Belgian Journal of Linguistics - Volume 35, Issue 1, 2021

Codeswitching in the Turkish migration settings in Western Europe has been studied almost since the beginning of the labor migration that formed these communities. The patterns of codeswitching have gradually become more complex, which is demonstrated with reference to data from the Netherlands. Initially it was limited to the insertion of Dutch words into Turkish utterances, largely needed to fill lexical gaps. Gradually, as a new generation grew up with Dutch education, living most of their lives in a Dutch social environment, codeswitching patterns grew in complexity, with an increase in alternation between Dutch and Turkish utterances. Often, these and other patterns are attested in the speech of the same people at the same time, within a single conversation: they represent a bilingual speech style.

The most recent investigations into Turkish-Dutch codeswitching show how this increasing intensity of bilingualism in the community has led to increasing integration of the languages in everyday in-group bilingual speech. Of all switches, a relatively small percentage is taken up by the insertion of content words into either language; yet, codeswitching is not always straightforward alternation between Turkish and Dutch utterances. Instead, utterances often contain chunks from both languages. This phenomenon will be illustrated in this article and explained through a usage-based perspective that privileges a processing-based over a structuralist account, but crucially ties this account to an understanding of the social motivations that make this kind of bilingual speech possible, or even desired.

Viewed from an interactional sociolinguistic perspective, this research addresses multilingual codeswitching practices of youngsters with North African roots from the super-diverse Flemish city of Ghent. Particular attention is paid here to their artistic – mainly interethnic – performances, as these play a vital role in constructing ethnic and social belonging by voicing “glocal” identities. In so doing, the study provides evidence of the importance of multilingualism and codeswitching in the accommodation, socialization, and emancipation of these youngsters.

The present article focuses on semantic-pragmatic and sociolinguistic factors explaining the use of non-French words and constructions in the exolingual communicative context of French exams. On the basis of an oral and written exam corpus, this article looks into different types of cross-linguistic interactions to question the boundaries of codeswitching (CS) and propose a prototypical approach. Special attention is given to the correlations between the embedded languages (in this case, English or Dutch) and the type of influence these languages have on the participants’ French discourse. By taking into account the distribution of cross-linguistic interactions in the discourse of different groups of participants divided according to the levels of the European Framework of Reference for Languages (CEFR 2001), this study shows that, even if lower level participants resort to other languages to compensate for lexical deficiencies in French, they also choose to resort to non-French lexemes due to semantic and sociolinguistic factors, as is the case for advanced level speakers. From the perspective of language assessment, the notion of lexical error is central. Based on a typology of embedded units, this study aims at identifying the factors of acceptability of different types of transcodic marks.

In previous research, there has been an emphasis on differentiating and distancing translanguaging from codeswitching, partly on the basis that the latter refers to the combination of two discrete systems that correspond to named languages. While this is the mainstream view, there are codeswitching scholars who have proposed alternative views that align with some of the same observations and criticisms that have been raised by proponents of translanguaging. In this conceptual paper, I provide an overview of translanguaging alongside opposing views of codeswitching, and I underscore important similarities that have thus far been absent from present discussions regarding translanguaging versus codeswitching. Drawing on data from the understudied Spanish/English codeswitching variety spoken in Northern Belize, I discuss how bilingual compound verbs lend support to alternative views of codeswitching. Despite clear differences in their empirical goals, research conducted by both codeswitching and translanguaging scholars compels us to reexamine fundamental notions about language and linguistic competence. This reevaluation will not only contribute to theoretical advancement, but it will further elucidate our understanding of the complexity and dynamicity that characterizes bi/multilingual speech production and processing.

Grammatical gender has been thoroughly investigated in codeswitching (CS) research. When (only) one of the languages involved in CS has grammatical gender, it has repeatedly been shown that (simultaneous) bilinguals assign default gender to a codeswitched noun from a genderless language (Liceras et al. 2008; Valdés Kroff et al. 2017; Parafita Couto 2019). For Spanish, “[t]here is little doubt that masculine is the unmarked or default gender” (Harris, 1991, 43). When looking at Belgian Dutch however, the picture is a bit more complicated, as it is not clear which determiner is unmarked for gender. Rooryck (2003) argues that the neuter gender is the under-specified category, while the masculine and feminine are marked for gender. This would predict that English nouns would be preceded by a neuter determiner when incorporated in a Belgian Dutch sentence. However, it seems to be the case that masculine, rather than neuter, is the preferred determiner for English codeswitches into Dutch. This paper reports on an acceptability judgment task in which simultaneous bilinguals were asked to rate (Belgian) Dutch sentences containing an English noun on a 7 point scale. Results surprisingly show a preference for feminine gender, in addition to neuter being dispreferred. I speculate that this unexpected result is due to influence from Standard Dutch.

This study focuses on unveiling the strategies involved in gender assignment in codeswitching between two gendered languages: Dutch (common/neuter gender) and Portuguese (masculine/feminine gender). We draw on naturalistic speech (n = 32 speakers), elicited production (n = 35) as well as intuitional data (n = 57) from Dutch/Portuguese bilinguals stemming from three communities in Paraná, Southern Brazil, aiming to disentangle the relative roles of linguistic and extralinguistic factors on gender assignment. In unilingual Dutch, we find that Dutch/Portuguese bilinguals overgeneralize common determiners and adjectives to neuter nouns, similarly to other Dutch bilinguals outside the Netherlands (Clyne 1977; Clyne and Pauwels 2013; Folmer 1991; Giesbers 1997). In codeswitched constructions, however, speakers assign common and masculine gender as defaults, in line with the prediction that speakers of language pairs with no gender values in common prefer gender defaulting in mixed constructions (Klassen 2016). While extralinguistic factors such as age and relative use of the languages shaped unilingual Dutch production, the patterns during codeswitching were conventionalized across the speaker sample.

The diminutive construction is formed and used differently in Spanish and English, which leads us to the question how this construction with different morphosyntactic and semantic-pragmatic characteristics in the input languages is governed in Spanish-English bilingual and codeswitching speech. Through the analysis of a dataset of diminutive constructions extracted from the Bangor Miami corpus, this paper contributes to a better understanding of how one and the same construction differently represented in the input languages is administered in bilingual contexts. As this is a first approach to studying diminutives in codeswitching, three well known structural codeswitching models serve as a primary theoretical tool against which the diminutive is tested. These are Poplack’s (1980) Universal Constraints, Myers-Scotton’s (2002) Matrix Language Frame Model, and Blom and Gumperz’ (1972) Metaphorical Codeswitching Framework. The results show that Miami bilinguals prefer the prototypical markers of each language, -ito and little (e.g. un partimecito, un little estante). Furthermore, while the data largely confirm Poplack’s Constraints, they refute our hypothesis based on Myers-Scotton’s MLF model. Regarding Gumperz’ theory, the use of diminutive markers in a particular language correlates with a certain meaning the speaker wants to communicate (i.e. quantitative or qualitative), which again provides support to the framework.