Evangelia Adamou (LACITO, CNRS, Paris)
Several studies on language contact report the use of content words and especially nouns in most contact settings (e.g., Thomason & Kaufman, 1988; Muysken, 2000; Myers-Scotton, 2002; Matras & Sakel, 2007; Matras, 2009; Haspelmath & Tadmor, 2009). Also, a vast literature on language contact suggests that the degree of borrowing is related to extra-linguistic factors such as the intensity and type of language contact, as well as on language attitudes (e.g., Thomason & Kaufman, 1988; Muysken, 2000; Winford, 2003; Matras, 2009; Haspelmath & Tadmor, 2009).
In this talk I will present evidence relevant to this discussion from collaborative research conducted within the French-German ANR-DFG project EuroSlav, “Electronic database of endangered Slavic varieties in non-Slavic speaking European countries”, co-directed with Prof. Walter Breu. This study is based on the analysis of free-speech corpora from four Slavic minority languages traditionally spoken in Austria, Germany, Greece, and Italy. The corpora were collected among fluent speakers of the Slavic languages in a language documentation perspective and are freely available online http://lacito.vjf.cnrs.fr/pangloss/.
A quantitative analysis of the corpora, totalling approximately 34,000 word-tokens, was conducted in a variationist perspective with respect to the study of language contact (see Poplack, 1993; Adamou, in press). It appears that the Slavic-speaking communities of Austria, Germany, and Greece use surprisingly low rates of word-tokens from the current-contact language, i.e., German and Greek, despite fluency of the speakers in both languages. These communities can be thought of as ‘low borrowers’ similar to the typology proposed in Tadmor (2009). In contrast, the Molise Slavic speakers from Italy are producing significantly higher rates of word-tokens from Italian, and thus fall under the category of ‘high borrowers’ (Tadmor 2009: 57). A Random Forests analysis (Breiman, 2001) identifies ‘language’ as the main predictor for the ratio of both borrowings and noun borrowings. This result can be interpreted as an indicator of the existence of borrowing patterns in bilingual communities (also see Poplack, 1985; Adamou & Granqvist, 2014; Adamou, in press; Travis & Torres Cacoullos, in press). Finally, comparison of the results with existing sociolinguistic studies on these Slavic minority languages allows us to consider that these patterns of borrowing depend on the intensity and type of contact in the past and, perhaps more importantly, on language attitudes that stem from the existence or not of century-long literary traditions.