Previous literature suggests that musicality benefits the perception of prosody in a second language. However, there is less literature that covers the effect of musicality on the production of...Show morePrevious literature suggests that musicality benefits the perception of prosody in a second language. However, there is less literature that covers the effect of musicality on the production of prosody in L2. This study examined 134 Dutch bilingual speakers of English with varying musical experience in their production of 11 English words during a reading-aloud task. This experiment was done in order to find out if Dutch native speakers acquire English prosody more easily when having experience with musical training. The pitch contours were measured and analyzed together with the amount of musical experience using a multi-level regression. The results that were found were insignificant and argue that the Dutch participants did not benefit from the musical training. This result contradicts previous literature, leading to an evaluation of previous results and theories.Show less
Ambiguity resolution has been a topic of debate in language processing models. The present study investigated the impact of punctuation and working memory in Garden Path structures and related...Show moreAmbiguity resolution has been a topic of debate in language processing models. The present study investigated the impact of punctuation and working memory in Garden Path structures and related these concepts with assumptions made under the Good Enough parsing approach. A word-by-word self-paced reading task was used to assess the effects. A backwards digit span and a reading span task were administered to measure the participant’s working memory capacity. The study found evidence for the presence of Garden Path effects in Late Closure structures. In the form of response accuracy to the comprehension questions, the data supports the Good Enough parsing approach’s assumption that ambiguities are not fully resolved. The scores on the backwards digit span task correlated positively with response accuracy on comprehension questions while reading span scores did not. Punctuation did not impact reading time of the disambiguating region nor response accuracy.Show less
Gesture annotation is a fundamental process in all types of gesture research. However, it is usually performed manually, making it a time- and resource-consuming process. Recent progress in...Show moreGesture annotation is a fundamental process in all types of gesture research. However, it is usually performed manually, making it a time- and resource-consuming process. Recent progress in automatic human motion tracking has sparked the development of tools that partly automate gesture annotation, such as the tools by Ripperda et al. (2020) and Ienaga et al. (2022). Although aimed at annotating signs, the tool by Fragkiadakis et al. (2021) could also be useful for gesture research. As the tools are only recently developed, it is not yet confirmed whether they can facilitate the annotation process. In this thesis, the extent to which the tools can be made operable is tested through User Acceptance Testing, and their performance and usability are evaluated based on criteria sets. It was observed that the tools by Ripperda et al. (2020) and Ienaga et al. (2022) still contained fatal errors and failed to generate annotations, whereas the tool by Fragkiadakis et al. (2021) was easily made operable. However, the accuracy of the produced annotations was low and robustness could be improved, for example by training a new model. Overall, the results indicate that the tools by Ripperda et al. (2020) and Ienaga et al. (2022) still require development, and that the sign language-based tool by Fragkiadakis et al. (2021) could potentially be used for facilitating gesture annotation.Show less
In linguistics, coming up with a certain continuation in a sentence before even reading or hearing that sentence is called prediction. People pre-activate upcoming possibilities when reading...Show moreIn linguistics, coming up with a certain continuation in a sentence before even reading or hearing that sentence is called prediction. People pre-activate upcoming possibilities when reading earlier words in a sentence. In this study a sentence completion study, a likelihood scale questionnaire and a reading time experiment are conducted to test this effect called prediction in a semantically constrained context. That participants can be lead to a certain semantic expected word is found in the sentence completion task. The likelihood scale questionnaire gave us insight in how likely the most frequent and less frequent given instrumental noun continuations were and provided us with the sentences for the reading time experiment. In this reading time experiment, there is found a significant effect, given a same specific constrained semantic contexts, that expected logical semantic instrumental nouns are read faster than unexpected illogical instrumental nouns in Dutch.Show less
