Polysemy and the nature of the mental lexicon

Terje Lohndal (Norwegian University of Science and Technology)

The mental lexicon is typically assumed to consist of lexical entries (‘words’). On a lexicalist view, these entries have a lot of information stored in them, information that typically projects into syntax. Anti-lexicalist views, on the other hand, assume a very different structure of the mental lexicon. For instance, approaches such as Distributed Morphology (e.g. Embick 2015) and Borer’s exoskeletal approach (Borer 2005a, b, 2013) hold that the mental lexicon consists of roots. Exactly what roots are differs among approaches, but they have no syntactic category and typically very little, if any, meaning. Carston (2019, 2021, 2023) argues that this anti-lexicalist view provides a fruitful foundation for understanding and modeling polysemy.

This talk will survey some of the evidence in favor of anti-lexicalist views of the mental lexicon and outline their basic assumptions. It will also address some challenges and avenues for future research.