Introduction

I think Kingston's “binding principle” is an extremely important concept for phonology which, once it is properly defined, will turn out to have application to a wider range of phonological facts than those he reviewed and will, moreover, constitute a major challenge to current phonological theory. In this comment I will attempt to (a) suggest a definition of binding which could survive the negative evidence Kingston found for it, (b) exemplify the wider range of phenomena which manifest binding, and (c) indicate how these phenomena challenge modern phonological practice and theory. I will also offer a brief digression on the history of the use of distinctive features in phonology.

The binding principle

As I understand it, the “binding principle” recognizes that the vocal tract, though consisting of individual parts (glottis, tongue, velum, lips, etc.), is nevertheless a unified integrated “whole” which requires the orchestration of the parts to achieve certain ulterior goals. Intermediate goals include certain aerodynamic targets but the ultimate goals are acoustic–auditory events (cf. Jakobson, Fant, and Halle's [1951: 13] dictum, “we speak to be heard in order to be understood”). More specifically, the principle states that the vocal tract is subject to physical constraints that require temporal coordination between different, sometimes “distant” articulators. Kingston focuses on the case of stop bursts – a highly salient, highly informative acoustic event – which, to be produced correctly, require the cooperation of the glottis and the supraglottal articulators (including the velum).