ABSTRACT
Sociologists have a difficult time keeping control of their concepts. Typically, an analyst devises a new term and applies it to some striking case. Often, because that case’s striking qualities seem self-evident, the analyst is able to skimp on the concept’s definition; after all, the example makes what is meant by the term self-evident. If the concept catches on, others adopt the term and extend it to additional cases. Quite often, these cases differ from that initial, striking case, so that the new applications expand the concept’s domain, via arguments that the new cases resemble—are analogous to, the equivalent of, or fundamentally like—that original case. Because the original definition lacked precision, it is hard to criticize these new, broader conceptions of the term as illegitimate. As the concept is applied to a wider range of phenomena, its domain stretches, and it becomes a more familiar term, even as its analytic usefulness is thereby diminished. For instance, once nearly every family can be labeled dysfunctional, then the term dysfunctional family becomes little more than another term for family, and whatever power the concept had to identify particular sorts of family interactions is reduced. Not infrequently, as a new concept diffuses, it crosses disciplinary boundaries, which only increases the likelihood that its meaning will morph in unexpected ways as a broader range of scholars apply the term. For instance, as the term social construction spread from sociology into, not just other social sciences, but the humanities, the natural sciences, and even into popular discourse, the various people who appropriated the term began using it to mean very different things (Hacking 1999).
