| Keywords [eng] |
psycholinguistics, code-switching, word frequency, switching costs, psicholingvistika, kodų kaita, žodžių dažnumas, kodų kaitos kaštai |
| Abstract [eng] |
This master’s thesis examined spoken-word comprehension in a multilingual environment involving Lithuanian, English, and Spanish. The aim of the study was to determine whether lexical frequency affects spoken-word recognition, whether code-switching slows down language processing through what is known as switching cost, and whether this cost differs depending on the specific language transition. The study focused on comprehension rather than language production, therefore participants were not asked to pronounce words but instead had to select the correct picture after hearing an auditory stimulus. Thirty multilingual participants took part in the study. Most of them identified Lithuanian as their native language, therefore Lithuanian can be considered the dominant language for the majority of participants in this study. The experiment consisted of two main blocks. In the first, monolingual block, words were presented in separate languages. This block was used to evaluate baseline processing speed in Lithuanian, English, and Spanish and to examine the effect of lexical frequency. In the second, multilingual block, the languages were mixed, which made it possible to analyse language switching. The results of the monolingual block showed that Spanish words were processed the slowest. Lithuanian and English response times were more similar, while Spanish stood out with longer processing times. The lexical-frequency analysis showed that the frequency effect was not identical across all languages. Spanish showed the clearest negative relationship between Zipf frequency and response time: more frequent words were recognised faster. Lithuanian also demonstrated a weaker but statistically significant frequency effect. In English, no reliable negative relationship between frequency and response time was found. It was found that code-switching costs were not identical across all language combinations. The strongest switching costs were observed when switching from Lithuanian to English and from Spanish to Lithuanian. In contrast, switching into Spanish did not show a reliable additional switching cost after extreme reaction times were removed. This suggests that, in multilingual spoken-word comprehension, it is important not only whether switching occurs, but also between which languages the switch takes place. Overall, it can be concluded that multilingual spoken-word comprehension depends both on lexical accessibility and on language-control processes. Lexical frequency influenced word recognition, especially in Spanish, while language switching caused an additional slowing of reaction times. The study also showed that switching cost is directional and depends on the specific language combination involved. Future research would benefit from including a larger participant sample, more detailed measures of language proficiency and dominance, and a more balanced structure of language transitions. |
