In a series of important papers concerning the nature and evolution of folk biological categories, Cecil Brown (1977Brown ( , 1979aBrown ( . 1979bBrown ( , 1981aBrown ( . 1981bBrown ( . 1982 Brown and Witkowski 1982) has attempted to show that Berlin and Kay's (1969) main observations regarding the domain of color are also applicable to "trees," "birds," and other so-called life-forms. Brown claims that a small set of highly inclusive plant and animal terms are like basic color terms both because they are linguistic universals and because they co-occur in such nonrandom patterns that one must assume they evolve as languages change. As with color, Brown concludes that the number of life-forms named is related to a society's cultural evolutionary development and that the content of life-form taxa reflects psycholinguistic principles (cf. Hays, Margolis, Naroll, and Perkins 1972;Kay and McDaniel 1977). Despite the sweeping nature of these claims, their publication in numerous widely read journals, and the obvious implications of findings parallel to the color-term discoveries for ethnosemantic research, Brown's studies have received little published criticism (but see Riley 1 980).Yet none of Brown's hypotheses apply very well to the field languages we have studied most carefully: Columbia River Sahaptin (Hunn), Southern Philippine Sinama (Randall), and Mayan Tenejapa Tzeltal (Hunn). In attempting to explain why his hypotheses differ so strikingly from our own observations, we have considered and rejected the possibility that we happen to have studied exceptional languages. Rather, we think (1) that Brown's sources are not sufficiently reliable nor detailed to prove or disprove his hypotheses; (2) that Brown's concept of "life-form" is too loosely defined to allow the evaluation of his hypotheses; and (3) that Brown's universal categories are not broadly comparable lifeforms but instead represent a diverse set of concepts defined by function, habitat, morphology, or the lack of such attributes.Standard hypotheses regarding folk biological lifeform evolution are not confirmed b y data from Sinama, Sahaptin, and Tzeltal. This is probably because such hypotheses have been developed from highly inadequate data. Empirical determination of the focus, range, and defining features of highly inclusive categories shows that such categories are not universal, as has been claimed. Frequently, such categories class what is useful or useless, and therefore reflect a closer correspondence between biological categories and socioeconomic factors than current theory admits. [cognitive anthropology, ethnobiology, folk classification, language universals, category exemplars]