The paper being reviewed puts the vernacular in the perspective of the linguistic repertoire of a speech community. It is suggested that the repertoire as a single system should be seen on a societal or individual level rather than on a linguistic level such that various codes are selected by members of the community according to socially determined rules of appropriateness. Labov's work suggests that for purposes of comparing different communities, whether bilingual or monolingual, the term "vernacular" should be restricted to a technical usage, applying only to the first-learned spoken languages of a group of speakers. Research on native New Yorkers' speech, comparative studies of children's speech in Philadelphia and New York, and Benji Wald's own work on the phonology of Swahili are offered to back up the critique. This approach to the vernacular avoids the problem of variability in judgment as to what is the vernacular. A second critique centers on Kachru's discussion of code-mixing. Instead of asking when there would be a new code, as Kachru does, the critique would ask when do we know we have a code in the first place. In addition to Kachru's call for a typology of constraints found across codes, the critique would add an appeal for a typology of constraints to distinguish code-mixing or switching in different multilingual communities. (AMH)
