Features, Classes and Systems
In examining the voiced stops and approximants in Spanish, we saw (see p. 24) that the distinction between them is allophonic, and accordingly, we wrote three rules which expressed the contexts in which the allophones of the bilabial, dental and velar phonemes occurred: Our claim was that a native speaker of Spanish perceives [b] and as instances of the same thing, and likewise and, and [g] and. We said in Chapter 1 that the rules we write should express the generalisations the speaker has unconsciously acquired. Expressing our claim in terms of three rules therefore amounts to claiming that the speaker has acquired three different generalisations concerning, respectively, the bilabials, the dentals and the velars. Do you agree that there is something unsatisfactory about this? We can make a single, more general, statement about these stops and approximants in Spanish: in this language, voiced stops are realised as approximants intervocalically, regardless of their place of articulation.
Unable to display preview. Download preview PDF.
- It is probably unwise for most students to begin examining alternative theories of distinctive features before getting the hang of the one presented here. Hyman (1975) provides a good overview of three such theories, none of which corresponds exactly to the one adopted here, which owes much to that given in Lass (1984a), Chapter 5. Hyman (1975) covers acoustically defined features, which we have avoided discussing here. Both Hyman and Lass deal with segment types which we have not covered in this book.Google Scholar