Three componential analyses of American kin terms—Wallace and Atkins' (1960), Romney and D'Andrade‘s (1964), and Goodenough's (1965)—are a principal focus of an investigation of American kinship semantics and of an enquiry into the epistemological foundation of semantic analysis in general. The psychological, semantic, and structural implications of these analyses are explored in terms of the distinction between necessary and contingent facts. These analyses are determined contingent upon individual interpretation of genealogical facts and consequently relate to individual or “private” meaning. Relational analyses such as those of Burling (1970) and Wallace (1970) are necessarily real, relating to the ordinary use of American kin terms rather than to the objects designated by them and therefore are more in line with Wittgenstein's definition of meaning as linguistic use of words. In this regard, the latter analyses are “better” accounts of the meaning of terms. An additional analysis of American terminology is derived, focusing upon reciprocal usage of terms and their ordinary use in connection with bound forms such as great- and step-. This analysis corresponds to the structural principle related to the definition of “family” in American society as described by Schneider (1968) and others. This is in accordance with the hypothesis that necessarily real analyses complement structural reality because such solutions and structural principles are of the same logical order.