Interdependent but autonomous: Theorising phonological interaction in the Jamaica continuum/diglossic situation

There is no way in which the assumption of underlying English phonology for JamC, by way of simplification and deletion rules, could account for the intermediate and even variant JamE forms. The frequent areas of overgeneralization in every aspect of the segmental phonological system discussed here suggests that JamE, for many speakers, is based on a JamC lexical input. A proposal for deletion rules, working from either JamE or from 17^th century British English, would not produce the overgeneralizations discussed. Rules of elaboration, involving the lexical marking of some items to receive this elaboration, are a crucial part of the link between the phonological systems of JamC and JamE. It is the absence of lexical marking that produces the initial overgeneralizations.

Simultaneously, in the area of the prosodic system, we have seen what happens in the process of conversion from complex to simple. Since written English largely provided the model for the development of JamE, the fact that Standard British English and other metropolitan varieties of English were simple in this area had no impact on the conversion process. In fact, the greater complexity of JamE, relative to its external models, has gone unnoticed by speakers of JamE themselves. This fact underlines the largely written character of the model on which JamE has been built. It also supports the notion that much of the special lexical marking for conversion from JamC to JamE comes from cues given by the spellings of words in conventional English orthography.

The jury is out on whether a complex aspect of the phonology of a natively used variety can be suppressed in the acquisition of a non-native variety which is less complex. We would theorise that this would not occur unless strong social stigma were associated with this more complex set of features. Since, in the Jamaica situation, all the users of the potentially less complex system are also native users of the more complex one, the likelihood is that the greater complexity would go unnoticed, which is what we suggest happens with the complex JamC prosodic system in JamE. In addition, the distinction has some functional load. Thus, failure to carry it over into JamE would leave a communicative gap which would be noticed by speakers since they regard JamC and JamE as varieties of the same language.

The fact that, in pairs of related phonological variables, one variant would have a high frequency of occurrence in the variety to which it does not ideally belong, and that another would not, suggests that we are dealing with two language varieties which interact with each other, sharing on a variable basis some features of the other, but reserving other features for its own exclusive use. We have a pair of phonological systems that converge with each other but that are, in a systematic manner, nevertheless kept apart by their users.