"explanatory pluralism" is the view that some events can be correctly explained in two distinct ways. (..) vs "explanatory realism" = the view that there is a single complete and correct explanation for every event (..) vs Kim's (1993) explanatory exclusion principle = two complete and independent explanations of a single event cannot coexist. (..) I identify two distinct (but compatible) styles of explanation in paleobiology. The first approach ("actual sequence explanation") traces out the particular forces that affect each species. The second approach treats the trend as "passive" or "random" diffusion away from a boundary in morphological space. I argue that while these strategies are distinct, some trends are correctly explained in both ways. Further, since neither strategy can be reduced or eliminated from paleobiology, we should accept that both strategies can provide correct explanations for a single trend. (..) defense of explanatory pluralism (..) four conditions are necessary and sufficient: 1. Two distinct explanatory strategies must offer explanations for a single event. Unless these strategies are truly distinct, any case of pluralism will be merely apparent. 2. Both strategies provide correct explanations for the same event. I do not intend to provide a full analysis of what it means to provide a "correct" explanation. Roughly, I mean that the explanation is "true" and complete. Showing that an explanation has heuristic advantages is not sufficient. 3. The explanatory strategies are compatible. To satisfy the realist, both explanations must be true. Thus, it must be possible for particular instances of both strategies to provide true explanations of a single event. Presumably, if the strategies are truly distinct, they will diverge in some cases as well. 4. Neither strategy is "eliminable." That is, neither explanatory strategy can be reduced or completely eliminated from our most complete theory of the domain. (..) McShea (1994) distinguishes two mechanisms that could cause an asymmetrical increase in variance. First, the trend could be "driven" by some force that consistently favors an increase in the trait under consideration. The "driving" force is typically natural selection, though other forces could (in principle) drive a trend. (..) The second approach views cladogenetic trends as "passive diffusion" of species away from a boundary. (..) One reason passive diffusion models (PDMs) are philosophically interesting is that they hold out the possibility of a new interpretation of "progress" in the history of life. Trends toward increased complexity have sometimes been cited as evidence of progress. But McShea (1991, 1993) has argued that complexity increases are not always driven. (..) Increases and decreases occur and both can be locally advantageous. (..) Although the well-known case of horse [size] evolution turns out to be a driven trend, other significant trends appear to be passive. (..) the best available evidence suggests that passive diffusion away from a boundary is a relatively rare (but not insignificant) pattern in the history of life. (..) Two Strategies for Explaining Passive Diffusion Away from a Boundary: (..) The first strategy is to explain the trend by appealing to a passive diffusion model (PDM). That is, we explain the trend by saying that whenever the founding species of a new clade is close to a boundary and the direction of speciation is random, a passive trend is likely. (..) The second strategy is the "actual sequence" explanation. providing a detailed account of the evolutionary forces affecting each species in the clade. (..) [ in summary: ] Two Kinds of Cladogenetic Trends A. Driven: Trend "driven" by some consistent force (e.g., natural selection) that affects nearly all species; direction of evolutionary change is biased by the driving force. B. Passive: Trend arises through "diffusion" away from a boundary in morphological space; direction of evolutionary change is "random." Passive bounded trends can be explained in two distinct ways: 1. Actual sequence explanation. Traces the specific forces affecting each species. 2. Passive diffusion model. Identifies the "structuring cause" of passive trends: random diffusion away from a boundary. (..) In spite of the fundamental differences between passive diffusion and actual sequence accounts, they are compatible: both types can correctly describe the same trend. (..) The pluralism I defend is intended to be a constrained form of pluralism that is rooted in a naturalistic approach to the philosophy of science. (..) We are forced to accept both explanations because neither can be eliminated. In the contest between a priori philosophical arguments for explanatory realism and a pluralism based on close analysis of actual science, I favor the latter. |