a decade ago, Dan Simberloff bemoaned the fact that essentialism and determinism were still dominant, if not rampant, in ecology. He attributed their persistence to ecologists' physics envy and to their diffidence in the face of the “apparent sloppiness” of their discipline (Simberloff 1982, p. 83). If the arguments of the previous chapters are correct, then Simberloffs criticisms are well placed. Essentialism appears to be driving some conceptions of “stability” (chapters 2 and 3) and of “species” (chapter 5) that run counter to the theory of evolution. Indeed, because of their heuristic power, deterministic models have been used to undergird some hypotheses about proposed community structures (chapter 3; see Simberloff 1982, p. 85). Yet, as we have argued, such hypotheses often fly in the face of the apparent uniqueness and randomness of ecological situations. Part of the reason why scientists appear to have underestimated this uniqueness and randomness, we suggested (chapter 4), is that ecology is permeated by a misguided and naive positivism and by the belief that science can avoid value judgments (see, for example, Peters 1982; see also Fretwell 1975; Rosenzweig 1976; Mclntosh 1982, pp. 27ff.). Our revisionist account of ecological method, as applied to conservation problems, is premised on the need to accept the shortcomings of positivistic, hypotheticodeductive science, even as we attempt to make ecology more rigorous, empirical, and testable. We believe that the uniqueness and randomness of ecological phenomena, as well as the lack of conceptual and theoretical clarity in the discipline, often suit it to a “bottom-up” rather than a “top-down” account of scientific explanation (see section 5.3 of the previous chapter).