On Certain Evidence for the Feature [Grave]

Supposed evidence for acoustic assimilations attributable to the feature [grave], presented in Hyman 1973 and Vago 1976, is shown to be evidence merely of deficient modeling of articulation in the SPE articulatory-feature system-rather than, as Hyman and Vago assume, proof for the acoustic basis of the processes they cite. Straightforward articulatory explanations are available for all the evidence adduced, but they are not expressible in the SPE framework. The necessary refinements of such a system of articulatory description are shown to require more than adjustments of the feature inventory.*