There is, therefore, some reason to think that teachers' stereotypes and predispositions may unfairly hinder children's school progress. The work of Rosenthal and Jacobson (1968) is relevant here. They claimed that, on the basis of manipulated information provided to teachers, the latter formed different expectations for what they thought were groups of bright and not-so-bright children. Rosenthal and Jacobson also claimed that these expectations were reflected in differential school progress made by the children. This work has been strenuously criticized (Dusek, 1975, provides a useful review) and the results have not always proved to be replicable. However, whatever the verdict on Such manipulated expectations, it is surely beyond doubt that, in the normal course of classroom life, teachers routinely hold expectations about their pupils. Such expectations are perfectly normal and, in most cases, are probably fairly accurate assessments. However, in instances in which there are differences between the social backgrounds of pupil and teacher, we would expect such judgements to be more fallible. Rist (1970) outlined here a self-fulfilling prophecy originating from teacher expectations of children. Teachers may feel that certain children (or groups of children) are less able than others; this may be communicated in subtle (and, as Trudgill, 1975, noted above, not so subtle) ways to the children (e.g. by non-verbal reactions; see Rosenfeld, 1978) who, sensitive to this,. respond in ways consistent with the expectation