Some observations on the phonology of Mandekan: Diachronic and synchronic regularity

An argument claiming that: (1) There is strong evidence from the comparative study of the dialects of Mandekan (the language including the dialects called Bambara, Maninka, Mandingo, Dyula, etc.) that the rules characterizing the lenition of initial consonants in the Southwestern Mande languages were at one time productive in the Northern Mande languages and most likely in Proto-Mande itself; and (2) Synchronic phonological solutions, which result in an inability to define synchronic analysis as optimal, are indeterminate. Given that a language or language group has a community of related rules, diachronic evidence that illustrates the similarity of those rules in all the languages is in support of what might be called "diachronic regularity." This kind of discussion can be called explanatory in that it accounts for why and how the language evolved to its present form. No attempt is made to contravert the claim that the child arrives at an optimally general (and thereby simple) grammar; however, such a conclusion does lead to the assumption that there is a unique maximally general synchronic analysis. Given discussions of Spanish phonology and work on Australian languages, it is evident that the notion of simplicity is more illusive than originally assumed. Generalizing a paradigm may in fact complicate a set of rules and, conversely, generalizing a rule may produce irregular paradigms. An attempt is made to demonstrate that there exist areas of indeterminacy unsupported by any internal facts which defy a unique resolution. J. Wehrmeister