This paper is based on data from homeland and heritage varieties of Croatian. Homeland varieties are looked at to see what has and can happen ‘at home’, particularly in non-standard varieties. Heritage Croatian data comes from a cross-national project spanning four generations and nine countries (Hlavac, Stolac 2021).
Four types of loan translation are looked at: (1) semantic extension; (2) word-for-word/morpheme-for-morpheme renditions of expressions/collocations/VPs; (3) copying of argument structure; (4) other instances of template-copying with conspicuous morphosyntactic features. The following observations can be made:
(1) is infrequent. An addition or change in semantic meaning is more readily achieved through global copying of lexical items that function as hyponyms.
(2) is frequent in both homeland and heritage varieties. Some calques were not included in the standard Croatian, but other ones created to replace global copies of foreign-origin lexemes. Amongst these are NOUN+NOUN structures which are rare but with implications as a model that may be increasing in use. In heritage varieties, the proportion of recorded loan translations remains relatively high, perhaps as a favoured strategy to avoid code-switching which some speaker informants may feel is dispreferred in recorded interviews.
(3) is rarely found in homeland varieties, more so in heritage ones (structure of complex NPs, sequencing of attributives according to foreign models). Forms of possessive marking also belong to this group.
(4) is found in heritage varieties only, such as changes in valency marking for some verbs and negative polarity rather than negative concord which is otherwise required in Croatian. The copying of other-language NP templates is recorded resulting in use of the numeral jedan ‘one’ as a nascent indefinite article amongst some speakers.
Hlavac, J.; Stolac, D. (Eds.) 2021 Diaspora Language Contact. The Speech of Croatian Speakers Abroad. Berlin: De Gruyter Mouton.