Variation and Convergence: Studies in Social Dialectology [Reprint 2013 ed.] 9783110851601, 9783110110456


239 42 8MB

German Pages 326 [332] Year 1988

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Introduction: Variation and Convergence as a Topic in Dialectology and Sociolinguistics
On the Interpretive Analysis of Historical Records: Linguistic Relations in Seventeenth Century Osnabrück
A Case of Convergence and its Interpretation: MHG î and û in the City Dialect of Constance
Sociophonology
Swiss German Dialects and Swiss Standard High German
Word Final /r/ in a Northern English Accent: An Interactional Account of Variable Production
Divergence and Convergence of Dialect and Standard from the Perspective of the Language Learner
Convergence, Discourse and Variation
Uncodified Code: A Look at Some Properties of the Dialects of Sicily and a Presentation of One Speaker
Conversational Microconvergences between Dialect and Language
Converging Divergence and Diverging Convergence: The Dialect-Language Conflict and Contrasting Evolutionary Trends in Modern Italy
Varieties, Variation, and Convergence in the Linguistic Repertoire of the Old Order Amish in Kent County, Delaware
Standardization Processes and Linguistic Repertoires in Africa and Europe: Some Comparative Remarks
Recommend Papers

Variation and Convergence: Studies in Social Dialectology [Reprint 2013 ed.]
 9783110851601, 9783110110456

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Variation and Convergence

Soziolinguistik und Sprachkontakt Sociolinguistics and Language Contact

Herausgegeben von / Edited by Norbert Dittmar

Band 4 / Volume 4

w DE

G Walter de Gruyter • Berlin • New York 1988

Variation anc. Convergence Studies in Social Dialectology Edited by Peter Auer and Aldo di Luzio

w DE

G Walter de Gruyter • Berlin • New York 1988

Printed on acid free paper (ageing-resistant - p H 7, neutral)

Library of Congress Cataloging-in-Publication Data

Variation and convergence : studies in social dialectology / edited by Peter Auer and Aldo di Luzio. p. cm. - (Soziolinguistik und Sprachkontakt; Bd. 4 Sociolinguistics and language contact; v. 4) Includes bibliographies. ISBN 0-89925-447-0 (U.S. : alk. paper) 1. Dialectology. 2. Language and languages-Variation. 3. Sociolinguistics. I. Auer, Peter, 1954. II. Di Luzio, Aldo. III. Series: Soziolinguistik und Sprachkontakt: Bd. 4. P367.V37 1988 87-36490 417'.2-dc 19 CIP

dP-Titelat^hahme der Deutschen Bibliothek

Variation and convergence: studies in social dialeaology / ed. by Peter Auer and Aldo di Luzio. - Berlin; New York: de Gruyter, 1988 (Sociolinguistics and language contaa; Bd. 4) ISBN 3-11-011045-8 NE: Auer, Peter [Hrsg.]; Soziolinguistik und Sprachkontakt

© Copyright 1988 by Walter de Gruyter & Co., Berlin 30. Alle Rechte, insbesondere das der Übersetzung in fremde Sprachen, vorbehalten. Ohne ausdrückliche Genehmigung des Verlages ist es auch nicht gestattet, dieses Buch oder Teile daraus auf fotomechanischem Wege (Fotokopie, Mikrokopie) zu vervielfältigen. Satz: Dörlemann-Satz GmbH & Co. KG, Lemförde Druck: Rotaprint-Druck W. Hildebrand, Berlin Bindearbeiten: Lüderitz & Bauer, Berlin

Contents PETER AUER & ALDO DI LUZIO

Introduction: Variation and Convergence as a Topic in Dialectology and Sociolinguistics

1

UTZ MAAS

On the Interpretive Analysis of Historical Records: Linguistic Relations in Seventeenth Century Osnabrück PETER AUER,

Constance

A Gase of Convergence and its Interpretation: MHG fand ü in the City Dialea of Constance SYLVIA MOOSMÜLLER,

IWAR "WERLEN,

43

Vienna

Sociophonology

75 Berne

Swiss German Dialects and Swiss Standard High German J . PETER FRENCH,

93

York

Word Final /r/ in a Northern English Accent: An Interactional Account of Variable Production O T T O STERN,

11

124

Zürich

Divergence and Convergence of Dialect and Standard from the Perspective of the Language Learner NORBERT DITTMAR & PETER SCHLOBINSKI,

Convergence, Discourse and Variation

133

Berlin 156

VI FRANK MÜLLER,

Contents

Frankfurt

Uncodified Code: A Look at Some Properties of the Dialects of Sicily and a Presentation of One Speaker ALBERTO A . SOBRERO,

Lecce

Conversational Microconvergences between Dialea and Language J O H N TRUMPER & MARTA MADDALON

194

(Cosenza)

Converging Divergence and Diverging Convergence: The Dialect-Language Conflia and Contrasting Evolutionary Trends in Modern Italy WERNER ENNINGER AND JOACHIM RAITH,

216

Essen

Varieties, Variation, and Convergence in the Linguistic Repertoire of the Old Order Amish in Kent County, Delaware . . . ALBERTO M . MIONI,

175

259

Padova

Standardization Processes and Linguistic Repertoires in Africa and Europe: Some Comparative Remarks

293

Contributors Peter Auer, FG Sprachwissenschaft, Universität, 7750 Konstanz, FRG Norbert Dittmar, FB Germanistik, Freie Universität, 1000 Berlin, FRG Werner Enniger, FB 3, GHS Essen, Postfach 103764, 4300 Essen, FRG Peter French, College of Ripon and York St. John, Lord Mayor's Walk, York, England Aldo di Luzio, FG Sprachwissenschaft, Universität, 7750 Konstanz, FRG Utz Maas, Universität Osnabrück, FB SLM, 4500 Osnabrück, FRG Marta Maddalon, Universita della Calabria, Cosenza, Italy Alberto Mioni, Dipartimento di Linguistica, Universita di Padova, 35137 Padova, Italy Sylvia Moosmüller, Institut für Sprachwissenschaft, Liechtensteinstr. 46a, 1090 Wien, Austria Frank Müller, Romanisches Seminar, Universität, Gräfstr. 72, 6000 Frankftirt,FRG Joachim Raith, FB 3, GHS Essen, Postfach 103764, 4300 Essen, FRG Peter Schlobinski, FB Germanistik, Freie Universität, 1000 Berlin, FRG Alberto Sobrero, Universita degli studi, Dipartimento di Filologia e Linguistica, 73100 Lecce, Italy Otto Stern, Deutsches Seminar, Universität Zürich, Rämistr. 74,8001 Zürich, Switzerland John Trumper, Universita della Calabria, Cosenza, Italy Iwar Werlen, Institut für Sprachwissenschaft, Gesellschaftsstr. 6, Universität Bern, 3012 Bern, Switzerland

PETER AUER & A L D O DI LUZIO (CONSTANCE)

Introduction: Variation and Convergence as a Topic in Diaiectology and Sociolinguistics As the first linguistic sub-discipline concerned with synchronic Variation, diaiectology has always remained somewhat outside the strong linguistic mainstream that elaborated, from the beginning of the Century up to the 60ies, the notions of 'system' and 'structure' as fundamental in linguistics and had little to say about lingustic diversity. Conceived as a correlate or/and a correction of diachronic reconstruction, diaiectology has also never quite gave up a historical dimension of linguistic description. As a consequence, structuralist/generative linguistics and traditional diaiectology with its 19th Century roots influenced each other relatively little in the first half of this Century. Attempts to incorporate structuralist notions such as "overall pattern" (Trager) or "diasystem" (Weinreich), or structuralist ways of explaining dialect change on systemic grounds (cf. Moulton's famous argument on the "nordostschweizerische Vokalspaltung"), came late and were only partly successful; this applies even more to generative diaiectology (Vasiliu). Regrettable as this partly independent development in the two fields may have been, it turns out today to have been a healthy safeguard against the reductionalist tendencies of a linguistics more concerned with ideal speaker-hearers and ideal languages than with empirical analyses. Labovian socio-linguistics was the first serious attempt to integrate the positive aspects of generative phonology and syntax such as systematicity and explicitness on the one hand, and empirical diaiectology on the other. Labov succeeded in overcoming the Opposition 'system' vs Variation' by incorporating Variation into lingustic description. He also showed that it was necessary for diaiectology to get rid of its selfimposed limitation to non-mobile, older, rural speakers and to turn to other sources of lingustic diversity than those on the geographical (diatopic) level. If we look upon socio-linguistics in the narrowest sense, i.e. as the study of linguistic diversity related to social class, it is true that, 20 years after Labov, interest in this field has declined. Whatever the non-linguistic

2

Peter Auer & Aldo di Luzio

reasons for this development may be, the linguistic ones are surely to be found in the unsatisfactory way in which this type of socio-linguistics dealt with the social part in the hyphenated discipline. Not only did correlational sociolinguistics fail to go beyond rather illfounded concepts of social stratification in society, it also failed to go beyond merely quantitative ways of combining the parts to the left and to the right of the hyphen. In retrospective, it seems that the very idea of amending structuralist/ generative linguistics and its notion of system by merely adding (Statistical) probabilities to rules of competence was misconceived. A more radical rethinking of the notion of system is necessary. Hints at this can be found in Gumperz' writings on dialectology and socio-linguistics. Gumperz points out that the notion of 'system' itself cannot be a theoretical "prime" in linguistics. Instead, linguistic systems ('varieties', 'codes', 'languages', 'dialects' or whatever) are the resultoi a speech community's more or less focussed (categorical) use of linguistic structures and processes. If this reconstmctive perspective is taken seriously, it means that linguistic description of any type, whether 'socio-linguistic' or not, has to take as its starting point the whole spectrum of linguistic means of a speech Community (i.e., its repertoire). It has then to ask if co-occurrence restrictions between the structures of this repertoire justify postulating one, two or more distinct linguistic systems. Questions of homogeneity are always empirical questions, and they will always have to be answered in terms of degrees. Due to their history, and to conscious political efforts, the important European languages are focussed to such an extent that it has become usual in parts of the discipline to disregard linguistic diversity for the sake of the analysis of a Standard variety in its institutionalised form. Although this way of proceeding is possible, and sometimes useful, it must not be forgotten that Standard languages are regularly not the languages of everyday linguistic praxis. A more realistic approach would therefore seem to consist in starting from the other end: i.e. from actual linguistic utterances. This means, for most speech communities, integrating 'dialea' or 'coUoquial' forms. Against such a position, Chomsky recently argued that the investigation of observable linguistic data and the Variation inherent in it (e.g. bilingual conversation), is, for a linguist, as unrewarding as the investigation of 'real' fluids, for instance, the poUuted water in a river, is for the (theoretical) chemist. In order to arrive at a high scientific Standard, the argument continues, one has to do research on the pure elements of the matter observed, for only this research can reveal how pure elements combine into complex ones. The analogy is doubtful, however. It implies that

Introduction

3

linguistic diversity is made up of, and reducible to, primary elements comparable to the natural elements, i.e. it takes 'codes' as primaiy, Variation as secondary. But a linguistic repertoire can be analysed without the prior discovery and description of 'underlying' homogeneous varieties - unlike the water of a river which may be difficult to describe without knowledge of the chemical elements; and in fact, there often are no 'underlying', 'pure' varieties at all. As a consequence, it is of theoretical, not only descriptive interest for linguistics to deal with Variation not reduced to homogeneity. The neo-biologistic view of language disregards what distinguishes language so thoroughly from the subject of the natural sciences: i.e. that the structures of a linguistic repertoire are meaningful to their users on most occasions. This meaningfulness allows us to analyse linguistic diversity, not as the uninteresting surface remnant of a clash or interference between two independent, regulär varieties, but as linguistic data in its own right. Members of the speech Community are able to establish referential moreor-less equivalences between items - such that habe, ham, habm, haben can all be seen as meaning 'to have' - , and they are also able to attribute social and conversational meaning to the uses of these variants - such as Southern German origin of the Speaker, colloquial or formal expression, etc. These meanings establish a net of relationsliips between structures which put us in a Position to analyse nonhomogenous linguistic data, although the tools we need for this purpose are more difficult to handle than those currently in use in less realistic approaches to language. Operations of equivalence used by members of the Community allow one to group the abovementioned items together and to look for (variable) phonological processes that might explain their occurrence. It is another, secondary question to see if - as happens sometimes, but not in all cases - categorical boundaries can be established that might justify varieties inside the repertoire that show less internal Variation than the repertoire as a whole. For instance, the distinction bersveen habe and the other variants {ham, habm, haben) can be proved to be of a different type, both on the basis of its social meaning and its phonological status, than the one between ham, habm and haben-, this, in turn, makes it possible to distinguish between w-apocoping and «-assimilating varieties of German. Co-occurrence restrictions between a number of such distinctions might justify positing a sub-variety of German called Alemannic, or the like. The way of proceeding is clearly opposite to that proposed by Chomsky: diversity comes flrst, focussing into standardised varieties comes second. We have argued that the study of Variation is much more at the core of linguistic research than socio-linguistics in the social-stratificational sense

4

Peter Auer & Aldo di Luzio

has realised. It is not merely a question of defining ways of speaking relative to social status or class. The notion of 'convergence' (implying, of course, its opposite, 'divergence') indicates this broader scope. 'Convergence' can be conceptualised in a number of ways. O n the most interactional level, convergence/divergence may be applied to the linguistic behaviour of two or more interactants who may or may not adapt to each other's speech behaviour. Although it is easy to deduce that these processes must be the foundation of larger scale iimovations (linguistic change), they have not been investigated much empirically. Existing studies on the face-to-face level tend to be more interested in social-psychological explanations ("assimilation") than in fme-grain interactional and grammatical analyses. For instance, it is not at all clear if divergence in face-to-face interaction is restricted to assimilation to the common Standard language (vertically, so to speak), or if it can also take place between local dialectal variants (horizontally). Neither do we know about the linguistic levels affected, the conversational loci in which the process begins or ends, and its small-scale functions. While becoming relevant in larger societal constellations and groups than the interactional dyad, structures and processes gain access to the converging speech community's repertoire. For the purpose of describing this take-over, a mediating concept has to be found between interaction (dyad) and speech Community. Often, network analysis is useful here. It is quite common that linguistic forms of a neighbouring Community (or a social class, caste, or whatever) are first taken over by members at its periphery, usually those who have networks open to both communities or groups. At later stages, the new variant will become prestigious for larger segments of the receiving Community, although it may not yet be part of the individual repertoire of all its members. The development may stop as soon as the new variants that converge with those of the neighbouring Community or group are integrated in the community's repertoire, either by eliminating old items from it, or giving them marginal status, or by redefining their meaning. Finally, total convergence would consist in abandoning the old parts of the repertoire completely. It should be obvious at this point that convergence may diminish or increase Variation in the individual's or the speech community's repertoire. Whenever this repertoire is uniform and 'focussed', convergence towards that of another Speaker or speech Community (or a variety in it) will produce additional Variation. O n the other hand, a Speaker or a speech Community that shows much Variation and litde focussing from the very Start will perhaps make use of the variety converged towards as a means of

Introduction

5

establishing a focus in the repertoire, thereby diminishing overall Variation. Veiy grossly put: stable repertoires may become less stable in convergence, loose repertoires may become less loose. Both on the macro- and on the micro-dimension, horizontal ('interdialectal') and vertical ('standard/dialect') convergence have to be distinguished. O n the macro-dimension, the possible outcomes of such developments may be koinai (Ausgleichssprachen) or Umgangssprachen. However, when distinguishing Umgangs- and Ausgleichssprachen one has to keep in mind that horizontal convergence is usually influenced by a co-existent Standard variety (if there is one), that is, it incorporates aspects of vertical convergence between the dialect(s) and the Standard as well. In a parallel fashion, vertical convergence usually diminishes differences between neighbouring dialect varieties and therefore also implies aspects of koine-iormation. This is evident from Schirmunski's classic description of the levelling tendencies between the German Settlements in Russia (Schirmunski 1930). Schirmunski showed koine formation between the Franconian and Alemannic dialects to depend mainly on two factors. First, "primary dialect features" are given up more readily than "secondary dialect features" (a distinction later worked out more precisely but also differently by Reiffenstein). Primary dialect features are "most characteristic" for a narrowly locally defined variety. When these features are given up in koine formation, this implies a convergence towards the Standard. But, secondly, Schirmunski observed that it was not always the variant closer to the Standard that 'won' over the neighbouring community's expression, that is, there was horizontal without vertical convergence. His examples suggest that such developments might depend on the phonological status of the Variation in question. Above all, it seems that his cases of purely vertical convergence are those based on natural lenition processes, whereas the primary dialect features are morphophonologically and lexically restricted. At this point phonological theory comes in and proves to be useful. Distinctions like the one made by Schirmunski can be stated more clearly when their phonetic (articulatory and perceptory) basis is taken into account. It has therefore turned out to be fruitful to apply the theoretical apparatus of Natural Phonology or another theory with a strong focus on phonetics to the study of Variation; for instance. Natural Phonology differentiates processes from rules, and this dichotomy is mirrored in the dialeaological distinction between processes and correspondence (or input Switch) rules. However, Umgangs- and Ausgleichssprachen presuppose that processes of convergence and divergence have come to a point where new, distinct varieties have emerged in the repertoire. Again, the horizon-

6

Peter Auer & Aldo di Luzio

tal and vertical processes of convergence are logically and empirically prior to such varieries. It is very well possible that convergence does not lead to the formation of intermediate varieties, as Gumperz & Wilson (1971) demonstrated in their Kupwar investigation. In this Indian village, they found Kannada, Marathi, Urdu (and Telugu) spoken side by side but clearly distinguished from the point of view of the villagers. Nevertheless, Urdu, Marathi and Kannada (as spoken in the village) had converged on the phonetic and syntactic levels to the degree of exhibiting the same syntactic surface structure, and of making morph-by-morph translation possible, although the languages are typologically quite distinct (Marathi and Urdu are Indo-European, Kannada is not). Most members of the Community had a passive competence of the three languages, Marathi was the language of active inter-group communication. We may therefore speak of a repertoire containing all the three languages as separate varieties, held apart by the Community members' efforts to keep them apart. Although the linguistic structure of the three languages had changed quite radically, the repertoire continued to have the same Organisation: Kannada, Urdu and Marathi coexisted, with no intermediate varieties having developed. Thus, convergence may lead to various outcomes. What is needed is not only its linguistic description but a typology of the repertoires possible and empirically observable in various sociolinguistic situations. One parameter for such a typology is given by the question as to whether a repertoire is parcelled out into distinct sub-varieties, or whether it orders linguistic structures on a continuum between two (or more) poles. Another parameter is the number of varieties or poles in the repertoire, a third, the existence or non-existence of intermediate varieties {Ausgleichs- or Umgangssprachen) in addition to the 'extreme' varieties, and there are others. As far as standard-dialect levelling is concerned, we are dealing with one of the most important linguistic features of modern industrial societies, characterised by high local and social mobility. Even classical dialectology could not disregard the fact that convergence towards the Standard takes place, sometimes merely changing, sometimes even erasing the traditional rural dialects. This fact makes it unrealistic to continue to work with traditional dialectological methodology, for some of its most basic assumptions are no longer true. As soon as standard-dialect levelling occurs the typical dialect Speaker is no longer a person who has only one way of expressing himself at his disposal. Instead, there is now a variety of forms in his repertoire, which leads to variable rule application. Somehow, this Variation has to be coped with by the dialectologist. On the other hand, the

Introduction

7

speakers of the speech Community do not behave uniformly any longer; therefore, interviewing one or a handful of informants per Community is no longer an adequate empirical basis for dialectological research. In addition, the notion of the isogloss, which was of central importance for dialectological cartography, is particularly ill-suited to capture noncategorical transitions, and offers no help whatsoever for representing diastratic and diaphatic (transsituational) Variation.

The contributions in this voIume We have tried to introduce a number of notions and of points of interest that constitute a new approach to Variation and convergence, such as: convergence/divergence in discourse, conversational meaning of Variation, network analysis and the spread of innovations, the formation of koinai and Umgangssprachen, repertoire typology, natural phonological theory, reconstructive Variation analysis, etc.. The contributions in this volume centre around these theoretical issues but also have a strong empirical concern with particular sociolinguistic Situation. The first two papers deal with macro-convergence in micro-analytic ways. Both are concerned with the meaning of a take-over of variants which clearly belong to another dialect area, but which are also nearer to the high, prestige variety. And in both cases, the meaning of this take-over is reconstructed on the basis of fine-grain textual analyses. Maas investigates the gradual spread of upper German forms in literary and legal documents from 17th Century Osnabrück. The transition from Low to High German in Northern Germany and the parallel formation of a national German Standard language have always been important research topics in historical linguistics of course. However, there is no methodology for the textual interpretation of Variation in written documents avaüable. Maas develops such a diachronic methodology (connotation analysis) which focusses on "inscriptions" of linguistic praxis in the historical manuscripts. Auer 6.ra.-ws his materials from quite a different context, i.e. the phonology of a present-day city vernacular (that of Constance). Whereas the quantitative part of the contribution deals with the development of certain parts of the phonological system which are ambiguous as to their interpretation (vertical or horizontal convergence), the qualitative analysis, based on intra-conversational switching and shifting, and on other functional uses of dialect features, reveals that vertical convergence plus phonological process regularities are at the heart of the development.

8

Peter Auer & Aldo di Luzio

Moosmüller discusses the relationship between dialectal Variation and phonological theory. She develops a sociophonological model which distinguishes input-switch rules, postlexical (natural) processes, dialect borrowing, and dialect assimilation processes. Empirical basis is the city dialect of Vienna. The author also gives a prosodic analysis of this dialect, contrasting it with the Austrian Standard variety of German. The following two contributions focus on the linguistic Situation in German-speaking Switzerland with its diglossic distribution of dialect and Standard. Stem adresses the question of Standard acquisition by dialectspeaking children, a topic that has hardly been investigated empirically, aithough it has important implications for the theory of L2 acquisition. He identifies two phases: a first one in which Standard forms are holistically taken over, mainly from television, and a second one in which Standard features are learned at school in a more systematic way that shows many of the characteristics of 'true' L2 acquisition. Derlen (for Switzerland), Dittmar & Schlobinski (for Berlin, among others), French (for Northeast Yorkshire) and Müller (for Sicily) all are concerned with conversational dialect Variation. Werlen investigates the forms and functions of code-switching and code-shifting into Swiss Standard German in Swiss dialect discourse. Dittmar & Schlobinski analyse conversational convergence in various social situations, setting off this type of Variation against "habitual speech behaviour" and "Strategie code-switching". Whereas Werlen and Dittmar & Schlobinski analyse shifting and complex variational signs, French treats Variation of a single phonetic parameter, i.e. syllable-final /r-Zdeletion in a semi-rhotic English dialect. He succeeds in explaining part of this Variation (i.e. before pausa) by linking it to the turn-taking systems: /r/-maintenance works as a turnholding device in this dialect. Müller's paper introduces the contributions dedicated to the Italian dialect Situation. Müller works on the conversational micro-level where he argues for an interdependence between the Sicilian dialects' phonotactic structure (juncture dissolution) and its almost exclusive oral mode d'etre. The argument is particularly important as it brings phonological evidence to bear on the literacy/orality discussion which has found so much interest in and outside linguistics in the last years. The larger-scale level of societal networks is the point of departure for Sobrero's analysis of phonological convergence in the Salentino. Sobrero combines this network analysis with the conversational analysis of processes of negotiation in the beginning of an interaction episode. On the still larger level of whole dialectal speech

Introduction

9

communities, Maddalon & Trumper explore the sociolinguistic Situation in various regions of Italy. They distinguish repertoire types by pointing to

different processes and outcomes of convergence on the veitical and horizontal level in Calabria and in the Veneto. The final two papers deal with somewhat more 'exotic' sociolinguistic situations: a Pennsylvania-German settlement of the Old Order Amish is the topic of Enninger & Raith's contribution. The repertoire of this Community includes a German Standard variety of restricted use, the Pennsylvania German vernacular, and American English. Contrary to the other papers, Enninger & Raith analyse convergence of Pennsylvania German, not towards the German Standard, but towards American English. Problems of standardisation in Africa are discussed in Mioni's article. The Situation in Africa with its young and sometimes little accepted Standards, and with its complex relations between local vernaculars, koinai, and colonial languages, is an ideal field for proving one of the basic assumptions discussed above, i.e. that Standard languages are the product of deliberate and sometimes artificial attempts to homogenise a repertoire. Finally, we want to add that most of the articles in this volume are based on papers read and discussed at the Workshop 'Tnterpretive Sociolinguistics n", held in Constance in May 1985. The help of the Deutsche Forschungsgemeinschaft who kindly supported this meeting is gratefuUy acknowledged.

References Chomsky, Noam (1985).Transcript of an informal interview with N. Chomsky on the topic of bilingualism, conducted by F. Grosjean May 29, 1985. Gumperz, John & Wilson, R. (1971). Convergence and creolization: a case from the Indo-Aryan/Dravidian border in India. In D. Hymes (ed), Pidginization and Creolization of Languages, 151-167. London. Labov, William (1972). Sociolinguistic Pattems. New York. Moulton, William G. (1961). Lautwandel durch innere Kausalität: die ostschweizerische Vokalspaltung. ZFM2S. 227-251. Reiffenstein, Ingo (1976). Primäre und sekundäre Unterschiede zwischen Hochsprache und Mundart. Überlegungen zum Mundartabbau. In Opuscula Slavica et Linguistica (= Festschrift Issatschenko), 337-347. Klagenfurt.

10

Peter Auer & Aldo di Luzio

Schirmunski, Viktor (1930). Sprachgeschichte und Siedlungsmundarten. Germ. Rom. MonatsschrifiXWm, 113-122 and 171-188. Trager, G. L. & Smith, H. L. (1951). An outline ofEnglish structure: studies in linguistics. (= Occasional Paper No 3). Norman, Okl. Vasüiu, E. (1966). Towards a generative phonology of Daco-Rumanian dialects. Journal of Linguistics 2: 79-98. Weinreich, Uriel (1954). Is a structural dialectology possible? Word 10:388-400.

UTZ MAAS, Osnabrück

On the Interpretive Analysis of Historical Records: Linguistic Relations in Seventeenth Century Osnabrück^ For Herbert Penzl

0 . My paper is made somewhat marginal by its subject matter, for most readers are probably not familiar with problems of Middle Low German and/or Early New High German. Thus I have provided some introductory remarks for those unfamiliar with German philology. The argumentation, however, deals chiefly with questions of methodology and will hopefully prove to be relevant to Interpretive Sociolinguistics; in the following I present a number of the methodological difficulties with the documents in our research project on linguistic relations in Osnabrück in early modern times; for the convenience of the argumentation, I will restrict the examples to the first half of the seventeenth Century^. 1. As in any research project investigating earlier linguistic relations in Germany, ours focuses on a single aspect of the global restructuring within the Continental linguistic Space which began in early medieval times and came to a resolution in the early modern period. The central European German language area of today was culturally polarised in the early

The term linguistic relations is a calque on German Sprachliche Verhältnisse, used as a cover term in philology since the nineteenth Century for the complex historical Situation under examination; the translation foUows a familiar pattem, as e.g. German Produktionsverhältnisse is rendered by relations ofproductions The research project has been funded by the Deutsche Forschungsgemeinschaft. Reports are published as Materialien zur Erforschung der sprachlichen Verhältnisse in Osnabrück in der frühen Neuzeit (Osnabrück: Universität, Fachbereich Sprache, Literatur, Medien). Two volumes have thus far been published, and a third is in preparation. The actual research has essentially been done by Judith McAlister-Hermann, from whose work most of the examples presented here are taken. I am responsible for the flaws in the argumentation. Cf. the case studies (McAlister-Hermann 1983a, 1983b, to appear) for further development of the line of argument presented here.

12

UtzMaas

medieval period. There was a noithern region centering around the North Sea, where even today the vernaculars show strong common traits (English, Frisian, Dutch, Low German and the Scandinavian languages as well), which stood in Opposition to the southern midlands and highlands, where Upper German dialects were (and are) spoken, and whose cultural development was deeply influenced by dose contact to bordering Romance cultures. Yet even at the onset of the written tradition in the ninth Century the beginnings of a linguistic restructuring are evident in the texts, which appear as the first signs of a German language area. The Integration of this area is the work of early modern political and economical developments; an important step in this process was the creation of a corresponding linguistic ideal, i.e., the Hochsprache ('High German') of the Baroque linguistic societies; and still this process of Integration is not entirely concluded. It is quite natural that the geographically intermediate regions (with their Middle German dialects) would have played a key role in this restructuring, for these varieties shared peculiarities of both Low and Upper German - this is the background of the still inconclusive debate on the regional origins of Early New High German. Most handbooks favor East Middle German from the region of Thuringia, and so-caUed Meißnisch in particular, as the primary source. I cannot go into these questions here; I only want to point out that this argument lacks plausibility at least for the Western Low German area under investigation here, whose economical, political, religious and cultural orientation in general was directed towards the West Middle German (Rhenish) area which (like Western North Germany in general) maintains dose relations to the Netherlands.^ 2. Compared with the other research projects presented in this volume, projects on the history of language show a general deficit in the theoretical framing of the central research questions, both in relation to sociolinguistics or the sociology of language and to their methodology, i. e., how these questions are to be implemented in concrete research. If one undertakes to do historical sociology of language, one must, as afirststep, reinterpret the Any handbook of the history of the German language will provide a more detailled picture, although developments in the north are graduaUy presented only very briefly. Northern Germany is often portrayed as an area that passively received High German from the south. It is this over-simpUfied picture that our research project attempts to redefine. For information on Low German see for example Sanders (1982) (with ample bibliographical references). A more detailled picture of the process of linguistic Integration in the north is given in Maas 1985a.

Linguistic Relations in Seventeenth Century Osnabrück

13

current theorems of the sociology of language in terms of historical research. One factor determining this deficit is surely the decline of historical research in favour of ethnographic work in recent years; the decay of the philological departments {Alte Abteilungen) has certainly been a decisive factor, for the prerequisite for historical research, the practical knowhow necessary to use historical records (including the analysis of paleographic aspects and of linguistic forms), is no longer necessarily concomitant with language studies. This corresponds to the current emphasis on spoken language, which is part of the self-image of modern linguistics. There is an obvious advantage to investigating speech; certain sophisticated procedures such as different kinds of interviewing and testing simply cannot be applied to historical records. Even exact quantifying as practiced in survey research is impossible with historical sources, because (written) tradition is to a very high degree determined by chance, and no question of representativeness can even be defined. For example, the archives of the Osnabrück chancellary burned several times, so that we cannot even make a guess as to the original extent of formal chancellary records, not to mention our uncertainty as to the amount of written tradition in domains not subject to bureaucratic formalisation. But these difficulties should not overshadow the fact that it is possible to do qualitative („interpretive") work with historical records. In principal at least, the task of investigation is not different for written and for spoken language. In both cases the texts under investigation are linguistically heterogeneous - and the resolution of this heterogeneity, the extrapolation of varieties, is, as long as it is done textimmanently (using distributional procedures), of comparable hypothetical validity. Subjective data will be used in both instances to confirm the analytical hypotheses. In the case of historical records this cannot be done by questioning informants, but evidence from self-correcting interventions in the texts are available. In speech we observe self-corrections in the form of prosodic breaks and interruptions and alterations of syntactical constructions; we can trace the corresponding phenomena in writing in more or less obvious graphic interventions. In Order to do so, we must read (written) texts as inscriptions of several levels of writing praxis; we must distinguish at least between the immediate act of writing and successive editorial 'campaigns'. In this editorial process of working through the text we can again distinguish several layers; there are some interventions that continue the linear ductus of writing and are the consequence of continuous self-monitoring of the writer who, e. g.,

14

UtzMaas

Grosses something out and goes on writing in the same line with a new stait. This is to be distinguished from subsequent editorial interventions, where the writer has gone back in the text making corrections: between the lines or in the margin or inserting proofreader's signs with corresponding corrections somewhere eise on the page, etc. Here again, there are different levels to be distinguished, e. g. between editorial interventions by the same 'hand' (with possible differences of ink, in the ductus of the Script, etc.) and those made by different hands. The first example should illustrate more clearly what is meant by this."*

8

H 10 11 12

13 14 15

6 7 8 9 10 11 12 13 14 15

Vid. Weimers Polman noie. Vid. Reprodücirte Citatione Co. ad videndü ülteriüs procedi. sistirte Gerhard Meyef daneben in persone Lüdeke, Dönies bat des relation d Intimatione und dem negst bat er über denselb daß gerichte, Lüdeke Dönies berichtet daß er ver gangen Sontage twish ein und 2 Uhre Beclagt vor des. hause ufr Brugke Zu Uphaüse Copei Citationie habe Inti mirt in persone.

Figure 1 A technical note on the mode of transliteration: in accordance with this model of writing praxis, we try to capture the first Version immediately written down as our primary text, and record editorial interventions in the notes. This directly contradicts current editorial praxis, which strives to present the final Version (the Text letzter Hand). This has been

Linguistic Relations in Seventeenth Century Osnabrück

15

In rough translation the text reads: 'In the name of the widow (Weimers) (the lawyer) Polman appears and presents the citation. Furthermore Lüdeke Dönies appears in person and reports on the case. He demands this case be taken up at court. Lüdeke Dönies reports that he personally presented the court citation to the accused last Sunday between one and two o'clock in front of his house at the bridge at Uphausen.' The passage of interest is in line 12 in the word zwischen, 'between': The scribe first wrote twi, then corrected himself by making the initial ? into a z, i. e., he substituted for the non-shifted Low German initial consonant t its High German counterpart z (as to the forms of the letters cf. the two f s in berichtet line 11 and the initial z in zu line 14). Isolated examples cannot bear the bürden of much proof, and editorial interventions that are relatively unambiguous in their Interpretation are unfortunately rather seldom in our material. Yet even so they give valuable clues as to the conflictladen linguistic praxis that has been inscribed in the texts. On the one hand they show what the writer wished to avoid, in this case not quite successfuUy: the revealing un-shifted consonant is evidence of the negative self-stereotype of Low German. On the other hand we see what the writer tried to attain: the ideal of a High German mode of writing. 3. This first example should show that questions of the sociology of language can very well be raised with historical records; in work with autographical material the analytical procedures are even rather analogous to those of ethnography. But taking my argumentation further, I want to show that an apparent inconvenience of historical work, i. e., being bound to written texts, can actually be an advantage that helps raise new questions and overcome current short-cuts in linguistic reasoning. This is especially true of the aforementioned emphasis on spoken language with the myth of immediacy and authenticity of oral communication which leads directly to the devaluation of writing as merely a more or less deficient picture of spoken utterances. Not only does this myth stabilise ethnographic practice, it also evokes a kind of self-stigmatisation in those who do analyse written texts. It is a commonplace of works in the history of language, even those which very painstakingly scrutinize the relations of written language, to regard the results of research only as a link in a complex chain of evidence which criticized by many Germanist coUeagues; in fact the text thus produced sometimes does not make sense, for editorial interventions are generaily meant to restore the sense. Since our editorial praxis corresponds to our technical aims and does not intend to produce readable texts for a larger public, we want to abide by it.

16

UtzMaas

ultimately should lead to understanding of contemporaneous spoken language. The obviousness of this reasoning is in inverse proportion to its plausibility. It is of course correct that writing constitutes a kind of fUter through which much of what conveys oral communication will not pass (elaborate systems of linguistic transcription make this evident ex negativo): intonation and all prosodic features; facial expression, gestures and all the other non-verbal modalities that control communication. But it is also quite evident that the communicative stress of face-to-face communication allows the participants only limited room to deploy their linguistic competence. No matter whether it is a question of wanting or being forced to maintain a turn, or wanting or struggling to obtain one, the mechanism of turntaking requires constant vigilance, Observation of the other, Interpretation of the symptoms in the behaviour of the other to determine whether s/he finds the utterances proferred interesting enough to go on listening to, or whether s/he is about to interrupt, if s/he does not understand you, etc.^ This communicative stress surely ties up huge amounts of psychic energy, not to mention all the kinds of resulting obstruction familiar to everyone who prefers to deal with any difficult Situation in writing rather than face oral inquiry. In other words: The necessity of Controlling the immediate social Situation during oral communication extracts cognitive energy otherwise available for the production of the text, thus leading to the specific orate style of speech.^ The strict linearity of speech forces all structuring activity on the short-term memory. In contrast short-term memory is relieved during writing, where two-dimensional space can be used. Despite these distinctions writing is not completely different from speech. Here too a relation to the addressee controls the writing praxis but at greater distance and only in a rather indirect way; this becomes especially evident when we understand by writing not only the immediate writing down, including its monitoring, but the complete production process of a written record, including later editorial interventions. These somewhat general and for some readers perhaps trivial remarks are intended only to point out the absurdity of the current opinion by which written texts are seen as derived from spoken texts and are scrutSacks, Schegloff & Jefferson (1974) have shown that there is no let up on this stress; despite the fact that there are places where breaks are supposed to occur in the sequence of utterances, at which points turn-taking would be preferable. The participants behave with anticipation, so that communicative stress is maintained even in the intervals where breaks should occur. See e.g. Tannen (1982). For the distinction between oral/written and orate/literate see Maas (1984).

Linguistic Relations in Seventeenth Century Osnabrück

17

inized as it were with X-ray vision for the 'underlying' oral structure. Of course this position can historically be seen as a reaction against an older view that regarded oral texts as reduced writing, thus attributing no value in itself to an analysis of oral linguistic praxis. But this controversy has only museal interest today and to continue it would be a kind of ritualized commemoration of the emancipation of linguistics from its role as the handmaiden of philology. Instead of developing the argument further here, I shall now axiomatically propose that the analysis of graphic as well as oral texts has to deal with their linguistic articulation (which means it has to be capable of being explicated in the framework of a theory of grammar); in the concrete analysis this linguistic articulation must be related to the particular conditions of oral or graphic realisation of the specific text. This is in turn to be distinguished from the difference between orate and literate style, which refers to an independent dimension of analysis, as the style categories can determine texts in the opposite modality (e.g., we can 'talk like a book', and oral Speech can be noted with naturalistic fidelity; linguistic transcription is a borderline case). These preliminary remarks determine the framework for the analysis of written texts. As I said before, these are to be read as inscriptions of a writing praxis, whose forms had to be leamed, whose cultural determinants are to be deciphered. This excludes scrutinizing texts in X-ray fashion for the underlying patterns of oral speech which alledgedly have been transcribed in them; the task of analysis is rather to discover the cultural patterns which are reproduced in the writing praxis and which the writer happened to have encountered before. The mediating instances of this process of cultural reproduction make this kind of analysis a highly complex process: it has to begin with the teaching, or better: the learning of writing/reading (as a formal Institution to this task, the modern school cannot be presupposed); it has to include subsequent individual reading history, and finally, since the corpus of read materials is heterogeneous, it has to go on to the complex mechanisms of evaluation that factorise the reading material. Following Hjelmslev, I make use of the notion of connotation to defme this structuring of experienced forms of praxis (see Maas 1985a). Each cultural form (in this case: form of writing - but the same is true for oral forms al well) represents (or should I say: appresents?) the Situation or the class of situations in which the subject who practices it (i.e. the writer/reader) has met it before. Just one further comment on this model of linguistic praxis, which may be used here as a cover term to encompass writing praxis: Linguistic praxis

18

UtzMaas

is, in this sense, accessible to conscious reflexion. This is, of course, the precondition for the specific mechanisms of social control: the Stigmatisation of certain forms, the establishment of certain forms as compulsory models in school, not to forget the informally institutionalized stereotypes below the threshold of the State apparatuses. The official discourse about linguistic problems institutionalises, for its part, the threshold of awareness for these regulations, which determine the subjects' reactions to them in different ways. Putting it simply, the official discourse of linguistic cultivation {Sprachpflege), as for example in teaching in school, creates areas of linguistic saliency easily accessible to conscious control that filter oral as well as written language via continuous self-monitoring. A t the level of the lexicon, grammar and even phonemics, this orientation towards the Standard language is easier to obtain than at levels below the threshold of this discourse; thus a Speaker with correct Standard grammar, even as far as phonemic distinctions are concerned, can be identified on a sub-phonemic level by the individual local or regional 'accent'7 Mutatis mutandis the same is true for writing, where choices involving the lexicon, syntax, orthography and the global ductus of writing are salient and consciously controlled - corresponding to the official discourse about writing involved in the process of teaching and learning to write/ read. But there is no such discourse for sub-graphemic structures; thus, these sub-graphemic patterns should prove to be a very fruitful area to explore in cultural analysis. But here paleographic problems begin to accumulate. Paleographists are familiar with the break in tradition that occurred in medieval times; during the early epoch the book-writing of the artisans was characterised by the meticulously identical reproduction of individual Script types, to the extent that it is generally impossible to identify individual scribes of a particular codex, although stylistic individuality of the various schools of writing (the scriptoria) remains identifiable. In later times, as the practice of writing became more and more widespread, idiosyncratic traits of individual hands began to appear even in book-writing, forcing the scriptoria to make regulations forbidding a shift

In the individual linguistic biographies, the consequence of being confronted with the official discourse of linguistic cultivation and its sanctions is dramatically evidenced by a change in learning flexibility. Before starting school, children are more or less malleable in respect to the local accent, as their reaction to relocation by their parents normally shows; as they begin the climb up the steps of school career, they internalize the linguistic discourse of school and increasingly lose this flexibility, which culminates in adults' well known difficulty in learning a foreign language.

Linguistic Relations in Seventeenth Century Osnabrück

19

of hands within a quire or at least not on page so as to avoid a heterogeneous impression (in earlier times this problem was simply non-existent). This differentiation increases rapidly with the growing demotisation of writing in later medieval times.^ But even with the later, extreme individualisation of handwriting which develops parallel to the new mode of standardising texts for reading by printing them with movable letters - every hand, no matter how idiosyncratic, is an adaption of culturally defined models, ideal script types, that can be identified in a text in the same way as we identify its grammatical varieties - and these ideal script types can be interpreted in terms of their cultural connotations. Although there may be no special problems in theory, there are tremendous practical problems of analysis here, for we know almost nothing about the existence of these ideal script types, since late medieval and early modern paleography is stUl in its beginnings.' In addition, individual texts and hands present a baffling amalgam of different script types; sometimes there are a dozen different forms for a single letter. Such technical problems have demanded a great deal of time and energy in our project. But I will not go into the details here, for they could be discussed fruitfuUy only within a paleographic framework and have no direct correspondence to the other projects presented in this volume. In the following I shall dwell instead on the graphematic and supragraphematic (grammatical) level, where parallels to the other projects are more evident. 4. With the following example I hope to demonstrate the scope of connotation analysis. The reproduction shows two pages from an Osnabrück printing of 1618. It is evident that contemporary printings have to be investigated for the study of the relations of written language in town; but here additional

See Maas (1985b) for a discussion of the term and historical illustrations. See Bischoff (1979). The research Situation will be improved when the comprehensive projea of a catalogue of dated manuscripts of the fourteenth to sicteenth centuries, begun in nine European countries in 1959 (see e.g. Scarpatetti 1983), has made further progress. For a case study of an individual chancellary praxis see Heinemeyer (1962) and also the stimulating study by Frenz (1981), who tries to correlate paleographic and grammatical traits in their areal distribution. As an introduction and for self-study, Gladt (1976) is to be recommended. For the general problems of linguistic interpretation of historical records, see the excellent introduction by Herbert Penzl to whom we owe much for a better understanding of the inscriptions of writing. Therefore I dare to dedicate this paper to the Altmeister of historical linguistics.

20

Utz Maas

^fniferfenrcbniarij, eoiil)« n'iin im Pontam". l-imcciiisrntArenus, aiicl) ©Palenberg fdireibn/tafi ^efclgf 6r«ii|f ® repiAhrirt tar(/atid) ein het]} ©omnter/ ®?a(,cmet/ gum anrcrntnüf }ii QSicnftlfidtrti(t/ ^JfittemfelbenwenmafimfrieiV ^ c n fem «»iirärn aufFfcin «gtn @d)fof / C3rf߀(l-faduentbet tn6bbie!t6en(? 3" 0.

s

nO

50

Peter Auer

For MHG fand ü, the following realisations were found: - ü\ apart from two isolated cases^, ü is realised by any of the sounds in the continuum - (a)u/0(u) - Au - m- au. (In the quantitative representation, the first two - symbolised as U - and the darkened ('verdumpft') variants Au/Ouku - symbolised as O U - have been conflated.) - i: apart from two reduced forms^, the sound is realised as one of the following: i:/i - Eihi - äi - n - ailoEloB. (Again, the three raised variants symbolised as EI - have been conflated for quantitative representation.) In addition, there is a small percentage of darkened diphthongs Ai/Oi'^. The following diagrams (1 and 2) show the percentage of monophthongal, diphthongal-dialectal (raised or darkened) and diphthongalstandard realisations'®. In order to avoid rash age-dependent interpretations, Speakers were ordered according to an Overall dialectalisation index. This index is a technical device for displaying the data and must not have attributed any theoretical importance". One thing is immediately apparent: the monophthongal realisation of MHG i is much rarer than that of MHG ü. However, the difference becomes less dramatic in the context of the lexical environments in which monophthongs have survived. All but three monophthongal realisations of MHG ä occurred in the dialectal version of one of the following words:

One most probably mispronounced uef (for auf, 'on'), and one slighdy raised ai in Ousrdme ('to clear out'). (Here, the aih one of the realisations of the M H G Umlaut iu, that is, the item is derived from the MG variant riumen, not the M H G mmen. Cf. Paul 1913, 40 on the influence of the following bilabial.) No difference is made between tense and lax variants here. Variation between the two has to be treated separately and merits a detailled investigation of its own. Two instances of gottsadank ('thanks heaven', sai sa). These diphthongs are a frequent realisation of M H G ei, as we shall see below. The spill-over from one into the other source of Standard German ai indicates a certain, although marginal fuzziness of the underlying differentiation between old i and ei. Only those parameters and informants were taken into account for which at least 10 items were available. Usually, the number is between 20 and 30. If, for an informant, the minimum is not reached, his/her age rank is is in round brackets in the diagrams. The index is based on the ranks of the individual informants for all investigated parameters for which a Statistical representation was possible for all 20 informants. These were: monophthongal vs. diphthongal realisation of M H G ü; monophthongal vs. diphthongal realisation of M H G t; dialectal (monophthongal and darkened) vs. Standard realisation of MHG. ü; dialectal (monophthongal and raised) vs. Standard realisation of M H G i; realisation of M H G eins dark Oi/Ai\s. other realisations. Age can be reconstructed from each informant's number which gives his or her rank on the age scale.

51

M H G i and ü in the City Dialect of Constance

I2SJ M H G i a s U

2

i 5 s « » +1 -u ftf! j äo! « « •(«

Diagram 1:

»

-W

2

MHG«asOU



MHG i as AU MHGoi dial. i: is only applicable to those lexemes which have - historically speaking - foUowed the NHG diphthongisation; in practice it may prove to be very hard to describe the constraints on such a rule.

102

IwarWerlen aar: nöügiirde oder das du aefach wotsch Art Neugierde, oder, daß du einfach willst (1/2;2,15)

shows in genao (instead of gnao) a frequently observed form of SHG-like pronunciation of prefixes. Nöügiirde represents a somewhat more complex case because nöügiir is documented in dialects whereas SHG Neugierde is not. It is a matter of debate whether this is a lexical transfer from SHG or a phonetic assimilation to the Standard. Phonetic shifting occurs mainly in prefixes, suffixes or in the shape of unrealised assimilations. It could be described on the basis of phonological rules, but this would require an extensive discussion of the problem of underlying form. The second much more effective form is lexical shifting; SHG lexemes are completely pronounced in dialect. An outstanding example is: (4)

S1

äbe du machsch psüchologii nim i aa eben, du machst Psychologie, nehme ich an /3 sec./ S 2 nimsch du aa nimmst du an S 1 iu ja S 2 ehe + [laughing] fääuaanaame Fehlannahme i ha jets aakno du machisch psüchologü ich habe jetzt angenommen du machst Psychologie S 1 aa nei [laughing] dsch o ne fäälannaam ah nein das ist auch eine Fehlannahme (1/1,1;5)

In this case SHG Fehlannahme ('false assumption') is phonetically realised in forms that make it look like a dialect word. Similarly: (5)

S2

nei i gloube richtig jedefaus isch nein ich glaube, richtig jedenfalls ist es mer rede wüu e + jaa as mer daß wir reden, weil e ja daß wir schpräche zunenang soo es + kchuasi sprechen zueinander so ein quasi ksichtslos schpräche zunenant i gloub gesichtslos sprechen zueinander, ich glaube, s het scho ne kwüssi richtikcheit es hat schon eine gewisse Richtigkeit chönti mier voorschteue jedefaus (1/1,1 ;5) könnte ich mir vorstellen jedenfalls

Swiss German Dialects and Swiss Standard High German

103

This not very coherent contribution uses the verb sprechen zueinander ('to speak to each other'), which appears in its dialectal form only as reden miteinander i^to talk to each other'); the use of a SHG-like style is confirmed by other items (e kwüssi richtikcheit). That both participants were conscious of this style and that they were taking this in account is manifest in the following interesting section: (6)

S2

S1 S2 S1 [S 2 S2

S1

S2 S1

ase ich säg der s iz aber weisch also ich sage dir es jetzt, aber weißt ich säg es immer esoo des t lüt ich sage es immer so, daß die Leute zeersch müend schtudiere. ich bin zuerst müssen studieren ich bin schtokchweerkchtekchtonikcherin am Stockwerktektonikerin am geoloogische inschtitüt + geologischen Institut schtokchwerkch Stockwerk tekchtoonikcheri [loudly] Tektonikerin tekchtonikcherin [low] Tektonikerin then explains that she is a charwoman] nae ich tue det aefach ds geoloogische nein ich tue dort einfach das geologische inschtituut raenige + ase ich bin Institut reinigen also ich bin aarbaeterin + oder Arbeiterin, oder ehe + + dasch ged wider e bewüs soo aha das ist gerade wieder der Beweis so aso + + + + + + aso so wie du retsch + hät also also so wie du redest, hätte i jez wüürkchli uf uf ä öper + + + + ich jetzt wirklich auf auf eh jemand uf vaas p hätsch jez - | tipet auf was hättest du jetzt getippt - eee

+ + uf e auf eine

schtudäntyn (6/1,2) Studentin

From the start, S 2 resists talking about her profession as this is irrelevant for her person; S 1 on the contrary thinks that knowing each other's profession would provide more clues for the conversation. After several

104

IwarWerlen

starts the above-mentioned section follows. S 2 uses lexical forms borrowed from SHG: the suffix -in and later, instead oi putzen ('to clean'), the SHG word reinigen ('to clean'). S1, on the basis of her speech, takes her for a Student. Later S 2 explains that she had consciously acquired her manner of speech through interaction with the students of the Institute. Another example shows a relatively rare case of syntactic shifting: (7)

S1

ich bi aefach da iekfüert woorde + ich bin einfach hier hereingeführt worden me hät mier ksaet de versuechslaeter man hat mir gesagt, der Versuchsleiter chunt + + + + dä du warschinlich bisch (1/2,2,1) kommt der du wahrscheinlich bist

The relative clause introduced by dä (SHG der) instead of wo is the most obvious and most often criticised case of syntactic shifting. Other cases are less easily detected. They concern the position of finite modal verbs with infinitive complements and finite verbs in subordinate clauses which vary considerably within dialects and the use of simple complementisers in subordinate clauses of the type ich weiß nit, warum dass er isch hei kange ich weiß nicht, warum daß er ist heim gegangen

with the complementizer warum daß which is, however, reduced occasionally in many dialects. Another example of syntactic shifting is (8)

S 1 je nu des isch jo sicher gewollt ja nun, das ist ja sicher gewollt das des soo isch oder (2,1,1 ;4) daß das so ist, oder

The construction des isch gewollt is certainly SHG as the participle gewollt does not occur in dialect at all (cf. Werlen 1985a). Returning to lexical shifting, a slightly different form is found in the use of SHG vogue-words which are part of a style defined by age. The two observed cases are dufte 'fantastic' and doof'hor'mg, stupid', the latter being more frequent: (9)

S2

genaao meer sii di kschprächsleiter genau wir sind die Gesprächsleiter sch duufte oder duufte oder (1/1,1 ;5) (das) ist dufte, oder, dufte, oder (10) S 1 des isch doch ao doof des schpüu + + + das ist doch auch doof, das Spiel

Swiss German Dialects and Swiss Standard High German S2

S1

105

nei es isch nit doof i finge s eis fo nein es ist nicht doof, ich finde es eines von de beachte schpiiu weisch isch wirklech den besten Spielen, weijh, ist wirklich "Schees interessanter keines interessanter das riist der doch t näärven us nid das reißt dir doch die Nerven aus, nicht

(7/l,2;13) Item (10) shows that doof'is not accompanied by other shiftings, it is rather to be Seen as a teenage vogue-word exactly like das riist der doch t näärven US.

Another form of lexical shifting approaches code-switching; we are dealing here with formulas of the length of whole phrases which are borrowed from SHG and which could almost be called citations: (11) S 2

aber lue oder duu muesch deer o aber schau, oder, du mußt dir auch im bewuste syy das e + + + das t im Bewußten sein, daß eh, daß du äbe + psüüche vom möndsch gaar nie eben die Psyche vom Menschen gar nie cha erfasst wäärde oder hesch nit kann erfaßt werden, oder hast (du) nicht ts kfüü so (4/1,2) das Gefühl so

Im bewuste syy is the important formula here. The first part of item (11) has entirely SHG-like form. A shift in the construction is followed by a passive erfasst wäärde with dialectal position of the modal verb; in psüüche vom möndsch the syntax is simüarly dialectal whereas the word is technical. Finally there is the widely observed monophthongisation in kfüü which ought to be pronounced in dialect as kfüeu. Another item shows a translated citation which is rather rare: (12) S 2

wei mer nid wider ga wollen wir nicht wieder gehen S 1 [laughing] pakche mer ts züüg du packen wir das Zeug, du S 2 [laughing] /12 sec./ [unintelligible] do chunt da kommt sicher kchene sicher keiner S 1 jetzt waarte mer no jetzt warten wir noch

106

IwarWerlen S2

ufwäär+pufgödoi auf wen auf Godot [both laughing] S1 Lja nei -I (7/,2,2;4) ja nein

Here the title of the play 'Waiting for Godot' is cited in dialect; besides pakche mer ts züüg is also an assimilation to the SHG idiom es anpacken. 4.2.2. Switching into SHG This form of Variation concerns - in the corpus - one single turn exclusively. We never observed a dialogue which changed into SHG-like forms over a longer period of time. Therefore, we are dealing rather with phenomena of transfer or code-fluctuation. It is only the length of the SHG-like forms that is different: it can vary from one word to a phrase to a whole sentence. I am adding examples of increasing length: (13) S 1 nei mier müen daa öpis mache nein, wir müssen da etwas machen gemaensaam mit wyter (2/2,3) gemeinsam, nicht, weiter

Here the pronunciation of gemaensaam by S 1 is distinctly SHG with a füll e in the prefix and a SHG diphthong. (S 1 usually uses [ei].) The appropriate word in the dialect would be zäme 'zusammen, together'. A few words later, S 2 cites the same form as kemaensaam. In both cases both speakers insist on doing something together, not individually. Echoing forms like these occur again once or twice. (14) . . .

aber ts reseltaat isch immer aber das Resultat ist immer dass es problem nid aefach vo vornheraen daß ein Problem nicht einfach von vornherein fernaent wiirt (3/1,3) verneint wird

Vornherein is articulated in SHG. The distinction between lexical and phonetic shifting is blurred, however. (15) S 2

i haa ja dii zeersch kfraakt süsch ich habe ja dich zuerst gefragt, sonst hat i ja kseit kchomm meer tüe des apzie (4/1,2) hätte ich ja gesagt, komm, wir tun das ausziehen

Komm (otherwise dialectal chumrri) is articulated very clearly in SHG; S 2 is arguing that she had asked S 1 whether she might take off her mask; had she been the investigator she would simply have given an order, komm...

Swiss German Dialects and Swiss Standard High German

107

The following example shows a foreign element being used as a foreign word and then 'translated': (16) S 2

si rede mitenang + aber sii hei kchee

sie reden miteinander, aber sie hohen kein

säg schnäu

sag schnell

S 1

weis

S2 S 1

rnid

weiss nicht Lfiidbäkch

feedback

kche fiidbäkch [very low] + [loudly] me cha

kein feedback

man kann

s eifa kches wisawii oder eifach

es, einfach kein Vis-ä-vis oder einfach

kche + e kche reakchtsion + i ha nit

keine

keine Reaktion

ich habe nicht

gäärn so Wörter [laughing]

gerne solche Wörter S2

e oder dii wäärde dr soo iitrichteret

eh, oder, die werden dir so eingetrichtert

dii chan + nä chöme si der aotomaatiisch

die kann nachher kommen sie dir automatisch i sinn

in den Sinn

S 1 S2

jaa fiidbäkch das heist ja zurückcherneeren

ja, feedback, das heißt ja zurückemähren jaa + es i es isch en aart ernäärig oni

ja, es ist, es ist eine Art Ernährung, ohne fiidbäkch geit gloub filicht o z grünt (7/1,1 ;10)

feedback geht (man), glaube, vielleicht auch zugrunde Zurückemähren is realised in S H G ; it is clear that the Speaker is relying on something like an entry in a written dictionary. The section is very interesting for metalinguistic reasons too as it shows reflection on the use of foreign words. (17) S 1

näär nimts eim gäng wunder we me

nachher nimmt es einen immer wunder, wenn man

gäng nume t ouge gseet u t haar u t aarme

immer nur die Augen sieht und die Haare und die Arme

wie das dee der ganz uusgseet [laughing]

wie daß dann der Ganze aussieht

weisch oone +

weißt, ohne

S2 S 1

[unintelligible] jaa /5 sec./

ja

108

IwarWerlen S 2 e dess s aa jaa dasch warschiinlich drum eh das sah, ja, das ist wahrscheinlich deswegen S1 waarschaenlichpow] (7/l,l;ll) wahrscheinlich

Here ohne and particularly wahrscheinlich are pronounced in SHG. The latter is spoken very softly, thoughtfully as if the Speaker was only talking to herseif; noteworthy is the fact that S 2 articulated the same word in her previous turn in dialect. (18) S2

ja Waas du machsch du laetisch aes ja was, du machst, du leitest eins S 1 nei i leite kcheiss + + + r das war i nein ich leite keines, für das wäre ich äuwä nift keaeknet ha-n-i soo s kfüel (3/1,2) wohl nicht geeiffiet, habe ich so das Gefühl

Nicht geeignet is realized in SHG. The Speaker contradicts to the assumption of being the investigator, by saying that he would not be suitable. Further items showing clear switching to SHG include tides of TV shows, books and plays. (19) S1

schpeethok Heist s (2/2,2) 'Spätholz'heißt es

The book title "Spätholz" is given in SHG. In dialect it would sound schpaathouz. (20) S 1 was i no geern los des isch ring was ich noch gern höre, das ist Ding si k t naachrichte-n end echo dr zaet (8/1,3) sind k die Nachrichten und "Echo der Zeit"

"Echo der Zeit", the name of a radio broadcast is cited in SHG. S 2 repeats the item a litde later in the same way. (21) S 1 [after a long pause] soo + wartn wir aof so warten wir auf • godoo + hm' • Godot S 2 L [laughing] -1 ja'a kämmer öpe no ne frischt (3/2,2) ja gehen wir etwa noch eine Frist

In this counterpart to (12) S 1 changes the tide of the play, pronouncing it in SHG; he comments on the Situation.

Swiss German Dialects and Swiss Standard High German

109

Example (21) leads us to a category of idioms or proverbs which are introduced in SHG: (22) [S 2 and S 2 understand that something is wrong in the Situation; then:] S 1 also [laughing confusedly] + + + + waarten

S2 S1

S2 S1 S2 S1

also waarten wür der dinge dii daa kchommen + wir der Dinge die da kommen jä chunt no öper + ja, kommt noch jemand ja ich ha kche aanik + + duu siigsch dr ja, ich habe keine Ahnung, du seiest der versuechsleiter ha-n-i + pmeint Versuchsleiter habe ich gemeint hesch duu kmaent [laughing] hast du gemeint jää ja ehä ich waes ase fo nüt nein, ich weiß also von nichts also isch prima also + waarte mer emaal (3/1,2) also, ist prima also, warten wir einmal

S 1 comments on the Situation. This is understood literally by S 2. His question leads to a discussion of the problems of the Situation itself. The discussion ends with the same comment, this time in dialect. (23) [S 2 says he intends to study history in order to learn something about polirics. Whereupon:] S 1 so nach dem moto würdsch du säge des + so nach dem Motto, würdest du sagen, das ist alls schon mal da gewesn + + wen uf alles schon mal da gewesen, wenn (du) auf politikch luegsch und mit der kschicht die Politik schaust und mit der Geschichte fergliichsch he (1/1,6;5) vergleichst, he

Later, S1 labels the motto alles schon mal da gewesen with "spitz" ('sharp') i. e. ironical, but he explains that it sounds somewhat right. S 2 then repeats this motto in dialect translating it lexically by using gsi instead of gewesen. (24) [S 1 talks about Aying hang-gliders] S 1 unt de wirt mer ao + dementschprächend und dann wird man auch dementsprechend en entschuldigung derfür sueche das me s eine Entschuldigung daßr suchen, daß man es

110

IwarWerlen nid wot leere + + u denn zum biischpüu saet + + nicht will lernen, und dann zum Beispiel sagt, ich find das lächerlich oder ich find das ich finde das lächerlich oder ich finde das idiotisch was die daa mached das isch /5 sec./ idiotisch, was die da machen, das ist jeedm das saene + + [unintelligible] [laughs] jaa we jedem das Seine ja wenn me nid fertik wird dermit saet me-n aefach nmn nicht fertig wird damit, sagt man einfach jaajaa jedem das saene söl mache jaja jedem das Seine, (er) soll machen was er wot (8/2,1 ;23) was er will

S1 cites the saying jedem das Seine twice in SHG and adds a translation; the motto is used as an assumed motivation for those who do not want to fly hang-gliders. Besides S 1 is distinctly using a SHG-Iike style (e. g. dementschprächend, ich find das... etc.) Further material on code-switching is not found in the corpus. 4.3. Functions of linguistic Variation in the investigator Situation The first intuitive impression of (1) to (24) suggests a very strong differentiation of two functional types. Shiftings from forms not or very little influenced by SHG to expressions resembling SHG on the phonetic, lexical or syntactic level obviously strongly depend on the definition of the specific Situation and on the way speakers define themselves. This Variation is for most speakers an integral part of a special verbal style which identifies the Speaker as an 'educated' person (see (6)). The self-characterisation as an 'educated' person is part of face-work and connected to the definition of the Situation. For whenever the subjects become aware of the fact that they are both subjects in the test, the Situation changes from a test Situation to a waiting Situation with small-talk; accordingly in various cases a number of characteristics disappear. In items (l)-(2) and many others the SHG-Iike forms occur whenever the profession of the other participant is mentioned for the first time seriously. It seems that this topic - profession - which also appears in item (6), is essential for the self-characterisation of the participants. This can also be observed in other utterances about profession, as in (25) S 1 was machsch du' + was machst du

Swiss German Dialeas and Swiss Standard High German

S2

III

ech bi medizinschtudänt (1/2,2;8) ich hin Medizinstudent

S 2 uses the official expression in dialectal form instead of circumscribing it {ich mache medizin ...). Another, different function is found in SHG-variants. Gumperz (1982: 75-80) discriminates the following six functions of code-switching: quotations, addressee specification, interjections, reiterations, message qualifications, personalisation vs objectivisation. The last function is the most interesting but also the least clearly defined one. In our corpus the second function is only observed in item (17) where wahrscheinlich has the character of an utterance directed to the Speaker himself. Formally we are dealing with a reiteration as well but not one of the same type Gumperz described. Reiterations occur several times in our corpus as in item (23) where the SHG-form is later repeated in dialect, or in (24) where the Speaker repeats himself in SHG before switching into dialect. Rather frequently, we find quotations of tides of books, films, plays and such, which occur in SHG in other contexts as well, as in (26) [S 1 says he would find it terrible to be forced to live constantly with a microphone] S 1 sch graauehaft das ist grauenhaft S 2 big brööer is wotsching juu (7/l,l;13) Big hrother is watching you

The quote form Orwell's 1984 is given in English. Quite another funcdon is observed in the case of Warten auf Godot, Warten wir der Dinge, die da kommen etc. Here the Situation itself is being commented on and the choice of SHG can be explained by Gumperz' last category. It is also to be considered that stereotypes may be used to legitimate behaviour as in item (24) jedem das Seine. Objectivisation in the sense of legitimisation by relying on the testsituation is the basis of the 'joint' acting in item (13); in a similar way the Speaker in (18) stresses his disability by nicht geeignet.

5. Medical Interviews 5.1. Corpus and Situation The corpus of medical Interviews contains 24 dialogues between doctor and patient of the department of rheumatology at the Inselspitalm Berne.

112

IwarWerlen

All doctors and patients are native speakers of a Swiss German dialect.'* Patients and doctors agreed to having their interview taped. They were wearing microphones; a certain uneasiness cannot be excluded. In the course of a general check up only anamnesis was taped. Doctor and patient are facing each other. The doctor follows a special pattern of questions and makes notes on symptoms, previous diseases, Operations and other important information. In general the patient's role is to answer those questions although there are patients who try to talk about topics they choose themselves. For the patient the Situation of anamnesis is less structured than for the doctor. Still, we may assume that most patients have experienced similar situations and dispose of certain definitions of the episode. They particularly accept their part as dependent and restrict themselves to following the doctor's Instructions. Even when they try to draw the doctor's attention to their own topics they allow him to intervene. For the doctor a medical interview is a form of nontechnical communication. He has to try to translate his technical expressions for the patient. Such translations sometimes imply code-shifting, e. g. when SHG terms are transferred phonetically into dialect. Another translation takes place when symptoms are related to technical terms, i.e., to a diagnosis. The patient, too, has to do translations: his first problem consists in the difficulty of localising and describing his specific pain; 'trained' patients already dispose of a technical vocabulary corresponding to the doctor's language. Medical staff, e. g. nurses or other doctors, might use specific technical terms to which - at at least in our corpus - the examining doctors do not answer. To talk adequately in the given situations means for the patient to choose a more formal register. This includes shunning taboo-expressions for excretory and sexual organs and procedures, the mentioning of which cannot be avoiding, expressions which are thought to be rude etc. Therefore it is to be expected that participants in this interaction switch to forms closer to SHG. This variety is abandoned only when the patient fmds enough time for longer, usually autobiographical talk, e. g. in response to questions concerning the patient's profession, with descriptions of accidents or psychological problems.

I wish to thank Francesca Roncoroni-Waser from the Institute of Applied Linguistics (University of Berne) for making the material she taped and transcribed avaUable to me.

Swiss German Dialects and Swiss Standard High German

113

5.2. Forms of linguistic Variation in the medical interviews 5.2.1. Shifting to variants dose to SHG Forms of phonetic, lexical and morphological shifting are found just like in the first corpus; I am therefore restricting the number of examples. Phonetic shifting: (27) P

aber är het o eh beschwärde mit em aber er hat auch Beschwerden mit dem mit em schnuufe-n oder isch äbe mit dem Atmen, oder, (er) ist eben e chlii e feschte (6,13) ein wenig ein fester (dicker)

The patient is using beschwärde with stress on the first syllable, the prefix; in such a way the word is recognisably SHG, i. e. a foreign word (foreign words having first-syllable stress in the patient's dialect). (28) P

ja i ha dispänz gha für sch schwär Ja ich habe Dispens gehabt für schwer z lüpfe oder (16,34) z« heben, oder

Here the patient pronounces sp instead of schp in the foreign word Dispens. All other items correspond to those dealt with in section 4. Morphological shifting is rarely found in the investigator Situation. In the medical corpus we get present participles which are not common in the dialect unless they appear in a phonetically changed form which renders them unrecognisable as participles: (29) P

(30) A P A P

. . . oder rächne mer uus s go turne drü oder rechnen wir aus das gehen turnen drei mau ir wuche oder i weis o nid mal in der Woche oder ich ich weiß auch nicht i miech eifach wider öpis wo rükebelaschtend ich machte einfach wieder etwas, das rückenbelastend isch das es de scho das es wider chämt (13,10) ist, daß es dann schon, daß es wieder käme händ si bi der aarbeit händ si dann müe haben Sie bei der Arbeit, haben Sie dann Mühe |-zytewys zeitweise Lebe dur das eben durch das (deswegen) zytewys scho es isch eifach behindernd (16,14) zeitweise schon, es ist einfach behindernd

114

IwarWerlen

Belaschtend'm(29), behindemd 'm (30) and other items show distinct signs of shifting usually combined with other elements, as in behindernd where be- is phonetically SHG. More frequendy we find lexical code-shifting which is, however, not always used consistently by the patients. An example is: (31) A

P

also herr XY + [clears throat] wa isch ihres Also, Herr XY was ist Ihr problem warum sind si dahäre cho? Problem. Warum sind Sie hierhergekommen? eeh miis problem isch eifach das i eh mein Problem ist einfach, daß ich vermehrt rükeweh ha ir letschte zyt vermehrt Rückenweh habe in der letzten Zeit un i gärn mau hat wöue wüsse was eintlech und ich gern mal hätte wollen wissen, was eigentlich was eigentlech los isch was nid normau isch was eigentlich los ist, was nicht normal ist i mym rüke das heisst i ha einiges in meinem Rücken, das heißt ich habe einiges gwüsst wüu my my fatter isch ja mediziner gewußt, weil mein mein Vater ist ja Mediziner

m gewesen A mh P • u mir hei das aus chind sehe mal und wir haben das als Kind schon mal untersuecht u denn sy gloub eh untersucht und dann sind, glaub, offeni wirbuböge hei si denn offene Wirbelbogen haben sie dann A mh [affirmative] P hei si denn diagnoschtiziert u das het haben sie damals diagnostiziert und das hat mi an sich aus chind nid wyter gschtört (21,1 f.) mich an sich als Kind nicht weiter gestört

The padent uses the non-dialect words vermehrt, einiges, mediziner, diagnostiziert, an sich and the construction ir letschte zyt (instead of Bernese ir letschti). He consciously is using a medical register. It is interesting to see that doctors translate patient-statements into their technical language as the following section shows. The doctor is asking for localisation of the pain: (32) P

geit scho fasch meh i ds gsäss abe eifach geht schon fast mehr in's Gesäß runter einfach

Swiss German Dialeas and Swiss Standard High German A P

115

mh' + schticht++

sticht A P

aso en schtächende schmärz also ein stechender Schmerz ja und nächär e ziet es sech de ja und nachher zieht es sich dann mängisch abe eifach dür s ganze bei (14,3) manchmal runter einfach durch s ganze Bein

The patient is using the non-taboo-word gsäss appropriate in the Situation. The doctor changes (es) schlicht xo e schtächende schmärz. This is accepted by the patient. Patients not always accept such medical expressions. The following example shows how a patient can insist on his version: (33) A P A P

und sind dann das dikchi belääg gsy oder und sind dann das dicke Beläge gewesen oder ja s git eifach so wi nes wi nes rüüfli ja es gibt einfach so wie ein wie ein "Rüüfli" ehä jawohl wi nes rüüfli (12,42)

The patient replaces the doctor's SHG-like belääg by his dialect-word, which might be translated by Schorf, 'scab' (scab on wounds are usually called ruuf/rüüfli). Doctor and patient often are aware that processes of translation within the technical language are being made; when patients report what a previous doctor said they use such translations with attenuating additions. In the following example the doctor explicidy formulates the problem. (34) [P says that a previous doctor has extracted her acid from the bone with a syringe] A i weis nu nid was er äü meint ich weiß nur nicht, was er wohl meint mit chnochesüüri zieh aber das macht mit 'Knochensäure ziehen', aber das macht das sch vilech het är öpis versuecht das ist vielleicht hat er etwas versucht mundart z säge-n oder (in) Mundart zu sagen, oder "was do' i wüsst nid was da ich wüßte nicht aso i bi au nid nachecho also ich bin auch nicht dahintergekommen (23,8)

116

IwarWerlen

Öpis mmdart säge means for the doctor to translate it for his patient; the doctor perceives that his colleague has unsuccessfully used a technique he himself accepts by tacidy assuming it. Reports written by colleagues (in SHG of course) are usually reformulated for the patient in dialect: (35) A

är schriibt mir do au si henged im einenachzgi er schreibt mir da auch Sie hätten im 81 emal entzündige vom + ellboge gha (22,1) einmal Entzündungen vom Ellbogen gehabt

There is not one case in this corpus where a doctor would quote in SHG; in item (35) the doctor cites in dialect and indirectly by using a conjunctive form. Another example may suffice to illustrate the abundance of lexical shiftings. (36) A

. . . isch dä schmärz z beüflusse dür + ist dieser Schmen zu beeinflussen durch zeigt dä schwankchige am morge we der zeigt der Schwankungen am Morgen wenn Sie ufschtööt zum biischpiu heit der ne denn scho aufstehen zum Beispiel, haben Sie ihn dann schon oder isch er denn sogar am schlimschte zeigt er oder ist er dann sogar am schlimmsten. Zeigt er irgendwie e verlouf dür e tag irgendwie einen Verlauf durch den Tag was verschtercht ne was schwecht ne ab (19,4) was verstärkt ihn, was schwächt ihn ah?

The entire turn clearly shows shifting of words and constructions like schwankige, zeigt er... e verlouf, isch.. .z beüflusse, verschtercht, schwecht ab. The difference is evident if we look at patients who describe their pain without such shiftings: (37) P

ja de gschpüür i dä cheib u näär git s ja dann spüre ich den Kerl und nachher gibt es de plötzlech aube wider e schlag dann plötzlich jeweils wieder einen Schlag ez isch s auso so passiert das das ig jetzt ist es also so passiert, daß daß ich irgendwie i der nacht oder eso chrampf übercho irgendwie in der Nacht oder so Krampf bekommen ha bis i d bei aabe do u und derno habe bis in die Beine hinunter da und demnach e beweegig gmacht und dis eine Bewegung gemacht (habe) und dieses

Swiss Gemun Dialects and Swiss Standard High German

117

i weles bei in welches

nvor auem i ds rächte Bein vor allem in s rechte -houptsä i rächte hauptsä(chlich) in (den) rechten im rächte jo ja aber links ziet dä cheib im rechten ja, ja aber links zieht der Kerl iz o es bitzi M (17,12) jetzt auch ein wenig an

This patient - who sees a doctor for the first time in his life - uses cheib to describe the pain he is feeling. cheib is considered to be a coarse word which is usually avoided. From that point of view the patient is an exception. Syntactic forms dose to SHG are also found. An example is item (36) above. Another very clear case is shown in the foiiowing passage: (38) [P P A P

talks about his brothers and sisters] di sii beidsiitig i de hüft operiert die sind beidseitig in den Hüften operiert mh' + wovon eine isch gschtorbe iz (10,35) wovon eine ist gestorben jetzt

The construction wovon is impossible in dialect; correctly the relative clause should have been introduced by vo dene. We conclude that these variations are very frequent with both doctors and patients; lexical shifting is prominent, phonetic shifting rather are. 5.2.2 Switching into SHG Switching into SHG is very rare in this corpus. If we disregard the case of doctors reading out in an undertone what they are writing - which happens several times - we only find one clear case and two unclear ones. We will Start with the latter: (39) [A and P are talking about F s hobby: beekeeping] A da hei mer no [unintelligible] da haben wir noch P irgend eh d ruu irgend eh die Ruhe A wie d nie wie, die Ruhe? P d rue je rujiger as me-n isch die Ruhe, je ruhiger daß man ist A ja P je weniger schtiche het mer (2,36) je weniger Stiche hat man

118

IwarWerlen

RUH could be understood as a SHG-form which the doctor first does not recognize. On his request P repeats the word in a dialectal version. (40) A

dihr heit vorhär no gseit bim böüge

Sie haben vorhin noch gesagt beim Beugen vom chopf heiget der o schmärze (3,13)

vom Kopf hätten Sie auch Schmerzen

Böüge in this form is SHG, but we probly have to assume a not totally successfull translation. A simüar example is P's schaurige schmärz (schauriger Schmerz). SHG schaurigis translated wrongly with o«instead of a long HH.

The following extract shows the best example of code-switching in the corpus: (41) A

eh vo bruef syt dihr bahnmeischter

von Beruf sind Sie Bahnmeister? P

jawohl

A

was macht me da genau

P

das isch eso i haa d überwachig +

was macht man da genau? das ist so, ich habe die Ueberwachung für houptsächlech mi houptuf

für hauptsächlich meine A P

p gab -

Hauptaufgabe Lja

-1

isch gleis u weicheaalage

ist Gleis- und Weichenanlagen A

uhm

P

ernöüerung erwaeterung und unterhalt

Erneuerung, Erweiterung und Unterhalt vo gleis und weicheaalage (10,25f.)

von Gleis- und Weichenanlagen [a few instances later] A

das isch eh eh ufsichtsarbeit

das ist

Aufsichtsarbeit

p eigetlich und planigs und ja ja ja ja

eigentlich und Planungs- und P

L ja jawohl ja bewachung planung und bewachung he (10,19)

Repeatedly P uses SHG terms like Erneuerung, Erweiterung, Unterhalt, Planung, Bewachung in their SHG form to describe his duties at work. The investigator-situation has already shown that the mentioning of professions leads to shifting. In the medical interview setting we find the same form of shifting (only -/«-forms do not occur) whenever professions are mentioned. The following extract adds to these examples an item with a slight switching character:

Swiss German Dialects and Swiss Standard High German (42) A B

119

was sind si vo bruef

was sind Sie von Beruß

ich bi fernmeldeassischtänt (16,14)

ich hin Femmeldeassistent

F. Roncoroni-Waser noted in her transcription that this is the only time the patient uses ich instead of i; so this is a form of switching into SHG (although without[g] for /x/). The word Temmeldeassistent'is phonetically shifted, in dialectal form ist would have to be fäämmäudeassischtänt where the word itself of course is still not dialect. 5.3. Functions of Variation in medical Interviews

Compared to the investigator-setting medical Interviews more frequently show shifting, particularly on the lexical level, switching being comparatively rare. The function of code-shifting is, on the one hand, part of the doctor's effort to make himself understood to patients not knowing his technical code. On the other hand he is forced to translate the patient's statements into an SHG form for purposes of writing. For patients codeshifting can be understood as a process of assimilation to a specific Situation. A patient can be faced with conflicting demands, the Situation asking for technical terms to verbalise his pain while his abilities for doing so are rather poor. By repeating the doctor's technical terms and by relying on previous experience with similar situations a patient may prove to be competent and cooperative. Here, code-shifting is a way of self-display. The same applies to code-shifting in questions about a patient's professional skills. Here the patient can build up the ritual balance by facing the doctor as an expert too who uses technical language to communicate successfully. The function of the single case of code-switching in the medical Interviews is that of objectivisation. In item (41) the patient quotes his duties at work in SHG and thereby legitimises his work objectively.

6. Conclusions The forms and functions of linguistic Variation sketched for two different settings ought to show that within a language Community looked upon as monolingual there is linguistic Variation which is relevant to conversation. I know that many questions (especially those concerning code-shifting) have to be taken up again. This is due to the fact that there are no norms for

120

IwarWerlen

the dialect as there are for SHG (at least tentatively). Therefore, the Interpretation of phonetic, lexical, morphological and syntactic shifting is not always conclusive. However, as there are no fruitful attempts better solving that problem available for the Swiss German dialects yet, I had to fill the gap as well as I could. To detect and describe the precise conditions on Variation is still a task for communicative dialectology.

References Ammon, Ulrich (1973). Dialekt, soziale Ungleichheit und Schule. 2nd edn. Weinheim and Basel: Beltz. Auer, J. C. P. (1984). Code-shifting: Phonologische und konversationelle Aspekte von Standard-Dialekt-Kontinua. Konstanz: Sonderforschungsbereich 99. (Nr. 88). Auer, J. C. P. (1985). O n the meaning of conversational code-switching. In P. Auer and A. di Luzio (eds), Interpretive Sociolinguistics, 87-112. Tübingen, Narr. Auer, J. C. P. and di Luzio, Aldo (1982). On structure and meaning of Variation in the Speech ofltalian mi^ant children in Germany. Konstanz: Sonderforschungsbereich 99. (Nr. 74). Baumgartner, Heinrich (1940). Stadtmundart. Stadtmundart und Landmundart. Bern. Felix, Sascha and Kühl, Dagmar (1982). Hierarchien phonologischer Regeln bei Dialektsprechern. In Th. Vennemann (ed), Silben, Segmente, Akzente, 61-86. Tübingen Niemeyer. Gumperz, John J. (1982). Discourse Strategies. Cambridge: University Press, (esp. chp. 4: Conversational code switching, 59-99). Haas, Walter (1981). Das Wörterbuch der schweizerdeutschen Sprache. Frauenfeld: Huber. Haas, Walter (1982). Die deutschsprachige Schweiz. In R. Schläpfer (ed), Die viersprachige Schweiz, 71-160. Zürich: Benziger. Ris, Roland (1979). Dialekte und Einheitssprache in der deutschen Schweiz. International Journal of the Sociology of Language 21, 41-61. Schwarzenbach, Rudolf (1969). Die Stellung der Mundart in der deutschsprachigen Schweiz. Frauenfeld: Huber. Sieber, Peter and Sitta, Horst (1984). Schweizerdeutsch zwischen Dialekt und Sprache. Kwartalnik Neofilologiczny 31, 3-40. Stirnemann, Knut (1980). Zur Syntax des gesprochenen Schweizer Hochdeutschen. Frauenfeld, Huber.

Swiss German Dialects and Swiss Standard High German

121

Strübin, Eduard (1976). Zur deutschschweizerischen Umgangssprache. Schweizerisches Archiv für Volkskunde 71., 97-145. Weber, Daniel Erich (1984). Sprach- und Mundartpflege in der deutschsprachigen Schweiz. Frauenfeld: Huber. Werlen, Erika (1984). Studien zur Datenerhebung in der Dialektologie. Wiesbaden: Steiner. (ZDL-Beihefte 46). Werlen, Iwar (1980). R im Schweizerdeutschen. Zeitschrift für Dialektologie und Linguistik 47, 52-76. Werlen, Iwar (1983). Stand und Tendenzen der Domänenverteilung zwischen Dialekt und deutscher Standardsprache (Südliche Hälfte). In Dialektologie. Ein Handbuch zur deutschen und allgemeinen Dialektforschung. 1. Halbband, 1130-1136. Berlin: de Gruyter. Werlen, Iwar (1984). Ritual und Sprache. Tübingen: Narr. Werlen, Iwar (1985). Die Einschätzung schweizerdeutscher Dialekte. In I. Werlen (ed.), Probleme der schweizerischen Dialektologie. Freiburg: Universitätsverlag. Werlen, Iwar (1985a). Gebrauch und Bedeutung der Modalverben in alemannischen Dialekten. Wiesbaden: Steiner. (ZDL-Beihefte 49). Wolfensberger, Heinz (1967). Mundartwandelim 20. Jahrhundert. Frauenfeld: Huber.

Appendix: English translations (1) (2)

(3) (4)

(5)

(6)

S 1 S 1

I'm a lab assistant . . simply there is a psychologist and a psychiatrist with us and two practitioners who are students " and S2 L a psychologist and a psychiatrist yes exacdy but probably it's just a sort of curiosity, isn't it, that you simply w a n t . . . S 1 just so, you're into psychology, I suppose (3 sec.) S 2 do you S 1 yes S 2 aha + (laughing) false assumption I was thinking that you were into psychology S 1 ah no (laughing) that's a false assumption too S 2 no I think, it's right in any case that we talk because eh - well that we speak to each other a sort of quasi faceless speaking to each other, I think, this has a certain justification I could imagine anyway S 2 well now I shali teil you but you know I always teil it such that people first have to puzzle. I am a floor teaonician at the Geological Institute S 1 floor

Iwar Werlen

122

S2 tectomcian 5 1 tectonician (...) 5 2 I mean I simply clean the Geological Institute - so I'm a worker - am I not well from the 5 1 aha - - that's just been another proof I mean you know way you talk - really I would have guessed you are 5 2 you would have guessed I am what S1 .eh - - a Student (7) S 1 I was simply introduced here - they told me the experimentor will come who is probably you (8) 5 1 well, this is certainly intended that it is like that isn't it (9) 5 2 exactly we are the Interviewers that's fantastic isn't it fantastic isn't it (10) 5 1 this is also stupid this game isn't it 5 2 no it's not stupid I think it's one of the best games you know there really is none which is more interesting . it drives you mad doesn't it S1 (11) S 2 but look you have to take into account that eh that you son of - can never totally capture the human psyche or don't you have this feeling (12) S 2 don't you think we should go 5 1 let's be going come on 5 2 ( . . . ) certainly nobody wiU come 5 1 now we wait for Gödot 5 2 for who . well no S1 (13) 5 1 (14) (15) 5 2 (16) S 2 51 52 51 52 51 52 (17) S 1

(18) (19) (20) (21)

S2 51 52 51 52 S1 S1 S1 S1

no we have to do something here together, don't we, let's go on but always the resuk is that the problem isn't denied from the very Start I first asked you otherwise I would have said come on let's take off this they talk to each other - but they don't have how do you say don't " know .feedback no feedback - you can simply no vis-ä-vis simply no - eh no reaction - 1 don't like these words eh they drum them into your head you can - after that they automatically come to your mind yes feedback that means to feed back yes - it's a kind of feeding without feedback you probably die I think after that you always wonder when you keep seeing just the eyes and the hair and the arms what the whole looks like you know without (...) yes (5 sec) eh that's ah yes that's probably why probably oh really, you do, you direct one no I do not direct one 1 don't think I would be apt for that I have a feeling it's called Spätholz (the book) what I like to listen to as well is ah is the news and Echo der Zeit so - let's wait for Godot - hm

Swiss German Dialects and Swiss Standard High German S2 (22) 5 1 52 51 52 51 52 S1 (23) S 1 (24) S 1

(25) S 1 S2 (26) S 1 (27) (28) (29)

(30)

(31)

A P A P (32) P A P A P (33) A P A P (34) A

123

yes let's give him some more rime so let's wait for things to come oh is somebody eise going to come well I don't have a clue - - I thought you were the experimentor did you yes no, I really don't know anything well that's fine well - let's wait sort of foUowing the motto you would say nothing new under the sun - when looking at politics and you compare it to history he and then you will also - find a good excuse accordingly that one is not Willing to learn it — and say for instance — I think it is ridiculous or I think it is stupid what they do that is (5.0 sec.) chaqu'un ä son goüt - - ( . . . ) yes when you can't cope with it you simply say yes chaqu'un ä son goüt let him do as he likes what do you do eh I'm a Student of medicin that's terrible but he also has complaints about breathing doesn't he he is a bit a thick one yes I was dispensed from lifting h heavy things you see or let's figure out going to the gymnastics three times a week or I don't know I simply Started to do things again that strain the back that it so that it would come again do you have, at work do you have difficulties ' at times . because of that at times I do it's just a handicap So Mr X Y - what is your problem why have you come here? eh my problem is simply that increasingly I have a backache lately and that I would have liked to know what what is going on what isn't normal with my back, that is I have been knowing a number of things because my my father was a doCTor hm and when I was a child we examined this and then I think then their diagnosis was open vertebra bows hm and äs a child I wasn't troubled by that actuaUy it simply almost goes down to the bottom hm it itches - so it is an itching pain yes and then it sometimes goes down simply all through the leg and was it a thick cover then or yes there is simply like a like a rüUfli aha yes like a rüüfli it's only I don't know what he means by extracting the acid from the bone but that doesn't that is maybe he wanted to say something in dialect you know I

124

Iwar Werlen wouldn't know

P (35) A (36) A

(37) P

A P (38) P A P (39) A P A P A P (40) A (41) A P A P A P A P

what well I couldn't find out either he writes to me that in 81 you had an inflammation of the - elbow . . . is it possible to have an influenae on the pain by - are there fluctuations for instance in the morning when you get up do you have it already or perhaps is it even worst then does it show a kind of progression over the day what intensifies it, what mitigates it yes then I can feel this chap and then there suddenly is a beat again lately it happened like that that that somehow in the night I got a cramp down to the legs there and after that made a movement and this in which leg above aU in the right one above all into the right one in the right one yes but left this chap is about to pull a linle too they had a hip Operation on both sides hm from which one of them died now there we still have 1" ( . . . ) L some eh tranquillity what, tranquillity? tranquillity, the calmer you are yes the less bee stings you get before you also said that you had pains when bowing the head eh by profession you are a rail worker that's right what exacdy does one have to do there it's like this I have the supervision - for mainly my main job L yes is the line and point installations uhm rebuilding extension and maintenance of line and point installations

(• • •) A is that eh eh supervision work essentially and planning and eh eh eh P yes correct yes supervision planning and supervision he (42) A what is your profession P I'm a telecommunication assistant

J . PETER FRENCH,

York

Word Final /r/ in a Northern English Accent: An Interactional Account of Variable Production"" 0. Introduction It is widely recognised that during the eighteenth Century a sound change began in English concerning the consonant Irl. This change, referred to as 'R dropping' (Wells 1982a: 218), involved the elision of Irl in nonprevocalic positions. Different accents of the British Isles have incorporated R dropping to different extents. At one extreme, there are accents such as those of the South West of England, East and West Lancashire, most of Scotland and Ireland, which have remained unaffected by the change. These accents, often termed 'rhotic', preserve the pronunciation of Irl both preconsonantally in, for example, start. North and farm and word-finally in such as star, moor and hetter. At the other extreme, there are accents which have undergone the change fuUy and have lost both preconsonantal and wordfinal /r/. These, the 'non-rhotic' accents, include Received Pronunciation, most of South East England, the Midlands and the largest area of the North'. Intermediate between the two extremes, there exists the much less extensive and lesser known phenomenon of 'semi-rhotic' speech. Wells has applied this term to the accents of Humberside and parts of Lincolnshire as well as that spoken by "fishermen in Flamborough, between Bridlington and Scarborough" in coastal North East Yorkshire. Here, he states, "his-

I am grateful to grateful to Stanley Ellis, David Langford, John Local and Jack Windsor Lewis for helpful comments on an earlier draft of this article. The usual disclaimers apply. Statements about features of accents extending over large socio-geographical areas are, inevitably, idealisations. See, for example, Romaine (1978) and Bertz (1975), respectively, for indications of non-rhotic sub-varieties of Scots and Irish speech. See Windsor Lewis (1979) for examples of the occurrence of pre-consonantal Irl in 'General British' pronun-

126

J. Peter French

torical Irl is lost preconsonantally, but retained (to some extent) in final Position tt . . ( 1 9 8 2 b : 368). I have two purposes in this article, one major, the other minor. My minor purpose is to suggest that this 'semi-rhoticity' is rather more widespread in North East Yorkshire than has generally been recognised, extending far inland among the older agricultural population^. My major purpose is to demonstrate that there are rather precise constraints upon the occurrence of word-final Irl. These constraints, however, are not to be found in the types of phonotactic or syllabic conditions that phonologists traditionally look towards in formulating rule sets for vowel and consonant distributions; nor do they consist in the sorts of 'situational variable' which sociolinguists have used to explain speakers' alternation between phonological forms. Rather, they concern aspects of the Organisation of conversational turn-taking. Some methodological implications of this finding are discussed in the final section of the article where it is proposed that, in Order to provide satisfactory explanations of the occurrence of certain ranges of phonetic features, linguists may have to show a willingness to take greater account of language as social interaction.

1. Data The analysis is based upon recordings made in the North Yorkshire village of Settrington. The village is situated on the Northern edge of the Wolds, three miles East South East of the town of Malton, some twenty miles from the coast. The recordings are of interviews on the topics traditionally investigated within dialect studies (outmoded agricultural processes, implements, etc.) and last in total around ninety minutes. Whilst only one informant is used - a retired farm worker who has lived in the region all his life - 1 have encountered the patterns described here in the speech of several older inhabitants of the area and believe them to be of some generality.

Although this point is most frequently overlooked in general statements about regional accent Variation, non-prevocalic Irl is, in fact, noted in the Survey of English Dialects (Orton etal{eds.), 1962-1971) for areas of East and North Yorkshire. Stanley Ellis (pers. comm.) informs me that a special remark was inserted into the S.E.D. fieldnotes for these regions by Dr Peter Wright stating that it may occur as either an r-coloured vowel or a distinct post-vocalic segment. He also teils me that the feature was even more apparent in the so far unpublished tape recordings he coUected in the area.

World Final /r/ in a Northern English Accent

127

2. Analysis 2.1. Preconsonantal Irl The characterisation 'semi-rhotic' provides a rough, rather than exact, fit with the distribution of Irl in the speech of the informant. Whilst Irl is not produced preconsonantally in the overwhelming majority of words, there is one word where it does occur in this position. One finds Irl absent from, for example, tum, form, curve, com, course, horse, harvest, yard, dark and Murton (place name): (I) (3) (5) (7) (9)

turnsit lovely curve of course harvest time dark

[tz:n?i^t] [lövle'"^ :khz:v] [avko's] [aivi-sthai'm] [dp:k]

(2) (4) (6) (8) (10)

farming museum com the horse rake foldyard Murton

[farmi'myöyzjsm] [ko'n] [djo:'sje:'k] [fp:ldja:d] [m3:'i?]

But, on two productions, Irl is present in fortunate: (II)

[fD^/anan]

The realisation of Irl here differs markedly from the norm for the speaker's prevocalic realisations. In rake, from, straight, correct, ground, heakfast, break, drawing, brayed and previous, the place of articulation is clearly post-alveolar: (12) (14) (16) (18) (20)

rake straight ground break brayed

[je:'k] [sJjeA] [gjeünd'] [bje:'k] [bje:'d]

(13) (15) (17) (19) (21)

from correct breakfast drawing previous

[fjom] [k'jek~it] [bjekfasit] [d'JD:''n] [^.leivjas]

However, in both instances of fortunate, Irl shows a marked degree of retroflexion, a feature which begins during the production of and 'colours' the preceding voweP. 2.2. Word-final/r/ Word-finally, the Situation is that described by Wells: Irl certainly is present "to some extent". However, its occurrence is by no means random and my concern is to establish the specific conditions under which it is ' The occurrence of preconsonantal /r/ in fortunate is an example of the kind of accent feature Wells (1970,1982) has termed 'lexical-incidential'. Such features are peculiar to restriaed sets of words and cannot be stated in terms of general phonological rules.

128

J. Peter French

found. The main thrust of the analysis will be thatfinalIrl is used to fulfil a specific social-interactional fiinction: that of signalling non-completion of speaking turns at points of pausing and hesitation. The grounds for attributing this significance to final Irl become clear from the foUowing examination of the data. First, it may be noted that /r/ does not occur word-finally where there is no pause before the onset of the following word. Thus, one finds no final Irl in manure in (22) and (23) where phonation is continuous through the constructions manure leading and manure was: (22) they would set you on manure leading... [m'njoijaleidin] (23) all the manure was made [m'njoij 3W3zmed]

But in (24) where manure isfinalin a tone unit and is foUowed by a pause and inbreath, final /r/ is produced: (24) and made you manure (0.7 secs.) and and you led this . . . [manjoijaj-] [sn'raja]

Similarly, in the unbroken sequence farmer would in (25),finalIrl is absent from farmer. (25) some old farmer would spot you . . .

[fa:m3wü'~i(J]

However, in (26), where farmer occurs directly prior to a pause,final/r/is again produced: (26) to that farmer (1.0 secs.) and e r . . . [fa:mar] [an'e']

In both (24) and (26) the pause after final Irl is followed by a word beginning with a vowel {and). And it is, of course, widely known that non-rhotic speakers produce final 'linking Irl' prior to words with vocalic onset: eg. better or worse, richer and richer, bigger of the two. However, the examples of final Irl found here are not to be regarded as instances of linking Irl which have been disjoined from following vowels by Speaker pauses. They do not fit this analysis for two reasons. First, they differ in place of articulation from linking /r/. WhilstfinalIrl in (24) and (26) is, like the preconsonantal Irl in fortunate, retroflex and preceded by a coloured vowel, the speaker's Standard linking Irl is post-alveolar and not preceded by anticipatory colouring:

World Final Irl in a Northern English Accent (27) forit (29) Shoulder and (31) your(h)orses

[fojit] [Jeü'dajan] [jajoisis]

129 (28) more of (30) overand (32) were out

[mo:J3v] [ovajan] [waj^üt]

Second, pre-pausal final /r/ is also produced where the word to follow begins with a consonant, as in (33) - (37): (33) all the year (0.9 secs.) but in harvest month . . . [öaji-ar] [bot] (34) that particular month September (1.3 secs.) when you would . . . [sep^itembsf] [wenjs] (35) stock was in there (0.9 secs.) messed amongst i t . . . [möe:rr] [mestamoqstit] (36) you were hired for the year (1.5 secs.) for a . . . [öajisj-] [fajs]

Thus, the occurrence of these word-final /r/s cannot be explained by reference to vowel Haison. Instead, their functional significance becomes clear as a social-interactional one when one compares the above examples with instances of speech where there is a potential for word-final Irl at the end of a speaking turn, as in (37) - (39) where the informant's turns end with the words year, chair and farmer. (37) Inf: . . . all that overtime you put in was m with that fifteen pound that you were hired for for a year [fo'fDjgjia] (0.6 secs.) Int: and what time are you talking about now what year would this be? (38) Inf: fetch yourself a chair [3tJe:B] (0.8 secs.) Int: thanks (3.0 secs.) Int: some picture of erm . . . (39) Inf: . . . and that fastening money that fest as we called it-(where-) i»always went by the name of fest ( ) er were- sealed your contract with that farmer [öaf fa:mB] (14 secs.) Int: did you ever live in the farms or did you . . .

In these, and other comparable cases, final /r/ is not produced. Word-final retrofiex /r/, then, is exclusively associated with positions directly prior to intra-turn pauses and, in the data under consideration.

130

J. Peter French

invariably occurs in this position where the potential for its articulation exists. It thereby operates as a signal not only that a pause is to occur, but that the turn is incomplete and that further talk is to follow'*.

3. Discussion In describing distributions of vowels and consonants in particular language varieties, phonologists have formulated phonotactic rules and tables which specify co-occurrence possibilities and restrictions within words and syllables. An example of such a rule for the accent under study is Wells's already quoted Statement: "historical Irl is lost preconsonantally, but retained (to some extent) in final position # . . ( 1 9 8 2 b : 368). The rule is, of course, stated in 'rough form', and the condition "to some extent" may be seen as needing further attention. Presendy, there are two well established options available to anyone wishing to refine the rule. The first, which is provided within the conceptual framework of phonology itself, would be to consider the occurrence of final /r/ present and final /r/ absent versions of the same words (eg. manure, farmer) as constituting sufficient grounds for deeming the phenomenon unsystematic and therefore beyond further comprehension. The 'refinement' to the rule produced by following this option could be seen as stylistic rather than substantive: 'final /r/ occurs infi-eeVariation with its own absence'. The second established option would involve looking beyond the speaker's phonological system for the possibility of situational factors governing selections from the system. Correlational sociolinguistic studies by Labov (1972), Trudgill (1974) and others have demonstrated that pronunciations in given accent communities vary, not only across subgroups of the popuIation in accordance with social class, age, gender and so on, but also within individuals - for instance, in response to different degrees of situational formality. Increased formality tends to result in a decrease in the frequency of the most regionally marked, non-standard phonological forms. Brought to bear upon the present data, this type of

Further evidence for this function of final Irl might involve finding cases where it is produced before a pause whereupon another Speaker attempts to take the floor. One would predict thefirstSpeaker would treat the other Speaker as making an interruption. It is perhaps because the Interviews were conducted in accordance with the guidelines for linguistic fieldwork which dictate that informants should be given maximum opportunities to talk that no such instances occur in the recordings.

World Final Irl in a Northern English Accent

131

analysis might involve one locating versions of words like farmer and manure with final /r/ as the non-standard forms and surveying the Interviews for evidence that fluctuations in final Irl frequency corresponded to shifts in the formality of the proceedings. If this land of anlalysis were to prove successful, and evidence of such correlations were found, then it would be possible to augment the phonotactic rule with information about situational variables which regulate its operation. As far as it is possible to determine, there is no such evidence, and this analytic option, like the first one, would fail to reveal a principled or systematic basis for final Irl alternation. In order to reveal this basis, the method adopted in the present analysis involved examining the data for possible associations between phonological forms and interactional activities. That is, the relevant questions turned out to be not 'What are the system-based constraints upon final /r/?' or 'What social information is the Speaker responding to in his alternations?', but 'What conversational task is being performed in producing/not producing the segment?; What information of an interactional kind is the Speaker signalling to his interlocutor in using or witholding it?' As yet, very little is known about the extent to which interactional information is carried by phonetic features. However, the little work that is available in this area suggests that interactional concerns may enter into phonological selections to a greater degree than one might intuitively judge to be the case. In particular, a recent paper by Local & Kelly reports that in an East London accent, not only are there particular distinguishing features of approximant consonants in the word we//when it introduces reported Speech and others when it is used in the environment of interactional repair (Speaker self-correction), but also that glottal closure sustained across silences seems "to correlate with holding of turns and the projection that there will be further talk to foUow" (Local & Kelly 1986)5. An informal survey of the present data suggests that the held glottal stop as a turn holding marker is not confined to East London speech, but occurs in For studies of the role of suprasegmental phonological features in signalling interactional information see French & Local (1983; 1985), Local, Wells & Sebba (1985) and Local, Kelly & Wells (forthcoming). The first two studies examine the use of prosodic features (principally polysyllabic pitch height and loudness) in signalling that interruptive speech is competitive for a conversational turn. The study by Local, Wells & Sebba (1984) focuses upon the uses of intonation in indicating that speaking turns are complete in London Jamaican English, and that by Local, Kelly & Wells points to the ways in which prosodic and segmental features (vowel centralisation and aspiration of final fortis plosives) interrelate in signalling turn finality in the speech of urban Tynesiders.

132

J. Peter French

North East Yorkshire too. Obviously, many intra-turn pauses occur after words where there is no potential for the articulation of final /r/, and it is to be expected that in these cases the functional bürden of final /r/ is carried by other phonetic elements^. The exact ränge of phonetic resources implicated in carrying this type of interactional Information has not yet been charted for any variety of English. However, it is clear that, if linguists are to establish füll accounts of the principles governing phonological choices, they will need to develop a greater awareness of language as social interaction^ One would not, of course, wish to propose that researchers of phonological patterning give exclusive attention to the possibility of interactionally related selections. It is abundantly clear that both systemic and situational factors form significant constraints upon phonological choices. In fact, the present proposal is a relatively modest one; it is simply that we add the interactional line of analysis to our stock of available options.

References Bertz, S. (1975). Der Dubliner Stadtdialekt. Eine synchronische Beschreibung der Struktur und Variabilität des heutigen Dubliner Englischen. I. Phonologie. Doctoral dissertation, University of Freiburg i. Br. French, J. P. & Local, J. (1983). Turn-competitive mcorrnngs. Journal of Pragmatics, 7 i 17-38. French, J. P. & Local, J. (1985). Prosodic features and the management of interruptions in English conversation. In C. Johns-Lewis (ed.), Intonation in Discourse. London: Croom Helm. Labov, W. (1972). The study of language in its social context. In P. Giglioli (ed.), Language and Social Context. Harmondsworth: Penguin. Levinson, S. (1983). Pragmatics. Cambridge: Cambridge University Press. Given that speakers often wish to pause (for breath, for utterance planning, for rhetorical effect and so on), then one would expect there to be mechanisms for making their intentions known to their interactional partners, and thereby avoiding the Situation where another Speaker takes over before the first has finished (Sacks, SchegloffficJefferson 1974). There is, of course, a rapidly growing body of literature on the Organisation and management of social interaction. This work, termed 'conversation analysis', originated within the discipline of sociology, rather than linguistics. However, the relevance of social interactional concerns to linguists is receiving growing recognition, and a useful summary of work in this area, together with an explanation of its relationship to linguistic study, can be found in Levinson (1983; Ch. 6).

World Final /r/ in a Northern English Accent

133

Local, J. & Kelly,]. (1986). Projection and 'silences': notes on phonetic detail and conversational structure. In G. Button, P. Drew & J. Heritage (eds.) Special edition of Human Studies. Local, J., Kelly, J. & Wells, W. (forthcoming). Towards a phonology of conversation: turn-taking in Tyneside English. To appear in Journal of Linguistics. Local, J., Wells, W. & Sebba, M. (1985). Phonology for conversation: phonetic aspects of turn-delimitation in London Jamaican. Journal of Pragmatics 9, iii, 309-330. Orton, H. et al (eds.) (1962-1971). Survey of English Dialeas Vols. 1-4. Leeds: Arnold. Romaine, S. (1978). Postvocalic Irl in Scottish English: sound change in prog-ess? In P. Trudgill (ed.) Sociolinguistic Pattems in British English. London: Arnold. Sacks, H., Schegloff, E. & Jefferson, G. (1974). A simplest systematics for the Organisation of turn-taking for conversation. Language 50,696-735. Trudgill, P. (1974). The Social Differentiation of Speech in Norwich Cambridge: Cambridge University Press. Wells, J. C. (1970). Local accents in England and ^^Xts. Journal ofLinguistics 6, 231-252. Wells,J. C. Accents of English Vol. 1: An Introduction. Cambridge: Cambridge University Press. Wells, J. C. (1982b). Accents of English Vol. 2: The British Isles. Cambridge: Cambridge University Press. Windsor Lewis, J. (1979). Pre-consonantal Irl in General British pronunciation. English Language Teaching Journal XXXIII, 3, 188-190.

O T T O STERN,

Zürich

Divergence and Convergence of Dialect and Standard from the Perspective of the Language Learner Standard language acquisition by the Swiss-German dialect speaking child

1. Introduction The controversial question, whether or not the German Standard language is a foreign language for Swiss-Germans, will never be answered in a satisfying way, since it concerns the speaker's subjective judgement on the degree of divergence of linguistic varieties. Because of differing linguistic biographies, these judgements will always vary gready from one Speaker to another. In recent years it was Haas (1982) who, under the heading of Mundart und Standardsprache, dealt with the question of linguistic divergence. He argues that an exhausting description of the differences between the dialect and the Standard is not possible, because of the great linguistic Variation within the Swiss-German dialects themselves. Therefore, in his explanations of divergence, he limits himself to a few examples which, however, should be typical for a large number of Swiss dialects (p. 93). The following general statements are based on Haas' introductory characterization of the different linguistic levels': Phonology: The greatest differences between Swiss-German (dialects) and the German Standard language (henceforth SG and StG) are found on the phonological level, the phonemes differing not only in quality but also in number (p. 93). Morphology: The inflectional system of StG is considered to be archaic and complicated. SG morphology is much simpler, although it has not yet • The following citations (translated by the äuthor) stand for a more detailed introduction to the linguistic Situation of German speaking Switzerland, exceeding the purpose of this paper. For more details cf. Lötscher (1983) and Sieber & Sitta (1984).

Divergence and Convergence from the Perspective of the Language Learner

135

reached the simplicity of English (p. 94). The StG relative pronoun der! die/das, e.g., is subject to gender and case inflection, whereas the only SG relative pronoun wo is invariable in all contexts. Syntax: The comparison between SG and StG syntax is difficult, because very often spoken SG is compared to written StG, with the result that typical traits of a spoken or written variety are considered to be differences between SG and StG (p. 96). Lexicon: One can speak of the fantastic richness of the lexicon of the SG dialects only if all the words of all the different dialects are taken together. One dialect alone is, of course, much poorer. The still unfinished Schweizerdeutsche Wörterbuch arrives at the astonishing number of 100,000 words only because of the numerous local synonyms, e. g. the about 75 regional expressions for Maikäfer ('lady bug'). A comparison with StG does not make much sense because every word found in written texts is considered to be part of it independent of its regional origin, i. e. whether it comes from Munich or from Hamburg. On the other hand it is false to claim that the dialect is poor in abstract expressions of scientific use because SG is the spoken language of everyone independent of social background and position. If a Swiss medical doctor speaks to his colleague of wi.ssi bluetkörperli (StG: weisse Blutkörperchen; 'leucocyte'), then this must be a dialectal expression; what eise could it be! Every language following strictly the purists' prescriptions would immediately die out. Thus, SG has exactly as many words as its speakers are using, and it enables them to borrow and adapt phonetically all those words they need. For most of the frequent SG words there exists a corresponding one in StG differing only in phonological details (Haas 1982: 98). The following citation from Schwarzenbach & Sitta (1983) concerning the status of SG as a language may add to a better understanding of the specific linguistic Situation of German speaking Switzerland in comparison with the greater German speaking territory: Whereas in other parts of the German speaking territory a colloquial variety has developed between dialect and Standard, it is missing in German-speaking Switzerland. For the understanding of the specific Situation it is important to note that, in general, speakers of different dialects have no problems in understanding each other, or, if there are major differences between two dialeas, the speakers are able to modify their speech for successful communicarion. Since the dialects are the only means of communication across different regions as well as across different social groups, dialects have a high prestige, i. e. they are evaluated positively and accepted in every Situation of public life (p. 64, my translation).

136

Otto Stem

Thus, German speakers are able to continuously move between a Standard and a dialect pole, but for Swiss-Germans this means a considerable jump that is feit like switching from one language to another, especially by the less trained speakers. How big is this 'jump'? How big is it for the least trained speakers, i. e. for children in kindergarten and in first grade who are about to make their first steps into the new variety? This paper tries to answer some of these questions. But, instead of doing so by a contrastive analysis of the two linguistic systems, we are going tp analyze processes of language acquisition. Thus, we are looking at the problem of divergence and convergence of dialect and Standard from the perspective of the learner. The investigation is based on data gathered in a number of pilot studies carried out by students of the Deutsches Seminaroi Zürich University (cf. references: Seminararbeiten), a Lizentiatsarbeithy Erny (1984) comparing text comprehension in SG and StG, and a corpus compiled by Stern (1978).

2. Spontaneous production of Standard German At play in kindergarten children very often use single expressions or entire utterances in StG, especially in situations of spontaneous role playing, e. g. (1) when calling the Space ship by radio^: ALLES BERÄIT ZUM

ABFLÜ:GEN!

(everything ready for take off)

(2) when playing cowboys: H E N D E H O : C H O D E R IG S C H I : S S E

(hands up or TU shoot)

(3) when playing mother and children: 1 2 3 4

chimb, gönd i s bett, (children, go to bed wän ar gga« händ, gönd ar as mittagsschlä:fli go macha after you 've eaten go and take a (afternoon) nap - - zwar marsch! (...) machsd rasch, marsch! (...) hurry up! do it quickly, hurry up! S O N S T B Ä I S T Öig DER H U N D ! - - A L L DAS I S T ES SigER. otherwise the dog '11 bite you! all that is sure

For the transcription Conventions cf. p. 154.

Divergence and Convergence from the Perspective of the Language Learner 5

jo, isch würklkh, gälbd. yes, it's true, really)

137

(Stern 1978: 17)

(4) t w o b o y s (P. and R . ) playing reporters; they speak directly into the mike o f the recorder: 1 P:

J A : , JETZT SE:EN SI: DEN R O G E R UND — J A U S L I N .

2

DER T H O : M A S UND ÖI R O G E R HABEN ZU:GLU:GT

3

AN DEM BAUM, KASCHTANIENBAUM.

(yes, now you see Roger and - - Jauslin T.

and

R.

have watched

at the tree chestnut tree 4

PATRICK IST ZU SCHPE:T GEKOMMEN.

P.

has arrived too late

5 R:

X : HI:R - - HAB IK MÄINEM MOTO GERÖSSLET,

6

Ä: UND HA-HAT DER R O G E R JAUSLIN ÄIN

7

AM PATRICK SÄIN WELO GENOMM(EN),

8

P.'s bike taken is ahm, AB DEM hünlihaum - - i hin i vorra abi gheit is down from the somersauk I am I in front fallen

here

I

and has

9 10

11 12 13

have on R.

my

moto raneled J.

a

ah ds WELO HAT ÄIN PURZELBAUM GENOMME,

from the bike has a somersauk taken un nachär HAB ICH ÄIN LOCH IM KÖPF gha:, and then I have had a hole in the head noch(dr) sim-im mitanandar häi, then we went home together u(n) noch(3) is da Patrick mit miy häi, and P. went home with me un nocha - - m nochsr, scRixsss. and then and then the end) (Stern 1978:20)

(5) five girls are playing scenes at the king's court ( C = the child, D = the maid, M = Monika, Q = the Q u e e n , T = Tanja): 1 C: 2 M: 3 C: 4 D: 5

ich hett immar web zum könig. (I always wanted to go to the king he:, du hettsch imnar abs mi.r web na:go:. you wanted always to follow me ja. [they pass the maid who is cleaning up the place] yes go:pf! ir döfad jetz nüm da dun lauß! gosh! you are not allowed to go through here any more iti händ-ar nur da: widar albs varschmbrt. now you messed up everything here again

138

Otto Stern 6 Q:

[commanding] JA! DAS MACHT MAN NI^T!

7 M:

yes you don't do that ou:! ds könig schimpft doch ga.r nü:d, the king does never scold you (...)

17 C: und mi hilf ad doch da königin imma and we are helping the queen always 18 Q: a: ja:! du hilfsch MIR GA:R NI: ! yes and you never ever help me 19

(un)d

s o WIRD ICH VERUCKT!

and so I'm getting mad 20 M:

NÄIN, LASS DAS, DAS KIND!

no 21 Q :

leave that child alone

NÄIN!

no 22 M:

wann

if 23

- - wann

DAS IMMER MI:R WILL NA:GE:N

if she always wams to follow me

DAS IST DOCH WÄIL —

this is because

WENN DAS KIND MICH LI:B HAT.

if that child loves me

2 4 Q : ja, si söt mi.r au hälfa. JETZT T U : T ÄINMAL WAS,

yes. she should help me too. now start doing something schüsch hau dar äis uf d-fmgar! or rll hit you on your fingers 26 M: ja, chum Marlen come on Marlen (the queen's name) 27 du we.rsch nüd so bö:s gsi: zu da chinda. (...) you wouldn't have been so angry with the children) (Stern 1978:20) 25

Evidently, the spontaneous production of StG is an integral part of roleplaying. For the children the roles of queens, cowboys, spaceship crews and reporters are firmly connected with StG. These roles are directly copied from TV which, for most of the preschool children, is the only source of StG. The linguistic forms of the StG acquired from TV, however, are not part of the Standard variety spoken in Switzerland. It is a register of Standard spoken in central and northern Germany, as is evident from articulation, Intonation and idiomatic expressions. This is not surprising, since Swiss TV for children is only cast in dialect (cf. Burger 1984:221), and the movies children watch (on German Channels) are all produced or synchronized in Germany. The role dependency of the use of StG appears clearly in (5). All the discussions about the roles are in SG. There is code switching as soon as the speakers go back to their roles. Especially commands, exclamations and arguments, i. e. authoritative Speech acts, are most likely to be pronounced

Divergence and Convergence from the Perspective of the Language Learner

139

in StG. As a matter of fact, StG words and expressions are used by adults too to add authority to an utterance (cf. the paper by I. Werlen, in this volume). As soon as the children try to produce more than single words or formulaic expressions in StG, we observe instances of code fluctuation, i. e. Single Parameters, such as phonological, morphological or lexical, vary in a non-systematic way between Standard and dialect, making it difficult to distinguish the two codes and the point of switching (cf. Auer 1 9 8 4 : 6 5 ) . Typically, lack of proficiency in the Standard appears in (4) and in ( 5 ; 2 2 ) where StG is used for narrating and explaining respectively. This type of Speech cannot be produced by ready made expressions, but has to be generated step by step from the semantic base.

3. Comprehension and lexical acquisition processes At the age of six a child has acquired a relatively large vocabulary for everyday life situations, although individual differences, due to varying socio-cultural environment, are considerable. The study of the natural acquisition of StG shows that for production and comprehension 6- to 7-year-old children rely strongly on the SG lexical base, transforming the phonological and morphological surface structure of SG into StG. A study of the acquisition of antonymous expressions by Weber & Arter-Lamprecht with SG and StG stimulus words shows very little differences in the responses to the two varieties, i. e. acquisition of a lexical item in one variety makes it available in the other too, as soon as surface correspondence can be established. Comprehension of SG and StG at the kindergarten level (age 6) has also been investigated by Erny ( 1 9 8 4 ) . Two groups of children heard the same Story either in SG or in StG. Comprehension was tested by retelling and questioning, and evaluated by means of a detailed propositional analysis. The results confirm the hypothesis that the linguistic form of the input, i. e. SG or StG, has only a minimal impact on comprehension. The comprehension of narratives depends much more on other variables of text processing than on the linguistic variety used (cf. Stern 1984: 141 f.). Comprehension is the result of the hearer's active generating of the meaning of the text, based on a number of simultaneous processes, which in turn are based on previous knowledge (world knowledge) and on expectations built up during text reception. These hearer directed comprehension processes work in interdependence with text directed processes

140

Otto Stern

like, e.g., comprehending of single words or of the morpho-syntactic structure. The hearer directed processes enable him to bridge momentary comprehension gaps on the text level and to dose them later. It is exactly this capability of bridging that helps children to understand StG texts. Thus, the general degree of development of text comprehension is important for the acquisition of StG. Let US now look at specific phenomena of StG text processing by children. It is known that the attention of the listener of a Störy is greatest at the beginning. Much more details are stored, and later available, at the beginning of a story, whereas towards the end attention is mainly directed to the understanding of the point. Erny (1984) found that children who listened to the SG version of a story conformed much more to this general law of text comprehension than those who heard the StG version: until about the middle of the text the SG hearers had stored more details than their StG partners. However, from the middle to the end the result was quite the opposite: StG hearers - against the general findings of comprehension testing - stored more details than their SG partners. These findings indicate that for the comprehension of StG texts a part of the attention is not available for normal meaning processing. It seems that the children listening to the StG version use part of the available attention span, at least at the beginning, to perform the phonological transformations (see below, section 4) allowing them to discover SG words behind StG forms, and to test whether the meaning found fits into the whole of the story. The problems they are facing have to do directly with the degree of divergence between SG and StG: how far may a StG phonological form differ from SG in order to remain recognizable for the child? The following example, in which the phonological divergence exceeds the band-width of being understandable, stems from a study on the acquisition of the StG preterite (cf. the study by Koch-Niederer mentioned below in section 5). Note that SG has no preterite tense, the only past tense being the perfect formed by means of the auxiliaries sein and haben. In the sentence (sie) trug den Zwerg zum See ('she carried the dwarf to the lake'), most children at the age of 6 translated the StG preterite form trug into SG present tmckt, meaning to push (under water). This shows how flexibly the children adapt the meaning of single words, looking for morphophonological correspondence between StG and SG. In the case of trug, the divergence was to great to find the correct form, the SG verb being trä:ge. Similar processes, of course, are observed in second language learners who often have to proceed with a lesser amount of sizable Information than

Divergence and Convergence from the Perspective of the Language Learner

141

Swiss StG learners. A good example is given by Erny (1984:59f¥.): the two only children of foreign LI in her sample retold the story of the Sehne:glöggli ('snow-drop') as the story of the Sehne:flöckli ('snow flake'). A Turkish boy transformed the StG verb verschomt ('the snow-drop has been spared') into gsehme, in the context of 'the sun shone'. The capacity to cover a large band-width of phonological Variation distinguishes the good/fast learner from the bad/slow one. It is noticed that children with a foreign language background, who learned SG as their L2, differ from Swiss children exactly in the capacity of using the füll bandwidth of morphophonological Variation for the semantic Interpretation of StG. This difference has to do with the differing degree of differentiation of the SG lexicon. A child who for the first time hears, e.g., the StG preterite sang will be able to immediately identify this form as being part of the paradigm of singen ('to sing'), if he already knows at least one of the SG words like gsang ('the singing'), gsangveräin ('choral society') or chilegsangbuech ('hymn-book'), which are the only forms in the semantic field of the verb singe in SG containing the ablaut radical vowel /a/. Note, however, that these words are part of a rather elaborated vocabulary. The capability to learn StG forms for SG words already acquired is considerable. This is shown by a longitudinal investigation (cf. Künzler et al.) on the possibilities to teach StG at kindergarten. In a single lesson the children learned in a playful way 60 StG forms for words they already knew in SG. The degree of divergence of the two linguistic forms varied from great to none, compare e. g. StG: kämm vs. SG: schtred ('comb'), zwiehelvs. zwible ('onion'), auto vs. auto. One week later the children were able to reproduce in StG - by naming pictures and translating - all the words learned without any difficulty. The study of the acquisition of the StG lexicon and the comprehension of StG texts shows that - despite of considerable phonological and lexical differences - most children have easily access to the StG variety. They acquire rapidly the capability to discover familiar words under a divergent phonological form. We think that this capability develops naturally in contact with the varying SG dialects during first language acquisition. Erny (1984:102) observes that her children were used to hearing stories told in a dialect differing considerably from their own Toggenburg/St. Gallen variety, since the teacher spoke Zürichdeutsch. Listening to different SG dialects obviously prepares for the understanding of the phonological Variation of StG. The capacity to deal with dialectal Variation in production also develops early. As a matter of fact, in kindergartens of bigger city agglomerations.

142

Otto Stern

mixed groups of children with very different dialectal backgrounds speak among themselves a relatively homogeneous local dialect, even though their teacher speaks still another, sometimes markedly different one (cf. Stern 1978). Summarizing the findings on the comprehension capacity in StG we can say that there is füll transfer of this capacity from SG. The children acquire phonological and morphological transformation rules, making it possible for them to understand StG forms on the base of the SG lexicon. We are now going to look at the children's production capacity of StG, in order to explain what these transformations are.

4. Phonological acquisition processes At about the age of six children are becoming aware of the linguistic characteristics of StG, but they need the help of the adult to make this awareness explicit. At kindergarten during an experiment in which a 'visitor' spoke only StG, the children realized that his speaking was different only when they were questioned about it (cf. Künzler et al). The question of which language the visitor was speaking they answered with tü:tsch ('German'), schwi:zertü:tsch ('Swiss-German') and ho:chtü:tsch ('high German') (p. 18). Questions about their experiences with StG, asked at different points of the longitudinal study, they answered with -

es isch nüd änglisch ('it's not English') das isch wie chine.sisch ('it's like Chinese') hä, tänk vom femse: ('of course from TV') wie mer i de schnei ret ('like you speak at school') ich hin emol z tüxschland gsi:, und döt han i au öppis verschtande ('once I was in Germany, and there I understood something too') - hochdeutsch, wriladuli:aliduli.. ./('high German,...')

These anecdotal observations illustrate the rather casual knowledge about StG children have at this age. The analysis of their spontaneous production of StG, however, shows how much more accurate their unconscious knowledge of StG is. Koch-Niederer (pp. 14f.) reports that during the experiment in kindergarten one child, who had refused to speak StG for a long time, suddenly started speaking 'fluently' a sort of StG having the foUowing phonological characteristics: - raising and unrounding of vowels - transformation of SG diphthongs in monophthongs and vice versa - suffixing of /n/ to SG words.

Divergence and Convergence from the Perspective of the Language Learner

143

The following is a closer analysis of these observations, illustrated by examples from the studies of Künzler et al, Stoll & Kunz, and the corpus in Stern (1978). 4.1 Change of vowels 4.1.1 Raising and unrounding Examples: SG:

(6)

the children's StG:

a)

himcP [y]

>

b) c) d)

sichB/ [i] Tomas' [j] hända [ae] hoxh odar ich schkssa [Y:] z schpoit [o:] mänggmal scho: [ae]

> siier{3) [i] > Thoimas (4) [o:] > hende [e] hoxh «f schi:sse (2) [i:] > zu schpe:t (4) [e:] > menggmol scho: [e]

e) f)

('dog') ('sure') ('hands up or I shoot') ('to late') ('sometimes yes']

These examples show that the children spontaneously discover some essential phonological rules of correspondence between SG and StG by raising and unrounding the SG vowels which in general are articulated very open and back. In b)-d) the transformations meet the target language, in a) there is a tendency to exagerate the raising. That these transformations are rule guided is evident from the frequent overgeneralizations exemplifyed in e), the StG target being spät, and in f), the target being manchmal In how far these phonological transformation rules are true also for idiomatic expressions like d) carmot be determined on the base of our data. Since formulaic expressions are acquired in a holistic way (cf. Fillmore 1979), holistic learning probably includes also the phonological form. Therefore, phonological transformation rules would not affect single phonemes of these expressions (cf. also the discussion of schü:ssa > schäissB below, p. 12).

The transcription i, ö, k stresses the very open articulation of the SG vowels, e.g. the /u/ in SG hund is almost pronounced /o/, whereas the StG articulation is quite dose. ch stands for the velar fricative /x/, f for the palatal allophone. Note that in SG only the velar fricative exists, whereas in StG /x/ and / j / occur in complementary distribution before back and front vowels respectively. In SG /t/ is not aspirated and short / o / is also much lower than in StG; long /o:/ is dose in both varieties. Numbers in parenthesis refer to the previously listed examples the displayed item belongs to; a second digit after a semicolon, e.g. (4;5), refers to the line number of the example.

144

Otto Stern

Evidently, similar phonological transformations are also characteristic for the transition from one SG dialect to another. The raising and unrounding of [x] to [e], in (6) f) produced as a StG form, is typical also for eastern Swiss dialects like St-Gall and Appenzell using the word mengmod. Changes of manner and place of articulatiori from SG to StG are also found in the acquisition of consonants. In (4;5) R. is trying to change the articulation of ich - in SG always pronounced with the velar fricative [x] resulting in ik. In this example the place of articulation is changed in the right direction (velar > palatal), the manner of articulation, however, is incorrect (occlusive instead of fricative). The acquisition of the StG allophone [?] is very difficult to learn for all Swiss chüdren, and many will never master it perfectly. At school the overgeneralization of [5] is very frequent, e. g. bewogen or magert (cf. Kunz & StoU). In the early stage of acquisition the allophone [9] is produced correctly in StG words that don't have the phoneme /x/ in the corresponding SG form, cf. e.g. the 'queen' in (5):

but: and:

nigtiq] ich [x] mich [x]

< <
> > >

SG diphthongs

zu3glu3gt

>

nig

>

liab tiünd

> >

the children's StG diphthongs häist ('to bite') säin ('his')

äin wäil

(3) (4;7)

(V)

(4;9)

('because')

(5;23)

the children's StG monophthongs zu:glu:gt ('watched') (zugeschaut) ni: ('never') li:b ('nice') tu:t ('do')

(4;2) (5;18) (5;23) (5;24)

Divergence and Convergence from the Perspective of the Language Learner

145

Because SG and StG words very often differ only in phonological form, the phonological transformations lead mostly to correct StG words. Only in the case of zmluagd > zu:glu:gt the transformation results in a word not existing in StG. A special case is the word hi:r (4;5). Although there exists a SG word hi3, it is only used in the expression hia und da: and never stands alone (at least not in the eastern dialects). In free occurrences, as a local adverb, it is always replaced by da:. The use of hirr'm(4;5) is formulaic, i. e. it is part of the reporter's formula: here speaks X in Y. Although the rules for the monophthong/diphthong transformation are correctly learned quite fast, there always occur errors caused by overgeneralization of the rule: (8)

umflä:gt ('falls down')

>

sumhrüh ('sunglasses')

>

gog» schi:fa:r9 ('to ski')

>

umflöigt (umfliegt, umfällt) (StoU & Kunz) sonnenbräile (Sonnenbrille) (Künzler et al.) gogen schäi - - xhi:fa.ren (skifahren gehen) (StoU & Kunz)

SG /ü:/ seems to be especially difficult. There are many cases of SG /ü:/ leading to StG /öi/(written eu or äü), e.g. ü:b ('owl') > Eule, hü:l3 ('howl') > heulen; but there are also some cases of SG /ü:/ corresponding to StG /i:/ (written ie), e.g. bü:gd ('to bend') > biegen, flü:gd ('to fly') > fliegen. It is difficult to determine when the children perform the monophthong-diphthong transformation for /ü:/ and when not. In (1) there is an example of unchanged use in the StG form, SG ahflü:g3 becomes StG abflü:gen. Note that this rule operates only on long SG /ü:/. Therefore, its use in the case of the above example sunebrüb is erraneous. Overgeneralizations of the diphthongization rule are recognized as erraneous very early, which is confirmed by frequent selfcorrections like in (8) schäi — schi:fahren. Anecdotal evidence for the difficulties children often have with monophthongs/diphthongs is given by the following dilemmatic Situation that occurred to my 6-year-old nephew when visiting friends in Germany. When playing cop-and-robber with his StG speaking friends, he repeated several times the expression hende ho:ch oder ich schäisse. To the playmates correcting him: "It doesn't say schäisse ('to shit'), it says schüsse ('to shoot')!", he answered: "I'm not allowed to say this word at home." When playing at home (in Switzerland) he used to pronounce correctly ich schi.-sse. To him, however, this was part of his SG repertoire, therefore it needed translation, i.e. diphthongization, when used in the actual StG

146

Otto Stern

Situation. This example also shows that single words and phonemes of a holistically acquired idiomatic expression are not immediately processable as isolated elements (cf. Fillmore 1979). The ambiguity of SG schisse was recognized by the child only when the word was isolated by the playmate's correction. 4.2 N-suffixing A striking phenomenon of early StG production is the frequent use of /n/ as an ending of SG words. The following transcription of a discussion between the experimenter and a 6-year-old boy illustrates this procedure: (9)

E:

Was machtest du mit der Sonnenblume^ (what did you do with the sunflower) Der Vögel gem. (SG: devögelgei) (give (it) to the birds) Was machten die Vögel? (what did the birds do) Die picken Kernen usen. (SG: die picked cheme-n-use) (they pick out the kernels) (Koch-Niederer, p. 32)

Evidently, we are witnessing the first attempts to deal with StG morphology. The suffix -n is very frequent in StG. For the children this must be a striking difference between the two varieties, since in SG, due to historical change, very few words have kept this ending. The acquisitional process confirms Slobin's (1979:108) Statement that "for various reasons of attention and memory, children find ends of words and utterances more salient than beginnings and middles". Searching for meaningful grammatical items children are following the "operation principle A: Pay attention to the ends of words". Although this procedure very quickly leads to the acquisition of the noun and verb inflection, it seems that in the early stage of n-suffixing we are dealing with a merely phonological process, allowing the learner to come immediately very dose to the phonological gestalt of StG (cf. also MacWhinney 1978). This is confirmed by the relatively unsystematic distribution of the n-suffixes over the different word classes. e. g. (10) Verbs gogen schi:fa.ren (SG: goge schi:fa:re) (going to ski) hat ghäien la:n (SG: hat gheie la:) (he let it fall down)

(Koch-Niederer, p. 28)

Divergence and Convergence from the Perspective of the Language Learner

147

Nouns The following are examples of words repeated by the children, after the experimenter's (correct) pronounciation for a drawing dictation in kindergarten (cf. Künzler et al, p. 26) Experimenter: Chüd: eine Feder ('a feather') (e fädere) eine Federn eine Zwiebel ('an onion') (e zwihle) eine Zwiebeln eine Schüssel ('a bowl') ein Schüsslen (e schüssle) ein Kran ('a crane') ein Kranen (en krame) Adverbs usen de:ten ahen

('out') ('there') ('down')

SG:

use de:te abe

StG:

raus dort runter

Of course, SG words ending on -e are priviledged to get a n-suffix, whereas occurrences pischamam ('pyjama'), or Schudöjfen ('schoehorn'), from SG schuelöffel, are rather exceptional. Thus, most n-suffixes occur on nouns and verbs, very often ending on -e, and the step to meaningful grammatical marking is only a small one. As the acquisition of the morphological system of StG progresses, the unsystematic use of the n-suffix decreases rapidly. In our investigations, 7-year-old first graders, contrary to the 6-year-old kindergarteners, produced only very few misplaced n-suffixes (cf. Koch-Niederer, p. 37ff., and recordings by Stern 1980). 5. The acquisition of the StG preterite The discussion of the acquisition of StG on the lexical and phonological level showed acquisition processes typical for the dialect-standard Situation which, however, are not typical in situations of greater divergence between LI and L2, i.e. in actual second language acquisition. The great degree of convergence berween SG and StG in the lexicon, and the systematic relation between the two varieties at the phonological level induces the learner to continue first language acquisition, i. e. to develop new linguistic registers in hisfirstlanguage. The new register (StG) is modeled on the SG base by means of adaptations and transformations. This procedure, however, is not adopted for the acquisition of the StG preterite. The preterite does not exist in SG; there is only a perfect past tense formed by an auxiliary vefb + perfect participle (cf. Haas 1982:95). Thus, for the preterite we observe typical second language acquisition processes, such as gradual approximation to the target form, simplification

148

Otto Stern

and regularization of irregulär target structures and slow progress with large individual Variation (cf. Wode 1983: 191 ff.). We assume that in the domains of morphology, in which StG and SG differ rather greatly, typical second language acquisition occurs, although, up to now, only the preterite has been studied in a more systematic way. The investigation by Koch-Niederer documents, by means of ad hoc experiments, the process of acquisition of the StG preterite in a small group of 6-year-old kindergarteners (5 children) and 7-year-old first graders (6 children). The investigation took place over several weeks. The data gathered in the experiments are completed by unsystematic observations and recordings. Because the preterite occurs mainly in written StG, the investigation of Koch-Niederer is based on a short rime, written especially for the purpose of testing. The rime contains 9 verbs in the preterite, 6 strong/irregular and 3 weak/regular forms: (11) Test-rime: Es war einmal ein Zwerg, der ging auf einen Berg, dort setzte er sich in das Gras und putzte seine lange Nas. Dann aß er zwei Melonen. und schleckte zwei Zitronen. Da kam die gute Fee und trug den Zwerg zum See. Der Zwerg schwamm schnell nach Haus. Jetzt ist das Märchen aus.

(There was once a dwarf, he went on a mountain, there he sat in the grass, and cleaned his long nose. Then he ate two melons and licked two lemons. Then came the good fairy and carried him to the lake. The dwarf swam quickly home. Now the Störy is finished)

The children learned the rhyme by heart repeating it together with the experimenter and the teacher on several consecutive days. The children's rehearsals were tape recorded at all stages. In addition the experimenter organized play activities (pantomime, puppets) and discussions, in order to provide opportunities for spontaneous reproductions of the preterite forms heard. It was assumed that during rehearsal the children would not be able to reproduce preterite forms they had not acquired yet (cf. Slobin 1979:104), or that they would change the forms following the actual stage of acquisition. Note that when talking about acquisitional stages, we do not assume that these are necessary actual stages of the individual child, but rather reconstructions of observable stages on the base of our data. In the beginning the children concentrated on the meaning of the verb stems. They understood the meaning of a verb in StG, if they were able to establish identity - by the means of phonological transformations - with a SG verb they already knew.

Divergence and Convergence from the Perspective of the Language Learner

149

Düring the first rehearsal of the rime the 6-year-olds reproduced the verbs as follows: (12) ging setzte putzte ass schleckte kam trug schwamm

gehe, gangt, gin^ setzt, sitzt, setzte putzt, putzten, putzte esst, isst, asst schleckt, schleckte kommt, kummt, kunnt, kam truckt, truck, trug sich, trug springte, schwamm

(Vent') ('sat') ('cleaned') ('ate') ('licked') ('came') ('carried') ('swam')

There is a strong tendency to reproduce StG preterite as SG present. This is confirmed by the reading of 7-year-olds tending to omit the seemingly meaningless ending -e of the regulär preterite form. In (12) the forms sitzt/setzt, putzt and schleckt are also SG present. The same transformation (StG preterite > SG present) occurs with the irregulär verbs, if the meaning is detected: kummt/kunnt and springt are SG present. For the children's reproductions of the rime, what is important is the meaning and not the form. The event the rime teils about is temporally situated by the introductory formula es war einmal ('once upon a time'), which was reproduced correctly by all children; in their view there seems to be no need to mark past tense (redundantly) on the following verbs. As a matter of fact, their focussing on meaning leads them to very forced interpretations (cf. above, in section 3, the comment on trug ('carried') understood as truckt ('to push down')). For a 6-year-old, the past tense - which during the linguistic development of the child emerges from the perfective aspect - is expressed by the means of the perfect. It is the auxÜiary that carries temporal Information in the first place. MacWhinney (1978:53), who studied first language acquisition of the perfect tense in German, states that although the auxiliary enters only after age 3, it is the first form to express perfectivity. Therefore, the lacking of the auxiliary in the StG text will be interpreted by the child as present tense. There is some evidence that for his first expressions of past tense in StG the SG chUd uses the auxiliary verb tun ('to do') + infinitive. In an experiment with pictures showing what the now big girl Anna liked to do when she was a child, i.e. representing past events, the

' The list contains all forms produced by düferent children of the sample, neglecting the frequency of occurrence.

150

Otto Stern

experimenter asked the children to comment on the pictures beginning with the phrase als Anna klein war ... Many of the answers were of the type tun + Infinitive. (13) Experimenter: Als Anna klein war... Child: tut sie flüschteren ('she does whisper') tut sie turnen ('she does gymnastics') tun sie gern be:hi spiden ('does she like to play doUs') (Koch-Niederer, p. 29)

The introductory clause in the past tense did not prevent the children from continuing in the present tense; however it is also possible that the actual Situation of describing pictures, here and now present, influenced much more the actual speech Situation than the verbal stimulus asking them to teil the Story of Anna's experiences in the past. Still, they also produced past tense forms in the perfect. The comparison between 6- and 7-year-olds shows a remarkable increase with age: age 6: 12 present age 7: 19

4 preterite 12

2 perfect 9 (ibid., p. 46)

The expression tun + Infinitive is also frequently used by the older children. It is produced simultaneously with present and past tense forms (perfect and preterite) in the same text by several children. When asking a 6-year-old to say the expression tut spielen ('do play') in other words, it was replaced spontaneously by spielte ('played'). To what extent tun + infinitive for these children contains temporal information cannot be determined on the base of our data. Evidently, it is a typical simplification strategy also appearing early in first language acquisition of German and SG, as well as in motherese and in foreigner-talk. Typically, this simplification strategy reappears during the acquisition of the StG past tense system revealing second language learning processes, long after these children have stopped using this form in their first language (SG) in similar contexts. In spontaneous production of StG, e.g. in (4), or when asked to teil personal past experiences, all children use exclusively perfect past tense, as they are used to in SG. Difficulties arise only in the past participle, which very often is phonologically and/or lexicaUy adapted from SG by means of (overregularized) transformations. The following is an example of a 71/2year-old girl after eight months in first grade (the StG production is transcribed in regulär StG orthography; if the form is incorrect, the corrected form is given in parenthesis: first form = correct StG form, second form = SG lexical base); the English translation is dose to literal).

Divergence and Convergence from the Perspective of the Language Learner (14) 1 2 3 4 5 6 7 8 9 10 11 12 13 14

151

Wir haben Femseh gelwgt (geschaut; glusgt) (we have TV watched dann ist ein Film gekommen von einem Delphin then has a movie come of a dolphin und der Delphin ist ganz stark gsein (gewesen; gsi:) and the dolphin has been very strong und ein Bub noch, der ist auch stark gsein (gewesen; gsi:) and a boy also, who has been strong too und dann hat 's Räuber gehabt, die sind ganz böse gsein (gewesen; gsi:) and then there have been robbers who were very wicked und denn ist ein Messer im Meer gsi:n (gewesen; gsi:) and then there has been a knife in the sea irgendwie angri:jfen go gangen (angreifen gegangen; go a:gri:jff) somehow (he) wem to attack und da hat er einen Schnitt überkommen (bekommen; übsrcho:) and then he got a cut ist am Meer gele:gt (gelegen; gläg^) was laying at the sea (the knife) und der Bub ist auf dem Schiff gsi:n (gewesen; gsi:) and the boy has been on the ship und de het es de Bub gesin (gesehen; gse:) and then the boy saw it (the knife) und ist zum Delphin (gegangen) and went to the dolphin und da hat er s gsi:n (gesehen; gse:), and then he saw it das er ein Schnitt gehabt hat. (...) that he (the dolphin) had a cut)

This is a typical example of direct transformation of SG into StG. Evidently, the perfect participles expressed by different words in StG and SG are difficult to translate. What is striking, is the coinciding of the two words gsi: ('been') and gse: ('seen') in the child's StG expression gsi:n (lines 6,10 and 11) Standing two times for i>een and once for seen. The phonological adaptation foUows the rules described in section 4. The coincidence is caused by omitting the diphthongization leading from gsein in lines 3, 4, 5, to gsi:n in lines 6, 10. These examples show that the StG participle is not acquired in a holistic way, but is generated anew, starting from the SG base each time it is used. By doing this, the children don't take the easiest way, i. e. they don't simply translate a SG participle in a StG one by the means of phonological (surface) transformations. From the early stages of acquisition it is evident that the system of participle formation is reconstructed radically. The following examples were produced by children who are one year younger than the one in (14), i.e. 6i/2-year-old:

152

Otto Stern

(15) StG: SG: wo mir oben gloft sind < aba gloffs sind (als wir hinunter gegangen sind; as we wem down) dann ist er hinauf gegangt < uß ggangg (gegangen; then he wem up) wir sind dort eingebrocht < ibbrochs (eingebrochen; there we broke in) wir haben ganz viel gestohlt < gschtoh (gestohlen; we stole a lot) (Künzleretd/.,pp. 17, 34, 40)

Although all these children master the distinction between weak/regular and strong/irregular in the SG dialect, they overgeneralize the weak form when producing a StG participle. Since SG and StG have virtually the same forms, it would be easy to translate a SG participle into correct StG by simply adding a -n to the SG base (cf. 4.2, n-sufFixing): SG:2gang3 > StG: gegangen ('gone') gschtoh > gestohlen ('stolen') Nevertheless, the learners fall back on earlier acquisitional stages and reconstruct the entire system from the beginning. Summarizing our fmdings of the acquisition of the StG past tense system more systematically, we distinguish four stages: Stagel: StG preterite is replaced by present tense forms. Stage II: Frequent strong/irregular preterite forms are acquired in a holistic way as new words, e. g. war ('was'), hatte ('had'), ging ('went'), kam ('came'). At this point we observe an interesting combination of SG perfect and StG preterite: as soon as the preterite forms of the auxiliary verbs sein ('to be') and haben ('to have') are acquired, they are immediately used to express StG past tense. The foUowing narrative by a 71/2-year-old girl (8th month in first grade) illustrates this procedure: (17) Mir waren zuerst auf dem Flughafen gewesen, und da hatten wir so (first, we had been at the airport, and then we had lang müssen warten, wil unser Auto war kaputt gewesen, und dann waited for a long time, 'cause our car had been broken, and then hatten wir müssen zum Automechaniker gehn, und hatten müssen das we had had to go to the mechanic and had had to have Auto la flicken, und dann hatten wir gewiss noch können gehn ... the car repared, and then, sure, we could have gone...) (She is searching for words and finishes after a long pause).

Divergence and Convergence from the Perspective of the Language Learner

153

Although all the verb forms are clearly pluperfect, it is evident that the child doesn't mean to use this tense, but simply tries to express past tense in StG. By doing so, she tries to integrate the dual past tense system of StG in the simple one of SG. Plupeifect is acquired only much later, and many SG speakers never reaUy master it. In an investigation of the syntax of written reproductions of a previously heard narrative by 13-year-old pupils (6th grade), Guyer (1985) states that the use of pluperfect is generally avoided. As a consequence, in subordinate temporal clauses expressing anteriority, the same tense is used as in the main clause. Stage III: The proper ending -te of the regulär preterite tense appears first with strong/irregular verbs, i. e. it is immediately overgeneralized: gangte {ging; 'went'), eßte {aß; 'ate'), stiegte {stieg; 'climbed'). Often the stem of the SG perfect participle is used to form the StG preterite: gangte < SG: gganga Cgone'). This procedure puts the preterite form closer to the perfect, thus confirming the above Statement that learners try to construct a simple past tense system - similar to SG - by merging perfect and preterite. Stage N: Finally, the correct target forms of StG preterite and perfect, and the correct use of regulär and irregulär verbs are acquired by most children at age 8 to 10, but at individually greatly varying pace. Exceptionally, as in the example below, even a 7 V2-year-old first grader is able to produce correct preterite and perfect tense forms: (18) a girl's narrative about St-Nicolas Day (6th of december) (...) Mit mir war er ganz zufrieden. Ich hatte - - und wo ich am (With me he was rather pleased. I had - - and when I was Abend daheim war, da hatte ich vergessen, den Stiefel heraus zu at home in the evening, I had forgotten to put out the stocking. tun. Und dann, am Morgen, hatte ich doch den Stiefel voll mit And then, in the moming, nevertheless, I had the stocking füll of Schokoladen. Und dann kam er nochmal in den Wald hinein, und dort chocolates. And then he came again into the forest, and there hat er uns noch ein Päckli gegeben, und dort waren noch ganz he gave us another present, and there was again viele Sachen drin ... lots of stuff in it)

154

Otto Stern

6. Closing remarks The spontaneous use of StG in role play shows that children integrate StG Clements quite easily and rapidly in their SG repertoire, as if they were part of their LI. Evidendy, it would be easy for them to acquire a conversational register of StG (cf. also Stern 1984:142). However, since TV is almost the ordy source of input of colloquial StG, opportunities to get sufficient input and practica are very limited. This leads to an interruption of the 'natural' acquisition process at the beginning of school, when the children start hearing and using StG for language learning purposes in a Swiss context and with a Swiss accent. In phonology, for example, especially concerning the raising of vowels, it is evident that in the early phase of 'TV-StG' the pronunciation is very similar to northern German Standards. Under the influence of schooling children change very quickly to a backened and lowered Swiss accent. But not only is the phonological quality changing, they are losing the other typical characteristics of (northern) 'TV-StG' as well, i.e. idiomatic terms, interjections, modal particles and fast speech forms. The analysis of acquisition processes on different levels of the linguistic system (phonology, morphology, lexicon) shows clearly that there is not a straightforward additive development of linguistic registers in LI. Although this might be true for the lexicon, on other levels we observe genuine L2 acquisition processes. At school, where the use of StG becomes crucial for successful language learning, the particular mixture of first language development and second language learning leads to a wide gap between passive and active linguistic competence. The ability to express oneself in StG does not correspond to the level of linguistic development reached in SG. This becomes evident when one compares narratives (13), (15), (16) to narratives in SG of children of the same age or even younger (cf. Stern 1984). The narratives in StG are pronounced very slowly, and intonational shaping is very limited. A good example of these difficulties is given at the end of (16) where the modal particle gewiss - - being part of a much more elaborated and flexible Performance - - plays a false note within the rigid and awkward narration.

Transcription Conventions

The transcription of the SG examples and the children's attempts to speak StG follows the Convention of Dieth's (1938) Schweizerdeutsche Dialekt-

Divergence and Convergence from the Perspective of the Language Learner

155

Schrift. The familiar aspect of StG writing is largely maintained, especially concerning the representation of consonants. All voiced consonants, however, are pronounced voiceless. Continuous StG discourse is transcribed in Standard orthography. gg

double consonants = geminate articulation,

h

e-g- ggangs (gegangen; gone). glottal constrictive; does not indicate vowel lengthening in Dieth's transcription.

Special Conventions for vowels: :

«

ä a

marks long vowels, e. g. ar ga:t {geht; goes). Open articulation, especially of long vowels, which otherwise would be dose, e. g. ich werr {wäre; 'would'). Short vowels have always an open articulation. In the children's attempts to speak StG there is sometimes exagerated closing of short vowels; this is indicated by a period under the vowel, e. g. hund ('dog'). very open e-sound, e. g. s galt {das Geld; 'the money'); also used for the very low diphthong /äU, e.g. wäisch {weisst; 'you know'). schwa.

Examples in SG and in StG are italicized. StG expressions within SG discourse are printed in small capitals. The English translation does not attempt to be idiomatic, but follows closely the structure of German, resulting in some places in inacceptable English sentences.

References Auer, J. C. P. (1984). Code-Shifting: Phonologische und konversationeile Aspekte von Standard/Dialekt-Kontinm. Arbeitspapier, Sonderforschungsbereich 99, vol. 88. Universität Konstanz. Burger, H. (1984). Sprache der Massenmedien. Berlin: de Gruyter. Erny,M. (1984). Textverstehen in Mundart und Standardsprache im Kindergarten. Universität Zürich: Lizentiatsarbeit, mimeo. Fillmore, L. W. (1979). Individual differences in second language acquisition. In C. J. Fillmore, D. Kempler, W. Wang (eds), Individual differences in language ahility and language hehavior, 203-228. New York: Academic Press. Guyer, G. (1985). Syntaktische Untersuchungen von Nacherzählungen nach Mundart und Standard Textvorlagen. Universität Zürich: Lizentiatsarbeit, mimeo.

156

Otto Stern

Haas, W. (1982). Die deutschsprachige Schweiz. In R. Schläpfer (ed), Die viersprachige Schweiz, 71-160. Zürich: Benziger. Lötscher, A. (1983). Schweizerdeutsch. Geschichte, Dialekt, Gebrauch. Frauenfeld: Huber. MacWhinney, B. (1978). The acquisition of morphophonology. Monographs of the Society for Research in Child Development, vol. 43/1-2. Chicago University Press. Schwarzenbach, R. and Sitta, H. (1983). Mundart und Standardsprache in der deutschen Schweiz. Bulletin CILA 38, 62-71. Neuchätel. Sieber, P. and Sitta, H. (1984). Schweizerdeutsch zwischen Dialekt und Sprache. Kwartalnik Neofilologiczny, XXXI/1, 3-40. Prag. Slobin, D. I. (1979). Psycholinguistics, 2nd edn. Glenview, III.: Scott, Foresman and Co. Stern, O. (1978). Schweizerdeutsch im Kindergarten: Korpuserstellung, linguistische Analyse, Computer-Verarbeitungssystem. Zürich: Diss. mimeo. Stern, O. (1984). Developing decontextualized language in children's narratives. I n j . C. P. Auer and A. Di Luzio (eds), Interpretive sociolinguistics, 129-149. Tübingen: Narr. Wode, H. (1983). An integrated view of language learning. In H. Wode (ed), Papers on language acquisition, language leaming and language teaching. Heidelberg: Groos.

Seminar Papers Seminar: DerAußau der Wortbedeutung bei Kindern, directed by Prof. H. Burger, Deutsches Seminar der Universität Zürich, 1983/84. Koch-Niederer, A. Der Erwerb des Imperfekts bei Kindern mit Schwelzerdeutscher Muttersprache. Längsschnittuntersuchung an 6- bis 7jährigen Kindern. Künzler, A., Loher, F., Hegglin, S. Hochdeutsch eine grausge Sprooch. Untersuchung über die Langzeitwirkung von hochsprachlichem Sprachgebrauch im Kindergarten. StoU, G., Kunz, J.-P. Linguistische Analyse spontan produzierter Hochsprache von Kindern. Weber, D., Arter-Lamprecht, L. Die Entwicklung der Bedeutung von Antonymen bei 3- bis 8jährigen Kindern.

N O R B E R T D I T T M A R & PETER SCHLOBINSKI,

Berlin

Convergence, Discourse and Variation

1. Convergence as an Interactive process In the current State of sociolinguistics there is no clear concept of the role of convergence as a source of linguistic Variation. Although first theoretical attempts have tried to "locate the accommodation model firmly as a psychological xheory" (Giles & Smith 1979:53) three problems have not been solved yet: (1) Giles & Smith (1979:46) State that convergence refers "to the processes whereby individuals shift their speech styles to become more like that of those with whom they are interacting". However, the degree to which a shift can take place, as well as the types of source and target varieties which can be involved in a shift, must be formulated in linguistic terms. (2) There is no clear distinction between unmarked (normal) and marked (convergent) speech patterns. (3) Because convergence does not function like a monolithic block, we have to find out what levels are affected and to what degree they are affected when speakers converge to other participants in interaction. The following discussion is a contribution to a socioHnguistic basis of convergence for future analysis. As a first step we will clarify the concept of convergence in contrast to other patterns of speech behavior. We refer to habitml speech behavior as the average linguistic Performance of an individual which is relatively independent of specific context and discourse. It is the prototypical and unmarked communicative behavior of a Speaker. The habitual speech behavior ensures that communicative efforts and demands can be foreseen, and that communication in everyday situations can be coordinated and checked against mutual expectations and norms. We claim that the individual has at her/his disposal a knowledge of the main characteristics of her/his normal way of speaking. As a rule, the habitual speech behavior is not used for Strategie ends. It has clear-cut

158

Norbert Dittmar & Peter Schlobinski

regularities. Put the other way round, the characteristics of these Speech habits can be predicted with high certainty. In contrast to habitual speech behavior linguistic convergence means a speaker's adaptation to the verbal register of another in the process of interaction. O n the one hand convergent behavior depends on the situational conditions and the social frames which covary with certain norms and expectations. O n the other it depends on the varieties used in verbal interaction, their prestige value for the interactants and the structure of the utterances. Minimal convergence is the case if at least one linguistic element of the source variety of a Speaker is adapted to the target variety of the interactive partner. We call such a case marked.^ The more convergence leads to a deviation from the source variety the more marked it usually is. The determination of both the source variety and the target variety is as much an empirical problem as the question to what extent and under what conditions convergence takes place; for example Intonation contours are more or less marked than lexical variants. In any case it is possible that speakers converge on several linguistic levels at the same time to the speech behavior of alter. In this case specific rules are in force which coordinate and integrate linguistic elements into the patterns of convergence. These rules describe constraints of cooccurrences. They are based on criteria of appropriateness and compatibility. Convergent speech behavior is a reaction to certain input conditions of the interaction; the following are salient and relevant: (1) The interpersonal relationship - at the beginning or in the process of interaction - is characterized by a high degree of openness with respect to the quality of interaction. (2) The regulation of the interpersonal relationship is important for the successful achievement of the communicative events/acts; it has to be negotiated. (3) There is, at least from the point of view of one of the interactants, considerable cognitive uncertainty about the appropriateness of communicative means relative to (i) status, power and symbolic prestige of the interactive partners, (ii) contents that have to be processed in the discourse, (iii) the preceding states of understanding and (iv) predictable consequences of understanding. In this view, convergent speech behavior has the function of constraining the availabUity of alternative ways of behavior in order to improve the ' " A marked category is signaled by adding something to an unmarked category. In language it may be voicing, nasalization, an affix, or a component of meaning." (Frake 1975:37)

159

Convergence, Discourse and Variation

interpersonal relationship. Its cost is deviation from individual habits; its benefit is the diminishing of emotional distance by accommodation to the other. Moreover, cognitive uncertainty is reduced. From this it foUows that cognitive uncertainty is a presupposition of convergent behavior; this uncertainty affects the interpersonal relationship of the interactants and can only be compensated for by direct stylistic markers (accommodation). Thus, convergent behavior controls the interpersonal relationship. The accommodation process usually is neither controlled nor conscious. Divergent Speech behavior is also marked and directed with regard to habitual behavior, but its directedness contrasts with convergent Speech behavior. We refer to divergent speech behavior as an activity which establishes differences between two or more interactants in the process of interaction. The process of divergence can be marked by elements on all linguistic levels. In general, this process occurs unconsciously, and is not controlled by the Speaker. Thus, states of uncertainty may be consequences of divergence. In conclusion, convergence weakens the boundaries of identity, whereas divergent speech behavior strengthens the maintenance of identity. Strategie speech behavior is teleological, i. e. it is first of all oriented towards ends and purposes of action, reflecting the individual's intention to pursue certain aims and goals. Thus, the process of interaction is determined by Strategie means and is subordinated to prescribed sequences of certain cognitive structures of actions. Having goals and purposes in mind, speakers plan their verbal behavior and control their speech style in this way. The prototype of Strategie speech behavior is argumentation. Strategie speech behavior is reflected more in the pragmatics than in the Table 1 Speech behavior convergent/ Components

habitual

divergent

Strategie

Language use Modus of communication

— marked polyfunctional

+ marked interpersonal

+ marked oriented towards aims and purposes

Aspect of cognitive stability Degree of speech control

certain uncontroUed

uncertain uncontroUed

certain

Degree of consciousness

unconcious

unconscious

conscious

intentional

160

Norbert Dittmar & Peter Schlobinski

logic of argumentation. The logic of argumentation only refers to the coordination and coherence of arguments, the pragmatics of argumentation deals with the question of how to present arguments effectively in the ongoing discourse. The more explicitly speech behavior is guided and controlled by goals and ends, the more Strategie it is. The less speakers aim at such teleological effects, the more we are concerned with habitual speech behavior. There is a whole scale of different states of Strategie versus habitual speech behavior. Although we have to realize that Strategie and habitual speech behavior share sets of Clusters of features we assume that there are prototypical configurations for each of the two. The same holds for the distinction between all three kinds of behavior under discussion. In conclusion we summarize the main aspects of the preceding discussion in the foUowing table that bears on the most relevant and salient distinctive features of convergent/divergent vs. habitual vs. Strategie speech behavior.

2. Gase Studies In this article, it will not be possible to validate our approach to convergence/divergence based on large-scale sociolinguistic data. Our intention is a modest one: we wül attempt to illustrate our concept of 'convergence/divergence' by analyzing particular cases and discussing critical points of our categorization. According to this perspective, sections 2.1 and 2.2 are devoted to the process of 'convergent vs. divergent behavior'. The first example is an interaction between a Speaker of the urban Berlin vernacular and two speakers from other West German dialect areas. Features of 'habitual', 'convergent' and 'divergent' speech behavior will be discussed on the paralinguistic, phonetic and stylistic levels. The second example illustrates extreme pressure for convergence as two interactants of different linguistic backgrounds try to communicate with each other under the condition of minimal knowledge of the language of alter. In order to 'survive' under these extreme conditions of communication, particular strategies of convergence have to be developed. Section 2.3 examines an example of 'Strategie speech behavior'.

Convergence, Discourse and Variation

161

2.1. 'Local identity' as a trigger of 'convergent' and 'divergent' behavior (a) The data The verbal interaction presented in (b) was recendy taped in Berlin by two students, Rüdiger Kneifel und Jürgen Warscheid. The interactants are an old native Berliner B and the students K (native Palatinate Speaker) and W (native Speaker from Westfalia). K and W know each other well; neither knows B. On a Friday afternoon, K and W approach B identifying themselves as West Germans and ask B about where to go in order to enjoy themselves in night clubs, discos, striptease bars, etc. (West Berlin is known for its night life). B tries to help them, although he clearly states that he is not an expert in these matters. The interaction goes on for several minutes with B presenting some places to go, in a way a father would give advice to his sons, or an aged, experienced person to youngsters. K, W and B share some knowledge of German culture, but beyond this, their backgrounds are quite different. (b) Analysi^ The following two hypotheses should be accounted for by our discourse analysis: (i) B increasingly shifts to Standard features when he follows the maxim 'help strangers and maintain a certain solidarity with those who do not know the local Community structure'.

(ü) B strongly shifts to Urban Berlin Vernacular (UBV), thus distinguishing himself clearly from the West-Germans by 'divergent' behavior when 1. he attempts to get rid of the two students; 2. he tries to distinguish the image of Berlin as some kind of "Las Vegas" (night clubs, bad reputation, etc.) from - in his eyes - the normal Berlin neighborhoods of the older days; 3. he defends local Berlin values against West-German values. Recurrent verbal patterns provide evidence for (i) and (ii). The transcribed conversation takes place between individuals who have never met before. The input conditions of the verbal interaction are therefore insecurity and uncertainty with respect to the interpersonal rela-

The analysis of this and the following examples is based on a füll and detailed transcript of the relevant passages of conversation. Because of limited space it is not possible to present these transcriptions in their fuU length.

162

Norbert Dittmar & Peter Schlobinski

tionship, the request for advice and B's willingness to react to it. Regarding the expression of 'identity' in the encounter, we expect the social and personal Information to be coded parallel to the propositional content of the utterances so that the speech styles of the individuals allow inferences about the attitudes the interactants may have towards each other. Because the interactants cannot make any predictions about the possible outcomes of the encounter, there is considerable insecurity and uncertainty about the interactive and communicative mode; ends and effects cannot be controlled consciously because of a major lack of shared knowledge and background Information. In what follows only the communicative behavior of B will be taken into account because B's discourse comprises enough turns and words to identify the 'habitual' as well as the 'convergent' and 'divergent' speech styles. A first analysis of the tape clearly shows that e. g. the following lines represent more or less the unmarked, habitual speech variety of B^: B: Früher war et am Potsdamer Platz war Cafe Vaterland, dit war unterteilt mit ne türkischet eeh Stube mit ne Mokka-Stube, denn jab's die Rheinterrassen da jab's alle halbe Stunde 'n 'n Jewitter zooch da uff r een Jewitter da wa een K: L Was? B: Panorama und da eeh eeh dit war wurde eeh eeh beleuchtet und denn zooch da een Jewitter uff alle halbe Stunde dit wa'n die Rhein-Terrassen un' denn spielten se demensprechende Musik un' denn jab's den Prater also alles der der der eeh ^ Stimmung und p der der der der Stadt W: • Ach so. L (Ach Wien?) B: den Stadt teil anjepaßt r ja und denn haben- hab ick anjepaßt K: W: •-Ja. B: mir jedacht, mein Jott, die denken alle die nach Berlin kommen wir Berliner leben jeden Tach so wat, nech? K: Ja. B: ((laughs)) Sowieso also jetz wat woll'n Se nu jetz machen? Jehn' Se jetz hier jeh'n Se die Keithstraße hoch und dann seh'n Se die Kaiser-Wilhelm-Jedächtniskirche und denn jeh'n se links den Kurfiirstendamm, denn werden se schon werden se schon wat finden.

^ '"" indicates primary stress, secondary stress in an utterance. The passage is spoken in a typical Berlin vernacular intonation panern with rapid speech flow.

Convergence, Discourse and Variation

163

(B: Formerly it was on Potsdamer Platz, there was Cafe Vaterland, that was subdivided with a Turkish eeh room with a mocca room then there was the Rhine Terraces there was every half hour a thunderstorm came up a thunderstorm there was a (artificial) panorama and there eeh eeh that was was eeh eeh illuminated and there a thunderstorm came up then every half hour there was the RhLne-terraces and then they played appropriate music and then there was the Prater, well everything fit to the atmosphere and the the the the city the district well and then I though to myself my god, they think, who comes to Berlin, we Uve like that every day, huh? Anyhow, weU now what do you want to do now? Go now here go up Keithstrasse then you'U see the Kaiser-Wilhelm-memorial church and then you go you go to the left along Kurfürstendamm. Then you'U then you'U find something.)

This is the normal, fluent and familiar style of a narrative (cf. Labov & Fanshel 1977). Psychologically speaking, B identifies with the past when Ufa in Berlin was not as alienated as it is now. We find the typical U B V markers: spirantization of / g / , final / s / in pronouns/determiners becomes [t], the diphthong / a u / in auf is realized as [u] (among other features; for a füll account of the dialectal features see Schlobinski 1987). Convergent behavior of B can be found in various places in the transcript. There is a semantic shift in the foUowing passage: B: Also weeßte Jungs da muß ich eu da muss ick Ihnen ehrlich s a g e n . . . (You know guys I have to teil you ("tu") I honestly have to teil you ("vous") . . .) With weeste jungs (you know, guys) and Euch B selects a level of style that presupposes a certain intimacy and mutual confidence. There appears to be a tendency to reduce solidarity as evidenced by the shift in address forms. The repair of Euch by Ihnen is due to the fact that politeness norms in encounters with strangers cannot easily be broken. A convergent pattern of intonation and speech rate is found in the following utterances/lines'': B: Ick bm zwar 'n Berliner aba da kann ick kann ick Ihnen nich die ((laughs)) diericht'je Antwortjeb'n. (ThoughI'maBerlinerIcan'tican't give you the right answer.)

B uses a more formal (Standard German oriented) register. His typical intonation pattern is weakened. The speech flow is slower. There are hesitations and repetitions; the idiosyncratic speech delivery converges more to the Standard; the style shows that the response to the initial question in the encounter is dispreferred.

164

Norbert Dittmar & Peter Schlobinski

B: Ick ick weeß ja nu nich überhaupt wät se kumma ick weeß nich übahaupt wat wat wat Sie sich so wat wat Sie so erleben möchten dit dit jeht mir nich innen Kopp rin ((laughs))... {Idon'tknow atallwhatyou yoH see, I don't know at all what what what you what you want to experience that that isn't sinking in...) B is very embarrassed about how to answer K's and W s question. His stuttering indicates convergence as a marker of 'high personal insecurity'. Repetitions, rising Intonation at the end of chunks and clauses and the dropping of broader UBV features produce the Impression of convergent behavior because of interpersonal insecurity. Thus, the feeling of not being able to help and the apparent desire to get rid of these two West Germans seems to push the Speaker into a more convergent behavior in order to overcome the divergence between request/obligation and personal need/ desire. The same level of style holds for the following utterances: B: Na also dazu bin ick zu alt ick bin über siebzig Jahre also weeß ick wo'n Schuppen is " ((laughs)) also dit tut mir K: - ((laughs)) W: L ((laughs)) B: leid also ick druck Ihn' Daumen daß'e noch wat finden K: Ja, ja. B: ja vi^ie lange w^ollen Se de " nn in Berlin K: L J r a so zwei drei Tage noch. W: L Ja (wir ham noch) ja. B: Na bitte sehr ((laughs)) ick also ick kann Ihnen sonst n-nicht helf 'en Sie K: Lja. B: kämen ja eigentlich hier ooch hin aber dit is warten Se mal jetz wo sind wa denn jetz h i e r . . . (B: You see, F m too old for that, I'm over seventy, well do I know where there is a joint (laughter) well I'm sorry about that, well I'Il keep my fingers crossed that you'U find something well how long do you want to stay in Berlin? Ah well then, I eh, I can't help you any further you could actually go to this place too but this is, hold on now, where are we now, then . . . )

B tries to dose the interaction, but corrects his opening up of a closing by a self-initiated question that keeps the conversation alive. These utterances have the same stylistic function as those above: 'convergence' in this passage is a move in the direction of the strangers to the city, an attempt to follow the demands and norms of politeness. B reacts to a request for help,

Convergence, Discourse and Variation

165

widely accepted as legitimate in the speech Community, although the 'urban native counselor' has apparently no interest in the topic; nor does he have a positive attitude towards the night-club milieu of 'modern' Berlin. We have seen that 'convergent' behavior of B is his preferred way of coping with the social demands of the Situation arising from his intimate knowledge of the city and the incompetence of the two West Germans in urban matters. Do we also find 'divergent behavior' on the part of B? Recall that we defined 'divergence' parallel to 'convergence'. It reverses the direction of deviation from the normal, habitual level of speech. This deviation is marked, salient and directed; it carries social meaning for interpersonal relationship. By 'divergent' speech patterns, B marks his local Berlin identity in a salient way. Several examples provide evidence for this. (a) The first type can be called 'strong Performance of Berlin bluster' with an ironical function of divergence. Dissociating himself from the Berlin prostitution and night-club milieu, B tries to get rid of K and W by giving them a final direction for their planned enjoyments: . .. und von der kaiser-wilhelm-jedächtnis-kirche ab jeht links der kurfürstendamm also dann sehn se himmel und menschen und denn jehn se noch mal den jeruch nach - ick drück ihn' (die) daumen ('and from the Kaiser-Wilhelmmemorial church the Ku-Damm branches off on the left, and then you can see sky and people and then you foUow your nose again - I'll keep my fingers crossed'). The 'divergence' is mainly represented by lexical items of the typical Berlin bluster and contrasts with more habitual expressions and denotations of the same propositional content. (b) The local identity of B is foregrounded in . . . nee nu macht mal halblang, et jibt ja ooch noch normale... ('No, don't exaggerate. There are stül some normal people . . . ' ) Here he contrasts the stigmatization of Berlin as a German Las Vegas with his own understanding of the city and his local identity by means of Intonation and broad dialectal features. This divergence marks a basic attitude of the native: 'don't touch positive values of my local identity'. (c) A third type, similar to (a), can be found in the course of B's discourse when he refers to the two 'worlds' in Berlin, the traditional, solid, good one and the modern, corrupt, bad one. The corrupt world (prostitution, drugs, alcohol, night clubs) is referred to as (i) . . . hässlichetPflaster. .. kann ick ihnen nich empfehlen . .. nejänz verrufene strdsse... ('a nasty place . . . I can't recommend it to you . . . a really notorious street');

166

Norbert Dittmar & Peter Schlobinski

(ii) ... da sind Spielhöllen un- und nütten und die sind uff de strässe schon ziemlich offenherzig... ('there are gambling-hells an - and whores and they are rather 'openhearted' in the street already'); (iii) ... da sehn se viele heromsüchtge und betrunkene... ('there you'll see a lot of herein junkies and drunks'); (iv) . . . dat is wat für Spieler und für leute also für für dü-dü-mädchen... ('that's something for gamblers and for people for for youyou-girls (whores)'); (v) . . . mein Gott die denken alle die nach Berlin kommen wir Berliner leben jeden tdch so wat... ('my god, they think, everybody who comes to Berlin we live like that every day'). Intonational divergence is - in this case - an identity marker. In these utterances the speech rate slows down, the pitch falls, the fluency is disturbed. Clear contrasts to the preferred (old) Berlin are established. Lexical items are also part of this. In this threefold way, emotional distance to persons, behavior, values and states is marked by the reverse direction of 'convergence'. Both fulfill the same function: moves deviating from the normal, habitual mode of speech in order to mark the social territory of identity. 2.2. 'Convergence' as a strategy of survival in extremely difficult communication situations (a) Data Below we will analyze the interaction between a Greek Immigrant, who had been living in Berlin for two months at the time of the recording and two native Germans. Janis, the Greek migrant, had never learned German before coming to Berlin. Düring the first two months of his stay, he learned only a couple of German words, namely those he was exposed to most frequently. Since Janis has a good job, he is interested in learning German; but, according to his own words, he likes to learn German in communication, not in the classroom-situation of foreign language teaching. One of the two Germans, A, knows only a few words in Modern Greek; the other, I, learned Old Greek some 20 years ago, but doesn't know any Modern Greek. The talk among these three is in German; as they don't know each other, the first topic of the conversation naturally is their biographical background. From a linguistic point of view, verbal means are needed for establishing different kinds of reference in the conversation: names, common basic nouns like Schule (school), Kind (child), Arbeit (work) etc., local and temporal reference; these expressions have to be

Convergence, Discourse and Variation

167

coordinated, of course, by some simple grammatical rules, i.e. copula, number, gender and case agreement, word order (among others); but even if there is little or no grammar available, understanding may be obtained by simply listing expressions / words in a linear order. Strategies are needed to bridge the gap between appropriate lexical items; the pragmatic problem then is this: what shared knowledge can the communication be built upon? The conversation is about G's place of birth (near Thessaloniki), his time in elementary school (6 years) and in high school (5 years). G did not pass the school-leaving exam because he was very much interested in boxing and had no ambitiöns in high school. Apparently, G had been boxing since he was ten years old. It seems that he was a good boxer. The conversation starts with the question why G did not become a great boxer after leaving school, but learned gastronomy instead. (b) Analysis On several levels the conversation shows what we can call 'pragmatic convergence'. The pragmatic problem is this: there is no shared body of words. We will examine the kinds of strategies by which the participants can survive in the conversation. The following lines pose a severe problem of tense and aspect: the question is whether and when the finite State of an activity (here: boxing) can be marked. I: (8.0) Warum (.) heute, heute, ne (.) heute? (2.0) Warum bist du nicht großer Boxer heute? (2.0) Warum heute (.) Restaurant und nicht boxen. (4.0) p G: In Berlin in Berl ün. Nein nein. I: [ Ja ja. Nein (.) nicht boxen (.) G: Nein. I: Nein. (1.0) Überhaupt nicht mehr. Schluß fertig (.) G: Kaputt kaputt I: Fertig kaputt? Ah ((smiling)) is kaputt, ja? I: G: I: G: I: G: I:

(8.0) Wry (.) today today. huh (.) todayf (2.0) why aren'tyou great boxer today. (2.0) why today (.) restaurant and not boxing (4.0) in Berlin in Berlin no no iyesyes no (.) not boxing (.) no no (1.0) not at all. the endfinished (.) finished ibroken broken broken?ah ((smiling)) is broken, yes?

168

Norbert Dittmar & Peter Schlobinski

In Order to 'survive' in this communication an expression has to be found that is able to converge on two different linguistic Codes. I tries to mark the finite State of boxing by using the words Schluss (end) and fertig (finished). Instead of picking up these expressions, G uses - apparently interpreting the Intention of I - the adverb kaputt (broken), one of the best known words in migrants' learner varieties of German. It should be mentioned that earlier on, both Germans already had the problem of marking the finite State of an activity by an expression that could be understood by G. Although they used several expressions, they didn't achieve their aim. However, with kaputt they now have a clear offer of convergence. I converges on kaputt by first equating/ertjg with kaputt and then using this expression for marking the finite State of boxing. From this passage on both Germans use kaputt to signal the finite State of something: I: kronos? Ja (.) boxen kaputt G: kronos. kaputt ja I: wann (.) wann (.) wieviel Jahr? (7.0)

ja wann (.) ja

I: kronos? Yes (.) boxing broken yes when G: kronos. broken yes L yes I: (.) when (.) when (.) how much year?

With kaputt, the Interviewers, however, gain only one joker 'm the process of understanding. Their aim is to get to know at what time G gave up boxing. I tries to elicit this by means of the following paraphrases wieviele Jahre (How many yearsf): I: G: I: G: I:

Aha, wann kaputt? wieviele Jahre? (2.0) achtzig (.) ehm ein'achtzig (.) zwei'nachtzig (.) kaputt? Kaputt, kaputt. Wann? Lange schon (.) kaputt (18.0) eh wie- wieviel Jahre? Wieviel Jahre? Das Jahr (4.0) kronus eh kronus G: öh eh I: ist das die Zeit} I:

aha - when broken? how many years? (2.0) eighty (.) eighty-one (.) eighty-two (.) broken? G: broken broken. I: when? for a long time (.) broken? (18.0) eh how how many years? the year (4.0) kronos eh kronos is that time? G: öh eh

Convergence, Discourse and Variation

169

In Order to convey the concept of 'yea.r', he adds achtzig, einundachtzig, zweiundachtzig (80, 81, 82) wann, lange 5cÄo«f'(Vhen, a long time ago'), chronus, ist das die Ze/t^Cchronus, is that the time'). With chronus ('time'), a Greek word that the informant understands, and boxen kaputt ('boxing Bröken'), the participants can converge semantically - finally - with respect to a mutually understandable meaning. Two concepts had to be transmitted: (i) the finite State of an activity and (ü) a point in time not given clearly. We can infer from the context, however, that G stopped boxing soon after leaving school and before beginning to work in a restaurant. 'Convergence' in this conversation is governed by pragmatic principles which make mutual understanding possible. Two ways of convergence can be distinguished: the foreigner and the native negotiate an expression from the native's first and foreigner's second language (in our case kaputt), or they agree on an expression from the foreigner's first language (i. e. loan words, projected meaning etc.). In this intercultural communication (with mutually exclusive languages) we find bidirectional convergence which is different from unidirectional convergence (i.e. to the Standard). There is another prominent feature of convergence in the passage presented above: foreigner talk. The two natives converge to the single word utterances of the migrant, thus simplifying their syntax, morphology and reducing semantic/lexical complexity. This kind of 'foreigner talk' has been described thoroughly by Ferguson & de Bose (1977) and Dittmar & von Stutterheim (1985). Foreigner talk can be misinterpreted in the sense that it downgrades the addressee (Bodemann & Ostow 1975). The above conversation, however, clearly shows that the Situation of not understanding each other as well as the pragmatic mode of communication are extremely strong constraints, so that foreigner talk can be interpreted as extremely convergent behavior. 'Divergence' cannot be found in the conversation above. The problem of mutual understanding overrides all sociopsychological factors involved in signalling social identity. This means that if there were no semantic and pragmatic convergence, there would be no understanding at all. 'Convergence' in this case is a necessary precondition of communication and understanding. The convergent behavior, we found, is not habitual; it is below the level of consciousness and therefore not controllable. It is marked and uncertain as well as interpersonal in the sense that it can never happen with a single person.

170

Norbert Dittmar & Peter Schlobinski

2.3 Strategie speech behavior The following analysis is based on a specific kind of 'propagandistic discourse'. In Berlin's biggest department störe we recorded a pretty thirtyfive-year-old woman, who was selling cheap Ornaments there for a twoweek period. As her income depended on the amount of goods sold, she tried to attract as many customers as possible. We made two recordings at an interval of one week. The transcription is given in Schlobinski (1982:216-234). Looking at its macro-structure, the 'propagandistic discourse' has two parts: (1) the discursive work' of the booster to get the attention of the people strolling by and (2) the acts of buying/selling. Together they constitute a discourse cycle. A discourse cycle may be defined as the "systematic ordering of recurrent and functional equivalent discourse segments within a discourse unit which are marked at the beginning and at the end of the discourse unit" (Schlobinski 1982:74). Part one stereotypically begins with the woman's focussing first on the imaginary interactants and then on herseif: Meine Damen und Herren, Freunde der Nacht. Ich mach'jetz was Besonderes. ('Ladies and gentlemen, friends of the night. Fm going to do something special'.) She closes her discourse by thanking the audience and asking them to recommend her: Ich sag mal Danke für's Zuschauen. Und machen Sie bitte für mich Reklame. ('I say thanks for your attention. And teil your friends and neighbors about me'.) After the first part of the discourse the acts of selling/buying follow which are overlapped by selling strategies. In the following analysis we want to summarize the prominent features of the linguistic repertoire of P. We will also argue that the rieh Variation in the linguistic repertoire of P is a part of Strategie speech behavior and a function ofa specific "linguistic market" (Bourdieu 1977). The normal and unmarked speech behavior of P is a variety dose to Standard German. It is used by P in private communication as well as in the propagandistic discourse. What sets the propagandistic discourse apart are the following recurrent and salient linguistic phenomena:

^ The term 'discursive work' refers to the activities of a Speaker to get the attention of an audience, to persuade and to involve (to 'captivate') interlocutors in discourse by means of certain speech pattems.

Convergence, Discourse and Variation

171

(1) Code-switching and/or code-shüting^ P has a more or less elaborated competence of different languages and dialects. She switches into (i)

Berlin urban vernacular (a) rounding of /i/ before s and liquids, e. g. [zYlbgr]. (b) regressive nasal asimilation, [habt;] ^ [han;i]. (c) monophthongization o f / a u / t o [o:] (d) 1-deletion in die adverbial ma/ (e) [kukn] - [ki±n] 'to look'. (ii) Swabian dialect (a) suffigation of [b] or [Ii], (b) deletion and assimilation in [weni9] - [weq] (c) tagquestion: [gel]. (d) intonation contours. There also seem to be some other features. We are not sure if all features are in fact Swabian, because, for example, the suffix Ii occurs in Swiss AUemanic, too. It may be a kind of Southern German koinL The switch into the forms listed above has the connotation [+ southern German] for Berliners. (üi) English The white pearlfor the white heart. (iv) French Regardezpour la demonstration (v) Spanish Senoras and senoritas. The switch into Spanish occurs only in adressing the audience. (vi) Unidentified 'Russian' particle As a kind of re- or prefocussing particle she uses [sitnotnik], which sounds Russian, but which is not a lexical variant of any Slavic language. (2) Idioms Silber ist gut gegen Haarausfall. Und tut der Mutti gut, wenn'sihrder Vati schenken tut. ('Silver is good for hair loss. And it makes mother feel good, if father gives it to her.') (see 11/169)

We believe that there is no clear distinction between code-switching and code-shifting in an analytic sense. In the broadest sense both terms refer to the switch of a Speaker from one consistent set of cooccurring rules to another.

172

Norbert Dittmar & Peter Schlobinski

(3) Rhymes Variation of a nursery rhyme: One, two, three,four,five, six, seven, wo ist meine Frau geblewen. ( ' . . . where's my wife?') There are many other interesting structures in the discourse of P, e. g. different rhetorical figures. We only want to mention the most salient features, however. In the first step above we listed some elements of P's stylistic repertoire. In the following section we want to approach the question of what function these variants have. Our hypothesis is that the use of the linguistic features listed above reflects the function of Strategie speech behavior. This means that the use of these features marks the contrast to the normal, unmarked speech behavior of P. It is planned and oriented towards purposes. There is a high level of speech control and consciousness, a high degree of cognitive stabiÜty characterizing the speech Situation. Let US Start with a language game. Imagine you have to Interpret the following sequence of interaction: P: A: P: A:

Do you speak English? No. Parlez-vous frangais? Un ( ).

The Information you have is that P is a white woman and A a black man. Without further contextual information we would assume that P is trying to Start a conversation in the language of A. The presupposed knowledge of P is that black people can speak English, and if not, they can speak French. In a hypothetical sense, the speech behavior of P is convergent, because P makes her speech variety similar to the one of