In this paper I analogize'language'to'population'in population genetics, hoping to account more adequately for some aspects of language restructuring (see next paragraph!) in contact situations, especially those associated with the varieties called'creoles'. 1 I focus on those lexified by European languages, particularly those spoken in the New World and the Indian Ocean, which the literature has presented as (proto) typical. However, nothing precludes the main thesis of this paper (stated two paragraphs below!) from applying to other con tact-induced language varieties, especially those which have been lexified by non-European languages but have also been called pidgins or creoles, disputably (Mufwene, in press). What matters first in all such cases is the contact origin of the varieties. The specific social histories of their developments ac count for variation in the outcomes of the restructuring of the same lexifier (Le Page & Tabouret-Keller 1985), as I argue below, invoking ethnographic ecol ogy.

I use the term vernacular in its original meaning as the language variety of the home, the basic form used for day-to-day communication. As for the term restructuring, I use it here in the sense of" system reorganization", which makes a creole different from its lexifier. The latter was primarily the colonial variety which was spoken by the European colonists and was itself developing from the contact of diverse metropolitan dialects. Consistent with the position proposed below on creole genesis, this reorganization often consists in modi-