Paradigms are fostered by (possibly) intersubjective social processes. This chapter begins by considering the role of ostensive definition in introducing a paradigm. An ostensive definition, however, presupposes a general category that the defined paradigmatic sample belongs to. As the accounts of Goodman, Schlick, and most importantly Wittgenstein show, this category can never be fully determinate. Next, utilizing Derrida's concept of arche-trace, I argue that the essence of a paradigm and the reason for its institution disappears. This means that the reason for the emergence of a certain paradigm, together with the practice of its application, cannot be accounted for by that practice. Hence, the originary paradigm must be retrospectively constructed as an ideal object of comparison. This frees us from any commitment to empirical evidence in support of the existence of a present or past paradigm.