I am interested in how a symbolic system like language affects conceptual representations. Words act as pointers to concepts (mental representations of categories), but they are not the concepts themselves. Interesting questions here include: What does having a word do to a concept? Is a concept less well defined when you do not have a word for it? Are we less likely to form concepts for which we do not have words? I investigate these questions by looking at cross-linguistic variation. I look at a domain (motion) where languages differ regarding which situations in the world get linguistically encoded (e.g., “skip”, “hop”, and “leap” all map into a single verb in Spanish, “saltar”). I combine vector space models of semantics with neuroimaging (fMRI) paradigms to explore how well the semantics of a language map onto the mental representations we can read out of the brain. This is possible today due to exciting developments in multivariate pattern analysis. The broad goal of my project is thus to understand the mapping between language and concepts.
Share this page