Under the auspices of the WORDS project, we studied the mental component of the sound of language from multiple points of view. These include theoretical, as well as usage-based, and experimental studies of many different languages.
Our theoretical work on the WORDS project was based on the Featurally Underspecified Lexicon (FUL) model of phonological perception. This framework prizes economy in the representation of speech sounds in the mental lexicon: our working hypothesis is that the distinctive features of speech sounds have specific acoustic and articulatory definitions (see Speech Recognition), and crucially that the inventory of features is as small as is necessary and sufficient to describe the phonological segment inventory of any language in the world.
We explored the implications of this working hypothesis for the theory of phonology generally, by examining previous analyses that present challenges to FUL, and by showing how FUL deals with the phenomena it analyses.
A number of different studies were conducted, including:
- Vowel harmony in Uyghur.
- The inventory of coronal segments in Tahltan.
- The representation of secondary articulations generally.
- Phonological analysis of stød distribution in Danish.
Published papers included:
The study of word accent distribution in Norwegian compounds: Wetterlin, A. & Lahiri, A. (2012). Tonal alternations in Norwegian compounds. The Linguistic Review 29, 279-320.
- The historical development of word accent in North Germanic.
- The history of umlaut in German.
- The development of geminates in West Germanic.
- Behavioural & EEG studies of word accent in Norwegian.
- Behavioural & EEG studies of geminates in Bengali and Swiss German.
- EEG studies of vowel duration in Bengali and British English.