Night Owls and Majestic Whales: Modeling Metaphor Comprehension as a Rational Speech Act over Vector Representations of Lexical Semantics

Abstract

While they are one of the few computational models that directly capture some of the pragmatic processes underlying language reasoning, current Rational Speech Act (RSA) models of metaphor do not align well with contemporary accounts of metaphor comprehension. The following research project leverages GloVe word vectors to capture pragmatic language reasoning in metaphoric utterances using an updated RSA framework. The model yields higher posterior probabilities for the same attributes that humans deem relevant in metaphoric utterances over erroneous ones in 89% of all cases, validating the use of word vectors to generate prior probabilities in an RSA framework. When presented with biased priors like listeners are in many naturalistic conversations, the model accurately matches human judgements of the top-most relevant attribute of a topic/target indicated by a metaphoric utterance 90% of the time.

Tasks

Reproductions