We agree with Woolard's argumentation and propose that WY constitutes further evidence for such a universal in language shift processes. When language choice depends on the ethnic group of the addressee, we are more likely to discover in-group identification with the L variety. The L variety is functionally intact and can fulfill all communicative needs. In moving away from an interlocutor-based language allocation, Western Yiddish subjected itself to the effects of structural interference from German, which led to WY obsolescence as the final outcome of dialectalization.
If redistribution of linguistic material over functions in a diglossic situation is interpreted as involving not just the physical or topical context, but also the interlocutor principle, the concept of diglossia with its principle of role compartmentalization remains useful in explaining and predicting patterns of language maintenance and shift. As noted by Woolard ( 1989: 366), in Fishman's initial formulation of his framework, "interlocutor was an integral component of the domain concept" and only later has his concept of domain been interpreted as physical or topical context. Maybe it is time to "rediscover" the interlocutor principle and to focus more on the individual speakers in linguistic theory in general, since after all a theory of language cannot ignore its primary agents. We hope to have convinced the reader that Western Yiddish is another case of language death which was characterized by the move from an interlocutor-bound language allocation to a context-bound language allocation and is thus interesting not only in its own right but also from a more general linguistic and theoretical viewpoint.