I will discuss aspects of the mapping problem for vocabulary: how learners discover which word expresses which concept. The commonly-held view is that this problem is solved by matching recurrent sounds (e.g., /dog/) to recurrent events or objects (e.g., dog sightings). In light of the loose and variable relations between word use and the passing scene, this observational procedure is assumed to operate piecemeal, with knowledge accruing probabilistically over repeated exposures. In the first part of this talk, I will argue that only a tiny subset of the lexical stock is acquired straightforwardly "by" observation, and that this observational-learning procedure, to the extent employed, is more one-trial than accrual. In the second part of the talk, I'll show a bit of how learners acquire the aspects of word meaning that are remote from perception.
To view a recording of this talk click here (You will need a Rutgers NetID and password) - NOTE: Talk recording is split into 2 parts