I don't know where the word originates; but I believe it to have been used by Aristotle as I've seen it referred to in relation to him. As I understand it, not all arguments are logical syllogisms. An Aristotelian enthymeme, "an incomplete or not-quite-air-tight syllogism," provides a useful (and ancient) type of argument that may be put to effective use by social scientists, as economists, sociologists, et al often work in approximations of reality. An enthymemic argument might, for example, include the qualification that a premise is, say, 80 percent true. When the evidence is gathered to support the 80 percent, the argument succeeds; an argument is thus not destroyed by the production of a single counterexample.
One paper in the recent academic literature that mentions the word, and specifically develops the idea of an enthymemic style of argumentation is “Is There Life after Samuelson’s Economics? Changing the Textbooks” by Arjo Klamer, Deirdre McCloskey, and Stephen Ziliak Post-Autistic Economics Review May 2007, 42(18): 2-7. N2e 04:44, 25 March 2009 (UTC)