Cedar Creek Natural History Area: Literature   Up   Home


Citation. Tilman, D. 1991. The schism between theory and ardent empiricism: a reply to Shipley and Peters. The American Naturalist 138:1283-1286.   [1176  LTER]

Introduction. In replying to Shipley and Peters (1991), I will focus more on their philosophical points than on a repetition of points already discussed (Shipley and Peters 1990; Tilman 1991). Shipley and Peters are members of a school of thought that has fostered a highly empirical approach to ecology. The detailed analyses of empirical data by members of this group have provided solid descriptions of many ecological patterns and thus contributed greatly to the discipline. I am concerned, however, that their papers (Shipley and Peters 1990, 1991) suggest the existence of a schism between theoreticians and empiricists. They accept as "true" the results of their own experiments, which, when analyzed by regression, explained only 4% of the variance in their data, which were only significant at the P< .05 level in one of two comparisons, which had questionable methods that may have biased their results (Poorter and Lambers 1991), and which contradict many published studies (e.g., review in Poorter 1989; Poorter and Remkes 1990). They asserted that this weak evidence totally disproved a theory, even though the suite of species they studied fell outside the stated bounds of that theory. Their willingness to accept weak empirical data but their demand that a theory be universally true suggest a lack of understanding of the role of theory. In proposing my theory, I said that the simplified version (used by Shipley and Peters) applied to plants that had much greater differences in morphology than in physiologically. The species chosen for study by Shipley and Peters were similar morphologically but likely had quite different physiologies (see Tilman 1990a). It is not at all surprising that their results did not fit the simplified version of the model.

Keywords. theory, empirical data, empiricism

For reprints or technical issues, please correspond with the author of the paper. For comments on the format or contents of the web site, please contact webmaster@cedarcreek.umn.edu