On the phonetic basis of phonological change (original) (raw)

Sound changes are changes in speakers' phonetic abilities The distinction in N. American English between the low labial vowel [0] of caught, dawn, hawk and the low non-labial vowel [a] of cot, Don, hock is an old one, going back to early English aw versus short o. But in dialects of Canada and various parts of the United States of America, the [0] is pronounced [a], and the two sets of words are no longer distinguished. This sound change [0] > [a] was a historical event, but it remains a synchronic constraint that characterizes the phonetic capacities of the speakers of the innovating dialect: without additional learning, they cannot perceive or pronounce the [0] of other speakers, and even if they learn to perceive it, they often continue to pronounce it as [a] (at least in unguarded moments) or as some other substitute, e.g. [0] or [09] or [oQ]. It is significant that children learning the conservative dialect also typically at first cannot perceive and pronounce [0]. When they learn to perceive it, they still use [0] or [0] or [09] or [oQ] until they learn also to pronounce it. Thus speakers of the conservative dialect acquire a phonetic ability that speakers of the innovating dialect do not. But some adult speakers of dialects with [0] in fact never learn to perceive it or to pronounce it without substitution. They are innovators, though their innovation lies in not doing something their fellow speakers do do, namely learn to perceive and pronounce [0]. And some speakers, having learned to hear and say [0] as children, may move to an innovative dialect area and, under peer influence, lose the ability to even to hear [0] as distinct from [a]. One such person, whose childhood distinction of [0] and [0] was well attested in a linguist'S diary, later forgot