It’s tempting to think that lexicography can go on, untroubled by the concerns of theoretical linguistics, while the rest of us plunge into round after round of bloody internecine strife. For better or worse, as Patrick Hanks makes clear in Lexical Analysis: Norms and Exploitations (MIT Press, 2013), this is no longer true: lexicographers must respond to theoretical and practical pressures from lexical semantics, and this lexicographer has very interesting things to say about that discipline too.

Hanks’s central point is perhaps that the development of huge electronic corpora poses enormous problems, as well as exciting challenges, for the study of word meaning. It’s no longer tenable to list every sense of a word that is in common currency: and even if we could, it would be a pointless exercise, as the vast output of such an exercise would tell us very little about what meaning is intended on a given instance of usage. However, these corpora provide us with the opportunity to say a great deal about the way in which words are typically used: and the theory that Hanks develops in this book represents an attempt to make that notion precise.

In this interview, we discuss the impact of corpus-driven work on linguistics in general and lexical semantics in particular, and discuss the analogy between definitions and prototypes. In doing so, we find for Wittgenstein over Leibniz, and tentatively for ‘lumpers’ over ‘splitters’, but rule that both parties are at fault in the battle between Construction Grammar and traditional generative syntax.