Previous chapters of this book have argued that linguistic diversity is more than a fact of behavior. Linguistic diversity serves as a communicative resource in everyday life in that conversationalists rely on their knowledge and their stereotypes about variant ways of speaking to categorize events, infer intent and derive expectations about what is likely to ensue. All this information is crucial to the maintenance of conversational involvement and to the success of persuasive strategies. By posing the issue in this way, one can avoid the dilemma inherent in traditional approaches to sociolinguistics, where social phenomena are seen as generalizations about groups previously isolated by nonlinguistic criteria such as residence, class, occupation, ethnicity and the like, and are then used to explain individual behavior. We hope to be able to find a way of dealing with what are ordinarily called sociolinguistic phenomena which builds on empirical evidence of conversational cooperation and does not rely on a priori identification of social categories, by extending the traditional linguistic methods of in-depth and recursive hypothesis testing with key informants to the analysis of the interactive processes by which participants negotiate interpretations.

Initially we approach the problem of the symbolic significance of linguistic variables by discovering how they contribute to the interpretation of what is being done in the communicative exchange. The hypothesis is that any utterance can be understood in numerous ways, and that people make decisions about how to interpret a given utterance based on their definition of what is happening at the time of interaction. In other words, they define the interaction in terms of a frame or schema which is identifiable and familiar (Goffman 1974).