The Philosophy of Evolutionary Theory

Many of us are familiar with the phenomenon of a box set. Successful musical acts collect their greatest hits into a single package that often includes previously unreleased or hard to find tracks. The box set serves as a nice analogy for Elliott Sober’s latest book on philosophical issues in evolutionary biology. Sober has been blazing a philosophical trail for decades through penetrating analyses of fitness, natural selection, species, systematics, and more. It is not an exaggeration to say that his work set the agenda in many ways for philosophy of biology. In this latest installment, we get the best of Sober’s work on these topics, with classic distinctions like “selection of” versus “selection for” (including the “selection toy” that celebrated its fortieth birthday last year), alongside material of more recent vintage, such as explorations of the likelihood approach to inferences about common ancestry. After summarizing the book’s well-executed themes, I offer a critical perspective, both in terms of methodology and themes, to situate its contributions.In evolutionary theory, it sometimes seems like all roads lead from Darwin. Sober adopts this conceit explicitly by introducing concepts from Darwin’s theory of evolution and then segueing to their more modern instantiations. His first chapter provides a short and helpful tour of these basic concepts. Subsequent chapters dive more deeply into specific topics: fitness and natural selection, units of selection, common ancestry, drift, mutation, taxa and genealogy (including a discussion of race), adaptationism, and “big-picture questions.” This last chapter touches on issues such as whether there are evolutionary laws, the status of determinism and indeterminism, and the contingency of the evolutionary process. Along the way, Sober introduces the reader to important terminology and key distinctions, giving simple but illuminating examples with line drawings and tables to unpack the ideas. The writing is clear and direct throughout. Sprinkled in the footnotes are questions for the reader to ponder on their own: “How can a bird that puts its eggs in two nests be reducing variance, as compared to a bird that puts all its eggs in one, since the first bird has more nests than the second?” (26). Sober graciously highlights places where others disagree with him (often in footnotes), but score settling is not a part of this book’s plan. I noticed a few places where others might demur on how their position was interpreted, but that is to be expected.A core contribution of this book—and of Sober’s scholarship more generally—is an explicit treatment of how general issues in philosophy of science manifest in evolutionary biology. These include instrumentalism, reductionism, underdetermination, and interpretations of probability. A good example is found in chapter 3 (“Units of Selection”), where Sober introduces Simpson’s paradox in the context of building an argument for why altruism can evolve because of group structure. This structure is invisible when viewing the fitness of the associated individuals in the entire metapopulation (i.e., aggregated together). Others include his analyses of conventionalism and realism with respect to genic selection and the difficulties of objective interpretations of probability in evolutionary models.Sober handles the complex models of population biology and systematics with grace and aplomb. He is especially perspicuous in discussing how the law of likelihood is applicable in the context of comparative hypothesis evaluation. However, he often undertakes these discussions without considering how these models are used in research practice. In the preface, Sober says, “My focus is on how these concepts ought to be understood; my project is normative, not purely descriptive” (xv). Still, one might be concerned that Sober’s analysis is not descriptive enough, relying largely on rational reconstructions abstracted from the wilds of scientific reasoning. His motivating questions take the form “What is X?” (e.g., What is fitness? What is a species?) but do not grapple with how evolutionary biologists conceptualize fitness or species in the practice of inquiry. He indicates that it is his “goal is to sharpen concepts so that they better serve the goals of theorizing” (xvi). What isn’t considered—but actual practices would suggest—is that concepts serve more than theorizing goals. It is an open question whether concepts sharpened for theorizing are also automatically sharpened for experimentation or data collection. One might also worry that there are different theorizing goals that require their own conceptual edges.From the outside, it might look like Sober is getting at most if not all the major conceptual issues at stake. However, there is a sense in which the book is quite narrow in what it covers. An interesting hint of this is in Sober’s acknowledgments: “I started work on the book you see before you with the idea of revising my book Philosophy of Biology” (xviii). What changed in recent decades for “biology” to become “evolutionary theory” in the title? In short, a lot. However, we do not need to document the tremendous growth of different strands of analysis in philosophy of biology beyond evolution (e.g., ecology or cell biology and development) to indicate the narrowness of Sober’s treatment, which is often reflected in citation practices (e.g., little mention of the now diverse literature on natural kinds and evolution). Even within evolutionary theory, much territory is not covered, perhaps most noticeably the question: What is evolutionary theory? This has been tackled by many other philosophers in part because the structure of evolutionary theory is not obvious and there is not a consensus on the individuation of its content.Regarding the content of evolutionary theory, consider the following question: What is a biological trait? Given that the models discussed throughout the book typically involve using traits as evidence for common ancestry or understanding how traits evolve under different conditions, you might think evolutionary theory should have an account of what counts as a trait. Sober talks as if traits are already known objects in the ontology of evolutionary theory. Although an ontology of traits may not be perceived as the responsibility of evolutionary theory, many have thought it was a task for evolutionary biology. Almost fifty years ago, in the context of one of Sober’s central topics (adaptationism), evolutionary biologists Stephen Jay Gould and Richard Lewontin worried about this openly: “We must omit an extended discussion of the vital issue: ‘what is a trait?’ Some evolutionists may regard this as a trivial, or merely a semantic problem. It is not” (Gould and Lewontin 1978: 585). More recently, Lewontin (2001: xvii) claimed that “No issue is of greater importance in the study of biology.” If evolutionary theory is foundational for all of biology (as many philosophers hold), it might seem odd to say that what counts as a trait doesn’t fall under its aegis. Since Sober ignores large regions of evolutionary biology, including biogeography, coevolution, evo-devo, quantitative genetics, and speciation, it would be helpful to have an account of how we know what is or is not part of the content of evolutionary theory, especially as the book advertises comprehensiveness (The Philosophy of Evolutionary Theory).In addition to these overlooked issues, another concern about Sober’s presentation is its occasional use of anachronistic labeling. Although Sober is not doing history of science, his rational reconstructions are not as benign as he envisions. He says in the preface, I also do not blush when I supply Darwin’s various theses with clarification and arguments that he never considered. When I do so ... I am evaluating the propositions that Darwin endorsed; these propositions are separable from the thought processes that let him embrace those propositions. This is not anachronistic. What would be anachronistic is attributing those clarifications and arguments to the man himself. (xvi)But then Sober goes on to say in chapter 6, “I quoted the analogy that Darwin drew between beneficial mutations and the stones that fall from a cliff... Darwin’s thesis is that mutations also have their causes, but they do not occur because they would be useful to the organism in which they occur” (141). How could Darwin endorse or even articulate a proposition about mutation when he did not have the concept (Sober refers to Darwin’s view or idea about mutation throughout)? The original analogy was between beneficial variations and stones. To substitute “mutation” for “variation” (silently) in these claims changes the meaning and theoretical implications of the propositions in view. The claim that mutation does not occur because it is useful to immediate organismal needs is not identical to the claim that variation does not occur because it is useful to immediate organismal needs. For example, organisms exhibit adaptive plasticity, such as not growing as large under conditions of food scarcity, and this variation manifests in response to immediate needs (though not due to mutation). It is especially problematic to ascribe a concept of mutation to other nineteenth-century thinkers with whom Darwin was in conversation (i.e., to the community), such as Asa Gray. In fact, later in life, Gray offered a similar viewpoint to Darwin on variation (not mutation): “Incipient variations are wholly vague and irrespective of ends—are as likely to occur in the direction of unfitness as of eventual fitness to the environment” (Gray 1883: 78). Claiming that “Darwin’s idea remains the received view in modern biology” (142) runs roughshod over important differences between these two concepts, some of which are relevant to active controversies in evolutionary biology, such as the compatibility of random genetic mutation with nonrandom phenotypic variation.One final concern is that Sober has not engaged with recent work on the significance of polysemy in the terms used to represent a concept (“gene” or “species”). In chapter 7, a thought experiment about whether a particular species concept is suitable for phylogenetics leads Sober to conclude, “Maybe we need to get used to the fact that the word ‘species’ is ambiguous. The term in evolutionary biology would then be like the term ‘bank’ in ordinary English” (182). However, philosophers like Philip Haueis and Rose Novick (among others) have given positive arguments for concepts having patchwork architectures that yield an account of the interacting relationships between multiple meanings associated with a concept, thereby moving beyond a bare acknowledgment of polysemy and toward a better understanding of how and why scientific concepts behave the way they do in the context of inquiry. Multiple meanings for scientific concepts are the beginning rather than end of philosophical investigation.Criticisms aside, you should buy the box set. Sober is a great philosophical musician, and there is good reason to be familiar with his melodies and virtuoso performance. At the same time, I would encourage philosophers to expand their range and sample from a wider diversity of styles and genres available in the philosophy of (evolutionary) biology. Altogether, they are a testimony to one of the most exciting fields in contemporary philosophy, one that owes a lot to Sober’s pioneering efforts. Keep an eye out for a live concert playing near you soon

Analogy · Darwin (ADL) · Evolutionary psychology · Evolutionary theory · Exaggeration · Natural (archaeology) · Phenomenon · Set (abstract data type) · Philosophy and History of Science

• The spandrels of San Marco and the Panglossian paradigm

  Open Access•Stephen J Gould, Richard C Lewontin•Proceedings of the Royal Society…•1979

  Add to list