The evolutionary community is as active as an alarmed beehive over the critique of inclusive fitness theory recently published in the journal Nature by Martin Nowak, Corina E. Tarnita, and E.O. Wilson. I do not agree with them in every respect but I’m glad that they have aroused the evolutionary community from its stupor. The general public and majority of evolutionary biologists have a pre-1975 understanding that hasn’t even kept pace with modern inclusive fitness theory, not to speak of the debates that will be taking place among the cognoscenti. This is an opportunity for everyone to take stock of the core issues at stake.
It is important to realize that numerous issues are at stake that must be examined one by one. It doesn’t help that Richard Dawkins continues to issue boneheaded statements about group selection, as I recount in my previous post. Inclusive fitness theorists should be joining me in pointing out the errors of these statements, just as I intend to join them in pointing out some errors in the Nowak et al. critique.
In this post I want to focus on a statement that Nowak et al. make in the caption to figure 3 that “inclusive fitness theory…is an alternative accounting method, but one that works only in a very limited domain.”
What does it mean for a theory to function as an accounting method? Consider the standard definition of genetic evolution, which is a change in the gene frequency of a population for any reason, whether selection, drift, linkage disequilibrium, and so on. It is important for the definition to include any kind of genetic change, just as it is important for a financial accounting method to include all monetary inputs and outputs. It is equally important for the accounting method to provide meaningful categories of change such as selection, drift, and linkage disequilibrium, whose relative importance must be determined on a case-by-case basis. After a lot of hard empirical research, we can say that guppy spots and finch beaks evolve largely by natural selection, that peacock tails evolve largely by sexual selection, that some base pair substitutions are selectively neutral and evolve by drift, that haplotypes reflect linkage disequilibrium, and so on.
It is important for a theoretical framework to function as an accounting method, but we haven’t made much progress merely by pointing out the fact. If I boast to you that my bank account keeps track of all deposits and withdrawals, your appropriate response would be “of course it does–if not you should switch banks”. Similarly, if I boast that my theoretical framework for studying evolution keeps track of everything that evolves, your appropriate response would be “of course it does–if not you should change your theoretical framework.”
Evolutionary theory includes a number of different frameworks such as multilevel selection theory, inclusive fitness theory, and what Dawkins calls gene selection theory, that all claim to function as good accounting methods. This is a bit like different banks that all keep track of deposits and withdrawals but differ in how they are categorized. One might categorize them according to dates, another according to their tax-exempt status, and so on. There might not be a single best accounting method, providing a justification for several to coexist. If so, then it becomes necessary to translate between accounting methods. The statement “this deposit was made on July 16, therefore it is not tax-exempt” is obviously stupid, but equivalent statements are made all the time by evolutionary biologists who don’t know how to translate among their various different accounting methods.
Against this background, I can begin to articulate my agreements and disagreements with Nowak et al. One of their claims is that the accounting method known as inclusive fitness theory has led to no useful insights whatsoever. That claim is too harsh. I’m happy to credit inclusive fitness theory with useful insights.
Nowak et al. also seem to contrast inclusive fitness theory with what they call “standard natural selection theory”, as if the point of inclusive fitness theory is to add some additional factor. This is poor phrasing on their part. What they mean to say, if I understand them correctly, is that inclusive fitness theory attempts to function as an accounting method. To the extent that it succeeds, every statement made within the rubric of inclusive fitness theory can be translated into the rubric of explicit models of natural selection in multi-group populations.
The next question is how well inclusive fitness theory functions as an accounting method. Here, Nowak et al. score some valid points about the simplifying assumptions made by inclusive fitness theory that limits its generality, or “works only in a very limited domain”, as they put it. Simplifying assumptions are a two-edged sword; all theories require them but they can be terribly misleading when a theory is applied to situations where the assumptions don’t hold. Few people who use inclusive fitness theory are aware of the long list of simplifying assumptions that result in Hamilton’s deceptively simple rule. Nowak et al. are performing a service to the evolutionary community by challenging the assumptions and forcing them to be openly discussed.
All this talk about alternative accounting methods ultimately becomes boring, like arguing over whose bank is better. The purpose of all accounting methods is to make sense of the evolutionary process, leading to the evaluation of factual claims such as “genealogical relatedness is required for the evolution of altruism” or “altruism is locally disadvantageous and requires a positive selective differential at a larger scale to evolve in the total population.” It is important to remember that there is such a thing as an empirically validated fact that all accounting methods must agree upon to remain correct. In my next installment, I’ll consider some of the factual claims associated with inclusive fitness theory, some that have survived the test of time and others that have turned out to be just plain wrong.