David Poeppel, Ph.D.

Lecture Abstract

From Vibration in the Ear to Abstraction in the Head:
A Brain's-eye-view of Language

The inputs to spoken language comprehension are continuously varying waveforms -- yielding vibration in the ear -- but comprehension is built on very abstract representations in the head -- roughly, words. I present neurobiological research, informed by studies using fMRI, EEG, and, principally, MEG, that suggests (i) that there is a way to fractionate continuous signals into linguistically interpretable chunks and (ii) that there really is abstraction in the head. These observations leave one with the considerable challenge of connecting the biology and the linguistics in a serious way. Given the granularity mismatch between the representational elements/primitives, how are these domains of inquiry supposed to be linked? I argue that a theoretically motivated, computationally explicit, biologically plausible model of the neural basis of language requires linking hypotheses between the cognitive and biological ‘alphabets’ that are most plausibly stated in computational terms of a certain type. Building on the (typically discredited) neo-phrenological approach associated with much of contemporary neuroimaging, I advocate a 'computational organology' approach to the neural basis of language. The types of computations I suggest include multi-time resolution processing and analysis-by-synthesis.