English language-Australian influence on serbian language

Abstract.

This paper explores change induced through contact with English in a variety of Serbian spoken by the children of Serb migrants, the so-called second generation, in Melbourne, Australia. Depending on the manner in which contact-induced change affects this immigrant variety of Serbian (AS), it is possible to differentiate between three categories of change:

(1) change that does not simplify the structure of AS and brings about increased similarity with English;

(2) change that simplifies the structure of AS but does not bring it closer to English; and

(3) change that simplifies an AS structure and makes it more similar to English.

Key words: contact-induced change, attrition, convergence, transference

1. Introduction

That an immigrant minority group will, as a result of prolonged contact, adopt the language of the dominant majority for in-group communication is commonly recognized as a truism in the literature on language shift (Romaine, 1989: 35; Grosjean, 1982: 103—07). Viewed diachronically, the shifting process is (generally) complete in the course of three generations: the immigrant first generation is dominant in the minority language, the second generation is bilingual yet generally displays a clear preference for the majority language, while the third generation is monolingual in the language of the dominant group.

Ethnolinguistic groups able to maintain an immigrant language into the fourth or fifth generation, although documented,1 are considered an exception rather than the norm. The minority language is initially the only language used across domains in interaction with group members; however, as bilingualism becomes prevalent, the majority language begins to encroach on spheres of interaction, such as the home domain, from which it had previously been excluded. The minority language is never replaced suddenly by the majority language as the language of interaction with group members; instead, language use gradually becomes variable. Viewed synchronically, language shift entails the use of both languages, for a time at least in less stable, immigrant communities, across the same domains. The source of language shift is in this 'synchronic heterogeneity': the linguistic diversity resulting from language shift is at once both the reflection of and impetus for continuing change (Appel and Muysken, 1987:
41).

As a language recedes, loses speakers and domains, it ultimately begins to lose structure. The loss of structure, phonological, morphological, syntactic, or discourse structure, is part of the gradual process of attrition. Since the lost linguistic material is not replaced by new material, for instance, from the dominant language, attrition is the most common linguistic route to language death (Thomason, 2001: 227). Attrition, however, is not the only kind of contact-induced change which it is possible to encounter in the course of language shift. Contact-induced changes in the minority language are often also due to interference, the direct importation of material and/or structures from the dominant language. Finally, many changes have multiple causes, with both attrition and interference from the majority language contributing to bring about the change. An example of a study which considers contact-induced changes in a receding language is Fenyvesi (1995).2 In her investigation of deviations from Hungarian as it is spoken in Hungary evident in an immigrant variety of the language spoken in the United States, Fenyvesi attempted to establish the sources of the structural changes she observed. She was able to identify three categories of changes:
(1) changes that made the minority language more similar to English without simplifying it, referred to as 'borrowing alone';

(2) changes that simplified the language without bringing it closer to English, referred to as 'attrition alone'; and

(3) changes that both simplified the language and made it more similar to English, referred to as 'both borrowing and attrition'.

Considering all three of the above types of change, this paper analyzes contact-induced changes in an immigrant language in a language shift situation, namely in Serbian as it is spoken by second generation Serbian/English bilinguals in Melbourne, Australia; in this it follows Fenyvesi (1995). A significant difference, however, between Fenyvesi's approach and the one adopted here is that the polysemous term 'borrowing' (cf. Poplack, 1980; Myers-Scotton, 1993; Thomason, 2001; Auer, 1991) is, in this paper, superseded in favor of the distinction between interference features resulting from the processes of transference and convergence (Clyne, 2003), respectively. Transference is the process whereby any „form, feature or construction [is] taken over by the speaker from another language, whatever the motives or explanation for this" (Clyne, 2003: 76).

Convergence, on the other hand, refers to the process whereby partial similarity between structures in one language (Serbian) in contact with another (English) increases at the expense of difference. Clyne (2003: 79—80) explains the basis on which, for example, syntactic convergence and syntactic transference are differentiated by citing two ways in which 'we went to school in Tarrington' might be expressed by German/English bilinguals:

a) Wir haben zu Schule gegangen in Tarrington we AUX + have to school go+PAST.PT in Tarrington

b) Wir haben gegangen zu Schule in Tarrington we have gone to school in Tarrington

c) Wir sind in Tarrington zur Schule gegangen. we AUX+be in Tarrington to.the school go+PAST.PT

A comparison of sentences a) and b) with sentence c), which is how 'we went to school in Tarrington' would be realized in Standard German, reveals a morpheme-for-morpheme correspondence with the English in sentence b), which thus constitutes an example of syntactic transference. Conversely, there is no morpheme-for-morpheme correspondence in sentence a), which converges towards the English.

The difference between the transference of syntactic constructions and syntactic convergence can be further illustrated by examples from the present corpus:

(1)
A svako počne da smeje mi
and everyone PR.3.SG that PR.3.SG I:DAT

Homeland Serbian:
A svi počnu da mi se smeju.
and everyone PL PR.3.PL that I:DAT R PR.3.PL
'And everyone begins to laugh at me'.

(2) Kad ja kažem svako smeje se na mene.
when I:NOM say R.1.SG everyone:NOM laugh R.3.SG R at I:ACC

Homeland Serbian:
Kad ja kažem svi mi se smeju.
when I:NOM sayR.1.SG everyone:NOM.PL I:DAT R laugh R.3.PL
'When I say [it] everyone laughs at me'.

The utterances in (1) and (2) feature two ways in which a second generation Serbian/English bilingual expressed the indirect object of the verb smejati se 'to laugh': the two realizations differ both from each other and from the realization in (monolingual) homeland Serbian (HS). In HS, when the indirect object is a pronoun, its unmarked realization will be a pronominal enclitic in the dative, rather than the full pronominal form; the indirect object will then precede the verb. This contrasts with the preferred English SVO word order. As a result, in (1) the enclitic appears in the position following the verb. The pronoun in (2) is not the indirect object but the object of the preposition na 'at': in fact, there is a morpheme-for-morpheme correspondence in (2) with the English 'laugh at me'. We conclude that (1) is an example of syntactic convergence, whereas (2) is an instance of semanticosyntactic transference.

Differentiating between the linguistic processes of attrition, transference, and convergence in the analysis produces a four-way distinction between changes resulting from: (1) transference only; (2) convergence only; (3) attrition only; and (4) both convergence and attrition, each of which will be considered in turn.