet 303 Alternant 336, 337 — modification 337 — extension 337, 339 — lexicalized 340, 341 Alternation 325, 326, 327, 328, 329, 330, 331, 332, 333, 334, 335, 336, 337, 338, 339, 340, 341, 342 — automatic 328, 331 — closed list 329 — combinatory 330 — free 330 — functional 331 — neutralization 330 — non-automatic 326, 331 — productive 327 — stability 328, 333, 334, 335, 336

— suppletive 328 Ambiguity 176, 177, 179 Amelioration 400 Analogy 6, 8, 54, 224, 225, 227, 228, 229, 230, 231, 232, 233, 235, 244, 337, 347, 400 — false 230 — morphophonological 335 — proportional theories of 353 Analogical — change 18 — extension 458 — formation 342 — processes 335, 339 Anaphora 180, 188 Anatolia 428 Ancestry 441 Anglo-Saxon Chronicle 107 Appendix Probi 113 Appropriateness — degrees of 11 Archiphoneme 328 Area — dialect 259 — focal 165, 259 — isolated 140, 262 — linguistic 264 — peripheral 141, 262 — relic 165, 260 — transition 165, 259 Areal — features 264 — linguistics 81, 82, 87, 141, 282 — norms 261 — studies 264, 471 — typology 476 Artificial — intelligence 181 — language 190, 509 — speech 184

536

Subject Index

Aryans 142 Assimilation — of loan words 284 Atlas 258, 261 Atlas linguistique de la France 257 Attitudes of speakers 527, 528, 529

Babbling 186 Babu language 510 Balkans 281, 283, 285, 286 Balkan — Carpathian area 146 — Hypothesis 146 — linguistic area 475 Baltic linguistic area 476 Bartholomae's Law 125 Basilectal 460 Beech 139 Beowulf 101 Benennungsmotiv 428 Bible 41 Bilingual 29, 290 Bilingualism 463, 466, 477, 478, 483, 485, 527, 528 — intrafamilial 288 — societal 527, 528 — stable 529 — unstable 529 Bioprogram hypothesis 79, 460, 510 Biuniqueness 16, 44 Black America 278 Black English vernacular 27, 290 Blending 471, 473, 479 Bohr's Paradox 461 Bolster 425 Borrowing 14, 45, 46, 49, 107, 122, 164, 244, 284, 409, 412, 531 — constraints on 79 — dialect 54, 444 — lexical 287, 289, 411 — morphological 47 — structural 282 Brand name 392 British Received Pronunciation 168 Broad a 168 Bronze Age 143

Caique 49, 285, 287, 391, 410 Canadian rising 171 Carolingian punctuation 102 Case — studies 484 — syncretism 354 Celticity — cumulative mutual 151 Centum 164 Chain 443 Chance 14 Chancery English 103 Chiefdom society 146 Child phonology 186 Chronology — of linguistic change 107, 113, 116, 118, 119, 121, 122, 125, 126, 127, 131 — relative 140 Cimbri 429 Clear cleavage 444 Class 73 Clause order 272 Coalescence 168, 304, 316 Code — code-mixing 530, 531 — code-switching 530 — standardized 478 Cognates 427 Cognitive development 186, 189 Cohort model 177 Common — development 164 — source 14 Comparative — method 17, 18, 300, 444, 445 — reconstruction 17, 18, 20 Comparison — systematic 162 Compartmentalization 461, 462 Competence — bilingual 287 — linguistic 25 — panlectal 25 Compounds 359, 427 — avyayTbhäva (indeclinable) 360 — bahuvrihi (possessive) 360 — dvandva (copulative, coordinate) 359

Subject Index — endocentric 360 — exocentric 360 — karmadhäraya (descriptive) 360 — tatpurusa (dependent) 360 Conflicting reconstructions 20 Consecutio temporum 375 Consonant shift — First (Germanic) 118, 126, 164, 170, 242, 298, 376 — Second (High German) 126, 129, 133, 164 Constant flux (of language) 7 Contact 44, 471 — dialect 120 — linguistic 117, 471 — morphological 48 — reconstructing mutual 476 — semantic 49 — situations 484 — studies 486 Contextual factors 528 Continuum — pidgin-creole 29 Contrasts 304 Convergence 167, 282, 458, 473 — areas 142 — grammatical 282, 286 — structural 286 Copula deletion 27 Correspondences 446 — sound 162 Creole 78, 81, 86, 460 — plantation 509 — settlers' 509 — studies 473, 479 Creolistics 465 Creolization 80, 286, 458, 465, 471, 507 — stratified 460 — fusion 460 Creoloid 80 Critical naturalism 459 Cross-linguistic 192 Cultural — forces 165 — interaction 471 — items 428 — transformation 144

Darwinian gradualism 461 Daughter language 14 Decomposition 316 Decreolization 458, 460, 466, 512 Deduction 465 Deductive method 234 Deflection 3, 4 Degeneration 400 Dephonologization 333, 334 Derivation 427 Descent 441, 442 Deutscher Sprachatlas 257 Development of word meaning 189 Developmentalism 458, 459, 460 Developmentalists 459 Diachronic — analysis 12 — "laws" 313 — linguistics 5 — semantics 400, 401, 402 Diachrony-in-synchrony 404 Diachrony-oriented linguistics 13 Dialect 266, 455 — area 259 — boundary 259 — colonial 509 — contact 120 — elite 28 — geography 7 — differences 161, 166 — dispersion 143 — map 260 — merger 457 — mixture 166, 168 — social 27 — variation 27, 402 — words 163 Dialectology — American 274 Dictionary of Old English 101 Diffusion 264, 289 — of agriculture 143 — lexical 318 — processes 266 — spacial 265 — structural 289 Diglossia 28, 79, 80, 86, 463, 478, 528 Digraphs

537

538

Subject Index

- Old English 99 Diphthongs - Old English 99 Diphthongization - Old High German 298 Discontinuity 461 Discourse 180, 181 - competence 189 - model 180, 181 Distinctive features 185 Divergence 458 Domain 529 Domestic hypothesis 512 Domestication of the horse 144 Domination by conquest 146 Doublets 55 Dove 170 Dravidian languages 283, 286 Drift 55, 57, 63 Dynamic phenomnenon 260

Early oil trade 429 Economy 314 Egalitarian peasants 151 Elevation 400 Ellipsis 390, 401 Enforcement 464 English Great Vowel Shift 113, 123, 124, 313, 315, 317 Episodic memory 182 Embedding Problem 377 Ethnic approach 474, 479 Ethnicity 73, 529 Ethnolinguistics 477 Etymological fallacy 402 Etymology 415, 427 — folk 284 — popular 400, 401 Euphemism 400 Evidence — contact 45, 46, 117 — diachronic 63, 124 — documentary 107, 123, 427 — hypothetical 38, 54 — intuitive 38 — lexical 45 — metrical 42, 116, 375

— orthoepic 113, 116 — orthographic 109, 116, 117, 130 — poetic 116, 130 — reconstructed 59, 60 — rhyme 42, 116 — semantic 59 — suprasegmental 42 — synchronic 57, 63 — typological 20, 56, 60, 61, 62 — theoretical hypothetical 38 Eye-rhymes 116 Extended Standard Theory 455

Family Tree Theory 125, 236 Features — areal 264 — individual 259 — irretrievable 17 — loading 458 — stripping 458 Fieldwork 257 Finns 425 First Grammatical Treatise 129 Flapped r 263 Flip-flops 462 Focus/topic 456 Foreigners' mixed speech 509 Form — derived {forme fondee) 348 — fundamental {forme de fondation) 348 — hybrid 410 — hypercorrect 109 — relic 163 — common underlying 315 French vowel nasalization 315, 316 Frequency 338, 342 — code 326 — high message 336 — message 326 Functional load 308 Fundamental unit of language 366

Gapping 367 Gender 73, 74, 86 Genealogical relationship 140 Generative

Subject Index — grammar 5, 313 — phonology 100, 282, 313, 314 — semantics 455 Generativists 303 Genitive — position of 371 Germanic — accent shift 299 — consonant shift 118, 164, 170, 242, 298, 376 — sleeping habits 425 — umlaut 301 Ghost languages 29 Glosses 101 — Erfurt 120 — Mercian 120 Glottochronology 139, 217, 218 „Glottometry" 457 Graffiti 127, 132 — of Pompeii 108 Grammar growth 190 Grammarians 49, 50 Grammatical development 187 Grammatology 465 Grapheme 98 Graphemics 50 Grassman's Law 317, 379 Greece 428 Grimm's Law 46, 126, 274, 314, 379 Group 329, 339

Haeckel's Law 460 Halaf Culture 145 Haplology 245 — heit 19 Hemp 428 Heterogeneity 479 Historical — change 404 Holophrase 187 Homonymy 54, 116, 117, 168, 403, 445 Homonymie — clash 404 Hybridization 145 — secondary 531 — tertiary 531 Hybrid 531

— form 410 Hypercorrection 113, 320 Hypocorism 515 Hypothesis — consistency of 21 — simplicity of 21

Iconicity 44 — constructional 458 Idiolect 25 Immigrants — recent 484 Implicational hierarchies 402 Importation 409 Incineration 146 India 428 Indo-European — expansion 143 — family 162 — homeland 138 Inductive method 234 Indus Valley Culture 151 Inferencing 180 Inflection 3 — inflectional pattern 427 Informants — interviewing of 258 Innovations 166 — diffusion of 140 — duplication 444 — in method 167 — phonological 119 — shared 220, 442, 443, 444 Integration 531 Interactive model 178 Interference 284, 463, 480, 531 — direct 286 — grammatical 284, 285 — graphic 46 — hypotheses 471 — indirect 286 — lexical 287, 410 — linguistic 285 — puristic attitudes 288 — syntactic 48 Intermediate stage 441 Internal reconstruction 20, 21, 54, 55, 123, 125, 126, 127, 300, 314, 444 Intonation change 375

539

540

Subject Index

Isoglosses 141, 164, 259, 263, 265 - bundles of 260, 261 Isomorphism 44 Isophone 266

Jargon 460, 508 - trade 509 - settlers' mixed 509

-keit 19 Kentish-Mercian charters 120 Koine 80, 464 Kurgan Culture 144

Labovianism 459 Labovian 457, 458, 460, 465 Lachmann's Law 252 Language Acquisition Device 254 — Language Universale 254 — Evaluation Metric 254 — Functional Constraints 254 Language — acquisition 186, 190, 191, 252 — acquisition process 314 — boundaries 483 — change 315, 472, 473, 475, 482 — classification 220, 446 — community 457 — contact 29, 78, 79, 471, 473, 475, 483, 485 — development 175, 186 — divergence 217, 218, 219 — family 441 — maintenance 77 — processing 179 — shift 74 — split 457 — subclassification 446 — subfamily 441 — subgrouping 217, 219, 220 — variation 404 Language death 340, 482 — dying minor languages 509 Language loyalty 288, 529 — mother-tongue loyalty 288

Laryngeals 21, 42, 60, 61 Laryngeal Theory 4 Laut— ersatz 44 — gesetz All — nachahmung 44 Lautverschiebung 224 — first consonant shift 118, 126, 164, 170 — High German consonant shift 126, 129, 133, 164 Lect 456 Lenition 110, 129 Levelling — paradigmatic 339, 356 Lexical — borrowing 287, 289, 411 — change 45 — fading 339 — field 402 — form (of morphemes) 315 — interference 410 — reassignment 338, 339 — representations 313, 315 — resegmentation 285 — search 177 — substitution 395, 399, 401 — truncation 390 Lexicalization 329, 330, 333, 340, 341, 360, 510 — lexifier language 513 Lexicon 175, 176 Lexicostatistics 139, 217, 218, 219, 220 — dating 217 Lexification 341 Lingua franca 460, 464 Lingua rustica 465 Linguistic — accomodation 287 — area 264 — atomism 165 — change 465, 475, 478, 487 — conservation 401 — contact 117 — crystallization 151 — geography 166, 257, 258, 259, 263, 265 — innovation 261, 401

Subject Index — input 191 — map 261 — minorities 482 — paleontology 143, 152, 416 — relationship 219 — relativity 458 — signs 13 — splits 219 — subgrouping 219 — variables 7 — variaties 478 Linguistica Spaziale 261, 262 Linguistics — autonomy of 402 — contact 481 — convergence 472 — developmental 458 — historical 12, 313, 483 Literacy 77 Literary — criticism 462 — studies 405 Litotes 400 Loan 117, 118, 145 — (-)blend 409, 410 — mistranslation 410 — (-)shift 409 — translation 49, 285, 287, 391, 410 — Turkish 461 — (-)word 409 Logogen 176, 177

Manuscripts 100 — Beowulf 102 — Old English 102 Markedness reversal 463 Marker 338 — concomitant (redundant) 327 — primary (relevant) 327 — productive 341 — tense 339 Maximal similarity 16 Mean length of utterance (MLU) 187, 188 Meaning 427 — change 399, 405 — universale of 401

541

Medicine -men 151 Medieval — manuscripts 97 — scribes 100 Memory 181 Mental model 180 Metanymy 401 Merger 109, 458, 459, 460, 461, 462 — conditioned 58 — dialect 457, 464 Metaphor 400, 401 Metaphoric — use 403 Metathesis 124, 316 Method — comparative 17, 300, 444, 445 — direct 257, 258 Methodology 223, 225, 227 — of research 152 Metonymy 390, 400 Metropolitan language 508 Migrant workers 481 Migration — routes of 145 — theory 146 Military aristrocracy 151 Minority 483 — speakers 274 Models — cladistic (tree) 441, 442, 443, 444, 445 — cohort 177 — generative 27 — geometrical 445 — glottochronological 445 — interactive 178 — mental 180 — time-based 26 Modularity 179 Monogenetic hypothesis 512 — replacement 460 Monophthongization — Gothic monophthongization of Proto-Germanic /ai, au/ 301 — East Scandinavian monophthongization of Northwest Germanic /ai, au, iu/ 301 — Latin 122

542

Subject Index

— Old English monophthongization of Northwest Germanic /iu, eo, au/ 301 — Pre-Old High German monophthongization of /ai/ and /au/ 298 Morpheme 285, 315 — alternations 319 — bound morpheme 285 — variants 314, 315 Morphological — blends 47 — change 56 — restructuring 308 Morphologization 320, 333, 334, 341 Morphology 366, 368 — derivational 358 — inflectional 358 Morphophoneme 331 Morphophonemics 242, 341 — morphophonemic representation 98, 99 Morphophonology 325, 326, 330, 331, 332, 341 Morphosyn tactic — change 373 — continuum 462 — loss of inflection 374 Multidimensionality 478 Multilingualism 289, 478, 479, 516 Mundart 463

Nativization 464 Nativizing 463 Natural Generative Phonology 252 Naturalistic determinism 459 Naturalness 458 Necker Cube 461, 462 Neighborhood effect 82 Neogrammarian 3, 84, 163, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 246, 399, 416, 472, 473 Neogrammarian hypothesis 38 Neolinguistica 84, 140, 261 — theory of language 74 Neolithic Period 143 Neumes of Gregorian chant 102 Noetic 459

Nomination 400 Notger's Anlautsgesetz 302 Nouns — abstract 19

Observers' paradox 28 Obsolescence 54 Obsolete words 394 Ockham's Razor 37 Ogam inscriptions 40, 50 Old Europe 144 Old European hydronomy 145 Onomasiology 351 Optimal patterning 334 Optimum grammar hypothesis 515 Orthoepie — commentary 107 — evidence 113, 116 — interpretation 131 Orthoepists 114, 115, 130 Oorthographic — alternation 99 — analysis 303 — convention 128 — evidence 109, 113, 116, 117 — representation 131 — variants 101 Ostension 456 Othello 116 Overlapping 443, 444, 445 — partial 410

Palatalization 124 — Slavic 316 Paradigm 25 — static 25 — dynamic 25, 26 — Principle 253 Paradigmatic — conditions 15, — conditioned change 18 — unity 125 Parent language 14 Parole 6 Pastoralists 144, 148 Patterne 259

Subject Index — geometric 257 — naturalness of 21 — regional 163 — status 168 Pejoration 400 Permutation 400 Philology 97 Phoneme 98, 241, 242 — loss 298 — merger 298 — monitor 176 — shift 298 — split 298 Phonemic — analysis 302 — change 241 — indeterminacy 302 — representation 98, 99 Phonemicization 302 Phonetic — change 282, 283 — Laws 3 — variants 314 Phonic interference 282, 283 — overdifferentiation of phones 282 — phone substitution 45, 282, 283 — re-interpretation of phonemes 282 — undifferentiation of phones 282 Phonology — abstract 249 — autosegmental 317 — child 186 — concrete 249 — generative 100, 282 — diachronic 297 — metrical 317 — structural 297 — structural comparative 477 Phonological change (sound change) 28, 45, 47, 250, 297, 308, 369, 416, 444 — allophonic variation 301 — mechanisms of sound change 297 Phonological — correlations 475 — development 186, 187 — features 167 — innovations 119 — space 307

543

Pidgins 81, 87, 531 Pidginization 80, 286, 471, 507 Pitch — accent 116 Plausibility — test of 21 Polygenetic hypothesis 512 Polysemy 403 Polysemie — clash 404 Polysytematism 479 Pompeii 107, 108, 112, 113 Postvocalic r 167, 168 Poverty of stimulus 190 Pragmatism 465 Prague theory of written language 100 Prehistory 13 Preliterate periods 13 Prescription 77 Prestige 73, 74, 75, 76, 86, 87, 288 Preterite — presents 337 — strong form 20 Primary language 284 Proper Inclusion Precedence 253 Prosodic — information 185, 178 — structures 317 Prothetic vowel 128 Proto-Germanic Vowel System 299 — a-umlaut 299 Protolanguage 17, 20, 125, 143 Pun 116, 117 Punctuated equilibrium 461 Push and drag chain 56, 59, 308

Quantifictional positivists 457, 459 Quantitative studies 273 Questionnaires 171, 257

Racist policies 138 R— dropping 263 — pronouncing 263 Reclassification 338 Reconstruction 64, 416, 427, 445, 446

544

Subject

Index

— comparative 17, 18, 20 — internal 20, 21, 54, 55, 300, 444 Reduplication 335 Regularity 14 — Principle 52, 223, 224, 229, 234, 243 — of sound change 161, 165 Relative units 260 Relationships — systematic 163 Relexicalization 251, 254 Relexification hypothesis 512 Renderings — of actually spoken language 12 Restructuring 298, 356 Retention 443 — rate of 140 — shared 220 Rhenish Fan 165, 261 Rhotacism 107, 128 Rig-Veda 40, 41 Rules 315 — addition 251, 252, 253 — bleeding 64, 253 — counter-bleeding 64, 125 — extrinsically ordered 249, 253 — exchange 318 — feeding 125, 253 — insertion 251, 252 — intrinsically ordered 249, 253 — loss 251, 252, 254, 352 — morphologized 320 — opacity 125, 252 — order 125, 127 — phonetic conditioning rules 28 — phonological 315 — readjustment 332 — reordering 125, 249, 251, 252, 254, 352 — simplification 5, 251, 252, 253, 254, 319, 352 — synchronic 313, 314 — transparency 125 — variable rules 27, 275, 276, 277, 278, 279, 317 — "via" 317 Runic alphabet 300 — acrophonic names 300

Sabir 513 Salmon 139 Sandhi 366 — sandhi change 376 Sarmatians 428 Satem 164 Schema 180, 181 Schriftsprache 463 Scribal — errors 127 — practice 303 — profiles 100 — tradition 110 Scribes — Babylonian 129 — foreign 127 Scythians 428 Secondary — European homeland 145 — language 284 Second Sound Shift 118 Semantic 399 — asymmetry 458 — change 29, 48, 368, 399, 400, 401, 404, 405 — degeneration 74 — fields 144, 427 — loans 410 — memory 182 — principles 140 — verification 182 Semantics — diachronic 400, 401, 402 — generative 455 — synchronic 404 Semasiological lexicon 430 Semilingualism 486 Sentence 366 — phonemics 241 Serial processing 177, 178 Series 329, 338 — subseries 329, 338 Sexism 86 Shadowing 178 Shibboleth 161, 168 Shortening 400 Similarity 13 — grammatical 264

Subject Index — linguistic 264 Simplicity 314 Social — network 73 — stratification (of language) 7, 121 Social subgroups 29 — college students 29 — cowboys 29 — criminals 29 — drug addicts 29 — entertainers 29 — homosexuals 29 — military 29 — prostitutes 29 — teenagers 29 Sociolinguistics 27, 457, 477 Sound Law (Lautgesetz) 223, 224, 225, 228, 229, 230, 231, 232, 233, 235, 236, 473 — exception to the sound law 231 South Asian — languages 27, 30, 283, 285 — subcontinent 142 Sparsamkeit 367 Speech — community 87, 457, 527 — errors 176, 180, 191 — folk 162 — perception 175, 183, 184, 185, 186 — production 175, 183 — rural dialect 262 Spelling — analogical 304 — clues 107 — consistent 111 — etymological 304, 305 — good 112 — naive 108, 109, 112 — poor 108, 112 — research 186 — reverse 304 — unconventional 127 Split 458, 459, 460, 461, 462 — conditioned 58 — language 457, 464 — phonemic 305 — primary 54, 58 — secondary 54

545

— segmental 316 Spoonerisms 243 Sprach— atlas 164 — bund 78, 142, 471 — geist 228 — mischung 78 Spread — of farming 146 Stammbäume 219 Standard language 456, 457 — subvarieties of 481 Standardization 77, 460, 461, 463, 464, 465 State of immobility (of language) 7 Statistical method 349 Stemmatics 442 Strassburg Oaths 107, 109 Strata — underlying 473 Stratalinguistic — historically-oriented approach 474 — research 473 — structurally-oriented approach 474 — view 474 Stratomania 486 Stratum — ad-44 — sub- 44, 78, 283, 471 — substratum approach 476 — substratum population 145 — super- 44, 473 Stress — accent 116 Structuralism 402, 461 Structuralists 303 Subgrouping 443 Substitution 319, 395, 399, 409 Suffixes — nominal 19 Subprasegmentals — Old English 102 Surface — representation 315 — structure 179, 183, 188 Switching — code 530 — metaphorical 530

546

Subject

Index

— situtional 530 Symmetry 308 Synchronic — linguistics 5 — semantics 404 — variation 27, 402 Synchrony-oriented investigations 11 Syncope 108, 112, 113, 119, 123 Syncretism 54 Synecdoche 400 Syntactic — change 27, 250, 365, 367, 368, 369 — development 187 — from OV to VO 372 — VO 370 Syntagmatic — conditioning 18 — conditions 15 Syntax — historical 367 — OV 370 Synthetic languages 62

Taboo 77, 87, 140 Target structures 445 Technicians of the sacred 151 Terminology 392 Text — structure 180, 181 — traditional 162 — understanding 180, 181 Thracians 429 Time depth 139 Topicalization 48 Transfer 281, 400 Transference 480 Transition — area 165, 259 — Problem 377 — zone 260 Trends — divergent 166 Trigger words 530 Triglossia 80 Typology 55, 445, 446, 460, 461 — ambivalent 370 — areal 476

— distinct feature 61 — morphological 471 — OV 370 — phonological 56 — structural 475 — subject-prominent 377 — syntactic 62, 471 — topic-prominent 377 — VO 370, 375 Typological — classification 445 — criteria 16 — factor 444 — relativity 458 — similarities 282

u-apocope 112 Umgangssprache 463 Umlaut 4, 326, 328, 334, 335, 340, 342 — allophones 307 — α-umlaut 299 — English 316 — velar 131 Universale 55, 56, 57, 62, 63, 64, 402, 444, 445 — of meaning 401 Universal G r a m m a r 191 Unmarking 463 Upgliding diphthongal reflex of syllabic a 169 Urban centers 260 Urnfield Civilization 143 Uvular r 263 — in Western Europe 82, 84

Values 77 Variables — ethnic-cultural 480 Variant — conditioned 314 — morpheme 314, 315 — paradigmatically conditioned 15 — positional 169 — syntagmatically conditioned 15 Variation 25, 29, 30, 72, 86, 458 — dialect 27, 402

Subject Index — discourse level 73 — free 25 — geographical 161 — language 404 — phonological 27 — stylistic 28, 121 — synchronic 27, 402 Verner's Law 126, 132, 274, 299, 314, 339, 340, 342, 379 Vespasian Psalter 120 Vocabulary — core 139 Vocal Tract Hypothesis 183, 184 Voice onset 185 Vowel — front 169 — fronting 170 — full grade long 21 — harmony 461 — prothetic 108 — raising 179

547

Vowel shift — Modern English 308 — Norwegian 308 — Swedish 308

Wandering loanwords 137 Wave of Advance Model 146 Wave Theory 7, 163, 236, 443, 445 Wechselwirkung 226 Wellentheorie 458 West Indian Creoles 290 Word formation 358 — compounding 358 — derivation 358 Word - atlas 258 - history 405, 428 Words-and-things (Wörter und Sachen) 402, 458 Written records 97

Language Index

Abaza 60 African languages 40, 219, 411, 446, 447, 450 — African pidgins and Creoles 81 — North African 476 — West African 460 Afrikaans 258, 261, 340, 456, 464, 509, 515, 516 Akkadian 42 — Old Akkadian 129 Albanian 40, 142, 264, 419, 429, 435, 450, 486 Algonquian 448, 449 American Indian 40, 219, 264, 450 American Sign Language 192, 412 Amharic 48 Anatolian 40, 142, 145, 148, 422, 431, 452 Anglo-Norman 461 Anglo-Saxon 42, 112, 458, 460, 461 Arabic 47, 60, 79, 284, 460, 475 — Spanish Arabic 50 Armenian 40, 41, 61, 145, 146, 419, 424, 435, 443 — Old Armenian 454 Aromunian 485 Asian languages 476 Athapaskan 170 Australian 40 Austronesian 219, 220 Avestan 40, 125, 421, 422, 424, 430 Aztec 63

Balkan languages 142, 283, 373, 457, 475 Baltic 40, 61, 419, 434, 453, 475, 476 — Baltic-Finnish languages 476 Bantu 80, 152 Basque 218, 476 Bavarian

— Central Bavarian 260, 266 — North Bavarian 260 — South Bavarian 260 Bengali 41 Berber 476 Brahui 286 Breton 119, 425, 476 — Old Breton 110 Bulgarian 47, 142, 258, 264, 265, 435

Canaanite 450 Catalan 258 Celtic 40, 48, 61, 110, 112, 118, 119, 122, 129, 151, 163, 418, 422, 429, 430, 432, 443, 449, 453, 473, 475 - British Celtic 110, 476 - Britonnic 110, 112, 119 - Goidelic Celtic 476 Chinese 40, 43, 44, 47, 192, 373, 375 - Mandarin Chinese 44 Chinook Jargon 87, 509 Chipewyan 170 Comanche 287 Cornish 119, 425, 476 - Old Cornish 110 Cushitic 48 Czech 46, 61, 390, 435

Danish 258, 263, 476, 484, 485, 487 Dardic 446 Dravidian 142, 283, 373 Dutch 75, 129, 192, 258, 260, 261, 264, 328, 340, 429, 433, 464, 484 — Middle Dutch 429 — Low Franconian 129 — Negerhollands 460 — Old Low East Franconian 129 — Old Low West Franconian 129 — Old Netherlandic 110

Language Index Eblaite 446, 448, 449 Egyptian 40, 41 English 3, 37, 41, 42, 46, 47, 48, 51, 52, 53, 57, 59, 61, 62, 74, 76, 77, 80, 98, 108, 111, 113, 117, 118, 123, 129, 130, 161, 163, 168, 192, 250, 258, 262, 263, 264, 283, 284, 290, 298, 308, 313, 314, 315, 316, 317, 319, 326, 327, 328, 330, 332, 335, 338, 339, 340, 348, 375, 378, 399, 402, 403, 405, 409, 410, 411, 412, 433, 441, 458, 461, 464, 465, 475, 476 — American English 43, 62, 167, 170, 318, 393, 411, 456, 458, 509 — Black English 76, 87, 275, 276, 278, 290, 516 — British English 163, 167, 168, 170, 245, 262, 393, 456 — Canadian English 167 — Chancery English 103, 166 — Cockney English 124 — Early Modern English 43 — Filipino Bamboo English 508 — Indian "Chichi" English 510 — Jamaican Creole English 277 — London Standard English 166 — Marathi English 284 — Middle English 47, 81, 109, 114, 115, 116, 117, 124, 166, 171, 298, 339, 415, 441 — Midwestern American English 73 — North American English 166 — Northern Middle English 244 — Old English 41, 43, 44, 46, 47, 48, 98, 99, 101, 102, 103, 109, 110, 111, 112, 120, 124, 245, 301, 302, 336, 356, 394, 424, 425, 458 — Pidgin English 508 — Southern American English 124, 169, 171 — Standard English 166, 171 — Standard Written English 103 — Substandard English 327 Esperanto 509 Estonian 315, 476 Ethiopic 50 European languages 476, 480

549

Finnish 46, 47, 263, 317, 425, 485 — American Finnish 410, 411, 509 — Baltic-Finnish 318, 476 — Volga-Finnish 476, 486 Franco-Provengal 456 French 46, 47, 48, 50, 61, 62, 63, 80, 87, 109, 117, 121, 122, 128, 129, 130, 131, 192, 257, 258, 263, 283, 315, 348, 368, 375, 390, 409, 456, 460, 464, 465, 476, 484, 485 — American French 509 — Canadian French 131, 411, 509 — Haitian French 460, 509, 514 — Lousiana French 165, 515 — Montreal French 72 — Norman French 460 — Occitan 456 — Old French 54, 122, 128, 129, 360, 403, 415 — Parisian French 165 — Picardian French 390 Frisian 484 — Mainland Northern Frisian 76 — North Frisian 485 — Old Frisian 424 — West Frisian 457, 464 Friulian 74, 456

Gaelic - Irish 476 - Scots 74, 476 Galatian 42 German 4, 19, 20, 42, 47, 58, 62, 74, 77, 80, 87, 98, 109, 110, 113, 122, 123, 130, 162, 163, 164, 171, 192, 231, 232, 257, 258, 261, 263, 298, 304, 326, 327, 328, 336, 340, 347, 372, 391, 410, 411, 422, 434, 441, 456, 463, 464, 484, 486, 487 - Alemannian 460 - High German 76, 79, 109, 129, 164 - Low German 429, 484, 485 - Middle High German 20, 62, 116, 301, 302, 304, 328, 336, 337, 338, 351 - Old High German 4, 41, 47, 54, 59, 62, 69, 98, 100, 103, 110, 118, 122,

550

Language Index

298, 301, 302, 304, 308, 328, 335, 337, 338, 339, 342, 351, 424, 425 — Pennsylvania German 87, 509 — Pidgin-Deutsch 482 — Pre-Old High German 298, 308 — Standard German 164 — Swiss German 79, 258, 335 — Texas German 74 — Volga German 509 Germanic 4, 20, 46, 55, 59, 61, 117, 118, 119, 122, 126, 131, 162, 163, 164, 170, 219, 225, 244, 246, 298, 299, 304, 307, 314, 328, 329, 332, 333, 334, 336, 337, 338, 339, 342, 418, 421, 422, 424, 425, 429, 432, 443, 446, 451, 473, 475, 479, 483 — Continental Germanic 485 — East Germanic 40 — Ingvaeonic 446, 451 — North Germanic 40, 335, 475, 476, 483, 485 — Northwest Germanic 301, 302 — Pre-Germanic 299 — Proto-Germanic 81, 110, 170, 298, 299, 300, 301, 308, 314, 484 — West Germanic 40, 120, 124, 164, 335, 483 Gothic 118, 164, 300, 301, 302, 314, 329, 335, 337, 338, 339, 342, 424, 429 — Pre-Gothic 302 Greek 3, 19, 46, 49, 50, 52, 54, 58, 60, 61, 75, 116, 118, 119, 123, 142, 161, 162, 219, 264, 286, 317, 355, 369, 375, 391, 395, 415, 416, 417, 423, 424, 427, 431, 443, 446, 448, 449, 451, 452, 454, 461, 486 — Aeolic Greek 444 — Attic Greek 123 — Cappadocian Greek 461 — Homeric Greek 41, 423 — New Testament Greek 41 — Modern Greek 41, 43, 124, 430 — Mycenean Greek 148 Gullah 277, 515

Hattic 145 Hausa 510

Hebrew 3, 28, 50, 62, 192, 411, 456, 464 - Biblical Hebrew 415 — Hebrew-Aramaic 460 Hellenic 40, 148 Hindi 28, 510 Hittite 20, 41, 46, 142, 148, 355, 424, 425, 430, 451 Huastek 50 Hungarian 48, 50, 61, 192, 258, 486 Hurrian-Urartean 145

Icelandic 219 - Old Icelandic 129 Illyrian 452 Indian 411 Indie 40, 49, 61 - Early Middle Indie 41 Indo-Aryan 4, 142, 283, 417, 430, 449 Indo-European 20, 40, 41, 43, 54, 59, 60, 61, 81, 123, 126, 141, 148, 162, 163, 164, 170, 218, 227, 230, 330, 334, 337, 342, 354, 367, 442, 443, 446, 447, 448, 449, 451, 452, 453, 462, 473, 476 - Pre-Indo-European 342 - Proto-Indo-European 18, 21, 60, 61, 132, 298, 299, 342, 430 Indo-Hittite 42, 60, 453 Indo-Iranian 125, 148, 422, 430, 446 - Iranian 125, 286, 424 Inuit Trade Jargon 509 Irish 50, 61, 76, 87, 111, 129, 258, 334, 476 - Middle Irish 50, 422 - Old Irish 60, 110, 334 Italian 50, 122, 128, 129, 161, 192, 258, 263, 264, 435, 441, 486 - American Italian 284 Italic 40, 142, 417, 422, 431, 443, 447, 453 - Proto-Italic 142 Italo-Celtic 142

Japanese 57, 62, 63, 64, 192, 368, 411

Language Index Kaluli 192 Kannada 76 Kashubian 476 KiSwahili 510 Konkani 76 Korean 55 Krio 515 Kwakiutl 60

Lapp - North Lapp 318 Latin 6, 19, 25, 40, 47, 49, 51, 52, 54, 58, 59, 60, 101, 107, 108, 109, 110, 112, 118, 119, 122, 123, 128, 129, 131, 141, 142, 161, 162, 164, 252, 304, 317, 342, 347, 371, 372, 375, 391, 403, 425, 426, 429, 441, 444, 446, 449, 484, 485 — Late Latin 43 — Medieval Latin 102 - Neo-Latin 410 — Provincial Latin 429 - Vulgar Latin 52, 128, 441, 473 Latvian 111, 425, 476 Lettish 46 Letzebuergersch 80 Lingua Franca 509 Lithuanian 127, 258, 422, 425, 476 Livian 476 Lydian 42

Macedonian 142, 264, 265, 485 Makran Baluchi 286 Malagasy 448 Malay 50 Malayo-Polynesian 40 Maltese 484 Manjaw 448 Mazateco 46 Mexican 50 Moxo 50 Mycenean 40, 448 — Linear Β 446

551

Nahautl 99 Navaho 60 Neo-Melanesian 219 Non-Indo-European languages 475 Nootka 60 Nordlinn 509 Norwegian 258, 263, 290, 308, 463, 476 - Bokmal 290, 463 — Dano-Norwegian 463 — Nynorsk 463 - Riskmäl 463

Old Church Slavonic 40 Old Icelandic 50, 51 Old Norse 47, 48, 244, 302, 337, 425, 432 Old Persian 40, 41 Old Saxon 63, 103, 338, 424, 433 — West Saxon 166 Old Slovenian 47 Osage 464 Osco-Umbrian 40, 141, 142, 426, 432, 444 Ossetic 422, 424

Pali 376 Papiamentu 460, 466, 514 Parthian 42 Persian 41, 42, 48, 422 Phoenician 42, 46 Phrygian 42 Polish 192, 258, 349, 435 Polynesian 460 Prakrits 376 Prussian 425, 447 Portuguese 410, 460 — American Portuguese 287 — Angolan Portuguese 509 — Brazilian Portuguese 258, 509 — Creole Portuguese 516 Quechua 100

Reunion Creole 81 Rhaetoromansch 74, 429, 456, 460, 487

552

Language Index

Romance languages 41, 42, 43, 81, 108, 113, 128, 129, 141, 161, 219, 220, 371, 378, 416, 417, 435, 442, 446, 451, 473, 475, 483, 486 - Proto-Romance 43, 52, 128, 129, 450 Romanian 48, 128, 142, 258, 264, 265, 486 — Meglenite Romanian 47 Romontsch 460 Runic 46, 300 Russian 46, 48, 265, 283, 316, 320, 348, 409, 422 - Old Russian 485 — Russian Church Slavonic 429

Saek 61 Salish - Proto-Salish 373 Sami 485 Samoan 192 Sanskrit 41, 43, 49, 58, 59, 127, 162, 355, 376 Saurashtri 289 Scandinavian 78, 163, 164, 411, 485 - East Scandinavian 301 Scythian 42, 429 Semitic 40, 41, 367, 373, 458 - North Semitic 145 - West Semitic 46 Semito-Hamitic 446 Serbo-Croatian 61, 142, 192, 435 Siamese 61 Sinhalese 377 Slavic 40, 41, 46, 57, 61, 162, 164, 316, 320, 348, 419, 422, 434, 435, 443, 447, 451, 453, 473, 475, 483, 484, 485, 486 - East Slavic 476, 485 - Old Slavic 425 - West Slavic 485 Slovak 61 Slovenian 485 South Asian languages 283 Spanish 47, 50, 74, 78, 99, 128, 258, 263, 287, 316, 317, 319, 409, 476 - Castilian Spanish 64 - Florida Spanish 287

— Latin American Spanish 64 — Mexican Spanish 509 — Old Castilian 61 — South American Spanish 258 — "Spanglish" 509 — Sranan Spanish 509 — Texas Spanish 74, 76 Sumerian 40, 429 Swahili 47, 80 Swedish 47, 258, 263, 308, 476, 485, 486 — Old Swedish 425

Tagalog 50 Tai 61 Tamil 28 Tartarian languages 476 Thracian 429 Tocharian 40, 61, 424, 430, 435, 451 Tok Pisin 81 Totonak 50 Tunica 63 Turkish 48, 192, 282, 461, 486

Ukrainian 258 Uralic 142 Urdu 28 Uto-Aztec 373

Vedic 3, 421, 424 Venetic 452 Vietnamese 410 Volapük 509 Votian 476 Votyak 55

Welsh 50, 114, 119, 258, 319, 421, 476 - Middle Welsh 112 - Old Welsh 110, 112 Wolof 76

Yiddish 46, 287, 351, 409, 411, 460 - Western Yiddish 510 - "Yinglish" 509 Yuman 373 Yrik 510

Author Index

Abdulaziz Mikilifi, Μ. H. 80 Abraham, Werner 29 Abramson, Arthur 184 Adam, Lucien 512 Adler, Max 511 yElfric 50 jEöelbald 131 Aizetmüller, Rudolf 434 Aleong, Stanley 411 Alfred 41 Algeo, John 399, 405, 409, 411 Allen, Harold 82 Allen, William Sidney 118, 122, 128, 303, 304 Almastae, Jon 78 Amos, Ashley Crandell 112 Andersen, Henning 255, 314, 341, 465 Anderson, James M. 30, 303 Anderson, Stephen E. 341 Anderson, Stephen R. 124, 125, 316 Anshen, Frank 27, 74, 86, 466 Anthony, David W. 146 Antilla, Raimo 30, 44, 61, 341, 351, 374, 405, 410, 465 Antonsen, Elmer H. 125, 126, 297, 298, 300, 301, 307 Apollo Claudius 107 Apollonius Dyskolos 49, 53 Aristotle 49, 456 Arlotto, Anthony 47, 48, 405, 410 Arndt, Walter W. 219 Ascoli, G. I. 228 Ash, Roberta 39 Ashby, William J. 29 Aslin, Richard N. 186 Atwood, E. Bagby 76 Autels, Guillaumes des 131

Bach, Emmon 316, 318 Bachellery, E. 432

Bailey, Charles-James 25, 26, 27, 81, 458, 465 Bailey, Guy 82 Baker, Linda 183 Baldi, Philip 40, 347, 354 Baldinger, Kurt 436 Bammesberger, Alfred 433 Barnhart, Clarence L. 409 Barnhart, Robert K. 409 Barrit, C.W. 99, 302, 305 Bartoli, Matteo 74, 84, 140 Bartsch, Renate 62 Battison, Robbin 412 Baugh, John 27, 111, 273, 276, 277, 278, 290 Baumann, Richard 289 Bebout, Linda 74 Bee, Pierre 465 Becker, Curtus A. 176 Beeler, Μ. 142 Behagel, Otto 63 Belasco, Simon 29 Bellugi, Ursula 188 Bender, Marvin 219 Benediktsson, Hreinn 129, 302 Benjamin, Steven M. 411 Bennett, William 126 Benveniste, Emile 63, 144, 401 Benware, Wilbur 228 Bergsland, Knut 219, 220 Berlin, Brent 402 Berman, Ruth 192 Bernerker, Erich 419 Berr, Samuel 433 Bertrand-Bocande, E. 511, 512 Best, Karl-Heinz 234, 235 Bever, Thomas 180 Bezzenberger, Adalbert 416 Bickerton, Derek 27, 29, 78, 79, 80, 81, 276, 460, 465, 466, 510, 512, 513

554

Author Index

Bigler, Kurt 182 Bjerke, Andre 463 Blackall, Eric Α. 463 Blank, Michelle A. 177, 185 Bliss, A. J. 411 Bloch, Bernard 25, 245 Bloch, Ο. 420 Blom, Jan-Petter 290 Bloomfield, Leonard 40, 42, 43, 47, 49, 75, 164, 242, 330, 366, 400, 405, 410, 512 Bock, Michael J. 179, 183 Bohannon, John N. 191 Boisacq, Emile 417 Bokamba, Eyamba 30 Bonfante, Giuliano 74, 84, 141 Bonner, Anthony 131 Bopp, Franz 223, 244 Bortoni-Ricardo, Stella Maris 73 Boshoff, S. P. E. 433 Botha, Rudolf P. 37, 38 Bowerman, Melissa 188 Bransford, Joan 182 Braune, Wilhelm 54, 225 Breal, Michel 399 Bregman, Albert S. 190 Bricker, Diane 187 Bright, William 77 Britto, Francis 80 Brockelmann, Carl 367 Brosman, Paul W. 129 Brown, Gillian 181 Brown, Roger 187, 188 Brückner, Aleksander 419, 435 Brugmann, Karl 40, 103, 223, 225, 227, 228, 230, 231, 232, 233, 366, 367, 368, 369, 373, 379, 443 Bruner, Jerome 466 Buchholz, Peter 425 Buck, Carl Darling 143, 354, 430 Burks, Arthur W. 38 Burn, John 53 Burrow, Thomas 49 Bybee, Joan L. 342 Bynon, Theodora 63, 410 Byrne, Frank 512

Cable, Thomas 97, 111 Cairns, Helen S. 179

Calder, George 50 Callary, Robert 72, 170 Camden 111 Cameron, Angus 101 Campbell, Lyle 87, 219 Cannon, Garland 74, 411 Capella, Martianus 54 Carrell, Thomas D. 185 Casagrande, Joseph B. 287 Caskey-Sirsons, Leigh A. 30 Cassidy, Frederic 512 Catford, Ian C. 319 Caxton 111, 130 Cedergren, Henrietta J. 27, 29, 275 Cercignani, Fausto 117, 299 Chafe, Wallace L. 319 Chaika, Elaine 29 Chambers, J. K. 27, 30, 81, 82, 83, 84, 85, 171, 465 Chantraine, Pierre 427, 431 Chaucer, Geoffrey 42, 161 Chen, Chin-Chuan 319 Chene, B. de 317 Cherubim, Dieter 235 Chevillard, Andre 274, 511 Chifflet 122 Childe, V. G o r d o n 146 Chomsky, N o a m 27, 30, 114, 115, 179, 188, 190, 234, 250, 279, 313, 314, 316, 317, 331, 459, 464 Chretien, C. Douglas 218 Chrysippos 49 Cicero 107, 108 Clark, Eve 189 Clark, Herbert H. 180 Clarkson, Iain 78 Clemoes, Peter 102 Clough, James C. 511 Clyne, Michael 78 Coelho, F. A. 511, 512 Cohen, Antonie 180 Cohen, Marshall M. 184 Cohen, Morris R. 37, 38 Cohen, Paul 274 Collinge, Ν. E. 308 Collins, Anne M. 176 Comhaire-Sylvan, Suzanne 512 Cooper 114

Author Index Cooper, William E. 28, 185 Copely, James 405 Corominas, Joan 436 Coseriu, Eugenio 218, 219 Costello, John R. 87 Coteanu, Jon 282 Courtenay, Baudouin de 341 Cowgill, William 41 Crain, Stephen 179 Craven, Robert R. 29, 76 Cretien, Michel 411 Crocker, Christopher 29 Crocket, Henry J. 167 Cromer, Richard 187, 189 Crosby, Faye 27 Cross, Ephraim 219 Curtius, Georg 229, 230, 416 Cutler, Anne 176, 178, 187 Cyril and Methodius 41 Cziko, Gary A. 79

Daines 114 Dal, Ingerid 307 D'Alquen, Richard 302 Dante 50, 161 Darius I 41 Darwin, Charles 442, 461 Das Gupta, Jyotirindra 30 Daunt, Majorie 110, 111 Dauzat, Albert 437 Davies, Mark 191 Davis, Carol 177 Davis, Lawrence M. 411 Davis, Monte 466 Davy, Adam 171 DeCamp, David 512 Delattre, Pierre 184 Delbrück, Berthold 224, 225, 227, 228, 229, 230, 231, 233, 234, 365, 373, 377 Dell, Gary S. 180 Democritus 415 Derrida, Jacques 462 Devlamminck, Bernard 431, 432 Devoto, Giacomo 40, 142, 143 Diebold, Richard 219, 281 Dietrich, Wolf 41

Diez, Friedrich 416 Dillard, J. R. 76 Dillon, George L. 404 Dines, E. R. 73 Dinnsen, D. A. 341 Dionysius Thrax 49 Dittmar, Norbert 73 D'jakonov, I. M. 145, 146 Dobson, E. J. 114, 115, 116 Domingue, Nicole 81, 527 Dorian, Nancy 73 Dorman, Michael F. 184 Dowling, Linda 103 Dressier, Wolfgang 122, 326, 332, 339 Drinka, Bridget 30, 107 Dror, Malcah 405 Dryden, John 42 Dumezil, Georges 151 Duns Scotus 49 Dyen, Isidore 218, 219, 220

Eakins, B. 74 Eakins, G. 74 Ebel, Η. 443 Eberhard of Bethune 49 Eckelsamer, Valentin 113 Ederyn the Golden-Tongued 50 Edwards, John 87 Egger, Emile 53 Egle, Beatrice Mendez 411 Eimas, Peter D. 186 Elcock, W. D. 128, 130, 131 Elias-Olivares, Lucia 78 Ellis, Alexander J. 163 Ellis, Andrew W. 180 Elphinston 113 Emeneau, Murray 282, 285, 286 Endzelin, J. 419 Entenman, George 315 Entwhistle, W. J. 141 Erbaugh, Mary 192 Erdösi 50 Emout, Alfred 417, 427, 431 Esau, Helmut 62 Esper, Erwin 228, 231, 234 Estienne, H. 131

555

556

Author Index

Fairbanks, Gordon 354 Falk, Hjalmar 416, 418 Fant, Gunnar 56, 61, 299 Farrell, Ronald A. 29 Fasold, Ralph 27, 277 Fay, David 176, 188 Feagin, Crawford 76 Fick, August 416 Feist, Sigmund 81, 138, 418, 432 Feldman, Carol 466 Felix, Sascha 190 Ferguson, Charles 27, 28, 30, 79, 187, 188, 191, 393, 511, 528 Feyerabend, Paul 358 Finkenstaedt, Thomas 411 Fischer, Franz 151 Fisher, John H. 103, 166 Fishman, Joshua A. 77, 79, 528, 529 Flores d'Arcais, Giovanni B. 178, 189 Florus 52 Fodor, Janet D. 179, 404 Forbach, Gary B. 182 Ford, Marilyn 179 Forsaidh, Fenius 50 Forster, Kenneth 177, 182 Förster, Max 110, 129 Foss, Donald J. 177, 185 Foster, Susan 189 Fouche, Pierre 117, 121, 122, 131 Fought, John 285 Fourquet, Jean 118, 342 Fowler, Carol A. 186 Fraenkel, Ernst 434 Franck, Johannes 419 Frank, Francine 74, 86 Franks, Jeffrey J. 182 Frazier, Lyn 179 Friedrich, Johannes 46 Friedrich, Paul 145 Fries, Charles 378, 410 Frisk, Hjalmar 54, 431 Fromkin, Victoria 180 Fromm, Hans 235 Fujisaki, Hiroya 184 Furrow, David 187, 191

Galton, Herbert 308 Gamillscheg, Ernst 437 Gamkrelidze, Thomas V. 132, 145, 146, 148, 422, 430 Ganz, Peter 235 Garcia, Erica C. 86 Garrett, Merrill 178 Geipel, John 411 Gelb, Ignace 46 Gell-Mann, Murray 400 Georgiev, V. 435 Gil 114 Giles, Howard 86 Gillieron, Jules 165 Givon, Talmy 48, 377, 461 Glass, Arnold L. 182 Gleason, H . A . 218 Gleitman, Lila 191 Gloy, Klaus 75 Glucksberg, Sam 182 Goldsmith, John A. 317 Gombocz, Zoltän 420 Goodman, Morris 511, 513 Görlach, Manfred 81, 87, 464 Gould, Stephen J. 461 Graesser, Arthur 181 Grassmann, Hermann 224 Gray, Louis H. 40, 41, 42, 49, 50 Greenberg, Joseph 41, 56, 57, 58, 63, 297, 371, 447 Greenough, James Bradstreet 405 Gregory of Tours 129 Greppin, John A. C. 435 Gimbutas, Marija 144, 145 Grimm, Jacob 97, 162, 163, 170, 224, 228, 231, 242, 379 Grosjean, Francois 180 Gudschinsky, Sarah 218, 220 Guiter, Henri 218 Gumperz, John J. 27, 30, 79, 290, 527, 530 Guyonvarc'h, Christian 432 Gysseling, Maurits 129

Gal, Susan 73 Galinsky, Hans 411

Haggard, Mark P. 184 Hahn, Victor 137

Author Index

557

Hall, Robert 43, 46, 51, 52, 84, 128, 129, 141, 219, 512 Halle, Morris 27, 56, 61, 114, 115, 250, 274, 299, 313, 314, 315, 316, 317, 331 Hancock, Ian 29, 507, 510, 511, 512, 513 Harms, Robert 273, 276, 279, 313, 316, 318, 319 Harris, Harold J. 191 Harris, James W. 316 Harris, Roy 464 Harris, Zellig 317, 374 Harshman, Richard 184 Hart 114, 115 Hart, John 51 Hatcher, Anna Granville 410 Haudry, Jean 152 Haugen, Einar 44, 51, 78, 129, 285, 287, 302, 409, 410, 463 Havranek, Bohuslav 282, 283 Havers, Wilhelm 365, 378 Hawkins, R. E. 411 Healy, Alan 218 Heath, Jeffrey 287, 288 Hedberg, N. L. 190 Heine, Bernd 81 Heller, Karin 78 Heller, Louis 405 Hellquist, Elof 418 Hellstrom, Robert W. 411 Henley, Nancy M. 86 Henry, Victor 418 Heraclitus 415 Herodian 49 Herskovits, Frances 512 Herskovits, Melville 512 Herzog, Marvin 26, 30, 255, 274, 372,

Hirt, Herman 138, 146, 380 Hjelmselv, Louis 512 Hoad, T. F. 433 Hoagabaum, Thomas W. 176 Hobbs, Jerry 180 Hock, Hans Heinrich 7, 127, 283 Hockett, Charles 28, 55, 56, 99, 297, 303, 307, 331, 410 Hodges 117, 130 Hoenigswald, Henry 54, 58, 63, 123, 298, 360, 375, 441 Hofmann, J. B. 417, 420 Hogg, Richard M. 252 Hoijer, Harry 218 Holden, Cyril 284, 320 Holmes, Virginia M. 178 Holthausen, Ferdinand 418, 420, 432 Holub, Josef 419 Holyeak, Keith J. 182 Hooper, Joan B. 252, 332, 342 Hopf, Maria 148 Hopper, Paul 132 Horn, Paul 417 Householder, Fred W. 341 Howard, Philip 405 Howell, Robert 87 Hübschmann, Heinrich 417, 419 Hudson, Richard A. 27, 30, 75, 465 Huggins, Andrew W. 185 Huld, Martin E. 435 Humboldt, Wilhelm von 458 Humez, Alexander 405 Hupsek, M. 276 Hyman, Larry 48

377, 465 Hesseling, Dirk C. 511, 513 Hesychios 41 Hickerson, Nancy P. 30 Hickman, Maya 189 Hiersche, Rolf 434 Hill, Archibald 241 Hill, Kenneth C. 465 Hines, Carol 82 Hinton, Leanne 30 Hinz, Η. 425

Ickelsamer, Valentin 52 Ingria, Robert 317 Irvine, Judith 76 Isidore of Seville 41, 54 Ivanov, I. 132, 145, 146, 148, 422, 430

Hymes, Dell 86, 219, 220, 510, 527, 528

Jaberg, Karl 171 Jackson, A. 177 Jackson, Kenneth 110, 112, 129 Jakobsen, Jakob 418

558

Author Index

Jakobson, Roman 56, 57, 61, 78, 185, 186, 283, 299, 328 James, Jennifer 29 Jankowsky, Kurt 223, 224, 228, 229 Jarvella, Robert 178 Jeffers, Robert 43, 46, 48, 58, 61, 308 Jespersen, Otto 51, 87, 97, 354, 463, 511, 512 Jolly, Julius 231 Jones, Sir William 162 Jones, William J. 99, 303, 304 Jongen, Rene 78 Johnston, James C. 192 Jonz, Jon G. 29 Joseph 50 Jucquois, Guy 431, 432 Jud, Jakob 171 Jusczyk, Peter A. 186 Justus, Carol 372

Kahane, Henry 39, 75 Kahane, Rene 39 Kahn, David 318 Kaliprasanna Sinha 41 Karmillof-Smith, Annette 189, 190 Karsten, Gustaf 226, 228, 233, 234 Kay, Paul 29, 276, 402 Kaye, Alan S. 81 Keenan, Janice M. 183 Kelkar, Ashok R. 284 Kelso, J. A. Scott 186 Kerswill, Paul E. 86 Kesselring, Wilhelm 117, 129 Kettunen, Lauri 315 Key, Marie Ritchie 86 Keyser, Samuel Jay 315, 316 Kiernan, Kevin S. 102 Kilbury, James 326 Kilian, Lothar 152 Killion, Τ. Η. 176 King, Robert 5, 30, 71, 98, 249, 251, 252, 253, 282, 303, 305, 314, 351 Kiparsky, Paul 71, 125, 252, 253, 315, 317, 320, 332, 349 Kisseberth, Charles W. 316 Kittredge, George Lyman 405 Klaeber, Frederick 101

Klatt, Dennis H. 184, 185 Klausenburger, J. 341 Klein, Ernest 425, 433 Klein, Flora 29, 290 Klima, Edward 188 Kloeke, G. G. 75, 87 Kluge, Friedrich 225, 418, 424 Knowles, Jane 80 Koerner, Konrad 235 Kökeritz, Helge 116, 117 Kolere, Paul A. 182 Kopecny, Frantisek 419 Korhammer, Michael 101 Kortlandt, F. Η. H. 380 Koschmieder, Erwin 365, 374 Koutsoudas, Andreas 58, 64, 253 Krahe, Hans 145 Kristal, Andres M. 29 Kristensson, Gillis 304 Kroch, Anthony S. 27 Kronasser, Heinz 431 Kroskrity, Paul V. 29 Kruszewski, 341 Kuhn, Adalbert 137 Kuhn, Sherman M. 99, 111 Kuiper, F. B. J. 283 Kukenheim, L. 50 Kurath, Hans 167, 170, 171, 304 Kurylowicz, Jerzy 6, 55, 58, 301, 347 Kuvelsky, William 74 Kyes, Robert L. 110, 303

Labov, William 7, 25, 26, 27, 28, 72, 73, 75, 86, 115, 116, 130, 245, 246, 251, 252, 255, 273, 275, 276, 277, 278, 279, 317, 372, 377, 457, 459, 460, 465 Ladd, D. 178 Ladefoged, Peter 184, 220 Lado, Robert 283 Laferriere, Martha 30, 73 Lakoff, Robin 27, 73, 354 Lambert, P.-Y. 432 Lass, Roger 25, 30, 114, 115 Lavandera, Beatriz 72, 73 Laviosa-Zambotti, Pia 143, 148 Leben, W. R. 341

29, 30, 167, 274, 320,

Author Index Leech, Geoffrey 404 Lees, Robert B. 218 Lehiste, Ilse 43, 46, 48, 58, 61 Lehmann, Winfrid 5, 8, 40, 42, 43, 46, 47, 48, 49, 51, 52, 60, 62, 78, 81, 86, 126, 274, 297, 299, 348, 365, 401, 402, 405, 410, 421, 432 Lehrer, Adrienne 29 Leibniz, Gottfried Wilhelm 393 Leis, Ernst 411 Lenz, Rodolfo 511, 512 LePage, Robert 73 Leskien, August 223, 225, 231, 233 Leumann, Julius 417 Leumann, Manu 128, 130 Levelt, William J. M. 180 Levine, Lewis 167 Lewis, C. S. 405 Lewis, John 274 Lewy, Ernst 78 Li, Charles N. 43 Liberman, Mark 317 Lichtveld, Lou 512 Lieb, Hans-Heinrich 57 Lightfoot, David 62 Lightner, Theodore M. 314, 316 Lindblom, Björn Ε. 186 Lindsay, Peter H. 184 Linn, Michael 82 Lipka, Leonhard 361 Lipski, John M. 288 Lloyd, A. 421, 434 Lommatzch, E. 128 Longacre, Robert E. 219 Longfellow 168 Lottner, Carl 224 Lowman, Guy 166,168, 169 Luke 41 Lyons, John 297, 401, 402, 403, 404, 405

MacBain, Alexander 418 MacKay, David G. 180 MacNamara, John 76 MacNeilage, Peter 184 MacWhinney, Brian 192 Makaev, E. A. 297 Malkiel, Yakov 57, 420, 426

559

Malmberg, Bertil 50 Manczak, W. 349 Manessy-Guitton, Jacqueline 423 Maniet, A. 107, 109, 128 Mann, Stuart A. 430 Maratsos, Michael 189 Marban 50 Marchand, Hans 359 Marchand, James 300 Markey, Thomas L. 455, 461, 464, 465 Maroldt, Karl 81 Marslen-Wilson, William 177, 179 Martinet, Andre 56, 59, 61, 129, 297, 307, 308, 330, 331 Mayerthaler, Willi 458, 465 Mazarin, Cardinal 121 Masica, Colin P. 283, 286 Masing, Ferdinand 225, 233, 234 Massaro, Dominic W. 184 Maurer, David W. 29 Mayrhofer, Manfred 430 McClelland, James L. 178 McCloskey, Michael 182 McClure, Erica 288 McCormick, Susan 325, 334 McDaniel, C. K. 276 McDavid, Raven 84, 161, 167, 168, 170 Mcintosh, Angus 100 Medawar, Peter B. 455 Meid, Wolfgang 78 Meigret, Louis 51 Meillet, Antoine 54, 107, 128, 141, 284, 367, 370, 379, 417, 421, 427, 431 Melich, Janos 420 Meriggi, Piero 235 Merwe, Nikolas J. van der 218 Meyer, Gustav 419 Meyer, Paul 511 Meyer-Lübke, Wilhelm 128, 417 Midgett, Douglas 30 Miklosich, Franz 419 Miller, C. 74 Milroy, James 77 Milroy, Lesley 77, 465 Misteli, Franz 227 Mitchell, Bruce 102 Moeser, Shannon D. 190 Mohanan, K. P. 332

560

Author Index

Moliere, Jean Baptiste 513 Montgomery, Michael 82 Moore, Barbara J. R. 74 Moore, Janet R. 466 Morgan, James L. 190 Morgenstierne, Georg 417 Moravcsik, Edith 78 Morris, Kenneth M. 400 Morrison, Philip 466 Morse, William R. 29 Morton, Jacqueline 177 Morton, John 176 Moulton, Charles 168, 297, 298, 302, 307, 308, 335 Mühlenbach, J. 419 Mühlhäusler, Peter 79, 80, 81, 87, 419, 458, 460 Müller, Fr. 417 Muysken, Peter 79

Nadkarni, M. C. 76, 286 Nagel, Ernest 37, 38 Nagy, Gregory 379 Napoleon 165 Naro, Anthony J. 29 Neide, Peter H. 78 Neu, Erich 40, 431 Neumann, Günther 78 Newman, Stanley 287, 288, 289 Newport, Elissa L. 190 Newton, Brian 284 Nienaber, G. S. 433 Nilsson, Lars-Göran 181 Noll, Craig 57, 64, 253 Nonnus 116 Nooteboom, Sibout G. 180 Norman, Donald A. 184 Norris, Dennis 177, 178 Notger Labeo 54 Nunberg, Geoffrey 30 Nurse, Derek 219 Nyquist, Linda 27

Obusek, C . J . 185 O'Cain, Raymond 171 Offa 131

Ohala, John J. 255, 319 Ohmann, Richard 73 Okeju, John 287, 288 Olbrei, I. 182 O'Neil, Wayne 98 Onions, C.T. 433 Orr, John 404 Orton, Harold 163, 166, 167, 169, 171 Osgood, Charles 179 Osthoff, Hermann 103, 223, 225, 227, 229, 230, 231, 233, 234

Palef, Sandra R. 182 Palmer, Frank R. 404 Palmer, Leonard R. 123, 128, 142, 405 Pandit, P. B. 286, 289 Pänini 49 Pascual, Jose A. 436 Patanjali 49 Patella, Victoria 74 Paul, Hermann 78, 97, 223, 225, 226, 227, 228, 230, 231, 233, 234, 235, 303, 304, 305, 367 Payne, Arvilla C. 28, 30, 289 Pedersen, Holgar 50, 244 Pedro de Alcala 50 Pei, Mario 508 Penzl, Herbert 41, 43, 44, 47, 48, 52, 59, 98, 100, 109, 119, 131, 297, 298, 301, 302, 303, 304, 308 Perelman, Eliezer 455 Perfetti, Charles A. 177 Perkell, Joseph S. 185 Petraeus 50 Pfaff, Carol 457 Pfister, Max 436 Philippson, Gerard 220 Piaget, Jean 187 Pickford, Glenna R. 167 Pictet, A. 416, 443 Pierce, Charles Sanders 465 Pierce, Joe E. 219 Piirainen, lipo Tapani 303, 304 Pike, Kenneth L. 410 Pinker, Steven 190 Pisani, Vittore 78, 141, 425, 426 Pisoni, David B. 184

Author Index

561

Pitts, Ann 87 Plato 49, 415 Pokomy, Julius 54, 422, 423, 430 Polome, Edgar C. 3, 29, 43, 52, 81, 87, 137, 415 Pope, Alexander 42 Poplack, Shana 29 Porzig, Walter 365 Postal, Paul 315, 318 Pott, August Friedrich 229, 416 Poyen-Bellisle, Rene de 511 Preece, A. 190 Prellwitz 417 Preobrazenskij, Antonin 419 Priestley, Joseph 455 Prince, Alan 317 Priscian 49 Probus, Valerius 52, 113 Prokosch, Eduard 376 Puhvel, Jaan 152, 431 Pulgram, Ernst 6, 112, 130, 142, 236 Putnam, Hilary 176 Pye, Charles 191 Pyles, Thomas 405, 411

Reis, Marga 235 Remigius Autissiodorensis 54 Renfrew, Colin 7, 148 Repp, Bruno Η . 1 8 4 Rex, Richard 400 Rickford, John 87 Roberge, Paul T. 333, 341 Roberson, Susan 74 Robins, Clarence 274 Robinson 115, 130 Robinson, Fred C. 101 Rocencvejg, V. J. 286 Rohlfs, Gerhard 52 Romaine, Suzanne 27, 72, 86, 87, 276 Rosch, Eleanor 176 Ross, A. S. C. 420, 426 Rousseau, Pascale 279 Royce, Josiah 464 Rubin, J. 79 Ruhlen, Merrit 40 Runge, Richard 87 Russ, Charles 130, 303 Ruyll 50 Ryan, Ellen Bouchard 86

Quilian, M. R. 176 Quintilian 108 Quirk, Randolf 62, 99, 111

Sadnik, Linda 434 Sa'id, Majed F. 284 Salmons, Joseph 71 Samuels, M. L. 39, 100 Sanders, Gerald 57, 64, 253 Sanderson, Steward 163 Sanfeld, Kristian 282, 283, 286, 373 Sankoff, David 26, 27, 29, 72, 218, 275, 276, 279 Santee, Jeffrey 182 Sapir, Edward 29, 56, 57, 60, 63, 78, 170, 284, 331 Saunders, P. Τ. 466 Saussure, Ferdinand de 4, 5, 6, 7, 21, 25, 26, 57, 60, 235, 318, 341, 347, 379, 459, 465 Sayer, Edgar 508, 511 Scarborough, Don L. 176 Schach, Paul 78 Schane, Sanford A. 243, 316, 332 Schaller, Helmut V. 282, 283, 286 Scherer, Anton 138

Ramanujan, A. K. 27, 283 Ramat, Paolo 126 Rao, G. Subba 411 Raphael, Lawrance J. 184 Rask, Rasmus 137, 223, 242, 243 Ray 114 Rayfield, Joan R. 288 Rayner, Keith 179 Rauch, Irmengard 37, 59, 62, 63 Raumer, Rudolf von 228, 229 Rea, John A. 217, 218, 219, 220 Recatas, B. 282 Reder, Lynne M. 182 Reed, David W. 283 Reichelt, Hans 132 Reichardt, Konstantin 423 Reinecke, John 466, 509, 513

562

Author Index

Scherer, Klaus 86 Scherer, Wilhelm 230, 231 Schiefelbusch, Richard 187 Schiffman, Harold 286 Schindler, Jochem 356 Schleicher, August 3, 163, 228, 230, 231, 236, 244, 416, 442, 450, 451 Schlerath, Bernfried 431 Schmalstieg, William R. 111, 347 Schmidt, Johannes 7, 163, 230, 234, 236, 443, 458 Schmidt, Karl-Horst 151 Schmidt, P.W. 368, 371, 372, 378 Schneidemesser, Luanne von 411 Schönfeld, M. 118, 119 Schräder, Otto 138 Schrodt, Richard 118 Schuchardt, Hugo 78, 234, 235, 284, 458, 509, 511, 512 Schulze, Ernst 511 Schulze, Ursula 304 Schwyzer, Eduard 374 Scott, Charles T. 42, 52 Scott, Sir Walter 394 Scotten, Carol Myers 287, 288 Searle, C. L. 185 Seebold, Elmar 421, 428, 432 Sehrt, Edward H. 433 Seiler, Hansjakob 462, 466 Sengal, C. J. 189 Serjeantson, Mary S. 117, 411 Shakespeare, William 111, 116, 117, 169 Shankweiler, Donald 185 Shapiro, Fred R. 74 Shapiro, Michael C. 286 Sharpe, William D. 54 Shattuck-Hufnagel, Stephanie 180 Shen, Yao 283 Sherzer, Joel 282, 283, 285, 286, 289 Shibatani, Masayoshi 316 Shipman, George Raymond 51 Shores, David 82 Shuy, Roger W. 27 Sievers, Eduard 225 Silva-Corvalän, Carmen 78 Simpson, Greg B. 177 Skeat, Walter W. 418 Skok, Petar 435

Sköld, Hannes 141 Slawski, Franciszek 435 Slobin, Daniel 187, 188, 190, 192 Smith, Carlota S. 175, 191 Smith, Donald L. 411 Smith, Philip M. 74, 75, 86 Smitherman, Geneva 507, 513 Snow, Catherine 191, 192 Solan, Larry 189 Sole, Yolanda 74 Solta, R. 435 Sommer, Ferdinand 107, 128 Sommerfelt, Alf 110 Southworth, Franklin 25, 27, 30, 281, 283, 286, 288 Spalding, Keith 77 Spenser, Edmund 43 Spilich, G. J. 181 Springer, O. 421, 434 St. Anselm 49 St. Augustine 393 St. Thomas 243, 393 Stampe, David 332 Stanley, Richard 316s Stanners, Robert F. 182 Steckler, Nicole A. 28 Steedman, Mark 179 Steever, Sanford E. 29 Stein, Gabriele 358 Steiner, Richard 115, 116, 130, 465 Steinmetz, Sol 409, 411 Stephenson, Edward Α. 128 Stern, Gustaf 399, 400, 401 Stevens, Kenneth, 184, 185 Stewart, William A. 27, 512 Stockwell, Robert P. 99, 114, 115, 116, 123, 124, 125, 302, 305, 314, 418 Stoel-Gammon, Carol 190 Stokes, Whitley 416 Strabo 119 Streitberg, Wilhelm 41 Stross, Brian 29 Studdert-Kennedy, Michael 183 Sturtevant, Edgar 25, 60, 74, 243, 244, 303 Suetonius 52 Summerfield, Quentin A. 184

Author Index Swadesh, Morris 39, 139, 217, 218, 219, 445 Sweet, Henry 98 Swift, Jonathan 42 Swift, K. 74 Swinney, David A. 177, 187 Szantyr, Anton 374 Szemerenyi, Oswald 61

Tabouret-Keller, Andree 73 Tacitus 51, 119, 425 Tagliavini, Carlo 511, 512, 513 Tanenhaus, Michael K. 177, 179 Tanouye, Ellen 192 Tartter, Vivien C. 186 Taylor, Douglas 512 ten Kate, Lambert 52 Thananjayarayasingham, S. 30 Thieme, Paul 139 Thom, Rene 466 Thomas, John J. 512 Thomason, Sarah G. 84 Thompson, R. Wallace 512 Thompson, Sandra 43 Thornley, G. C. 102 Thurneysen, Rudolf 417 Timm, L. A. 288 Tischler, Johann 220, 431 Titkin, Η. 437 Tolomei, Claudio 51, 52, 161 Tomkin, Elizabeth 512 Toon, Thomas E. 37, 38, 120, 131 Toporov, V. N. 434 Torp, Alf 416, 418 Townsend, David J. 180 Trabasso, Thomas 180 Träger, George 25, 46 Traugott, Elizabeth 44 Trautmann, Reinhold 419 Treitler, Leo 102 Trier, Jost 433 Trubacev, Oleg N. 422, 434, 435 Trubetzkoy, N. S. 78, 330, 331, 341 Trudgill, Peter 27, 28, 30, 78, 80, 82, 84, 85, 87, 257, 465 Tschirch, Fritz 41 Ts'ou, Benjamin K. 30

563

Tucker, Susie I. 405 Tulving, Endel 182 Turgot, Anne Robert 3 Turner, Lorenzo Dow 411 Turner, Ralph L. 417, 430 Tyler, Lorraine 179 Tyma, Stephen 465 Twaddell, W. Freeman 304, 372 Twain, Mark 161 Uhlenbeck, C. C. 417, 418 Ullman, Stephen 56, 57, 62, 63, 401 Untermann, Jürgen 152, 426, 432 Ureland, Sture 78, 471 Väänänen, Veikko 109, 118, 122, 128, 130 Valdes Fallis, Guadalupe 288 Valdman, Albert 513 Valkhoff, Marius 512 Van Coetsem, Frans 299, 325, 329, 333, 334, 335, 340, 342 Van Deyck, Rika 131 Van Eerde, John 73 Van Ginneken, J. 57 Van Loey, Adolphe 86 Van Name, Addison 511, 512 Van Wijk, N. 419 Van Windekens, Albert Joris 435 Varro 49, 108 Vasmer, Max 422, 435 Velleius Paterculus 119 Vendryes, J. 422, 432 Vennemann, Theo 57, 62, 252, 317, 320, 332, 335, 354, 369, 371, 372 Vercoullie, Jozef 419 Verner, Karl 224, 242, 298 Victorius 50 Villa Dei, Alexander de 49 Villar, F. 355 Villon, Frangois 121, 131 Vipone, Douglas 181 Visser, F. T. 53 Vogt, Hans 219, 220 Voyles, Joseph 404 Vries, Jan de 420, 425, 433 Vygotsky, Lev S. 187

564

Author Index

Wackernagel, Jakob 365, 511 Walde, Alois 142, 417, 420, 423, 431 Waldron, R. A. 405 Walker, Alastair 76 Walley, Amande D. 185 Wallis 114 Walpole, Η. 393 Walshe, Μ. O'C. 434 Wang, William S.-Y. 318, 319 Wanner, Eric 191 Warden, David A. 189 Warren, Robert E. 185 Warren-Leubecker, Amye 191 Wartburg, W. von 420, 436 Wasserzieher, Ernst 434 Watkins, Calvert 143, 372, 433 Waugh, Linda R. 328 Webb, Charlotte 316 Wechssler, Eduard 223 Weinrich, Uriel 26, 29, 30, 44, 46, 47, 75, 78, 255, 274, 275, 281, 282, 283, 284, 285, 286, 287, 288, 372, 377, 410, 465, 466, 529, 531 Weekley, Ernest 418 Wells, Rulon 97 Wenker, Georg 163, 164, 171 Werth, Paul 404 Wheeler, Benjamin Ide 233, 234 Whinnom, Keith 512, 513 Whiteley, W. H. 47 Whitney, William D. 44, 228, 231, 232, 378 Widdowson, John 163 Wiener, N. 393

Wilkins 114 Williams, Edna 404 Williamson, Robert 74 Wilson, Robert 290 Winter, Werner 11, 436 Wolfe, Patricia M. 115, 130, 315 Wolff, Dieter 411 Wolfram, Walt A. 27 Wolfson, N. 276 Wood, Richard E. 464 Woodcock, Alexander 466 Wrede, Ferdinand 171 Wright, Joseph 163, 169 Wright, Nathalia 163 Wulfila 40, 41 Wurzel, Wolfgang U. 332 Wüst, Walter 417 Wyld 52

Yaeger, Malcah 116, 130, 155, 465 Yaska 49 Yeni-Komshian, Grace H. 187 Yenne, Herbert 29 Yule, George 181

Zengel, Majorie 77 Zentella, Ana Celia 288 Zeps, Valdis 40 Zetterstrom, Rolf 186 Zgusta, Ladislav 389 Zohary, Daniel 148 Zwaenepoel, Romana 131

m

Philip Baldi (Editor)

Linguistic Change and m Reconstruction Methodology m m m

m m

m m m m m m m m m m

1990.15.5 x 23 cm. XII, 752 pages. With 5 maps. Cloth. ISBN 311011908 0 (Trends in Linguistics. Studies and Monographs 45]

This collection of papers on historical linguistics addresses specific questions relating to the ways in which language changes in individual families or groups, and to which methodologies are best suited to describe and explain those changes. In addition, the issue of 'long distance' relationships and the plausibility of recovering distant linguistic affiliations is discussed in detail. Material by specialists is presented not only from the well-documented Indo-European family, but from Afroasiatic, Altaiac, Amerindian, Australian and Austronesian languages as well. Despite many claims to the contrary, the method of comparative reconstruction, based on the regularity of sound change, is the most consistently productive means of conducting historical linguistic enquiry. Equally important is the demonstration that the comparative method has its limitations, and that linguists must be cautious in their postulation of large super-families, whose existence is based too heavily on lexical evidence, and whose scientific foundation is difficult to establish.

mouton de gruyter Berlin · New York

m

Ian Terje Faarlund

m

Toward a Theory of Historical Syntax

m

m m m

m m m m m m m m m m m

Syntactic Change 1990.15.5 X 23 cm. X, 222 pages. Cloth. ISBN 3110126516 (Trends in Linguistics. Studies and Monographs 50)

This research monograph presents a theoretical study of the causes and mechanisms of syntactic change in natural language. Historical syntax has been the focus of growing interest in recent years, and this interest has raised important issues concerning the relationship between syntactic theory and explanations of syntactic change. Based on a theoretical and metatheoretical discussion of the locus of linguistic change and the notion of explanation in historical syntax, this work provides an in-depth study of a variety of syntactic changes in Germanic languages. Those dealt with are mainly related to grammatical role and basic sentence structure such as the change from OV to VO order, the introduction of the verb-second constraint and the expletive topic in Germanic, as well as changes in the assignment of grammatical roles and semantic roles and the change from non-configurational to configurational sentence structure. The conclusions are based on data from many, including older, Germanic languages.

mouton de gruyter Berlin · New York