ABSTRACT The claim that social class differences in educational success are related to language and language use of a cognitively relevant kind is discussed and criticised. Various aspects of the relationship between rationality and language are examined and it is argued that the contention that deficiencies in rationality are related to language is an ambiguous one which involves different kinds of claims. When the claims of verbal deficit theorists have been clarified, it has been found that the weakest of these (...) claims have not been proven, while the strongest are empirically possible but extremely unlikely. (shrink)