The aim of this article is to reflect on and explore questions of truth and objectivity in the sociology of educational knowledge. It begins by reviewing the problems raised by the social constructivist approaches to knowledge associated with the "new sociology of education" of the I970s. It suggests that they have significant parallels with the pragmatist ideas of James and Dewey that Durkheim analysed so perceptively in his lectures on pragmatism. The article then considers Basil Bernstein's development of Durkheim's ideas. We argue that despite his highly original conceptual advances Bernstein seems to accept, at least implicitly, that the natural sciences remain the only model for objective knowledge. This leads us to a discussion of Ernest Cassirer's idea of symbolic forms as a more adequate basis for the sociology of knowledge. In the conclusion, the article suggests how an approach to knowledge in educational studies that draws on Cassirer's idea of "symbolic objectivity" can come to terms with the tension between the concept of truth and a commitment to "being truthful" that was left unresolved, even unaddressed, by the "new" sociology of education of the I970s. (Contains 18 notes.)