Articulating the Perceptual Assimilation Model (PAM): Perceptual assimilation in relation to articulatory organs and their constriction gestures.

Abstract

A core premise of the Perceptual Assimilation Model of non-native speech perception (PAM) [Best (1995); Best & Tyler (2007)] is that adults perceive unfamiliar non-native phones in terms of articulatory similarities/dissimilarities to native phonemes and contrasts. The implied attunement to native speech emerges early: As infants begin to discern the articulatory organization of native speech, language-specific effects in non-native speech perception appear (≈6–10 months). Given that non-native phones, by definition, deviate phonetically from native ones, how can we characterize articulatory similarity in concrete, testable ways? The Articulatory Organ Hypothesis (AO) [Studdert-Kennedy & Goldstein (2003); Goldstein & Fowler (2003)] offers a possible approach, positing that infants decompose the oral-facial system into distinct articulatory organs (e.g., lips, tongue tip, tongue dorsum) and are sensitive to their actions in producing vocal tract constrictions. Thus, between-organ contrasts should be easily perceived/learned by infants and adults, whereas detection of within-organ contrasts must become attuned to the distribution of differing constriction locations/types by that organ in input speech. We discuss articulatory, attunement modeling, and perceptual evidence consistent with these predictions, and present a revised version of PAM that incorporates the AO Hypothesis and related principles of articulatory phonology [Browman & Goldstein (1991)]. [Work supported by NIH.]