The nominal domain in Dutch-English code-switching: insights from acceptability judgments

Objectives/research questions: This thesis looks at Dutch-English determiner-noun code-switches and investigates whether there is a preference for the language of the determiner, whether the position of the switched nominal construction and language dominance affect these preferences; and whether these preferences support the predictions of the Minimalist Program (MP), which predicts that the languages with more phi features provides the determiner, or the Matrix Language Frame Model (MLF), which predicts that the determiner will come from the Matrix Language (ML). This thesis also investigates how grammatical gender assignment is resolved in a code-switch involving a Dutch determiner followed by an English noun, and how language dominance affects this. Methodology: The participants (N = 68, aged between 20 – 77) were split into a Dutch-dominant group, an English-dominant group, and a balanced Dutch-English bilingual group, based on their dominance score, which was calculated from...Show moreObjectives/research questions: This thesis looks at Dutch-English determiner-noun code-switches and investigates whether there is a preference for the language of the determiner, whether the position of the switched nominal construction and language dominance affect these preferences; and whether these preferences support the predictions of the Minimalist Program (MP), which predicts that the languages with more phi features provides the determiner, or the Matrix Language Frame Model (MLF), which predicts that the determiner will come from the Matrix Language (ML). This thesis also investigates how grammatical gender assignment is resolved in a code-switch involving a Dutch determiner followed by an English noun, and how language dominance affects this. Methodology: The participants (N = 68, aged between 20 – 77) were split into a Dutch-dominant group, an English-dominant group, and a balanced Dutch-English bilingual group, based on their dominance score, which was calculated from their global language score as measured by the Bilingual Language Profile questionnaire. Participants completed two two-alternative forced-choice acceptability tasks (2AFC); one where they evaluated the acceptability of sentences with code-switches between the determiner and the noun that reflected the predictions of the MP, the MLF, of both, or none. The second task tested which Dutch grammatical gender the participant assigned to an English noun. Data and analysis: Results from the first 2AFC were analysed using Thurstone’s Law of Comparative Judgment, while the results of the second 2AFC task were used to create a frequency distribution of Dutch grammatical gender marking. Results: A general preference for an English determiner followed by a Dutch noun was found, though this preference is influenced by the matrix language of the sentence, the position of the code-switch in the sentence, as well as language dominance of the bilingual. The first task provided evidence supporting both the predictions of the MLF and the MP, as participants preferred a Dutch determiner followed by an English noun in pre-verbal position, which is what the MLF predicts, but participants preferred an English determiner followed by a Dutch noun in post-verbal position, which is what the MP predicts. Results from the second task suggest the common determiner de is preferred for English nouns by all participants, regardless of language dominance and regardless of the grammatical gender of the translation equivalent. Originality: Following Parafita Couto and Stadthagen-González’s methodology (2019), this study is one of the first studies to use the 2AFC acceptability judgment task analysed using Thurstone’s Law of Comparative Judgment to directly contrast the predictions of the MP and MLF, and is the first study to do so while controlling for language dominance. Furthermore, this study investigates a language pair that has yet to be researched extensively, in addition to investigating how grammatical gender assignment is resolved. Implications: Results provide evidence for a default gender assignment strategy, as well as evidence for a difference in code-switching preferences depending on language dominance. Furthermore, the results support Parafita Couto and Stadthagen-González’s (2019) suggestion that a theory that combines both the MLF and MP would be more useful to explain Dutch-English code-switches than either framework separately.Show less

Leiden University Student Repository

Documents

In Collections

The nominal domain in Dutch-English code-switching: insights from acceptability judgments