Abstract It is argued that as a scientist one does not justifiably employ analogies to ethnographic observations for the "interpretation" of archaeological data. Instead, analogies should be documented and used as the basis for offering a postulate as to the relationship between archaeological forms and their behavioral context in the past. Such a postulate should then serve as the foundation of a series of deductively drawn hypotheses which, on testing, can refute or tend to confirm the postulate offered. Analogy should serve to provoke new questions about order in the archaeological record and should serve to prompt more searching investigations rather than being viewed as a means for offering "interpretations" which then serve as the "data" for synthesis. This argument is made demonstratively through the presentation of formal data on a class of archaeological features, "smudge pits," and the documentation of their positive analogy with pits as facilities used in smoking hides.
Stored only when both papers are in the corpus — 310 citations exist upstream.