We are interested in understanding learners’ biases with regards to the relative ordering of nominal modifiers (specifically demonstratives, numerals, and adjectives). Learners faced with a new artificial language tend to assume orders of these modifiers that correspond to the linguistic typology. We test the robustness of this effect by examining diverse linguistic populations in an attempt to understand if these preferences derive from an underlying cognitive bias, or are rather transfer effects from learners’ L1.
The aim of this project is to examine different parameters that come into play during the learning of different kinds of phonological rules (specifically phonetically natural and typologically abundant rules compared to unnatural, unattested ones), and the different levels at which a learning bias can be observed. We explore this “naturality” effect in different tasks (based both in perception and production), and different populations, and by modulating variability in the input, and exploring differential effects of memory consolidation after sleep. We also study the role of bias in transmission over time with the help of computer simulations.
As phonological contrasts can be lost over time, so too can they emerge. This project focuses on a current emergence in Dutch (namely /ɡ/) with the specific aim of exploring the different levels at which phonological emergence can occur (perception, production, the lexicon), and what social (foreign language knowledge/use, education, region) and linguistic (phonological and phonetic specificities of the language) factors may be at play.
This project focuses on asymmetrical processing of phonological
features during speech perception, both in prelexical phonological
processing, and in word recognition. Its goal is to tease apart
lexical bias from low-level acoustic bias. It includes a
computational component focusing on measuring lexical
Looking at the classic bouba-kiki effect, we explored the role of different segment types (specifically consonants vs. vowels) on sound–shape associations in adults. We found that in line with research on lexical access, consonants seem to have a much more important role in these associations. We also explored the effect in prelexical infants and found that 5–6 month-olds were not sensitive to such sound symbolism.