Marshall (1995) once argued that Wittgenstein's social constructivist view of mathematics is not "idealistic", "relativistic" or 'subjectivistic" but rather is "non-idealistic and objective". Wittgenstein is not idealistic because he attacks the prioritizing of mental states over linguistic accompaniments of those internal states. What he emphasizes is not intuition or mental process but the use of language . This, says Marshall , is an objective criterion, for although mathematics is "invented" rather than "discovered", the manner of its invention is in the form of discursive construction and in this sense it is independent of the individuals who use it as are the criteria of the truth and falsity of its propositions . It is thus non-foundational in the Russellian/Fregian senses. Rather, its objectivity is guaranteed by understanding mathematical objects within a formal language system. Truth in this sense depends on correct derivation in terms of the rule structure of the "language game" relative to a "form of life". Truth is thus "internal to a scheme". In this article, I compare Wittgenstein to Foucault with respect to the issues of idealism , scepticism and language to highlight some of the main issues which seem to me central to any serious consideration of the limits and possibilities of social constructivism. These relate to (1) the central differences between radical constructivism and social constructionism, and (2) the extent to which the problem of relativism is overcome by Wittgenstein and Foucault making comparisons between the two thinkers. In completing these tasks, I will also consider some of the contributions of Foucault to the constructionist debate.