Bestemt og ubestemt : En korpusbasert studie av morsmålets rolle for bruk av bestemte og ubestemte former hos norskinnlærere uten grammatisk bestemthet i morsmålet

This thesis presents a corpus-based study of first language influence (L1 transfer) on the use of definite and indefinite forms of the noun. It investigates texts written by learners of Norwegian whose first language is without grammatical definiteness (-GD-learners). The aim is to investigate two potential effects of L1 transfer: Investigation 1 examines a quantitative effect on the use of the definite form in obligatory contexts, and investigation 2 examines a qualitative effect on the choice between definite or indefinite forms in different linguistic contexts. The study makes use of a set of methodological criteria for transfer research proposed by Jarvis (2000, 2010), which involves the investigation of four different transfer effects in learner language: within-group similarity, between-group differences, cross-language congruity and intralingual contrasts. When these four effects can be found together in learner language, they constitute evidence of transfer by proving that a specific linguistic behaviour is motivated by the first language and that it is also specific and representative for the group of learners that are studied. The learner language data are drawn from the Norwegian learner corpus ASK, and the study is to a large extent corpus-based, but also includes additional manual annotation of parts of the data. Investigation 1 is based on a corpus of texts written by learners with Polish, Russian or Serbo-Croatian (-GD) as their first languages, and this group of learners is compared to a group of learners with English, Dutch and German (+GD) as their first languages. Investigation 2 studies a smaller corpus of learners whose first languages are Polish (-GD) and English (+GD), and also includes texts written in L1 Norwegian. While investigation 1 simply looks at the percentages of correct suppliance of definite suffixes in obligatory contexts, investigation 2 investigates transfer through two different hypotheses on how L1 Polish may influence the choice between definite and indefinite forms in the Polish informants’ learner language. The first hypothesis is about transfer of pragmatic word order as a signal of a noun’s identifiability, including an L1 based assumption that identifiability is already signalled when pragmatic word order is used. This hypothesis is based on earlier work on how communicative redundancy affects the use of articles in learner English, especially Young (1996). Like Youngs’ study, Hypothesis 1 predicts that definite suffixes and indefinite articles will be considered redundant when their function (identifiability) is already signalled by word order, but suffixes and articles will be supplied when identifiability is not signalled, that is when pragmatic word order is not used. The second hypothesis is about transfer of the functional distinctions underlying the choice of form within the grammatical category case in Polish. Hypothesis 2 predicts that definite and indefinite forms will be used to express the distinction between functionally marked and unmarked cases, that is between the nominative and the other cases (non-nominatives). This means that in the Polish learners’ texts the indefinite forms will be used in linguistic contexts associated with the nominative in Polish (grammatical subjects), and that definite forms will be used in linguistic contexts associated with non-nominatives in Polish. The results of the study are positive for investigation 1: There are significant differences between -GB-learners and +GB-learners when it comes to how often definite forms are used in obligatory contexts, and it is concluded that this is due to L1 transfer. Investigation 2 comes out negative for both hypotheses: The definite and indefinite forms in Norwegian are neither used in accordance with redundant and non-redundant contexts, nor to mark the distinction between nominatives and non-nominatives. Instead, it is revealed that the Polish learners to a great extent use forms in accordance with the rules of the target language: Indefinite forms are used when the nouns have indefinite reference, definite forms are used when the nouns have definite reference, and both forms are used for generic reference. It should be emphasised that the Polish learners did have linguistic choices that were in accordance with the predictions of the hypotheses about L1 transfer in some of the contexts. But when these L1 like tendences were compared to the linguistic choices of the Norwegian L1 group, it was revealed that the Polish learners’ choices occurred in contexts where the implications of the hypotheses were the same as the typical choice of form in the target language. In linguistic contexts especially elicited to highlight where the implications of the hypotheses were not the same as in the target language, the Polish learners’ choice of form was not in accordance with the hypotheses about L1 transfer. This finding proved to be of special importance for the conclusion that the Polish learners’ use of definite and indefinite forms probably was influenced by the target language, rather than their first language. A secondary finding of the study was two cases of L1 transfer in the English learners’ choice of form. The first case was an overrepresentation of the indefinite article when reference was indefinite, and the other case was an underrepresentation of definite forms when reference was generic. This transfer was explained by two differences between English and Norwegian: Indefinite articles are more often used in English than in Norwegian when reference is indefinite, and definite forms are less often used in English when reference is generic