Are all evolutionary phenomena products, in some significant sense, of natural selection? This is a question of considerable importance in contemporary biological theory. I have shown that the answer to this question is a "yes" with several reservations. Yes, in all likelihood, any characteristic of any living creature is likely to have been influenced by natural selection. This does not mean that selection determines all characters or (what comes to the same thing) that selection can explain all evolved phenomena. Some contemporary evolutionists have noted that selection cannot explain every biological fact (e.g., Gould and Lewontin 1979); other evolutionists have appeared to believe that selection will explain all (Dawkins 1976). Both sides are partially right and partially wrong. Natural selection probably does hold universally for all living things, but this in no way means that selection explains all of biology (or sociology and psychology, for that matter). This pro- and antiselection debate has been marred by an implicit confusion between the applicability of selection, which is universal and exceptionless for living things, and its explanatory scope, which is circumscribed. One group of contemporary evolutionists (e.g., Dawkins 1976; Wilson 1975) seems to assume that any and all characteristics of organisms can be explained with a selectionist account. If this means that a selectionist account of any biological phenomenon is likely to be a sufficient explanation, it is surely wrong, for that is like saying that gravity alone will explain the trajectory of a feather falling from a bird in flight. To be sure, gravity (like natural selection) should not be overlooked, but the conditions under which the law is operating (e.g., the point where the feather started to fall) or other relevant laws (e.g., those of aerodynamics) must be taken into account for a satisfactory explanation to emerge. To use a more biologically relevant example, those who argue that organisms must maximize their inclusive fitness (Hamilton 1964) are assuming that feathers will fall as fast as rocks. Our statements of laws, whether in physics or biology, are inevitably idealizations, but we cannot allow these idealizations to obscure the empirical conditions of the situations we are trying to explain. Other contemporary evolutionists have concluded that the explanatory limitations on natural selection imply that it is not a factor of universal importance in the evolutionary process (Gould and Lewontin 1979; Lewontin 1978; Platnick and Rosen, in press). This is like arguing that, because the law of gravity cannot explain how a rocket flies away from earth, gravity is probably not a factor operating on rockets. In all probability, if there were no longer any selection, then mutation would produce evolutionary change-but without selection evolution would be degeneration toward higher entropy, if the second law of thermodynamics is correct. Note that selection does not violate the laws of thermodynamics, or any other physical laws, but it does counteract them. Presumably psychological and sociological laws exist which will counter the laws of biology and physics, without violating them. Evolutionists have yet to determine the nature and extent of the role played by natural selection in the evolutionary process. One large stumbling block has been an inability to treat natural selection as what it is, a law of nature. Once the lawfulness of selection is appreciated, then progress will be made in relating selection to evolution.