Gå til hovedinnhold

You are now on UiA's old website. The information you find here may be outdated.

Visit our new website

0
Jump to main content

Spoken word production in Norwegian-English bilinguals. Investigating effects of bilingual profile and articulatory divergence

Malin Toften Mangersnes photo

In this project I investigated how individual differences in bilingual profile and articulation affect bilingual word production.

Malin Toften Mangersnes

PhD Candidate

Malin Toften Mangersnes will defend her thesis Spoken word production in Norwegian-English bilinguals. Investigating effects of bilingual profile and articulatory divergence for the PhD degree in Humanities and Education 8 December 2023

Summary of the thesis: 

During bilingual word production both languages are to some degree active, and control mechanisms are therefore required to produce the intended word. These control mechanisms have been investigated in language switching experiments. Previous studies have shown that general language proficiency differences affect the speed and accuracy of bilingual word production (e.g., Costa & Santesteban, 2004). Cognates are words that share form and meaning across languages, for instance hammer in Norwegian and English, and often facilitate word production (e.g., Costa et al., 2000; Hoshino & Kroll, 2008). Cognate effects have been partly attributed to cross-linguistic phonological similarity, but the degree of similarity between cognates is subject to individual variation. Therefore, this thesis particularly focused on individual differences in the domains of phonology and accent.

In this project I investigated how individual differences in bilingual profile and articulation affect bilingual word production. The participants were native Norwegian speakers of English (aged 18-35). In two experiments they named pictures in Norwegian and English. First, pictures were named separately in each language before the participants completed a language switching task. In the language switching task symmetric switch costs were observed, i.e., naming latencies were similar when switching from Norwegian to English and from English to Norwegian. In the single-language context, naming in Norwegian was faster than naming in English, but in the language switching task a reversed dominance effect was observed as naming was faster in English than in Norwegian. This supports previous findings of a control mechanism which affects the first language. Previously reported effects of cognate status, language and language switching were also reproduced with a new group of bilingual participants. 

Several methods were employed to assess individual differences. Four different tests measured language proficiency and phonological processing in each language. A bilingual profile questionnaire (Marian et al., 2007) was augmented to obtain more information relating to accent, phonology, and language switching. The responses to the questionnaire were factor analysed and questions relating to phonology and accent grouped together, suggesting that the augmentations contribute to the assessment of bilingual language profile. Two factors extracted from the questionnaire, one relating to general English proficiency and the other to English accent, were used as individual difference measures in the word production analysis. There were no significant effects of general English proficiency. Faster naming in the simple picture naming task was associated with higher values on the English accent factor. In the language switching task the effects were marginal. Two individual difference measures of articulatory divergence were calculated for two pairs of Norwegian and English vowels. There were limited effects of articulatory divergence on bilingual language production. The observed effects in the picture naming task were in the direction of faster naming with more divergent production.

The findings confirm the general facilitatory effects of cognates in word production. However, they do not appear to be strongly modulated by the measures of individual differences assessed here. The results suggest that bilingual word production is to some extent influenced by individual differences in the domains of phonology and accent, but for clearer answers the measures used need to be further developed. This project constitutes a first step in the investigation of how individual differences in bilingual profile, particularly relating to the domains of phonology and articulation, affect bilingual word production.  

Find more information about time and place for the doctoral defense.