Introduction

The framework that Keating has presented for describing articulatory movements has a number of desirable features. It incorporates the view that the positioning of articulatory structures to implement a particular feature or feature combination can occur within a certain range of values or “window,” and that the movements that occur when there is a feature change need only follow trajectories that do not violate these windows. Whether, as Keating suggests, there is no preference for different target positions within a window might be open to debate, since there is at least some evidence that certain acoustic characteristics of clear speech show greater strength or enhancement than those for conversational speech (Picheny, Durlach, & Braida 1986). That is, there appears to be some justification for the notion that the strength of the acoustic correlate of a feature may vary, and one might conclude, then, that some regions within a window (as defined by Keating) are to be preferred over others.

In Keating's framework, a target region is characterized not only by a range of values for an articulator but also, through the specification of horizontal parallel lines, by a time span within which the articulatory dimension must remain as it follows a trajectory defined by a feature change. The implication is that an articulatory structure should remain within a target region over a time interval in order to implement adequately the required feature or feature combination. Thus, for example, a narrow window would require an almost stationary position for an articulator over a substantial time span.