The Causal Structure of Natural Selection

If you’re interested by my contribution to the Cambridge Elements series, here is a brief introduction to the text. The Elements series, edited (in the philosophy of biology) by Michael Ruse and Grant Ramsey, is designed to be a set of what one might call “opinionated introductions” to an area of contemporary research.

In my case, I’ve started from a discussion – at the very least around forty years old, and, as I’ve actually argued elsewhere, as old as our discussions of statistical theorizing about natural selection – about the question of causation in natural selection. Natural selection is an interesting sort of theory. We have properties of individuals, of the kind that Darwin was so interested in: being stronger, being faster, and so on. Those individuals interact with one another, and form populations. Certain kinds of events that can happen to those individuals – certain ways in which they can systematically live, mate, die, and so on – lead to reproducible and explicable kinds of change at the population level, like the generation of adaptations.

So: how should we understand the general structure of all of that?

In the contemporary philosophy of biology literature, a whole host of questions have been picked up that roughly cluster around these issues, and unfortunately have tended to be discussed simultaneously. What are the correct definitions of concepts like natural selection and genetic drift to be drawn from biological practice? Are they processes, or population-level outcomes? If they’re processes, are they causal? How should we define fitness? Is it a property of individual organisms, or of traits? Can it play a causal role, or not? Where is the “causal action” within an evolving population? At the individual level, or the population level, or the genetic level (or all or none of the above)? How do evolutionary explanations work? What is the role of the subjective interests of observers in the creation of those explanations? How should we understand the relationship between individuals and populations, and the sort of “multi-level” system that they produce? What role do mathematical models play in evolutionary inferences?^[1]

As you can see, this is a pretty bewildering variety of issues. For better or worse, since the early 2000’s, they’ve tended to be grouped together into two major “positions,” which have come to be known as the “causalist” and “statisticalist” views. For the causalist, evolutionary factors like selection and drift are causal processes that act upon populations (and, for some causalists, have analogues that work on individuals, too). The causal action is therefore at both the level of individuals (mating, reproducing, and dying) and the level of populations (natural selection, genetic drift, migration, and so on). Evolutionary explanations are perhaps particularly complex, but fundamentally aren’t unlike other kinds of causal explanations in other special sciences.^[2]

For the statisticalists, on the other hand, natural selection and genetic drift are not causes; they are non-causal statistical abstractions away from the real causal action at the level of individual organisms. Fitness is nothing more than a growth rate in a certain kind of population, and evolutionary explanations are crucially constructed around the subjective decisions of individual researchers to abstract away from certain kinds of details about organisms.^[3]

I think it’s fair to say that this debate has somewhat stagnated in the last few years. Both causalists and statisticalists continue to publish (and I say this as someone who has published a number of papers broadly in the causalist vein), but the arguments have a persistent sense of not engaging with one another. There’s also, I think, a sense that both these positions are more complex and interesting than they seem when they’re simply described as being “causalist” or “statisticalist” – or, put differently, there’s many different ways that one might be a causalist or a statisticalist, and to describe these works as only coming from one tradition or the other seems to elide over interesting and philosophically rich differences of opinion.

I’ve tried in this work to find a new way to “un-stick” the debate, in two different ways. First, I’ve worked to extract a particular subset of these questions, surrounding causal structures – which I think are both more complicated than people often admit, and haven’t been very clearly discussed in isolation so far in the literature. And second, having extracted these questions, I think, gives us a way to find connections between this question of causal structure in the biological sciences and other discussions of causal structure in places like the philosophy of physics and the philosophy of mind. I make an attempt to draw some preliminary such connections near the end of the book.

If that’s the kind of thing that sounds interesting to you, watch this space! The book is quite short, and, if Cambridge University Press continues the same tradition that they have with other books in the series, it should be available as a free PDF download for the first few weeks after they finish typesetting it. Keep an eye on my various social media channels and I’ll distribute the link as soon as I have it.^[4]

And more than this, besides, which I don’t have the space to get into in a short introduction like this one! ↩︎
For paradigmatic causalist papers, see, for instance, Hodge, M. J. S. 1987. “Natural Selection as a Causal, Empirical, and Probabilistic Theory.” In The Probabilistic Revolution, Volume 2: Ideas in the Sciences, edited by Lorenz Krüger, Gerd Gigerenzer, and Mary S. Morgan, 233–70. Cambridge, MA: Bradford Books; Millstein, Roberta L. 2006. “Natural Selection as a Population-Level Causal Process.” British Journal for the Philosophy of Science 57 (4): 627–53. https://doi.org/10.1093/bjps/axl025; Pence, Charles H., and Grant Ramsey. 2013. “A New Foundation for the Propensity Interpretation of Fitness.” British Journal for the Philosophy of Science 64 (4): 851–81. https://doi.org/10.1093/bjps/axs037. ↩︎
For paradigmatic statisticalist papers, see, for instance, Walsh, Denis M., Tim Lewens, and André Ariew. 2002. “The Trials of Life: Natural Selection and Random Drift.” Philosophy of Science 69 (3): 429–46. https://doi.org/10.1086/342454; Matthen, Mohan, and André Ariew. 2002. “Two Ways of Thinking about Fitness and Natural Selection.” Journal of Philosophy 99 (2): 55–83; Walsh, Denis M., André Ariew, and Mohan Matthen. 2017. “Four Pillars of Statisticalism.” Philosophy, Theory, and Practice in Biology 9: 1. https://doi.org/10.3998/ptb.6959004.0009.001. ↩︎
I’d be remiss not to signal my many thanks to a whole host of people who have provided fantastic comments and revisions over the course of the process, especially Marshall Abrams (whose own forthcoming book on similar subjects, with a slant toward questions in the interpretation of probability, will be a must-read), and to detailed draft comments from Brian McLoone, Victor Luque, and an anonymous reviewer. ↩︎