American Scientist 76:572-573 (1986) Notes on empirical ecology To the Editors: In their article "Competition, Scientific Method, and Null Models in Ecology" (Am. Sci. 74:155-162, Mar.-Apr. 1986), Edward F. Connor and Daniel Simberloff make their latest plea for rigorous application of a particular and powerful version of the scientific method. I have always had difficulty imparting this message to naive students, because they so readily adopt the rubrics of a philosophy without thinking about its substance, let alone about the substance of the biological questions that the philosophy is supposed to explore. Several years ago, J. Chester Farnsworth [Horn's "alter ego"], a local countertenor and satiric iconographer, wrote the following lament. I have found it to be an excellent point of departure for discussion of Connor and Simberloff's message and for constructive criticism of the literature of empirical ecology. Chorus: I've got the Kuhn & Popper knee-jerk philosophy of science blues. I've read my John Rader Platt, and strong inference is what I use. With alternative hypotheses, how can I lose? At the ".05" significance the data will choose. I've got the Kuhn & Popper knee-jerk philosophy of science blues. Verse 1: Now a "natural experiment" is a contradiction in terms. And a "structured observation" is a whole 'nother can of worms. Factorial design won't be found on Nature's shelf. If you want to do it right, just manipulate it yourself, And pray for homogeneous variance and linear interaction terms. Verse 2: If you publish random papers, with your students all working in teams, 5% of what you say will be statistical truth, or so it seems. (Vamp 4 measures.) That the "Origin of Specious, by Selection of Natural Means." Verse 3: Well, there ain't much tenure out there, And if you want to get yourself some, You can make a lot of mileage, with a couple of rules of thumb: "You've got to dump on someone famous if you want to get ahead, And the safest choice is always somebody who's dead"; And "Any hypothesis can be falsifiable, if it's sufficiently dumb!" Readers who want the music and the complete tablature of Farnsworth's mildly bizarre guitar accompaniment should send me a stamped, self-addressed envelope. Henry S. Horn Department of Biology [now EEB] Princeton University Princeton, NJ