Language in Ethnicity: A View of Basic Ecological Relations [Reprint 2012 ed.] 9783110862805, 9783110106886


188 31 6MB

English Pages 296 [300] Year 1986

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Preface
1. Language and ethnicity in a network of ecological relations
2. The role of language in processes of ethnic fusion and fission
3. Language planning, prestige planning, and the limits of planning activities
4. Bilingual profiles and changing bilingual patterns – an approach to formalize bilingual identity
5. The adoption of foreign cultural patterns as reflected in massive contact languages
6. The use of foreign languages as symbols of prestige in Japan – problems of ethnic identity in modern Japanese society
7. Language in ethnicity: conclusions and perspectives
Bibliography
Subject Index
Index of Ethnic Groups, Languages and Countries (Administrative Territories)
Index of Names
Recommend Papers

Language in Ethnicity: A View of Basic Ecological Relations [Reprint 2012 ed.]
 9783110862805, 9783110106886

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

Language in Ethnicity

Contributions to the Sociology of Language 44

Editor

Joshua A. Fishman

Mouton de Gruyter Berlin · New York · Amsterdam

Language in Ethnicity A View of Basic Ecological Relations

by Harald Haarmann

Mouton de Gruyter Berlin · New York · Amsterdam

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

Library of Congress Cataloging in Publication Data Haarmann, Harald Language in ethnicity. (Contributions to the sociology of language ; 44) Bibliography: ρ Includes indexes. 1. Anthropological linguistics. 2. Sociolinguistics. 3. Ethnicity. I. Title. II. Series. P35.H24 1986 401'.9 86-18179 ISBN 0-89925-097-1 (alle, paper)

CIP-Kurztitelaufnahme der Deutschen Bibliothek Haarmann, Harald: Language in ethnicity: a view of basic ecological relations / by Harald Haarmann. — Berlin ; New York ; Amsterdam : Mouton de Gruyter, 1986. — (Contributions to the sociology of language ; 44) ISBN 3-11-010688-4 NE: GT

Printed on acid free paper.

© Copyright 1986 by Walter de Gruyter & Co., Berlin. All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced in any form — by photoprint, microfilm, or any other means — nor transmitted nor translated into a machine language without written permission from Mouton de Gruyter, a Division of Walter de Gruyter & Co., Berlin. Typesetting: Arthur Collignon, Berlin. — Printing: Ratzlow-Druck, Berlin. — Binding: Lüderitz & Bauer Buchgewerbe GmbH, Berlin. Printed in Germany.

Table of Contents Preface VII 1. Language and ethnicity in a network of ecological relations . 1 2. The role of language in processes of ethnic fusion and fission .

37

3. Language planning, prestige planning, and the limits of planning activities 83 4. Bilingual profiles and changing bilingual patterns approach to formalize bilingual identity

an 119

5. The adoption of foreign cultural patterns as reflected in massive contact languages 155 6. The use of foreign languages as symbols of prestige in Japan — problems of ethnic identity in modern Japanese society . . . . 209 7. Language in ethnicity: conclusions and perspectives

257

Bibliography

267

Subject Index

277

Index of Ethnic Groups, Languages and Countries (Administrative Territories)

281

Index of Names

286

Preface The flow of books and articles which have been published on ethnicity related subjects in the last decade is astonishing and can hardly be fully covered by any one scholar. A comprehensive bibliography in the field is long overdue, as is a systematic outline of the major theoretical positions taken in interdisciplinary research on ethnicity. In order to avoid misunderstandings about the nature of the present study: it neither will nor can be either of these. Both the annotated bibliography and the systematic outline should be produced by groups of researchers representing such different disciplines as sociolinguistics, sociology, psychology, political science, etc. There is a great amount of literature on ethnicity problems written in English and in Russian (to say nothing of other major languages of the world, such as German). Researchers in the Western world who have little or no knowledge of Russian tend to discuss subjects related to ethnic identity in isolation from and ignorance of the views on these issues in socialist countries. At the same time, Soviet researchers and those from other Eastern bloc countries prefer to see the situation in the Soviet Union and in socialist countries in general as unique and uncomparable to the situation in the West. For the most part, it is the language barrier of Russian which impedes the reception of Soviet publications in the West, while it is the ideological barrier which presents an obstacle in the East to the balanced exchange of scientific views about ethnicity problems. I hope that the present study, in evaluating both Western and Soviet positions, will contribute to an effective exchange of ideas. There are, however, other aims more central to this study. First, it is intended to provide a general background for the understanding of ethnicity problems. Based on the assumption that ecological relations constitute the most basic network of interaction in human society, ethnicity is assigned a specific place in that network. The discussion of ecology in general and of language related aspects in particular (language ecology) is not intended to be a definitive treatment of these subjects, but merely to serve as an introduction to the problems treated in this volume. I am hoping that my analysis of ethnicity in a framework of ecological rela-

VIII Preface tions will also challenge preconceptions and encourage a new appraisal of sociolinguistic reasoning. Although there is already abundant research available on ecological problems, there are very few studies questioning currently accepted methodologies in the field. Despite many sociolinguistic and sociological theories about the role of language in its environment there are too few attempts to evaluate methods and categories. Although offered only tentatively, I hope that my framework of ecological relations will be regarded as and employed as a tool for further research on the covariance of language use and the dynamics of environmental factors. Differing from previous attempts is the view presented here that ethnicity related factors should be evaluated as internal ecological variables distinct from external variables like demographic or political factors. The discussion of ecological problems mainly aims at the clarification of concepts and ideas concerning ecological relations (including ethnic identity) as well as of their adequate application in ethnicity research. A second aim of this study is to suggest that research about patrimony (distinctive cultural pattern), up to now primarily carried out in the field of ethnicity, be carried into the area of microsociolinguistics or — depending on the standpoint of the researcher — to specify sociolinguistic aspects on the microlevel analysis which are relevant for the study of cultural patterns and their transmission. As a matter of fact, attempts have already been made to correlate language data with data which are part of the value system of a speech community (phenomenology). One example of this has been the correlation of data about dialectal variation in a language to extralinguistic attitudes of the speakers. Opening the field of patrimony related research also should be understood as an attempt to correlate language data and (ethno)cultural data in an area of ethnicity studies where this has been hitherto neglected. In this connection, special attention is paid to the problem of how to formalize the concept of "acculturation", a term which is used here to mean the adoption of foreign cultural patterns. Evidence of acculturation in massive contact languages is analyzed for a number of such languages and their structures (i. e., the system of numerals and of deictic categories as well as the lexical fields of parts of the body and kinship terms). Research on these subjects may provide a broader insight into phenomena of changing patrimony (that is, of changing cultural patterns). Given the diversity of my two mentioned aims, a third synthesizing and overall aim of this volume has been to generally broaden the horizons of ethnicity related studies. In this regard, the methodological discussion of ecological relations should be seen as an attempt to clarify the

Preface

IX

purpose and range of ethnicity research on an extreme macrolevel, whereas the formalization of acculturation phenomena in sociolinguistic terms is an attempt to shed light on special implications of ethnic identity on a specific microlevel. It is precisely the most general macrolevel and the most specific microlevel which are presently neglected in publications on ethnicity. Other chapters in this volume deal with basic problems of ethnicity, and they refer to ancestry (chapter 2) and patrimony (chapter 4) as well as phenomenology (chapter 6). The discussion about language planning and prestige planning (chapter 3) focuses on problems which illustrate close links between features of patrimony (language as a cultural pattern) and of phenomenology (prestige as a catalyst of planning). In this volume findings from a broad range of ecological settings in Europe and Asia (and also occasionally Africa) are evaluated. It is hoped that this study will provide some stimulating ideas for scholars and students in the field of ethnicity research. I am grateful to the Alexander von Humboldt-Stiftung (Bonn) as well as to the Japan Society for the Promotion of Science (Tokyo) for granting me a fellowship to carry out a long-term sociolinguistic research program. This study includes some findings and evaluations related to that program (see chapter 6). I would also like to express my thanks to Beverly Nelson, Michael D. Ashby, and Dr. Beverly Hill for their kind editorial assistance as well as for their critical comments on the text. Tokyo, April 1985

Harald Haarmann

1. Language and ethnicity in a network of ecological relations

A great deal of research work has been carried out and further investigation will be needed to elaborate existing relations between language and ethnicity. Although many isolated features in the supposed relationship have been described accurately, the larger context seems to have been neglected in the discussion of this complex subject. When discussing ethnicity problems many scholars tend to concentrate on the analysis of identity factors in ethnic processes rather than on the assignment of ethnicity to a given place in the broader network of human relations. In the vast amount of publications on ethnicity, studies tend to be limited to monographs on social, political or language-oriented subjects. This is not to say that there have been no attempts to give an overall look at ethnicity (see p. 9ff.). It is significant, however, that such general outlines either neglect language relations or fail to consider political implications except in passing and then simply subsumed under "societal factors". In this chapter I will try to provide a description of a network in which ethnicity as a cluster of several factors has to be assigned a specific place. A general and comprehensive framework of human relations can best be established in terms of ecology. Following the basic assumption that the interaction between ethnic groups is the result of environmental factors influencing their members, phenomena which are related to a collective ethnic body's identity have to be analyzed in terms of ecological relations. The ethnic identity of any ethnic group comprises elements which are the reflection of a sum of experiences in the group's ecological settings. This statement implies that the identity of an ethnic group cannot be adequately described when environmental factors which have shaped ethnicity to a decisive degree have been left out. Whereas the concept "ecological factor" it mostly understood as "outside environmental factor" is has to be pointed out here that such a view actually is an inadequate simplification of the basic concept of "ecology". The isolation of an ethnic group may be due to geographical separation from other groups. This is an outside ecological factor. The isolation of an eth-

2

Language in ethnicity: a view of basic ecological

relations

nie group, on the other hand, may be due to the people's aversion to contact with other ethnic groups. In this case isolation is what can be called an inside ecological factor. The network of ecological relations outlined below comprises inside as well as outside ecological factors. As language is involved in interethnic relations, it is only a language-oriented view of ecology which can provide the key for understanding ethnicity and its place in the network of ecological relations. As the result of intensified studies in the field of the sociology of language during the last few decades, a number of special areas of study have emerged where research has concentrated on specific problems and subjects. Thus, research is carried out in increasingly specialized fields like ethnolinguistics, contact linguistics (study of language contacts), the study of language politics and policy, language planning, ethnicity (study of self-categorization and identification of ethnic groups), multilingualism, etc. Some of the fields mentioned are sometimes referred to as independent disciplines (e. g., ethnolinguistics, contact linguistics or the study of multilingualism). If this is aimed at emphasizing research activities with different orientations (more societally or more linguistically oriented research) then such divisions are useful but it should not be forgotten that these seemingly separate fields of study are actually only different approaches to a common subject, namely language as a social phenomenon in human society. Language ecology is not specified as a special subject or area of language sociology. The ecology of language is not a field of research with a limited sphere of application or of objects investigated. Rather, it is based on principles of sociolinguistic analysis. Thus language ecology is valuable as a methodology in the sociology of language dealing as it does with the principles underlying the sociological study of language. The concept of "ecology" was first used in the natural sciences. Thus, in biology one speaks about the ecology of plants or animals, about ecological systems, etc. Hawley first transferred that concept into sociology in 1950, and since then ecological principles have been given increasing attention. This transfer of the principles and methods of ecology from the natural sciences to the social sciences has required special adoptions to make them and associated concepts amenable to the research of social subjects. Whereas it is comparatively easy to describe a lake as an ecological system in terms of plant and animal ecology, it is much more difficult to analyze environmental influences on the behavior of two contacting speech communities and their languages. A lake as an ecological system is less complex than the contact setting of ethnic groups. It can be

Language and ethnicity in a network of ecological relations

3

said that any setting involving social networks in human society is much more complex than a setting without such social relations. When Haugen (1972) first applied the concept of "ecology" to the field of sociolinguistics it was with some reservations because of the complexity of society related issues. It should be pointed out, also, that scholars of the sociology of language had already investigated environmental conditions affecting language before, though not using the term "ecology". "Linguists have been concerned with it [i.e., the ecology of language] in their work on language change and variability, on language contact and bilingualism, and on standardization." (Haugen 1972, 327)

One can view language ecology as an attempt to find ecological principles applicable to the social study of language, an attempt to construct models of ecological relations for the purpose of elaborating a general theory about such relations. With such a theory, language ecology could serve as an integrating research perspective, one which promises progress in reaching the general goal of sociolinguistic studies. If one shares the view of Grimshaw that this general goal is the "examination of the interaction of language structure and social structure and of the interimplications of speech behavior and social behavior" (1971, 93), then the application of ecology related principles is likely to become a favored methodology for sociolinguists. Although the previous studies mentioned by Haugen contain valuable material and useful evaluations of empirical findings, there is in them a lack of ecological reasoning and methodological discussion. Haugen himself took the first decisive step in that direction. Language ecology should cover the whole network of social relations which control the variability of languages and their modal speakers' behavior. The overall network of social relations, however, is not a clearly defined set of factors which can be mechanically employed by scholars in the social sciences. As social relations (including language) are characterized by a high degree of complexity, efforts have to be made to distinguish ecological factors in a range of ecological functions (see p. 7ff.). Also, as even the range of environmental factors is not clearly defined and as the sociology of language as a scientific discipline cannot be separated from other fields of social studies and linguistics proper, language ecology has to cover the whole range of subjects studied in different areas of language sociology.

4 Language in ethnicity: a view of basic ecological relations When referring to the "interactions between any given language and its environment" (Haugen 1972, 325) it may seem obvious at first which relation is indicated thereby. Thinking about the consequences resulting from such a statement, however, the specification of environmental factors appears quite difficult for the analyst. Besides the problem of deciding whether ecological variables are linked to both the ethnic group and the corresponding national language, or to the language only (see p. 18 for either predominantly group related or language related variables), it is not easy to find a solid basis for specifying ecological variables according to their different functional ranges. In this connection, functional ranges have to be considered for single ecological variables which constitute the tentative inventory of ecological factors (see p. 11 ff.). The problem of environmental factors even affects the general question of isolating and identifying which social relations shape the network of conditions determining the existence of natural language. The specification of environmental factors according to their functional range is thus a methodological problem of general sociolinguistic importance as it is directly related to the theoretical foundations of the sociology of language. The dispute about how to define micro-and macro-sociolinguistics as fields of study (that is, whether research should be more linguistically or societally oriented) has its counterpart in the ecology of language with the possible difference that language ecology has also to concentrate on the dynamics of variables in an ecological system. In my opinion, there is no question but that political factors and their functional range should be included. The same is true for ethnolinguistic variables. Ethnolinguistics is sometimes considered as a field of study separate and independent from that of sociolinguistics and language ecology. This view appears to me too formalistic, leaving out the many areas of interdependence between sociolinguistic and ethnolinguistic phenomena (e.g., in the study of language contacts). The problem of specification is therefore linked to the general question of whether or not to provide a general (overall) network of ecological factors. And then we must ask where the limits of a general ecological system should be set. As the fundamental variables of language ecology are linked to the speakers of a given language (see p. 18), I consider the following basic relations to be the most comprehensive as a general framework for an ecological system: INDIVIDUAL : G R O U P : SOCIETY : STATE

Language and ethnicity in a network of ecological relations 5 This string of concepts leading from the most specific (individual) to the most general (state) could be interpreted as a hierarchical structure with different levels of complexity. The links between the different levels are each of a special nature. In Haarmann (1980a, 39ff.), I pointed out that the study of language contact and bilingualism with reference to groups of speakers is of a different nature than similar phenomena in the language behavior of an individual speaker, and that the methodologies employed to study them should reflect this difference. There are also certain factors that are only relevant to group or to individual language relations. This is the case, for instance, with diglossia. Diglossie or polyglossic settings can only be referred to in terms of group relations in speech communities and not in terms of individual relations or conditions of individual speakers (see Haarmann 1983 a, 375 ff.). It is also important to note the conceptual difference between the terms "group" and "society". Society in this connection is the general concept comprising the subordinate concept of group. Thus society is considered to be the most complex organization of social groups (ethnic, professional and political groups, speech communities, etc.). At first sight the relation between society and state (or state organization) seems obvious but in the traditions of sociolinguistics proper it is seen to be a relation of a special nature. Most theoreticians consider society to be the most complex social organization, implicitly including all political relations that maintain the functioning of the society. The hierarchical structure of basic ecological relations given above indicates, however, that the political implications of state organization are not included in the concept of society but represent a component of their own. This view also implies that the societal organization is considered to be subordinate to the political relations of a given state. A society cannot exist without or outside the governmental leadership of a state. One speaks of a society as democratic when it has a democratic government, democratic political institutions. Conversely, a society under totalitarian rule cannot be considered to be democratic. But, as the political organization of a given state may exist without support from the society (/. e., a military regime), it is more practical to set up a special relationship between the society and the state (as in the above string of concepts) rather than to subsume political relations totally under societal conditions. The différenciation of society and state as separate concepts has important consequences in the specification of environmental factors (see p. 11 ff.). Of course, this interpretation of the links between society and state is particularly suited to conditions in modern industrialized societies, but in

6

Language in ethnicity: a view of basic ecological relations

general for the purpose of making an inventory of variables, the identification of special political variables in addition to a variety of societal factors is useful. Language is not mentioned in the string of basic ecological relations. This is due to a special interpretation of the relations between language and its environment. As any natural language has no abstract existence but is always linked to individual speakers or groups of speakers, there is no direct relation between language and its environment. Instead, language is affected by ecological factors only in the sense that the speakers of that language are influenced by different factors of their environment. The general description of language ecology presented by Haugen (see p. 4) must therefore be considered as an elliptic expression leaving out the special position of the speakers. Although the concepts (t. e., individual, group, society, state) are not language related they nevertheless imply basic links with language. Language ecology is primarily concerned with language in its fundamental forms of existence which correspond to the different levels in the above string of concepts: language behavior of the individual speaker, the role of language in group relations, the functional range of language(s) in a given society, and language politics in a given state. Other special aspects of the interdependence of language as used at different levels of the hierarchy are discussed below. Another stipulation has to be made before specifying the functional ranges of environmental factors. It has already been pointed out that language as the means of communication for an individual speaker is of a different nature than language within group relations in a speech community. The proposed basic or general inventory of ecological variables in this study comprises those factors which are related to groups of speakers rather than to individual speakers. In my view, a general framework of ecological factors can only be realized on the basis of intra- and intergroup relations which are indicated by the concepts of group, society, state. Individual relations can easily be separated from the general group relations and may be illustrated in a special ecological model which is based on the general model of ecological processes (e. g., individual relations like language skills among bilingual individual speakers, language choice among bilingual speakers for communication with friends, etc.). Besides this methodological explanation there are technical reasons for concentrating on group related variables rather than including both group and non-group related factors. As it is generally more complicated to specify factors influencing the individual speaker, and also because there is a greater variety of non-group related varia-

Language and ethnicity in a network of ecological relations

7

bles, it would be difficult to illustrate special variables influencing individual speakers and general factors affecting groups of speakers in one and the same inventory of ecological variables. Some misunderstandings arise from earlier attempts to set up an inventory of factors affecting bilingualism which failed to clearly distinguish the two sets of variables. Mackey (1976, 275 ff.) includes group related and non-group related variables but the priority given by him to the bilingual speaker is not explained. The same is true for an earlier version of a general ecological framework by Haarmann (1978, 54ff.). In order to make the inventory of variables most comprehensive, the following wide range of functions should be taken into consideration: 1. 2. 3. 4.

Ethnodemographic variables Ethnosociological variables Ethnopolitical variables Ethnocultural variables

5. Ethnopsychological variables 6. Interactional variables 7. Ethnolinguistic variables

(see p. l l f f . for the specific variables included in each category).

1. Ethnodemographic range of ecological functions The relevance of factors in the network of ethnodemographic functions for language ecology results from the general importance of demographic factors in the evolution of communities. With respect to ecological relations, the general demographic factors must be related to the différenciation of ethnic groups and thus be transferred from the general field of study, demography, to the ethnically oriented sub-field, ethnodemography.

2. Ethnosociological range of ecological functions The influencing factors which are combined in this range of functions refer to the social conditions of ethnic groups in their concrete contact settings. This factor range carries important implications for the principles and subjects as studied in population sociology. The focus on ethnic groups as reference groups makes ethnosociology a specific field of research, different from general population sociology which deals with communities without reference to ethnicity.

8

Language in ethnicity: a view of basic ecological

relations

3. Ethnopolitical range of ecological functions Those variables important in shaping the relations between the social structures of ethnic groups and the political structures of the state organizations they exist under (variables of the interrelations between social and political structures) are specified here as ethnopolitical factors. The dynamics of these factors have affect on the political status of ethnic groups and their languages.

4. Ethnocultural range of ecological functions Ethnic groups are distinct because of their specific cultural traditions and behavioral norms. Ethnocultural variables constitute a network of relations which are vital for the evolution of cultural patterns and cultural profiles of ethnic groups and which also includes variables of cultural exchange in interethnic contacts. The cultural pattern does not only comprise ethnically specific features or conditions but also specific cultural activities of ethnic groups {e.g., language cultivation).

5. Ethnopsychological range of ecological functions Group solidarity of individuals towards ethnic groups as well as intragroup communication (interaction among members of the same ethnic group) and intergroup communication (interaction between members of different ethnic groups) are controlled by a variety of attitudes about and evaluations of the roles of the reference group and other contact groups. These control mechanisms are specified here as ethnopsychological variables.

6. Interactional range of ecological functions The network of dynamic factors which control interaction in a speech community is specified as a distinct functional range of ecological factors. The factors refer to the conditions of intra- and intergroup interaction as well as to communicational capacities (including multilingual mobility) of single speech communities.

Language and ethnicity in a network of ecological relations

9

7. Ethnolinguistic range of ecological functions The variables in this range of ecological functions comprise factors which are directly related to the language (communicational means) of an ethnic group and its structure. There is a variety of ethnolinguistic features but only those variables are included in this network which — according to hypothetical assumptions based on empirical findings — play a key role in varying structures of verbal behavior in a speech community. Factors determining the profiles of speech acts under differing conditions of language contacts are taken into consideration. Note: The specification of different ecological ranges (or functional ranges of ecological variables) is an important preliminary step towards setting up a general (or basic) model of ecological relations. The factors by which the integration into the different ranges can be illustrated are specified below (see p. Uff.). It should be pointed out that there is a close interdependence between all ecological ranges and that no range dominates the others. The only value of concentrating on the analysis of factors in a special range is in technically limiting the area of research in order to gain deeper knowledge about variable relations in one category. Findings then have to be correlated for perspectives on the inter-range dynamics of ecological factors. Several previous attempts have been made to clarify the network of ecological variables in some of the categories specified above. The approaches chosen tend to reveal the scientific background of the scholars who set up a given tentative inventory of variables or ethnicity-influencing factors. Giles, Bourhis and Taylor (1977) tried to specify the role of language in ethnic group relations. Their approach was ethnopsychological. However, valuable as their theoretical framework is, it neglects to mention specific language relations, with only one special factor "language status" mentioned in the model of "structural variables affecting ethnolinguistic vitality" (Giles, Bourhis and Taylor 1977, 309). Language status as it is described, has to be viewed as a cluster of variables rather than as a single variable. Although for the purpose of a macrolevel analysis such a lack of specification of the full range of language related factors may be acceptable, the value of the theoretical discussion will remain limited when applying the network to problems on the microsociolinguistic level. When Giles, Bourhis and Taylor use terms "ethnolinguistic group" and "ethnolinguistic vitality" they apply them

10

Language in ethnicity: a view of basic ecological relations

both to general language-ethnicity relations and to features of language structure (and status). This use, however, is misleading, because the structural features of language have first to be shown to be related to social phenomena before lumping them together with ethnicity components. If Giles, Bourhis and Taylor were linguists or ethnolinguists they would most probably have preferred to define language-oriented variables more accurately and to separate ethnolinguistic variables clearly from other clusters of societally relevant factors. The extension of the meaning of "ethnolinguistic" to a kind of general cover word for language in ethnicity makes the term useless. The way Giles, Bourhis and Taylor use the term "prestige" seems to indicate the same tendency toward an over-generalized meaning, in this case covering the whole variety of status factors. As a matter of fact, prestige is only one special aspect in the whole network of ecological variables and has to be assigned a specific place in relation to phenomenology (see also chapter 3 and 6 for prestige related phenomena). Prestige as one element in the value system of ethnic groups has furthermore to be viewed in a general ecological framework where self-identification and the categorization of others play the role of a filter in interaction (see model of ecological relations on p. 27). Any theory of language ecology must specify the ecological variables affecting language in its environment. Although several attempts have been made to establish an inventory of ecological variables, most previous classifications lack specific différenciation and can only be employed for general descriptive purposes. This is true of Ferguson's socalled sociolinguistic profile formulas which are a "summary description of the language situation" (1966, 309), Stewart's "sociolinguistic typology for describing national multilingualism" (1962 and 1968), Kloss' criteria for the establishment and identity of language communities (1966), Fishman's inventory of variables of language behavior (1965), Mackey's "variables du bilinguisme" (1976, 371 ff.), Allardt's criteria for ethnicity (including language related variables) (1979, 52ff.), the "structural variables affecting ethnolinguistic vitality" established by Giles, Bourhis and Taylor (1977, 309), and several other attempts. Differences between these classifications mainly result from different methodological approaches and are due to different ideas about the aims of such inventories. Most of the ecological inventories thus far assembled could be considered as partial, as panels in a broader general panorama of ecological variables, and will be treated as such in the following discussion.

Language and ethnicity in a network of ecological relations

11

A crucial choice in establishing an inventory of language ecology is that of whether to take into consideration all possible variables which could indirectly as well as directly affect language choice and behavior in groups (i.e., among members of speech communities), or to restrict the inventory to only language related categories. I consider an overall inventory more adequate because it allows the inclusion of general as well as specific variables, and at the same time illustrates the links between variables referring to the general conditions of ethnic groups and those referring to their language. In Haarmann (1980b) I proposed a general inventory of basic ecological variables, the categories of which shall be discussed below.

1.

Ethnodemographic variables

1.1. The size of an ethnic group (number of members in a community, i.e., language community, national minority, etc.) 1.2. The polarity between focused and dispersed population in ethnic groups (concentration versus dispersion as features of settlement) 1.3. The polarity between ethnic homogenity and heterogenity in the area of an ethnic group's settlement (monoethnic versus polyethnic area of settlement) 1.4. The polarity between urban and rural settlements within an ethnic group 1.5. The polarity between static settlement and migration movement in an ethnic group 2.

Ethnosociological variables

2.1. The polarity between stability and dynamic change in the ethnic profile of areas of settlement (e.g., the distribution of speech communities in a region) 2.2. The distribution of the population in an ethnic group by sex 2.3. Age-group distinctions as an ecological variable influencing language choice and speech behavior 2.4. The specifics of social stratification in an ethnic group 2.5. The specifics of family relations in the social structures of an ethnic group (e.g., endogamy versus exogamy as features of family relations) 3.

Ethnopolitical variables

3.1. The ethnos-state relation 3.2. The speaker-language-state relation (group- and non-group related bilingualism) 3.3. The institutional status of a community's language (cf., categorizations like language of the state, official language, language for administrative usage, etc.)

12

Language in ethnicity: a view of basic ecological relations

3.4. The reproduction potential of a community's language (referring to its special status as a medium of instruction or as a subject at school) 3.5. The characteristics of the division of labor (hierarchical versus segmented division of labor in the relations of an ethnic group with other ethnic communities in a state) 4. Ethnocultural variables 4.1. Ancestry (descent) as a criterion of group solidarity 4.2. The polarity between ethnocultural patterns and social distance in interethnic relations 4.3. The relevance of cultural and/or political organizations for the promotion of a community's interests 4.4. The relevance of a language's ausbau status 4.5. The specifics of a language's sociocultural potential (e.g., characteristics of the written language according to different categories of usage, as for science, commerce, etc.) 5. Ethnopsychological variables 5.1. The relevance of enculturation for ethnic identification 5.2. The relevance of self-categorization (self-identification) among the members of a community 5.3. The relevance of and ways of categorizing other ethnic groups among the members of a community 5.4. Language maintenance as a measure of ethnic identity 5.5. The attitude of the members in a community towards interaction with members of contacting ethnic groups (inclination towards interethnic communication versus rejection of contact) 6. Interactional variables 6.1. The relevance of communicational mobility in a language community (lowlevel mobility of monolingual speakers versus high-level mobility of multilingual speakers in a community) 6.2. Interactional determination in the use of communicational means (cf., language varieties in diglossie and polyglossic settings) 6.3. The relevance of intra- and interethnic role relations for interaction 6.4. The degree of routine interaction with members of other ethnic groups (degree of familiarity with interethnic communication among members of a community) 6.5. The degree of publicity (publicness) of speech settings 6.6. The relevance of topic for intra- and intergroup interaction (with topics ranging from general political to special private subjects) 7. Ethnolinguistic variables 7.1. The relevance of linguistic distance between contact languages (problem of contacting languages with different degrees of linguistic distance/abstand)

Language and ethnicity in a network of ecological relations

13

7.2. The relevance of ethnically specific pragmatic strategies of verbal interaction 7.3. The role of grammatical determinism within the framework of deictic categories (specifics in the system of deictic categories and their usage in contacting languages) 7.4. The caracteristics of language contacts with respect to the sociocultural status of the contacting languages (sociocultural categorization of language contacts) Note: Henceforth, when the above ecological variables are referred to in this study, they will be indicated by using the numerical code given here.

The above mentioned environmental factors are called ecological variables because their influencing effect causes variations in multilingual settings (especially in the language behavior of speech communities). There is no space here to explain all possible variations of all environmental factors. Only those variations which require additional information beyond what is given in the above inventory will be mentioned. Variations of single factors will be illustrated with concrete examples for better understanding. Variable 2.1. (Variations) Strong population growth (high birthrate) in one ethnic group as a triggering factor for changes in the ethnic profile of a multinational region (e. g., Uzbeks in Uzbekistan) A stagnating population in a given ethnic group as a factor for slow change in a contact region (e.g., Sorbs in Lower Lusatia) An ethnic group which is characterized by a high degree of negative growth (high rate of population decline) (e.g., Karelians in the Karelian A.S.S.R.) Variable 3.1. (Variations) The ethnic group is given autonomy rights in a multinational state (e.g., the autonomy of the Swedish-speaking population on the Aland islands as an autonomous region of Finland) The ethnic group has partial autonomy in a jurisdiction in which nationality is promoted (e.g., the Welsh community in Wales) The ethnic group has partial autonomy in a jurisdiction in which nationality is tolerated (e.g., Friulians in Italy) The ethnic group does not possess any special right to maintain its language and culture (e.g., the Kurds in Turkey)

14

Language in ethnicity: a view of basic ecological relations

Variable 3.5. (Variations) The members of an ethnic group are represented by all social groups or classes in a country's society (e.g., segmented cultural division of labor as illustrated by the situation of Finland Swedes in Finnish society) The members of an ethnic group are socially represented only in the private not in the public sector (e.g., partially segmented cultural division of labor as among the Poles in Germany before 1914) The members of an ethnic group are only represented in lower social classes (e.g., hierarchical cultural division of labor as illustrated by the situation of Moroccan guest-workers in West Germany) Variable 4.2. (Variations) There is a strong social distance between two ethnic groups because of great differences in their ethnocultural patterns (e.g., the relations of Turks and Germans in West Germany) There is a moderate social distance between ethnic groups (e.g., the relations between Castillans and Catalans in Spain) There is a weak social distance between ethnic groups (e.g., relations between Finns and Estonians in the Estonian S.S.R.) Variable 5.1. (Variations) Ethnic identity on the basis of an ethnically specific monocultural education/upbringing (e.g., Faringians on the Faroese islands as an autonomous region of Denmark) Ethnic identity on the basis of a bilaterally specific bicultural education/upbringing (e.g., Finland Swedes in bilingual areas in southern and western Finland) Ethnic identity on the basis of an ethnically unspecific mixed education/upbringing (e.g., culturally mixed education as among members of the second generation of Greek guest-workers in West Germany) Variable 6.2. (Variations) A language only functions as a high variety (Η-variety) in the functional distribution of languages (language varieties) in a society (e.g., French in Belgium as the only official language in the 19th century) A language only functions as a low variety (L-variety) in the functional distribution of languages (e.g., diglossia in Switzerland with Schwyzertütsch as L-variety) A language has a special function in a polyglossic setting involving several varieties (e.g., Gagauz as an Li-variety in a polyglossia involving also Moldavian as a L2-variety and Russian as a Η-variety in the southern part of the Moldavian S.S.R.)

Language and ethnicity in a network of ecological relations

15

Variable 6.3. (Variations) The ethnic group plays a leading role in a multinational society (e.g., Jews in the U.S.A., Russians in the Soviet Union) The ethnic group plays an equal role together with other groups in a multinational society (e.g., Flemings and Walloons in Belgium) The ethnic group plays a non-dominant role in a multinational society (e.g., Kurds in Iran) Variable 6.5. (Variations) A language is used in all domains of public and private life (e.g., French in Québec, Canada) A language is used preferentially in the public sector (e.g., French in Luxembourg) A language is used preferentially in the private sector (e.g., Letzeburgish in Luxembourg) A language's use is restricted to the private sector (e.g., Breton in France) Variable 7.1. (Variations) A language is characterized by a great linguistic distance from a contacting language (e.g., the distance between Basque and Spanish, the contacting language, in Spain) A language is characterized by a medium degree of linguistic distance from a contacting language (e.g., the distance between Occitan and French, the contacting language, in France) A language is characterized by a small linguistic distance from a contacting language (e.g., the distance between Frisian and Dutch, the contacting language, in the Netherlands) Variable 7.2. (Variations) The verbal strategies for interaction differ strongly among contacting ethnic groups (e. g., the language behavior of Eskimos in contrast to that of Franco-Canadians in northern Québec) The verbal strategies for interaction differ moderately among contacting ethnic groups (e.g., the language behavior of Galicians and Spaniards in Northwest Spain) The verbal strategies for interaction differ only slightly among contacting ethnic groups (e.g., the language behavior of Occitans and Catalans in southern France) Variable 7.3. (Variations) The system of deictic categories in a language is strongly affected by a contact language (e.g., the deictic categories of Romany under the influence of various co-territorial languages; see chapter 5)

16

Language in ethnicity: a view of basic ecological relations The system of deictic categories in a language is moderately affected by a contact language (e.g., the deictic categories of Izhorian under the influence of Russian; see chapter 3) The system of deictic categories in a language is scarcely affected by a contact language (e.g., the slight influence of French on the deictic system of Breton)

The variations illustrated here can be further refined for each ecological factor. A refinement, however, mainly depends on the experience resulting from the application of the inventory to a variety of concrete settings. Empirical findings for a number of ethnic groups may reveal special cases of variations which do not yet show up in the present categorization. Empirical research based on the inventory of variables should be in two directions. On the one hand, it is necessary to apply the inventory of variables to single groups with the aim of identifying as many factors with their specific variations for a given group as possible. This would make possible broad comparisons between single groups. On the other hand, research into single ecological factors is needed to discover potential variations in a great number of compared settings. That is to say, study of a single variable in a variety of settings is needed. The advantage of this kind of multilateral comparison which is restricted to one or a few criteria is that it enables researchers to extend and elaborate the inventory of categories of potential environmental factors. The number of possible variations for a given variable may be specified. I have shown four possible variations for the factors 3.1. and 6.5., and have labeled them 1, 2, 3 and 4. In other cases variables may show a three-fold variation. For convenience, an ethnic group could be characterized on the basis of the inventory as a community with the typical features 3.1.1., 3.5.1., 5.1.1 or 7.3.3. These particular features are associated with settings like that of the Swedish-speaking population on the Aland islands (geographically situated between Finland and Sweden, culturally strongly related to the Swedish community in Sweden and politically integrated as an autonomous region into the Finnish state). There are many advantages to applying a numerically coded inventory of variables to concrete settings of ethnic groups in their ecological relations. Details of such an application are discussed in connection with the implications of a basic model of ecological relations. The most far-reaching application of this kind of formalized inventory of ecological factors in terms of their variations is the production of a typology of patterns for language maintenance, language shift, etc. In the framework of such

Language and ethnicity in a network of ecological relations

17

types of language behavior, variations of single ecological factors may be seen to be clearly associated with behavioral patterns, making the network of conditions easy to identify. For an understanding of typological perspectives and other implications, further explanations are needed with respect to the nature of the variables. The inventory includes ecological variables with different conceptual associations, one of the main polarities being the relation to the speech community, on the one hand, and to the language on the other. When referring to the conceptual differences between the functional ranges (e. g., between the ethnodemographic and the ethnosociological ranges) one can see the conceptual polarities they describe as a kind of superstructure. As can be clearly seen in the following table (cf. Tab. 1), this superstructural polarity of community and language related variables normally does not fit neatly into the framework of ecological functional ranges. Whereas the ranges of ethnodemographic and ethnosociological variables (1 and 2) are one-sidedly community related, language related variables predominate in the ethnolinguistic range (7). The other functional ranges (3, 4, 5 and 6) are characterized by both community and language related variables. It does not seem appropriate to split up the latter ranges only because they include both kinds of variables. The distribution of variables in ranges 3, 4, 5 and 6 illustrates the interdependence of variables related to the speech community (see 3.1, 3.5) and to the mother tongue (see 3.2, 3.3 and 3.4). In fact, the polarity of variables with different conceptual concentrations within the same functional range stresses the fact that an overall description of multilingual settings in ecological terms must always take into consideration not only the language but also the speakers of the language (i. e., those who are representatives of the living language). It has already been pointed out that there is no direct relation between the language and its environment but rather an indirect relation involving the speech community which is directly influenced by environmental factors. As the main aim of the inventory of ecological variables here introduced is to represent group relations, all variables which do not concentrate on the language itself refer to the ethnic group or speech community, and not to the individual speaker. Relations between the speech community and its individual members have to be illustrated separately. When referring to language group phenomena, phenomena of language behavior related to groups rather than individuals are indicated. This is a basic presupposition for the relations in Tab. (1).

18

Language in ethnicity: a view of basic ecological relations

Tab. 1. Relations of ecological variables to the speech community and their language Conceptually related to the speech community

Ecological variables

Conceptually related to the mother tongue

XX XX XX XX XX

1.1 1.2 1.3 1.4 1.5

X X X X X

XX XX XX XX XX

2.1 2.2 2.3 2.4 2.5

X X X X X

XX X X X XX

3.1 3.2 3.3 3.4 3.5

X XX XX XX X

XX XX XX X X

4.1 4.2 4.3 4.4 4.5

X X X XX XX

XX XX XX X XX

5.1 5.2 5.3 5.4 5.5

X X X XX X

X X XX XX X X

6.1 6.2 6.3 6.4 6.5 6.6

XX XX X X XX XX

X X X X

7.1 7.2 7.3 7.4

XX XX XX XX·

Language and ethnicity in a network of ecological relations

19

When specifying single ecological variables, setting up a systematic inventory of ecological factors, and characterizing ecological relations in a model, one encounters the problem of how to illustrate the potential influence exerted by different factors on the behavior of ethnic groups. By potential influence I mean that any specified ecological variable may influence the language behavior of a given group. Under certain conditions in concrete ecological settings, however, the influence of some variables may be blocked by the dynamics of other variables or their influence may be negligible. The given inventory of ecological variables comprises factors that, from a theoretical standpoint, are potentially relevant for intergroup relations. The actual influence of any given variable can only be tested by the application of the inventory and the related model in characterizing and analyzing concrete settings. I have found, in using the inventory of variables and relating the findings to the variable network of the basic model, that in a given concrete setting (e. g., variant structures of bilingualism) some variables exert a clearly distinguishable influence while the influence of other factors seems to be insignificant. In order to illustrate the phenomenon which I will call the potentiality of influence or the potential dynamics of ecological variables, let me refer to two concrete settings of language maintenance (preservation of the mother tongue) among ethnic groups in the Soviet Union. These are the Moldavians and the Gypsies. In the diagrams illustrating variant structures of language maintenance among the two ethnic groups, two different variables are taken into consideration. These are one ethnodemographic (1.4 The polarity between urban and rural population) and one ethnosociological variable (2.2 The distribution of the population within an ethnic group by sex). Both variables are specified with respect to different regional settlements of Moldavians and Gypsies. As in this context the discussion concentrates on the problem of potential influence (or dynamics), the specific rates of language maintenance for single

4 Nòte: As explained in the text, the terms "speech community" and "the language" (or mother tongue) do not indicate an opposition between two extremes but rather a dichotomy of concepts which are interdependent. The presentation above should therefore not be misunderstood as a selection of opposite features. The number of crosses merely indicates the strength of impact of single ecological variables in the inventory.

20

Language in ethnicity: a view of basic ecological

relations

groups are not relevant. Neither is the comparative level of language maintenance among Moldavians and Gypsies relevant here. I will merely comment on the dynamics of the applied variables. A comparison of the diagrams representing demographic milieu (i. e., the différenciation between urban and rural population) immediately reveals that this ecological factor leads to different structures of language maintenance for the two ethnic groups. The significance of this factor is clearly illustrated by differing proportions or rates of language maintenance among the urban and rural populations. At the same time, the comparison reveals that proportional differences are specific for the Moldavians and also specific (that is, specifically differing) for the Gypsies. This results from differing dominances of language maintenance in urban or rural areas. Whereas there is a clear polarity among the Moldavians (with a dominance of language maintenance in rural areas), shifting dominances and proportions can be seen in the Gypsy settlements (see graphs on p. 2Iff.). A comparison of the diagrams representing differing ratios of male and female population also reveals differences in language maintenance among Moldavians and Gypsies, but these differences are due to special restrictions in the dynamics of this ethnosociological variable. Significant differential structures in connection with the sex distinction only occur among the Moldavians, with varying dominances of language maintenance for males and females in different regions of settlement. Among the Gypsies there is only a minimal variation of rates for language maintenance by sex group, and the variable is of negligible significance. In comparing rates of language maintenance, differences of up to 1% or 1.5% are considered insignificant. It is immaterial for this discussion what special conditions, in the case of the Gypsies, caused the blocking of the working of this variable (sex distinction as an ecological factor). I only mention it to illustrate a case of insignificance of an ecological variable. The tentativeness of the factor "sex distinction" is indicated by the fact that variations in language behavior result from it among Moldavians, but not among Gypsies. When the whole network of ecological variables is applied to concrete settings, one has to keep in mind the above described phenomenon of "potentiality" of influence. A priori judgments as to which variables may or may not reveal a significant influence are not possible and any attempt in that direction would distort the results. Such judgments can only be made after application of the variables to a concrete setting, and then any statement as to the potential influence of a single variable will be

Language and ethnicity in a network of ecological relations

21

Tab. 2. Language maintenance of Moldavian among the Moldavian urban and rural population in different regions of settlement, (see Haarmann 1979 b, 96)

Rural population

%

Urban population

Regions of settlement

Language maintenance Urban population Rural population

(1) Moldavians in the Ukrainian S.S.R.

57.1%

92.9%

(2) Moldavians in the Russian S.F.S.R.

68.6%

77.4%

(3) Moldavians in the Kazakh S.S.R.

64.8%

66.9%

(4) Moldavians in the Georgian S.S.R.

71.9%

74.4%

(5) Moldavians in the Azerbaydzhan S.S.R.

62.3%

85.0%

(6) Moldavians in the Belorussian S.S.R.

54.5%

63.9%

22

Language in ethnicity: a view of basic ecological relations

Tab. 3. Language maintenance of Moldavian among the Moldavian male and female population in different regions of settlement, (see Haarmann 1979b, 97)

Female population

%

Male population

Regions of settlement

Language maintenance Male population Female population

(1) Moldavians in the Ukrainian S.S.R.

81.5%

85.1%

(2) Moldavians in the Russian S.F.S.R.

76.2%

59.4%

(3) Moldavians in the Kazakh S.S.R.

68.9%

61.2%

(4) Moldavians in the Georgian S.S.R.

77.6%

66.1%

(5) Moldavians in the Azerbaydzhan S.S.R.

68.5%

56.0%

(6) Moldavians in the Belorussian S.S.R.

60.7%

46.3%

Language and ethnicity in a network of ecological relations

23

Tab. 4. Language maintenance of Romany among Gypsy urban and rural population in different regions of settlement, (see Haarmann 1979c, 78) Rural population

%

Urban population

Regions of settlement

Language maintenance Urban population Rural population

(1) Gypsies in the Russian S.F.S.R.

73.8%

78.9%

(2) Gypsies in the Ukrainian S.S.R.

55.6%

45.5%

(3) Gypsies in the Moldavian S.S.R.

89.2%

67.0%

(4) Gypsies in the Belorussian S.S.R.

78.1%

72.9%

(5) Gypsies in the Lithuanian S.S.R.

88.2%

97.3%

(6) Gypsies in the Latvian S.S.R.

70.8%

76.8%

(7) Gypsies in the Kazakh S.S.R.

84.6%

88.1%

(8) Gypsies in the Uzbek S.S.R.

52.6%

45.9%

24

Language in ethnicity: a view of basic ecological relations

Tab. 5. Language maintenance of Romany among Gypsy male and female population in different regions ot settlement, (see Haarmann 1979c, 79) Female population

%

Male population

Regions of settlement

Language maintenance Male population Female population

(1) Gypsies in the Russian S.F.S.R.

75.7%

77.1%

(2) Gypsies in the Ukrainian S.S.R.

52.1%

53.2%

(3) Gypsies in the Moldavian S.S.R.

79.1%

79.0%

(4) Gypsies in the Belorussian S.S.R.

76.7%

77.6%

(5) Gypsies in the Lithuanian S.S.R.

89.4%

90.5%

(6) Gypsies in the Latvian S.S.R.

71.7%

73.4%

(7) Gypsies in the Kazakh S.S.R.

85.2%

86.3%

(8) Gypsies in the Uzbek S.S.R.

49.2%

49.2%

Language and ethnicity in a network of ecological relations

25

valid only for that specific setting. In my experience, the blocking of the sex distinction variable in the case of the Gypsies may be considered to be an exception. Normally, variant structures of language behavior do result from the working of this factor (e. g., the strong observed influence of sex distinctions in the Moldavian case). The tentative nature of the influence of ecological variables is not only important for empirical analyses (/. e., when applying ecological variables of the inventory to concrete settings) but is also relevant for theoretical aspects of interrelations between single variables and variable clusters. Much empirical work has to be carried out on the basis of what must be remembered are a provisional inventory and a provisional basic model of ecological relations in order to find true regularities in the dynamic interrelations of factors. Based on broad empirical data (i.e., after applying the ecological model to a variety of concrete settings), it may be possible to characterize the potential influence of variables in terms of their regular or irregular interrelations. But for the time being, the data needed for an empirically based ecological theory is not yet available. Presently, attempts to use the ecological model for sociolinguistic typologies (see chapter 3) tend to focus on characterizing regularities in the dynamics of single ecological variables. It has been pointed out that, for the moment, only provisional statements can be made concerning the dynamics of single ecological variables. And this is all the more true for statements concerning the interdependent working of several factors (including variable clusters) influencing ethnic groups and their languages (see chapter 4). The identification of single variables as relevant group- or language-influencing factors, and their integration in an inventory are dependent on illustrating how the variables fit into a network of ecological functions (i.e., into an ecological system). The list I have proposed of components for a model of ecological relations, representing a tentative overall ecological system (or system of language ecology), is not exhaustive. I am not presenting here a complete ecological theory. The model and its components represent instead a provisional framework or orientation intended to illustrate what I assume to be interdependent, influencing factors. But although this model is provisional and may undergo modifications in the course of application to real ecological settings, such a model is needed in order to show the links between it and special models of ecological relations derived from it. The central point or focus of the basic model is the ethnic group of reference, that group to the behavior of which the ecological variables

26

Language in ethnicity: a view of basic ecological relations

refer (n.b., RG = reference group or ethnic group of reference, for which I used the abbreviation BE which corresponds to "Bezugsethnie" in my German text; see Haarmann 1980b, 201). The placing of the reference group within the network of ethnopsychological variables (see functional range 5 of ecological variables) is based on the assumption that the factors within this range of functions play a special role as a filter in relations within the reference group and between it and other ethnic groups. It is assumed that any interaction of the reference group is controlled by self-categorization {i.e., subjective self-identification), categorization and evaluation of the contact groups (CG in the model, corresponding to KE "Kontaktethnie" in the German terminology), as well as by other ethnopsychological factors. This means that the group of reference never interacts with another ethnic group without preconditions but, rather, is ready or able to communicate only in terms of its ethnopsychological support system (functional range 5). A special difficulty arises when trying to illustrate the effect on group behavior of ethnopsychological variables because these factors and their potential influence can only be detected indirectly. Categorizations and identifications linked to ethnicity are always implicated, but indirectly, in the dynamics of interethnic group relations. As the ecological relations set forth in the basic model refer to the group of reference as the focus and always reflect its position, the filter function of ethnopsychological variables has been pointed out for this group. It should be noted that the filter function of ethnopsychological variables also is at work on the behavior of a given contact group which itself can at any time be studied as a reference group. The distinction made between reference group and contact group is one that should be considered relevant only from the standpoint of the viewer. The contact between Finnish-speaking and Swedish-speaking groups in the urban area of Helsinki may be analyzed in such a way that the Finland-Swedes are the reference group interacting with the Finns as the contact group. This viewpoint is clear in the study presented by Allardt, Miemois and Starck (1979), which is very useful for its analysis of the filter function of ethnopsychological variables affecting the Swedish-speaking urban population which is its reference group. A similar analysis could have been carried out with the Finland-Swedes in Helsinki as the contact group and the Finns as the reference group. For the puposes of the basic model of ecological relations (cf. Tab. 6) it is therefore not necessary to show the functional range (5) of ethnopsychological factors with respect to the contact group. It is more important to illustrate the general role of eth-

Language and ethnicity in a network of ecological relations

27

Tab. 6. Basic model of ecological relations (or relations of ecological variables with different functional ranges)

Remarks: RG = reference group, ethnic group of reference CG = contact group, ethnic group in contact with the reference group 1 2 3 4 5 6 7

= = = = = = =

ethnodemographic range of ecological functions ethnosociological range of ecological functions ethnopolitical range of ecological functions ethnocultural range of ecological functions ethnopsychological range of ecological functions interactional range of ecological functions ethnolinguistic range of ecological functions

All variable ranges with their corresponding ecological functions refer to the behavior of groups, not individual speakers.

nopsychological variables as they control intergroup relations. And this role is clearly indicated in the diagram. Whereas the effect of ethnopsychological variables is specific for the reference group and specific for the contact group - processes of categorization always result in specific profiles of ethnic identification for single ethnic groups - the interaction of the reference group as well as that of the contact group is determined by ecological variables within the range of interactional factors (see functional range 6). Considering only interactional factors (the network of ecological variables as they affect intergroup interactions in situational settings), the conditions of interaction are equivalent — which is not to say equal - for both the reference and the contact group, different as their roles may be in the processes of

28

Language in ethnicity: a view of basic ecological relations

verbal interaction. Looked at this way, potential contact groups as well as the reference group are inextricably caught in the network of interactional factors, as is indicated by functional range 6. The distribution of ethnodemographic (1), ethnosociological (2), ethnocultural (4) and ethnolinguistic (7) variables illustrates the different concentrations of functional ranges with respect to the groups interacting. The interdependency of the variable ranges is also graphically indicated by connecting the ranges. One can easily identify the special filter function of the interactional variables of range (6), a function which is different in nature from the filter function explained for the ethnopsychological variables of range (5). I have assumed that the effects of ecological variables (1), (2), (4) and (7) are exerted indirectly on the reference group and any contact group through the filter of interactional variables and their dynamic effect. The functional distribution of the languages spoken by the reference group and the contact group in the society (i.e., the profile of diglossia or polyglossia), has a direct effect on the language behavior of both groups because it shapes their interactional relations (functional range 6). Whether or not a reference group in a given society is primarily urban (functional range 1), or their language is an ausbau language (functional range 4), affects their interactional behavior indirectly. The same is true for other variables like the age-group différenciation of the reference group (functional range 2), or the specificity of language contacts (functional range 7). The functional range of ethnopolitical factors (3) is best understood as an embracing category comprising all other functional ranges. This indication of an outer framework of ecological relations implies that political factors influencing the behavior of ethnic groups form a general background for all other societal and intergroup relations. This has been pointed out already in connection with the establishment of an inventory of ecological variables. As all social conditions of ethnic groups are bound to the political organization of the society in a given state, the components of the political systems must be indicated separately. It is assumed that the effect of ethnodemographic, ethnocultural or other factors on the behavior of reference and contact groups can only be represented in an overall ecological system when these have been integrated into the general framework of a society's political foundation (or organization). The above discussed basic model of ecological relations, not only because of the internal correlations it stipulates but also because of the inclusion of both ethnicity and language related variables (see 4), should

Language and ethnicity in a network of ecological relations

29

be considered as a model for illustrating fundamental relations in the behavior of reference and contact groups. The ecological system represented in the model refers to the shaping of communicational systems (at the level of language knowledge) as well as to the verbal interaction of ethnic groups (at the level of language use). This fundamental dichotomy has been discussed in Haarmann (1980 a, 34ff.). It is because of this that the above model of ecological relations differs from most other tentative models which mainly refer to the behavior of groups with respect to their interaction. One such model is that proposed by Cooper (1969,202-3), which shows the special relations of interaction constructs. The relations described in that model could be considered to be a special case of the more general or basic ecological relations represented in the model (cf. Tab. 6) discussed above. A fundamental problem always comes up when one attempts to establish a model, and it is of a methodological nature. Any model, if it is to be widely applicable, has to be abstract. The degree of abstraction depends on the aims of the model. A comprehensive or basic model requires a higher degree of abstraction than a model representing special relations. In my experience, a model is good only insofar as its degree of abstractness is adequate for the application intended. On the one hand, any model has to be at least abstract enough to go beyond a single concrete (ecological) setting. Any model has to be abstract in order to be applicable to various concrete settings. Moreover, a large number of concrete settings must be capable of being adequately illustrated or represented by the model. If the degree of abstractness of a model is too low, it can only be applied to a restricted number of concrete settings. If the degree of abstractness is too high, the subtler distinctions to be found in various concrete settings cannot be adequately represented in the model. The conceptions underlying the constructs (e.g., the relation between RG and CG, the position of ethnopsychological variables with their filter function) which constitute the model are what mainly determines the abstractness of the whole model. In the model proposed here the different functional ranges of ecological variables as well as the indicated interrelations between the sections are theoretical constructs and therefore both determine the abstractness of the model. The ecological relations indicated by graphs in the model are themselves constructs. One of the main aims of a fundamental theory of language ecology should be the setting up of hypotheses about regularities implied by the proposed model. The formulation of hypotheses is also needed in connection with the links between the basic model and special models.

30

Language in ethnicity: a view of basic ecological

relations

Methodology must also be mentioned in relations to the graphic representation of basic ecological relations in the model. The proposed model shows these relations in a two-dimensional diagram. Although this is the method of representation most widely adopted in models constructed for widely differing purposes, it may be too restricted for the special problems and complexities of ecological relations. One could imagine a three-dimensional diagram in which the graphic representation of several interdependencies between functional ranges of ecological variables and the interrelations between ecological factors and the reference group could be more adequately illustrated. Such a project, however, cannot be undertaken until further progress has been made identifying and analyzing those interrelations. Further developments in a general theory of language ecology could therefore make possible or, indeed, necessitate the transformation of the proposed two-dimensional model into a three-dimensional model with room for the more refined strings of interdependent factors that will have been identified. A general or basic model of ecological relations is needed in order to formalize the interrelations of different variables in their association with single functional ranges. The application of basic variables, as they have been discussed as an abstract network of ecological relations to concrete settings, also requires insights about the interaction of factors and functional ranges as they are provided by such a model. A basic model, however, reveals ecological relations in a general framework but does not illustrate specific relations. Such specific relations are those between single variables or between different functional ranges. Although all specific relations are included in the general relations, they are not made explicit in the basic model. For the purpose of making specific relations transparent, specific models of ecological relations are also important. Whereas this is primarily of a theoretical nature, specific models are also practical for empirical research because they provide the investigator with an idea about how the applied ecological variables are actually functioning in the network of ecological relations. Thus it is not only important to apply the set of variables to concrete settings (i. e., by using a detailed questionnaire), but one should also be well aware of the relations between single variables and functional ranges (i. e., by basing one's analysis on specific models of ecological relations). In some cases of specific relations, it is not possible to correlate and evaluate data adequately unless the interdependence between factors has been specified. Such a specific model is presented in Table 7. This model illustrates the basic relations between factors in the ethnopsychological range,

Language and ethnicity in a network of ecological relations

31

Tab. 7. Basic relations between self-identification (subjektive Identifizierung) and the categorization of other ethnic groups (Fremdkategorisierung) (see Haarman 1980 b, 202)

Note: The relations on the right side of the graph form part of the infrastructure of phenomenology. In this specific model, different factors are specified for the value system which forms the background for any kind of interaction between groups of different ethnic origin (e.g., position of the ethnopsychological range of ecological functions in the basic model).

namely between self-identification and the categorization of others. The component indicated as external factors actually implies a variety of ecological factors and thus has to be considered a bundle of variables in the ethnodemographic, ethnosociological, ethnopolitical, ethnocultural, as well as interactional ranges. The specific model is linked with the general model of ecological relations by that component. Further details of this specific model have been discussed in Haarmann (1983 d. 31 ff.). One could consider this specific model a formalization of the basic relations in ethnicity such as stressed by Fishman (1977, 16): "Ethnicity is rightly understood as an aspect of a collectivity's selfrecognition as well as an aspect of its recognition in the eyes of outsiders". The interrelations between different factors as represented in the inventory of variables and integrated in the basic model of ecological relations need further explanation. One way to make progress toward hypotheses about the interdependence of single factors and their potential influence on language behavior is to measure ecological variables. Regularities and irregularities of different factors working together can be best detected when measuring the influence of variables in quantitative terms. The ecological relations in the basic model have been established

32

Language in ethnicity: a view of basic ecological

relations

on the basis of provisional hypotheses which have to be elaborated. As the provisional hypotheses correspond to a kind of a priori decision "to collect information from empirical investigations dealing with similar or related problems in order to support or negate a concept" (Kreppner 1975, 22), they may therefore be transferred to exact hypotheses by measuring ecological relations. The popularity of quantitative methods in all areas of the social sciences results from the need for objectivity in data analysis and in the evaluation of information. Applying quantitative methods - already proposed in the field of linguistics (especially in the field of general typology) by Gabelentz (1901) - provides the possibility to supervise theoretical concepts by means of quantitative concepts. The concepts used in the inventory of ecological variables are of a qualitative nature. Measuring as the "empirical process to determine a quantitative unit" (Stegmüller 1970, 46) presupposes metricalization as the "introduction of a quantitative concept for a range of objects" (Stegmüller 1970, 46). The main presupposition underlying measuring is therefore the transfer of qualitative and quantitative concepts. As a means of linking qualitative and quantitative concepts, comparative concepts have to be utilized, thus extending the simple polarity of qualitative versus quantitative concepts (see Haarmann 1980a, 119ff.). With respect to the general aim, namely establishing hypotheses about the interrelations between single variables and variable clusters, a quantitative-oriented analysis is quite feasible. In this connection, it should be pointed out that an overall quantification may meet with difficulties because of the non-quantitative nature of many concepts (e. g., the functional range 5 of ecological variables). Thus, ideal results cannot necessarily be expected to result from measuring. There are restrictions (especially in range 5) such that several qualitative concepts may not be easily converted into quantitative concepts, or such that the qualitative findings in connection with the influence of a variable may reveal to be less practical in quantitative terms. On the other hand, quantitative approaches certainly allow for greater accuracy of the findings in several sections (e.g., functional ranges such as 1, 2, 3, and others). Even if an overall quantification of all ecological relations (of the entire ecological network illustrated here) proves to be impossible, the conversion of at least some sections of functional ranges is of value. So, with due reservations regarding quantification and its feasibility, measuring the influence of variables is, nevertheless, a relevant step toward a basic theory of language ecology.

Language and ethnicity in a network of ecological relations

33

One method to convert qualitative concepts into comparative concepts and to metricalize these by introducing quantitative concepts shall be presented here. It should be pointed out from the start that the proposed network of comparative and quantitative concepts is a provisional one which will undergo refinement. Such a refinement of categories involving a further différenciation of quantitative concepts may be required when applying this network of variables to a variety of concrete settings. This refinement may be revealed to be necessary in case of single variables as well as of variable clusters (i. e., total functional ranges of ecological variables). But such a refinement can only be adequate as an a posteriori means on the basis of broad experience. With respect to the conversion of comparative to quantitative concepts for variable (1.1), I used a more refined scale of metrical units in Haarmann (1983 a, 31) than in Haarmann (1980b, 220), as the special subjects of investigation in the former required a network of quantitative concepts that was more differenciated than was necessary for the more general purposes of the latter. Earlier I outlined variations for some ecological factors in order to illustrate how the dynamics of single factors may vary in concrete settings. The variations associated with concrete settings may be specified as (1), (2), (3), etc., thus providing the investigator applying the inventory of variables a means to characterize any group in terms of a numerical code. A n d it is upon such a numerical code, with the value of a formalized description of sociocultural conditions, that the measurement of ecological relations is based. T h e association of single speech communities and their ecological setting with a numerical unit is a matter of correlating a theoretical construct (metricalized variations of ecological factors) and real conditions. This correlation, however, must not be misunderstood as an absolute equation. A concrete setting is related to a theoretical construct under the condition that most features represented by the construct are found in the setting. For the purposes of a general outline of measurement, comparative concepts such as " a v e r a g e " , "higher or lower degree", and others used in the description of single settings are converted into quantitative concepts which constitute a scale ranging from three to eight units for each ecological variable. In theory, any given ethnic group may be specified by applying the proposed scale of quantitative units to its sociocultural conditions and language behavior (thus including language preferences, maintenance, assimilation, second language choice, etc.). The language behavior of any given speech community may thus be characterized in

34

Language in ethnicity: a view of basic ecological

relations

the form of a numerical code with respect to different functional ranges of ecological variables. As there are many preliminary studies yet to be undertaken which would provide material for the whole range of comparative concepts, in referring to single groups an overall measurement of ecological settings for a variety of different speech communities is not possible at this time. One of the aims of a long-term project on small languages and speech communities in Europe (including the Soviet Union) is the elaboration of data for different functional ranges of ecological variables, in order to make the basic model illustrated above applicable and ecological relations measurable for a greater number of speech communities (see Haarmann 1979a, 1983a, 1983b). Any ethnic group in any concrete contact setting may be described on the basis of the framework of quantitative concepts elaborated above. The status and language behavior of a given speech community can thus be analyzed in terms of a specific numerical code. As the inventory of ecological variables covers a broad panorama of ecological relations, it is most likely that the numerical code identified for different ethnic groups will prove to be ethnically specific and unique for any single community. A future plan to apply the inventory of quantitative concepts to a variety of European speech communities, as part of a long-term research program, will make possible a general assessment of similarities and divergences. Comparisons may pertain to the speech communities as a whole, or to specific ecological relations (comparing, for instance, single functional ranges of variables). Such large-scale comparison will also make possible conclusions concerning different patterns of language behavior. The inventory of quantitative concepts may also be used for purposes other than to merely describe speech communities in terms of ecological relations. One aim will be to use the indices for the development of typologies. As the inventory is quite complex, and since up until now no single speech community has been totally analyzed in ecological terms, any typology in connection with patterns of language behavior remains within the framework of a priori assumptions about specific features. The assumed features of certain patterns may well be proved accurate by later empirical findings, or they may require adjustments in order to characterize single patterns more precisely. A very effective method of establishing typologies is to choose ideal features in order to characterize a pattern (or type). In Haarmann (1980a. 119ff.) I illustrated the use of ideal typologies within the field of sociolinguistic studies. A typology generally differs from any actual classification because it is more complex and flexible and, also, provides more information about the features

Language and ethnicity in a network of ecological relations

35

of particular objects. In chapter 3,1 elaborate a set of ideal features for one pattern of language behavior (i. e., language shift as a consequence of assimilation pressure). This special type of setting is illustrated by concrete contact settings for which external conditions (environmental relations of ethnic groups in contact) as well as internal conditions (influences in the linguistic structures of contacting languages) are specified. In that connection, the characterization of the pattern "language shift in progress" concentrates on the typical features, and as there is only space in this study to discuss a single case of a concrete setting, I leave out the numerical code which could easily be associated with the variations of ecological factors attributed to that pattern. What is referred to here as types of language behavior is an expression referring to typological principles and their application in the field of language maintenance, assimilation, second language attraction, etc. As any typology aims at setting up stable relations between the conditions of real settings and the corresponding theoretical constructs in terms of variables in the inventory of ecological factors, any given formal approach to characterize a type of language behavior has to be based on a priori assumptions about such relations. After the measurement of a great number of ecological variables in different functional ranges has been carried out, the a priori assumptions may be transferred to exact measurements on the basis of a posteriori evaluations of data received in empirical studies. A priori assumptions for a formal characterization of assimilation, or any other phenomenon of language behavior, also require the use of quantitative concepts as already outlined.

2. The role of language in processes of ethnic fusion and fission

During the course of the 70's researchers in the field of ethnopolitical and sociological studies as well as social psychology began to pay more attention to problems of ethnicity (ethnic identity), resulting in a large and increasing body of literature to deal with. And since definitions have been offered to specify the term "ethnicity" from various standpoints (see Isajiw 1974, and Liebkind 1984, 42 ff. for a wide range of definitional approaches), there seems to be only one point where opinion converges. It is generally accepted that ethnicity is a phenomenon that cannot be characterized in any single feature but must be described as a cluster of features which are different in kind and which include language oriented as well as non-language oriented criteria. When discussing single features, isolating them for the purpose of definition, one may feel discouraged by the lack of agreement about how to identify features of ethnicity and how to establish priorities in the evaluation of single criteria. A successful attempt to clarify the relations between different features of ethnic identity (especially with respect to the role of language in ethnicity) and pinpoint the problems of evaluation was presented by Fishman (1977) whose concept of ethnicity is based upon a threefold différenciation of criteria. Of all the components which shape ethnicity, namely (and these are Fishman's terms) (1) paternity (the ethnically oriented ancestry of an individual), (2) patrimony (the distinctive cultural pattern or ethnocultural pattern of education), and (3) phenomenology (the selfidentification of individuals and groups), Fishmann (1977, 17) calls paternity the "key element in ethnicity". Allardt (1979) has recently opposed giving paternity such weight in the process of ethnic identification. According to Allardt (1979, 31), who refers to paternity as descent, this "is not a necessary condition for individual membership although it is an elementary criterion". In this connection Allardt, stresses that ethnicity is subject to fluctuations, thus following a recent tendency to view ethnic identity not as a stable unchangeable cluster of features but, rather, as a

38

Language in ethnicity: a view of basic ecological relations

variable which depends on changing conditions which affect the relative importance of identification and categorization of others. The use of terms like paternity and patrimony is crucial for the weighting of such features in the shaping of ethnicity, but these expressions connote overtones which might arouse criticism (cf. paternity versus maternity, patrimony versus "matrimony"). Elements of neutral terminology such as ancestry or distinctive cultural pattern shall be predominantly used in this study. In Haarmann (1983 d), in which I concentrated on identifying relevant components in the ethnic identity of ethnic groups as whole communities, I pointed out that in reality ethnicity can be observed to be a cluster of group-specific dominant single features or components. This means that the ethnic identity of a given community (as represented by its members) is shaped by a variety of features, some dominant and others non-dominant, that form a network which differs from one ethnic group to another. I rejected an a priori evaluation of given single features as dominant or non-dominant and argued for an a posteriori evaluation (on the basis of concrete analysis of ethnic identity as it is manifested in groups). In this connection it must also be stressed that language - as it is empirically found in concrete settings — is not a necessary criterion of ethnicity and therefore does not play a fundamental role in all processes of ethnic fusion and fission (cf. examples given under the typology of ethnic boundaries in 1.3., 2.1. and 2.3.). When facing the wide variety of ethnic identities existing in the world, one becomes aware of the importance of dividing lines between single ethnic groups and their corresponding ethnocultural patterns. On the one hand, ethnic boundaries can be considered as a phenomenon resulting from a community (or reference group) singling itself out and thereby giving shape to its ethnic identity. On the other hand, (regarding the problem from the standpoint of interethnic relations), the shape of ethnic identity often appears in the form of elements stressing mutual separation, creating "distance" between ethnic groups. These are two aspects of the one and the same phenomenon, but they emphasize contrasting points of view, and opinions based on these views are not necessarily in opposition. The components of self-identification and categorization of others (phenomenology), which comprise evaluations and stereotypes about one's own ethnic group and other contacting groups, can be considered as bricks, that is to say, the building blocks of which ethnic walls (boundaries) are built. All these factors which are related to evaluations about oneself and about others in the mentality of people who belong to

The role of language in processes of ethnic fusion and fission

39

the ethnic group could be grouped together as a system supporting identity. Most scholars would agree on that the dividing lines between ethnic groups are either caused by or themselves produce a social and psychological support system. "Attitudes toward other ethnic groups, along with pride and self-awareness, play significant roles in each nationality's collective support system. These regulatory support factors, especially in the defense manifestations — group intolerance, mistrust, indifference, and prejudice — often greatly strain relations between ethnic groups" (Kazlas 1977, 232).

The view that attitudes toward other ethnic groups as well as features of self-categorization shape the ethnic identity of an ethnic group is supported by the assumption that evaluations of the group's own role and that of other groups in interethnic contacts are elementary factors in all interethnic relations: " . . . , ethnic groups are formed to the extent that actors use ethnic identities to categorize themselves and others for the purpose of interaction" (Allardt 1979, 28). The sources of ethnic conflict are defensive elements of the support system such as intolerance, mistrust, etc. Wherever and whenever relations between ethnic groups are strained, this is due to the dominance in a group's support system of defensive manifestations which tend to isolate the group from others. Ethnic boundaries between a given ethnic group of reference and other ethnic groups grow stronger in those settings where the ethnic profile of the reference group is characterized by negative attitudes toward contacting groups. Ethnic boundaries are the consequence of differently shaped ethnic identities in different groups, and these boundaries are not fixed but deeply dependent on the attitudes and activities of ethnic groups, which may serve to weaken or strengthen them. Objective features like differences in language, cultural heritage (including religion) or ancestry (including racial differences) mark ethnic boundaries. However, it is the perception of such features and the judgment about such boundary markers among the members of an ethnic group which are decisive for the shaping of attitudes between a given group of reference and contact groups. Boundaries, thus, may be perceived as positive {e.g., pride in one's language and culture) or they may be experienced as negative and passively suffered (e. g., fading of self-awareness among minorities under extreme pressure of assimilation). This circumstance, namely the fact that objective features are perceived by the members of an ethnic group to form their value system, explains the somehow hybrid nature of ethnic boundaries. On the one hand, they are the cause of ethnic

40

Language in ethnicity: a view of basic ecological relations

identification, on the other hand, they may function as the effect of ethnic groups' attitudes toward other groups. It would be too negative a definition of Catalan ethnicity to argue that the existence of the Catalans as an ethnic group can only be explained by the ethnic boundaries separating them from the Castilians and other ethnic groups (Basques, Galicians, etc.) in Spain. The fact that the Catalans are different from the rest of the Spanish population should be only a secondary consideration to the student of ethnicity. More decisive are the elements of identity and group solidarity in the forming of the Catalan community (/. e., consciousness of the value of their own language, awareness of the prestige of the old cultural tradition, striving for cultural and political autonomy). The awareness of a long cultural tradition (which in the case of the Catalans dates back to the Middle Ages) as an element of pride is also strong among other ethnic groups such as the Occitans, Irish, Welsh, Armenians, Ukrainians, et al. In Haarmann (1983 d, point 7), I refer to this special feature as the "historische Dimension der ethnischen Identität" (the historical dimension of ethnic identity). The cohesiveness of an ethnic group's identity may also be the function of the level of threat from outside (e.g., political discrimination). Ethnic boundaries, however, as a nexus of various features are an important subject in the study of ethnopolitical conflicts. In Europe, where the existence of ethnic groups is almost exclusively marked by language distinctions (n.b. the use of the term "Sprachnationen" (language nations) by Kloss (1969 a), the special relations between ethnicity and language are of vital importance in the analysis of ethnicity problems. Allardt (1979) therefore pays special attention to the role of language in the movements of "ethnic revival" among ethnic minorities in Western Europe. The problem of categorizing ethnic boundaries or markers of ethnic boundaries has been investigated from different perspectives. In this connection, special attempts have been made to classify variants of boundary crossing (/. e., the implications of crossing ethnic boundaries through assimilation and other processes). The classification presented by Horowitz (1975) can be considered a big step in this direction. Horowitz (1975, 116) proposed a categorization of ethnic boundaries based on the different conditions under which they emerge. In other words, he tried to categorize the processes of ethnic fusion and fission, incorporating the element of "time" to indicate the process of evolution. The classification presented by Horowitz, however, is not comprehensive and

The role of language in processes of ethnic fusion and fission

41

therefore does not include all the possible types of ethnic fusion and fission. This may be due to his limiting the discussion to interethnic relations and evolutionary processes of ethnic boundaries in India, or to conceptual weaknesses which become apparent when applying his system to concrete settings. Based on the assumption that ethnicity is a dynamic multifaceted phenomenon and is subject to potential changes (either in the course of a generation or over longer periods of time), I would like to set up a comprehensive categorization for ethnic fusion and fission, taking into account the conditions of ethnic boundaries as they vary diachronically. An extensive overview of the potential changes in ethnic identity and the evolutionary processes of ethnic boundaries would have to include the following variants: Principal changes

Variants of changes

1. Assimilation or integration (fusion)

1.1. Amalgamation 1.2. Incorporation 1.3. Conglomeration

2. Différenciation or segregation (fission)

2.1. Profilation 2.2. Separation 2.3. Proliferation

Abstraction(s) A + Β> C A + Β> A A + Β > A + Ba A A A (A

> > > +

Ai + A 2 ( + A 3 . . . ) Β+ C A + Β B > A + B + C)

Amalgamation, incorporation and conglomeration are phenomena of ethnic fusion (indicating processes of assimilation or integration) whereas profilation, separation and proliferation are variants of ethnic fission (indicating processes of différenciation or segregation). The use of alphabetical formulas to indicate what is essentially a theoretical construct for the purpose of abstraction proves to be practical when discussing concrete settings. At the same time, these formulas offer a basic for formalization of statements about boundary profiles. The discussion of examples in the following passages mainly concentrates on processes of ethnic fusion and fission in Europe, but settings outside Europe are also used for illustration. The varieties of ethnic boundaries and the processes of their emergence may be classified and characterized by the above presented scheme in the following way.

42

Language in ethnicity: a view of basic ecological

relations

1.1. Amalgamation The process of assimilation (fusion process) can be characterized as a amalgamation when neither of the ethnic groups involved in a fusion process is dominant with respect to culture or language, and when ethnicity in the new formation represents an amalgam of elements of the former independent ethnic identities. The emergence of Basque ethnic identity (C) illustrates this kind of fusion. Although the exact evolutionary stages of this assimilation, which had already begun in Roman times and was completed in the early Middle Ages, cannot be traced because of the lack of documentary evidence and the scarcity of information available about the pre-Roman languages, it is widely assumed that the formation of the Basque people resulted from the fusion of two antique ethnic groups. According to this hypothesis, partly romanized Iberians (A) from the upper valley of the river Ebro fused with people of Cantabrian stock (B). Neither the Iberian nor the Cantabrian language is the main component of the Basque language but both have largely contributed to the formation of this element in the Basque distinctive cultural pattern. It is most probable that the cultures of both the Iberians and the Cantabrians have been partly integrated into the new ethnic identity without a dominance of either. The acculturation of the Iberians (under Roman influence) is well documented by the many lexical borrowings from Latin which still survive in the Basque language and which constitute a broad layer in the Basque vocabulary (cf. Haarmann 1979d, 131 ff.). The Latin loan-words in Basque also include a number of archaic elements which did not survive in any of the Romance languages. The fact that there had not only been a cultural exchange among Iberians and Romans but that the Iberians were in the process of acculturation before fusing with the Cantabrians is indicated by the adoption of borrowings in the lexical field of parts of the body. The integration of Latin loan-words into this special section of the Basque vocabulary is illustrated in the following list. It seems reasonable to interpret the evolution of English ethnicity (C) also in terms of a fusion process where Saxons (A) and Angles (B) amalgamated to form a new ethnic profile. In the initial phase (raids of the Germanic tribes on the coastal areas of Roman Britain in the 4th century A. D., mass immigration and conquest of Celtic Britain during the course of the 5th and 6th centuries), Saxons and Angles represented independent ethnic groups. Their areas of settlement in Medieval Britain were initially separated in the common manner of tribal segregation. But

The role of language in processes of ethnic fusion and fission

Basque expression (borrowing)

Meaning of the Basque expression

Latin source of borrowings

boronte kikirista kukula, kukulu mazela, masailla mizpira kaiku gorputz khoro fuñ golko, kolko mamul, manbul musu, mosu tekadi makila, makhila eskurr pantika, mandika maskulu magina

'forehead' 'crest of birds' 'crest of cock' 'cheek' 'nostril' 'having a big nose' 'body' 'spine' 'inner part of bones' 'breast' 'female breast' 'face' 'finger' 'long thin leg' 'gland' 'stomach' 'bladder' 'vagina'

FRONTE (M) CRISTA CUCULLU MAXILLA MESPILA CAUCU CORPUS CORONA FÜNE(M) COL(A)PU/*COL(A)CU MAMMULA •MUSU DIGITU BACILLA (A)ESCULU PANTICE(M)/*PANTICA(M) VASCULU VAGINA

43

whereas the boundaries between Saxons and Angles initially were quite distinct they soon dissolved in the fusion process that followed the military invasion and settlement in England. In a general comparison, examples of amalgamation represent a special, rather rare variant of ethnic fusion. Incorporation, which will be illustrated next, is in all parts of the world a much more common variant of assimilation (of massive language contact). This is due to the fact that there is usually a dominant group in interethnic relations. Thus only a limited number of concrete settings show a balance in the cultural impact of contacting groups.

1.2. Incorporation Most processes of assimilation in the European framework of comparison have to be characterized as processes of incorporation. This is true for the assimilation of ethnic groups already completed in earlier phases of European history as well as for assimilation in progress nowadays. The acculturation and subsequent assimilation of the Slavs (B) who

44

Language in ethnicity: a view of basic ecological

relations

settled in the coastal areas of the Baltic Sea between the Elbe and Oder rivers, into German language and culture (A) is such a process of incorporation which was essentially completed in the course of the late Middle Ages. Although some isolated settlements of Slavs did survive until the beginning of the 19th century (e.g., the Polabs in the area of Dannenberg, now West Germany), the Sorbs in Lusatia (Lausitz) of all the Slavs in the Eastern part of the German territory (to be identified with East Germany in this context) are the only ethnic group which has preserved its identity. Other historically completed fusion processes that could be cited are, for instance, the total assimilation of Finno-Ugric tribes like the Merians and Muroms into Russian culture, the transformation of the Ugric Bashkirs into an ethnic group with Turkic identity, the romanization of the Western Goths in Spain and the Franks in Northern France, the latinization of the Etruscans, etc. There have been so many incorporations in the course of European history that there is little point in trying to name them all. Many individual cases of ethnic fusion remain shrouded in the past because, although we know something of them as they were recorded by historians (e.g., the assimilation of Celtic tribes in Central Europe, in the Alps and in Balkan countries), we lack any direct evidence. Hundreds of ethnic fusion processes have taken place since ancient times in different parts of Europe. Although ethnic fusion processes completed in the past can only be partly reconstructed, it is fair to assume that the general sociocultural conditions leading to incorporation may be discovered by processes of assimilation taking place today among modern ethnic groups. I have paid special attention to the incorporation now in progress in a number of mini-speech communities (e.g., Izhora or Izhorians, Vots, Livs, Veps, and Lapps on the Kola peninsula in the USSR, Frisians in the district of Saterland in Germany, Celts on the Isle of Man, et al.) in a long term research project. One case of incorporation, namely the assimilation processes among the Izhora (B) under conditions of massive contact with Russian language and culture (A), has been thoroughly investigated in Haarmann (1983 b, 25 ff.). Processes of incorporation are closely linked to the degree of dominance of a contact language which exerts a situational pressure on the mother tongue of a smaller ethnic group. The dominating language in a process of total assimilation tends to take over every communicative function of the language which is under pressure (e. g., the function of a language as a means of communication at the working place, at school, in the neighborhood, at home). Once the dominant language has absorbed

The role of language in processes of ethnic fusion and fission

45

all the functions of the smaller language, the ethnic group under pressure tends not to reproduce its original mother tongue in the course of generation change.

1.3. Conglomeration This term denotes a special phenomenon of assimilation which can be described as a partial fusion of features in one ethnic identity (B) with those of another ethnic identity (A). The result of such a part fusion process has to be understood as a shift of an original B-identity to a Ba-identity. Partial fusion in this connection does not refer to the ethnic group as a community, implying that only parts of the community shift to another identity. The term partial fusion reflects a shift in the identity features of the whole community. Examples of such assimilation, which Horowitz (see above) has not taken into consideration at all, clearly illustrate that ethnicity is a multidimensional construct and that only some in a panorama of features may be involved in a given fusion process. Thus, in partial fusion, like conglomeration, language as a feature of ethnic identity may be dislocated from other markers of ethnicity. This is especially true when studying problems of ethnic identity among the Jews, where many cases of secondary national affiliation may be observed. Ethnic diversity among the Jewish communities can best be described by making use of the term conglomeration. It is undoubtedly true that Yiddish has to be considered the national language of the Ashkenazic Jewish communities, but in the case of language relations among the Oriental Jews the conditions of ethnic identity are much more complicated. The Oriental Jews in the Caucasus speak Georgian (Grusinian) in Grusinia, Armenian in Armenia (as well as in the northwest of Iran), and Tat in Dagestan (where they are known as mountain Jews) as their mother tongues. The Oriental Jews have largely assimilated into the surrounding languages of other ethnic groups. The process of partial fusion in Oriental Jewish communities includes language as a feature of acculturation and assimilation but it does not involve other features like family-transmitted cultural tradition, religious practices, etc. This dislocation of language from other ethnic features that can be observed in the Caucasus is also true for Oriental Jews in Soviet Central Asia (who speak Tadzhik and Uzbek as their mother tongues), in Iran (where most of the Jews speak Farsi), and in the Arab countries (where Jews have assimilated the local variants of Arabic).

46

Language in ethnicity: a view of basic ecological

relations

Although integrated into the surrounding speech community from the standpoint of the language, Jewish identity is nevertheless clearly marked among the Jewish communities by the preservation of Jewish religious practices and the related cultural traditions. By their habits of living, family relations, religious ceremonies, evaluations and attitudes, the Jews who speak Armenian as their mother tongue are clearly distinguished from Armenian Christians. Jews speaking Uzbek lead a life different from that of Uzbek muslims. The special cultural shape of the features of Jewish identity is also reflected in their language. There are a number of lexical and syntactical features in each of the foreign mother tongues adopted by the Oriental Jews that make them real Jewish languages. It is therefore possible to distinguish a special Jewish variant of Tadzhik (Judeo-Tadzhik). Similar observations are true for other variants such as Judeo-Grusinian, Judeo-Armenian, Judeo-Tat, etc. (cf. Rabin 1981 for the role of Jewish languages in Jewish ethnic identity). Even more complicated are the conditions of contact and ethnic identity among the Sephardic Jews who had already completely assimilated the Romance languages in Spain and Portugal before their expulsion at the end of the 15th century. Since the 16th century, the Spanish and Portuguese languages spoken by these Jews have been exposed to various influences from the languages of new homelands (Jews from Portugal mostly fled to the Netherlands, Jews from Spain emigrated to southeastern Europe. The Sephardic Jews (Ba) in Balkan countries like Yugoslavia, Greece, Bulgaria and Rumania are as Jews (B) distinct from surrounding Christians (C) and Muslims (D), but also distinct from speakers of the Spanish language in Spain (A) whose language they had adopted earlier (this language is indicated by the component a in symbol Ba). Manifold local influences can be traced as Serbo-Croatian, Turkish, Rumanian and other strata of loan-words in the vocabulary of Ladino (Judeo-Spanish). The genocide policies of the Hitlerists during the Nazi occupation of the Balkan countries during the Second World War also strongly affected the Sephardic Jews. Most Ladino-speakers now live in Turkey or Israel but their language has tended to vanish in all communities (see Harris 1982 a for discussing of the reasons for the decline of Judezmo). Another case of complex contact conditions can be given here. As complicated as the conglomeration process among the Sephardic Jews is the partial fusion among the so-called Urum or Greeks of Mariupol' (today Zdanov in the Donee district on the northern coast of the Black Sea in Ukrainia). This Greek population had formerly lived on the Krim pe-

The role of language in processes of ethnic fusion and fission

47

ninsula during the rule of Crimean Khanat and had assimilated the Turkish language. The Turkish-speaking Greek Christians were evacuated from the Krim to Mariupol' in the second half of the 18th century under the rule of Catherine II (called Catherine the Great) of Russia. The Urum kept their faith and preserved their Turkish language in the new surroundings. This fact is striking because the Urum have lived since the 18 th century in the same neighborhood as other Greek Christians who have maintained their Greek mother tongue. There was no secondary fusion process (the Urum did not assimilate the Greek language, which was the mother tongue of their remote ancestors). Although the Urum speak Turkish (Ba) they are clearly distinct from Turkish-speaking Muslims (A) and, as Christians, have much in common (namely their ethnocultural tradition) with the Greek-speaking Greeks (B) of their neighborhood. The Turkish language (a) of the Urum (B) has been strongly influenced, however, by Ukrainian and Russian in the new settlements in and around Mariupol'. The process of assimilation as conglomeration, or a partial fusion, among Oriental and Sephardic Jews as well as among the Urum dates back to earlier times and can hardly be reconstructed. There are other examples, however, in which the partial fusion process can be better observed. The Coptic Christians in Egypt, for instance, have preserved their faith and related sociocultural patterns as features of their ethnic identity (B) up to the present. Beginning in the early Middle Ages, however, the Copts gradually assimilated with the Egyptian Arabs. The result of this partial fusion, in the course of which the Hamitic Copts adopted the language of the surrounding Semitic population (A) as the a-component in their Coptic ethnic identity (Ba), was the decline of Coptic as a mother tongue and spoken language. Since the 13th century the Coptic language has been used for sacral purposes only (similar in function to Hebrew as a sacred language among Jewish communities).

2.1. Profilation The term profilation as applied to processes of ethnic fission characterizes phenomena of internal diversification within an ethnic community. Profilation of characteristics of a group's ethnic identity depends on special sociocultural conditions of contact. In the situation of diasporacommunities (e.g., the scattered Jewish or Gypsy settlements), profilation is easily understood as a process of diversification into ethnic sub-

48

Language in ethnicity: a view of basic ecological

relations

groupings. When speaking about Jews or Gypsies one has to specify the different subgroupings that are the consequence of differing specific local contacts. There are, of course, general features of Jewish ethnic identity and of Gypsy ethnicity. At the same time, the awareness that they are internally differentiated, which is found both among Jews and Gypsies, may itself be considered to be a feature of ethnic identity in both of these ethnic communities characterized by diaspora. The dispersion of the Gypsy settlements has resulted in the proliferation of regional variants of ethnic identity. Gypsies in Germany are distinct from those living in Hungary, and the Romany-speaking Gypsies in Rumania are in a different subgrouping than the Gypsies in Latvia or Estonia, etc. There are many subgroupings which do not coincide with tribal distinctions or sociolinguistic différenciations. The distribution of Gypsies in European countries reveals the following profilation of ethnic subgroupings (see Tab. 8). The subgroupings shown on the left side of the table may themselves be subdivided (see right side of the table). This often leads to the somewhat surprising situation in which two Gypsy communities speaking the same second language belong to different subgroupings. In Haarmann (1979 c, 187 ff.) I have discussed the problem of diversification among the Gypsies in Transcarpathia (the western part of Ukrainia) who do not represent one speech community but are divided into several differently bilingual communities. In the following graph (see Tab. 9) I distinguish different levels of affiliation: nationality or a national group comprising the general features of Gypsy identity, a contact group with similar regional conditions of interethnic contact, and speech communities with specific sociocultural patterns as reflected in bilingual structures. The relations between these criteria may be represented in a hierarchical order. The phenomenon of assimilation of surrounding languages among communities of Oriental and Sephardic Jews has already been pointed out (see 1.3.). From another point of view, the term profilation can also be applied to these communities. The diversity of Jewish speech communities does not only mean differences in the language-oriented cultural patterns but also ethnocultural variations of the Jewish religion. Within the framework of secondary national affiliations there are Ashkenazic, Oriental and Sephardic patterns of practice of the Jewish religion and related customs. This means there are variations with respect to the interpretation of religious texts as well as to habits and customs in daily life. The separation of the Occitan speech community in Southern France into regional groups with distinct language varieties and ethno-

The role of language in processes of ethnic fusion and fission Tab. 8. Profilation of ethnic subgroupings among the Gypsies in Europe Subgroup Balkan jerlides

ursári

karamítika romá drindári Wallachian cace1

kalcbrári

lovári

gurbéti Ukrainian sárvi (Left-Bank Ukrainian group) volóxuja (Right-Bank Ukrainian group) Carpathian sórvika romá

Area of settlement

Contact languages

Western Bulgaria Macedonia Southern Serbia Rumania Moldavian S.S.R.

Bulgarian Macedonian Serbo-Croatian

Ukrainian S.S.R. (Crimean peninsula) Central Bulgaria

Rumania Moldavian S.S.R. Rumania Moldavian S.S.R. Ukrainian S.S.R. (Odessa, Transcarpathia) Slovakia Scattered settlements in other countries Rumania Ukrainian S.S.R. Scattered settlements in other countries Yugoslavia (Bosnia, Hercegovina)

Rumanian Moldavian (variant of Rumanian ) Ukrainian Crimean-Tatar Bulgarian Rumanian Moldavian (variant of Rumanian) Rumanian Moldavian Ukrainian Hungarian Slovak

Rumanian Ukrainian

Serbo-Croatian

Ukrainian S.S.R. (East of the Dniepr) Russian S.F.S.R. (adjacent to the Ukrainian S.S.R.) Ukrainian S.S.R. (West of the Dniepr)

Ukrainian

Slovakia (northern and eastern parts)

Slovak

Russian Ukrainian

49

50

Language in ethnicity: a view of basic ecological relations

Tab. 8. Profilation of ethnic subgroupings among the Gypsies in Europe Subgroup

Area of settlement

Contact languages

úngrike romá

Southern Slovakia

Slovak Hungarian Hungarian Ukrainian Hungarian

Northern Hungary Ukrainian S.S.R. (Transcarpathia) Baltic rúska romá2

Northern parts of the Russian S.F.S.R. lotfítka romá Western Latvian S.S.R. Estonian S.S.R. Eastern Estonian S.S.R. lajenge roma (village of Laius; all Gypsies exterminated by Hitlerists during World War II) pólska faldítka Poland (central and southern parts) romá Finnish3 fintilo rómma

German sintí sasítka romá

Welsh volsanánge kalá

Russian Latvian Estonian Estonian

Polish

Finland (western and southern parts) Sweden (some families)

Finnish Swedish Swedish

Germany, Austria, Northern Italy France Poland, Czechoslovakia, Yugoslavia Soviet Union (in the Volga area until 1941, afterwards in the Kazakh S.S.R.)

German Italian French Polish, Czech, Slovak Serbo-Croatian, Slovene German Russian

Wales

Welsh English

Sources: Haarmann 1979c, 30/31, Comrie 1981, 158

The role of language in processes of ethnic fusion and fission

51

Tab. 9. The conceptual-terminological différenciation between national group (national collective community), contact group and speech community illustrated by the variety of Gypsy communities in Transcarpathia (Ukrainian S.S.R.)

(cf. Haarmann 1979c, 200)

The diversity of speech communities among the Gypsies in Transcarpathia is illustrated in the following tables and diagrams. The polarity of first and second language in these communities is represented in terms of ethnostatistical categories. The rates of language maintenance, assimilation and second language attraction specified in the tables are transferred into profiles of bilingualism for each community. Profiles of bilingualism are a graphical means to describe bilingual structures and variations of bilingualism in speech communities on the basis of quantified data (see Haarmann 1980 a for methodical implications, also chapter 4 of this study). The graphs below are taken from Haarmann (1979c, 202-205).

Remarks: (1) cace may be differenciated into: lindurári, zbtári, kakavjári (2) including xoladítka romá (3) linguistically most closely related to the dialect of the lajenge roma (Gypsies of Laius in Estonia)

52

Language in ethnicity: a view of basic ecological relations

Tab. 10. Variety of Gypsy communities in Transcarpathia (Ukrainian S.S.R.) Speech community i: Bilingual Gypsies with Romany as a first language and Ukrainian as a second language

Language

Population total urban

Romany 21.6% Ukrainian 13.5%

28.4% 18.5%

100

rural 7.4% 2.2% 60

30

28,4%

21,6%

total

urban

rural population

Correlation of rates for Romany (1) and Ukrainian (2) Speech community2: Bilingual Gypsies with Romany as a first language and Russian as a second language %

Language Romany Russian

Population total urban 21.6% 28.4% 16.2% 22.1%

100 rural 7.4% 4.1% 60

28,4%

30

total

urban

rural population

Correlation of rates for Romany (1) and Russian (2)

The role of language in processes of ethnic fusion and fission

53

Tab. 11. Variety of Gypsy communities in Transcarpathia (Ukrainian S.S.R.) Speech community3: Bilingual Gypsies with Hungarian as a first language and Ukrainian as a second language

Language

Population total urban

Hungarian 73.5% Ukrainian 13.5%

68.9% 18.5%

100 rural

83,4% 73,5%

83.4% 2.2%

( 1 )

60

30 • (2) 18,5% 13,5% total

2,2%

(2) urban

rural population

Correlation of rates for Hungarian (1) and Ukrainian (2) Speech community4: Bilingual Gypsies with Hungarian as a first language and Russian as a second language

Language

Population total urban

Hungarian 73.5% Russian 16.2%

68.9% 22.1%

% 100

rural

83,4% (1)

73,5%

83.4% 4.1% 60

30

22,1% 16,2% total

urban

4,1%

rural population

Correlation of rates for Hungarian (1) and Russian (2)

54

Language in ethnicity: a view of basic ecological relations

cultural patterns (i. e., Gascons in the Gascogne, Languedoks in the Languedoc, Provenzals in the Provence) could also be characterized as a profilation of the Occitan community. Subgroupings resulting from profilation are also a salient feature of the Lapp (Saamish) community. The subgroupings reflect variations of Lapp ethnocultural patterns as a result of different conditions of interethnic contacts in different parts of Scandinavia. There are Norwegian-, Swedish-, Finnish- and Russian-Lapp ethnocultural varieties. This subgrouping does not follow tribal lines. There are many dividing lines cutting in various directions within the Lapp ethnic group. One finds profilation according to tribal distinctions, differing ethnocultural patterns, and region which may influence such socioeconomic criteria as occupation — some are reindeer breeders, other hunters or fishermen. Nonetheless, the awareness of being a member of the Lapp community and feelings of solidarity for the entire ethnic group are both strong among most Lapps.

2.2. Separation Incorporation has been pointed out as the most frequent type of ethnic fusion (the variant of assimilation occurring most frequently). In the case of ethnic fission, it is separation, which can be considered the most frequent variant of différenciation. In considering language-oriented processes of fission, many historical evolutions of branching in European language groups must be identified as processes of separation (e.g., segregation of Romance, Germanic, Slavonic, Finno-Ugric and other languages). During the evolution of separation, not only differences between languages but also between related ethnocultural patterns emerged. The emergence of new ethnic identities through separation of former ethnic units is a characteristic feature of European history since ancient times. The process of separation is widely known and there is no need in this contribution to further illustrate this phenomenon. Motivations for separation, the initial phase, the interim phases as well as the result of such a process have to be specified for single cases.

2.3. Proliferation Proliferation is a special phenomenon of ethnic fission. Proliferation is the fission process resulting in a new identity of one segregated group, the former community not shifting, but retaining its ethnic identity. This

The role of language in processes of ethnic fusion and fission

55

process is clearly distinct from separation where new ethnic identities emerge while the former is dissolved (i. e., différenciation of Eastern Slavonic tribal communities into Russian, Belorussian, and Ukrainian ethnic identities). Utilizing this concept of proliferation as a basis, phenomena of segregation can be adequately characterized while purely linguistic or political criteria fail to explain such a process. To take an example of the phenomenon, the growing awareness of being distinct and the strengthening of intragroup solidarity among the Corsicans have promoted a strong regional movement which includes a militant wing of separatists. Although an integrated part of French territory, the island of Corsica has close language and cultural ties to the Italian mainland. Corsica was colonized by Italian settlers during the Middle Ages. This tendency toward segregation among the Corsicans is an example of proliferation. The Corsican ethnic identity, (B), has split from the originally common Italian identity, (A). The separateness is equally underscored by the political separatist movement (contra France) as well as by the use of Corsican as a written language with norms distinct from standard Italian (contra the Italian community). Other cases of proliferation include the différenciation of a formerly common Bulgarian ethnocultural identity, (A), into an existing Bulgarian (A) and an emerging Macedonian (Macedo-Slavic) ethnocultural community, (B), and the segregation of a medieval Galician(-Portuguese) community, (A), into a specific Galician (A) and a Portuguese (B) community. The identification of ethnic groupings and identities by the letters (A) and (B) corresponds to the process of proliferation in each case and thus the two indications may not be reversed. (B), as has become obvious, indicates the group segregated in the process of proliferation and thus refers to the emerging secondary ethnic identity. The emergence of a so-called Bosnie ethnic identity in Yugoslavia is an example of proliferation in which religion constitutes a criterion for group identification. The main feature of the Bosnie nationality, (C), seems to be membership in the Islamic community. People who claim a Bosnie identity do not speak a hypothetical "Bosnie language" since the Bosnie community includes speakers of Croatian (A = Croatian variant of Serbo-Croatian) and speakers of Serbian (B = Serbian variant of SerboCroatian). There are other cases of ethnic fission in which the assumption of proliferation depends on political or ideological views. In this connection, there are the special problems of segregation involved in the formation of a Moldavian (Moldavian-Rumanian) and of a Dungan (Dungan-Chi-

56

Language

in ethnicity:

a view of basic ecological

relations

nese) ethnic identity. According to the official Soviet view (i. e., from the standpoint of Soviet national policy and language politics), the ethnic fission which resulted in the segregation of a Moldavian identity, (B), from a formerly common Rumanian (Daco-Rumanian) identity, (A), must be identified as a process for which the concept proliferation would be an accurate term. The Western view, however, differs in that it links the segregation of Moldavians (on the Soviet side of thè Rumanian-Soviet border) from the Rumanians in Rumania to the political separation of the two communities. Most non-Soviet opinions (including also recent Rumanian official statements) tend to characterize this duality as a profilation (see explanation of the term under 2.1.) within the Rumanian community. Things become even more complicated when taking into consideration the existence of a so-called Rumanian minority and a Moldavian nationality on Soviet territory (see Haarmann 1979b, 15ff.). Similar differences between a Soviet interpretation of a proliferation and a Western view of a profilation arise in connection with the evaluation of ethnic fission among the Chinese. The Soviets claim that the Dungan (Chinese Muslims living in two main settlements in Eastern Kazakhstan) are an independent ethnic group with their own national language (Dungan, written in the Cyrillic script since 1953) distinct from the Chinese community. The Soviet standpoint thus interprets the separateness as a polarity of a (common) Chinese identity (A) and a Dungan Islamic identity (B). The Western view is characterized by the assumption of a polarity comprised of a common (non-Islamic) Chinese identity (A) and an Islamic-oriented Chinese identity (Ai). On the basis of the above outline about ethnic fusion and fission, any given process of shifting ethnic identity can be described and analyzed in terms of the proposed typology of ethnic boundary crossing. Although I mainly referred to ethnic fusion and fission in the European framework of comparison, the conceptual inventory can also be applied to any nonEuropean setting. As the concepts and related terms are theoretical constructs, this implies that single cases of fusion or fission in concrete settings may differ from each other but that a cluster of general features is typical of several concrete settings, making it possible to apply one and the same term. The examples of ethnic fusion discussed for incorporation (see 1.2.) or conglomeration (see 1.3.) may differ because of the special conditions in each concrete setting. The general characteristics, however, allow a categorization of some cases as incorporation and others as conglomeration. In connection with the categorization of concrete processes of fusion or fission on the basis of the proposed typology, one

The role of language in processes of ethnic fusion and fission

57

may observe that in many cases the process of identity shift has remained uncompleted, perhaps because an evolution is still in progress or because the process came to a standstill before it could be completed. The situation of the Serbo-Croatian speech community may be described as an uncompleted ethnic fusion which has to be categorized as a partial amalgamation (thus linking this phenomenon to the type of ethnic fusion illustrated under 1.1.). It is difficult to determine whether to consider this partial amalgamation as the result of a process that has come to a standstill, or as a case of ethnic fusion in progress. Many tend toward the former. The creation of a Serbo-Croatian standard language in the nineteenth century promoted a partial ethnocultural amalgamation of the two communities (A = Catholic Croats, Β = Orthodox Serbs), but the process of amalgamation has never been completed. Most ethnopolitical conflicts — at least in the European framework of comparison — result from the crossing of ethnic boundaries, which is typical of fusion processes categorized as incorporation (see 1.2.). Processes of intense assimilation are in progress in all parts of Europe, including Western Europe (e.g., incorporation of the Basque into the Spanish and French communities; of the Bretons, Occitans, Flemings, and others into the French community; of the Romansh-speakers into the Swiss-German community; of the Frisians into the German community; of Albanians, Greeks, Friulians and others into the Italian community; of the Welsh and Scottish Gaels into the English community; of the Lapps into the Swedish, Norwegian and Finnish communities; etc.) and Eastern Europe (e.g., incorporation of the Sorbs into the German community; of the Hungarians outside Hungary into the Rumanian, Serbo-Croatian, Slovak, and Ukrainian communities; of the Macedonians into the Greek community; of Turks into the Bulgarian community; of the Serbs, Tatars, and others into the Rumanian community; of the Germans into the language communities of many East European states; of many smaller ethnic groups into the Russian community; etc.). In Haarmann (1983 a, 99 ff.) I specified two hundred speech communities in Western and Eastern Europe among which processes of incorporation into dominant contact communities are typical. The intensity of the incorporation process differs considerably in the variety of single cases. A common phenomenon in all the concrete settings, however, is a partial incorporation of people who assimilate into a dominant contact group. The dynamics of partial incorporation with a tendency to enter the final stage of total incorporation is stronger in smaller communities than in those ethnic groups which are comprised of a larger number of

58

Language in ethnicity: a view of basic ecological

relations

speakers. The number of members in a speech community, however, is only one factor in a framework of several dozen ecological variables which promote incorporation as a process of ethnic fusion under special conditions of interethnic relations (see Haarmann 1980b for a comprehensive inventory of ecological variables). An investigation of processes of ethnic fusion and fission in Europe (with special respect to incorporation) on the basis of a theory concerning ecological relations is currently under way. Many relations, however, have yet to be elaborated. This contribution is a partial result of research into the smaller language communities in Europe. Given the conceptual framework of categories outlined above one might think of a dynamic interpretation of the proposed typological categorization of fusion and fission processes. The dynamics of evolutionary stages in ethnic processes are very important in this regard. In the framework of ethnicity-related categories like amalgamation, incorporation, proliferation, etc., each category represents an independent type of shift in ethnic identity. These types, however, compose different evolutionary stages when viewing processes of ethnic fusion or fission as a continuum with extreme polarities and many intermediate stages of development. On the side of ethnic fusion, incorporation clearly marks the most extreme evolution toward a total shift of a former ethnic identity. On the side of ethnic fission, separation with the consequence of newly emerging identities is a clear marker of the other extreme evolution. The other four types of ethnic shift represent intermediate evolutionary stages on an imagined broad continuum between both extremes. For the time being, only provisional statements can be made about the relations between the various evolutionary stages of ethnicity shift because such relations have yet to be thoroughly investigated. It is nevertheless challenging to set up hypotheses about those relations for the purposes of multilateral comparison and evaluation of ethnic processes in concrete contact settings. It is most likely that conglomeration is a preliminary evolutionary stage leading to incorporation (when the process of ethnic fusion continues uninterruptedly in a given setting). In the course of ethnic fission, it seems probable that profilation is an antecedent to separation. As incorporation and separation are considered to be extreme positions on the continuum of potential evolutionary variations, the above hypotheses imply that the direction of development is irreversible. This means that incorporation cannot precede conglomeration and separation cannot anticipate profilation. The exact positions in the continuum of

The role of language in processes of ethnic fusion and fission

59

stages such as amalgamation or proliferation are as yet unknown, and more research has to be carried out to clarify their relations to the other stages of ethnic identity shift. The complexity of ethnic fusion and fission processes has been illustrated with a number of settings, all of which serve as examples of the specification of different types of boundary crossing. Although it has been pointed out that language is an elementary but not a necessary criterion of ethnicity, a more detailed illustration of this may be appropriate here. For this purpose, reference shall be made to some Jewish communities and their settings in multilingual surroundings. The following is based on an interlinguistic perspective of Jewish languages and communities as stressed by Fishman (1981 a). Thus, it is the role of Jewish and other languages for Jewish identity which is treated here (a controversial subject) and not so much the languages themselves. Much has been written about Jewish identity and the development of Jewish languages in different communities (see Fishman 1985 as the latest comprehensive approach). One of the most controversial areas of scholarly study is, doubtless, the settings of Jewish communities in the Soviet Union. Unique in many respects (see below), these settings reveal different kinds of ethnic processes including and excluding language as a factor of Jewish identity. Usually Jewish community settings are investigated by Jewish scholars and the tendency often is to isolate these settings because of their "uniqueness". Although it is hard for a non-Jew to fully understand the problems of Jewish identity, there seem to be, perhaps, more comparable features between Jewish and other ethnic processes than Jewish scholars are ready to admit. One aim of an interlinguistic view of ethnicity problems among Jewish communities is the elaboration of a framework of comparison which includes a number of other ethnic groups. Following this aim, reference will be made in the following outline to non-Jewish diaspora groups such as Gypsies and Greeks living outside Greece. Although this comparison is based only on a view of basic features, it nevertheless provides some insight into ways that Jewish settings may be compared on a larger scale. Such an interlinguistic perspective is needed to establish parameters for the measurement of the so-called uniqueness of Jewish settings. Investigations which take this principle of research as a starting point will certainly cqntribute not only to the understanding of ethnic processes among Jewish communities and their members themselves, but also stimulate the interest and attraction among non-Jews for vital Jewish affairs.

60

Language in ethnicity: a view of basic ecological relations

The phenomenon of diaspora or territorial dispersal as a feature of Jewishness can best be illustrated by the ecological settings of Jewish communities in the Soviet Union. There is no other country where such a variety of different speech communities exists as among Soviet Jewry. In Israel, representatives of each Jewish speech community live together, thus presenting a greater variety of ethnic subgroupings than in the Soviet Union. But these groups do not live in isolation in Israel, but — given the smallness of the country's territory — are integrated in the societal structures of the population. The Soviet Union is unique in that its ecological settings of Ashkenazic and a variety of Oriental Jewish communities reflect the state of Jewry, namely living in diaspora. Other countries such as Turkey, Iran, Iraq, etc., illustrate the phenomenon of Jewish diaspora in their territorial populations, but the diversity of Jewish speech communities is far more limited than in the Soviet Union. The diversity of Jewish speech communities has its historical roots in the manifold contacts between Jews and members of different ethnic groups. The great variety of co-territorial ethnic cultures and languages with which Soviet Jewry has established different kinds of contact is also a unique feature in a world comparison of Jewish ecological settings. The term diaspora has two meanings when referring to Soviet Jewry. On the one hand it means "living in dispersal", which indicates the state of living in separation. On the other hand, it implies "isolation", which is a typical feature of most Jewish communities in the Soviet Union. The term diaspora in this sense refers to the fact that the Ashkenazic as well as the different Oriental communities have little or no contact with each other. There are a number of other ethnic groups which, like the Jews, are characterized by being split up into different speech communities. This is the case of the Gypsies, who represent a variety of languages (other than Romany), but at the same time, do not regard themselves as members of the speech community of the language they have adopted. There is a Gypsy community in Transcarpathia (Western Ukrainia) whose members speak Hungarian as their mother tongue. Many Gypsies in Moldavia and southern Ukrainia have been brought up in Moldavian (a variety of Daco-Rumanian), there are Uzbek and Tadzhik speaking communities of Gypsies in Soviet Central Asia, etc. Another case of diversity is illustrated by Greek settlements. Not all Greeks in the Soviet Union speak Greek or have assimilated to local languages such as Russian, Ukrainian, or Grusinian, etc.; there is also a Turkish-speaking Greek community (the Greeks of Mariupol, modern Zdanov on the northern coast of the Black Sea).

The role of language in processes of ethnic fusion and fission

61

Although Gypsies, Greeks, and other groups with a similar diversification of locally bound speech communities seem to illustrate a state of diaspora similar to that found in Soviet Jewry, they also differ from the latter in many respects. Ethnicity among the Jews is associated with a cultural tradition unique among all co-territorial cultures; this tradition, of course, is Judaism. Being a member of the Turkish-speaking Greek community only means speaking a different language; a Turkish-speaking Greek is a Christian just as is the Greek-speaking Greek, both, therefore, have much in common with many co-territorial communities such as the Ukrainians, Moldavians, etc. Another criterion of Jewishness often neglected in discussions about ethnic identity but which separates Soviet Jewry from the Gypsies, Greeks, and others, is the fact that Jewish communities have produced an effective elite. Without an elite, one could not imagine that Yiddish, Judeo-Tat (Iranian language), or JudeoTadzhik (Iranian language) would have survived the first decades of Soviet rule. Nowadays, as there is little use made of the Jewish written languages in the Soviet Union, being a member of the Jewish elite is more a matter of advanced education (predominantly in Russian). Advanced education itself may be considered not only as a feature of the Jewish elite but also as a feature of Jewishness in general. This feature may be illustrated by ethnostatistical data. Whereas the Russians in general show a comparatively high degree of higher and special education, the Jews in many settings rank in first place with their rates of advanced education. This is true for Jewish communities in the Russian S.F.S.R., Ukrainian S.S.R., Moldavian S.S.R., Lithuanian S.S.R., in the Soviet Central Asia republics, as well as in Dagestan (see Haarmann 1980c, 49ff.). The criteria of the cultural tradition of Judaism and an elite, whose representatives often identify themselves in terms of global Jewry, are features which make Soviet Jewry unique even among ethnic diaspora groups in the Soviet Union. Soviet Jewry in its diversity of speech communities does not fit into the molds of Soviet nationality-related legislation, and only in the first decade of Soviet rule was it actively accommodated by Soviet language planning. According to Soviet views — which, in this respect, have not changed since Lenin's times — there is one national language acting as a common identity for a nationality's collective body. In the case of the Jews, Yiddish is considered to be their national language. Planners of Soviet multinational society thus ignore différenciations between Ashkenazic and Oriental Jews, as well as distinctions among the Oriental Jews themselves. The only group of Oriental Jews who seem to be in favor of

62

Language in ethnicity: a view of basic ecological relations

special recognition in Soviet nationality politics are the so-called mountain Jews of Dagestan, who speak Tat. But it seems as if speaking Tat prevails over recognition of features associated with Jewry. In all Soviet publications about Yiddish, this language is given the status of the national language of the Jews. But this status, valid only for the Ashkenazic Jews, has a more symbolic significance than a practical one. The only language which has the value of a pan-Jewish element in ethnicity is Hebrew. But with its status as a sacred language and cultural vehicle of Judaism, Hebrew finds no place in Soviet nationality law and politics. At the same time, Hebrew could be viewed as a vehicle to stabilize Jewish communities and to strengthen Judaism. According to the principles of Leninist national politics, such functions of Hebrew would not contradict Soviet claims that the cultures and languages of Soviet nationalities shall be promoted. The fact that Hebrew has been campaigned against by Soviet officials and that this language has lost its former sociocultural potential among Soviet Jewry has obviously resulted from prevailing Soviet fears that Hebrew could become the symbol of Jewish irredentism or Zionism. It is due to imagined dangers of deviation from Soviet ideology that many cultural activities among Soviet Jewry have met with suspicion from the officials rather than free acceptance. It is a paradox that in a country like the Soviet Union where there is a long and vital tradition of Jewish culture and history, Jewishness nowadays is mainly a matter of attitude or ancestry and not a matter of religious practice, cultural activity in Jewish organizations, or daily interaction in a Jewish language. Soviet officials do not register Jews according to their inclination to Jewishness or their actual confession of Jewishness. Official figures of the number of Jews in the Soviet Union refer to the criterion of paternity (according to Fishman 1977, 17, a "central experiential concept" and a key element in ethnicity) or ancestry. A Soviet citizen is registered as a member of the collective ethnic body of Jewish nationality if his parents and grandparents were themselves Jews. In the case of ethnically mixed families, it is the nationality of the mother which decides the definition of a Jew. In the framework of ethnostatistical categories, which play a significant role for Soviet society planners, the criterion applied by Soviet officials would correspond to the indication of ethnic nationality (see Kloss-McConnell 1974 a for this and other categories). On the one hand, Soviet censuses which base their data on this objective criterion of ancestry as a definition of nationality offer valuable information about the objective distribution of the population according to ethnic nationalities. On the other hand, such sources lack information

The role of language in processes of ethnic fusion and fission

63

about free choice or inclination toward one or the other collective ethnic body entirely. As Soviet Jewry is unique in many or most interethnic relations, it also appears to be unique in how it is handled by Soviet statisticians. Any sociolinguist attempting a macro-overview of Jewish settlements in the Soviet Union must rely on Soviet censuses as primary sources. Western criticism with respect to the manipulation of data in Soviet statistics (including ethnostatistics) are practically groundless; ethnostatistical data in particular - that is, data collected on Soviet nationalities and their languages — show a high degree of complexity and accuracy (see Comrie 1981, 2). Manipulation of data is not a characteristic of the tables and numerical listings presented in official sources, but rather the result of omitting data in tables where Western statistics would include them. Thus, Soviet Germans were not specified in the Soviet census of 1959 for Kazakhstan, although they constitute one of the most numerous nationalities in that republic. Germans were specified, however, in the censuses of 1970 and 1979 for the Kazakh S.S.R. The omission at earlier times in official sources was due to the political taboo of that ethnic group, and their later specification was a reflection of the official rehabilitation. The Chinese who are considered by Soviet ideologists to be a nationality distinct from the Dungans were mentioned as a minority in Kazakhstan in the census of 1959 but, for obvious political reasons, were not included in the census of 1970. Such manipulations of data by omission can be illustrated for other groups in certain regions as well. Problems arise in connection with figures for the Jews because their subgrouping of Ashkenazic and Oriental communities is not reflected in census tables. Besides this lack of différenciation, there is the gap between figures for "ethnic" Jews (according to principles of Soviet nationality identification) and figures for those who actually feel like Jews or who actively profess their Jewishness. All ethnostatistical data for every Soviet nationality indicate ethnic nationality and leave out information about inclinations of the members in collective ethnic bodies toward the group with which they are associated. In the case of the Jews, in particular, there seems to be a discrepancy between those two criteria that is much wider than in the case of other ethnic groups. This is not meant to lead the analyst to a rejection of Soviet ethnostatistical data. Although there is no information available on nationality preference or inclination as an attitude, the figures presented are nevertheless very useful if the evaluation is based on the principles according to which the data were collected, within the framework of ethnic nationality. The figures in Ta-

64

Language in ethnicity: a view of basic ecological relations

ble 12 indicate Jewishness in terms of ancestry and can only be evaluated as such. The data reflect certain trends in the development of Jewish settlements since 1959 which could be characterized as a general decrease in the number of "ethnic" Jews in the last decades.

Tab. 12. Distribution and Development of Jewish Settlements in the Soviet Union between 1959 and 1979

Republic

1959

Russian S.F.S.R.

875,307

Ukrainian S.S.R.

840,311

Belorussian S.S.R.

Soviet censuses 1970

1979

growth rate (1959-1979)

807,915

700,651

-20.0%

777,126

634,154

-24.5% -

150,084

148,011

135,450

Uzbek S.S.R.

94,344

102,855

99,908

+ 5.9%

Moldavian S.S.R.

95,107

98,072

80,127

-15.8%

Azerbaydzhan S.S.R.

40,204

41,288

35,497

-11.7%

Latvian S.S.R.

36,592

36,680

28,331

-22.6%

Grusinian S.S.R.

51,582

55,382

28,298

-45.1%

14,697

-40.4%

14,667

+ 18.1%

Kazakh S.S.R.

28,048

27,689

Lithuanian S.S.R.

24,672

23,564

Tadzhik S.S.R.

12,415

14,615

Kirghiz S.S.R.

8,610

7,680

Estonian S.S.R.

5,436

5,288

Turkmen S.S.R.

4,078

3,494

***

Armenian S.S.R

1,024

1,048

***

9.8%

***

4,966

*** Ethnostatistical data are not yet available. Sources: Itogi 1962, 1963, Itogi 1973, Nacional'nyj sostav 1980

-

8.6%

The role of language in processes of ethnic fusion and fission

65

The general decline in the number of Soviet Jews is reflected in overall figures for all the Soviet Union (e.g., 1959: 2,267,814 Soviet Jews, 1970: 2,150,707, 1979: 1,810,876), as well as in most data for single republics. The sharpest decrease can be observed for the period between 1970 and 1979. The Uzbek S.S.R. and the Tadzhik S.S.R. seem to be the only regions where the number of Jews increased between 1959 and 1979 (although this tendency was weakening during the 1970s). The most far-reaching ethnodemographic change has taken place in the Georgian, or Grusinian, S.S.R., where there was a decline in the population of Soviet Jewry of 45.1%. The only region with considerable growth in the Jewish population is Tadzhikistan, where the number of Jews increased by 18.1%. The growth rate, however, almost entirely reflects an increase between 1959 and 1970, whereas between 1970 and 1979 there was stagnation. The situation in the Tadzhik S.S.R. has to be considered an exception to the general steady decline of the Jewish population during the 70s. Although there was growth in the population between 1959 and 1970 in regions such as Uzbekistan, Moldavia, Azerbaydzhan, Latvia, and Grusinia, the number of Jews declined (in some regions rapidly) in the course of the 1970s. Figures are not yet available for Kazakhstan, Kirghizistan, Turkmenia, and Armenia, but the general trend seems to indicate that in these regions, too (except for Armenia), there was steady decline in the 70s. Soviet Central Asia in general and Tadzhikistan in particular are the regions with the highest degree of stability in the demographic structure of Jewish settlements. As there is no distinction made between Ashkenazic and Oriental Jews in the Soviet sources, one can only make some general remarks about how differently the subgroupings of the Jewish "nationality" were affected by population changes. As the great majority of Soviet Jews belong to the Ashkenazic subgrouping (the so-called European Jews), the decline has most strongly affected the Ashkenazim. The tendency of a declining population can be observed in the Russian S.F.S.R., in Ukrainia, Belorussia, Moldavia, and in the Baltic areas of Latvia, Lithuania, and Estonia. Ashkenazic Jews are living in the other republics (either in the neighborhood of or separated from Oriental Jewish communities), but their demographic development cannot be exactly determined. The impression from the available data, as well as from findings with respect to language maintenance and language shift (see Haarmann 1980 c, 123 ff. for recent developments in Soviet Central Asia), allows us to conclude that in areas where both Ashkenazic and Oriental Jews reside (Central Asia, Caucasus) the negative growth is more marked in the

66

Language in ethnicity: a view of basic ecological

relations

Ashkenazim than in Oriental Jewish communities. On the other hand, one can point out that the decline of Ashkenazic Jewry is stronger in central regions of the Soviet Union (e.g., Russian S.F.S.R., Ukrainian S.S.R., or Lithuanian S.S.R.) than in peripherical areas (e.g., Moldavia, Estonia, Azerbaydzhan, the Asiatic republics). The Oriental Jewish communities (especially in Central Asia) generally reveal a higher degree of stability. The most significant reasons for ethnodemographic change among the Jews are ethnic boundary crossing (shifting ethnic nationality) through interethnic marriage, and emigration. Both factors sometimes affecting Jewish communities interdependently — have a differing impact on the Jewish population in different regions. The ecological settings of Jewish communities are most complex in regions where Ashkenazic and Oriental Jewish settlements have been established in the course of history. Modern history has changed the profile of Jewish communities considerably in Soviet Central Asia. Before the Second World War, the majority of these communities consisted of Oriental Jews, but due to evacuation from Western parts of the Soviet Union, the Ashkenazim have been more numerous since the early 40 s (see Haarmann 1980c, 19ff.). In Table 13, I present a specification of Jewish communities in Central Asia and Kazakhstan. Where two subgroupings are mentioned, the one given first indicates the majority population. Thus in Buchara and Samarkand, there is a majority of Oriental Jews and a minority of Ashkenazic Jews, whereas in Fergana and Taskent, the Ashkenazim are more numerous than the Oriental Jews. Ashkenazic and Oriental Jews behave differently in terms of language maintenance and second language attraction. In Haarmann (1980 c, 219ff., 265 ff.) I outlined the basic features of this behavior. A salient feature of Jewishness is the variety of Jewish languages. This warrants some explanation. It is the variety of Jewish languages that is a salient feature of Jewishness, not the languages themselves. As most Soviet Jews have assimilated to co-territorial languages (mostly Russian), Jewish languages have for decades been a feature of ethnicity among only a minority of Jews. This realization is vital for understanding the fact that language is a minor element of Jewish identity in the Soviet Union. The ecological settings of Soviet Jewry clearly illustrate that there is only a partial relationship between language and ethnicity. There are two basic perspectives on how the diversity of Jewish languages can be described in terms of ethnicity. In the outline about ethnic fusion and fission (see above) I differenciate between profilation as a phenomenon of ethnic fission and conglomeration as a phenomenon of

The role of language in processes of ethnic fusion and fission

67

Tab. 13. Jewish communities in urban centers in Soviet Central Asia (including Kazakhstan (source: Haarmann 1980c, 25)

Territory

Urban center

Jewish communities

Kazakh S.S.R. (KazSSR)

(1) Karaganda (2) Kzyl-Orda (3) Cimkent (4) Alma-Ata

Ashkenazic Jews Ashkenazic Jews Ashkenazic Jews Ashkenazic Jews

Kirghiz S.S.R. (KirgSSR)

(5)

Frunze

Ashkenazic Jews

Uzbek S.S.R. (UzSSR)

(6) Andizan (7) Fergana (8) Taskent (9) Buchara (10) Samarkand (11) Karsi

Ashkenazic Jews Ashkenazic/Oriental Jews Ashkenazic/Oriental Jews Oriental/Ashkenazic Jews Oriental/Ashkenazic Jews Ashkenazic/Oriental Jews

Tadzhik S.S.R. (TadSSR)

(12) Leninabad (13) Dusanbe

Ashkenazic/Oriental Jews Oriental/Ashkenazic Jews

Turkmen S.S.R. (TurkSSR)

(14) Cardzou (15) Aschabad

Ashkenazic/Oriental Jews Ashkenazic/Oriental Jews

ethnic fusion among Jewish communities. The term profilation refers to the state of an ethnic group being represented by communities with differing ethnocultural variants. Profilation thus indicates ethnic subgroupings which may be specified in terms of language varieties. With special

68

Language in ethnicity: a view of basic ecological

relations

reference to the variety of Jewish languages in the Soviet Union, profilation can be illustrated by the following formula: A

Ai + A 2 + A 3 + A 4 + A 5 . . . (+ A„)

A = Soviet Jewry, Ai = Yiddish-speaking Ashkenazic communities, A2 = Tat-speaking Oriental communities, A3 =Tadzhik-speaking Oriental communities, A4 = Uzbek-speaking Oriental communities, A5 = Grusinian-speaking Oriental communities. A n in the above formula indicates that the number of languages may vary considerably when applying the formula to settings of different ethnic groups. In the case of the Jews in the Soviet Union, the undisputed number of Jewish languages amounts to three (for the problem of Karaim, see below). Profilation characterizes the state of Jewishness as reflected in the variety of languages from the standpoint of the Jewish collective ethnic body. The other view of Jewish ethnicity has to do with the interpretation of certain assimilation processes as a conglomeration. Conglomeration is a phenomenon of ethnic fusion (partial ethnic fusion) among only the Oriental Jews. Single Oriental communities could be described according to the following formula: A + Β

A + Ba

A = Oriental Jewish community, Β = Tadzhik-speaking community (Tadzhik speech community), B a = Oriental Jewish community (characterized as A3 in the formula for profilation) with Tadzhik as mother tongue. Integrating all the single Oriental Jewish communities in one formula would make this rather complex. For the purpose of this contribution, the above formula may suffice as an illustration of conglomeration. Whereas profilation as a phenomenon of ethnic fission within Jewry is well known (though not under this special term), the aspect of conglomeration may seem unfamiliar even to experts. This may be due to the fact that most categorizations in the area of Jewish ethnicity are provided by Jews, thus implying mostly a Jewish perspective. In the framework of an interlinguistic analysis, however, one has also to take into account aspects of interethnic contacts, which requires a neutral standpoint, and that is a non-Jewish view. What is indicated as B a in the formula of conglomeration refers to the special status of Tadzhik-speaking Jews in the Tadzhik-speaking language community. The Oriental Jews who distinguish themselves clearly by their Judaic tradition from the surrounding Muslim population nevertheless form a portion in the collective body of

The role of language in processes of ethnic fusion and fission

69

those who speak a variant of Tadzhik. Judeo-Tadzhik has to be seen as a social variety (sociolect) of Tadzhik, because only this interpretation leaves room for the understanding of both past and present integration processes. Although the concept of conglomeration as presented here may at first sight give the impression of being a non-Jewish view (which could be rejected as an outsider's standpoint), it is a necessary tool in the set of categorizations with respect to Jewish identity and ethnicity-related issues. When Lewis (1972,139) speaks of "secondary national affiliations" among Soviet Jewry he refers to the phenomenon of profilation (see above). Although Lewis did not intend it, his description also allows for an interpretation of conglomeration. Given the fact that conglomeration is a phenomenon limited to Oriental Jewish communities, one could speak of profilation as a primary (or basic) feature of Soviet Jewry, whereas conglomeration is a secondary (or special) feature of Oriental Jewishness. There are several ways to define the term "Jewish language". Here, I would like to quote Weinreich (1980, 166): "The essence of a Jewish language, we postulate here, is the fact that it is a fusion language". The concept of fusion is indeed a salient criterion of every Jewish language and is thus adequate to serve as a characteristic of any Jewish language. One has to be aware, however, of the fact that the degree or extent of fusion may differ considerably from one Jewish language to the other. In the following outline, I limit myself to a description of some basic features in the ecological settings of Jewish languages in the Soviet Union to which Weinreich's general definition applies.

1. Yiddish The uniqueness of Jewish settings in the Soviet Union has been underscored in several connections in the preceding. The postwar situation of Yiddish as a communicational means of Ashkenazic Jews also reveals features of uniqueness. Recent developments in Jewish communities indicate a steady decline of Yiddish. Although this language is the most widely spread Jewish language in all the Soviet Union and is spoken or known by several hundred thousand Jews, Yiddish is at the same time the weakest communicational means of all the Jewish languages. This paradoxal state of the language bears some explanation. According to the Soviet census of 1979, Yiddish was a first language among 257,813 Soviet Jews. This would correspond to 14.2% of total Soviet Jewry. As

70

Language in ethnicity: a view of basic ecological

relations

Ashkenazic as well as Oriental Jews are included in this census figure, the percentage is actually less than 14% for the Ashkenazic communities alone. This low level of language maintenance, which fell from 21.5% in 1959 to 17.7% in 1970, is among the lowest rates for Soviet nationalities. Indeed, the degree of language maintenance among Ashkenazic Jews is in the same range as the language loyalty rates (rates for the preservation of the national language as the first language) among the smallest speech communities (e.g., the Izhorians, Veps, Karaims, Itelmens, etc., and their mini-languages). Yiddish is at least known by Jews in all Soviet republics. Moreover, Yiddish is much weaker in the central regions, and better preserved in the periphery. Thus the rate for Yiddish as the first language is only 9% in Ukrainia, 10% in Russia (Russian S.F.S.R.), and 11.2% in Belorussia, but 28.3% in Latvia, 33.1% in Moldavia, and 41% in Lithuania. There are also communities where the maintenance rates exceed 50%. With respect to the Jewish communities in Grusinia, about 33% of all Jews have preserved Yiddish as their first language according to the figures for 1979. As the total number of Jews in that region includes both Ashkenazic and Oriental Jews, one can estimate the maintenance rate for the Ashkenazic communities in particular to be over 50%. Detailed figures for Soviet Central Asia are not yet available. According to the 1970 census there are even instances of a preservation of Yiddish by over 70% of Ashkenazic Jews in some regions (see Haarmann 1980 c, 225 ff. for communities in Uzbekistan). The Ashkenazic community in Dagestan is unique with an average maintenance rate for Yiddish of 86.7% in 1970 and of 86.1% in 1979. As the data in the Soviet census refer to first language acquisition (first language status in chronological perspective), they do not say anything about the use of a communicational means adopted in early childhood. It is most probable that in the case of Yiddish, the figure given above for the Soviet Union reflects passive knowledge rather than actual ability to speak among Ashkenazim. Most of the Jews who have adopted Yiddish as their first language do not use it actively, but rather maintain as adults a passive knowledge of what was a childhood language. The total number also includes many cases of only a little passive knowledge of Yiddish in cases where the second language (mostly Russian) has been dominant for a long time in an individual's daily interaction. Given the fact that passive knowledge supersedes active use of Yiddish in many communities, the number of Yiddish first-language users must be considered a remnant of a formerly flourishing Yiddish cultural life among

The role of language in processes of ethnic fusion and fission

71

Ashkenazic Jewry. The few books and booklets published in Yiddish (and on Yiddish), as well as the newspaper Birobidzhaner Shtern (appearing since 1955 in Birobidzhan) and the magazine Sovietish Heymland (begun in 1961 in Moscow) are cultural relics (see Haarmann 1980c, 73 ff. for further details). When the newly established Jewish Musical Theater was looking for actors some years ago, the organizers had to teach younger team members Yiddish first. Although the figures for Yiddish in the census are much higher than for any other Jewish language among Soviet Jewry, this language is nevertheless the one most endangered by decline. One main reason for the dynamic process of language shift among Ashkenazic Jewry is the fact that there is no school instruction in Yiddish and no curriculum which includes Yiddish as a subject. In this respect, Soviet postwar school policies differ from prewar regulations, which provided Yiddish language education among Ashkenazic communities in Central Asia. As there has been no education in or through Yiddish for decades, knowledge of this language has steadily faded among the Jewish postwar generations. Another reason for the decline of Yiddish, beyond the passiveness or even reluctance to promote Yiddish on the level of Soviet language planning, is surely the high degree of dispersion of Ashkenazic settlements. Most communities of Oriental Jews are less dispersed and their languages find more favorable ecological conditions for their survival than Yiddish. Russian is the most widely spread communicational means of Ashkenazic Jewry, but it does not function as a cultural vehicle of Jewishness. Even in those circumstances where Russian is used in Jewish publications, the impression is more of a vehicle expressing cultural items of Sovietism rather than Jewishness. For example, to assure access of non-Yiddish-speaking Jews to Sovietish Heymland, a summary of the contents of every issue is given in Russian.

2. Judeo-Georgian The situation of the Oriental Jewish community in the Georgian S.S.R. is, to a certain degree, unique not only among Oriental groups, but also with respect to Soviet Jewry as a collective body. Grusinia is the only region with Jewish settlements where the Jewish language has a greater number of speakers than Russian. In terms of first language acquisition, this means that the number of Jews who have adopted Georgian as their mother tongue (10,020 in 1979) is greater than the number of

72

Language in ethnicity: a view of basic ecological

relations

Ashkenazic as well as of Oriental Jews who have assimilated to Russian (8,777 in 1979). It must be realized, however, that this refers only to one particular aspect of Oriental Jewry, and must not be misunderstood as a sign of stability in the Jewish community. Jewish communities in Georgia lost nearly half of their members between 1970 and 1979 due to de-ethnization (as the result of assimilation), migration to other regions of the Soviet Union (mostly to towns in the Russian S.F.S.R.), and emigration (to Israel). The number of Oriental Jews speaking Georgian as their mother tongue decreased from 32,032 in 1970 to less than one third this number. Formerly, the Oriental community of Georgian Jews was numerically the strongest among Oriental Jewish communities in the Soviet Union. And, although Oriental Jews assimilated to Georgian long ago, they live more or less isolated from the Georgian-speaking communities (see Neustadt 1970). Oriental Jews have also developed some social distance toward Ashkenazic Jews living in Georgia. On the whole, isolation as a feature of an ethnic group's collective support system seems to be as typical of the Oriental Jews in Georgia as it is of the behavior of other Oriental Jews in Asia. Thus, the co-territorial language (Georgian), although the basis of the Jews' communicational means, does not play the role of an interactional link, integrating them into the corresponding community. Perhaps because of these isolational characteristics, some scholars tend to stress the independence of Judeo-Georgian as a Jewish language, which, actually, in terms of linguistic affinity, is a variant of Georgian.

3. Judeo-Tat or Judeo-Tatic (Dzhuhuric) In some southeastern regions of Caucasia live the so-called mountain Jews who speak a variant of an Iranian language, namely Tat. The Hebrew component makes this variant a Jewish language (Judeo-Tat). Most mountain Jewish settlements are found in Dagestan, which is an A.S.S.R. in the territory of the Russian S.F.S.R. A smaller portion of Oriental Jews speaking Tat live in northern Azerbaydzhan. In the Soviet censuses, Tats as an Iranian people (nationality) and the mountain Jews are not distinguished, so that ethnostatistical data refer simply to "Tats". As most Tats in Dagestan are mountain Jews, one can specify the number of Tat-speaking Oriental Jews in that region at about 6,800 (1979), which corresponds to a language maintenance rate of 91%. There is no reliable estimate, however, for the number of mountain Jews

The role of language in processes of ethnic fusion and fission

73

in Azerbaydzhan speaking Tat. The census of 1979 gives 3,965 Tats (j. e., Tats and mountain Jews) who claim to have adopted Tat as their first language. Comrie (1981, 164) argues that there is an imbalance in the number of Tat speakers because many Tats in Azerbaydzhan tend to claim Azeri as their mother tongue. Thus one could assume that the above number of Tat speakers, which corresponds to only 44.8% of the Tat population of Azerbaydzhan, probably comprises a majority of Jews Tab. 14. Alphabets for the two standard varieties of written Tat (Tat proper and Judeo-Tat)

Taza sldfbi Tati. HOBMH Ta-rcKBH

ΑλφΛΒητ.

Aa

Bb

Ce

Çç

Dd

Κ

η



Ί

Ί

9a

Ff

Gg

0|οι

Hh

Tifi

Κ

S

1

Π

Π

Μι

li

JJ

Ρ Kk

LI

Mm

b

Ö

y

*

Nn

Oo

1

Rr

Ss

Κτ

Pp S

η

0

Tt

Uu

Vv

Xx

η

1

η

D

yy 1



D(t

ff

Zz »

Source: Cobanzade (1930, 129)

*) In Cobanzade's original table D d is erroneously represented by the same Hebrew letter as the one used for R r. This error has been corrected here.

74

Language in ethnicity: a view of basic ecological

relations

with Tat as their first language. Gijunberg (1963) doubted the official figure for Tats in the Soviet census of 1959 (i.e., 11,463) and estimated their number at around 20,000 to 30,000. There is, consequently, much room for speculation about the total number of Oriental Jews speaking Tat since neither the census nor other sources contain precise figures. In the late 1920s, Soviet language planning promoted the creation of a modern literary standard for a number of minority languages in the Caucasus region. The Latinization campaign, that is, the application of the Latin script for writing Turkic and Caucasian languages, had a tremendous impact in Azerbaydzhan with its variety of smaller languages. For Tat, too, the NTA (Russian, novyj tjurkskij alfavit, referring to the Latinization of the national language in Turkey in the early 20 s) was adopted. Under early Soviet language planning, it is important to note that the Latin script was restricted in usage for written Tat to the Muslim Tat communities, whereas the Hebrew alphabet became the basis for Tat as a written language of the mountain Jews (Russian, gorskoevrejskij literaturnyj jazyk). Cobanzade (1930,128 ff.) reported on the development of the two varieties of written Tat in the 20s (see Table 14). A number of school books, a newspaper, and pamphlets concerning political, social, and economic issues were published in the years following the creation of the written norms. As Tat has not been used as a medium of instruction since the 1940s, publications in Tat have remained outside educational sectors. Until recently, most literature in Tat, which is published in Makhachkala (Dagestan) and in Moscow, appears in the Hebrew alphabet and thus is meant mostly for the mountain Jews.

4. Judeo-Tadzhik (Bokharic) Oriental Jews speaking Tadzhik, an Iranian language closely related to Persian, live in various parts of the Uzbek S.S.R. as well as in Tadzhikistan. The cultural center of the Oriental Jews in Central Asia has been Bokhara (Buchara) in Central Uzbekistan (see Table 13 for the location of Jewish settlements in the Asiatic regions). Exact figures for Tadzhikspeaking Jews can only be given for Jewish communities in the Tadzhik S.S.R. According to the 1979 census, 5,155 Jews in Tadzhikistan had adopted Tadzhik as their first language. The number of Yiddish-speaking Ashkenazic Jews in that region is comparatively low (2,905 in 1979). Both Ashkenazic and Oriental Jews in this area have assimilated to Russian, the number of first-language speakers amounting to a total of

The role of language in processes of ethnic fusion and fission

75

6,564. It is reasonable to assume that the percentage of assimilated Ashkenazim is much higher than that of Oriental Jews. Detailed figures for the Uzbek S.S.R. are not yet available. The number of Oriental Jews speaking Tadzhik in Uzbekistan, however, can be estimated at several thousand (see Haarmann 1980c, 138ff., for a discussion of ethnostatistical data in the 1970 census). Judeo-Tadzhik, written using the Hebrew alphabet, can be considered to be a variety of Judeo-Persian (see Lazard 1971). Weinreich (1980, 73) calls Bokharic (Judeo-Tadzhik) and Dzhidi (Judeo-Persian as used by the Jews in Iran proper) "two dialects of the same language". The cultural tradition of Judeo-Tadzhik is one of the oldest among Oriental Jewry, yet the background of this tradition as well as its origins are not well known. Much remains to be learned concerning the historical relations of the Persian-speaking Jews in Central Asia with the Oriental Jews in the Caucasus. "Are there direct linguistic associations between Dzhuhuric, Bokharic, and Dzhidi beyond their common link to Persian? To what degree are there discoverable threads extending from these three linguistic groups to the ancient Parsic? These questions have not yet been touched by scholarship" (Weinreich 1980, 73 ff.).

Research is greatly needed in order to determine the conglomeration process among Oriental Jews and to shed light on the chronological stages of their ethnic fission. Although the above description cannot provide the reader with more than a short overview (leaving out Jewish languages like Judeo-Crimean Tatar, and Judeo-Kurdish), it may nevertheless be of some use. Weinreich (1980, 73), in his outline of Yiddish in the framework of other Jewish languages, names only two language varieties of the Oriental Jews, namely Dzhuhuric (see 3 above) and Bokharic (see 4) for settings in the Soviet Union. All Jewish languages among Soviet Jewry are in decline with respect to their number of speakers as well as their functional ranges and sociocultural potential (i. e., as written languages). The dynamics of decrease are different in different single communities and vary according to language. Thus the decline is much stronger among the Yiddish-speaking Ashkenazic Jews than among the Judeo-Tadzhik speaking Oriental Jews. Even when speaking of stability in the Oriental communities of Central Asia (see Haarmann 1980c, 298ff.), such statements refer to a relatively slower pace of assimilation than among the Ashkenazic communities but they do not imply any notion of growth.

76

Language in ethnicity: a view of basic ecological

relations

All Jewish languages in the Soviet Union are characterized as fusion languages, although the impact or the degree of fusion varies cosiderably from one variety to another. The fusion character of Yiddish (see Weinreich 1980, 599ff.) is much more complex and extensive than, for instance, in the case of Judeo-Georgian or Judeo-Tadzhik. The impact and the consequences of fusion in each Jewish language may perhaps be described best in terms of a fusion continuum with different degrees of intensity. The Hebrew language plays a key role in this fusion and is the most typical determinant in all Jewish languages. Thus Hebrew is not only the common cultural vehicle for all the "Jewish subcultures" (Weinreich 1980, 74), but is at the same time the common linguistic determinant of fusion processes in Jewish language varieties. These observations are important in a discussion about Jewishness, but they do not address the basic problem of how to define or specify the role of Hebrew as a feature of Jewish identity. There is at least one clear case of a language variety with Hebrew influence which traditionally has not been considered a Jewish language, and that is Karaim. All definitions of Jewish languages which are based on the characteristic fusion and the role of Hebrew as criteria (e. g., Weinreich's definition) fail in connection with the specification of Karaim. Karaim, which, according to the 1979 census, is spoken as a first language by only 535 persons (corresponding to 16% of the total population), is the national language of the Karaim (ethnic term) or Karaites (religious designation) who live in small communities on the Crimean peninsula, in southwestern Ukrainia and in Lithuania. The Karaim have preserved their religion, Judaism, which Comrie (1981, 47) calls "a unique survival of the adoption of Judaism as the official religion of the Khazar empire". Karaim, written with the Hebrew alphabet, is a fusion languages with a Hebrew determinant and is linguistically as distinct from other Turkic languages as Judeo-Georgian is from Georgian proper or Judeo-Tadzhik from Tadzhik proper. Yet the Karaim are not considered Jews and their language is not considered a Jewish language. The linguistically oriented definition of a Jewish language is such that the fusion character as well as the Hebrew determinant do not by themselves indicate Jewishness, but only Judaistic tradition. The criterion of ancestry determines traditional views of what are Jewish languages. Thus a fusion language with a Hebrew determinant is implicitly understood as being spoken by ethnic Jews and not by a community of (ethnically) Turkic descent. Fishman (1981 a, 5 ff.) calls any language "Jewish" "that is phonologically, morpho-syntactically, lexico-semantically or orthographically dif-

The role of language in processes of ethnic fusion and fission

77

ferent from that of non-Jewish sociocultural networks and that has some demonstrable function in the role-repertoire of a Jewish sociocultural network". Jewish ancestry as an element of identity among the speakers of a Jewish language is not explicitly referred to in this statement, although further explanations indicate that this component is included. Thus, according to Fishman's definition, Karaim could not be included among the Jewish languages in the Soviet Union. As the element of Jewish ancestry is missing, the fusion character of Karaim as well as the Hebrew component are features of Judaic but not Jewish identity. Less convincing are categorizations such as those proposed by Rabin (1981). Any categorization of Jewish languages which focuses on the role of Hebrew in a diglossie setting must fail because it contains the same weakness as the definition given by Weinreich. Hebrew as such is not a unique feature of Jewishness either because it is a fusion determinant in Jewish languages or because it functions as a sacral language in Judaism (with the role of a high variety in communities). Both criteria are valid for the settings of Karaim speakers. Thus it is the correlation of the Hebrew element with the (ancestry-oriented) criterion of the Jewish community which seems to be the clue to the definition of a Jewish language. The settings of Jewish communities amidst a variety of co-territorial cultures and contact languages can only be fully understood if one does not only concentrate on the characteristics of Jewish languages as single units, but also takes into consideration the contact conditions. This enables the investigator to discern the status of Jewish languages in different communities and the mobility among Jews in using different communicational means for the purposes of interaction. To illustrate the importance of this, I shall discuss some aspects of Jewish interlinguistics (or interlinguistics of Jewish languages). In the Soviet census of 1970, a distinction was made for the first time between first and second languages. With such ethnostatistical data (available also in the one for 1979), the sociolinguist can specify basic types of communicational patterns (e. g., monolingualism, bilingualism, multilingualism) for single Soviet nationalities. With respect to the Jewish communities, ethnostatistical data may be used to illustrate the communicational network (or network of communicational means) among Ashkenazim and Oriental Jews. Most Jews are monolingual, having adopted Russian as their first language. Although there is a tendency toward assimilation in all communities, Russian first-language attraction is the strongest among Ashkenazim. This general spread of monolingualism reflects the decline of Jewish languages. Monolingualism involving a Jewish language can be excluded

78

Language in ethnicity: a view of basic ecological relations

altogether as a communicational pattern among Jewish groups. It is plausible that the knowledge of a second language with a wider functional range is necessary for a child in a Jewish family brought up speaking a Jewish language. The knowledge of a Jewish language enables the speaker to communicate only with family members or close relatives, although in the Oriental communities, the Jewish language variety provides him access to the co-territorial speech community (e. g., Georgian, Tat, Tadzhik). As school education for Soviet Jews is almost entirely in Russian, it is this communicational means which plays a key role as a second language among Ashkenazim and Oriental Jews. The general conclusion, then, is that all Soviet Jews speak Russian, which they have adopted either as a first language (Jewish assimilation) or as a second language (speakers of a Jewish language with Russian second language knowledge). The question remains as to the extent Jews master Russian in terms of language skills (speaking, understanding, reading, writing). It can be assumed that not only do most Jews have a developed passive and active knowledge of Russian, but they also frequently (if not exclusively) use it for daily interaction in intragroup communication (interaction among Jews themselves) as well as for the purpose of intergroup communication (interaction in interethnic contacts). In multilingual contact areas, the Russian-speaking Jews are often categorized as Russians by members of local ethnic groups, making them a target of anti-Russian resentment. "Nationalism is also growing among other minority nationalities. Jews are often caught in the middle; local nationalists regard them as alien Russianizers and sovietizers and Russians regard them as Jews (i.e., non-Russians)" (Katz 1975 a, 379).

In addition to monolingualism, which is the most widespread communicational pattern among Soviet Jewry, bilingualism with Russian as a second language component can be found in each Jewish community. This type of communicational pattern is called national-Russian bilingualism (see Haarmann 1985 b), because it comprises the national language of an ethnic group as the first language component and Russian as a second(ary) constituent. As this variant of group bilingualism appears in many specific manifestations among all non-Russian nationalities in the Soviet Union (e. g., Estonian-Russian bilingualism among the Estonians, Armenian-Russian bilingualism among the Armenians, etc.), it is characterized as "mass bilingualism", carrying great societal impact (see Deseriev 1976b). Given the divergence of Jewish subcultures, there is a

The role of language in processes of ethnic fusion and fission

79

variety of manifestations of group bilingualism among Soviet Jewry (e.g., Yiddish-Russian bilingualism among Ashkenazic Jews, JudeoGeorgian-Russian bilingualism among Oriental Jews in Georgia, JudeoTat-Russian bilingualism among Oriental Jews in Dagestan, Judeo-Tadzhik-Russian bilingualism among Oriental Jews in Tadzhikistan, etc.). Although ethnostatistical data are only available for some of these communicational patterns, structures of bilingualism can be specified for a number of Jewish communities in different surroundings. As contact conditions vary, from setting to setting, language maintenance rates as well as Russian second language attraction rates correlate differently among local communities. In Table 15, the main constituents of national-Russian bilingualism are illustrated in a basic model of group bilingualism. Such a model may be applied to all settings where Russian is the only (or only significant) co-territorial contact language for the local Jewish population. National-Russian bilingualism is a communicational pattern significant for interaction in all Jewish communities where a Jewish language is maintained as a mother tongue. Given the fact that national-Russian bilingualism is a configuration in the network of a whole communicational structure (i. e., components in the basic model of group bilingualism), the dynamics of language maintenance and (Russian) second-language attraction are not only determined by the direct correlation of these two factors (i. e., 1:2'), but their relation is also controlled by additional factors, such as the dynamics of assimilation (i.e., 2 influencing 1). There are some regularities in the correlations which can be illustrated with ethnostatistical data. Some of these regularities shall be mentioned as typical features of national-Russian bilingualism among bilingual Jewish communities. A widespread phenomenon is the correlation of low rates of language maintenance (i. e., preservation of Yiddish as the first language) and high rates of second language knowledge (Russian with the status, 2'). As can be illustrated by bilingual structures in Oriental Jewish communities, normally a high level of maintenance rates (i. e., Tadzhik as the first language) correlates with a comparatively lower degree of second language knowledge. From empirical findings from a variety of non-Jewish speech communities, there is much evidence making it plausible to assume that a stronger impact of Russian as a first language (i. e., assimilation pressure exerted from the position 2 unto the position 1 in the model) correlates with a higher degree of Russian second-language attraction. Whereas the degree of language maintenance is directly determined by the dynamics of assimilation, it is indirectly controlled by

80

Language in ethnicity: a view of basic ecological relations

Tab. 15. A basic model of group bilingualism (involving Russian as a contact language)

Explanations: (1) characterizes the status of a (non-Russian) community's national (nonRussian) language as a first language (language adopted in early childhood); e.g., 70,204 Jews in the Russian S.F.S.R. with Yiddish as their first language. (1') characterizes the status of the national language as a second language among non-Russian assimilants who have adopted Russian as their first language (i. e., bilingual Jews who were brought up in Russian but learned Yiddish as a second language); e.g., 34,458 Jews in the Russian S.F.S.R. with Yiddish as a second language. (2) characterizes the status of Russian as the first language of non-Russian assimilants (for instance, Jews who have adopted Russian as their first language); e.g., 627,371 Jews in the Russian S.F.S.R. who have assimilated to Russian. (2') characterizes the status of Russian as the second language of bilinguals who have adopted their national language as a first language (status 1, for instance, bilingual Jews with Yiddish as their first and Russian as their second language); e.g., 62,261 Yiddish-speaking Jews in the Russian S.F.S.R. with Russian as a second language.

the correlating second language attraction. A weaker impact of (2) correlates with a weaker impact of (2') — and vice versa. When language maintenance is comparatively strong in a Jewish community, it can be observed that both the assimilation impact (Russian as a factor of language shift) as well as the second language attraction are comparatively weak. There are, however, a number of specific divergences which make the bilingualism issue a complex subject for Jewish communities. There is no space here, however, to elaborate on empirical findings.

The role of language in processes of ethnic fusion and fission

81

Table 16 illustrates the contact conditions in multilingual settings where at least three languages are involved (one being a local non-Russian language). Settings characterized by complex contact conditions can be found among Ashkenazic communities (e. g., Lithuania, Moldavia) as well as in Oriental communities (e.g., Dagestan, Uzbekistan). Multilingualism as a communicational pattern is most widely spread among Jewish communities in peripheral regions. It can be generally stated that multilingualism is steadily declining and multilingual structures are more rapidly dissolving than bilingual structures. Bilingual and multilingual structures as communicational patterns are best preserved among older people, thus making bilingualism, multilingualism, and the maintenance of Jewish languages a characteristic of the older Jewish generations. The phenomena of group bilingualism and multilingualism among Soviet Jewry above described cannot be explained but by the dynamics of influencing ecological factors of language choice, preference, maintenance, etc. (see chapter 1). The mentioned age factor above is an important ecological factor within the functional range of ethnosociological variables. Although the amount of ethnostatistical data with respect to the distribution of Jewish languages according to age groups in the Jewish population is limited, some general observations can be made. As it is reasonable to assume that the preservation of Jewish languages among Soviet Jewry is closely linked to the upper age groups, there must be some evidence available in the ethnostatistical material provided by the census(es). In the census of 1970 (see Itogi 1973, 373), an age-group distinction is provided for the Jewish communities in the Russian S.F.S.R. The evaluation of the data in the tables reveals a strong disproportion in the distribution of language maintenance rates for Yiddish according to age group. These disproportions reflect extreme conditions of pressure exerted by Russian on the Jewish communities. The rates for Yiddish as the mother tongue (first language status) are not as high as 20% in any age group. Thus, a strong assimilation process is occurring in all age groups. Considerable disproportions can be identified with respect to the total number of Yiddish-speaking Jews in that republic. Most Jews who adopted Yiddish as their first language are over fifty. Nearly 40% of all Yiddish speakers are distributed in the upper age groups (60 years and older). This indicates that the assimilation to Russian is extremely strong in the lower age groups and that the number of speakers of Yiddish will rapidly decline from one decade to the next (see Haarmann 1985 b, 3.3. for further details on age-group distinctions with respect to language maintenance rates in different speech communities).

82

Language in ethnicity: a view of basic ecological relations

Tab. 16. A basic model of group multilingualism (involving Russian and another contact language)

(1) characterizes the status of a (non-Russian) community's national (nonRussian) language as a first language; e.g., 6,022 Jews in the Lithuanian S.S.R. with Yiddish as their first language. ( Γ ) characterizes the status of the national language as a second language among non-Russian assimilants; e.g., 298 Jews in the Lithuanian S.S.R. with Yiddish as a second language. (2) characterizes the status of Russian as the first language of non-Russian assimilants; e.g., 7,990 Jews in the Lithuanian S.S.R. speaking Russian as their first language. (2') characterizes the status of Russian as the second language of bilinguals who have adopted their national or another contact language as their first language; e.g., 4,380 Jews in the Lithuanian S.S.R. with Russian as a second language. (3) characterizes the status of a (non-Russian) co-territorial language as the first language among non-Russian assimilants; e.g., 619 Jews in the Lithuanian S.S.R. speaking Lithuanian as their first language. (3') characterizes the status of a (non-Russian) co-territorial language as the second or third language of bilinguals; e.g., 4,579 Jews in the Lithuanian S.S.R. having Lithuanian as a second language.

3. Language planning, prestige planning, and the limits of planning activities It is obvious that there are close links between language planning and the ethnic identity of the related speech community. Most scholars who have been working in the field of language planning tend to assign to language an exclusive role of shaping a community's ethnic identity. This seems reasonable when investigating problems of language planning, but one has to bear in mind that language is an elementary, not an exclusive, criterion of ethnicity (see examples of nonlanguage related identity in chapter 2). Such a broader understanding of the role of language in ethnic processes is necessary to assign language planning a specific role in the self-identification of speech communities as well as in the categorization of these communities in the eyes of outsiders. Concentrating on the language aspect in ethnicity, however, one can assume that any kind of successful planning of the national language has to be seen as a reflection of the community's ethnic identity. This simple statement may be easily illustrated in terms of a negative delimitation. Whenever language planning activities fail, it is because of a lack of ethnic identity in the speech community. One such example with respect to a minority group on the verge of total incorporation shall be discussed in the latter part of this chapter. As language planning is always related to the phenomenology of a group's identity, it seems reasonable to discuss the implications of such a relationship. There are, in fact, several aspects of this relationship which have so far either been underestimated in their relevance or simply neglected. It cannot be the aim of this discussion to specify the whole network of ethnicity related implications in language planning; for such a task, more elaborations are needed than can be provided here. I should like, instead, to comment on the importance of a specific string of features by which ethnic identity and language planning are directly linked. Although this string of features is sometimes referred to in connection with active language planning or scholarly research in the field, the longrange effect of the cluster of criteria has not yet been fully outlined. With that aim, the string of features will also be assigned its place in the system of ecological relations.

84

Language in ethnicity: a view of basic ecological

relations

Language planning is comprised of a cluster of concepts; as each of these phenomena is quite complex, one is not astonished that a great number of scholars have already tackled various individual problems of language planning processes, focusing on linguistic, sociological, societal, and political implications. Although some overviews exist (e.g., Wood 1980 as an example of a comprehensive approach), an annotated bibliography which could provide a panorama of worldwide activities and research in this field is absent. The term language planning is often employed ambiguously, and there is a need, therefore, to discuss methodological aspects of the concept and terminological usage. In this connection, I shall lay emphasis on some oft-neglected principles in the sociology of language. Language planning emerged in the 20th century, although there are instances of planning activities (exceeding mere cultivation aims) in earlier periods. One could mention the successful attempt made by a group of scholars in the 15 th century to provide the Korean language with a script (called Hangul) independent of Chinese characters. The emergence of a number of new written languages in Europe in the course of the 16th century was mainly due to the stimulant idea of the Protestant movement to promote the vernaculars and to provide them with a written standard. In some respects, this large-scale cultural innovation could be considered language planning (see Haarmann 1983a, 270ff., for details of this development). There was also the efforts among certain writers and scholars in the early 19th century to create a unified SerboCroatian literary standard. This initiated a strong unitarist movement. With regard to this process, here, too, features of planning are evident. One must stress that language planning in the 20th century first developed under the conditions of newly industrialized societies (especially in Europe). Language planning processes in the developing countries of Africa and Asia are a comparatively modern development. Language planning, in the initial phase of emergence as a societal movement promoted by statai authorities, was typically a European phenomenon. The modernization of Turkish society under the leadership of Ata Türk after World War I was a movement to adopt Western European standards. One consequence of this renewal was the decisive step to abandon the Arabic writing system and adopt the Latin script for writing Turkish. This shift had much more significance than the mere exchange of writing systems; it has to be categorized in terms of state ideology. The modernization of Turkish was an integrated component of Turkey's planned Europeanization. Farther to the east, large-scale language

Language planning, prestige planning, and their limits

85

planning for a variety of languages was effected for the first time in the Soviet Union, where since the early 1920s long-term policies were applied which affected the corpus (alphabet, literary standard, etc.) as well as the status (official use, medium of instruction, etc.) of many languages (e.g., Finno-Ugric, Turkic, the Iranian languages, etc.). The beginning was, as in Turkey, marked by a general movement toward Europeanization. The adoption of the Latin script as the basis for the writing system, a step which for Lenin was equivalent to a "revolution in the East" (referring to the Asiatic regions of the Soviet Union), was a reflection of the European tradition of social ideas. The idea of Latinization, the basis of Soviet language planning in the '20 s (see Isaev 1979), had its roots in the thinking of the 19th century. It is striking, however, how little even qualified researchers in the field of language planning know of these early stages of planning activities in Europe or in Asia (under European influence). Pool (1979, 5), in a contribution about language planning and identity planning, gives, at the very start, a misleading impression: " A few years ago both terms in the above title would have been novel". He claims that only recently has language been thought of as an object of planning. Such a statement leaves out the comparatively long tradition (more than six decades) of theoretical and empirical approaches to language planning in Europe and thus ignores an important part of the history of such processes. At this point, let me elaborate some features of language planning and its effects, especially exploiting the long-term experiences of Europe and the Soviet Union where planning activities have been vital for several generations. The components of the cluster of concepts that make up language planning can most adequately be illustrated by applying principles of language ecology. One main purpose of a systematic overview of ecological relations (see chapter 1) is to differentiate complex environmental influences by singling out special variables and functional ranges. In a general sketch, there is no position for a variable like "language planning" because this phenomenon, in terms of language ecology, is not a single ecological factor but a cluster of ecological relations. In Haarmann (1980b, 209ff.), single members of this cluster are specified and an ideal typology of planning activities is illustrated. Applying such a systematic framework of ecological relations reveals some relevant conceptual différenciations. The consequences are also relevant for language planning as a field of study. In the following table 17,1 have tried to illustrate the ecological implications of language planning processes while also includ-

Language

86

in ethnicity:

a view of basic ecological

oo e CA " 2 u e 00 5 c .2 « α « ^ c

relations

prestige

planning

2P 2

.2 3, " S β £ 3 οο •Η 3 ο ο οο ε ε

^ I •β β

S3

S

3 g op £ •ε « J Λ

as o

*¡jj

•S

>

2 Π

* °

-J

O « Ξ

i

u S g· δ-

α ι 3 ·. O o

> 00 J) β 00 • 3 Cd

S ε cd

3

oo ε

:

E

S

I

">

M §· ο

Έ 2 -α 3 g1 « § I

cd

u iô SP c 3 .2 0 0 CA

M

c .SSP

¿2 ε



« .a •S ö O.

O (A a cd

,

«J « S i 0) Ϊ 3

ε

2

g

g

5

"S.

υ οο cd

I 3 cd u >· ü á I

«> f

i• 5 « «

•a o gS

C S

Β O —

β O„

3 0«

£ 'ôô u O •S o o. o

VÌ o

00

•S 3 3 Β JS

cd Ή

ω

g

s

I g » .rtΞ . S ~ e

α> ' S

1 J o o. •a u cd op υ cd

cd 00 ε _ 'ε ο- S ε -2 α

» S 2 c

cd 'S S 3

O 3 O . cd « Χ cd £ 3 i: S 8 s 3 cd ·« U » « . • S c I s >

Language planning, prestige planning, and their limits

87

ing features of language planning that have heretofore been neglected. Language planning is displayed as a mosaic of different ecological variables, functional ranges, and focusing centers. For the purpose of a general overview, not all ecological variables composing language planning processes can be mentioned. One disadvantage in the arrangement of components given is they are coordinated and subordinated. The relations between single components and functional ranges would be best indicated using a three dimension graph. The distinction between the spheres of language corpus planning and language status planning is widely known and is based on the classical différenciation of language planning activities proposed by Kloss (1969 b, 81 ff.). According to this view, language planning can focus on the corpus on the one hand (referring thus to the linguistic structure and the elaboration of norms for a written standard, functional styles, etc.) and on the status on the other (with respect to the political status of a language and its functional ranges). The categorization proposed by Kloss was quickly adopted (see Fishman 1971 a, etc.) and is now a firm component of sociolinguistic terminology in the field of language planning. The term "prestige planning" has to be considered a new item in connection with language planning, although it has some relation to previous attempts to establish an interdependence between language planning and its so-called secondary effects (see Lamy 1979 for the role of identity in planning processes), which, traditionally, are not included in planning perspectives. Indeed, features like self-categorization and categorization of others (referring to ethnic groups as well as to their languages) have not been adequately dealt with, having been neglected, it seems, as factors of secondary importance. When specifying variables of language ecology (see chapter 1), it becomes clear that ethnopsychological factors cannot be treated as being dependent on the effect of language planning activities, but have to be analyzed as equivalently independent variables which reveal some special language-oriented relations in the range of language planning processes. What is described here as prestige planning may be defined as a cluster of individual ecological variables which imply evaluations of language planning activities by the speech community. The term prestige planning could be paraphrased as "planning with regard to elements of evaluation in the ethnic identity of a speech community". Thus prestige planning is not the same as identity planning, as ethnic identity includes many other features which are not language related (see Haarmann 1983a, 322ff.). It must be admitted, however, that it is as difficult to define the term

88

Language in ethnicity: a view of basic ecological relations

"prestige" adequately in this sense as it is in the case of identity. In the early contributions, sociolinguists tried to avoid conceptual confusion by restricting the use of the term to special conditions. Weinreich (19'53, 79) proposed defining prestige as "a language's value in social advance". Until now, it has been in this sense that the term has been applied. Fishman (1964, 54), commenting on Weinreich's proposed definition, pointed out that new problems arise when attempting to evaluate or define "social advance". According to Fishman, the term prestige is needed for the analysis of language preferences or of factors linked to language maintenance. On the other hand, he stresses the dilemma which results from specifying a term whose meaning and use have been ambiguous: "It may be precisely because "prestige" obscures so many different considerations and has been used with so many different connotations that the relationship between prestige data and language maintenance or language shift data has been rather more uneven than otherwise be expected" (Fishman 1964, 54).

This warning of Fishman's with respect to the ineffective use of prestige should always be taken into consideration. In a recent contribution, Neide (1982) analyzed attitudes involving prestige in a bilingual urban community in western Europe (German speakers in Belgium). In this case, as in similar examples of language contact in multilingual settings of western Europe, the U.S., or Canada, the word prestige may appropriately be used in explaining language preferences, leading sometimes to language shifts. It should be pointed out, however, that an interpretation such as Weinreich's must not be assumed to be valid for multilingual settings in general. There are other conditions of language contact in the areas of southeastern Europe, the Baltics, the Caucasus, Soviet Central Asia, China, Brazil, and other parts of the world; in many of these settings, the application of the term prestige in its conceptual orientation as "social advance" would not be appropriate. The evaluation of prestige features differs considerably in the settings just named from those Neide (1982,141) referred to in western Europe: "In a multilingual area the use of one language will always imply more prestige". One has to take into consideration the fact that the concept of prestige is a different one in settings where socio-economic conditions, traditions, and/or stable (/. e., impenetrable) social barriers do not promote an ideology of social advance. There are also ecological settings in western Europe which lack the language preference attitudes usually associated with the concept of prestige. One example would be the social

Language planning, prestige planning, and their limits

89

isolation of the Gypsies, for whom group solidarity seems to be more important than social advance. At the same time, the Gypsies are firmly restricted from possibilities of social advance because of the wide-spread prejudice members of this ethnic group have to face when in contact with co-territorial ethnic groups. More appropriate for the purposes of a comprehensive view of language planning processes in terms of language ecology is a consideration of the concept of "prestige range". Prestige range consists of a network of identity elements and evaluations, including many stereotyping components of self-identification and categorization of other ethnic groups and their languages. The term as applied here refers to the whole mechanism of categorization which controls intragroup and intergroup relations. On the basis of such an overall concept, "prestige settings" would have to be specified within the framework of concrete ecological settings. This is also true for the definition of prestige planning, which as a special sector (or component) in language planning processes may have a decisive effect on the results of corpus or status planning. What is important with regard to the concept of planning is one should not mistake prestige planning for planning activities which aim to consolidate prejudices or stereotypes, thus reinforcing overreactions of the collective support system in different ethnic identities. On the contrary, the general aim of prestige planning must be to overcome stereotyping categorizations which often include intolerance, mistrust, or even elements of hostility. This means that prestige planning should function as a kind of control for corpus and status planning activities in order to keep up a balanced planning level (thus neither unilaterally promoting a minority language nor unilaterally fixing the status of a dominant language in terms of prestige). Ideally, prestige planning in a setting of contact between a minority language and a dominant language is balanced so that the potential circumstances of ethnic friction and conflict can be reduced to a minimum. Such a general perspective of prestige planning may seem modest at first sight, but according to the many experiences and results of language planning activities in different multilingual settings at different times, such a goal represents an ideal balance which is very difficult to achieve. As most planning activities tend to be partial (in unilaterally promoting the language the planning is focusing on), the planning process often carries the risk of creating new sources of friction in the future between the members of the promoted community and members of other communities. Overcoming stereotyping categorizations is not equivalent to the

90

Language in ethnicity: a view of basic ecological

relations

destruction of basic elements in a nationality's collective support system. It is not enough to unroot ethnic prejudices or stereotyping identifications in terms of propaganda, thus leaving behind a vacuum of frozen ethnic frictions. Such an approach can be found in some multinational socialist states where the government attempts to resolve ethnic conflicts by labelling them historical relics of a bourgeois society and therefore nonexistent in an advanced socialist society. Instead, prestige planning itself has to provide the ideological background for a setting where harmonious interethnic contacts outweigh any steady friction or opposition. This may appear to be an ideal aim, too, but it is the only genuine rationale for successful planning. Prestige planning as it is described here has to be considered as a delicate sector in planning activities which is extremely difficult to handle, and any language planner has to be very sensitive when integrating corpus and status planning items in the framework of prestige planning control. The preceding explanations indicate that I propose a différenciation of corpus planning, status planning, and prestige planning as a part of future perspectives in the field of language planning processes. To my knowledge, there is no instance in which the active application of all three of these aspects or ranges of language planning have occurred in a concrete setting. The unilateral promotion of the modernization of the Turkish language in the 1920s succeeded in transforming that language into a vehicle of the modern world, but at the same time, language planning failed to reduce the ethnic frictions between the Turks and the Kurds, whose language has been deliberately neglected in Turkey. This is only one example of many where language planning activities not only show a lack of a broad, far-sighted perspective, but also tend to be evaluated as generally successful in the narrow view of some sociolinguists. Before trying to apply prestige planning as a practical means of control in language planning activities, one has to be aware of the theoretical findings in the field of ethnicity related research. Some recent achievements in the discussion about ethnic identity must be included in the conceptual framework of prestige planning. There have been some theoretical contributions concentrating on ethnicity problems which provide some sociological insight into intergroup relations in multilingual contact areas. With special respect to contact settings in modern industrialized societies, one has to stress the importance of multiple identities (i. e., ethnic, social, political identity) as components of potential prestige planning activities. The identity of a speech community is not only shaped by features of ethnicity (intragroup solid-

Language planning, prestige planning, and their limits

91

arity on the basis of an ethnically specific network of categorizations), but also by other components such as the existence of cultural or political organizations to promote the community's interests, the group's integration into the network of social security in the state, as well as the group solidarity in the environments of daily work, etc. The promotion of multiple identities could be considered a major target of prestige planning activities. "It seems reasonable to assert that patterns of multiple identities will stimulate the maintenance of ethnic loyalties on one hand, but also weaken conflicts on the other" (Allardt 1979, 40). Prestige planning could be exemplified as a future activity for a great number of concrete settings where ethnic friction still prevails. For the purposes of this contribution, however, I will concentrate on the elaboration of settings in two different countries, namely France and the Soviet Union. The ethnicity related movement in Western Europe, which may be labeled "ethnic revival" (see Allardt 1979) or "renaissance des cultures régionales" (see Petrella 1978), has met with some special problems in France because the prestige of the language of the state (French) has been based until recently on the strong patois ideology. Language politics in France concentrates mainly on the assimilation of the peripheral ethnic minorities by maintaining certain clichés regarding the regional languages. One of these clichés involves the concept of patois (see Fourquet 1968), the central element of French chauvinism. Patois has the negative connotation of an "uncultured language variety", a judgment which differs from the functions of an L-variety (low variety) in a diglossie setting. According to prestige standards in France, speaking a patois is not equivalent to speaking an L-variety (colloquial language use, informal style) because patois implies a low level of prestige compared with the status of the H-variety (high variety). Η-variety in France refers only to French in all parts of France. Patois has the value of an ideological concept in those situations where not only local varieties of spoken French are labeled as such, but also languages of ethnic minorities (e. g., Occitan, Catalan, Breton, Basque, Flemish, Alsatian, Corsican) are called patois. The low prestige attributed to patois and their speakers is not only linked to intralingual polarities like the functional distribution of L- and Η-varieties in the French language, but is extended also to interlingual polarities like the function of regional languages as L-varieties versus French as the Η-variety. The case of intralingual distributions of varieties would represent settings of

92

Language in ethnicity: a view of basic ecological

relations

in-diglossia, whereas interlingual distributions would be considered as settings of out-diglossia (see Kloss 1966, Haarmann 1983 e on the term "diglossia"). Fishman (1964) has pointed out that prestige related evaluations do not always directly refer to features of the foreign culture, but they often crystallize in the form of language stereotypes: "However, in multilingual settings, particularly in those in which a variety of "social types" are associated with each language that is in fairly widespread use, languages per se (rather than merely the customs, values and cultural contributions of their modal speakers) are reacted to as "beautiful" or "ugly", "musical" or "harsh", "rich" or "poor", etc. Generally speaking, these are language stereotypes (Fishman 1964, 60)".

This tendency is easily discerned in the "patois settings" in France. In the categorization of people from northern France, that is, the Occitans, Bretons, or Alsatians, they are looked upon as patois-speakers who live in the "provinces" (equivalent in its usage to a "remote, underdeveloped area far from French cultural centers") and speak "bad French". Such an evaluation has for a long time exerted a strong influence on the selfidentification of minorities in France. There has been a growing feeling of inferiority among many speakers of local languages because of their categorization by the French people as patois-speakers. This feeling could be called a language-related collective inferiority complex. The local language is often considered by the speakers themselves to be an obstacle to social advance or even a stigma (see Schlieben-Lange 1971, 177 ff.; Alvar 1976 for details about such attitudes). The attitude, or patois ideology, as I shall call it, of the French, has its roots in the ideology of the national language, which has been promoted since the French revolution (see Balibar-Laporte 1974 for early stages of this development). The language situation as it is in France precludes the possibility of elaborating written standards for the regional languages. Yet, setting aside the principle of the French national language as the only cultural vehicle in modern France, there is no practical reason to believe that the coexistence of French and other written languages in France would lead to confusion or would increase tensions among the different ethnic groups. On the contrary, the kind of ethnic frictions that prevail in France today have not been created by the existence of different written languages, but by the fact that other written media have been stripped of their prestige by promoters of the patois ideology. It is reasonable to provide regional languages like Breton, Catalan, Basque, Occitan, and

Language planning, prestige planning, and their limits

93

Corsican with a modern written standard. The attempts to elaborate these media as ausbau languages have a long history. Especially impressive is the development of modern written Occitan, which is used in southern France by a dialectally highly diversified speech community. Kremnitz (1974) has analyzed the various attempts since the nineteenth century to create a modern written standard for Occitan. The patois ideology of French nationalists did not prevent the creation of written norms for regional languages, but it had a negative effect on the sociocultural potential of these standardized languages. Sociocultural potential has been elaborated and exemplified by Kloss (1978a, 46ff.), who based his explanations on a typology of a written language's applicational range. H e explicitly stresses the importance of nonliterary prose for the stability and the prestige of a standard language. And it is just this applicational range which is poorly developed among the regional languages of France. Whereas French is used for every written purpose, other standard languages are scarcely used, for example, as a medium for the daily press or scientific writings. At the same time, these languages are only used as a written means by a small elite in the speech community (mostly by members of the academic world). Corpus and status planning with respect to the minority languages in France has practically come to a standstill, or has reached the limits of its effectiveness, because of the numerous barriers created by the widespread belief that such efforts will be fruitless in the face of the patois ideology. Prestige planning as a factor of control in planning activities would have to effect some kind of balance for the situation in France, so characterized by its extreme imbalance. Much effort has to be made to dissolve the patois ideology, an obsolete overreaction, it seems, of French nationalism. In this connection, language planners would have to address French mother-tongue speakers as well as the many members of the regional speech communities who suffer from the inferiority complex imposed on them by the patois ideologists. While prestige planning should aim at diffusing the patois ideology as an active attitude among French mother-tongue speakers, it should, at the same time promote, a consciousness of the cultural heritage and the regional languages as an ethnocultural pattern among non-French communities in France. Such far-reaching prestige planning is not partisan planning, which promotes only the interests of the local-language speakers, in potential confrontation with the patois ideologists; correctly, prestige planning in France must cover the whole panorama of diversified interests in all speech communities.

94

Language in ethnicity: a view of basic ecological

relations

Such an overall attempt of prestige planning may seem extremely difficult and may have only a slight chance of success. No serious language planner would agree that the conditions for planning activities are favorable in France. Yet, obviously little progress can be made if one clings to a belief in easy successes. Prestige planning, under extremely difficult conditions as in the settings of France, necessitates, also, a sophisticated means of argumentation to cope with the chauvinism among patois ideologists. One argument in this connection would consist in stressing the fact that the prestige of the French language will not increase by maintaining an artificial prestige barrier toward the regional languages of France but, on the contrary, will only keep regional activists on the alert and incite strong resentment toward the French and French culture in those regions. It cannot possibly be in the interests of French cultural policies to create an image of French as a colonial language in its homeland, France. The French elite in particular should be challenged by prestige planners with such an argument. The aim of such activities would be the creation of a movement in the French community itself to promote an anti-patois attitude. A change like this in the general attitude would prove to be the most favorable background for planning activities with respect to the elaboration of the regional languages' functional ranges. The special view presented here may appear to be some sort of ideal vision with little prospect for success. Nevertheless, I consider such a view to be a positive alternative to the confrontation which has dominated until now and which has, indeed, proved to be the greatest obstacle to balancing the functional ranges of the languages in France. Confrontation, which has been treated as a principle of planning strategy among politically and culturally active regionalists, is, in fact, not a strategy but an energy-consuming, ineffective tactic which only creates mutual resentment and ill will among both regionalists and patois ideologists, alike. The necessity of including principles of prestige planning in language planning activities may also be illustrated for settings in a socialist society, namely, the Soviet Union. Phenomenology, a term used by Fishman (1977) in his definitional approach to ethnicity, occurs in its variants "self-identification" (Selbstidentifizierung) and "categorization of others" (Fremdkategorisierung) more or less openly among ethnic groups in democratic societies {e.g., settings in France). But in the case of the Soviet Union, with its totalitarian regime, features of phenomenology are overshadowed by Soviet language-planning propaganda. This means that the ethnocultural stereotypes, language clichés, and other elements

Language planning, prestige planning, and their limits

95

of evaluation which shape the ethnicity in the more than 130 speech communities, are not publicly articulated and can only be traced indirectly from tendencies in language behavior. Basic features of phenomenology among non-Russians can often be identified by observing the willingness or reluctance to adopt Russian as a primary or secondary (first or second) language. Such attitudes are related to variant structures of bilingualism which can be illustrated on the basis of ethnostatistical data (see Haarmann 1985 b for ecological settings in the Soviet Union). In the early period of Soviet rule, language planning was based on the principles of Lenin's nationality politics and was characterized by a high degree of democratization throughout the 1920s. Although it has been claimed that the Leninist principles of language planning have been fully preserved to the present day (see Lewis 1972 for a general outline of Soviet language planning activities), there has been some change in the actual policy, especially with respect to minority languages (see below). There can be no doubt that the initial impact of Soviet language planning efforts has had a long-lasting effect on the consolidation of languages with well-established written forms like Ukrainian, Georgian, Armenian, and Uzbek, and the elaboration of a written standard for a number of less widespread languages (see Isaev 1979 for the changing strategies to apply the Latin and Cyrillic scripts). Corpus and status planning activities have been integrated in an overall scheme of planning in Soviet society (see Haarmann 1983 a, 203 ff.). The implications of societal needs in language planning aims resulted in a special task for Soviet language planners; namely, to introduce the concepts of Marxism-Leninism into the lexical structures of the many languages and to extend the societal functions of Soviet languages. With respect to the newly written languages in particular, efforts were made to create a Soviet-internationalistic inventory of terms for different sectors of public life. Specifically, some objectives of language planning in this framework of internationalizing the vocabulary of the various languages have been incorporating Marxist-Leninist terminology in the social and political fields (including the social sciences), military terminology (including terms of the arms-producing industries), and lexical inventories for different areas of technology, for higher education, and for special professional and scientific fields. Russian loanwords like socializm, kommunizm, sovet, kollektivizacija, front, divizija, technikum, and others are found in the vocabulary of most languages in the Soviet Union. Normative dictionaries for a number of lesser standard languages with a limited functional range comprise thousands of internationalistic terms of Rus-

96

Language in ethnicity: a view of basic ecological relations

sian origin. The intensity of Russian influence, which predominates over tendencies to coin neologisms on the basis of indigenous lexical elements by means of word formation, is extremely high among the lesser languages because not only single terms but whole inventories of terms were introduced to expand their lexicons. The areas of vocabulary which refer to the conditions of modern socialist society have been extremely Russianized in many languages. The internationalization of languages under the special aspect of social functions in Soviet society had much impact on the modernization of standard languages, but had little effect on the Sovietization of non-written languages. Although there are a number of Soviet neologisms in the vocabularies of languages like Izhorian, Vot, Liv, Veps, Karelian, Lappish, Karaim, Tofalar, Itelmen, and other non-written Soviet languages, the amount of new terms cannot compare with the extensive inventory of Soviet expressions in the written languages. Russianisms have a weak representation in the socio-political as well as in the technical vocabularies of the non-written languages because most have not developed special societal or technical terminologies. The non-written languages have a restricted functional range; many societal functions typical of written languages cannot be realized by spoken vernaculars. Most mothertongue speakers of Izhorian, Karelian, or Itelmen do not use these languages as a means of communication during their daily work or in other domains of public life. For most communicational purposes, these speakers use Russian. Whereas Russian has taken over in most of social functions and become integrated in the structure of national-Russian bilingualism among speakers of Izhorian, Karelian, Itelmen, and others (see Haarmann 1985 b for variants of group bilingualism in the Soviet Union), their mother tongues do not possess lexical inventories for social functions which they do not perform. In some cases, not only Russian but other contact languages play a limited role in the modernization of non-written languages. Liv, for example, has adopted a number of Latvian terms to refer to aspects of modern life. Further discussion on this subject can be found in Haarmann (1983b, 389ff.). While Soviet language planners hail their effectiveness in accounts about successful achievements, the special development as well as the special problem of non-written "mini-languages" stay outside the mainstream of planning activities and, thus, remain in the shadow of efforts directed to the standard written languages. The only partial modernization of Soviet mini-languages itself illustrates a basic lack of flexibility in Soviet language planning. Prestige planning is worthwhile only in con-

Language planning, prestige planning, and their limits

97

nection with the dissemination of propaganda stereotypes about the brotherhood of Soviet nationalities and the equality of Soviet national languages, and, in addition, with respect to planning activities concerning the promotion of national-Russian bilingualism in school education. All these activities overshoot the needs and real problems associated with the ecological conditions among small ethnic groups and their nonwritten languages. School education is dominated by Russian and there is no educational means to further language skills for the next generation of mini-language speakers. The small speech communities are standing in the cross-fire of forces that have been intensified in modern Soviet language planning. An important element of modern Soviet nationality politics is the doctrine of "mutual inclination of nations and languages toward each other" (Russ., sblizenie nacii i jazykov), which aims at shaping one "Soviet socialist nation" (Russ., sovetskaja socialisticeskaja nacija) in the future. The related activities of language planning to achieve this goal have only had a negative effect on the small ethnic groups as the assimilation pressure exerted by the Russian language has increased. This pressure has brought about a tendency of displacement (i. e., active dissolution of ethnic identity) among most small nationalities as more and more members shift to Russian as a first language. On could speak here of a Janus-like strategy of Soviet language planning. On the one hand, there is the declared will to promote all non-Russian languages in Soviet society. And, indeed, there were attempts to create literary standards for mini-languages like Izhorian, Veps, Karelian, and others in the 1930s. Such attempts failed, however, partly because Soviet planners did not succeed in making the standard languages popular among the speech communities. On the other hand, there is the aim of an overall integration of all Soviet nationalities, which, in practical terms, means the dynamic Russification of all non-Russians, and those most vulnerable among them are members of the mini-language speech communities. The ecological conditions described regarding mini-languages in the Soviet Union illustrate a broad potential range of prestige planning activity which is not covered by Soviet language planning. It is not covered because there are ideological barriers and these barriers prevent effective planning activities for the special needs of small speech communities. Leninist ideology of integration of nationalities into one societal body does not imply, logically, "integration equals Russification". It would be reasonable and in complete accordance with official Soviet ideology to promote a movement for the preservation of mini-language

98

Language in ethnicity: a view of basic ecological

relations

groups and their cultures. An initial task of such a movement, the aims of which would not contradict the principles of Leninist nationality politics, would be the strengthening of ethnic identity among members of small nationalities. The integration of all Soviet nationalities is an additional component which requires the extension of ethnic identity to a multiple identity including a supranational element. Forcing Russification upon non-Russian nationalities is a misunderstanding (or a deliberate falsification) of Leninist principles on the part of Soviet planning ideologists. Although theoretically there is much space for planning activities which would promote the prestige of small languages and their speakers, practical Soviet policy has for some decades definitely followed a rigid course of Russification. The potential advantage of promoting a concept of multiple identities, which seems to be the most advanced and most adequate means of avoiding ethnic frictions in modern industrialized societies (see Haarmann 1983 d, 31 ff.), has either been rejected or not even taken into consideration by Soviet ideologists. Although Russification is the dominating characteristic of Soviet planning processes, one must admit that the needs of local written languages are respected (at least as long as any demands by non-Russian nationalities do not interfere with the larger aim of establishing a uniform system of communications). In the case of non-written languages, however, there is a clear attitude of passive tolerance but not active promotion on the part of Soviet language planners. Under present conditions, a movement for the promotion of a balanced multiple identity, including such components as self-identification and the categorization of others, cannot be established in the mini-language communities because it would have to be organized in dissociation with the State (i. e., outside State control). But the organization of a mass movement outside the control of the government has no chance of success in the Soviet Union. The unilateral promotion of Russian-orientated priorities in language planning has created a paradoxical situation. On the one hand, the Soviet Union is the country with the most extensive regulations and means of language planning politics (compared with other European states), yet on the other hand, it is this country which reveals, comparatively, the strongest possible impact of the language of the state (Russian) on its ecological settings. Kreindler (1982 a, 27) asked the question whether the Russian language will be able to fulfill its assigned role as a language of integration. In my mind, it is most doubtful that the aims attached to this role will be ever achieved. There was a time when the Czarist administration

Language planning, prestige planning, and their limits

99

promoted assimilation among non-Russians (end of 19th, beginning of 20th century). The long-term results proved to be contrary to the intended goals; increasing nationalism among non-Russians was the result. Similar developments seem to be in progress nowadays in the Baltic region (especially among the Estonians and Lithuanians), as well as in Soviet Central Asia (among, for example, the Uzbeks, Kirgiz, Turkmens, Tadzhiks, and other nationalities). Some examples of settings in countries with differing societal systems have been discussed, clearly illustrating a deficiency in prestige planning as a crucial part of planning processes. Many other ecological settings could be mentioned here which are also characterized by a serious negligence of prestige planning. Based on the principles of a three-fold distinction between corpus planning, status planning, and prestige planning, proposals can be articulated for active prestige planning in government policies toward languages in a number of multilingual societies throughout the world. Prestige planning has to be undertaken as a special component of overall language planning. Prestige planning warrants broad application because it can function as a strategy of activating those attitudes toward ethnicity which are likely to weaken ethnic conflicts and create patterns of multiple identities in speech communities. Since prestige stereotypes are mostly associated with a community's language rather than the community itself, prestige planning relates directly to activities of corpus and status planning. With respect to scientific research, prestige planning may best be regarded as a subject of contact linguistics, with close relations to theoretical and methodological concepts of language ecology. Language planning in general and prestige planning in particular have natural limits about which little is known, even among experts in the field. Any kind of planning is subject to the demographic factors of a speech community. In many cases of successful planning, the size of the target community does not seem to be of importance; this results primarily from the fact that the speech community is strong enough to guarantee an active response to planning measures. Contrarily, many languages of small ethnic groups remain in the shadow of planning activities because the severely limited number of speakers cannot assure a successful reception of planning aims in the community. Several languages of the Soviet Union mentioned previously are relevant examples. In the following discussion, one from among the mini-languages referred to shall be introduced along with its special problems of survival. The point here is

100

Language in ethnicity: a view of basic ecological

relations

ethnic processes, like those among Izhorians, reveal conditions which are likely to blockade language planning activities to a decisive degree. What are the implications of these processes and the underlying conditions? There are many aspects of the Izhorians' communicational needs as an ethnic minority in a multilingual contact area. In an area like Ingermanland, where different ethnic groups live together in neighboring monoethnic or in ethnically mixed settlements, language contacts are determined by the interactional conditions of intra- and intergroup relations. On the one hand, there is the desire to overcome language barriers for the purpose of interethnic contact (the principle of intergroup communication). Language shift, as a phenomenon of assimilation in connection with the change of generations, and the adoption of second language knowledge are dynamic factors in the process of transcending communicational barriers. On the other hand, there is the need among the Izhorian speech community to retain conditions of internal communication and to identify themselves as Izhorians (i. e., non-Russians). This latter need is served by language maintenance as a means of support of the community's identity and for the purpose of internal interaction (the principle of intragroup communication). Language maintenance and self-categorization are, however, very weak in the Izhorian community. The number of Izhorians able to speak their national language exceeded 400 persons according to the Soviet census of 1970, but had fallen below 400 by 1979. The figures include persons who have adopted Izhorian as their first (or primary) language and those who picked it up as a second(ary) language. Many Izhorians who have adopted Russian as their first language also acquire second language knowledge of Izhorian under the special conditions of ethnically mixed families, in which one spouse is Izhorian, the other Russian. Most Izhorians (71% of the entire ethnic group) speak Russian as their first language; less than 27 percent preserve Izhorian as a primary language. As I have already pointed out in connection with the analysis of multilingualism in the Soviet Union, the terms primary and secondary language refer to the acquisition of knowledge but not to the predominance of use in daily interaction. Regarding the latter criterion, Russian is by far the more dominant language used by all Izhorians, because it covers most of the functional ranges of everyday communication. The predominance of Russian in the Izhorian speech community is evident not only in the number of assimilated Izhorians, representing almost three quarters of the entire population (see above), but in the nature of bilingualism among izhorians as well. Below I reproduce a profile

Language planning, prestige planning,

and their limits

101

Tab. 18. Villages with Izhorian-speaking populations in Ingermanland around 1960

Tab. 19. Villages with Votic- and Izhorian-speaking populations in Ingermanland around 1960

of Izhorian bilingualism, which, together with other profiles of this kind, was discussed in Haarmann (1979a, 271 ff.). Language maintenance is generally low, but varies considerably from group to group within the population. Whereas 31.5 percent of the female rural population has preserved Izhorian as a primary language, only 19.8 percent of the male urban population has done so. This comparison, together with other

102

Language

in ethnicity:

a view of basic ecological

relations

findings, indicates both that Izhorian is better maintained in the rural areas and that it is better preserved by Izhorian women than men. A total of 84.5 percent of those Izhorians who have adopted Izhorian as their mother tongue are bilingual, speaking Russian as a second language. There is considerable variation between men and women also with respect to the level of Russian second language ability. Women obviously have a better knowledge of Russian than men (note higher rates for women in Table 20). The contrast is most striking for the rural population, where Izhorian women represent the highest figure for Russianlanguage ability (88.8% of the mother tongue speakers have adopted Russian as a second language), and men exhibit the lowest proportion (75.0%) relative to all groups in the Izhorian speech community. What has been said about the Izhorians so far is also more or less true for the Veps, who live partly in the district of Leningrad, partly in the district of Vologda, and partly in the Karelian A.S.S.R. The situation of the Veps under Russian influence, however, is not identical to that of the Izhorians. Specifically, the bilingual profile of the Veps, illustrating their level of language maintenance as well as the extent of Russian as a second language in the speech community, reveals the same principal tendency toward assimilation as in the profile for the Izhorian community, but the Veps profile also makes clear the existence of ethnic-specific

Tab. 20. The status of Izhorian and Russian among the Izhorian population (see Haarmann 1979 a, 309)

Izhorian population

Language status Izhorian as a first Russian as a second language language

Total population

26.6%

84.5%

male population

25.2%

76.7%

female population

27.1%

87.3%

Total urban population

22.3%

84.2%

male population

19.8%

79.2%

female population

23.2%

85.7%

Total rural population

31.4%

84.8%

male population

31.0%

75.0%

female population

31.5%

88.8%

Language planning, prestige planning, and their limits

103

Tab. 21. The status of Veps and Russian among the Veps population (see Haarmann 1979 a, 308)

Veps population

Language status Veps as a first Russian as a second language language

Total population male population female population

34.3% 28.7% 37.6%

94.9% 94.1% 95.2%

Total urban population male population female population

34.7% 26.0% 39.5%

95.8% 93.8% 96.5%

Total rural population male population female population

33.7% 31.7% 35.1%

93.6% 94.3% 93.2%

rates for each section of the population. The graph (e.g., Table 23) is from Haarmann (1979a, 304). The language situation in the Soviet Union as it has been partly illustrated by the bilingual profiles reveals strong assimilation, tending toTab. 22. Correlation of rates for Izhorian as a first language and Russian as a second language in different groups of the Izhorian population % 100 87 3%

84,,5% (2)

'

*\76,7%

85 7% 84,2% ' 84,8% · 79,2%^-· ·— 75,0%.

88,8%

(2)

60

30




104

Language in ethnicity: a view of basic ecological relations

Tab. 23. Correlation of rates for Veps as a first language and Russian as a second language in different groups of the Veps population 100

94,9%

94,1%

(2)

95 2%

·

95 8%

'

*

93,8%

96 5%

-



93,6%

94,3%

··

·-

· 33,7%

•·31,7%

93,2%



(2)

.

O

60

30 •

(1). 34,3%

total

. ~~· 28,7%

37,6%

male

female

39,5%

3 5 1 %

26,0%

urban

male



r-

female

rural

male

female

ward language shift among both Izhorians and Veps. The implications of this kind of extreme assimilation (incorporation) can be specified in terms of language ecology. Based on the theoretical framework of ecological relations elaborated in the general model (see chapter 1), as well as on empirical findings from investigations of a number of small ethnic groups, the following (hypothetical) framework of extreme conditions for language shift can be established. Such a framework of ecological conditions favoring language shift in a speech community has the value of a theoretical construct. This means that in concrete settings of assimilation there may be some divergence concerning one or more of the conditions listed. The fact is, however, a majority of concrete settings are characterized by most of the specified features. When applying the above set of extreme conditions for language shift to the settings of Izhorian-Russian and Veps-Russian language contact, clearly both settings, despite their slight differences, manifest most of the hypothetical features of assimilation in progress. Nevertheless, there are some divergences. Although the most "ideal" condition for language shift would be a highly dispersed community (functional range 1), this is not true for the Izhorians, most of whom live in the rural areas of Ingria (Ingermanland). This condition does indeed apply, however, to the settings of Veps-Russian contact. A second extreme condition which is not represented in the setting of the Izhorian speech community is a weak linguistic distance of the minority language from

Language planning, prestige planning, and their limits

105

Tab. 24. Extreme conditions for language shift in a speech community Functional range of ecological variables

Special condition of ecological relations

(1) Ethnodemographic variables

The speech community comprises less than 1,000 speakers The community's area of settlement is highly dispersed The speech community mostly settles in single diaspora groups The majority of all speakers live in urban areas The speech community is characterized by strong migration because of working conditions

(2) Ethnosociological variables

The speech community is characterized by a strong decline in the number of its members Language behavior with respect to first and second languages differs between the male and female populations The speech community comprises mostly higher-age members The speech community holds a low(er) social stratum in a hierarchical ethnocultural division of labor The speech community is characterized by a great number of ethnically mixed families

(3) Ethnopolitical variables

The speech community has no special groupsupporting rights The community's second language is the language of the state (which is the only official language) The community's mother tongue (or first language) has no official status The community's mother tongue is neither the medium of instruction nor a subject in school The community's potential for social advance is limited

(4) Ethnocultural variables

The majority of all members of the speech community are of polyethnic origin There is a weak social distance between the speech community and the dominant ethnic contact group

106

Language in ethnicity: a view of basic ecological relations

Tab. 24. Extreme conditions for language shift in a speech community Functional range of ecological variables

Special condition of ecological relations

There are no cultural or political organizations to support the interests of the speech community The community's first language has no ausbau status The community's first language only functions as a spoken medium (5) Ethnopsychological variables

The community's ethnic identity is based mainly on a strong tendency of acculturation Most members of the speech community tend to reject their mother tongue in their selfidentification The members of the speech community usually praise the second language as a language of high prestige The community's collective will to maintain the mother tongue is very weak The speech community shows a high level of interaction with members of the ethnic contact group

(6) Interactional variables

The community's interactional mobility is based mainly on the second language component of its bilingualism The mother tongue only functions as an L-variety in a diglossie setting The speech community plays a subordinate role in societal interaction The ethnic contact group is known to the speech community through regular interaction The use of the community's mother tongue is even restricted in informal (non-official) domains The community's mother tongue is used mainly for daily interaction among family members and thus its use is reduced to the home domain

(7) Ethnolinguistic variables

The community's mother tongue exhibits a relatively narrow linguistic distance from the contact language

Language planning, prestige planning, and their limits

107

Tab. 24. Extreme conditions for language shift in a speech community Functional range of ecological variables

Special condition of ecological relations

The community's interactional strategies are comparatively weaker than those of the ethnic contact group The deictic system of the community's mother tongue is similar to that of the ethnic contact group, or strongly affected by the influence of the contact language Intergroup language contacts are characterized by unilateral relations on the part of the speech community with a dominant contact language

the dominant contact language (functional range 7). Izhorian and Russian are characterized, instead, by a great linguistic distance, revealing contact between two highly different language structures. Nevertheless, most of the other conditions are typical of the Izhorian community. The application of the theoretical construct to these concrete settings illustrates an important principle; namely, that no process of language shift can be characterized by only one or a few features, but has to be explained in the context of a large cluster of various conditions which together make this process transparent. Also, this easily does away with any questions about differences of a single feature or two. Despite the deviation in some concrete conditions from the theoretical construct (see above), the basic features of assimilation are clearly revealed in the Izhorian settings. The interethnic relations of the Izhorian speech community have been characterized mostly by unilateral contact with the dominant neighboring languages. Izhorian has been, in most situations of contact, the influenced and not the influencing language (Votic and local Finnish dialects being the only contact languages exerting influence of Izhorian while also receiving influence from this language themselves). Russian has been predominant for centuries. Prehistoric contacts among the Finnic and Baltic language variants on the one hand and the Old Germanic language variants on the other, are not considered here because such contacts can only be understood in terms of language-group contacts rather than among individual languages. Laanest (1975,116ff.) describes these

108

Language in ethnicity: a view of basic ecological relations

ancient contacts and the loanwords of Baltic and Germanic origin which have survived as elements of the lexical substratum in the vocabulary of the individual Balto-Finnic languages (see Laanest 1966, 114ff. for old loans in Izhorian). Besides Russian, Estonian, Finnish, Votic, and Swedish have influenced the Izhorian language at different times. However, the survival of Izhorian as a mini-language depends exclusively on the dynamics of the Russian impact in Izhorian-Russian group bilingualism and on the structure of the Izhorian language. Regarding this impact, Russian, in both its spoken and written varieties, has influenced the Izhorian language. That is, colloquial Russian, a dialectal variety of Northern Russian, and literary Russian have been sources of impact. Russian influence can be perceived in nearly every reach of Izhorian grammar, phonology, and vocabulary. The phonological structure of Izhorian, its morphology, syntactic structure, and lexicon — all show traces of Russian influence. In Haarmann (1983 b, chapter 1), I illustrated the impact of Russian on Izhorian in a graph which is reproduced below (see table 25). The graph characterizes the contact of language varieties on a sociolinguistic level (note dialect, standard language, colloquial variety as subsystems; Germ., Subsysteme) as well as the distribution of transferential elements in different sections of language structure on a typological level (note phonology, morphology, syntax, lexicon as partial systems; Germ., Teilsysteme). Some main features of the Russian influence on the Izhorian language structure are outlined below.

The influence of Russian on the language structure of Izhorian 1.

Russian influence in the phonological system of Izhorian

1.1. Introduction of additional phonemes Iii in all positions (where it occurs) initial position: faltta, familia, fevrali internal position: lafka, opsefka, kafedra /t's/ in all positions initial position: t's'asnoi, t's'ernila, t's'esnoi internal position: prit's'ka, polut's'ka, stävet's'ka In the integration of Russian borrowings, additional phonemes like the above mentioned Iii and /t's/ were introduced into the Izhorian set of phonemes and thus extended the phonemic structure of the influenced language (see Laanest 1966, 104).

Language planning, prestige planning, and their limits

109

Tab. 25. A model of Russian-Izhorian language contacts M A C R O S Y S T E M O F T H E H I S T O R I C A L L A N G U A G E A (Russian) PART SYSTEM 2 Morphology PART SYSTEM,

SYSTEM PART,

I SYSTEM 1 PART 2

¡"SYSTEM I PARTJ

PART SYSTEM3

PART SYSTEM 4

Phonology

Nomina1

1 Pronominal

j Verbal

Syntax

Lexicon

SUBSYSTEM,

SUBSYSTEM 2

SUBSYSTEM 3 SUBSYSTEM

Standard Language

Colloquial Variety

Dialect A

4

Social Variety A

SUBSYSTEM 5 Technical Variety A

Subsystem^ and subsystem^

Subsystem , and subsystem 2 have not been elaborated

have not been elaborated

for Izhorian

for Izhorian SUBSYSTEM3

-

-

PART SYSTEM 4

PART SYSTEM3

Lexicon

Syntax

Dialect

A

-

Verbal

I Pronominal

SYSTEM PART3

! SYSTEM ¡ SYSTEM PART 2 Ι PART, A. JMorphology

! Nominal

-

PART SYSTEM, Phonology

PART SYSTEM 2 M A C R O S Y S T E M O F T H E H I S T O R I C A L L A N G U A G E B (Izhorian)

1.2. Extension of the distributional range of phonemes Extension of the distribution of Izhorian Ibi, làl, /g/ to word-initial position b—: boba, bol'nitsa, buit, butto d—: da, drästui, durakka, dessonkka g—: glad'ittâ, grëven'i

110

Language in ethnicity: a view of basic ecological relations These sounds were not introduced as new phonemes under Russian influence (note existence of consonant clusters in final position); the distributional range of the corresponding Izhorian phonemes, however, was extended to initial position under Russian influence (see Laanest 1966, 103 ff.).

1.3. Introduction of additional distinctive features Palatalization of consonants in all positions initial position: d'äd'ä, 1'ohkoiD, l'übo, t'uima internal position: budil'nikka, l'ul'kka, sentt'aBri final position: an'is' Palatalization frequently occurs with liquids and in consonant clusters with liquids (e.g., pl'üDo, pl'üd'etska, sl'upka, etc.). 1.4. Extension of the associativity of consonant phonemes Consonant clusters in initial position; examples of Russian borrowings: flakku, hräbroi, kruGa, kloppi, mrämori, plëtti, plodnikka, plüGa, prostina, prostoi, skiletti, stroppi, strumentti, trahtori, treBaDa träpkä, etc. Consonant clusters originally did not occur in initial position in Izhorian words. This possibility of consonant association was introduced into the phonemic system of Izhorian under Russian influence. Although many of the phonological characteristics mentioned above are restricted to lexical borrowings from Russian (Izhorian IV, for instance, only occurs in Russian loanwords), one has to consider the integration of those phonological features as an extension of the Izhorian phoneme system. Such an interpretation seems appropriate because of the large amount of Russian borrowings in Izhorian and their widespread distribution in the lexical structures of the Izhorian vocabulary (see Laanest 1966, 103ff., 1975, 26ff., etc.). 2.

Russian influence in the morphological system of Izhorian The impact of Russian prevails in Izhorian word formation as a special facet of syntagmatic morphology. The following types can be distinguished:

2.1. Substantive-forming loansuffixes: Izhor. -noi (Russ. -nyj) e.g., aktsïZnoi, rajonnoi, päGernoi Izhor. -noi (Russ. -noj) e.g., t's'etvernoi Izhor. -(s)sikka/-(s)sikka (Russ. -cik) e.g., izvassikka, perevossikka, pomessikka, roznossikka

Language planning', prestige planning, and their limits

111

Izhor. -(s)sikka/-(s)sikka (Russ. -scik) e.g., jamssikka, kâmensikka, oBissikka, paissikka Izhor. -nikka (Russ. -nik) e.g., konuznikka, plodnikka, tvornikka, mutnikka, peretnikka Izhor. -ista (Russ. -ist), -äri (Russ. -ar/-jar), -ikka (Russ. -ik), and other suffixes are only used in a limited number of derivations. 2.2. Adjective-forming loansuffixes: Izhor. -noi (Russ. -nyj) e.g., formennoi, jasnoi, strasnoi Izhor. -noi (Russ. -noj) e.g., korenoi, otstavnoi, rodnoi 2.3. Loanprefix: Izhor. pol(u)- (Russ. polu-) e.g., polupaltto, polussappöskaD, polZaposkaD A number of loansuffixes are restricted to Russian borrowings and do not occur in genuine Izhorian words (e.g., -noi, -ari). On the other hand, other word-forming elements of Russian origin are very productive in Izhorian lexical formation (e.g., -(s)ikka, -nikka, etc.). 3.

Russian influence in the syntactic system of Izhorian The influence Russian has exerted on the syntactic structures of Izhorian can only be pointed out here in general. Russian influences on Izhorian syntax are not surprising when taking into consideration the dominant role of Russian as a contact language and its part as a second language in IzhorianRussian bilingualism (see Haarmann 1983 b, chapters 6, 8). Further preliminary contrastive studies of Izhorian-Russian text analyses (that is, analyses of recorded spoken Izhorian compared with the corresponding Russian translation equivalents) have to be carried out in order to reach a high level of comparison. The Russian impact on Izhorian syntax, however, can easily be illustrated with respect to borrowed syntactic constituents such as conjunctions, interjections, particles, etc. This aspect of Izhorian syntax is also of special interest from the standpoint of bilingual interaction. Here, some links may be discernible for a pragma-linguistic interpretation of the Russian influence.

3.1. Conjunctions borrowed from Russian a (Russ. a) 'but'; 'however' da (Russ. da) 'and' i (Russ. i) 'and'; 'also' liBo (Russ. libo) 'or' nesto (Russ. ne cto) 'even less'

112

Language in ethnicity: a view of basic ecological relations no (Russ. no) 'but' sto (Russ. cto) 'that' ta (Russ. da) 'and' tö (Russ. to) 'or'; 'or . . . too'

3.2. Interjections borrowed from Russian amin (Russ. amin') 'amen' hoZpod'i (Russ. gospodi) 'oh my goodness' kul'u (Russ. gul'-guP) (signal for calling hens) passiBo (Russ. spasibo) 'thank you' por'a (Russ. boran) (signal for calling sheep) trastui (Russ. zdravstvuj) 'hello, nice to see you' t's'ur (Russ. cur) 'stop' (used by children when playing) urrà (Russ. ura) 'hurrah' voD (Russ. vot) 'look at that' 3.3. Particles borrowed from Russian hoD (Russ. chot') 'indeed' no (Russ. no) 'well now' 4.

Russian influence in the lexical system of Izhorian There are two main types of loans regarding borrowed lexical innovations:

4.1. Integration of a lexical innovation as a loanword e.g., Izhorian puDilnikka: Russ. budil'nik 4.2. Integration of a lexical innovation as a loan shift e.g., Izhorian näkkiäZmëZ, after the Russian sample expression, svidetel' ('eye-witness')

The Russian influence is most apparent in the area of lexical innovations of the Izhorian vocabulary. For a better understanding of the position of Russian elements in the Izhorian vocabulary, it is necessary to give some explanation about the origin and status of lexical borrowings in Izhorian. Loanwords in Izhorian can be identified as elements of different strata. Whereas borrowings like Izhorian ehki 'perhaps' (from Estonian), juhlia 'to hold a feast' (from Finnish), or surra 'to mourn' (from Votic) are components of the adstratum in the Izhorian vocabulary, Baltic or Germanic loanwords have to be classified as elements of the lexical substratum in Izhorian (e.g., Izhor. hammaZ 'tooth', kulta 'gold', and

others). From the standpoint of medieval Izhorian-Russian language contacts, the Old Russian loans are elements of the adstratum. The influence of Russian, however, has, in the subsequent centuries, increased and has exerted such an impact that the Russian borrowings constitute a

Language planning, prestige planning, and their limits

113

broad layer of the superstratum in the Izhorian vocabulary. There are over 1,450 lexical borrowings from Russian which have been integrated into the Izhorian vocabulary since the Middle Ages. The variety of Russian loanwords illustrated in Haarmann (1984 a, part II, in an alphabetical as well as conceptual order) is presently the most comprehensive collection of borrowed lexical innovations in a single non-written language of the Soviet Union. The number of loan shifts, which constitute an integrated part of the Izhorian vocabulary, has not yet been completely determined. Several variants of loan shifts have been discussed in Haarmann (1983 b, part II/7). The Russian loanwords in Izhorian belong to different chronological strata. Some borrowings are very old. Evidence from the historical development of Izhorian sounds indicates that Izhorian sövvatta 'Saturday', verBa 'Palmsunday', and prihoi 'bride', also 'young man', for example, are old borrowings (elements of the oldest layer of Russian loanwords). These loanwords were adopted in the tenth century, or at least by the beginning of the eleventh century (see Nirvi 1961,120). Elements of the most recent layer of Russian loanwords are, for instance, Izhorian jeropläni 'airplane', telifföna 'telephone', and trahtori 'traktor', which are semantically, as well as phonetically, very obviously modern borrowings. At any rate, most Russian loanwords have been adopted into Izhorian since the eighteenth century. The total number of lexical innovations borrowed directly or indirectly from Russian far exceeds 2,000 words if one includes not only loanwords and loan shifts, but also derivations from Russian borrowings. The following list contains a number of those internal derivations formed from loanwords using Izhorian wordformation elements. Processes of lexical internationalization and modernization determine, to a decisive degree, the sociocultural potential of a language to serve the needs of modern living conditions. In the case of non-written languages with a small number of speakers, these processes develop under the special conditions in relation to a dominant contact language. This means that for mini-languages such as Izhorian, Votic, Livonian, and Veps in the Baltic area, the coining of modern terms depends almost entirely on the lexical structures of Russian and its loanwords in those languages. For Livonian, Latvian is an equally important contact language. A special problem of these mini-languages is that their vocabularies have been enriched (internationalized) by international elements (mainly of western European origin) through the Russian language for centuries, but, at the same time, there has been no modernization in the sense of new

114

Language in ethnicity: a view of basic ecological

relations

Tab. 26. Derivation of Russian loanwords in Izhorian Russian equivalent

Izhorian equivalent

derivation or composition

kisa

kiZa

kalkun (dial.) kvas kopit' gorn

kalkkuna vässa koBittä korna

gruzd' ladit' muka

krusti lätia mükka

bljad' pravit'

pläDi prävittä

tërka kaplja bomba

torkka kabli pomBa

kisakko s. kiZukkain s. kalkkunakukka s. vässavilla s. koBittua v. kornä v. kornata v. omakrusti s. lätlssa v. müGaDa ν. mükkahuZ s. plättiä ν. präviDoZ s. prävihua ν. prävihussa s. torGaDa ν. kablikkain s. pomBittä v.

terms being adopted in order to differentiate the societal functions of languages in a modern industrialized society. As main reasons for the lack of modernization of the mini-languages, one could point out the small number of speakers (some hundred or a few thousand only), the restriction to the function of a home language, as well as the fact that former planning activities to provide these languages with a written standard failed. It is most appropriate, however, to correlate the lack of modernization with the entire cluster of conditions which have been elaborated for characterizing extreme conditions of language shift. To see that this is true, consider the following regarding the Izhorian vocabulary. Although the processes of modernization of the Izhorian vocabulary are weak and therefore not fully comparable with the coining of new designations for modern concepts in Estonian, Lithuanian, Ukrainian, Grusinian, or Armenian (languages with a strong sociocultural potential and interactional range), nevertheless, the question of internationaliza-

Language planning, prestige planning, and their limits

115

tion remains relevant. Russian borrowings in Izhorian compose a great number of indirect loans from the languages of western Europe. Although Izhorian has never been in direct contact with German, French, English, Latin, or Greek, its vocabulary contains words of Western civilization which were adopted via Russian. Most words of western European origin have entered Russian since the time of Czar Peter the Great (to be exact, after his return from Holland and England in 1699). Thousands of borrowings were incorporated in the course of the eighteenth and nineteenth centuries, and many of these were integrated into the vocabularies of minority languages like Izhorian, Veps, Karelian, etc. The following list (see Table 27) presents an outline of the strata of loanwords in the Izhorian vocabulary adopted from Russian, subsequent to being borrowed from languages in western Europe, beginning in 1700. So, for example, Izhorian possesses international lexical elements such as masina, muzyka, architektor, charakter, biblioteka, kilometr, klub, cement, cifra, maska, moda, and many others. Thus, even a non-written language like Izhorian has taken part in the long process of Europeanization of parts of the vocabulary common to many Soviet ethnic languages (via Russian). It might be noted that implications of this process for written languages are discussed in Akulenko (1972, 136ff.). Izhorian also offers interesting examples of international designations in a non-written language, in contrast to non-international designations in a written language such as Finnish. Finnish possesses a modern vocabulary (unlike Izhorian) to meet the communicational needs of a modern industrialized society. Yet many widespread international elements are absent from the Finnish language. The following list (see Table 28) illustrates the polarity represented by international terms in Izhorian versus non-international terms in Finnish. One main reason for this difference is the strong cultivational movement which has significantly shaped the vocabulary of the Finnish language (see Hakulinen 1969). Since the nineteenth century, language cultivators in Finland have tried to avoid the introduction of international terms by coining designations using Finnish cognate words or applying derivational means to form new, native Finnish terms. As processes of lexical innovation in Izhorian have been dependent entirely on the influence of Russian, without restrictional mechanisms of language cultivation, international terms have freely entered the Izhorian vocabulary. It has been pointed out, however, that these processes of internationalizing the vocabulary no longer have any practical effect at present for daily language use as the language is in à steady course of assimilation.

δ I

t 10 * s »» Ν " S S Sì 'S" à « i ω "ÇJ O b] β 1 rv ω β, s* χ * u cd « ex α χ

§

o X ο, >> Η

cd cd ^ to u M ja ς

p ST 2 c s-, l e^ 2 ex S 5 g

s Q

M * S g

ss S 2 β. υ -θ- Β

Β

ω

1 1 &0 s r « $ -S

1 ο à S β o

cd α. cd H α ex υ υ sí Χ Χ χ cd cd cd S S S

cd « o H

cd

C s

s η υ .

Η >Χ χ 'S υ Cd υ Ο. 2 Χ p x f - ( 0 c d J « 2 4) ρ C X f - c Ì Ì I - R t ì S u ^ e d e d O C e S cd cd c d \ 0 * 0 ( 0 * 0 * 0 « 0 0 * 0 v o β η £

u

Ü

2i

cd M

•5 ?

c

£ «

H CX O 4 * H t? υ M *d £ S C β, Β H υ s S - I_ I X 5 c ς α α ο Μ « . . cd S- α α, χ cd cd c d » 0 V 0 > 0 * 0 * 0 a: ω s

ex

o à c Q .g

υ

Ö cd s S s a s S & c3d u S S

υ 0

à

β u f- χ Ο (β o Cd & Π « * Β ·&·& 3

υ

Χ cd cd Χ Ut Χ *

ex Χ ex IO cd cd «

Η

Η Η _ 5 b « O £ ^ cd cd rt >o η η t.

υ ο ex H

£ cd 2 5 i o ¿> 3 O e? ΙΟ Λ ο β ex

Η μ ? m§

Tab. 28. Comparative Izhorian-Finnish glossary with special reference to contrasts in processes of lexical internationalization Izhorian equivalent

Meaning

Finnish equivalent

adressi

'address'

osoite

apreli auGusti

'April' 'August'

huhtikuu elokuu

fäbrikka fevrali flótta fwiDamentti

'factory' 'February' 'fleet' 'fundament'

tehdas helmikuu laivasto kivijalka, perusta

jeropläni

'airplane'

lentokone

karahteri konDrahti

'character' 'contract'

luonne sopimus

massina mattëri muskuli

'machine' 'material' 'muscle'

kone aine lihas

nerva

'nerve'

hermo

palanssi palhhoni penZi pibliottèkka pibliä

'balance' 'balcony' 'pension' 'library' 'bible'

tasapaino parveke eläke kirjasto raamattu

reniti

'rent'

vuokra

stiletti sorttu

'skeleton' 'sort'

luuranko laji

-4 Remarks: (1) When two variants of lexical elements are mentioned (e.g., mmacaïc/ m o w c a K , K B a p T H p a / c J j a T e p a , ρεκργτ/Ηεκργτ), this indicates the duality of written standard and dialect form. The variant in brackets has been adopted into Izhorian. (2) The number of loanwords from west European languages after 1700 is far greater than the number of borrowings adopted by Russian before 1700. Such proportions are also reflected in the international inventory of the Izhorian language. (3) In most cases of borrowings whose sources cannot be precisely identified, the alternatives seem to be either the French or the German language as the source. In some cases, the international element may have its origin in English.

4. Bilingual profiles and changing bilingual patterns — an approach to formalize bilingual identity Many approaches have been made toward a formal characterization of bilingualism in quantitative as well as typological terms. Such approaches include analyses of the language-state relation (see 3.2. in the inventory of ecological variables); relevant examples include the concept of sociolinguistic profile formulas (Ferguson 1966), the typology of languages in a multinational state on the basis of status criteria (Stewart 1968), an inventory of language-state variables in ethnopolitical terms (Kloss 1966), etc. In addition, there have been considerations focusing on the specification of functional ranges of languages in a bilingual speech community (see Fishman 1964, 1965 for a discussion of the term "domain" and the use of dominance configurations). Other contributions deal with the polarity of language maintenance and language shift as a phenomenon of group bilingualism (see Fishman 1966; Haarmann 1979b, etc.). There are other aspects to the issue which have, however, been neglected until recently. One problem, for instance, concerns the specificity of the correlation between the status of the first language and that of a second language in a bilingual community. The particular subject of investigation involves the dynamics of group relations in terms of first language adoption and second language attraction. Moreover, whereas analyses of the language behavior of bilingual speakers on the performance level provide insights about the languages' functional ranges as well as their modality of application, the study of the distribution of different languages in a speech community on the level of knowledge or competence may reveal that the language repertoire or communicational volume (cf. the German term "Kommunikationsvolumen", used in Haarmann 1980 a) is an indicator of a speech community's collective communicational mobility. The fact that the latter aspect has been widely neglected in research on multilingualism may be explained partly by the difficulties one meets when trying to measure elements of competence. As most quantitative

120

Language in ethnicity: a view of basic ecological

relations

analyses concentrate on the evaluation of performance-related data, little progress has been made in specifying phenomena on the competence level. It is much easier to base research on performance data as all phenomena of performance are open to quantitative analysis (see Kreppner 1975,23). This is true for the social sciences in general and for measurements of language behavior in particular. Knowledge of languages in a community of bilingual speakers is not open to direct observation and can only be specified indirectly. In Haarmann (1978), I attempted to formulate special methods for the correlation of ethnostatistical data on first and second language status (language maintenance figures versus second language attraction rates). The data, which are figures representing knowledge of languages (excluding data which are performance oriented), can be correlated to form a bilingual profile. In Haarmann (1979a, 210ff., 1980a, 205ff.), I elaborated the bilingual profile as a quantitative tool of sorts in macrosociolinguistics. Some of the main aspects of my proposed bilingual profile, which also serves as a means to identify variant structures (or differential structures) of group bilingualism, shall be discussed here. In multinational states where bilingualism is a common feature in societal intergroup relations, the communicational structures of ethnic groups are shaped according to a network of language components corresponding to first and second language status. The terms "first language" and "second language" in this connection characterize the communicational means in terms of their chronology of acquisition by the speakers in a bilingual community. Thus, the language adopted first in the course of a child's primary socialization is referred to as the first language; the second language is the language adopted after the acquisition of the first. The second language could be a language adopted at preschool age, or that used as a means of instruction at school. The special purpose of a bilingual profile is to illustrate the combination of first and second language knowledge in terms of dominance and nondominance among speakers in a bilingual community. The profile itself is based on a quantification of the range of communicational means in the investigated community. In order to explain the effectiveness of this approach, I refer to multilingual settings among ethnic groups in the Soviet Union. These settings are often quite complex and pose many challenges to the investigator's ability to categorize phenomena of group bilingualism. The language conditions in Soviet society are mainly characterized by a communicational pattern which can be specified as active bilingualism (active, meaning speakers' level of knowledge adequate for active use of

Bilingual profiles and changing bilingual patterns

121

languages), with the components of the "national language" having first language status and Russian having second language status. This variant of group bilingualism represents the only communicational pattern with a wide functional range in all the Soviet Union (see Haarmann 1985 b, 2.1.). This means there is, in fact, no other variant of bilingualism in the Soviet Union with an interregional range of application. As this variant of group bilingualism is found in each of the more than 130 Soviet speech communities (excluding the monolingual Russian community), it has attracted the attention of Soviet language planners who consider it a vehicle for integrating all ethnic groups into one Soviet nationality. In Soviet sociolinguistics and society-oriented planning activities, this variant is referred to as "national (language)-Russian bilingualism" (Russ., nacional'no-russkoe dvujazycie). Some representative concrete settings are Estonian-Russian bilingualism among the Estonians, UkrainianRussian bilingualism among the Ukrainians, Grusinian-Russian bilingualism, Armenian-Russian bilingualism, etc. National-Russian bilingualism is the most widespread communicational pattern among most non-Russians and is accordingly categorized as "mass bilingualism" in Soviet language planning. "The development and interregional spreading of mass bilingualism with Russian in the function of the second language has an extremely great importance in the living conditions of the multinational Soviet state (Deseriev 1976 a, 3; translated by the author)".

Although national-Russian bilingualism is often described and analyzed by Soviet language planners, there exists no formal approach in Soviet sociolinguistics to measure bilingual structures in different ethnic groups. Such measurements, however, could provide Soviet planners with a valuable, if not necessary, basis for the optimization of planning aims. Outside the framework of Soviet ideology, quantification of national-Russian bilingual structures is potentially an eminent field of research which could provide insights into patterns of differential bilingualism under varying ecological conditions. Analyses on the basis of data for nationalRussian bilingualism in the Soviet Union thus can be useful for general sociolinguistic discussions on multilingualism. In particular, the application of bilingual profiles and their evaluation in terms of a typology of bilingualism reveals ethnically specific structures of national-Russian bilingualism for single ethnic groups, and thus allow conclusions about their language choice under special ecological conditions.

122

Language in ethnicity: a view of basic ecological

relations

For the purposes of a wide-ranging evaluation of bilingual profiles, I base my observations and conclusions on a variety of data for ethnic groups in different parts of the Soviet Union (especially in the European and Central Asiatic parts) and of differing status (cf. autochthonous nationality versus national minority). In Haarmann (1980 a, 232 ff.), I also included findings regarding other variants of group bilingualism (i.e., either excluding Russian as a second language or excluding the national language as the first language in bilingual communities). There I referred to bilingual structures Ukrainian-Russian among the Polish minority in Ukrainia, and Russian-Ukrainian among the Greek minority in Ukrainia. The so-called national-Russian variants of bilingualism offer a broad panorama of different non-national contact languages which have been adopted as the first language by a number of speakers who represent portions of their corresponding ethnic groups (e. g., the role of Hungarian for Transcarpathian Gypsies, of Turkish for Greeks in Southern Ukrainia, etc.). In connection with a survey of group bilingualism in the Soviet Union (see Haarmann 1985 b), I have discussed the spread and social functions of non-national-Russian variants of bilingualism. Some of those bilingual communicational patterns will also be illustrated here. The différenciation of profile types as it is presented below has the methodological value of a genuine typology. The effectiveness of a typology is much greater than that of a simple classification. As a classification aims at categorizing objects according to classes, and as classes exclude each other because of their mere existence (an object of class A cannot be an object of class B), such a method of categorization is less flexible and inclusive than a typology. The schematism of a classification is avoided in the framework of a typology because the "type", as a theoretical construct of the features of concrete objects, is based on a variety of criteria, whereas the concept of "class" is usually defined on the basis of one or two single criteria. Concrete objects are associated with a type according to the prevalence of a cluster of features which are theoretically attributed to the type. Therefore, a typology is more flexible and allows for the specification of so-called floating feature patterns which cannot be described accurately in terms of a classification. Floating feature patterns or "mixed" patterns which often appear in concrete settings can be adequately specified only in typological terms. In Haarmann (1980 a, 119ff.), I pointed out main differences between classifications and typologies in a methodological discussion. The profiles as they are presented here include a number of variables which, together, produce a cluster of features in each profile. The eth-

Bilingual profiles and changing bilingual patterns

123

nostatistical data on which each profile is based vary considerably according to the dynamics of different variables, and such variations are reflected in the shape of the curves in the graphical representations. A comprehensive bilingual profile includes variations of the following factors: (a) relation profile) profile) (b) relation (c) relation (d) relation

between first language rates (see curve/component 1 in the and second language rates (see curve/component 2 in the between urban and rural populations between male and female populations between average (general) and partial (special) rates

The distribution of first and second language knowledge in different groups of speakers (speech communities or portions of them) is reflected in the configuration of the two respective curves. As the transposition of varying data results in different proportions, and as each profile consists of a particular cluster of features, there will be no single bilingual profile which is identical to another. Not unexpectedly, this means that every speech community has its own specific bilingual profile. The purpose of a typology of profiles is to categorize similarities and divergences in concrete configurations and to examine hypotheses about language behavior relative to the specified types. The term curve (the graphical component in a profile) is used as an abstraction to represent the pattern of distribution of first and second language knowledge (combining rates for language maintenance and for second language attraction) in a speech community. On the basis of such profiles, communicational patterns as of national-Russian bilingualism or other variants of bilingualism can be graphically illustrated. The application of this graphical means provides much information and insight concerning the vitality of single languages as well as the stability of bilingual structures in different groups (e. g., urban population, female population, etc.). Such information usually remains hidden in the flood of ethnostatistical data collected and evaluated for various ethnic groups. Bilingual profiles can be set up for ethnic groups as a collective whole, or for local groups within a certain speech community. This variation could also be included as an additional feature (e) in a comprehensive bilingual profile. In Haarmann (1979a, 210ff.), I discussed bilingual profiles which illustrate bilingual communicational patterns for entire speech communities (i. e., collectives of speakers of a national language). Additionally, a bilingual profile can be utilized to characterize communica-

124

Language in ethnicity: a view of basic ecological

relations

tional patterns for local groups within one and the same speech community. And of course the same can be done for local groups in different regions of a particular area (regarding this for the Soviet Union, see Haarmann 1979c, 155ff., 1980c, 211ff., etc.). In the present contribution, reference to an entire speech community is indicated by the words "general overview"; when referring to local groups, the region is specified {e.g., Lithuanian S.S.R.). Special problems in connection with bilingual profiles of local groups are discussed below. Specifying certain configurations of curves in bilingual profiles as types has the value of labeling basic structures of bilingualism according to the shape of curves resulting from correlations of ethnostatistical rates. More important than the labeling of profile types, however, is the interpretation of the hypothetical features of language behavior which are associated with the type. Such a view of the value of typologies in a sociolinguistic framework is similar to the approaches in the area of general typologies of language, where in this case too, the contents of a type is given priority over the name or label. In connection with quantitative approaches to language typology, it has even been stated that the name of a type is completely unimportant: "Which name is given to this "type" is completely unimportant. . . . It is the linguistic information associated with the type which is of importance for the application of the typology (Altmann—Lehfeldt 1973, 48; translation by the author)".

Although many bilingual profiles have been elaborated on since I first introduced this means as a technique in applied language ecology, I still maintain my original fourfold categorization of profiles. This is because the four basic types which are distinguished are inclusive and thus can serve the needs of general as well as specific analysis. In Haarmann (1980a, 21 Iff.), however, I extended the interpretation of characteristic features of single types. Here, the profile types are specified with respect to the synchrony of bilingual settings. Problems of transitional development between different types as stages of bilingualism are discussed below. Let us turn now to the four basic profile types.

Bilingual profiles and changing bilingual patterns

125

1. The open type of bilingual profile with curve (1) representing first language knowledge as the dominant component in the configuration This type of bilingual profile is characterized by an "open" configuration where there is no crossing of the components in any section. Curve (1) with its function of representing percentage rates of first language knowledge (equivalent to figures for language maintenance) constitutes the upper part of the profile. The bilingual structures of the following ethnic groups are characterized by this type: Lithuanians (general overview), Ukrainians (general overview), Poles in the Ukrainian S.S.R., and Gypsies in the Lithuanian S.S.R. These bilingual profiles represent the communicational pattern of national-Russian bilingualism for Lithuanians, Ukrainians, and Gypsies as well as of non-national-Russian bilingualism for the Poles in Ukrainia. The configuration of the components (1) and (2) differs in each profile according to the particular proportions in the ethnostatistical data for each group. Thus, whereas the profiles may show a number of divergences in different sections, the same type of profile is nevertheless represented in all the configurations in as much as the same (congruent) pattern of basic features of language behavior can be found in each of these groups.

Tab. 29. Correlation of rates for Ukrainian as a first language and Russian as a second language in the Polish community (Ukrainian S.S.R.) % 100

85.6%

85.5%

85.7%

(1 ) 68.6%

69.6%

fO October jüichi-gatsu 11 b'n· j November jüni-gatsu 12 Cui IZA1-? December

t,

10 jü

b' φ i

11 jü-ichi

Uw> Ί

12 jü-ni

lí Φ 9 1-

13 jü-san

the distribution is characterized by a mix of Japanese and Sino-Japanese. Whereas in the numbers up to twenty, the expressions jü-yon (hybrid term) or jü-shi (genuine Sino-Japanese term), and jü-nana (hybridterm) or jü-shichi (genuine Sino-Japanese term) may be used alternatively, the Japanese component is stable in the higher numbers (e. g., yon-jü, 40; nana-jü, 70; yon-hyaku, 400, etc.). Counting beyond thirteen, therefore, is thus characterized by an application of the mixed system. There are many other aspects to this multifaceted system of Sino-Japanese and Japanese numerals and their combination in different inventories. One special aspect is the synonymity of shichi (Sino-Japanese) and nana (Japanese) in special applications of the mixed system of numerals. Reference has already been made to the ambivalence of jü-yon/ju-shi and jü-nana/jü-shichi for counting fourteen and seventeen respectively. When telling time, nana-ji and shichi-ji may be used for "seven o'clock". To denote a lapse of seven hours, either nana-jikan or shichi-jikan can be used. For the month of July, only the term shichi-gatsu (lit. "7th

The adoption of foreign cultural patterns

Tab. 60. Mixed system of Sino-Japanese and Japanese numerals

person 1 hitori

l>t η

2

futan

3

san-nin

4

yo-nin

«ti; A.

5

go-nin

it ζ A

6

roku-nin

ò < iz A

7

nana-nin ù ù l z A shichi-nin l t>iz A,

8

hachi-nin

9

kyü-nin ku-nin

10 jü-nin

169

Ό ZAlzA

H^lzA èwjtzA < iz A Lì w i izA

Japanese

long thing e. g. pencil, bottle 1

ip-pon

i^olfTu

2

ni-hon

izllA

3

san-bon

4

yon-hon

5

thin thing e. g. paper, plate

book

ichi-mai

i ^ t ιλ

is-satsu

^ o è o

ni-mai

lift,*

ni-satsu

t; è o

san-mai

è A, i v^

san-satsu

ϊ/ι,ί-5

&AIS.A

yon-mai

Je

yon-satsu

kAZ-o

go-hon

.T'tiA,

go-mai

go-satsu

6

rop-pon

3 o ISA

roku-mai h < £

roku-satsu 5
$

has-satsu

ti

9

kyü-hon

§ φ ò tí λ.

kyü-mai

kyü-satsu

T W> ? £ O

10

jup-pon

is a»-o ISA jü-mai

Japanese

á vi

ΰ w> ì £

jus-satsu

i è

è

170

Language in ethnicity: a view of basic ecological

ship

1

is-seki

2

ni-seki

3

san-seki

4

^o-tí-S

relations

house

money

ik-ken

^ottA-

ichi-en

ni-ken

t ~ /o

ni-en

(c i A-

s/u-tüt

san-gen

è A. If A.

san-en

£ A. ¿ A.

yon-seki

J;A,-tì-§:

yon-ken

i A. It A,

yo-en

x. A.

5

go-seki

i-tt §

go-ken

ittA.

go-en

¿"i A

6

roku-seki δ < -t

rok-ken

roku-en h < χ. A,

7

nana-seki

nana-ken £ £¡tA.

nana-en & & i Α.

8

has-seki

hak-ken

l± o It A,

hachi-en li

9

kyü-seki

kyü-ken

£ κ» i It λ

kyü-en

10

jus-seki

liotj

is φ o -tí- i juk-ken

^¿A,

L; κ» o It A, jü-en

£ Α.

φ ?i A

Japanese

1

ik-kai (ichi-do

v>¿>¿*)

2

ni-kai

3

floor of a

order

frequency

building

ichi-ban

νίί,ϋ'Α.

ik-kai

i^/i'M

CAn*

ni-ban

i: If A.

ni-kai

IZÂHÎ

san-kai

è ΑΛη^

san-ban

è A l£A-

san-gai

4

yon-kai

i; ΑΛΗ^

yon-ban

J; Λ,ίίΑ,

yon-kai

¿A^AH*

5

go-kai

zlfrv*

go-ban

•!"(£ A

go-kai

8

hak-kai

li-^^i»

hachi-ban li h If A-

hak-kai

9

kyü-kai

^ «J> ') Λη>

kyü-ban

£ w ì If A.

kyü-kai

# ·Φ ? AHI

juk-kai

TMO^H»

jü-ban

is Φ 9 If A,

juk-kai

ΰ Φ O AH >

10

roku-ban

< If A.

The adoption

of foreign cultural patterns

171

Expressions for time duration hour

minute

day

month

ichi-nichi

ik-kagetsu

one day

one month

ni-fun U^A.

futsu-ka

ni-kagetsu UA'lfo

san-jikan èAb

san-pun è Λ,ΛΑ

mik-ka

san-kagetsu £ Α,Λ· If o

4

yo-jikan A

yon-pun Α. Λ A

yok-ka J;

yon-kagetsu •fcA-ri'lfo

5

go-jikan r'bfr/u

go-fun ¿"ΛΑ.

itsu-ka

go-kagetsu Oit'-?

6

roku-jikan Ι'Λ'Α.

rop-pun h -> ΛΑ

mui-ka t

nana-fun 65ΛΑ.

nano-ka

ichi-jikan i>*> b'^A one hour

ip-pun ΙΛ-3 -ΚΛ, one minute

2

ni-jikan e Ι'Λ'Λ.

3

1

nana-kagetsu

nana-jikan 7

shichi-jikan I U-frh

8

hachi-jikan b^A.

hap-pun ΙίοΛΑ. hachi-fun

9

ku-jikan < U:Λ>Α

jü-jikan b φ ì b ii· Κ

10

rok-kagetsu ·6οτί>Ίίο han-toshi !±A¿ I

Japanese

shichi-kagetsu l *>A>!fo yö-ka •t i fr

hak-kagetsu liTÄ'lfo

kyü-fun ^ H> Ί ΛΑ

kokono-ka

kyü-kagetsu

jup-pun b H> η ΛΑ.

tö-ka

juk-kagetsu

172

Language in ethnicity: a view of basic ecological relations Telling time fun J* A, minute

ji b o'clock 1

2 3

ichi-ji b one o'clock

1

ni-ji t: b

2

ni-fun li^A,

3

san-pun è Λ^ν^Α-

san-ji

ip-pun one minute

4

IQ-ji ¿b

4

5

go-ji rb

5

6

roku-ji b

6

7

shichi-ji UÜ

7

8

hachi-ji b

8

hap-pun, hachi-fun ti-^-áí/u,

9

ku-ji ? ΙΛΑ b

11

jüip-pun bφ?

12

jüni-ji b n> 9 íc b

15

jügo-fun b ι® j

Japanese

20

nijup-pun b φ -3.&A.

25

nijügo-fun ι; b φ i

30

sanjup-pun

The adoption of foreign cultural patterns T h e days o f the m o n t h jushichi-nichi

tsuitachi 1

17

bΦ 9 L

i:

the first day of a month 2 3 4 5

6 7

8 9 10 11 12

.

futsu-ka

jühachi-nichi Ιίιφ

ì\i*>\-*>

mik-ka

iq

jüku-nichi

yok-ka



hatsu-ka

itsu-ka Poi1

91

nijüichi-nichi 'lit® 9 nijüni-nichi ι; b φ ? t; iz t>

mui-ka tf·'^ ..

nano-ka

nijüsan-nichi I : b Φ ι È Λ, Ι : niiüvok-ka Ι; b Φ 9 r> i?

yö-ka kokono-ka zzmfr tö-ka ttsfr

Ί

,

nijügo-nichi [LB Φ Ί ¿IZt>

,

nijüroku-nichi \ζ B Φ 9 3 < I :

2B

jüichi-nichi

nijüshichi-nichi

b Φ ι

I : b Φ ι b ÍB I : h

..

jüni-nichi

nijühachi-nichi

U Φ 9 I-t-

ι; b Φ ?

jüsan-nichi

nijüku-nichi

i; n> ? è Λ, u £

ι; b Φ î < t:

14

iüvok-ka C Φ 1 J: ο Λ1

sanjü-nichi

15

jügo-nichi b-Φ î

16

jüroku-nichi C φ ? δ < ι: %

13

_ sanjüichi-nichi -51 έ λ ι ; « )

Japanese

174

Language in ethnicity: a view of basic ecological

relations

Cardinal numerals over thirteen 14 15 16 17 18 19 20 21

iû-von jü-go jü-roku jü-nana jü-hachi jü-kyü ni-jü ni-jü-ichi

c bΦ 7

30 40 50 60 70 80 90

san-jü von-jû go-jü roku-jü nana-jü hachi-jü kyü-jü

è ^ bφ ? Χ Λ. b m ì ¿ bφ7 b 7 r ¿% bΦ 7 t*> bm> ? ΙΦΗΙ»?

100 200 300 400 500

b n> 7 X L bφò Ζ b i9 7 -5 < b «6 i % % b KD 7 l±*> b w> j S φ -j

hyaku ί>* < ni-hyaku san-byaku S / u t » < yan-hyaku Χλ,Ό·* < go-hyaku
t sen ni-sen san-zen von-sen go-sen roku-sen nana-sen has-sen kyü-sen

-Ö-A, c-tt/i Ζλ,-tfA,

r-a-A, ZZ-itA ê

Φ

7

ΐλ-

10,000 ichi-man b Φ 7 i Α. 100,000 jü-man 1,000,000 hyaku-man < f A. 10,000,000 sen-man 100,000,000 ichi-oku

•tì•^í ^ libisi

Japanese

month") is used, but it is possible to use either nana-kagetsu or shichikagetsu to indicate a period of seven months. No special study is yet available which would shed light on the synonymity of the terms in the discussed examples by applying variable rules to colloquial Japanese. It has been pointed out that influences in the system of numerals are an indicator of the character of the influenced language as a massive contact language. A substantial influence in the system of numerals constitutes ethnolinguistic evidence of an acculturation process. As the cases of Romany and Japanese show, the degree of acculturation may be different, but it is evident that in both languages there is no pressure being exerted by the dominating contact language (most borrowed numerals in

The adoption of foreign cultural patterns

175

Romany are of modern Greek origin, in Japanese, of Chinese origin) that could give rise to language shift. Thus the existence of Modern Greek terms for numbers in Romany and the predominance of Chinese numerals in Japanese do not indicate a shift of the speech community to the influencing languages (Romany to Greek, Japanese to Chinese). Together with other factors (ethnolinguistic and other ecological variables), however, the influence exerted in a system of numerals may indeed be an indicator not only of a process of acculturation but also of language shift. Rjagoev (1977, 221), for example, explains the frequent use of Russian numerals by Karelians in the district of Leningrad as a sign of language shift in the Karelian speech community. Such an interpretation cannot be disputed, but it is only valid because there are many other indicators of language shift (Russ., mena jazyka) in Karelian and the community of its speakers. Thus, the use of Russian numerals appears in a framework of several assimilating factors and has therefore to be treated as one of several criteria of the general trend toward language shift. Clearly, acculturation does not automatically imply a tendency toward language shift. As there are various degrees of acculturation and different evolutionary stages in processes of acculturation, only extreme tendencies (or a strong degree) of acculturation are linked to a total assimilation or to the dissolving of ethnic boundaries in terms of ethnicity (see chapter 2).

(2) Borrowings in the deictic system as an acculturation indicator Borrowed constituents of speech acts, such as conjunctions, interjections, particles, pronouns, and elements expressing modal functions as well as other constituents play a keyrole in converging verbal strategies in speech communities with strong interethnic contacts. Thus the transfer of such constituents is directly linked to the process of acculturation, and the integration of foreign elements into the deictic structure of a language is an indicator of such an acculturation. The convergence of verbal strategies in the formation of speech acts is a special concern of ethnolinguistics, one which should also be given attention by researchers in the field of pragmatics and semiotics. As contact linguistics deals with intergroup communication, the constituents listed above should not be analyzed as separate units (according to the traditional view of grammarians), but rather as a whole. The investigation of deictic elements in massive contact languages has been neglected and there is, it seems, as yet

176

Language in ethnicity: a view of basic ecological relations

little interest in this subject, except for the study of specific, single cases of influence in the deictic system of a particular language (see, for example, Weinreich 1980, 527 ff. with respect to the development of aspectual relations in the verbal system of Yiddish under Slavic influence). Whereas in non-massive contact languages, borrowed conjunctions, pronouns, or other speech act constituents constitute a sporadic phenomenon and a rare exception (e.g., Spanish ojalá "my goodness!" (int.) from Arabic), borrowings in the deictic system of massive contact languages are a common feature, although the extent of the acculturation of a speech community as reflected in the system of speech act constituents depends on the concrete conditions of contact (i. e., structural specifics of the contacting languages, etc.). Thus, the degree of integration of foreign elements and categories in the system of deictic categories may vary considerably from one massive contact language to another. I would like to discuss in detail this complex problem, referring especially to deictic categories in Romany. An investigation of the deictic system in Romany with special regard to transfers as a result of language contacts has not thus far been carried out. The inventory of borrowed speech act constituents in Romany represents a broad layer of elements and is highly dispersed throughout the deictic system. There are six major categories which have been affected by contact languages. A variety of elements adopted by Romany at different times and from different languages is illustrated in Table 61. In the section on adverbs, only a few representative are listed because the total number of modal adverbs in Romany that have been borrowed from different sources far exceeds one hundred elements. There is a strong influence in the pronominal and conjunctional systems of Romany (see sections 1 and 4, Table 61). Such a concentration of foreign elements does not represent only a certain number of "lexical" borrowings (as it may appear in the table), but also reveals a significant influence of the syntax as well as of verbal strategies in the formation of speech acts in Romany. The intensity of influence exerted by contacting languages cannot be explained but by the fact that the Gypsies are bilingual. In the course of adopting a coterritorial contact language (in the status of a second language), and frequently using it as a means of intergroup interaction, the second language steadily increases its impact on the verbal strategies of the first language (Romany), which then serves as the special mode of Gypsy intragroup communication. This restriction often results in Romany being considered a secret language, deliberately used as such by the Gypsies themselves. As the coterritorial language

The adoption of foreign cultural patterns

177

Tab. 61. Borrowed deictic categories in Romany 1. Conjunctions German borrowing: und 'and' Polish borrowings: ale 'but', ani... ani 'neither . . . nor', bo 'as', 'because', i 'and', choc 'although', puko 'until' Rumanian borrowings: fincha 'as', 'because', châ 'that', 'because', nici 'neither' Serbo-Croatian borrowing: neg 'than' (after comparative) Slovak borrowings: alebo 'or', esli 'if, I'ebo 'as', no 'but, instead', ci 'in case', 'if Hungarian borrowing: waj 'or' Finnish borrowings: fani 'but', ja 'and', jos 'if, jotta 'in order to', 'so that', ko 'in case', mutta 'but', tai 'or', va 'but only', vahka 'although' Borrowings from other languages: kadzei 'although' (Lithuanian), nina 'also' (Modern Greek) 2. Particles German borrowings: gar 'not', nit 'not' Hungarian borrowings: unga 'isn't it?', ware- 'at all' Finnish borrowings: eija 'yes', -han (emphatic particle), kai 'may be', -ki 'too', kyllä 'yes', ni 'so', -pa 'just' 3. Interjections German borrowings: ach 'oh', ne 'well' Rumanian borrowings: faide 'let's go', märe 'hello', 'hey' Hungarian borrowing: ëljen 'cheers' Finnish borrowing: just 'like that', 'so' 4. Pronouns Polish borrowings: kazdo 'everyone', samo 'same' Rumanian borrowings: fitesáoro 'everyone', 'each', nishte 'something' Serbo-Croatian borrowings: nist 'nothing', sako 'everyone' Slovak borrowings: isto 'same', niko 'nobody', onä, ône 'both', samo 'such', co 'what kind?' Finnish borrowings: kömbi 'which one?', millo 'when?' 5. Modality of sentence In the dialect of the Gypsies of Finland {fintikd rómma), the modality of negation is expressed by a partial adoption of the Finnish construction with an inflected negation verb, e.g.: etteiväs (Finn, että 'that' + eivät 'they not . . . ' ) ei... na (Finn, ei '3rd person of negation verb + Romany na 'not') eikö ... na (interrogative form of negation by means of the negation verb and the Finnish enclitic interrogative particle -ko)

178

Language in ethnicity: a view of basic ecological relations

Tab. 61. Borrowed deictic categories in Romany 6. Adverbs German borrowings: blank 'only', gern 'with pleasure', son 'already', etc. Polish borrowings: darmó 'in vain', modze 'perhaps', tyl(k)o 'only', etc. Rumanian borrowings: ewe 'hardly', cham 'about', preá 'very', etc. Serbo-Croatian borrowings: dosta 'enough', niposo 'by no means', etc. Slovak borrowings: esce 'yet', nacin 'necessary', pocilku 'until now', etc. Czech borrowings: dycky 'always', ledva 'hardly', prétsa 'for sure', etc. Hungarian borrowings: bizo 'certainly', majd 'soon', cak 'only', etc. Finnish borrowings: ehkä 'perhaps', hyvin 'very', kohta 'soon', etc.

Remarks: (1) The selection of borrowings above is based on the material in the dictionaries of Wolf (1960) and Valtonen (1972). (2) The panorama of contact languages which have influenced the deictic system of Romany can be extended to also include Swedish, Russian, Ukrainian, and others. The given outline of borrowings is an overall sketch of Romany in general and can be further split up into specified sets of deictic elements in local dialects. Specified sets of elements would reveal significant variants in the deictic inventory between single dialects.

tends to become the means of communication most frequently used in daily (intergroup) interaction, it is natural that verbal strategies of the second language familiar to the Gypsy speaker would be transferred into the mother tongue and gradually shape the syntax (i. e., the role of conjunctions), the reference system (¿. e., pronouns), and the modal implications of speech acts (i. e., the role of interjections, modal adverbs, formation of negative sentences). I should point out that the table illustrates the impact of contact languages on Romany regardless of local dialects and their specific inventories. The number of borrowed elements in the deictic system of each Romany dialect naturally is smaller than in the total inventory of all dialects. It is plausible that the coterritorial language most frequently used is the one that most strongly influences the deictic system of Romany. In the language of the Gypsies in Rumania, borrowed deictic elements from Rumanian are predominant, whereas the Gypsy language in Germany is most strongly influenced by the German language. Romany as spoken by the Gypsies in Finland, not unexpectedly, contains a great deal of Fin-

The adoption of foreign cultural patterns

179

nish elements in its deictic system. At the same time, however, other borrowed deictic elements occur in each dialect, thus reflecting influences from non-coterritorial contact languages. In the language of the German Gypsies, for example, Rumanian conjunctions and interjections are also used. These elements do not reflect the influence of a coterritorial language, but are remnants of historical contacts dating back to times when the Gypsies reached the Balkans on their way from Asia to Europe in the late Middle Ages. In the language of the Gypsies in Finland, one can also find deictic elements from Rumanian, Slovak, German, and Swedish, thus also reflecting previous influences from coterritorial languages. In Finnish Romany, both Finnish and Swedish play important roles as coterritorial languages, as the Gypsies live not only among Finnish speakers, but also among Finland Swedes with Swedish as their mother tongue (in southern and western Finland). Romany is a very clear example of a massive contact language which displays a complex of influences from various contacting languages, as reflected in a multifaceted network of borrowed deictic categories. There are few such languages with such a variety of transfers, although Yiddish at least is another. There are many massive contact languages in which the impact of the influencing languages in the deictic system is much more limited than in the case of Romany. Languages such as Izhorian, Karelian, and a number of others, which are predominantly influenced by Russian, nevertheless clearly demonstrate their character as massive contact languages by having a variety of borrowed deictic elements (see Haarmann 1983b, 42ff. for Izhorian; Rjagoev 1977, 168ff. for Karelian, etc.). It would be a great achievement if this special field of study were met with greater interest by researchers of language contact phenomena and some progress could be made in comparing massive contact languages and their deictic systems.

(3) Borrowings in the lexical field of kinship terms When discussing lexical transfers in the area of kinship terms, one is often confronted by various prejudices similar to those in connection with the system of numerals or parts of the body (discussed in the following section 4). The prejudices are based on the assumption that kinship terms represent a section of the vocabulary which is restricted to cognate words and in which borrowings are seldom found. Thus, a famous scholar of the Russian language declares that borrowed kinship terms are

180

Language in ethnicity: a view of basic ecological relations

"unlikely" (Kiparsky 1975, 54). Such a view is widespread among scholars of Indo-European studies. Prejudicial statements like this one are also frequently found in the field of Altaic studies. One such may be reproduced here as an example of this sort of careless evaluation. In connection with research on lexical relations between the Turkic and Mongolian languages, Kenesbaew and Sarybaew (1975, 84) make the following claim regarding kinship terms and designations for parts of the body: "Both lexical strata belong to the ancient impenetrable or little penetrated sphere of a language". It is interesting to learn, however, that the languages Kiparsky (Russian) and Kenesbaew and Sarybaew (Kazakh, Buryat, Khalkha-Mongolian, and others) refer to possess lexical borrowings both for kinship terms as well as for parts of the body (see sect. 4 for Russian examples). It is not after thorough investigation of loanwords, but rather a prejudicial evaluation of these lexical fields that has shaped this misleading concept of lexical conservatism. It is inappropriate to apply categories of rational thinking (i.e., any concept which is well known in a speech community will be designated by a cognate word) to process of lexicalization. Lexical innovation, under conditions of language contact, is based on a variety of principles which do not reflect a schematism as represented in the above statements. Research in the area of massive contact languages and their ecological settings clearly reveals that kinship terms are a major indicator of acculturation (which must not be misunderstood to mean assimilation, language shift, or language death). Most likely, kinship terms are borrowed in the course of interethnic contacts between the members of one speech community and the members of another. There are hundreds of ecological settings in Europe and Asia where family relations and the structures of their lexical representations in the corresponding languages are firmly involved in intergroup contacts and interethnic relations. Kinship relations are affected by the conditions of such contacts, and borrowings in this sphere indicate different evolutionary stages of acculturation. Kinship terms as borrowings appear sporadically also in the vocabularies of non-massive contact languages (e.g., Finnish äiti 'mother" from Germanic; German Onkel "uncle" and Tante "aunt" from French, etc.). Such borrowings are, however, not sporadic but typical of massive contact languages. One example of an acculturation process as reflected in the lexical field of kinship terms shall be illustrated here. Izhorian is a language which reflects a high degree of acculturation in its structure. Details concerning the Russian influence in Izhorian have been provided in chapter 3. Kinship terms have also been affected by the

The adoption of foreign cultural patterns

181

Tab. 62. Russian borrowings in the spheres of Izhorian kinship terms and related concepts 1. Kinship terms of Russian origin d'ottä (cf. Russ. Tëra) 'aunt' d'äd'ä (cf. Russ. HHHH) 'father' köma (cf. Russ. KyMa) 'cousin' marna (cf. Russ. MaMa) 'mother' (children's speech) mättuska (cf. Russ. MaTymica) 'godmother' päBa (cf. Russ. 6a6a) (ironical term for one's wife) plemmänikka (cf. Russ. ruieMHHHHK) 'nephew' plemmänitsa (cf. Russ. njieMaHHHija) 'niece' rodnoi (cf. Russ. ponHoñ) (having the same father or mother, i.e., real brother or sister) svätta (cf. Russ. ceai-) (kinship term to characterize the relation between fathers-in-law) svojakka (cf. Russ. CBOHK) 'brother-in-law' t'ät'kä (cf. Russ. THTbKa) 'father-in-law' täDä (cf. Russ. ΛΗΑΗ) 'uncle' vunukka (cf. Russ. BHyic) 'grandchild' 2. Terms for marriage and family relations of Russian origin mesta (cf. Russ. MecTo) 'dowry' prihoi (cf. Russ. ΠΡΗΓΟΧΗΗ) 'bride' prittàna (cf. Russ. npn^aHoe) 'dowry' senihka (cf. Russ. xceHHx) 'bridegroom' svättaissa (cf. Russ. CBaTaTbca) 'to promise marriage to a girl' trûska (cf. Russ. npyacica) 'someone leading the bride to the altar during the wedding ceremony' v'entsäissä (cf. Russ. BeHiaTbca) 'to marry in a church ceremony' 3. Terms for bachelor of Russian origin holostoi (cf. Russ. XOJIOCTOH) 'bachelor' koppëli (cf. Russ. Koôejib) 'bachelor' 4. Terms related to the individual of Russian origin familia (cf. Russ. (})aMHJiHfl) 'surname' including 130 personal names of Russian origin used among the Izhorians (see Haarmann 1983 c). Remark: The Russian term RKRSI has been borrowed into Izhorian twice (/. e., the Izhorian equivalents d'äd'ä and täDä). The older borrowing, taDä, has spread more widely in Izhorian than the more recent d'äd'ä.

182

Language in ethnicity: a view of basic ecological relations

impact of the Russian language. About two dozen expressions have been borrowed from Russian (see Table 62), and if one also includes lexical items which are conceptually associated, the number increases considerably. Kinship terms are not isolated in the vocabulary of any language, including Izhorian. Designations of kinship relations play a keyrole in the framework of social relations or social contacts, and these relations, in ecological terms, are the main channels through which acculturation is transferred. Izhorian words like täDä, vunukka, senihka, holostoi, etc. are indicators of the acculturation of the Izhorians because these words are used to designate relations among the Izhorians themselves. Thus the acculturation is indicated by the fact that the borrowed kinship terms play a role in intragroup relations within the Izhorian community. This may seem obvious, but there are cases where borrowed terms are partly restricted to designate kinship relations in the ethnic groups from which the borrowed terms came and not in the "borrowing" community. Many borrowed kinship terms in Romany refer only to family relations in nonGypsy communities. The Izhorian terms, however, are not restricted to a designation of kinship relations among Russians, but have been integrated into the system of denominations reflecting social structures in the Izhorian community. On the other hand, the borrowed terms are naturally also used for expressing family relations in contacting groups such as the Russians, Vots, Finns, and Estonians. There are close conceptual associations between kinship terms and designations related to marriage (including related concepts like 'surname', 'dowry', etc.). Izhorian-Russian interethnic contacts have been close for centuries, as have family ties between Izhorian and Russians. Intermarriage has brought about various significant changes in this area of Izhorian. Indeed, the most obvious links between acculturation as a social phenomenon and language behavior are demonstrated in the attitudes toward Izhorian, on the one hand, and toward Russian, on the other. There has been a tendency traditionally among the Izhorians (especially as a reflection of settings in ethnically mixed families) to willingly adopt Russian as their first language and drop Izhorian as the language of the family. In addition, it is not uncommon to adopt Russian names as personal and family names. There are approximately one hundred thirty personal names that have been borrowed from Russian and have become popular among the Izhorians. Russian surnames are also found in the Izhorian community. Whereas personal names usually indicate preferences or inclinations on the part of members of a community under cultural influence and reflect a natural evolution of acculturation, the

The adoption of foreign cultural patterns

183

adoption of surnames is quite different. Adopting surnames carries juridical consequences for the sociopolitical status of an ethnic group. In Soviet society, the patrilinear principle (as in many other societies) is dominant. This means a family's surname will be Russian when the husband is a Russian and the wife Izhorian. Surnames thus not only indicate the degree of interethnic marriage and the resulting acculturation, but also the status of the group's collective identity in juridical terms. An extreme case of total acculturation with respect to surnames is illustrated by the ecological setting of the Karelian community in the district of Leningrad, where only Russian surnames are in use. This indicates that there is no Karelian family in that area where all members are of Karelian ethnic origin. In all families, instead, there is at least one living male Russian member or a Russian ancestor. I have discussed the role of personal names and surnames adopted by Izhorians and Karelians elsewhere (see Haarmann 1983c). Borrowed kinship terms occur as an indicator of acculturation in many other settings (e.g., Latin-Welsh, Latin-Albanian, Latin-Basque, Russian-Moldavian, Russian-Ukrainian, Polish-Karaim, Rumanian-Hungarian, Bulgarian-Turkish, Turkmen-Beludzhi, Japanese-Korean, etc.). The borrowing of kinship terms in all these cases is a reflection of close interethnic contacts. The fact that a dominating ethnic group is transferring its cultural patterns onto an indominant ethnic group (as a minority group in the contact area) through direct and mostly long-term interaction, can be regarded as a precondition for this phenomenon of acculturation. There will be no transfer of kinship terms unless two speech communities are in close interethnic contact. In this respect, the adoption of social relations as a cultural pattern is different in nature from other phenomena of acculturation, such as borrowings in the system of numerals. The case of Chinese-Japanese language contact illustrates that the transfer of numerals may take place without the precondition of interethnic relations. It was the mere prestige of the Chinese language (predominantly in its written variety) in Japan which was the source of this kind of acculturation in the Japanese speech community.

(4) Borrowings in the lexical field of parts of the body In this connection, I would like to discuss the bias widespread among native speakers of any language as well as among linguists. According to the prevailing view of language and language behavior among members

184

Language in ethnicity: a view of basic ecological relations

of a given speech community, the parts of the body are designated by cognate (i. e., indigenous or autochthonous) forms, and this part of a language's vocabulary is not affected by the influence of a contact language. There are some logical explanations which seem to support this view. As all parts of the body can be seen or touched and therefore experienced directly in the speaker's everyday life, the belief arises that all designations are original. This view is based on the assumption that a concept which is known - particularly as intimately as body parts - to the speaker will not be designated by a borrowed word. Although "logical", this explanation of lexicalization completely fails. The assumption is due mainly to a more or less schematic view of language contacts and lexical transfers. To consider the parts of the body to be a part of a presumed "basic" vocabulary comprised of only indigenous words is a language stereotype (on the native speaker's side) and a linguistic misconception (on the linguist's or sociolinguist's side). Here I would like to dismantle this false presumption and clarify the process of lexical transfers in language contact. There is enough evidence of lexical borrowings in the area of parts of the body such that this ethnolinguistic issue cannot be treated as a minor one, but should be considered a question of major importance for the study of interethnic relations in a general framework of language ecology. A number of European and non-European languages have been analyzed with special regard to loanwords in the area of parts of the body. These languages include Welsh and Breton (designations of Latin origin), Basque (designations from Latin), Albanian (also from Latin), Finnish, Karelian, Veps (designations of Germanic, Baltic, and Slavic origin), Izhorian (designations from Russian), Romany (designations from various non-Indic languages), Russian (designations from Germanic, Tatar, Greek, and others), etc. Many examples of borrowed designations are also discussed for a number of Romance, Turkic, and other languages. The findings from these and other examples (including some new accounts in this chapter on Iban, Urdu, and other languages) allow for some general statements on this subject, which is not one of "rare etymologies" but, instead, a major ethnolinguistic phenomenon resulting from language contacts under varying sociocultural conditions. There is no need to repeat the many misconceived statements of those who consider borrowings in the lexical area of parts of the body to be rare or unusual exceptions. This view is especially pronounced in connection with the notion of a so-called "basic vocabulary" in the tradition of historical linguistics. The comparative study of Indo-European lan-

The adoption of foreign cultural patterns

185

guages has given rise to the assumption that a precultural lexical foundation, in which only cognate words are found, exists in any language. This tradition, largely determined during the course of the nineteenth century, continues to prevail today (see e.g., Meillet 1965, 94ff., 1966, 33 ff.). Even though as early as the beginning of this century (Zauner 1903) the phenomenon of borrowed designations for parts of the body had been illustrated for the Romance languages (e. g., words of Germanic, Greek, and other origin in French, Spanish, etc., such as French hanche "hip", estomac "stomach", etc.), little consequence for further onomasiological studies resulted. In Haarmann (1979d, 66 ff.), I attempted to point out the relevance of Zauner's original findings for the study of language contacts in Western Europe. Nevertheless, statements such as the following by Kiparsky (1975, 54) concerning the historical development of the Russian vocabulary are not uncommon: "It is unlikely in the first place that names for parts of the body and kinship terms are borrowed...". In his word lists, however, he cites a considerable number of loanwords in this section of the Russian lexicon: glaz "eye" (Germanic); kulak "fist" (Tatar); baska "head" (lit. "turnip", ironically used for "head") (Tatar); morda "mouth" (of animals) (Iranian); sycug "stomach" (of animals) (Tatar); spina "back" (Polish, indirectly from Latin); stomach "stomach" (Greek); etc. Assertions of this sort only illustrate a lack of understanding of lexical transfers under varying conditions of language contacts from one language to another and of attitudes of bilingual speakers toward a dominant contact language. In the following, an attempt is made to analyze some universal features of the ethnolinguistic issue under discussion. The findings include remarks on the extent of this type of lexical borrowing, on the nature of the lexical innovational processes involved, as well as on the sociocultural conditions of the corresponding language contacts. It may be considered necessary to further elaborate some of the tentative "universal" statements (statements on presumed "universal" phenomena). Further studies on this subject will perhaps allow for the formulation of some "laws" on the basis of an enlarged inventory of borrowings in other languages. As there is already enough evidence to formulate the "universale" below, further findings can only lead to additional universals and to further refinement of single claims, but not to a falsification of the theoretical framework already achieved. There are many implicit consequences in the following statements which require a thorough elaboration for different fields, including the study of lexical structures in historical linguistics, the evaluation of lexical transfers un-

186

Language in ethnicity: a view of basic ecological

relations

der qualitative aspects, the analysis of synonyms under the conditions of language contacts, etc. As those aspects constitute special problems of linguistic methodology, they can only be indicated in this contribution, which concentrates on language contact universals. Although the universal principles below have been formulated on the basis of empirical findings about lexical transfers in the area of parts of the body, they nevertheless are abstractions and do not represent statements about empirical similarities or common features. The principles are abstracted from empirical findings in such a way as to refer to the mechanisms of language contact, as well as to the strategies involved in lexical transfer. The universals, therefore, deal with principles of language contacts, whereas the substance (concrete lexical transfers) has only the value of illustrating concrete realizations of the detected principles. Some universals (i.e., 4, 7, 8, and 9) seem to have a wide range of application and further research may reveal that they apply to other types of contact and transfer in addition to those discussed here. The following ethnolinguistic principles are basic, I believe, for the potential transfer of terms to designate parts of the body.

1. Borrowings in the lexical area of parts of the body may occur in languages of different genetic origin and genealogical affiliation. This may be considered to be the most general of all the findings concerning this subject. It is important to stress the fact that the phenomenon is not restricted to a single group of languages or to a language family. Otherwise, this ethnolinguistic phenomenon could be considered to be one of languages with a special affiliation, and therefore be restricted to a certain area and to a special type of language contact. As there are no such restrictions, however, it is appropriate to state that this phenomenon does not depend on a special genealogical grouping and is therefore not dependent on the genetic storage of cognate words or the distribution of cognates and loans in the vocabulary of a language. The investigations carried out so far have revealed borrowings of this kind in languages with different stocks of cognate words and varying lexical structures: Romance languages (see Zauner 1903) Celtic languages (see Haarmann 1970, 19ff. for Welsh, Haarmann 1973, 86ff. for Breton) Albanian (see Haarmann 1972, 48 ff.)

The adoption of foreign cultural patterns

187

Basque (see Haarmann 1970, 183 ff., 1983b) Balto-Finnic languages (see Haarmann 1983b for Izhorian, Veps, Karelian, and Finnish) Turkic languages (see Akylbekova 1966, especially for Uzbek and Uighur) Tungus languages (see Myreeva 1975 for Evenki) Indo-Aryan languages (see Haarmann 1984d for Romany, for Urdu, see below) Malay languages (for Iban, see below) Swahili (see below) The phenomenon of borrowed designations for parts of the body has so far been illustrated for a variety of languages from four main language families (Indo-European languages: Romance, Celtic, Indo-Aryan languages, Albanian, etc.; Uralic languages: Izhorian, Veps, Karelian, Finnish, etc.; Altaic languages: Turkic, Tungus languages; and Malay languages). When including Basque, Japanese, and other isolated languages (the genetic affiliation of which is uncertain or not known), the variety becomes even greater. Further analyses may detect this ethnolinguistic phenomenon in a number of other language families and groupings.

2. Borrowings are characterized by a non-selective occurrence among designations for the human body as well as that for animals (that is, borrowings may occur among terms for humans and for animals). When analizing the loanwords in single languages in this special lexical category, a general différenciation between designations for the human body and those for the body of animals has to be taken into consideration. Usually there are designations in any natural language that can be classified according to a threefold superstructure. Thus, cognate words as well as borrowings can be grouped according to these three distinctions: 1) terms for parts of the human body as well as that of animals, 2) for the human body alone, and 3) for the body of animals alone. A list of Latin borrowings in the Welsh language is given in Table 63 to illustrate this categorization. The examples under (1) refer to terms for parts of the body which can be used in reference to both man and animal. Thus Welsh coes "leg" can designate the leg of a human in one context and the leg of a bird in another. The same is true for the other examples. Welsh caul designates the stomach of an animal, whereas the stomach of a human is referred to

188

Language in ethnicity: a view of basic ecological relations

Tab. 63. Borrowed designations for parts of the body of Latin origin in Welsh 1. Designations for parts of the human body as well as of animals Latin origin

Welsh equivalent

Meaning

ARCA

arch

1. 'body' 2. 'trunk' (of the body)

ARTIC(U)LU

erthygl (Middle Welsh)

BAC(U)LU

bagl

1. 'joint' (general) 2. 'joint' (of hand and arm) 3. 'ankle' 1. 'leg' 2. 'shank'

CANDÉLA

cannwylly llygad

'pupil' (lit.: 'light of the eye')

CIPPU CONSOL(I)DU

cyff

'upper part of the body' 'knuckle'

CORPUS

corff

CORPORA

corffor

COXA

coes

CUNEU FONTANA

cyn ffynnon

'penis'

MANICA PANTICE(M)

maneg pennyg

'vagina' 'internal organs'

PATELLA

padelleg (deriv.)

'knee-pan'

PERMEDIU *PÏPA

perfedd

'intestines'

PÜPILLA

pubell (Middle Welsh)

cyswllt

cymhibau (deriv.)

1. 'body' 2. 'trunk' (of the body) 'body' 1. 'leg' 2. 'shank' 'gland'

'lungs' 1. 'eye' 2. 'pupil'

2. Designations for parts of the human body alone Latin origin

Welsh equivalent

Meaning

ARCA

arch

'chest'

BARBA

barf

'beard'

¡'he adoption of foreign cultural patterns

189

Tab. 63. Borrowed designations for parts of the body of Latin origin in Welsh

BRACC(H)IUM BUCCA

braich boch

CANÀLE(M)

canol

CAUPALU CAVELLA

ceubal cawell

CAVITÀTE(M)

ceudod

CISTA

cest

C(0)ÀG(U)LUM

caul

CORÓNA

corun

CULLEU

cwll (Middle Welsh)

FURCA

fforch

PALMA PLËNU

palf plwyniau (deriv.: Middle Welsh)

arm 1. 'cheek' 2. 'mouth' 'waist' 'stomach' 1. 'belly' 2. 'breast' 1. 'womb' 2. 'belly' 1. 'belly' 2. 'big belly' 'belly' 1. 'rist' 2. 'upper part of the skull' 1. 'bosom, female breast' 2. 'belly' 'fork-like shape of the human body' (front or back view: trunk + legs) 'palm' (of the hand) 'hair on an adult's body' (breast, etc.)

3. Designations for parts of the body of animals alone Latin origin

Welsh equivalent

Meaning

»ASCELLA

asgell

BRACC(H)IUM

braich

1. 2. 3. 1.

CAT(H)EGRA cadair C(0)ÀG(U)LUM caul CORNU CUNEU

corn cyna (deriv.)

'wing' 'feather' 'spine'(of fish) 'wing'

2. 'foreleg' 'udder' (of cow or sheep) 'stomach' 'horn' 'to be in heat' (of dogs)

190

Language in ethnicity: a view of basic ecological relations

Tab. 63. Borrowed designations for parts of the body of Latin origin in Welsh CÜPELLA

cibell

1. 'skin' 2. 'fur'

PALMA

palf

PAR(I)ËTE(M)

parwyden

'side of an animal fenced with

(deriv.: Middle Welsh)

ribs'

PLLJMA

pluf, plu

'feathers'

SPÀT(H)A

ysbawd (Middle Welsh)

3. 'crust' (of water animals) 1. 'paw' 2. 'small paw'

1. 'shoulder-blade' 2. 'shoulder-bone' 3. 'shoulder'

Source: Haarmann (1970, 19ff.)

as ceubal. Some designations fall under both (2) and (3) respectively, with, however, a special meaning attached to each occurrence (e.g., braich meaning "arm" in reference to the human body (2), but "wing" in reference to birds (3), and "foreleg" in reference to quadrupedes (3), or palf meaning "palm of the human hand" (2) but "paw" in reference to an animal (3)). There are also cases of morphological and phonetic variations of borrowings which have been integrated into several sections of the vocabulary. For example, Russian Ijazka has been adopted as Izhorian I'aska, with the meaning of "thigh" for humans (2) and as Izhorian l'äska, meaning "loin", "hip", referring to animals (3). Another example is Russian lëgkoe/lëgkie, which has been adopted as Izhorian I'ohkoiD (from Russian lëgkie, plural form) "lungs" of humans (2) and as Izhorian lohko (from Russian lëgkoe, singular form), meaning "lungs" of animals (3)· The distribution of borrowings in many other languages also indicates this kind of threefold categorization. Thus borrowings, when integrated into the vocabulary of a given contact language, are usually treated like cognate words with similar conceptual distributions.

The adoption of foreign cultural patterns

191

3. Borrowings are not restricted to the onomasiological "parts of the body" but also likely occur in adjacent lexical areas. Investigations into the lexical structures of languages in contact reveal that the potential influence of a contact language is not limited to the area of parts of the body, but usually also influences adjacent lexical areas (or sections of the vocabulary conceptually associated). When there is a number of borrowings to designate parts of the body in a given language, it is most probable to also find loanwords to designate defects, functions, and other characteristics of the body. Thus, the lexical area of parts of the body is not isolated in the vocabulary, but bound in a network of associated concepts. Some examples of borrowings in Iban (a Malay language in Sarawak, in the northwestern region of Borneo/Kalimantan, Malaysia) and in Romany serve to illustrate this:

Sections of the vocabulary

Examples from Iban

Examples from Romany

jari (

~

»

M ¡S O 5 O. öfi

-W « S? ^ « ?

cs ce T3 C ce

r¡ • i r-

relations

ε o o

U-

Ή 1) Ό Ό > l·« Ih S o o

•Í5 e« ni

u 2 CA X S 3 Λ Λ

SP !Λ* [Λ* W

C ,ο

3 Χ Έ

υ •α ce

Ο ο

Τ3 Ο

Xt ce ο.

I

I

Ô

U •ο 3 ο Λ

Ό CS

>ο τ-Η r-'

υ

υ •Ό -α ce rS

m CS

c ο υ u

£

3

υ

o

οο

M

c IS

ß

8

υ

od

'C

•o s-, Ό O X

o X υ

JS M

ra

e 3

OH

ω 2 3 o

M

o

o

en

es o

Ό Β 3 O (Η 0 0

o

.y S e Β

τ-Ι

Tf

VO Ti; ι—3 CS

r»· TtVO

en

es

es oo

CS 00 TT

ω c υ

M ®

Β S S 'C Ü

g> S g) fi

ω S1 3 W) S ¿3

σ\

CS CS od

Ü



υ υ e •"fr >Λ)

3 υ υ Ο

SO m

en oo

Os (N

in vi

fr υ e

•—s u υ ce * > υ 0 0 cu CA c ce ω Uh c υ •5 3 Ό 0 ô Un c Un w

C

s

υ υ e ce α. ο. cd υ

υ cd

ε c o

o o.

es es

Tt Ti-

O es

fi υ c u

ε tH Cl

ε

e '3

a ω χι



a. o •C

υ JS

235

236

Language

in ethnicity:

a view of basic ecological

relations

between product type and choice of a particular type of foreign element to promote it. The highest percentage for the category "foreign scenery", for instance, is 37.5%; this is the figure for the product range of transportation and machinery (10 in Table 73). The number of commercials in this range, however, is comparatively low. This is also true for the corresponding percentage in the case of fashion goods (5 in Table 73). The most we can say is that, among all the strategies available to advertisers for promoting products and services, the use of one or more foreign elements is often, but by no means always, favored. The highest percentages for foreign elements are found with respect to the following categories in different product ranges: 50.5% for characters in commercials for fashion goods, 44.6% for characters in settings for transportation and machinery, 39.3% for the written language in commercials for the latter product range. Percentages for foreign elements at a level above 30% are found in commercials for publications (6), general industrial equipment (7), transportation and machinery (10), shops and department stores (13). Overall, the most-preferred across the 16 product ranges is the participation of foreign characters. The break-down of figures in Table 74 thus reveals the same tendency as has been pointed out for the relations in Table 73 (see statement 2 in particular). When the product ranges "food and drinks", "cosmetic and hygienic articles", and "transportation, machinery" are further broken down, this same trend toward foreign characters in Japanese commercials overall becomes clearly visible in Table 74. The popularity of foreign characters in commercial settings — that is, foreign VIPs presenting Japanese products, foreign actors in action scenes or associated with European, American or Australian scenery - may be partly explained by the intention of commercial managers to make the spectators feel like members of a "cosmopolitan" community (see statement 18). It is noteworthy that the highest rates of foreign elements are observed in commercials for products which are among the most typical goods in a consumer's society. Cars, drinks and fashion goods (clothes and accessories) are meant for consumers regardless of sex or social status. Agegroup limitations are perhaps relevant in the case of cars, alcoholic drinks, and cosmetic articles. On the whole, however, such products are among the goods with the widest range in the Japanese consumers' society. These products thus differ in character from office equipment, raw material or other goods with special purposes. Since the highest rates of foreign elements are found in product ranges which are the most sensi-

T h e

use

o f f o r e i g n l a n g u a g e s as

υ Xi Λ •C α

Ο ~

s¿

en en tí

s y m b o l s o f p r e s t i g e in

es es