This chapter is based on my response to a widely cited paper by Moore and Muller (Moore and Muller 1999). They argued that both the early sociology of knowledge which drew on ideas from Max Weber and the symbolic interactionists and informed the new sociology of education (Young 1971) and the later poststructuralist variants influenced by writers such as Foucault and Lyotard were both forms of ‘voice discourse’ which reduce knowledge to knowers, their standpoints and interests. Reading their paper was a formative experience for me as it challenged the basic assumptions of my earlier work. It forced me to ask what exactly I meant by the claim that knowledge is ‘social’. Does it, as Moore and Muller appeared to argue in that paper, inevitably lead to relativism and therefore have no implications for the decisions that are made about the selection of knowledge in curricula? The chapter begins to develop an alternative social theory that does not reduce knowledge to the practice of knowers. It represented my first attempt to articulate a social realist view of knowledge and to work out some of its implications. Later chapters of this book take this argument much further by drawing on the much neglected social theory of knowledge first proposed by the great French sociologist, Emile Durkheim, and later developed in the second half of the twentieth century by Basil Bernstein.