Confusing the Tool with the Object of Study

There is an interesting controversy in biology over the concepts of eusociality and inclusive fitness.

This blog posting by Carl Zimmer on the Discover magazine site sums up the state of the controversy and identifies the key paper that is driving the current intellectual battles. Zimmer sides with the mathematical modelers. The other side is led by E. O. Wilson and is represented by this paper "The Evolution of Eusociality" in the Nature journal.

But I worry that the argument of inclusive fitness is an example of science mistaking the tool for the object of the science. Sure, if you make the assumptions of inclusive fitness you "explain" a large number of eusocial species. But I fear that models, as tools to tweeze out principles and help understand complex phenomena, get reified into "solutions" when in fact they are idealizations and can't handle all cases.

Carl Zimmer is dismissive of the E. O. Wilson paper, but it appears to me that Wilson has a valid criticism of the modelers. His point is the eusociality, while being extremely successful as a strategy, is rare in nature because it requires some unlikely series of steps and especially because it requires multilevel selection (a viewpoint that is anathema to the inclusive fitness crowd and many geneticists). But nature is complex and the drive to find simple models boxes you in conceptually. It is as if physics only allowed ideal gas laws and ignored any explanatory mechanisms that went below the level of an idealized billiard ball (i.e. models that attempt to describe real gas behaviour).

In a multicellular organism you have "eusociality" at a level of cells. Clearly you have reproductive division of labour, overlapping generations, and cooperative care of the young. And in this case, this is a degenerate and simplistic case of "inclusive fitness" because the genes in all cells are exactly the same. The odd cells that don't fit this pattern, e.g. blood cells, in fact do have the same genes until late in their "maturation" where they discard the nucleus with the DNA and go into a robo-mode. So in fact they do fit exactly the assumptions of inclusive fitness. But what about symbionts? How do lichen flourish? How does this cooperation arise since the originating algae/cyanobacterium and the fungus share no DNA? It does despite no fitting the mathematical model of inclusive fitness.

I like Wilson's attempt to understand the rise of eusociality in five phases:

1. Formation of groups in a freely mixing population.


2. Historical accidents that lead to the accumulation of traits that make the change to eusociality more likely (i.e. preadaptions).


3. The rise of eusocial alleles.


4. Natural selection operating on emergent traits among the cooperating organisms in a eusocial group.


5. Natural selection operating on competing eusocial groups with their variant emerging eusocial traits.

That explanation has the richness to ring true to me. It is a much more credible account than the simplistic inclusive fitness model which is a simple calculation that can't explain how eusociality arises and changes over time.

I enjoy Carl Zimmer, but I think he has made a mistake in picking a "winner" in this controversy. Sure the inclusive fitness guys may have the "numbers" to be persuasive. But science isn't a political contest. Ultimately it is effectiveness of a theory in explaining the real world. It may take some time, but I expect that Wilson and his crowd will win this argument, not the simple modelers of the inclusive fitness crowd. They have a very useful tool for analysis, but (in my humble opinion) they've made the mistake of confusing the tool with their object of study.