Arguing that the vagueness with which the relationship between social interaction and language is often treated in the literature makes it difficult to explicate and evaluate different views of that relationship, this paper poses four questions designed to differentiate positions and to provide a system for organizing data potentially relevant to evaluating them. The questions posed by the paper are as follows: Are the acquisition of social knowledge and linguistic knowledge simultaneous? Are there facilitating effects of one system on the other and are they unidirectional or bidirectional? How much is the relationship mediated by the internal properties of the child? and Is the relationship based on structural commonalities? As a way of examining the data bearing on the answers to these questions, the paper outlines and evaluates four different kinds of relationships between social interaction and syntax that are distinguishable on the basis of the different sorts of answers they provide to the questions. The paper concludes that the set of reasonable relationships adheres to the following constraints: (1) while some social knowledge is acquired earlier than linguistic knowledge, it is unlikely that much of the acquisition of one system occurs prior to the other; (2) it is unlikely that facilitation is bidirectional, if it occurs at all; and (3) there is at best only a partial structural commonality between the two domains, suggesting that internal properties of the child are central to a characterization of the relationship. (FL)