On schwa

Abstract

Everybody knows what schwa is or do they? This venerable Hebraic equivocation, with its later avatars like "neutral vowel", MUT'melvokaZ, etc. seems to be solidly established in our conceptual and transcriptional armories. Whether it should be is another matter. In its use as a transcriptional symbol, I suggest, it represents a somewhat unsavoury and dispensable collection of theoretical and empirical sloppinesses and ill-advised reifications. It also embodies a major category confusion. That is, [8J is the only "phonetic symbol" that in accepted usage has only "phonological" or functional reference, not (precise) phonetic content. As we will see, there is a good deal to be said against raJ as a symbol for unstressed vowels, though there is often at least a weak excuse for invoking it. But "stressed schwa", prominent in discussions of Afrikaans and English (among other languages) is probably just about inex ·cusable.