Consonant cluster reduction

The reduction of stop-final syllable-coda consonant clusters such as west to wes’, find to fin’, and act to ac’ is a well-known process affecting a wide variety of English dialects. Whereas all dialects of English reduce clusters preconsonantly, as in west side to wes’ side or cold cuts to col’ cuts, in prevocalic position consonant cluster reduction (CCR) is quite sensitive to ethnic and language background. Wolfram, Childs and Torbert (2000) maintain, for example, that extensive prevocalic reduction can usually be traced to language contact situations involving transfer from a source language not having syllable-coda clusters. It is also a wellknown feature of creolized varieties of English, including creole languages of the Caribbean (Holm 1988/89; Patrick 1996) and North America (e.g. Gullah), as well as ethnic varieties exhibiting such substrate influence. Both Holm (1980) and Schilling (1978, 1980) note extensive consonant cluster reduction as a characteristic of both black and white Bahamian English varieties.

The quantitative analysis of two outlying Bahamian communities in the Abaco region of The Bahamas, one exclusively Afro-Bahamian and one exclusively Anglo-Bahamian, suggests that there is an ethnolinguistic divide in the relative incidence of consonant cluster reduction. Afro-Bahamian communities tend to apply cluster reduction at much higher frequency levels than their Anglo-Bahamian cohorts. At the same time, Anglo residents in The Bahamas have higher levels of CCR than Anglo speakers in the US or in England. For example, Anglo-Bahamian speakers tend to reduce clusters more than vernacular-speaking white speakers in the Northern or Southern US, although their levels of reduction are not equal to those of their Afro-Bahamian cohorts (Wolfram and Schilling-Estes 1998). This pattern suggests that there has been some quantitatively based accommodation to the vernacular phonological norms of Black Bahamian speech by Anglo residents of The Bahamas.

As with other dialects of English where consonant cluster reduction applies, it can affect both monomorphemic (e.g. guest to gues’; mist to mis’) and bimorphemic clusters (guessed to gues’ and missed to mis’), with CCR favored in monomorphemic clusters. For basilectal Afro-Bahamian speech, however, this pattern is confounded by the incidence of grammatically based unmarked tense. That is, the lack of inflectional -ed suffixation may result from a grammatical difference in verb morphology as well as the phonological process of cluster reduction. The confluence of the grammatical process and the phonological process may thus have the effect of raising the overall incidence of past tense unmarking. It also makes it impossible to determine if a particular case of a past tense verb form (e.g. missed as miss’; guessed as guess’) results from the phonological or the grammatical process. This type of additive effect does not apply to Anglo-Bahamian speakers, who do not have grammatically based past tense unmarking and tend to have quite low levels of prevocalic CCR for bimorphemic clusters.