The sociolinguistic turn of the 60's has been remarkably successful: variability of language is no longer an issue open to debate. But studies of variation have by and large been restricted to the level of expression. This volume offers a critique of present day sociolinguistic, arguing that since meaning is critical to all contexts of life in society, ignoring it has led to a number of serious problems, foremost among them a mis-conception of the nature of sociolinguistics itself. By examining the possibility of systematic variation at the level of meaning, the volume makes a positive contribution to opening up a debate about the possibility of the neglected/misrepresented notion of semantic variation.Most chapters of the volume present an account of certain aspects of an empirical research which strongly supports the view that systematic variation in the choice of semantic features occurs across different social groups. Mothers and their young children less than four years old showed in their naturally occurring conversation a highly systematic, statistically significant, orientation to distinct styles of meaning, which correlated with their social positioning, and/or with the gender of the children. The comparison of kindergarten teachers' ways of meaning with those of mothers' provides empirical proof that teacher talk is an exaggerated version of middle class mothers' talk. The volume is relevant on the one hand to theoretical issues in linguistics and sociolinguistics and on the other, to any serious discourse about equitable education.