Sam Bond

The ability to extract statistical information from speech is an important part of language perception. We use this mechanism, statistical learning, to locate word boundaries in speech or determine rules, like grammar. We also acquire rules governing language sounds, its phonology, based on statistical properties (e.g. English words do not end in a /h/ sound). However, in phonology, not all statistical cues seem equally salient suggesting constraints on this learning mechanism. This PhD project, titled 'A Detailed Exploration into the Constraints on Statistical Learning', aims to explore such constraints using artificial languages.