Focus on Language and Ethnicity: Essays in Honor of Joshua A. Fishman (Focusschrift in honor of Joshua A. Fishman on the occasion of his 65th birthday)

LANGUAGE AND ETHNICITY

Professor Joshua A. Fishman Distinguished University Research Professor Emeritus Yeshiva University, New York

LANGUAGE AND ETHNICITY FOCUSSCHRIFT IN HONOR OF JOSHUA A. FISHMAN ON THE OCCASION OF HIS 65TH BIRTHDAY Volume II

Edited by

JAMES R. DOW Iowa State University

JOHN BENJAMINS PUBLISHING COMPANY AMSTERDAM/PHILADELPHIA 1QQ1

Library of Congress Cataloging-in-Publication Data Language and ethnicity / edited by James R. Dow. p. cm. -- (Focusschrift in honor of Joshua A. Fishman on the occasion of his 65th birthday ; v. 2) Includes bibliographical references and index. 1. Anthropological linguistics. L Dow, James R. II, Fishman, Joshua A. III. Series. P35.L268 1991 306.4'4'089 -- dc20 91-699 ISBN 90 272 2081 6 (Eur.) /1-55619-117-0 (US) (alk. paper) CIP © Copyright 1991 - John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher.

Preface

In the academic world we often pause to honor and celebrate a colleague's birthday or passage into retirement with a Festschrift. On rare occasions we attempt to focus such volumes on the topic which the celebrant pioneered and devoted his life and work to. Just such an occasion, Joshua A. Fishman's sixty fifth birthday, has caused us to produce not only a Festschrift, but also this three volume Focusschrift. It was indeed necessary to divide the effort, first because Professor Fishman has touched and influenced so many people around the world, that only through multiple volumes could we even begin to include all those who wanted to participate. Secondly, however, and far more importantly, it was necessary because Joshua Fishman has in fact pioneered, worked in and on, and he has contributed an enormous body of literature to each of the areas in this trilogy, so that it seemed only appropriate to call for focused studies on each of the topics he devoted his academic and personal life to. This three volume set thus focuses on Bilingual Education edited by Ofelia García, Language and Ethnicity edited by James R. Dow, and Language Planning edited by David Marshall. Each volume contains recent studies by well-known scholars from around the world, and each one attempts to document at this point in time (1990) where we are in the study of these topics. With these volumes we want to indicate to the contemporary world of linguistics, sociolinguistics, and sociology just how great an impact Joshua Fishman has had on all of us. For future generations we want to leave a record of our respect for this man, by furthering areas of academic research which he helped found, promote, and shape. Here then are three collections on topics which Fishman researched and wrote about throughout his entire academic career. All of us, editors and authors alike, hope that our work will contribute still more to our understanding of these complex but very gratifying subjects, and that they will thus be challenging and stimulating to the readers. Ames, IA New York, NY Grand Forks, ND July 1991

Table of Contents

Preface

v

Introduction

1

Marion Lois Huffines: Pennsylvania German: "Do they love it in their hearts?"

9

Werner Enninger: Linguistic Markers of Anabaptist Ethnicity through Four Centuries

23

KathrynA. Woolard: Linkages of Language and Ethnic Identity: Changes in Barcelona, 1980-1987

61

Joanna Courteau: Language and Ethnicity: The Case of Rosalía de Castro

83

Carol Myers-Scotton: Making Ethnicity Salient in Codeswitching

95

Andrew Gonzalez, FSC: Cebuano and Tagalog: Ethnic Rivalry Redivivus

111

David E. Lopez: The Emergence of Language Minorities in the United States

131

Calvin Veltman: Theory and Method in the Study of Language Shift

145

Alan S. Kaye: A So-Called Dialect of English

169

Rakhmiel Peltz: Ethnic Identity and Aging: Children of Jewish Immigrants Return to Their First Language

183

Rolf Kjolseth: W(h)ither Ethnic Languages and Bilingual Education in the US? Crisis & the Struggle between Hegemony & Humanism

207

References

225

Contributors

245

Index

251

Introduction

About one month after being asked to assemble studies on "Language and Ethnicity" as part of a trilogy in honor of Joshua A. Fishman I received a copy of his new book entitled Language and Ethnicity in Minority Sociolinguistic Perspective (1989). The book does not represent new work by Fishman, but is rather an anthology of his own studies on the topic since 1972 along with some preliminary and concluding summarizations written especially for this selection. My reactions to the tome were many, not the least of which was a sense of inadequacy in carrying out my charge to invite and draw out scholars from around the world to contribute their most recent research and thinking on this same topic, for inevitably the new collection would be compared to Fishman's own collected research on the topic. Joshua Fishman has been and still is one of the most profound and productive academicians to deal with the study of language and ethnicity, and he is a passionate scholar of language as perhaps the most significant marker of ethnicity. The task at hand was thus a formidable one. It was not only necessary for me to solicit contributions for the new work, but also to read or reread Fishman's own collected research on the topic in order to assure that the studies to be included here would be new and, hopefully, would further or challenge primary theoretical work done by Fishman himself. In the Introduction to his new collection Fishman admits that his own views on language and ethnicity are not unbiased, stating that he would not take seriously any one else's claim to an unbiased viewpoint on this subject. He then says: The best I can hope for is (a) to state my biases, (b) to seek as much contact with empirical evidence as is possible, since such contact may lead me to revise my views, and (c) where no such contact is really possible, as is the case with many of our most central convictions, to accept the fact that social scientists (including myself) are only human, i.e. that they have values and that these values do show, must show, whether they are on my side of the fence or on the other. (Fishman 1989: 2)

2

James R. Dow

Having read these words, the task of assembling a new volume on the topic of language and ethnicity, a topic which virtually consumed Joshua Fishman throughout his forty year academic career, seemed to come back into focus. There are, after all, numerous scholars around the world, some of whom are his own students, who have also devoted much of their academic careers to this emotion laden topic. Many write and teach with a similar passion, and most indeed are on the same "side of the fence" even though others clearly are not. One group of those who are on the other side is singled out by Fishman in his Introduction when he speaks about some few sociolinguists who have "'discovered' the principle of 'self-serving interest' in commenting on minority efforts to foster their language and ethnicity goals" (Fishman 1989: 1). The thoughts and the research results of many scholars have found their way into this present volume, most of whom are clearly in sympathy with Fishman's own conceptualization of language and ethnicity. Many other scholars indicated interest, but were not able to contribute for any number of reasons. The letters and telephone messages from contributors as well as potential contributors and well wishers read in some cases like testimonials to the intensity of feeling concerning the topic. Even a cursory reading of the titles of the studies included here will immediately indicate that, even though the commission to the contributors was to focus sharply on the concept of language and ethnicity and to offer their latest research on the subject, what resulted were widely divergent studies. It was, however, most gratifying to note that many of the papers reflected the variety of approaches Fishman himself had assumed over the years, indeed the section headings in his new book on Language and Ethnicity parallel many of the papers included here. An attempt was made to solicit research papers from around the world so that the reader would gain an international perspective on recent research devoted to the matter of language and ethnicity. There are in fact papers which deal with Europe, Africa, and the Philippines, but more than half are devoted to the language of ethnic minorities of the United States. It is distinctly within the realm of possibility that the political and social upheavals around the world during the years 1989 and 1990, when these studies were being solicited, caused researchers from elsewhere to deal more pragmatically with their own local realities than with their research. Those studies which finally did find their way into this collection all have in common the study of the language practices and behavior of groups which constitute ethnic minor-

Introduction

3

ities within larger and linguistically complex societies. It is an impressive collection of theoretical perspectives and empirical data, which is after all international in scope. A composite bibliography has been assembled and appears following the essays. In the opening two articles Lois Huffines and Werner Enninger look closely at the ethnic identity of speakers of Pennsylvania German. Even though this minority language has served the communicative needs of its speakers for over two centuries, it is clear to all scholars who deal with it actively that it is indeed dying. Huffines, who has had her finger on the life pulse of Pennsylvania German for over a decade, moves beyond the simple life-death metaphor and asks quite poignantly how important is it to be Pennsylvania German and does the language itself express that ethnicity. She looks closely at sectarians and non-sectarians, specifically at the slow death of Pennsylvania German as it proceeds along two different avenues, in order to answer the question of how PG does or does not express ethnicity. Werner Enninger, the foremost European scholar dealing with Pennsylvania German, has long dealt with synchronic perspectives, building on extensive field work among the Old Order Amish of Delaware. In the recent past, however, he has turned most of his attention to diachronic investigations, i.e. back to the European origins. In his lengthy study he reconstructs the linguistic markers of ethnicity among the Swiss Täufer, from their very beginnings down to the contemporary Amish/Mennonite culture of North America. He traces the concept of "peoplehood" through distinctive ways of speaking and conversational strategies, later through biological descendance, thus identifying a "new ethnicity" added to the original "old ethnicity". Following their migration to other language areas this "old ethnicity" comes back into play because of their language minority status in their new homeland, but Enninger is still not satisfied with the "revived original ethnicity", and thus pushes his theme of language and ethnicity into the realm of deliberate and systematic deviation from established rules of appropriateness with a resulting de-canonization and re-canonization of established rules of social appropriateness. Here he comes into agreement with Huffines and suggests that even anglicization, i.e. language shift, does not automatically mean de-ethnization. Two of Spain's linguistic minorities are next subjected to close scrutiny, once again from a contemporary perspective and also from an historical one. Kathryn Woolard has been closely monitoring Catalan, once a threatened minority language, now making a bid to become a principle public language.

4

James R. Dow

Traditionally Catalan was used only by native speakers, but full catalanization of an autonomous Catalonia now depends on recruiting immigrants as second-language speakers. Woolard's ethnographic and experimental data are used to explore the extent to which policy changes have led to a restructuring of the language-identity link. My colleague here at my own university, Joanna Courteau, has long functioned as a literary scholar looking at both Spanish and Portuguese literatures. It is only natural then that she would devote considerable attention to a Galician poet, Rosalia de Castro, who nevertheless wrote most of her poetry in the dominant Castilian language of Spain, perhaps, according to Courteau, with the motivation of revealing her own ethnic identity as a Galegan through the dominant language of Spain. The two papers which do not deal with Europe or North America treat the minority languages of Kenya, and the long-standing ethno-linguistic rivalries in the Philippines. Carol Myers-Scotton, who has been one of the most productive scholars on the phenomenon of codeswitching, devotes her study to the motivation of negotiating the saliency of ethnicity in a conversation, using the multi-ethnic society of Nairobi, Kenya as her data source. Here she singles out, from the four social functions for codeswitching which she has identified in her recent research, one specific function for closer analysis. Codeswitching as a marked choice is investigated as a means for bringing the speaker's ethnicity into the rights and obligations characterizing conversational exchanges. Andrew Gonzalez' sensitive treatment of the ethnic rivalry between the speakers of Cebuano and Tagalog in the Philippines, sees language as a symbolic form of resistance to Tagalog domination and imperialism on the part of Cebuanos. The latter have obviously associated their identity with their language, but they have not threatened secession, not yet anyway. With a strong movement toward a Tagalog-based Pilipino (now called Filipino) national language they see themselves becoming disadvantaged, visa-vis the Tagalogs and even becoming second-class citizens. The issue becomes more complex when the self-proclaimed nationalists countered that the system of education should become nationalized in Pilipino as soon as possible. Several models for both federalization and language development are proposed, particularly the Swiss and Singapore models for what is locally called an "aggrupation" of ethnic tribes. The final five papers all deal with ethnic minorities in North America. David Lopez and Calvin Veltman, both of whom have published extensively on language maintenance and shift in the United States, contribute from their

Introduction

5

most recent research. Lopez sees language patterns and other cultural aspects of ethnicity as having been less significant than racial or religious attributes in the past. Nevertheless, Asians and Hispanics have increasingly become defined as "language minorities" during recent decades. His paper traces how language, rather than other possible characteristics, have emerged as a central defining attribute for these groups. He reviews three areas of language policy: bilingual education, multilingual election services, and the recent upsurge of "English Only", demonstrating that an approach that emphasizes conflict, group interests and the exercise of power best explains the intersection of ethnic group interests and public policy. Calvin Veltman reduces his long experience and extensive writing on ethnic minorities down to a specific theory of the measurement of language shift, and then addresses the methodology which needs to be employed in order to correctly interpret the data obtained. He divides his paper into four parts: a brief discussion of the concepts retained for analysis, the development of appropriate indicators of language origin and current language practice, the procedures by which language shift data may be correctly analyzed, and the implementation of demographic models for minority language populations. Two studies are devoted to Jewish languages. Alan Kaye explores JudeoEnglish or Jewish English as a so-called dialect of English. His paper examines the problems associated with these terms and shows them to be ambiguous, imprecise and misleading. Rakhmiel Peltz, who does empirical research on spoken Yiddish, looks closely at the children of Jewish immigrants who are increasingly returning to their first language as a statement of their Jewish ethnicity. His fieldwork was done in South Philadelphia and in several smaller cities in Massachusetts, and his archival work comes from autobiographies found in the American Jewish Archives. For many Yiddish speakers exposure to the language generally diminished over the years. Recently, however, these people have come to view Yiddish as a cornerstone of their Jewish identity, and they respond enthusiastically when Yiddish is spoken. Peltz discusses these findings in relation to our understanding of identity development during the aging process and its significance as an element of ethnic identity. The final paper included here is Rolf Kjolseth's reflective and speculative discussion of ethnic languages and bilingual education in the United States. His paper reads more like a lengthy editorial on the topic of language and ethnicity, and it is one which could indeed have found its way into any one

6

James R. Dow

volume of this set, since it ponders an entire range of questions on ethnic languages, bilingual education, and thus finally also deals indirectly with the question of language planning. His thesis is that the current "hegemonic" world view in the United States tends to negate both ethnic languages and bilingual education programs, and that only a paradigmatic shift to an alternative "humanistic" world view offers the possibility for supporting ethnic language use and bilingual education. Finally, a personal note. I too had planned to contribute a paper on my work with one of the two German-speaking minorities in Iowa which I deal with, the Old Order Amish or the Amana Colonists. The papers by Huffines and Enninger, however, convinced me that still another paper on Pennsylvania German would have given this volume too many studies on one specific ethnic group. My recent work with the historical origins of the Amana Colonies, which is heavily oriented toward the ethnic identity of this group during its earliest years, would certainly have fit, but I decided to include it in another volume on Minority Languages / Language Minorities (1991). I wanted to mention this work and this volume, because in a large sense Joshua Fishman is indirectly responsible. When I spent the summer at Stanford working with Fishman as a participant in an NEH Summer Seminar, I read heavily in the area of German minorities in the United States and continually confronted the name of Werner Enninger. Professional and personal contact ensued and intensified, and as a result I have just spent a semester in Essen, as a Guest Professor, at the invitation of Enninger, where we taught together, discussed in detail our mutual interest in German-speaking Americans, and also carried out a small conference on "Minority Languages". My paper on Amana, which might well have been included here, appears instead as part of the proceedings of that meeting (Dow 1991). With this brief comment I want to thank Joshua Fishman for encouraging me to continue my work with the German minorities of Iowa, thereby putting me in direct contact with a group of German researchers who are also most interested in the concept of language and ethnicity. Finally, several expressions of gratitude are in order to all who have helped in preparing the manuscript for this volume. Primary thanks go to the contributors themselves who endured endless letters and telephone calls reminding them of dates, text preparation guidelines, and even finally some threats of non-inclusion if I did not receive their papers in time to edit and prepare the final electronic manuscript for the publisher. I am also deeply

Introduction

7

indebted to Werner Enninger for supplying me with working space and atmosphere, secretarial help, personal and professional support while I worked to prepare the final copy of the manuscript during a Guest Professorship at the Universitat-Gesamthochschule Essen in the summer of 1990. Susanne Zinner and Heidi Kroheck of Essen were most helpful in preparing the composite bibliography and retyping one of the manuscripts when a disk didn't appear on time. Hans-Peter Mai of Essen and my chairman at ISU, James S. Ruebel, helped me enormously with technicalities of converting from various word processing programs so that the texts would be unified and the composite bibliography could be automatically sorted. To all my sincere thanks.

Pennsylvania German: "Do they love it in their hearts?" Marion Lois Huffines Bucknell University

I remember very clearly sitting in Joshua Fishman's office one sunny afternoon during a LSA Summer Linguistic Institute describing to him how Pennsylvania German was dying among the nonsectarian Pennsylvania Germans. Looking up at me and smiling, Fishman asked, "Do they love it in their hearts?" Since that moment, the life-death metaphor as applied to language has taken on for me a less menacing finality. Do Pennsylvania Germans love Pennsylvania German in their hearts? The question is a more meaningful one than simply asking whether Pennsylvania German is dying. The answer lies in how important speakers perceive their ethnicity to be and how strongly they wish to express it linguistically. For minority languages in the United States, language death is an almost inevitable outcome of contact with American English. The promise of social and economic advancement offered by mastery of English eventually overcomes the most fervent desire to retain another language. The number of social contexts in which speakers can use the minority language steadily declines. Without continued immigration from the language homeland, the number of fluent speakers gradually decreases, and eventually no social context remains in which it is appropriate to speak the minority language. The language is said to be "dead". Pennsylvania German, although it has enjoyed a long history in America, is no exception to this general process. How is Pennsylvania German dying? As suggested above, the life-death metaphor does not serve the study of language usage very well. The more meaningful and complex question is one of ethnicity: how important is it to be Pennsylvania German and does Pensylvania German express that ethnicity? The purpose of this study is to investigate the mechanisms of language death by contrasting Pennsylvania German usage and form in two environ-

10

Marion Lois Huffines

ments: 1. among Amish and Mennonite sectarians, where Pennsylvania German is spoken as the language for daily discourse; and 2. among the nonsectarians where Pennsylvania German is dying, and the community has almost completely shifted to English. In both instances, Pennsylvania German is in intimate contact with English. This study argues that language death is proceeding along different avenues in each community type. Among the nonsectarians, transmission has simply ceased. Parents do not speak Pennsylvania German to their children, who are, in turn, nonfluent in Pennsylvania German if they acquire any at all. The variety spoken by these younger Pennsylvania Germans exhibits characteristics which demonstrate faulty acquisition. Among sectarians, although it is used extensively for daily discourse, Pennsylvania German exhibits features of convergence to English. Sectarian communities prescribe Pennsylvania German usage, but not Pennsylvania German form, and the Pennsylvania German forms are changing in ways which facilitate both code-switching and translation tasks. Linguistic evidence will be drawn from three different grammatical structures: the Pennsylvania German dative case, the progressive verbal aspect, and the infinitive complement. Each of these points offers perspectives on language death and ethnicity, and each suggests ways to answer the questions: How is Pennsylvania German dying? Do they love it in their hearts?

Procedures for this Study The following observations are based on interviews with 52 Pennsylvania Germans who live in central Pennsylvania: 33 nonsectarians and 19 sectarians. The nonsectarians live in the farm valleys at the nexus of Northumberland, Dauphin, and Schuylkill counties, Pennsylvania. They are classified into three groups: Group N: Native speakers of Pennsylvania German (n=13). The nonsectarian native speakers of Pennsylvania German range in age from 35 to 75 years; all but four are 60 years old or older. All but the two youngest speak Pennsylvania German to their spouses and generational peers, but all speak English to their children. Group 1: First in the family native English speakers (n=9). Speakers in Group 1 are the first in their respective families to speak English natively. They range in age from 32 to 54 years. They speak Pennsylvania German to certain

Pennsylvania German

11

elderly members of the family and community but English to their spouses and children. Group 2: Second or later in the family native English speakers (n=ll). These native speakers of English range in age from 22 to 65. They understand Pennsylvania German, some with difficulty, but seldom speak it. The Pennsylvania German sectarian sample consists of Mennonites and Amish, who range from 24 to 65 years of age. Both groups are characterized by horse and buggy transportation, distinctive dress, and limited education to the eighth grade. All the sectarians are bilingual and speak Pennsylvania German natively. Group M: The Mennonites. The Mennonite group consists of 10 members of an Old Order Mennonite community, also called "Team Mennonites", a term which refers to the horses used for transportation and work power. Group A: The Amish. The Amish group consists of 8 members of a conservative wing of the New Order Amish and one member of an Old Order Amish community. The New Order Amish differs from the Old Order Amish in the interpretation of what it means to be separate "from the world". The New Order Amish in this sample have electricity in their homes; the Old Order Amish informant does not. The interview consisted of three parts: free conversation, translation of English sentences into Pennsylvania German, and description of pictures. The picture description task was especially successful in eliciting comparable grammatical structures without overt reference to English. The topics for all three parts centered on growing up on a farm, farm chores, recipes, butchering, home remedies, and one-room schoolhouses.

Language Usage Pennsylvania German, popularly known as Pennsylvania Dutch, is a German dialect closely related to the dialects spoken in the German Palatinate along the Rhine River. Pennsylvania Germans settled in America during colonial times in farm communities across southeastern and central Pennsylvania. Secondary settlements arose later in Ohio, Indiana, Illinois, and the Virginias. While many of these communities have long ago completely and irrevocably

12

Marion Lois Huffines

assimilated into mainstream American society, some have maintained their peculiarly Pennsylvania German culture and language to the present time. Although all Pennsylvania Germans share many traits and values, the Pennsylvania German population is not homogeneous. It consists of many subgroups, each having a different relationship to mainstream American society, and for each, Pennsylvania German fulfills different communicative and symbolic functions. The death of Pennsylvania German has been predicted for decades, but Pennsylvania German is still spoken natively by Old Order Amish and a majority of Old Order Mennonites. Members of these sects speak Pennsylvania German in their homes and communities; they learn English in school and use it when conversing with outsiders. English also serves the sectarian communities for reading and writing and will be used in families with children who suffer from learning disabilities (Huffines 1980b). The increasing number of preschoolers who speak some English indicates that English is being used in the home at least to some extent. The Old Orders also use an archaic form of standard German for the scriptures and hymns in their church services. No one converses in this standard German form, and even the passive knowledge of this German variety is gradually fading, especially among the young people. Other parts of the church service are carried out in socalled Amish High German, a term coined by Frey (1945; see also Enninger 1986a). Amish High German is essentially an upper register of Pennsylvania German, infused with formulas and phrases from the archaic standard German variety just mentioned. Some ministers manage to use more German in their sermons than others. Less conservative Amish and Mennonites tend to use more English in their daily lives and religious services than do the Old Orders. Among the nonsectarian Pennsylvania Germans (often referred to as the "church people" because of their membership in traditional church demoninations), the shift to English is pervasive. The number of native speakers is decreasing. The vast majority of Pennsylvania Germans do not speak Pennsylvania German; some can understand it, but most are monolingually English. In nonsectarian communities Pennsylvania German functions in very limited circumstances; it is used to speak with certain elderly members of the family and neighborhood and in the attempt to keep secrets from children and grandchildren. Children have little opportunity to learn Pennsylvania German because they hear it so seldom.

Pennsylvania German

13

Linguistic Evidence Receding languages typically converge toward the languages with which they are in contact. Native speakers and nonfluent speakers of the dying language adopt features from the dominant, usually prestigious language in order to elaborate the language which no longer fulfills all their communicative needs. If speakers must switch frequently from one language to the other or if the languages fulfill overlapping functions, speakers may integrate linguistic forms and functions of one language into the other. Such convergence may extend to all parts of the grammar, and in this way a language may alter its grammatical structures to match the structure of the language with which it is in contact (cf. Gumperz & Wilson 1977). As Pennsylvania German is used lessfrequentlyand the number of native speakers decreases, proficiency in Pennsylvania German varies widely. Native speakers who continue to use the language with family members and friends exhibit full mastery of it, but younger speakers are nonfluent and make frequent errors. How is Pennsylvania German changing as people use it less? How and under what circumstances does Pennsylvania German converge toward English? Linguistic evidence drawn from the dative case, the progressive aspect, and the infinitive complement provides insights into both language loss due to the interruption of transmission and the circumstances of convergence. Prescriptive rules are derived from Buffington and Barba (1965), Frey (1981), and Haag (1982). Case Usage: Dative of Possession In Pennsylvania German, the case system for nouns, indicated by the endings on noun modifiers, consists of a common case, which fulfills both nominative and accusative functions, and a dative case, which among other functions expresses possession. In Pennsylvania German possessive constructions, the possessor occurs in the dative and is followed by a possessive adjective which agrees with the possessed, the noun it modifies. For example, the sentence "Where is Daddy's book?" is rendered wu is em Daadi sei Buck. The viability of the dative case in this and other functions varies across Pennsylvania German groups, as speakers often use the common case in dative functions. Table 1 shows the distribution of cases used to express possession.

14

Marion Lois Huffines Table 1 % Case of the Possessor (Translation Task) Group

dative

N 1 2 M A

49 43 18 7 2

common 41 39 20 74 76

-s 4 14 37 12 9

other

(Total #)

6 5 24 7 13

51 44 49 42 45

0 0 67 0 7

7 7 6 11 15

(Free Conversation and Picture Descriptions) N 1 2 M A

57 43 0 0 0

43 57 0 100 93

0 0 33 0 0

Nonsectarian native speakers (Group N) respond most frequently with dative forms to express the possessor. The native English speakers (Groups 1 and 2) show considerable variation. Group 2 demonstrates little mastery of the possessive construction, resorts most frequently to the English genitive -s, and also produces the most aberrant forms and other faulty constructions lacking grammatical agreement (listed as "other" in the table). In Group N, common case forms are given by the two youngest speakers, whose usage profiles parallel those of speakers in Group 1. The sectarians consistently use the common case to express possession, especially when not translating. The speech of members in Groups 1 and 2 demonstrates attempts to apply the Pennsylvania German rule, even when the English s intrudes; for example, Groups 1 and 2, in attempts to translate "in my aunt's house", produced for the correct in meinre Aunt ihrem Haus such formulations as in meim Aunt ihre Haus, in mei Aunt sei Haus, and in mei Aunts ihre Haus. Groups M and A use the common case: in mei Aunt ihre Haus. These formulations preserve both the appropriate word order and the correct number of

Pennsylvania German

15

requisite lexical slots. While using the English genitive -s, as in meim Aunts Haus, Group 2's efforts to employ the dative case are apparent. The variant forms of the possessive construction indicates the difficulty which Groups 1 and 2 experience in attempting to acquire Pennsylvania German rules. Their strategies in forming the possessive result in aberrant forms and extensive variation within the groups. Speakers in Groups 1 and 2 do not achieve the conservative norm of Group N because they lack opportunities to hear and practice it. The dative case remnants in their speech point to the loss of a community norm for dative usage due to interrupted transmission. In contrast to the nonnative speakers of Groups 1 and 2, the sectarian speakers (Groups M and A) exhibit a strong community norm; their choice of the common case to express possession is strikingly uniform. Sectarian speakers remember that parents and grandparents said it differently and in a way which they now perceive to be old-fashioned; the change is, therefore, a recent one. A merger of dative and accusative forms and functions is not unusual. Many varieties of American and European German exhibit that feature. Although the German of the Rhine Palatinate maintains the dative/accusative distinction, one cannot necessarily ascribe the loss of the dative case in sectarian Pennsylvania German to the influence of English, but the resulting nominal system parallels that in English more closely. The Progressive Aspect The Pennsylvania German progressive is formed by the auxiliary sei "to be" plus am followed by the infinitive of the main verb. Rules govern the placement of modified and unmodified noun objects in the progressive construction: modified objects occur before am and the infinitive, unmodified noun objects between am and the infinitive; for example: sie is am Buck lese and sie is es Buck am lese. In contrast to Group N, the nonsectarian Groups 1 and 2 treat am plus the infinitive as a constituent unit which they are reluctant to separate (see Tables 2 and 3). Speakers in Group 2 are least likely to place objects between am and the infinitive, and they seldom modify nouns in that position. The sectarians, Groups M and A, expand the field between am and the infinitive and frequently modify objects in that position (see Table 3). In the formulation of the progressive aspect, the placement rules for objects vary across groups. For Groups 1 and 2, the acquired rule entails a restriction

16

Marion Lois Huffines Table 2 % Position of Noun and Pronoun Objects Group N 1 2 M A

am obj V

obj am V

Total #

32 72 61 42 29

133 106 69 110 111

68 28 39 58 71

Table 3 % Modifiers of Noun Objects in the am obj V Position Group

None

Modifiers

*Total #

N 1 2 M A

75 93 92 48 47

25 7 8 52 53

91 30 26 64 79

* The number of items reported for the am obj V position will not correspond across tables because structures containing errors were ignored in table 3.

of options: the unit am plus the infinitive forms a template which enables speakers in Groups 1 and 2 to deal more consistently with troublesome Pennsylvania German word order by eliminating syntactic options in the field between am and the infinitive. In addition, speakers can utilize the template as a structure which contrasts with the progressive in English and thereby distinguish their variety of Pennsylvania German from English. The sectarians, on the other hand, are effectively de-emphasizing that contrast with English by merging the syntactic fields preceeding and following am. These speakers also typically reduce am phonetically from [am] to schwa plus [n], which

Pennsylvania German

17

occurs without sentence or word stress. In the resulting structure, the object placement rule becomes void. One further observation points to Mennonite and Amish innovation: in contrast to nonsectarian native speaker usage, the progressive occurs in a larger variety of grammatical forms. The Mennonites and Amish produce progressive forms in the passive voice, in past tenses, and with the verb duh "to do". The expansion of the progressive usage into these grammatical structures as well as the reduction (or loss) of am as a constituent in the progressive construction result in forms which parallel those in English. The Infinitive Complement The formation of the Pennsylvania German dependent infinitive construction also varies across groups. In prescriptive terms, the Pennsylvania German infinitive takes zu when it depends on nouns and adjectives and when the infinitive complements verbs which are not modals or one of a small set of specified verbs; the infinitive of purpose (the counterpart to Standard German um ... zu) is expressed by far... zu plus the infinitive. In today's Pennsylvania German, the use of zu to indicate or mark infinitives which do not complement modals or the small set of specified verbs is low for all speakers. It occurs more frequently in the speech of the nonsectarians than in the speech of the sectarians, but in no group does the percentage of infinitives marked by zu reach 18% of the total number of marked infinitives. With few exceptions, infinitives marked by far ... zu are present only in the speech of nonsectarian nonnative Pennsylvania German speakers (Groups 1 and 2). The use of far to mark infinitives is frequent for all groups. See Table 4. Table 4 Infinitive Marking Group

zu

% total

N 1 2 M A

26 18 18 17 5 8

12% 10% 10% 13% 13% 3% 3% 6%

far... zw zu

% total total

far far

% % total total

3 11 11 6 0 0

1% 6% 6% 4% 0% 0%

193 161 161 112 192 122

87% 85% 85% 83% 97% 94%

18

Marion Lois Huffines

All groups use infinitives marked by far most frequently in dependent relations with nouns: no hen mir di Chance griegt far sie doch kaafe 'then we got the chance to buy it anyway' For the nonnative speakers in Groups 1 and 2, the second most frequent use of far with infinitives is to complement verbs: sie browiertfar ihn ufflange 'she tries to reach him' s is yuscht am schterde far bliehe 'it is just starting to bloom' The Amish and the Mennonites (Groups A and M) differ from the nonsectarian speakers in that infinitive marking seldom occurs when the infinitive functions as a verbal complement. See Table 5. Table 5 Percentage of Marked Infinitives by Function Group N 1 2 M A

w/ Noun w/Noun 53 45 45 50 50 56 56 58 58

w/Adj. 14 14 14 12 12 16 16 15

Purpose 15 18 18 13 13 19 19 17

Verb Compl. 18 18 24 24 25 10 10 10 10

In addition to the relatively low low occurrence of marked infinitives which function to complement verbs, the Amish and and Mennonites contrast with the nonnonsectarian native speakers (Group N) in their greater use of ««marked umarked infinitives as verb complements: hat der der Daed Daedals alsgeschtart geschtarts schneide s schneide 'dad always started to cut it'

Pennsylvania German

19

no hat ar als browiert Hickory Holz griege 'he used to try to get Hickory wood' The contrast becomes even more pronounced because nonsectarians, especially nonnative speakers, not only mark infinitives which complement these verbs, but also mark infinitives of verbs which historically and prescriptively require no infinitive marking: un hap ghelfe far di fountains butze 'and helped to clean the fountains' ich hap gelanntfar Hinkel schlachde 'I learned to butcher chickens' Among the Amish a confusion seems to exist involving the infinitive complement and the progressive aspect of the verb. As described above, the Amish reduce the progressive constituent am to schwa plus [n], and on occasion, that reduction is not phonetically present. In such instances, the infinitive (without am) occurs without marking in the same position it would normally occupy as a verbal complement (without far): sie sin der Reger gucke 'they are watching the rain' ar is draus Blumme blanze 'he is outside planting flowers' Reed (1948: 243-244) also recognizes the emerging form similarity between the Pennsylvania German progressive aspect and the infinitive construction. He ascribes the problem to the influence of the English "-ing" translation for both. All groups exhibit variation in the infinitive complement construction. All groups clearly opt for far as the primary infinitive marker. Members of Group N preserve the use of zu, but the form is recessive. For the nonnative speakers (Groups 1 and 2), the use of zu alone may be supported by the English counterpart 'to', but far ... zu effectively separates Pennsylvania German usage by making it more complex than the English formulation. The (far) ... zu construction provides these nonfluent speakers with a useful template based on a generalization of an acceptable structure heard from native speakers. Speakers in Groups 1 and 2 generalize infinitive marking beyond those

20

Marion Lois Hufftnes

contexts in which it is required. The template helps them to deal with Pennsylvania German word order while differentiating Pennsylvania German structures from English structures. These strategic uses of (far) ... zu for acquisition purposes account for the relatively frequent occurrence of zu in their speech. The Mennonite and Amish are least likely to mark infinitives, especially when the infinitives complement verbs. Speakers in these groups generalize the Pennsylvania German rule which specifies that marking not occur when the infinitive complements modals and a small set of specific verbs. The sectarians expand the application of this rule by not marking infinitives complementing other verbs and thereby obviate one feature which distinguishes English and Pennsylvania German.

Discussion Specific linguistic behaviors characterize each group. The native nonsectarian speakers (Group N) exhibit the most conservative linguistic norm. Speakers in that group preserve a vigorous dative case, apply the placement rules for objects of verbs in the progressive aspect, maintain remnants of the infinitive marker zu, and mark infinitives which complement the prescriptively appropriate verbs. Groups 1 and 2 lack access to the conservative norm of Group N and display partial mastery of Pennsylvania German rules. Speakers in Groups 1 and 2 produce various more or less aberrant forms of the dative and resort to template building in the use of both the progressive aspect and the infinitive complement. In comparison with the other native speakers in Group N, the Mennonites and Amish (Groups M and A) exhibit substantial change in their variety of Pennsylvania German: the dative has merged with the common case, the placement rule for objects of verbs in the progressive no longer applies, and the rules regulating the nonmarking of infinitive complements of verbs have been drastically expanded. Each of the sectarian changes results in forms which more closely parallel English usage. Groups 1 and 2 deserve special attention. These nonnative speakers of Pennsylvania German are intent on learning better control of Pennsylvania German in an environment with ever-decreasing access to Pennsylvania German and few acquisition opportunities. Group 1 demonstrates mastery of the dative case although they are not successful at every attempt to use it. The progressive am plus infinitive and the far ... zu constructions provide them

Pennsylvania German

21

with templates which contain structures heard from native speakers. Speakers in Group 1 apply these template-based Pennsylvania German rules more consistently than the native speaker norm. Used judiciously, the template strategy enables them to solve problems of speaking Pennsylvania German and maintain a formal as well as psychological distance between their variety of Pennsylvania German and English. Group 2 exhibits obvious acquisition limitations. They do not control the dative case and adopt the progressive aspect and the far ... zu templates with a vengeance. Group 2 is unable to determine as successfully as Group 1 when templates are applicable, and they produce structures which are frequently unacceptable to native speakers. Given their limited access to native speaker norms, it is amazing how seldom speakers in Group 2 opt for relexified English structures in their efforts to speak Pennsylvania German. Their attempts at producing acceptable Pennsylvania German discourse indicate faulty application of Pennsylvania German rules, not the imposition of English rules. The effort engenders much frustration; the final resort to English structures is often accompanied by paralinguistic and gestural signs of disappointment and helplessness. The sectarians show a substantial amount of syntactic change which results in structures appearing to converge toward English. Frequent code switching within a stable bilingual community provides a hospitable environment for convergence of a minority language toward the language of the dominant society. This is especially true within a sociolinguistic context in which the rules for language behavior prescribe that Pennsylvania German be spoken, not how it be spoken. The influence of English on the sectarian variety of Pennsylvania German is not direct. While the social context may provide the impetus for linguistic change, such as the necessity for frequent code-switching with outsiders and for translating both within and without sectarian communities, the contact with English does not necessarily dictate the specific linguistic accommodation. As mentioned above, case merger has been documented for German dialects without contact with English. Changes in the object placement and infinitive marking rules can be described as changes in Pennsylvania German rules rather than the adoption of English rules. The contact with English may influence the direction of the linguistic changes, but at least in this case convergence toward English seems to be the result of linguistic analogies operating internally in Pennsylvania German. These analogies indirectly produce linguistic convergence. The larger issue is the social one: how the sectarians, as a separated people in an economically complex environment, accommodate mainstream society.

22

Marion Lois Huffines

One can now return to the original questions: how is Pennsylvania German dying and do they love it in their hearts? Among the nonsectarians, Pennsylvania German is dying because younger speakers do not have the opportunity to hear and practice it. They receive little feedback on their flawed linguistic formulations; quite the contrary, older speakers will switch to English for discourse with them. Do they love it in their hearts? Yes, of course. There is no other way to account for their continuing efforts to apply linguistic rules in the face of societal and familial resistance (cf. Dorian 1980). Their linguistic performance is poor, but their ethnic awareness is quite active, and they have developed a vision of what it means to be Pennsylvania German. Among the sectarians, Pennsylvania German is dying in spite of daily usage and continued transmission. Sectarian Pennsylvania German exhibits substantial change in comparison to the native nonsectarian speaker norm. However indirect, the change is contact induced and results in convergent structures which parallel English. Although socially living apart from mainstream society, sectarian interaction with that society is increasing because of economic pressures. Even within the sectarian community, speakers must use English when writing to each other. More importantly, English is replacing varieties of High German used in liturgical settings. The use of English in church services and the observed acquisition of English by preschoolers presage a full shift to English in these communities. Do the sectarians love Pennsylvania German in their hearts? Probably not. Many sectarians say that they will speak Pennsylvania German as long as German is used in the worship services. Many Amish and Mennonite communities in Pennsylvania and other states have already shifted to English in that setting. Ethnicity is clearly an issue secondary to their religious identification. As Kloss (1966: 206) states, "In the United States their point of departure has always been religion rather than nationality or language." The use of Pennsylvania German in sectarian communities is dependent on the use of Amish High German in worship services. When English replaces Amish High German in worship, English will also replace Pennsylvania German within the family and community.

Linguistic Markers of Anabaptist Ethnicity through Four Centuries Werner Enninger Univeristat-Gesamthochschule Essen

0. Introduction This paper deals with the development of linguistic markers of changing concepts of ethnicity in one particular branch of the Anabaptists, i.e. the Swiss Brethren and their successors in parts of Europe and in North America. Broadly speaking, "ethnicity" is here seen as an open set of traits such as shared and distinctive values, common ancestry, a collective consciousness and a self-perception as being different from others, "all of which are implied in 'a sense of peoplehood'" (cf. Rose 1988: 168). For the purposes of this preliminary sketch we resort to Fishman's definition who uses the term "ethnicity" to denote "a bond (self-perceived and/or ascribed by others, with or without objective justification) to a historically continuous authenticity collectivity" (1983: 128). In order to illustrate what ethnicity means and does in collectivities in which ethnicity is a salient focus of perception, the passages following Fishman's definition are helpful. Thus, ethnicity assists inviduals in coping with the existential question of 'Who am I?' and 'What is special about me?' by contextualizing these questions in terms of putative ancestoral origins and characteristics. These questions are therefore illuminated in terms of 'Who are my own kind of people?' and 'What is special about us?' and come to be answered at the level of peopleness being (biological continuity and, therefore, triumph over death), peopleness doing (behavioral fealty even in the course of behavioral change) and peopleness knowing (i.e. ethnicity includes not only native philosophy but historiosophy and cosmology: a Weltanschauung or world view). (1983:128)

24

Werner Enninger

Ethnicity is thus co-terminous with the awareness of membership in, and affiliation with an ethnos, the awareness of which is based on a sense of sharing those traits in and through which a given ethnos constructs its identity and its history. In order to become socially effective the imperceptible awareness of membership in, and affiliation with an ethnos must be projected in perceptible signals. It is true that the sense of ethnicity can be (unintentionally) signified and/or (intentionally) communicated by any perceptible behavior that is subject to cultural constraints such as eating habits, clothing customs, hair and beard styles as well as modes of transportation. All these signaling systems permit the projection of some modicum of ethnicity. The specific design features of human language (Hockett 1960a, 1960b; Hockett and Altman 1968; Thorpe 1972) make linguistic performance the prime medium (also) for the projection of ethnicity. No other signaling system has the design feature of displacement which permits, for example, reference to absent phenomena (such as the past) and the construction of abstract configurations such as cosmologies. No other signaling system has a combinatory mechanism that permits the production of indefinitely long texts which may be extensive reports of ethnicity or folk tales of ethnic origins (cf. Beck 1990), for example. Conversely, even the smallest segment of the speech chain or an intonation contour suffice to evoke - in an all or nothing fashion - the full set of ethnic attributes in the mind of the hearer (Giles et al. 1973; Giles 1973), irrespective of whether the speaker intends to communicate (i.e. to "give") his ethnic identity or whether he (unintentionally) signifies and thus "gives-off" (Goffman 1959: 14-16) his identity, even against his intention. The following quote formulates this view succinctly. In language however we are offered, by the society we enter, and we offer to others, a very overt symbolization of ourselves and our universe, not only in the various grammars and lexicons and prosodies we can create for various domains of that universe, but also through the social marking which each occasion of use carries. Language is not only the focal centre of our acts of identity; it also consists of metaphors, and our focussing of it is around such metaphors or symbols. The notion that words refer to or denote 'things' in 'the real world' is very widely upheld, but quite misplaced; they are used with reference to concepts in the mind of the user; these symbols are the means by which we define ourselves and others. (Le Page and Tabouret-Keller 1985: 248)

Anabaptist Ethnicity

25

In Goffman's terms, whenever people interact they convey not only the content of their message, but also an image of their "selves" (1959) by a combination of nonverbal and above all verbal means including both linguistic items and ways of speaking. On the basis of these assumptions and definitions this paper seeks to reconstruct the development of linguistic markers of changing Anabaptist profiles of ethnicity from the Zürich of 1525 to present-day North America. Since the earlier markers of Anabaptist ethnicity have attracted less attention than the more recent ones, the former will here receive broader treatment. The conclusion will suggest some principles of change in the realm of ethnicity marking which may apply to more than just the one case under examination in this paper.

1. Linguistic Markers of Peoplehood during the Founding Phase According to Giles "one important source of variance existing between different ethnic groups contexts is the baseline linguistic repertoires of the groups concerned" (1979: 253). However, during the founding phase of the Swiss Taufer [Baptists] movement in the Zürich of the fifteen-twenties, the Taufer did not - and could not - have a distinct linguistic repertoire. The "Brüderliche Vereinigung" of 1527 - better known as the "Schleitheim Confession of Faith" - codified the distinct value system of the Taufer movement (nonresistence, nonconformity with the world, adult baptism preceded by the experience of conversion) and marks the transition from the territorial people's church (ecclesia) to the small free church (ecclesiola) and thus the retreat into a particular and separate brotherhood of believers. From a Mennonite perspective Redekop concluded in 1984 that in points three and four of the Schleitheim Confession of 1527 "the peoplehood idea is implied" (Redekop 1984: 133). Redekop supports his assessment by quoting from another Mennonite author: "To us then the command of the Lord is clear when he calls upon us to be separated from evil and thus will be our God and we shall be his sons and daughters" (Wenger 1947: 209). Friedmann (1963) catches this idea of Anabaptist peoplehood in the doctrine of the two kingdoms which coexist side by side in this world. Ever since its founding phase the Taufer movement has existed in a theology-based dual-value situation, i.e. a dinomia-situation.

26

Werner Enninger

The earliest ethnicity profile of the Täufer can be described by slightly modifying the definition of ethnicity which Rose gives for Mennonite ethnicity. The definition that best fits for the early Täufer is one that is exclusively based on value dissensus (that is, the pursuit of major values not shared by others) plus an in-group ideology of shared values and beliefs and a collective consciousness through which purposive actions are possible. This ideology, which is encompassed in Anabaptist theology and differentiates them from other Christians, includes a belief in personal responsibility that participants translate as discipleship; freedom of conscience or conscientious objection to war; adult baptism and the individual responsibility and choice therein implied; and practical piety that manifests itself in mutual aid, stewardship and a simple lifestyle. Thus the identity of the early Anabaptists - their ethnicity - is exclusively defined by ideas and ideology (cf. Rose 1988: 169). In more general terms, we have an early case of what Gumperz and Gumperz-Cook - with reference to Glazer and Moynihan's Beyond the Melting Pot (1975) - call "new ethnicity" that is based on differences which distinguish one newly-emerging group from another. New ethnicity depends upon a process of reactive group formation whereby a group reasserts selectively certain historically established, yet neglected distinctions within the common polity shared with the other groups. Individuals build upon residual elements of shared culture to revive common sentiment upon which to found ethnically based interest groups. Ethnic identity thus becomes a means of eliciting political and social support in the pursuit of goals which are defined within the terms of reference established by the society at large. Because of the complex communicative environment in which individuals must exist, the cohesiveness of the new ethnic groups cannot rest on co-residence in geographically bounded or internally homogeneous communities (Gumperz and GumperzCook 1982: 5-6). All of the values which constitute the ethnicity profile of the emerging Täufer ethnos are values of early Christianity which had been lost or deemphasized in the course of church history. Their foregrounding in the emerging Täufer thought created the dinomia situation. During the founding phase this dinomia-situation was not reflected in distinct linguistic repertoires of the two kingdoms. The leaders of the traditional spiritual forces as well as of both the Zwinglian reformers and its radical wing, i.e. the (proto-)Anabaptists were highly educated theologians and humanists with a thorough knowledge of Latin, Greek and Hebrew (cf. Horsch 1950: 30-33). Conrad Grebel, for example, spent one year (1514) at

Anabaptist Ethnicity

27

the University of Basel, three years at the University of Vienna (1515-1518) and two more years at the University of Paris. "Altogether he spent six years in the best universities of his time, receiving a thorough training in humanistic studies and becoming a master of Latin and Greek" (Bender/Smith 1976: 26). Part of Grebel's correspondence with Dr. von Watt (Vadianus) of St. Gall - his brother-in-law - was in Latin. Balthasar Hubmaier, who played an outstanding role in the so-called Second Disputation of October 1523 in Zürich, was a doctor of theology. Felix Mantz, the first Anabaptist martyr (d. 1527) also had a university education. Michael Sattler, who supposedly drafted the Schleitheim Confession, was a former prior of a monastery in the Black Forest and he knew the Scriptures in the original languages (cf. Bender/ Smith 1976: 26-29). Ciasen (1965: 145) states that up to 1528 there were at least 40 intellectuals in the Täufer communities in Switzerland. Furthermore, various documents show that in the 1540s a discussion of doctrinal principles between Täufer from Saloniki and a "Schweitzer Gemein" in Pausrom/ Moravia was conducted in Latin (cf. Kadelbach 1971: 36-38). From the sources it is not clear whether Latin served as a lingua franca or whether it was selected as the appropriate language for the traditional disputado of theological issues. Thus not only the leaders of the ecclesia, but also of the ecclesiola can be assumed to have used Latin plus "oberlendisch deutsch" on formal occasions, certainly in the written mode. The Zürich Bible, also called the Froschauer Bible, of 1536 (first 1524) attests to the use of a distinct variant of German on formal occasions, as do contemporary letters, tracts, pamphlets and the records of the City of Zürich. Since the middle of the thirteenth century German had begun to replace Latin in legal documents, and the cities of Zürich and Basel had taken the lead of this development (cf. Lötscher 1983: 52). Furthermore, it can be assumed that in the oral discourse of informal situations a vernacular prevailed among the Täufer and their Reformed neighbors. Since the contemporary writings against the Täufer - as for example those of Zwingli and Bullinger which accuse the Täufer of all kinds of differences, for example in dress - do not mention any linguistic distinctness of the group on the level of language systems, one may - in an argumentum e silentio - conclude that the Täufer shared their linguistic repertoire with that of their neighbors as described by Studer (1981). The cultural and linguistic situation of this phase of the Täufer movement can - with some confidence - be hypothesized as suggested in table 1.

28

Werner Enninger Table 1 Cultural Situation Group

Individual

Linguistic Situation

+Dinomia

+Diglossia

values ours theirs

varieties shared between our and their group

+Biculturalism

+Multilingualism of our and their educaed elites: Latin, (Greek, Hebrew) Upper German / Schwytzerdytsch

+Bilingualism of our and their average member: Upper German / Schwytzerdytsch

With regard to values, the Täufer and the mainstream culture were distinct, yet they shared one identical linguistic repertoire. Their shared repertoire reflected the educational stratification of that society as a whole, but it did not distinguish the brotherhood-ethnos from mainstream culture. The brotherhood-ethnos was a representative cross-section of the society as a whole (cf. Kreider 1953), and its linguistic repertoire mirrored that of mainstream culture. In order to distinguish this linguistic repertoire from the one which was to emerge in the course of the migration of the Täufer to other language areas, we will call the above pattern a pattern of stratificational diglossia which reflects the educational strata of one social body rather than the cultural differences of majority and minority co-existing in the same social body. This indistinctive pattern of stratificational diglossia does, however, not mean that the Täufer did not act out their knowledge of peoplehood, or, more precisely, their knowledge of a religious brotherhood in and through linguistic means altogether. A good deal of evidence shows that even in seemingly homogeneous speech communities people use linguistic means in order to locate themselves in multidimensional social space. The early Täufer were no

Anabaptist Ethnicity

29

exception. Their group membership was not mediated through a distinct linguistic repertoire, but rather through a distinct speaking system. In this respect we will seek to provide descriptive evidence for a position taken by Hymes. My contention is that people who enact different cultures do to some extent experience different communicative systems, but not merely the same natural communicative condition with different customs affixed. Cultural values are in part constitutive of linguistic reality. (Hymes 1966: 166) The validity of this contention becomes clearest on the level of performative speech acts. Central doctrinal points of the Täufer belief system such as adult baptism and the rejection of oaths modified traditional performative speech acts and established distinctive variants of these acts. The total rejection of any oaths as formulated in article seven of the Schleitheim Confession ("... Christus, der die volkummenheit des gsatz leert, der verbüt den synen alles schweren ... Sichend zu, darumb ist alles schweren verbotten..." (Jenny 1951: 12)) deleted this performative speech act from the speaking economy of the Täufer. In terms of semiotics, it created a significant zero-sign which contrasted with the performative speech act of mainstream culture. In terms of pragmatics, the significant absence of this speech act had dire perlocutionary consequences. In a society which required oaths of allegiance to the temporal lords every year, the Täufer disqualified themselves from the status of full citizens and from all offices requiring the taking of an oath. Social boundaries began to emerge. While the rejection of oaths excluded the Täufer from the political community, the replacement of infant baptism by adult baptism excluded them from the religious community. The rejection of the doctrine of the original sin permitted the postponement of baptism to later years; the idea that baptism had to be preceded by the deliberate separation from a worldly life and the concomitant declaration for a deliberate renewal of life necessitated the postponement of baptism to adulthood. Again, the Schleitheim Confession of 1527 formulates this doctrine and sets it as the felicity condition of this performative speech act: "Der touff so[ll] geben werden allen denen, so gelehrt sind die buß und endrung des lebens und glouben in der warheit, das ire sünd durch Christum hin weg genommen sigent, und allen denen, so wollen wandlen in der ufferstannung Jesu Christi" (Jenny 1951: 10). In regard to the

30

Werner Enninger

theory of speech acts it should be noted that the performative act of baptizing (into the Name of the Lord) is here separated from the act of performative nomination. Nomination takes place right after birth, while baptism into the faith takes place during adulthood as a fully-fledged rite of passage - which Zwingli had already postulated. Again, the perlocutionary consequences were exclusion, now from the territorial church. In the present context it is worthy of note that membership in the ethnos as understood at that time, i.e. the religious brotherhood of believers, was an acquired status exclusively. It was acquired by the demonstration of a godly life, the explicit rejection of worldly life and by the declared intention of sealing the transition to a new life by deliberately applying for and by accepting baptism into the religious brotherhood. The acquired membership in the religious brotherhood and the boundary between the brotherhood and the rest of the world were acted out linguistically not only in the domain of ritual, but also in everyday life, and that in various ways. Apparently the ways of speaking in non-ritual and non-ceremonial domains of life were soon distinctive enough to make speaking an identity badge of the Täufer, Sebastian Franck's Chronica of 1536 makes the separatist motivation of their distinct speaking rules explicit. "Etliche gerathen dahin/dz sy nichts mit den Heyden gemein woellen haben/ ... Und haben wie die münche regel in essen/trincken/schweigen/reden/kleyderen" (1536/1969: CXCIIY). These (unidentified) normative regulations appear to have had their analogs on the plane of speaking behavior, at least in the beginning of the seventeenth century. Bossert summarizes the 1608 report of a priest from Göppingen as follows: "Dann etliche Jahre hernach sei ein Bote aus Östereich gekommen, den er für keinen Wiedertäufer gehalten habe, denn weder Kleider noch Rede seien wiedertäufisch gewesen" (Bossert 1930/1971: 806). Since this messenger later turned out to be a Täufer (cf. Ciasen 1965: 86), we seem to be dealing with the attempt of a Täufer at passing for somebody else (Goffman), an understandable communicative strategy at times of persecution. In any case, both documents attest in a general way to the existence of distinctive rules and ways of speaking among the Täufer. Other documents reveal the material differences between the mainstream and the Täufer system of speaking. A particularly sensitive realm for choosing between speaking and silence - the two linguistic phenomena mentioned by Sebastian Franck (cf. above) - is the initial face-to-face contact. With few exceptions (cf. e.g. Basso 1972; Hymes 1966) all cultures have rules for

Anabaptist Ethnicity

31

either uttering some routinized formula with greeting function (or some gesture, or both) and thus promote the co-present to a social with, or for remaining silent and thus leave alter in the status of a co-present who is (also) denied access to talk and interaction generally. The so-called "Zürich Kinderzucht" of 1539 (first printed by Eust. Froschauer in Zürich; cf. Weiß 1933) shows that offering a greeting was the expected behavior when coming face-to-face. Also the fact that the deletion of any tokens of acknowledgement of alter as a with is worth topicalizing and attests to the existence of such expectations. With regard to the Täufer, Johannes Keßler notes in his Sabbata written between 1519 and 1539: Derglichen woltend sy niemat mer, der in begegnet oder wohin sy wandlend, grutzen nach gruizend lassen, sunder schlichend hin mit beschlossnem mund; vermeintend, diewil sy undüchtig und onkreftig zu allen dingen und das, so sy wünschend, nit fergen noch geben kundend, wellend sy on grutz und gewünscht lassen; es möcht villicht einer am ainen guten tag wünschen, Gott welt im villicht den nit gonnen, so könne er den im nit geben; sam christenliche liebe nit so gescheftig sy: was sy (nit) mit worten und werken nach mit vermügen lasten mag, begere und wünsche doch von dem geben werden und verlangt, der sy gnug mächtig und gwaltig ist; sam Paulus in sinen briefen im anfang und end kaines grutzends beflissen hab, der och weder gnad nach frid fergen mocht, begert aber von Gott dem vatter und unserm hailand Jesu Christo verlichen werden. (Fast 1973: 617) Keßler explains the deletion of the greeting formulae by the Täufer in two ways. The first reason relates to the fact that the act of greeting very often takes the (surface) form of a wish (with a first person subject). Whereas one might take this as an instance of an indirect speech act - which "says" one thing while it "does" another - the Täufer take the literalism-stance. If the fulfillment of the (surface) wish is beyond the power of the speaker, and potentially against the will of God, the use of the formulae violates the sincerity condition. To speak the truth is, however, a supreme value, and where one cannot speak the truth, one must not speak at all. The insistence on the sincerity of greetings is confirmed, from an insider-perspective, in Peter Riedemann's (d. 1556) Rechenschaft, an early Hutterer defense of their variant of the Täufer faith. Es soll aber des Herrn Gruß nicht allein mit den Lippen und leicht-fertigem Herzen gegeben werden, sondern mit ganz völligem Herzen und im festen Glauben und mit solchem Vertrauen, daß Gott gewiß solche Worte

32

Werner Enninger und guten Wunsch erstatten werde, Mark. 13; Luk. 12; ja also als ob es Gott selbst durch ihn redet. Das Andre aber soll es auch wiederum in solchem Vertrauen und herzlichem Begehren aufnehmen, so wird Gott, der Herr das Gedeihen geben, und solchen Gruß und Wunsch (daß der Herr mit ihnen sei) erstatten, Richt. 6; Matth. 28, und ihnen allezeit beiwohnen, daß ihnen auch solcher Gruß (wie der Gruß Maria der Elisabeth) zu Freuden geraten wird. Wo es aber mit leichtfertigem Herzen oder Unaufmerken geschiehet, da ist es Sünde, Luk. 1; 2. (Riedemann 1988: 118)

This insistence on the sincerity condition is particularly important in social bodies which reject oaths and in which simple statements are the strongest acceptable form of assertions. The use of such formulae violates, second, the principle of humility; its use is a sign of arrogance or pride insofar as it interferes with the will of God (cf. Bauman 1983 with regard to seventeenth century Quakers). Beside these religious motivations of deleting greetings the social effects of such behavior are worthy of note. In cultures whose rules of speaking require that the act of greeting be expressed in some form, the deletion of any perceptible form in that slot is marked behavior. It is a significant zero-signifier given bald-on-record with the pragmatic force of 'impoliteness'. Ciasen, for example, states for 1540: "Der Heilbronner Wiedertäufer Endris Wertz, z. B. dankte plötzlich nicht mehr, wenn ihm jemand auf der Straße einen seligen Morgen oder Abend wünschte" (Ciasen 1965: 84). This Täufer adopted a new pattern insofar as he ceased to return a greeting. Not to return a greeting, i.e. withholding the second-pair part of a strict adjacency pair, is marked behavior. In the case of withholding the expected 'normal' second part of a greeting pair, the behavioral zero threatens the other's face and comes close to an overt insult. The behavioral zero is responding with an insult to what may have been intended as a friendly acknowledgement of a co-present as a social "with". In his "Geschichte der Straßburgischen Wiedertäufer in den Jahren 1527 bis 1543" Röhrich quotes the sixteenth-century Alsatian Knight Eckard zum Trübel as follows: "Die Täufer vermeinen ihre Heiligkeit in Den vor der Welt zu beweisen, daß sie Niemand grüßen, danken und wie stettige Ochsen in aller Unfreundlichkeit gegen andere menschliche Creaturen Gottes leben" (Röhrich 1860: 9). In this respect, Bauman's analysis of the greeting practices of the seventeenth century Quakers (1983: 44-46) also applies to the sixteenth century Täufer. Bauman notes that the Quakers' refusal to use formulae such as "good morning", "good evening", etc. was seen as a serious lack of civil courtesy. Furthermore, he adduces seventeenth century evidence

Anabaptist Ethnicity

33

for their refusal to exchange any greeting formulae at all: "... they will go or ride by them as though they were dumb, or as though they were beasts rather than men, not affording a salutation, or resalutation though themselves saluted" (Higginson 1653: 28; quoted from Bauman 1983: 44). The general analyses of greeting behaviors which Bauman adduces for the Quaker case also apply to the early Täufer: Greetings and salutations are part of the social duty of fully socialized people; to fail to use them is the mark of someone not fully human, either lacking the ability to speak at all or a beast. They are also ceremonial acts (Goffman 1967: 54), conventionalized means of communication by which an individual expresses his own character and conveys his appreciation of the other participants in the situation. To refuse to greet someone, especially someone who has offered a greeting first, is not only to mark oneself as unsocialized, but a lack of social regard for the other person, a serious face-threatening act (Brown and Levinson) in a society and a period in which much emphasis was placed on elaborate etiquette. (Wildeblood and Brinson 1965: 177) (Bauman 1983: 44-45) While those arguments interpet the deletion of greetings within the framework of the normative institution of social etiquette, closer scrutiny of deleted greetings may reveal their religious motivation and their effect for acting out ethnicity conceived of in terms of the religious brotherhood. A large set of historical documents shows that the Täufer did not delete greetings in general (as one may assume on the basis of the above quotes) but that they rather applied greetings selectively. Some alters are not greeted and remain thus in the status of co-presents, while others are greeted and thus promoted to social withs (cf. below when forms of address are discussed). In his Sabbata Keßler states: Es warend etliche von den ersten widergetouften, die nit mit so vil manungen befleckt, sunder hieltend an den artikeln allain, die sy von dem Cunradt Grebel erlernet. Die grutzend och kainen unwiedergetouften; aber nit uß ietz erzeltem grund, sunder wie Joannes in siner epistel leret; dann sy uns für unglobige haiden hieltend; woltend sich mit irem grutzen unserer sünden nicht teilhaftig machen. (Fast 1973: 617) The spiritual basis of this rule of speaking is the Second Letter of John 10: "If any one comes to you and does not bring this doctrine, do not receive him into the house or give him any greeting; for he who greets him shares his

34

Werner Enninger

wicked work." In his account concerning the separation ot the Täufer from the world, which the Schaffhausen Täufer Martin Weninger (called Lincki) wrote (before 1535), we find a paraphrase of this Biblical position: "Wer nit die 1er Christi bringt, den nehmend nit ze huß und grutzend inn nit (2. Jo. 1: 10 f.). Wer inn gruetzt, hatt gmainschafft mit sinen bösen wercken" (Fast 1973: 111). Riedemann, a Hutterer "Vorsteher" (d. 1556), also restricts the exchange of greetings: Grüßen ist an ihm selber Gutes wünschen, darum man denen allen, so des Guten begehren, auch Gutes wünschen soll. So nun in der Kirche ein Glied derselbigen zum andern kommt, soll es ihm auch den guten Wunsch, die holdselige Gabe, Luk. 1; Matth. 10; Luk. 10; Joh. 14, den Frieden des Herrn, den Christus hier gelassen und gegeben hat, anbieten, denn also lehret er auch seine Jünger: Wo ihr in ein Haus kommet, so sprecht: Der Friede sei mit euch. Ist jemand darinnen euch gemäß, so wird sich euer Friede wieder zu euch wenden. Aus diesen Worten lernen wir, daß der so grüßet und der so gegrüßet wird beide des Friedens Kinder sein müssen, soll anders Gott das Gedeihen geben. Denn welcher dem Evangelium nicht gemäß ist oder zu sein von Herzen begehret, da kann der Friede Christi nicht haften. (Riedemann 1988: 117-118) What is interpreted as impoliteness with reference to the norms of social etiquette can be explained as a religion-based, meaningful enactment of the theme of separation of the religious brotherhood from the rest of the world. Peter Riedemann's "Rechenschaft" is an exposition and a defense (of the Hutterer variant) of the Täufer faith. Martin Weninger's "Rechenschaft" is thematically focused on the separation of the two kingdoms, and one of the instruments of doing separation is "not greeting" outsiders. To be regarded as impolite is the social price to be paid for living up to one's faith. The scriptural rule of speaking provides for the socially effective communication of the doctrine of the two kingdoms in a frequent everyday situation. It has, furthermore, the perlocutionary effect of a) preventing the opening of an interaction across the boundary of the brotherhood ethnos, and of b) creating a social gap between 'us' and 'them' generally. Further documents show that this selective greeting pattern was not limited to the particular type of Täufer in Appenzell, to whom Keßler refers in the above quote. In his Chronica of 1536 Sebastian Franck states with regard to the Täufer: "Wer aber jrer Sect nit war / den grueßten sy kaum / boten auch dem kein handt / ..." (Franck 1536/1969: CXCIIY). Laurenz Boshart's chron-

Anabaptist Ethnicity

35

icle of 1529 to 1532 contains the following description of the Täufer: "Sy ... reden mit nimant und gruetzend kein, der nit ir sect ist..." (Fast 1973: 714). This document topicalizes not only the deletion of the greeting formulae, but also its discursive consequences, i.e. the abstention from any verbal interaction. This is more than Article Four of the Schleitheim Confession had set as a norm. While there we find an itemized list of prohibited kinds of interaction, this document maintains the general abstention from verbal interaction with anybody outside the brotherhood. The assumption that greetings were applied selectively to members of the brotherhood is supported by the wealth of documented greeting formulae (plus their second-pair parts) for brotherhood-internal use. In this respect we have to present evidence which appears to be in conflict with a document and its interpretation presented further above. It was argued that the Täufer refused to use greetings which have the surface form of wishes. In the light of the following data, we will have to modify this statement. If one goes by Neff (1937) and Ciasen (1965: 85), the Täufer appear to have developed and canonized distinct and elaborate greeting formulae (in the surface form of wishes and adjacency pairs) for group-internal interaction. Neff notes that the formulae (pairs) served the mutual identification of the Täufer, and were even meant to serve this purpose. While the regional variation is considerable, the format of the wish is a prototypical characteristic (cf. table 2, p. 36). The last pair appears to be the most widely used, whereas the others had rather local or areal currency. The widespread use of formulae topicalizing the wish of God's peace is also attested in a report of the inquest of Nikolaus Guldin in 1529: "Sie haben kain warzeichen, dabi si ainander kennen, sondern si wunschen ainander den friden gots" (Fast 1973: 439). It is interesting to note that article eleven of the Straßburg church discipline (Ordnung) of 1568 sets a norm for greeting behavior. It reads as follows: Ein Bruder oder Schwester soll je eins das andere empfangen mit dem Kuss des Herrn, die aber nicht aufgenommen sind, die soll ein Bruder oder Schwester nicht mit dem Kuss empfangen, sondern sagen, der komme dir zu Huelfe. (Bender 1927: 61) Both the holy kiss and the verbal formula are alloforms of the act of greeting, and both fulfill the function of establishing some social relationship. However, the complementary distribution of the alloforms over membership

36

Werner Enninger

Table 2 Greeting formulae in group-internal use First Pair Part

Second Pair Part

Area

"Gott grüße dich in dem Herrn" (after doffing the hat)

"Ich danke dir in dem Herrn"

Franconia

"Lieber christlicher Bruder" (with a handshake and a hug)

?

Northern Thuringia

"Der Friede des Herrn sei mit dir!"

?

ditto

"Der Friede des Herrn sei mit dir!"

"Amen"/"Das sei wahr" (then: handshake plus kiss of peace)

Friesland

"Der Friede Gottes sei mit dir!"

"Das muß Amen sein"

Friesland

"Die Gnade des Herrn sei mit uns"

?

Neckar valley

"Got gries dich, bruder im hern"

"Danck dir got, mein bruder im herrn"

Augsburg

"Got sei mit mir"

"Der fried sei mit uns"

Augsburg

"Der fried gots sei mit dir"

"Danck dir got"

Göppingen

"Grüß dich gott, christlicher bruder"

"Danck dir got du christlicher bruder"

Linz

"Der ewige vater beschirme dich mit seinem ewigen frieden und sei mit dir"

?

Thüringen

"Lieber christlicher bruder"

Eisleben

"Amen"/"Mit deinem geiste"

Southern and Middle Germany

"Lieber christlicher bruder" (plus handshake) "Der friede des herrn sei mit Dir"

(cf. Neff 1937 and Ciasen 1965: 85)

Anabaptist Ethnicity

37

categories makes a distinction in the relationship which obtains between ego and alter, i.e. religious peers vs. non-peers. Furthermore, it can be claimed that the verbal formula "Der Herr komme dir zu Huelfe" also has meaning beyond the mere social meaning: it is a wish which is directed at the seeker who needs God's help in finding her/his way into the true church. At this point two conflicting kinds of documentary evidence of the greeting patterns of the early Täufer require attention. First, the wealth of the greeting formulae used for group-internal purposes conflicts with the sweeping statement (of some documents) that the Täufer did not at all participate in the then extant greeting patterns. Second, the fact that the greeting formulae (as well as the leave-taking formulae not topicalized here) have the prototypical surface-forms of wishes appears to conflict with Keßler's assessment that the Täufer refrained from greetings for this very reason. Let us take the second issue first. Closer scrutiny reveals that the Täufer greeting formulae taking the surface form of a wish do not have the propositional format of 'I wish you x', but 'May God grant you x'. The decanonization of the traditional wish format of the greeting and the apparent re-canonization of a new wish format reflected and enacted in everyday life the theological distinctness between the brotherhood and "them" out there. Second, the dual pattern of refraining from greeting when meeting ousiders and of using a wealth of greeting formulae in group-internal interaction is associated with the dual function of greetings. Here we follow Bauman's argumentation in regard to seventeenth-century Quakers. 1) Greetings are instances of phatic communion insofar as they are almost devoid of referential meaning and have only the pragmatic force of raising spatial co-presents to the status of social withs. 2) They have the function of opening access to talk. "It follows naturally that if one has not real need to talk to another person, greetings are to that extent rendered unnecessary..." (Bauman 1983: 45). We suggest that the conflicting statements concerning the use and the non-use of greetings can be reconciled when one considers them as motivated in their (complementary) distribution over meeting Täufer and outsiders, respectively. With regard to outsiders, both the phatic and the talk-opening function of greetings are dysfunctional: not only are they redundant, but they are even in conflict with the value system of the Täufer. On the other hand, greeting other Täufer is not only functional within the conversational machinery, but it is even mandatory in view of the value system. Greetings are confirmations of the brotherhood of true believers. The distribution of the use versus the deletion of greetings over

38

Werner Enninger

intra- and intergroup encounters is another mode of acting out the themes of 'strengthening the true church, i.e. the body of Christ' on the one hand, and of 'separation from the world' on the other. This boundary marking of the particular religious ethnos by a selective application of (any) greeting is supported by the prototypical format of invoking god's grace in group-internal interaction and the particular formulation used when applied to outsiders "Der Herr komme dir zu Huelfe". Before we leave the linguistic marking of ethnicity through greeting patterns we will dwell for a moment on one document which provides explicit insights into the early concept of Täufer ethnicity. The specific ethnic profile of the Täufer in the sixteenth century becomes clear from the description of the opening of an encounter of 1578: "Meinen (halb)bruder Sebastian hab ich auch bei dem brauhaus angetroffen, der mir aber, wie auch der Christmann (whom the chronicler had met before; W.E.) die hand nit geben wollen. Seine erste worte waren: ich sei ein falscher prophet; darumb dörf er mir die hand nicht bieten." (Bossert 1930/1971: 1107). While the refusal to shake hands is stated and explained, the absence of a verbal greeting must be inferred from the statement "seine erste worte waren...". Thus without a verbal greeting or its non-verbal analog of a handshake and without a mitigating preface Stephan Gerlach's half-brother Sebastian opens the encounter with the hearercostly statement that Stephan (the addressee) is a false prophet and that any signifiers of a relationship must be avoided, because they are no brethren in faith. Taken together with other evidence presented above, this document grants insight into both conversational strategies and the ethnicity concept of the early Täufer, First, in intergroup communication bald-on-record contributions appear to supersede off-record ones. The former threaten both the negative face (the need to be left alone) and the positive face (the need to be approved by others) of the outside interactant. Here conversational directness appears to be preferred to indirectness. Broadly speaking, impoliteness is used as a boundary marker of the ethnos. The document shows, second, that religious ties supersede blood relationship. In the present context this underlines that initially peoplehood was exclusively defined on the basis of the deliberately acquired status of membership in the religious brotherhood. Complementarity, ascribed statuses of kinship were explicitly excluded from the defining criteria of ethnicity. The basic questions of "Who are my kind of people?" and "What is special about us?" were answered in terms of religious brotherhood with membership exclusively based on adult baptism and confession of faith.

Anabaptist Ethnicity

39

Before we follow up the development of different profiles of ethnicity, we must dwell on other markers of the earliest profile of ethnicity. Of all, we will focus on forms of address, because they reflect almost exclusively the relationship obtaining between speaker and hearer. Pragmatics and conversational analysis have taught us that the interactional partners do not function as individual speakers and hearers, but as incumbents of social positions and roles of their respective social worlds. The relationships to in-group and to out-group interactants, respectively, are therefore likely to be marked by alternative choices of address-forms, which choices thus become linguistic boundary markers of an ethnos. The background against which the address system of sixteenth century Täufer may be put into profile is provided by Metcalf (1938: 11-63). At the beginning of the sixteenth century abstractive honorifics or designatory markers such as Majestät, Gnade, Liebde as well as Ihr were in use among the highest ranks of society; for polite address, Ihr was given to, and received by the nobility and the upper middle class (including the artisans) except where honorifics were in use. Du was exchanged by persons of the lowest ranks, but also among intimate friends, even among the highest classes. Furthermore, Du remained the sole means of addressing supernatural beings; but Du was also used to express anger or contempt (cf. Brown and Gilman 1960: 274-276 for the "thou of contempt"). During the century the frequent use of Ihr caused it to lose its value for polite address so that it began to give way to a nominal construction with Herr and a third person plural form Sie. The Zürich book of etiquette of 1539 quoted above (p. 30), Trümpy (1963: 158159) as well as many contemporary documents (cf. Fast 1973) support the validity of Metcalf 's findings also for Switzerland. In the following we will, first, compare the actual choices among the nominal and pronominal options as reflected in letters in order to show how each letter writer used the address system in order to locate himself in social space. This comparison will, at the same time, show the emergence of addressorial ethnicity markers among the Täufer. The first example is a letter of Stefan Zeller, Landvogt at Adelfingen to the City of Zürich (dated 28 February 1535). It illustrates the addressorial extremes to which an inferior might resort when writing to his superior. Both, addressor and addressees belong to mainstream authorities. Fromen, vesten, fursichtigen, ersamen, wyßen herrn, sunders günstig, genedig, lieb herren, uver wyßheit sye min underthänig Ghorsam, willig

40

Werner Enninger dienst und, was ich eeren und gut vermag, nach schuldigen pflichten allezit bereit zu vor. Gnedigen lieben heren ... Ever wyßheit gib ich hiermit in underthänigkeit zu erkennen, das... (Fast 1973: 77)

The above complimentary opening has its formal analog in the equally complimentary closing. Deictically, the use of Du would be inappropriate, because Zeiler addresses more than one person. The use of Ihr would be appropriate deictically, but it is avoided throughout the letter, probably because Ihr, when used to address more than one person, is the form which is used to address both, inferiors and close equals as well as distant superiors. Zeiler avoids this ambiguity by resorting to what Metcalf calls "abstractions" incorporated into the sentence structure: Euer wyßheit gib ich hiermit... Social propriety dictates, and the inferior's motivation to fulfill the superior's expectation of tokens of positive face-work suggests both the selection of the most deferential options and the accumulation of honorific attributes and titles in the opening of letters to superiors in the (mainstream) power network. At first sight, a letter written by the Täufer leader Felix Mantz to the Council of Zürich written in December 1524 follows the same pattern. We will select only those passages from a modern German version which are relevant for the discussion of locating oneself in social space through choices of address forms. Weise, fürsorgende, gnädige, liebe Herren und Brüder! Eure Weisheit wissen wohl, daß viele ungewöhnliche Gepräche stattgefunden haben ... Da man mich nun (wenn auch ohne Grund) so einschätzt, halte ich es für nötig, Euch, meinen gnädigen, lieben Herrn, Rechenschaft ... zu geben. ... Zum folgenden will ich Eure Weisheit ... gebeten haben: Bitte hört ohne Ansehen der Person...; ermeßt wohl, was da angeführt ist, und laßt Euch die kurze Zeit nicht leid sein! ... Deshalb will ich Eure Weisheit freundlich und auf das allerdringlichste gebeten haben: Bitte nehmt mein Schreiben im besten Sinne auf.... Ich möchte Eure Weisheit auch ermahnt haben, daß Ihr Euch an den Streit über die Götzen erinnert ... Ich möchte Eure Weisheit auch daran erinnert haben, daß ... Deshalb möchte ich aufs fleißigste gebeten haben: bitte besudelt Eure Hände nicht mit unschuldigem Blut und meint, Ihr tut Gott einen Dienst, wenn Ihr einige tötet oder verjagt... Ich möchte Eure Weisheit auch um folgendes gebeten haben. (Fast 1962: 28-35) In the opening address Mantz uses unintegrated designatory markers, but unlike Zeller in the preceding letter, he does not accumulate them to the

Anabaptist Ethnicity

41

extreme; the difference is one of degree, not of kind. In a letter to the authorities, the term Brüder seems unusual. Fast (1962: 28) takes the use of this term as an indicator of the fact that at that time Mantz acknowledged the later persecutors of the Täufer still as brethren, although Zwingli and the Täufer had been drifting apart since the October disputation of 1523. Thus the initial address reflects Mantz's dual relationship to his addressees: citizen to magistrates in the power network vs. brother to brethren in the solidarity network. In the body of the letter he uses both, abstractions integrated into the sentence structure ("Eure Weisheit wissen wohl ... Ich möchte Eure Weisheit auch um folgendes gebeten haben") and Ihr, also in oblique cases and implicitly, as in the imperative nehmt While the former expresses deference owed to superiors, the latter - when used to address more than one person - is neutral with regard to power and solidarity. In comparison with the Zeiler letter (which does not contain any token of Ihr), the alternation between deferential and neutral forms of address indicates that Mantz is thoroughly aware of his dual, and probably ambiguous relationship with the addressees. His address forms locate him in just this double-faced position between the two (emerging) social worlds. The famous letter from Konrad Grebel (and associates) to Thomas Müntzer of September 5, 1524 indicates the general direction in which the Täufer forms of address were to develop in the intra-group network. Dem wahrhaftigen und getrüwen deß evanngelli Tome Müntzer zu Altstett am Hartz, unserem getrüwen und lieben mitbruder in Christo etc. Frid, gnad und barmhertzikeit von Gott unßerem vatter und Jesu Christo unserem herren sy mit unß allen, Amen. Lieber bruder Toman, laß dich umb Gotz willen nit wunderen, daß wir dich ansprechend on Titel und wie ein bruder ursachend hinfür mit unß zehandlen durch gschrift, und daß wir ungeforderet und die unbekant habend gedörfen ein gmeinkünftig gsprech ufrichten. (Franz 1968: 437) In a period, in which the address systems was characterized by "a gradual but constant tendency to veer off farther and farther from the familiar form Du: first to the second person plural (in addressing one person), next to the third person singular, finally to the third person plural" (Silverberg 1940: 510), Grebel moves in the opposite direction. He addresses Müntzer without title and with Du, although they had never met before. The comment (on the meta-level) shows that Grebel and associates are aware of the expectations which they violate: laß dich umb gotz willen nit wunderen, daß wir dich an-

42

Werner Enninger

sprechend on Titel und wie ein bruder. Throughout the letter Grebel and associates use consistently tokens of TU (du, dich, din; implicit TU in finite verbforms such as wellist, berichte, lere) in order to address Müntzer. Ihr forms occur only where Grebel et al. address more than one person, i.e. Müntzer and Karlstadt. Throughout the letter there is not one token of deferential Ihr or of any "abstractions". Thus this letter is not only the earliest summarizing statement of doctrinal points of the Swiss Täufer (who reject some of Müntzer's views), but at the same time the letter documents the emergent marking of the emergent brotherhood ethnos through an invariant Du for singular address. How radical this innovative use of the invariant Du for singular address inside the religious brotherhood was becomes clearer by its contrast to the maintenance of the traditional address forms between brothers-in-law. The same Conrad Grebel who on 4 September 1524 addresses Müntzer, his (half-) brother in faith with Du, on 23 November and on 15 December of the same year 1524 addresses his brother-in-law (Vadian) with honorific Ihr and title: Min schwagerlichen grutz zwor. lieber herr doctor, wüssend mir üwer brief.... So ir aber nüt vernommen hettind ... will ich üch schriben.... (Muralt/Schmid 1952: 28-29) Lieber herr doctor und schwager, daß ir mich gebetten hand, kond ich nit leisten... (Muralt/Schmid 1952: 29) Both letters use ir/iiwer, i.e. tokens of VOUS consistently. Thus while religious ties came to be signaled by solidary TU (and significantly absent honorifics), certain family ties continued to be reflected in polite VOUS plus honorifics. Not only in addressed written communication (personal letters, corporate epistles etc.), but also in group-internal face-to-face encounters tokens of pronominal TU appear to have been the rule. The list of greeting-formulae (cf. above) attests to that. Ciasen (1965: 87; Jacobs 1899: 484-485, 505) confirms the view that TU forms were the unmarked option of intergroup pronominal address. Outsiders perceived this usage as so marked that it roused their suspicion: Die Stuttgarter Regierung wurde schon argwöhnisch, als man ihr meldete, daß die Täufer einander duzten und Schwester und Brüder nannten. (Ciasen 1965: 111) This quote shows that the nominal equivalent of pronominal Du in the Täufer intra-group network was bruder/schwester. An abundance of docu-

Anabaptist Ethnicity

43

mentary evidence from, for example, Sebastian Franck's Chronica (die Täufer hießen einander Brüder), Wappler (1913: 396, 399, cf. below), Nicoladoni (1893: 224), Jacobs (1899: 484-485, 504-505) gives credence to d a sen's generalization: "Die Täufer selbst redeten sich nach ihrer Bekehrung nur noch als Brüder und Schwestern an" (1965: 84). This Standard form of address appears to be optionally followed by the qualifier "im Herrn", as in the opening and concluding addresses which frame the "Schleitheim Confession": brüder und Schwestern, Lieben brüder und Schwestern, Lieben brüder und Schwestern im Herrn!, Lieben brüder (cf. Jenny 1951). The qualifier explicitly selects the subsense 'related in faith' of the otherwise polysemous bruder/schwester, i.e. 'sibling' versus 'related in faith'. Since all of the above addresses are co-referential, the qualifier-free form must be refential-semantically compatible with the qualified address. As in the case of the selective greeting behavior, we are again made aware of the fact that during the founding phase the ethnicity of the Täufer was based on shared beliefs alone, and that to the explicit exclusion of the criterion of relationship in blood. The perceptual foregrounding of faith before family is evidenced in the records on the arrest of Täufer Orlamünde in Thuringia on 21 November 1535: "Als sie nun gefenglichen vorfast, hat der muller Hans Poißker zu seinem weibe geredt und sie Schwester geheißen.... hinwidder inen das weib bruder geheissen" (Wappler 1913: 399). The same tendency is repeated a few pages later. "(Martha...) saget, Gott hat sie underweist, das sie sich zu den brudern halden sollte, welche auch den vatern einen brudern und der vater sie widderumb eine Schwester genannte" (Wappler 1913: 399). If the definition of ethnicity in terms of faith - and not in terms of family ties - needed a confirmation which also reflects the bald-on-record conversational strategies the document contains one: Item sie wollen von iren freuntschaften und blutsfreunden nichts hoeren, sagen, sie haben auf erden keine freundschaft, begeren nicht irer hulf noch rat, weniger irer vorbitt, und sunderlich ist der wechter einer des mullers Hansen Poißkers gefatter, der inen im gefenknus gefatter geheißen, der er geantwort, er wußte von seiner gefatterschaft gar nichts, er were des teufeis gefatter. (Wappler 1913: 399) Further distinct rules of speaking could be discussed as linguistic ethnicity markers during the ethnogenesis of the Täufer cosmos. Limitations of space, however, prevent discussing the function of rules for the topicalization

44

Werner Enninger

of speech content in general or constraints on identifiable speech acts (such as slander, libel, derision (cf. Jacobs 1899: 485), litigation, talking-back, etc.) as linguistic markers of emergent "new ethnicity". In view of the growing importance of speaking rules in the Täufer cosmos, the pertinent documents deserve further attention in separate studies as for example Dietrich Philips' Enchiridion or Hand Book (repr. 1978; first 1569), Johann Arndt's Paradies Gärtlein (first 1612) or the anonymos Lust=Gärtlein Frommer Seelen (repr. 1970; first c. 1770), which on pages 219-241 anticipates the core content of Grice's maxims of quantity, quality, relation and manner (cf. Enninger et. al. 1989: 162-165). Instead, section 1 has exclusively focused on rules of greeting and address in their function of ethnicity marking, and has not even covered this realm exhaustively. The 'thou' of contempt as for example documented in Muralt/Schmid (1952: 39), Bossert (1930/1971: 1050), Ciasen (1965: 104) deserves attention also in regard to its role in (aggressive) boundary marking of the emergent ethnos. As a more recent study of the variable use of one address system by two co-territorial Mennonite groups (Howell and Klasen 1971) shows, ways of addressing are worthy objects of research beyond the ethnogenesis of the Täufer cosmos. This and the fact that rules and ways of speaking have been continually topicalized in the doctrinal writings of the Täufer leads us to assume that distinct conversational strategies have continued to play a role in the linguistic marking of Täufer ethnicity beyond the founding phases. Instead of following up pertinent leads we will in the following focus on other linguistic markers of ethnicity.

2. Linguistic Markers of Peoplehood during the Post-Foundation Phases Section 1 dealt (selectively) with the emergent linguistic marking of peoplehood, of "new ethnicity" during the ethnogenesis of the Täufer. The following section will deal with concepts of peoplehood and their linguistic marking in the post-foundation phases. The end point of what is meant by "post-foundation phases" is clearly the present - which includes also the migration to other language areas. The consequences of this migration for the linguistic marking of ethnicity are here disregarded and will be treated in section 3. Instead, we will here focus on an additional component in the concept of ethnicity and its linguistic marking, both of which are independent of migration to other lan-

Anabaptist Ethnicity

45

guage areas. Admittedly, the weak point of the following section is the exact point in time at which the new concept of ethnicity began to take shape. Only for the sake of convenience the year 1550 is suggested, because by that time harassment and persecution had forced the originally aggressively proselytizing urban movement to withdraw to a quietist rural movement which turned inward on itself. As was repeatedly stated in section 1, the dissensus was initially based on religious values exclusively. From the studies of religious-based schisms it is well-known that sometimes the differentiation, though at first defined in religious terms, broadens to include a whole way of life (cf. Petersen 1980: 241). This is particularly likely to happen, if the central doctrinal positions are to be lived out in life - as is the case in the existential Christianity of the Täufer (cf. Friedmann 1973: 31). In such a context membership can no longer be based exclusively on the acceptance of a closed list of doctrinal points in baptism, but also (or rather) on an ethnic sentiment which can only be absorbed in an extended socialization process within both the family and the closely-knit social network in which like-minded people interact frequently or even regularly. The group's separate religious past is reinforced in the present by physical-geographic proximity as well as occupational, political, friendship and familial ties. Such a group is likely to turn inward, and encourage or even enjoin endogamy as a boundary-protecting instrument. These considerations account for the inclusion of common ancestry and endogamy into most definitions of ethnicity: "... and common ancestry or place of origin ... Endogamy is usual" (De Vos 1975: 9). With the exception of a shared and distinctive linguistic past, we are here faced with the ensemble of components which are, as a rule, considered to constitute ethnicity, or - more precisely - "old ethnicity" (cf. Gumperz and Gumperz-Cook 1982: 5). In the case of the Täufer this turning inward did not only develop from internal motivations but it was at the same time enforced by outside powers. They were first forbidden to proselytize, then complied with the pertinent legislation, and ultimately some (conservative) branches canonized non-proselytizing as one of their norms. In due course, the brotherhood became co-extensive with a kinship-network and a new concept of ethnicity emerged. In analogy to Rose (1988: 168 who, in turn, refers to Francis 1955: 25) one can say: Within a few generations the Täufer passed from an exclusively religious brotherhood based on adult baptism and confession of faith to a society identified additionally by characteristics of descent and shared cultural traits. This means to say that the primary

46

Werner Enninger

locus of identification remains the system of Täufer beliefs, but that the medium of their transmission, i.e. parentage/ancestors develops into a second locus of identification. In such a cultural context exogamy is likely to be considered an indicator of assimilation into the dominant culture and "marriage out" might endanger the unbroken transmission of the Täufer belief system to the following generations. In fact, endogamy has remained the rule, - as one can infer from two linguistic markers of ethnicity alone. To the extent in which the proselytizing outreach receded the Täufer ecclesiola became almost coextensive with an endogamous race. Family names attest to that. Hard-core linguists may hesitate to include onomastics into their discipline, because the semantics of the various kinds of names (familiy names, first names, nicknames; Enninger 1985a) gives them a particular status among words which makes them peripheral to the language system. On the other hand, it is equally clear that names do not only identify individuals, but that they - given their inherited status or their anchoring in traditional first-naming patterns - are markers of ethnicity that cannot very well be assigned to non-verbal markers of ethnicity. Above all family names reflect a group's distinctiveness and they should therefore be included into the (verbal) vehicles of ethnicity (cf. Epstein 1978: X). The fact that among the 85,000 Amish in 1980 there were only 126 different family names attests to the prevalence of endogamy. Only "forty-three names (33 per cent) are American in origin and represent converts to the Amish faith, but they constitute a small proportion of the population. At the present time, 18 of the 43 names represent only single households" (Hostetler 1980: 241). In analogous fashion, thorough demographic analysis of one specific Old Order Amish settlement revealed that 1,304 persons living in or born to the 170 families with Old Order Amish household heads had only 20 different family names (Enninger and Wandt 1980). In the present context this means that the second criterion of ethnicity, i.e. kinship and endogamy, is saliently reflected in family names. While the clustering of family names in a narrow segment of the onomastic potential reflects the biological effects of conceptualizing the religious brotherhood also in terms of an historically continuous and endogamous kinship network (Pennsylvania German: die Freindschaft), the preoccupation of later successors of the Swiss Täufer with genealogies indicates the deliberate attempt to answer the ethnicity questions of 'Who are my kind of people? What is special about us?' (also) at the level of biological continuity. Again,

Anabaptist Ethnicity

47

while hard-core linguists would clearly exclude genealogies from their field, our definition of "linguistic markers" of ethnicity is broad enough to include this text-linguistic genre. Its inclusion is even warranted because it helps to distinguish the construction of historical continuity within the concepts of new versus old ethnicity. New ethnicity, we suggest, constructed historical continuity exclusively in religious terms, i.e. in the Nachfolge Christi (the discipleship of christ): Braght's Martyrs' Mirror (1660) reflects this view: Gleich wie man zweierlei Volk, zweierlei Versammlungen und Kirchen findet, die eine aus Gott und vom Himmel, die andere aus dem Satan und von der Erde, so findet man auch zweifache Nachfolge und Fortpflanzung derselben. (Braght 1973: 19) In the transition from new to old ethnicity biological continuity was foregrounded - and reflected in a small cluster of family names and the salience of family histories among all text types produced by the Täufer descendents. The case may serve to support and to modify Hansen's "deliberately overdrawn" hypothesis of "third-generation nationalism" and its markers. The hypothesis suggests a three-generation typology of ethnic marking, in which the third generation wants to remember their roots, which the second generation - caught in the strange dualism of being a native of foreign extraction wanted to disremember. Hansen suggests that after approximately 60 years that is, two generations - the immigrant population makes organized efforts to revive specific elements of its root culture (cf. Petersen 1980: 239). Among the successors of the Swiss Täufer revival efforts did not focus on Swissness or Germanness, but rather on the history of the brotherhood and its transmitters, i.e. the largely endogamous chain of procreation. To what extent the rise of genealogies fits Hansen's 60 years and three-generation cycle remains to be seen. It seems as if the genealogical interest among the Täufer-successors was due to a postponed search for roots - much like the ethnic revival in the US in the seventies and in present-day Europe. As Springer (1972) shows, genealogies and family histories became a major concern in the wider Mennonite world in America during the nineteenth century. The first books appeared in 1858. "From 1870 to the present there has been a continuous growth in American Mennonite genealogical publications with an early peak in the 1890's and a slight recession during the 1930s" (Springer 1972: 457). Springer's analysis of the genre brings to light that the distribution of such books over Mennonites in the wider sense on the one hand, and the Old Order

48

Werner Enninger

Amish on the other varies with time. The increase in the number of American genealogies published between 1952 and 1972 can be attributed largely to the Old Order Amish (and the Russian Mennonites). In August 1982 no fewer than 460 Old Order Amish genealogies had appeared (Luthy 1982; personal communication). This reflects the present-day concern of the more traditional segments with 'old ethnicity' at a time when more liberal Anabaptist groups have reverted to the original 'new ethnicity' (cf. section 3.2). Together with the absence of other widely produced historical studies by the population at large, this can be taken as an indicator of the fact that in the historical thought of the Täufer successors the procreational chain has been foregrounded against the implicitly understood and taken-for-granted background of the meanwhile diversified religious brotherhood. Today, there is hardly an issue of the culture-internal papers that does not advertise a (new) family history and/or that does not solicit and publish genealogical information. Also, in oral discourse, genealogy has been "the Mennonite game", as one Mennonite aptly put it. (For further details on doing ethnicity "through specific genres", cf. Enninger 1986b, 1988.) It should be noted, that constructing historical continuity at the level of family history makes sense in a group with extreme horizontal mobility: one's own procreational chain is the most stable historical dimension. In concluding this section it should be noted that the additional component of the Täufer concept of ethnicity, i.e. the unbroken procreational chain leaves immediate linguistic traces only in onomastics and in the genre of genealogies - and thus those markers should not be ommited, even if they are marginal linguistic phenomena. Since an unbroken procreational chain favors language maintenance, maintained languages may be considered as secondary markers of endogamy. They may become particularly salient markers of ethnicity in general after migration to other language areas where they may become a component of a linguistic repertoire not shared with mainstream society. These questions will be addressed in section 3.

3. Linguistic Markers of Peoplehood during the Migration to Other Language Areas The social distance between in-group and out-group - at first only actively sought - was as of 1527 increased by severe persecution by state and church.

Anabaptist Ethnicity

49

This persecution, later abating to harassment, lasted for almost 200 years. The early migrations were mostly the result of this persecution. Since then most branches of the Täufer movement - even the agrarian branches - have been horizontally mobile groups. For centuries many of their branches have been in continual transit from one region to another, from one state to the next, and from one language area to another. These migrations created the language-contact situations in which the inherited varieties could and, in fact, did become the most salient linguistic ethnicity markers. It is in these language-contact situations that the basic tenet of Giles unfolds its full heuristic value, namely that "one of the important sources of variance existing between different ethnic group contexts is the baseline linguistic repertoires of the groups concerned" (1979: 253). Giles proposed three types of adjustment strategies leading to different situations: language choice situation, accommodation situation, and assimilation situation (1979: 255). He furthermore suggested four typical ethnolinguistic profiles emerging in language-contact situations: a) monolingualism in the ingroup language, b) monolingualism in the outgroup language, c) bilingualism in the in- and outgroup languages, and d) bilingualism in the ingroup language and a lingua franca. Of these concepts, four are particularly helpful in the description of linguistic ethnicity marking across the Täufer-spectrum through history: linguistic accommodation resulting in bilingualism in the inand outgroup languages and linguistic assimilation leading (ultimately) to monolingualism in the outgroup language. Before we discuss these in turn we hasten to add that the above terms stand for prototypes with fuzzy edges rather than for strict categories with clear boundaries. 3.1 Linguistic Accommodation: Diglossic Bilingualism in the Ingroup and the Outgroup Languages In language-contact situations the assimilative and the isolative forces obtaining within a given ethnos (cf. Kloss 1966) can produce quite different ethnolinguistic profiles and quite different timings of language transitions to the ultimate terminus ad quem predicted by the majority of ethnolinguists, i.e. monolingualism in the outgroup language. Thus it comes as no surprise that the particularly strong isolative motivations of some Täufer branches should even today, i.e. sometimes after several hundreds of years of migration

50

Werner Enninger

through other language areas, exhibit a distinct bilingualism and sometimes even a trilingualism pattern supported by diglossia. Examples are: 1) Täufercongregations in the francophone Swiss Jura that have maintained their Bernese dialect. 2) The German-speaking Mennonite communities in Paraguay, Uruguay, Belize and Mexico in Spanish-speaking areas, and in Portuguese-speaking Brazil. 3) The Hutterer Bruderhöfe in the US and Canada. 4) The Old Order Amish in the US and Canada. 5) The Old Order Mennonites in the US. In order to provide a comparison with the largely hypothetical ethnolinguistic profile of the sixteenth-century Zürich presented in table 1 we will here give the linguistic profile of the Old Order Amish (OOA) settlement in Kent County, Delaware, which is representative of most OOA settlements, except those stemming from nineteenth-century immigration from the Switzerland, from the Pays de Montbéliard and Alsace. In those areas (high-) alemannic dialects close to the Bernese dialect takes the place of Pennsylvania German in table 3. Note that AHG stands for Amish High German and AE stands for American English. Note also that this example of linguistic accommodation contains a reflection of linguistic assimilation: "In a single generation in the Palatinate the Zürich and Bernese Anabaptists had completely adopted the speech of the new land" (Gratz and Geiser 1973: 671). In the context of the above quote this means that the immigrants gave up their Zürich and Bernese dialects for the Palatinate dialect, which later was to serve as the basis of present-day Pennsylvania German. Table 3 Cultural Situation Group

+Dinomia ours theirs values

Individual

Linguistic Situation

+Biculturalism

+Diglossia varieties values identity prestige

our

AHG PG high high

+Multilingualism

high low

their AE low high

Anabaptist Ethnicity

51

Table 3 reveals that the transit from Switzerland, Alsace and various parts of Germany to America was accompanied by a) the maintenance of Amish High German and the substitution of Pennsylvania German for Schwytzerdytsch and by b) the complementary association of the components of the linguistic repertoire with "our" and "their" values: AHG is associated with high prestige and high identity, PG is associated with low prestige and high identity values. The latter means that - in contrast to the original Zürich situation, the terminus a quo - the dinomia-situation is paralleled by or reflected in, a linguistic repertoire which, insofar as it is not shared with mainstream society, is distinctive of the minority. Both situations are depictable in terms of diglossia and bilingualism; but whereas in the original situation the plus and minus distinctions operated on the vertical dimensions or "our" and "their" elite vs. "our" and "their" average person, they now operate on the horizontal dimension of "us" versus "them": "we" share "their" variety AE, but "they" do not share "our" varieties AHG and PG. The table thus suggests that in the transit from Europe to the language-contact situation in America the linguistic repertoire of the Täufer, which was indistinguishable from that of mainstream culture, "transited" to the non-shared and distinct ethnolinguistic profile of the 0 0 A today (for further details, cf. Enninger 1986c). With regard to the signaling of identities in speech this diglossic pattern implies: 1. In the intra-group interaction network 'old ethnicity' is signaled by the use of one of the inherited varieties, either the low-prestige plus high-identity variety of PG, or the high-prestige plus high-identity variety of AHG in their respective domains. Two comments are appropriate here: On the part of the speaker, the use of these varieties is not equivalent with "free choice", insofar as the speaker is subject to the norms of appropriateness. Therefore, it would not be correct to say that the speaker intentionally communicates 'old ethnicity'. We would rather say that - in following the conventionalized rules of appropriate variety-use - s/he is socially obliged to signify the societally important 'old ethnicity'. The other point to be mentioned is that the heavy borrowing from AE into PG, as well as the interference from PG in AHG, and the borrowing from AE into AHG do not endanger the signification of 'old ethnicity' as long as the varieties are perceived as distinct and identified as "their" versus "our" varieties.

52

Werner Enninger

2. In the intergroup network the use of AE is again not a question of deliberate choice, but one of communicative necessity. After all, AE is, as a rule, the only variety shared with group-external interactants. The fact that AE is used signals the accommodative attitudes to mainstream society and the cooperativeness with mainstram interactants in secondary roles of power. The way in which it is used, signifies 'old ethnicity' through a minimal amount of learner interferences from L1 (PG) on L2 (AE) (cf. Eninnger et al. 1985). The order in which the components of the L2 (AE) are affected by interferences from L1 (PG) is - as can be expected - the inverse of the order in which the components of L1 (PG) are affected by borrowing from L2 (AE) (cf. Thomason 1986). In speech, the AE lexicon of the OOA is not marked by learner interference from PG, i.e. their speech contains no lexemes which are not also contained in the speech of their socially comparable territorial monolinguals. Switches to PG words or phrases are no evidence to the contrary. However, a questionnaire based on Kurath (1949) revealed a preference for originally German lexemes which have become established areally as AE words by substratum interference/borrowing, if such items co-exist with indigenous AE options. Thus they prefer smear case over cottage cheese, thick milk over clabber/clabber milk, sheep buck over ram, clook over setting hen, den/overden over loft. The same preference applies for options with a cognate in PG/HG as in pancake over hotcake, and options which are morphologically transparent to Pennsylvania Germans: frying pan ( < brotpann) rather than skillet. However, since the preferred items are also (dispreferred) options in the AE speech of AE monolinguals, their preferred use by the OOA can be interpreted as a hardly noticeable, and therefore weak marker of 'old ethnicity'. In view of the well-documented intralingual grammatical ethnicity markers in Black English (cf. Giles 1979: 262-263) the question arises, whether or not the marking of ethnicity is any stronger on the level of morpho-syntax. On this level fourteen types of deviation from the school booksanctioned norms of AE were found by the Essen Delaware Amish Project Team (EDAPT): 1) loss of plural marker, 2) "those" is replaced by ''them", 3) past participle is replaced by preterite, 4) preterite marker is deleted, 5) present perfect and past perfect "have"/"had" are deleted, 6) absence of subjectverb concord, 7) negation after negated verb, 8) negation after/before "hardly", 9) adverb is replaced by adjective, 10) non-standard prepositions, 11) de-

Anabaptist Ethnicity

53

letion of article, 12.1) deletion of relative pronoun, 12.2) relative pronoun is replaced by "what", 13) deletion of genitive "of, 14) non-standard insertion of "yet", "make", "get". However, a comparison with socially comparable monolingual speakers of AE reveals that - with the exception of type 14 their speech exhibits the same types of morpho-syntactical deviation from the school-book norm. Thus the morpho-syntactical component reflects areal features (Bynon 1977: 244) or even national features associated with informal registers or less educated speech styles rather than markers of old ethnicity. On the level of phonology the EDAPT search for ethnicity markers among the OOA of Kent County, Delaware was more successful. On this level the following pecularities could be observed: sometimes loss of final /r/ and loss of sonority in final [ b, d, g ]; more frequently [ d3 ] was replaced by [tf], [ e i ] b y [ e : ] , [ o ] b y [ o ] or [ A ] , and final [ i ] b y [ I ] , [ o ] , or [ e ] . None of these features occurred regularly, but rather at a level of low relative frequency; furthermore, these features were not evenly distributed across the OOA population but clustered rather with a few individuals. Given the insistence on "good English" in the OOA parochial schools (which are English language schools and not geared toward the maintenance of the ethnic varieties, cf. Enninger 1985b), and given the fact that a systematic contrastive analysis of PG (L1) and AE (L2) predicts the above peculiarities, we regard them as cases of interference, and thus as instances of (unintentionally) signifying rather than as instances of (intentionally) communicating and thus of asserting 'old ethnicity'. A matched-guise test (reading a text) supported the descriptive results insofar as the informant judgements indicated a slight, but not a straightforward tendency to identify OOA on the basis of pronunciation alone. What the linguistically naïve listeners perceived as the most salient cue for ethnic categorization was an undetermined intuition about "their strange melody", but not so much segmental units. This is in accordance with the pertinent literature (such as Gumperz; Taylor and Simard; von Raffler-Engel; Lass, Mertz and Kimmel; Dickens and Sawyer; Wächtler). The quality of the tapes, unfortunately, prevented a descriptive analysis like the one which Huffines did in her pioneer work for the less conservative Pennsylvania Germans. Summary: In general, the above linguistic findings for the Kent County, Delaware OOA confirmed the earlier results of Raith (1981a, 1981b) for Lancaster County, PA and of Huffines (1980a, 1984a, 1984b, 1984c, 1984d, 1986) for

54

Werner Enninger

various counties of Pennsylvania. In the language-contact situation of the US (and Canada) these fairly conservative descendants of the Täufer have opted for the accommodation paradigm. They signal 'old ethnicity' predominantly through the continued use of two German-based varieties group-internally, rather than through the way in which they use the outgroup variety AE in intergroup encounters, as Huffines had noted before. (Note that all written communication is in English. For ethnicity markers in written texts, cf. Enninger 1984, 1986b, 1987, Enninger et al. 1985.) 3.2 Linguistic Assimilation: towards Monolingualism in the Outgroup Language In many cases, the transit of the Täufer to other language areas was accompanied by a fast or slow language shift so that the linguistic repertoires of minority and mainstream culture have become, or are becoming, identical. Examples of this linguistic assimilation are: 1) The Zürich and Bernese Täufer in the Palatinate who shifted to the Palatinate dialect within one generation (cf. above). 2) The Dutch to German shift by Anabaptist-Mennonite refugees from Holland settling in Emden, Krefeld, Hamburg, Lübeck, and later in Danzig, Elbing, Königsberg and Graudenz. The transition from Dutch monolingualism through Dutch-German bilingualism to German monolingualism was completed everywhere but in Emden by the end of the nineteenth century (cf. Bender 1973a: 291). 3) The German to English shift by the 18th century Mennonite immigrants to Pennsylvania from Switzerland and the Palatinate (including subsequent settlements in Ontario, Virginia, and Ohio) and by the immigrants from Switzerland, Alsace, and Southern Germany to the area west of the Allegheny Mountains as far as Illinois. Both groups completed their shift to English monolingualism largely in the second half of the 19th century (cf. Bender 1973a: 291). In all these cases the ultimate terminus ad quem of the immigrant repertoire, namely monolingualism in the language of the country of destination, was reached in a complete linguistic assimilation. In a few other cases this shift has been almost completed, such as in the French (speaking) Pays de Montbéliard (formerly part of the Duchy of Württemberg) where only the oldest generation is still bilingual. The same applies to the Les Bulles and the Pruntrut/Porrentruy congregrations in the francophone Swiss Jura (Gratz and Geiser 1973: 671) and the Mennonite enclaves west of the (earlier) French-German language-border fol-

Anabaptist Ethnicity

55

lowing the ridge of the Vosges south across the Sundgau to Biel/Bienne in Switzerland (cf. This 1888; Sommer 1913). Today, the Birkenhof/Altkirch congregation in Southern Alsace supplies simultaneous translations into French for the sake of the younger generations, if the service is held in German - and vice versa. For a close-up picture of the language transition in Kansas, cf. Buchheit 1982. Below we will discuss the case of the more liberal Täufer descendants in Pennsylvania, who are at various points of the transition continuum between bilingualism and monolingualism. In many cases the linguistic assimilation to the out-group language caused serious friction within the Täufer ethnos, because the conservative segments of the groups involved perceived the linguistic assimilation not only as linguistic de-ethnicization, but rather as religious de-ethnicization. They insisted on language maintenance by making claims of higher spiritual values and forfeiture of group principles and even faith in God in the case of the surrender of the mother tongue (cf. Bender 1973a: 291, Buchheit 1982: 111). To the conservative wing, the inherited varieties were not only linguistic markers of ethnicity or instruments of doing ethnicity (by raising the fence against inmarriage and against outsider participation in intra-group domains of interaction), but the inherited varieties came close to being considered as a component of the ethnicity concept itself. Such considerations played a role during the Amish schism between 1850 and 1880 and the Mennonite schisms between 1870 and 1900. Though the language question was not the sole or even the central cause of the ultimate schism, it certainly did play a role. In order to stay with dem alten Gebrauch (the old custom) and in order to strengthen the isolative orientation, the Old Order Amish and the Old Order or Wisler Mennonites opted for the linguistic accommodation paradigm described above and implicitly for 'old ethnicity'. Bender's little article provides at the same time the key for understanding the emergence of a different concept of ethnicity and - in its wake - the emergence of the paradigm of linguistic assimilation in other branches of the Täufer descendants. After isolating the advantages of maintaining inherited languages in other language areas (separation from the surrounding culture, strengthening the sense of nonconformity and internal solidarity), Bender points out a disadvantage: "On the other hand, the language breach has usually prevented a program of active evangelism and outreach..." (Bender 1973a: 290). For ethnicity the evangelistic programs which were started in the 1870s by some Mennonite leaders in the wake of the revival movement in

56

Werner Enninger

some protestant churches (Bender 1973b; Bender 1973c) caused the opening of the (then endogamous) ethnos to outsiders with a non-Täufer biological, migrational and cultural history. Biological continuity was deemphasized for the sake of re-conceptualizing the ethnos in terms of a religious brotherhood of believers - as during the foundation phase. This re-conceptualiziation required the readjustment of the so far distinct ethnolinguistic profile to an indistinct one - as it had existed during the foundation phase. After centuries of linguistic separation the return to the original concept of ethnicity necessitated the deliberate linguistic assimilation, which, however, cannot be equated with de-ethnicization. The Emmental Täufer congregations of today (which never moved to other language areas) and the many Mennonite settlements that in other language areas went through linguistic accommodation and linguistic assimilation attest to the fact that ethnicity - if defined exclusively as a religious brotherhood - is independent of using a separate language. After all, religious principles are not language-specific (cf. Buchheit 1982: 112). However, during the shift toward monolingualism in the outgroup language, intralingual markers in the outgroup language resulting from previous or ongoing language contact may, at least for some generations, carry the linguistic distinctness of the ethnos. Towards the end of the nineteenth century the speech economy of the Mennonites was roughly identical with that of the Old Order Amish today. During the first quarter of the twentieth century one finds several instances of Mennonite scholarly and professional organizations to maintain the Mennonite speech economy in the status quo. By the thirties, however, such voices were hardly heard any more (cf. Buchheit 1982: 112), and more and more communities started on their way towards a monolingual speech economy, i.e. the more extensive use of English for more purposes with concomitant blurring of the boundaries of the domains for the use of each variety and, ultimately, the exclusive use of the outgroup language AE. In such groups which are at various points of the language shift continuum towards ultimate loss of PG and AHG, the question of what language they use is ethnolinguistically less relevant than the way in which they use the language towards which they are shifting, i.e. AE (cf. Huffines 1986). With regard to segmental phonology, Raith (1981a and 1981b) reports for his sample from Lancaster County that the more liberal Anabaptist groups betrayed almost all of the segmental interferences in AE which might be predicted by systematic contrastive analysis. This scope contrasted with that of

Anabaptist Ethnicity

57

the OOA: 1. initial 0 > s (thin), 2. final r > 0 (floor), initial v > w (very), 4. u > o (joke), z > s (bells), 6. initial w > v (want), 7. final v > f (leave), 8. d3 > tf (jug), 9. ei > e: (teenager), 10. A > D (O) (jug), 11 final r before s > 0 (teenagers), 12. final b, d, g > b, d, g (ñb, wood, jug). This is more than just a list, it is rather an implicational hierarchy. Speakers who exhibit interference number one are likely to exhibit also all the higher number interferences. Here it should be remembered that the OOA of Kent County exhibited none of the features one to seven, and that they exhibited none of the higher interferences regularly: interferences eight to ten occurred frequently, and interference twelve occurred only occasionally (Enninger et al. 1985: 5-7). Similarly, in Raith's interference hierarchy the OOA clearly clustered near the low-interference pole. The reduced participation of the Old Orders, and the ample participation of the nonsectarian Pennsylvania Germans in the interference hierarchy was also found by Huffines (1986). In the present context of ethnicity marking this means that the more liberal Anabaptists and nonsectarian Pennsylvania Germans mark ethnicity in their AE speech more clearly through segmental phonetic interferences than the Old Orders do. Summary: In the context of linguistic ethnicity markers this means that the replacement of 'old ethnicity' (endogamous brotherhood of believers sharing the same socialization process and thus the cultural and linguistic traditions) by the revived original concept of the brotherhood of believers necessitated the disuse of the separate inherited varieties; in turn, their function as ethnicity markers was taken over by the way in which the mainstream language is used. The interference-laden variant of English used by the ethnos evokes on the part of the member the same kind of solidarity and on the part of the outsider the same kind of social distance as would the use of one of the separate inherited varieties. "Nevertheless, it has been found that the use of an outgroup language, but with a distinctive ethnic accent, does not detract from the speaker's perceived ethnicity in the eyes of others" (Giles 1979: 256-257; cf. Giles 1973; Huffines 1986). In most cases, such ethnicity markers will be cases of unintentional signification (interference), but the member of the ethnos who has an interference-free variant of English at his disposal may choose to use it and thus totally converge on the outsider's speech. However, he may also choose to intentionally communicate and thus to assert his ethnicity by diverging from the outsider's speech. Whereas the former speaker will be located in social space by his interference, the latter speaker can locate

58

Werner Enninger

himself in social space by selecting among his options. He may try to pass for a mainstream person, or he may assert his ethnic identity.

4. Conclusion This paper sought to show how the ethnicity concept of the Swiss Täufer and its linguistic markers have changed through four centuries. The 'new ethnicity' emerging in the sixteenth century was exclusively defined by a closed set of religious principles not shared by others. Membership was acquired by the deliberate acceptance of the distinctive ideas. Linguistically, the peoplehood idea was not reflected in a separate language, but it was signaled by distinctive ways of speaking and conversational strategies. Within a few generations this original ethnicity concept was complemented by the ascribed criterion of biological descendance and by an open set of cultural traits absorbed in a socialization process within the kinship/brotherhood network. 'New ethnicity' gave way to 'old ethnicity'. Linguistically the ascribed criterion of biological descendance was reflected in an almost closed list of family names. During the (enforced) migration to other language areas 'old ethnicity' was additionally signaled either through the continued use of inherited languages by the side of the outgroup language and the concomitant linguistic interference in the outgroup language if the ethnos opted for linguistic accommodation, or through heavy interference in the outgroup language during language transition, if linguistic assimilation was the choice. In language contact situations linguistic assimilation is the only choice, if the original peoplehood concept of a brotherhood of believers is maintained or revived, as for example among nineteenth or twentieth century liberal Täufer churches in America for the sake of proselytizing outreach. The revived original ethnicity concept, the reversal to the ethnicity concept of the Reformation, required deemphasizing biological continuity of the ethnos. Beside the language transition to the outgroup language, a widening list of family names can be taken as an indicator of the successful implementation of "revived original ethnicity". This presentation of linguistic markers of Anabaptist ethnicity deviates from the extant publications on the same subject insofar as it includes ways and rules of speaking among the linguistic ethnicity markers. Most other studies appear to be based on the assumption that ethnicity is marked in speech by the use of a separate language or/and by the use of features of a

Anabaptist Ethnicity

59

separate language while speaking the language shared by the ethnos and the mainstream group. This line of thought presupposes that ethnicity is associated with the coexistence of different languages in one polity. Such a view is reductive insofar as it disregards the fact that linguistic ethnicity marking is only a specific aspect of locating oneself in social space through speech. Even in seemingly monolingual "language" communities there may be a variety of "speech" communities in the sense that within the one "language" community there may be social differences with respect to what is deemed appropriate to say and how it is deemed appropriate to say it. It is sharing of conversational strategies that creates the feeling of satisfaction which accompanies and follows successful conversation: the sense of being understood, being "on the same wave length", belonging, and therefore of sharing identity. Conversely, a lack of congruity in conversational strategies creates the opposite feeling: of dissonance, not being understood, not belonging and therefore of not sharing identity. This is the sense in which conversational style is a major component of what we have come to call ethnicity. (Tannen 1982: 217) Such considerations are, first, particularly relevant for establishing 'new ethnicity' by an ethnos that emerges within its own polity and which therefore shares its linguistic repertoire with outsiders. While in such a context creating a new and distinct language is unrealistic, the creative use of available linguistic items as ethnic distinguishers is not only feasible, but may also be dictated by the distinct value system of the emergent ethnos. Modified or deleted performative speech acts (adult baptism and the oath, respectively), the substitution of ceremonial events by others (Holy Mass by a congregation meeting (Versammlung, Gemeinde, Gemay)), the selective application of greetings, distinct greeting formulae and forms of address, distinct constraints on speech acts (derision, applause, talking back etc.) and topics, the handling of speech versus conversational silence, the handling of conversational (in-) directness etc. can be derived from the distinct value orientations of the Täufer. Such ways and rules of speaking and conversational styles are more readily transformed to linguistic distinguishers of ethnicity in a three phase process. First, the deliberate and systematic deviation from established rules of appropriateness. Second, the de-canonization of the established rules of social appropiateness through continued rule-infringement. Third, the recanonization of one's own continued usage (Hahn 1987: 28; Assmann 1987). It bears repeating that this de- and recanonization of rules of speaking is not

60

Werner Enninger

only quickly feasible, but also mandatory in view of the innovated value system. Religious value orientations help to constitute rules of speaking, whereas religious value orientations are not associated with or reflected in specific language systems. Tannen's tenet quoted above is, furthermore, of particular importance during and after the final phases of language shift, - in our case - for example during the linguistic "anglicization" of the less conservative descendants of the Täufer in the United States and Canada, because "conversational style is more resistant to change than more apparent marks of ethnicity such as retention of the parents' or grandparents' language" (Tannen 1982: 230). Tannen supports her contention by citing Seaman's study of language transition among Greek-Americans (1972: 204) who - though their Greek was "practically extinct" in the third generation and will be "totally extinct" in the fourth generation - may not have lost, or not lost entirely, Greek communicative strategies. Unfortunately, the ways of speaking of the liberal end of the Täufer spectrum are even less explored than those of the conservative wing, as for example the OOA. If "anglicization" is not "de-ethnicization", but rather a system-linguistic reflex of revitalized original ethnicity, the religious value system should be intact and it should have its reflexes in the ways of speaking even after completed "anglicization". Even limited observation of the ways of speaking of so-called liberal Mennonites suggests that this is more than an abductive speculation.

Linkages of Language and Ethnic Identity: Changes in Barcelona, 1980-19871 Kathryn A. Woolard University of California, San Diego

The links of a language to ethnic identity play a paradoxical role in patterns of language maintenance and shift. On the one hand, a key role in defining or symbolizing ethnic identity has sometimes been seen as contributing to the maintenance of minority languages and nonstandard vernaculars (see, for example, the comparative discussion of American immigrant groups in Conklin & Lourie 1983: 171-177, and Fishman 1972c: 52, 185-187). The importance of such languages in defining community membership and esteemed or rewarded personal identity can make them valued resources that speakers protect. But the linkage of language and ethnic identity has been seen to work differently for dominant languages. One explanation that has been given for the remarkable assimilative power of American society in relation to immigrants, and of the English language in the U.S. traditionally, is that this society and language are primarily "nonethnic" in character. So, "American nationalism was primarily non-ethnic or supra-ethnic in comparison to the nationalisms of most of Europe. ... it did not obviously clash with or demand the betrayal of immigrant ethnic values...." (Fishman 1965: 149, cited in Grosjean 1982: 111; see also Fishman 1972c: 59). "Similar observations may be made concerning the role of the English language in American nationalism. Just as there is hardly any ethnic foundation to American nationalism, so there is no special language awareness in the use of English" (Fishman 1966: 29-30). These disarming features of the American conception and ideology of language have been seen to promote bilingualism and then language shift among immigrants, making them less consciously defensive about language and culture,

62

KathrynA. Woolard

more accepting of the seemingly value-neutral language of upward mobility (Fishman 1972c: 61).2 To the extent that these two principles are both valid, then the rare threatened minority language that survives and revives well enough to present a hegemonic challenge, making a bid to become a principal public language, may be in a paradoxical position. Language planning is often intimately involved with what has been called "identity planning" (Pool 1979). Ethnic signalling value, a quality that can contribute to survival under conditions of subordination, may be a limiting factor when acquisition and use by a larger population become a goal. Such is the situation of Catalan in the autonomous community of Catalonia in Spain. During centuries of state-sponsored subordination to Castilian, and decades of state repression under Franco, the Catalan language survived as a vibrant vernacular, in large part because it has been the key defining criterion of a prestigious ethnic identity associated with high social status (Woolard 1989). Traditionally, most autochthonous Catalans held to a principle by which language practices determined ethnic identity; appropriate use of Catalan was essential to the exercise of a claim to Catalan identity. In post-Franco Spain, the political community of Catalonia (autonomous since 1979) is officially bilingual, and the goal of many Catalan language loyalists and policy makers is for Catalan to become the primary medium of communication in the community, a role long held by Castilian. The attainment of such a goal depends not only on successfully recapturing high and formal domains of language use for Catalan, but also on extending its use to a large population of Castilian speakers. Maintenance of Catalan among the autochthonous Catalan ethnolinguistic group (henceforth "Catalans") is not enough to support the full catalanization of regional institutions, since Catalonia was swamped in the postwar period by Castilian-speaking immigrants. These immigrants and their children (henceforth "Castilians"), many from impoverished sections of Andalusia in southern Spain, the majority concentrated in the working class, form about half the population of the metropolitan area of Barcelona. Recruiting them as second-language speakers of Catalan, under the largely voluntary conditions mandated by the official bilingualism permitted by the central state, is a serious challenge to language planners and educators. The ethnic symbolism of the Catalan language appears to have aided over the years in the recruitment of some speakers among immigrants who were

Linkages ofLanguage and Ethnic Identity

63

socially and psychologically situated to make a leap in identification, renouncing a view of themselves as Castilians and establishing a strongly Catalan identity through their new language behaviors. But ethnographic and experimental research I conducted in 1979-1980 demonstrated that this language-identity symbolic link may have hindered the use of Catalan by more people than it encouraged. An etiquette of language choice was practiced by the great majority of Catalans (virtually all bilingual in Castilian) which led them to switch to Castilian when addressing native speakers of Castilian. This was an automatic, often unconscious exercise of politeness, but it had the force of an ethnic boundary-maintaining mechanism. The exclusionary effect of such "politeness" was apparent in the repeated experience of many learners whose accents revealed their non-native status. Such individuals were often chagrined to find Catalan interlocutors unable to maintain Catalan-medium conversation with them after detecting their Castilian origins. A person not only had to understand and to speak Catalan in order to be addressed in that language, but very often, also had to be taken as a native speaker. In effect, the traditional norm allowed only "Catalans" - native speakers of Catalan - to be spoken to in Catalan. A social-psychological "matched-guise" experiment on language attitudes I carried out in Barcelona in 1980 (based on Lambert et al. 1960), corroborated the social significance of this etiquette. The test found that as judged by young people, the use of Catalan made speakers sound better than did the use of Castilian on a set of traits I have labelled a "status" dimension (intelligence, cultivation, leadership, etc.). This was true regardless of the ethnolinguistic origins of the speaker or the listener. However, on a measure of solidarity, only Catalan listeners valued the speaking of Catalan very highly, and only for native Catalan speakers. The solidarity ratings of these same speakers dropped significantly when they spoke in their Castilian guises. But the speaking of Catalan won no increase in solidary feelings from Catalan listeners when the speaker's accent revealed that she was a native Castilianspeaker. Catalan listeners were indifferent to the language choice of such Castilians. And importantly, Castilian listeners penalized their fellow Castilians for the use of Catalan, with their solidarity scores dropping significantly. The matched guise test showed that Castilian speakers had little to gain in cementing relations with Catalans by attempting to speak Catalan, while they had much to lose in solidarity and support from co-members of their own native

64

KathrynA. Woolard

ethnolinguistic group. In 1980, the use of Catalan was most often taken as a claim to be autochthonously Catalan. This symbolic value of the language worked against its adoption by non-native speakers, since they could not make good on the claim. These positive sanctions for the maintenance of Catalan by native speakers and negative sanctions against its use by Castilian speakers helped explain patterns of language proficiency and use. A survey in 1983 (Direcció General de Política Lingüística 1984) found that, of those born in Catalonia of parents born in Catalonia, 93% claimed Catalan as their principal language. This shows the remarkably minimal attrition of the Catalan language group. However, the demographic structure of Catalonia has changed significantly over the twentieth century, from a largely native-born to a massively immigrant population by the 1960's, and then with economic stagnation in the 1970's, returning to an increasingly native-born population. Immigration has virtually ceased, and among the 15-20 year olds in the DGPL sample, 87% were born in Catalonia, while over half of some older age brackets were immigrants. Among these 15-20 year olds, however, 34% were second generation immigrants, Catalonia-born children of immigrants. Table 1 shows that while Catalonia is again becoming much more native and less immigrant in character, its native-born are much less likely to be Catalan-speaking.

Table 1 Linguistic Profile of Adults in Barcelona Urban Area, 1983* Age Group

N

Catalonia-born

15-20 21-30 31-40 41-50 51-60

122 245 237 192 209

87% 68% 48% 45% 53%

CT-dominant 40% 40% 39% 42% 55%

* CT = Catalan Based on Direcció General de Política Lingüística 1984

Speak CT frequently 43% 49% 44% 46% 58%

Linkages of Language and Ethnic Identity

65

In older generations, we see a very close relationship between birth in Catalonia and the claiming of Catalan as a dominant language (and we also see that indeed, among the oldest group, some immigrants switched their language affiliation to Catalan). But in the younger generations, this relationship between birthplace and language has disappeared. Moreover, in spite of the increasingly native origin of the population, Catalan was not acquiring any more speakers who use it often as a second language than in past decades; in fact, it appeared to be acquiring fewer than ever. (See Woolard 1989 for a discussion of the difference between first and second generation acquisition of Catalan.) These were the sociolinguistic circumstances around the time of transition to political autonomy for Catalonia, when efforts would begin in earnest to restructure the linguistic profile of the region. According to the analysis of the matched-guise test, a successful strategy for extending use of Catalan to nonnative speakers could capitalize on the existing prestige of the language, but ultimately would depend on attenuating its ethnic symbolism. It would be necessary to create a perception of Catalan as a "public voice", rather than a distinctive and private ethnic voice, and the purpose would be to create not simply institutional access but also emotional access to the language for nonnative speakers. Acquisition of Catalan needed to be recast as a practical matter for instrumental purposes, compatible with maintenance of ethnolinguistic identity and the mother tongue. In the following sections of this paper, I will explore the question of whether political and linguistic policy changes in Catalonia have led to a restructuring of the language-identity link that might enable the use of Catalan by new sectors of the population, particularly among teenagers. This evaluation of changes is based on ethnographic and experimental research carried out in Barcelona in 1987.I will first briefly outline significant changes in language policy between 1980 and 1987. Drawing on a variety of ethnographic data, I will then sketch a somewhat impressionistic picture of changes in patterns of language use. Finally, I will present a summary discussion of the results of a more rigorous analysis of changes in language attitudes, as measured by a replication of the matched-guise test.

66

KathrynA. Woolard

A Brief Review of Some Major Linguistic Policy Changes The General Directorate of Language Policy was established within the Department of Culture of the Catalan government in 1980, the year that control over a variety of administrative areas began to be transferred from Madrid to Barcelona. Among other efforts, the Directorate launched a publicity campaign in 1981-1982 to encourage popular acceptance of a plan to "normalize" the use of Catalan. This campaign stressed gradualism, voluntarism, and cooperative efforts to make Catalan the shared language of the newly-recognized community (Woolard 1986). The slogan of the campaign, "Català, cosa de tots" ("Catalan is everybody's") underlines the goal of extending identification with Catalan to those not usually ethnically identified as Catalan. In 1983, the Law of Linguistic Normalization in Catalonia was passed by the Catalan Parliament, establishing a legal frame for full co-official use of Catalan in government, the judiciary, education, public signage, etc. The law also authorized government financial encouragement of Catalan-medium cultural activities. Demands for a Catalan-medium television channel figured in campaign platforms in the 1980 Catalan parliamentary elections, and the Catalan TV-3 was finally established in 1983. Although its alleged political bias toward the governing Catalan party is subject to criticism and debate, the channel quickly won audiences that rival those of the two Spanish-state channels. Changes in educational language policy are of particular interest given the special focus of this research on language use and attitudes among teenagers. In 1979-1980, the only legal mandate for Catalan in the schools specified three hours of study per week of Catalan language and literature. Public (as well as private) schools were allowed to teach in Catalan, but had to complete a lengthy certification process. In that school year, only about 9% of Catalonia's primary schools were registered as offering instruction wholly or partially in Catalan. Catalanization of the schools began in earnest in 1983. The teaching of at least one basic subject matter in Catalan was required for "middle cycle" primary school, and two basic subject matters for upper cycle elementary and secondary schools; since then, the Catalan requirements have been increased. By the 1984-1985 school year, 86% of schools in Catalonia used Catalan as the medium of instruction in at least some subjects (Arenas i Sampera 1987: 87), and by 1986-1987, over 62% of elementary schools had instructional

Linkages of Language and Ethnic Identity

67

programs the Education Department qualified as Catalan (whether exclusively Catalan, progressive immersion, or parallel Castilian-Catalan lines).

Changes in Linguistic Choices? Official public uses of Catalan are important in and of themselves, and as symbols of Catalan's "arrival". For many Catalan speakers, a principal desire is to know that Catalan is a language used for all communication purposes, at all levels of social and political organization that affect or involve Catalan speakers. But there is also a notion that official status bears a relation to the health and maintenance of a language as a vehicle of primary communication, that the support of formal institutions is essential to the survival of languages as vernaculars in literate societies. This assumption about links between public language policies and a population's everyday language practices merits examination, for empirically the relationship is not simple and direct. Latin is of course the example, par excellence, of a language that maintained not only institutional use but hegemony, even as it was lost as a vernacular. In general, I found Catalan to be significantly more visible and audible in central Barcelona in 1987 than it had been in 1980. Not only native Catalan speakers, but also Castilian speakers produced this effect. On television, radio, and in public meetings, I noted non-native but fluent versions of Catalan. Catalan appeared consistently in public sector signage - street signs, subway instructions, etc. - and sometimes monolingually in information distributed by the "autonomic" government (of Catalonia) and by the municipal government. Catalan was also evident in advertisements and signage for commercial establishments. These announcements were most often bilingual, rather than monolingual Catalan, but even this co-presence was a notable change from 1980, when Catalan was rarely evident in commercial writing. It was a testimony to the advances that had been made that Catalanist discourse now derisively attacked "bilingualism", rather than monolingual Castilian usage, as the situation that must be redressed in favor of Catalan. There was also, however, substantial basis for the view that things had not changed much under the new policies. Although the transition to a state of autonomies was in a structural sense complete by 1987, at least in relation to the "historic" community of Catalonia, there was considerable evidence that linguistic policy in the new Spain was still transitional rather than "post-tran-

68

KathrynA. Woolard

sitional". Contradictions and ambiguities in the framework for language policy provided by the Spanish Constitution of 1978 and the Catalan Statute of Autonomy continued to be confronted in the courts. Most policy problems stemmed from conflict between the recognition of linguistic pluralism within the autonomous communities (but not the state), and the still largely centralized system for assigning civil service personnel, particularly in the schools and the court system. Within Catalan social and political life the potential for ethnolinguistic divisiveness was restrained much as it had been in transition years. One striking change was a growing public recognition of the cultural presence of Andalusians in Catalonia. For example, considerable news coverage was given to the growing "Romería del Rocío", a several-day long quasi-religious festival, and to the secular Feria de Barbarà, a massively-attended week-long festival of food, song and dance modelled on the Feria of Seville. At the Feria, one might dance the night away with hundreds of thousands of fellow celebrants, but hear hardly a word of Catalan (a million people were reported to attend the Feria over all in 1987, surely an exaggeration, but one that is interesting in itself). Political parties nervously milled around this consolidating popular-cultural base with its potential for political mobilization, and competition arose among organizers over location and control of the festival. Nonetheless, the festivities maintained an apolitical tone, and some young Castilians criticized them as "bread and circus" distractions from the real issues of life in Catalonia. In spite of the folkloric resurgence, public expressions of Castilian linguistic backlash were still minimal and effectively marginalized. As reported in both the Catalan and Castilian-language news media, the prototypical pursuers of Castilian language rights in Catalonia were "ultra-rightists" such as the lawyer Esteban Gómez Rovira, who represented most of the cases of Castilian language rights claims in Catalonia, and eccentrics such as Carlos Obregón, self-styled leader of a tiny Andalusian Party of Catalonia, portrayed as unstable and given to dramatic hoaxes. No real movement among a traditionally leftist, working class sector of the population could build behind such personalities. Public "convivencia" (getting along together) remained the keynote of relations between these ethnolinguistic groups in Barcelona in 1987 as in 1980, and there was remarkably little political polarization. In the Catalan elections of May 1988, language policy was nearly a nonissue. Only one traditionally Catalanist party, Esquerra Republicana (Republi-

Linkages of Language and Ethnic Identity

69

can Left), pressed for official Catalan monolingualism, and it failed to increase its very small percentage of the vote. On the other hand, in the precampaign period, a conservative and a centrist party, both identified with a centralist state, tested the appeal of the language issue, floating complaints that the Catalan government didn't respect and observe official bilingualism. Both parties backed away from this issue during the actual campaign, apparently after it failed to arouse sufficient interest. The major contenders, the Socialist Party of Catalonia and the victorious Catalan nationalist party Convergence and Union, largely ignored language as a campaign issue, burying mention of language policy deep in their literature. Such studied inattention does not necessarily signal that language policy was established and accepted; it could also mean that mainstream politicians did not control the forces that could be mobilized around the issue. Organized concern for Catalanist language policy was alive and concentrated among younger, newer, and less mainstream political groups, but any grass-roots Castilian discontent with language policy that might exist had not been officially organized.

Language Choices in Informal Domains In informal arenas, Catalan was also more audible from sectors of the population where it had not been found in 1980. Particularly in the center of the city, I heard shopkeepers and clerks, especially the young, who accommodated customers' language choice with fluent Catalan marked by Castilian-speaking origins, an infrequent occurrence in 1980. Moreover, it appeared that a tenacious traditional norm, that two-party conversations should be monolingual, was beginning to erode. This monolingual norm had been strongly associated with Catalan accommodation of Castilian speakers. But more than once in 1987 I heard bilingual conversations among co-workers (e.g. in shops, or among school teachers) in which each spoke his or her own language. On trains I occasionally overheard such bilingual conversations among middle-aged passengers who seemed to be newly acquainted. It appeared, then, that making Catalan a public language, used routinely in formal institutions, had worked to loosen the bonds of traditional etiquette and allow more use of Catalan, by both native and second-language speakers, in inter-ethnic exchanges.

70

KathrynA. Woolard

The Ethnographer's Experience Reactions to my own language use were an indicator of sociolinguistic norms and the language-identity link. In 1980, I usually had trouble convincing interlocutors that I was not Catalan when I was speaking Catalan, even though my control of the language was far from flawless. (The phonetic and syntactic interference from American English was unlike the telltale interference from Castilian that would trigger accommodative switches to that language.) Interlocutors assumed that I was from a far corner of the Catalanspeaking world (most often Mallorca), or that I was not quite bright, but they were often shocked and confused to learn that I was not Catalan. Questions and praise invariably followed the revelation of my foreign origins. Accustomed to this misunderstanding, in 1987 I offered my usual preface, "I'm not from here", in responding to a question from a new acquaintance at a large private party. "Not from here! With that accent she thinks she has to tell me she's not from here!" was the laughing response. Another time, I stopped at a political campaign booth to pick up material. Hearing my response to a question from him, the gentleman in charge shrugged "Oh, you're not from here", and turned away, uninterested in further conversation, since I was not a recruitable voter. Neither was I the oddity worthy of further interest that I had been in the past. While these reactions were a bit of a blow to my ego, they constituted a sign that Catalan was indeed viewed as more "normal" - a language not restricted to an ethnic ingroup, that outsiders might learn like any other, a phenomenon that did not merit explanation or extended comment.

Changes in Linguistic Choices among Students Changes in educational linguistic policy are intended to increase knowledge and use of Catalan among the younger generations. Other official uses of Catalan may also be expected to contribute to changes in language choices, by altering not the knowledge base but the symbolic meaning of the language. Young people, both Catalans and Castilians, are now hearing the proficient non-native Catalan speech of authoritative models, as described above. Teachers, politicians, and mass media announcers, performers, and guests are among those Castilian speakers who routinely can be heard employing Catalan. The understanding of such behavior as a violation of a group norm might have changed accordingly.

Linkages of Language and Ethnic Identity

71

Additionally, young Castilian speakers, whatever their own language habits, are now more accustomed to hearing themselves personally addressed in Catalan (especially by teachers), or counting themselves among an audience addressed in Catalan (e.g. for TV-3). In 1980, most young Castilians took an interlocutor's use of Catalan as a signal that the talk wasn't addressed to them, or that it was hostile. But in 1987, being addressed in Catalan had become a part (even if only a small part) of most school-aged Castilian speakers' day. Official speakers are not necessarily taken by young people as models for language behavior, but their daily uses of Catalan do change its connotations. The 1986 census reflects the effect of instruction on knowledge as reported by parents. For pre-schoolers, 2-4 years old, only 55% are reported to understand Catalan and 26% to speak it. But among 10-19 year olds, 96% understand, 75% can speak, 78% can read, and 60% can write Catalan, as reported by heads of household (CIDC 1987, Tuson 1988). From the census reports, the new educational policies seem to have had a positive effect on knowledge of Catalan. What is less clear, and much harder to ascertain, is how the program of catalanization has affected language use, especially outside the school. The reports of teachers and my own ethnographic observation indicate that there have been some significant changes, but that these do not constitute the fundamental and widespread shift in everyday language choice that some language professionals hoped to achieve. In carrying out the language attitudes test, I once again came in contact with a number of teachers and many students in the five schools in the experimental sample. Moreover, in 1987 I spent nearly four months engaged in extended participant observation in an academic-track secondary school, focusing on one first-year group of 36 students with whom I attended classes, made field excursions, and carried out extended interviews. Some of the teachers of Catalan whom I met, particularly those who worked in schools where the overwhelming majority of students were of Castilian-speaking background, suffered from the professional disillusionment and exhaustion known among teachers in the United States as "burnout". They despaired of any success for their efforts to teach students Catalan. "They never speak it", several told me. This surprised me, because it differed from the classes I had observed. Don't the students speak Catalan in class? I would ask, pursuing the meaning of "never". Ah, well yes, students speak Catalan in class, to the Catalan teacher, "because they have to". But in the

72

KathrynA, Woolard

halls, among the students themselves, one hears only Castilian, teachers told me sadly. The teachers' observations seemed essentially correct, but it was their interpretation that was of interest to me. I remembered well that in 1980, many students didn't speak Catalan even in Catalan class, so I believed there had been more change than the teachers noticed. Underlying their evaluation of the situation were unexpressed assumptions about the goal of Catalan instruction. The explicit goal of educational linguistic policy is that by the end of studies, every student be able to understand, speak, and write both official languages (Arenas 1987: 30). But some teachers of Catalan implicitly assumed another goal for their efforts toward "normalization": to make Catalan the language normally used by their students. Anything short of this would seem to them failure. The teachers' implicit goal was indeed shared by many language planners. But this goal had not been made explicit, and pedagogic strategies were also not explicitly designed to achieve the goal (and indeed, it is not clear from sociolinguistic research that school policies ever could achieve such a drastic change in social practices). In particular, several teachers noted that the emphasis of the secondary curriculum was the traditional one of language studies, on grammatical analysis and correct written forms. Nothing in the secondary school curriculum acknowledged that they were teaching Catalan not only to non-native speakers, but to non-speakers in many cases. (This was less true of primary school, where a Catalan immersion program had recently begun to be promoted with considerable fanfare.) The overt recognition of existing ethnolinguistic boundaries that a special curriculum would demand was usually avoided in official spheres, as for example in the linguistic census, in which no question about mother tongue was asked. The teachers' expectation of hearing new language practices among peers was extremely dramatic, an expectation counseled against by sociolinguistic theory. If young people are to acquire a new language, one that has constituted an ethnic marker, and shift progressively toward its regular use, one of the last domains in which we can expect to find such shift is in peer relations between co-members of the same ethnolinguistic group. Yet this was precisely where some teachers looked for evidence of change, overlooking and disvaluing other concrete signs of progress, and condemning themselves to a sense of failure and futility. Additionally, some teachers hoped for students to be motivated to use Catalan through the emotional and political symbolism that had led the

Linkages of Language and Ethnic Identity

73

teachers themselves to their own profession. A large proportion of teachers of Catalan were, quite understandably, ardent Catalan nationalists; they had chosen to teach Catalan because of their great love and concern for the language as a symbol of their identity and nation. (This was beginning to change, with some younger teachers explaining to me that they had chosen to specialize in Catalan as one might choose to specialize in English, because there would be a good selection of jobs. While such motivation may seem cold to the older guard of Catalanism, it is also a fairly strong sign that the language is becoming "normal".) Catalanists still hoped their students would rise to the defense of Catalan, a sincere motive for the teachers but not conducive to success with young Castilians, particularly since Catalan was not obviously oppressed under conditions of political autonomy. For example, in one university preparation class of nearly 100% workingclass Castilian speakers which I was generously allowed to observe, the teacher discussed diglossia and the possibility of normalization with students (part of the regional curriculum). Characterizing Castilian as the high prestige language in Catalonia, the teacher asserted that bilingualism was abnormal and that bilingual schooling was impossible (positions that I heard repeatedly in 1987). She linked the speaking of Catalan to identity, stating that it reflected whether one feels Catalan or Spanish. (The teacher no doubt spoke from personal experience, since she was a native Castilian speaker herself. Yet she never mentioned her own transformation to her students, and I was warned that the topic was too delicate to broach casually with her.) A student objected strongly to this construction of the language issue, asserting that it created problems for her and her fellow students, forcing them to choose one identity or the other. She wanted to maintain both. Many of these students rejected the teacher's vision of the meaning of language behavior in contemporary Catalonia, and they had their own sociolinguistic analysis of the difficulties of language shift. They explained to me that they never use Catalan except in school classes because they are "socially marginated". However, they expected and planned to use it when they got to the university, because they believed that there they would be among Catalan speakers. The students might well be right. But this would mean that the majority of their young neighbors who would never experience university life were equally unlikely to find a use for the Catalan language. In the most concentrated Castilian-speaking communities on the periphery of Barcelona, particularly in vocational schools, the teachers were quite

74

KathrynA. Woolard

justified in their despair. In those schools, there were still large groups of students even among the youngest classes who appeared basically untouched by linguistic normalization and Catalan teaching. Discourse not only in the halls, but also in Catalan class was dominated by Castilian, a Castilian heavily tinged with the accents of Andalusia even though the overwhelming majority of students had been born in Catalonia.

Case Study: First Year High School Not all schools presented the same pessimistic picture. In public secondary schools around the metropolitan area where students were of more mixed linguistic background, considerably more second-language use of Catalan could be heard. I carried out extended observations in a public high school in the metropolitan area, known for its Catalan orientation, but offering mixed-language instruction to a student body in which Castilian speakers were heavily represented. My case study class of first-year students cannot be said to be representative of the school, particularly since Catalan speakers made up well over half of the group observed. But in that data I have tentatively identified three different patterns of response to catalanization among the fourteen native Castilian speakers in the group, patterns that indicate both continuity and change from 1980. First, as in 1980, there were some students of Castilian language origin who spoke Catalan fluently and frequently. Two boys, David and Toni, had taken on a primarily Catalan identity, deliberately changing over to Catalan language habits in the summer before beginning their studies at this school. These two used Catalan habitually in all their relations with each other and with Catalans, and preferentially with bilingual Castilians. David and Toni followed a pattern found in small numbers in 1980, whereby Castilian speakers (usually immigrants themselves rather than second-generation immigrants like these boys) were found to "catalanize" their linguistic repertoire, and where this change in language habits also constituted a claim to a new Catalan identity. David told me that he had consciously made the language shift because he wanted to be fully identified with his primary circle of friends, all of whom were Catalan speakers. His relations with these friends were close and comfortable, with no comments made about his linguistic identity, although Castilian was used for direct interchanges with him. David

Linkages of Language and Ethnic Identity

75

felt that this use of Castilian signalled incorrectly and unnecessarily a contradictory message, that he was not quite one of them. He chose the moment for his linguistic transformation deliberately, since the change from primary to secondary school offered him an opportunity to make the change in language choice convincingly. In the summer before beginning high school, David made an agreement with Toni, who was changing schools with him, that they would speak only Catalan. Indeed, I found that they used Catalan together at all times, even in murmured exchanges as lab partners in science class. So successful was the transformation that some of David's classmates did not realize he was not a native Catalan speaker until it was mentioned in a class discussion I conducted on the topic. At the other end of the spectrum, there was a group of four female students in the same class who retained strong and self-conscious "Castilian" and "Spanish" identities, although they were all born in Barcelona. Their Castilian identities were at least as strong if not stronger than any I had encountered in 1980. "I'm Catalan, I ought to say so" admitted Susana (meaning she had been born in Catalonia), even as she claimed to feel only Castilian and Spanish. Some (but not all) of these girls felt they had been discriminated against, and recalled instances of insulting epithets. Some belonged to Andalusian cultural clubs and flamenco dance groups, and all were ardent fans of the "Real Madrid" (Royal Madrid) soccer team, a powerful summary statement of Castilian identification in a region where support of the Barcelona soccer team has long been an expression of Catalan identity. This group of girls had acquired some competence in Catalan, but spoke it only when called on in Catalan class. To all of their peers and most other Catalan-speaking teachers, they generally spoke (very comfortably) in Castilian. (Interestingly, it was with me, a non-Catalan adult, that they were most likely to initiate conversations in Catalan). These two types of response to the ethnolinguistic challenge faced by Castilian speakers - either catalanization or consolidation of a Castilian identity - echoed patterns found in 1980. But a new response was displayed by a third subgroup in the class, three second-generation immigrant girls led by Rosa. These were fluent Catalan speakers who used that language consistently with Catalan-speaking teachers, outsiders like myself, and Catalan-dominant peers. But they openly maintained their primary identity as Castilian speakers, publicly speaking Castilian to each other unlike David and Toni. Rosa and her friends switched back and forth between Catalan and Castilian

76

KathrynA. Woolard

as they worked on their science lab assignments in ways that neither the native Catalan speakers in the class nor the catalanized David and Toni did. While all claimed to feel more "Catalan" than "Castilian", this group of girls appeared to be welding a new, more bilingual identity than had been available to Castilian language-origin students of their age in 1980. As before, friendly relations were possible across the ethnolinguistic boundaries. Now it was somewhat less predictable which language would be used in such relations. Castilian girls appeared to accommodate Catalanspeaking teachers and fellow students somewhat more frequently than did the boys. But as before, close friendships fell largely within the linguistic boundaries, except in the instance of one pair of close male friends. (As a student in another school said, "Sure, there's no problem with language choice among classmates, because [the two groups] don't talk to each other".) Susana's and Rosa's cliques had each absorbed a Catalan-speaking girl, and these girls took on the language habits of the circle: nearly monolingual Castilian in Susana's group, and bilingualism with heavy in-group use of Castilian in Rosa's group. For all of these young people, school language policy appeared to play an important role by providing the linguistic knowledge base and creating conditions for some use of Catalan, and most of the students mentioned this in interview. For those socially motivated to make an identity switch, primary school instruction in Catalan had given them an invaluable preparation and made such a switch possible, although it did not directly motivate or cause it. Among those less inclined to adopt a Catalan identity, school still enabled them to learn and practice Catalan to a point where they could express themselves fluently and comfortably, with little hesitation even thought they might incorporate many castilianisms. And lastly, even among the most resistant, complete receptive proficiency had been established through instructional use (which at first had caused them some difficulty), and these students claimed to find it perfectly acceptable that teachers addressed them in Catalan, although they might respond in Castilian. In private interviews, I could not elicit many hints of the linguistic resentment that was so often expressed in 1980. This brief profile shows that there was no uniform response in 1987 to language and identity decisions among young Castilian speakers, even those who found themselves in the same environment. But a new resolution did seem to be becoming available, mitigating the dichotomizing force of ethnolinguistic identity in the earlier period. To the extent that the different resolu-

Linkages of Language and Ethnic Identity

11

tions found in the same schoolroom are predictably patterned, it is in their correlation with residence and socio-economic status. The boys and girls who learned and frequently used Catalan, whether as a habitual language or a fluent second choice, all lived in more central neighborhoods where more Catalan speakers were found, and their parents had middle-class occupations. The resistant, Castilian-identified group were all children of workers living in a Castilian-dominant peripheral neighborhood.

Language Attitudes These ethnographic observations sketch an ambiguous picture, from which it is difficult to extrapolate a more general characterization of the effects of language policy on young people's language choice. Changes in language use on a large scale are of course the ultimate concern in planning and evaluation, but actual language behavior is notoriously difficult to document on a large scale. Given the mediating import of symbolic values, we can usefully consider changes in language attitudes and values even when behavioral changes are not (yet) apparent or are not readily documented. Have changes in the official functions of Catalan affected the connotations of speaking Catalan? As mentioned above, for new policies to be most effective, they need not only to maintain or enhance the greater prestige value of Catalan found in 1980, but to generalize the solidarity value of Catalan as well, extending its benefits to second language speakers. The matched-guise experiment, with the same taped voices, was used again in 1987 to gauge the extent to which such effects had been achieved among the adolescents described above and their peers in four other schools. Although not identical, the 1980 and 1987 student samples were deliberately matched very closely on a number of social dimensions. (See Woolard 1989 and Woolard & Gahng 1990 for details of the experiment and a full presentation of the statistical analysis.) The results for the Status measure are similar in 1980 and 1987, showing a higher evaluation of Catalan than Castilian in both years which, loosely speaking, is more marked in 1987 (see Table 2). All listener groups give higher Status ratings to speakers in their Catalan guises than in their Castilian guises, although the preference was considerably more marked for Catalan listeners than for Castilian listeners and those from bilingual homes.

KathrynA. Woolard

78 Table 2

Mean Status Scores by Language Guise

1980 1987

Catalan

Castilian

.0415 .0951

-.0415 -.0941

While Status did not show any major differences in patterning between the two years, Solidarity did. In 1980, the highest scores went to co-members of the listener's ethnolinguistic group using the ingroup language, and the lowest scores went to these same co-members using the outgroup language. Relatively indifferent, middling scores went to speakers not from the listener's ethnolinguistic group, no matter which language they used. In 1987, the highest scores are still reserved for co-members using the ingroup language, but they are no longer so heavily penalized for using the outgroup language. Moreover, listeners are no longer indifferent to the language choice of speakers from the other ethnolinguistic group, who now fare much better if they use the listener's language than if they use their own (see Table 3). Table 3 Mean Solidarity Scores, Use of Speaker's Dominant Language x Use of Listener' s Language

Speaker uses Listener's language ?

Speaker uses own dominant language? yes no

1980 yes no

.2874 -.0512

-.0369 -.1992

yes no

.1984 -.1656

-.0246 .0626

1987

Linkages of Language and Ethnic Identity

79

The Solidarity scores of the two native Castilian speakers in the speaker sample are of particular interest. Angeles is an Andalusian immigrant whose accent in both languages betrays her origins. Dolores is a Barcelona-born Castilian speaker who speaks a very standard form of Castilian, unmarked by typical Catalan interference, and a Catalan with some phonetic traces of Castilian. In 1980, these two young women's use of Catalan did not win them increased Solidarity ratings from Catalan listeners. But in 1987 it did, and significantly so for Dolores, a marked change. In turn, the strong, statistically significant preference among Castilian listeners in 1980 for the Castilian guise of these same Castilian-dominant speakers was gone in 1987 (see Table 4).

Table 4 Difference Scores on Solidarity, Speaker x Listener's Language f Speaker Angeles (CS)

Dolores (CS)

Montse (CT)

Catalan Castilian

-.1425 -.7200**

.0865 -.5091**

.4889** .0440

.0539 -.1225

Catalan Castilian

.1495 -.1545

.3944** -.1480

.3488** .0723

-.0975 -.2660**

L istener 's Language

Núria (CT)

1980

1987

f

*

Difference scores are derived by subtracting score for speaker's Castilian guise from score for her Catalan guise. Positive scores favor Catalan, negative scores favor Castilian. (CT) or (CS) under speaker's name indicates her dominant language. p