In the first of two pieces discussing the evolution of co-operative behaviours, I lay out the background to the orthodox solution.
The question of why creatures co-operate to the extent that they risk their own lives presents itself as something of a dilemma when the evolution of life is approached from the gene’s eye view. Looking solely from the perspective of genes invites the assumption that there is a problem with any behaviours that act against the propagation of a particular gene variant – if a behaviour works against a gene being passed on, how could the genes that facilitate this behaviour have originated and why would they have persisted?
It’s important to appreciate that in the orthodox gene centred view, co-operation isn’t strictly problematic. If two animals co-operate for mutual benefit, so much the better for both their genes. The assumption that the inherent background of competition will block co-operation is misguided, and not just because the importance of competition in the history of life is frequently overstated. Taking into account only self-interest, it is often possible to gain additional benefit from co-operating: cleaner fish of various kinds (for instance, remora that attach themselves to sharks) gain nutrition from cleaning bacteria and algae from the skin of other fish, while their “clients” attain better health by having dead skin and infectious agents removed. Since everyone benefits, this kind of co-operation is unproblematic from the gene’s eye view.
The problem comes when facing situations whereby one animal sacrifices its chance to have offspring or risks its life for another. This looks (at first glance) to be ruled out by gene centred arguments, yet it happens all the time in nature. As examples of the former, consider social mammals such as meerkats or dwarf mongooses where within any pack usually only one pair breeds; or ants nests, where worker ants are sterile and only the queen has offspring. As examples of the latter, consider alarm calling in squirrels, whereby the loud noise puts the squirrel in question in danger of being caught by an attacking predator but helps guard other squirrels nearby; or the white tail of deer and rabbits, whose bobbing motion attracts attention and allows one animal to lead a predator away from the rest of its group.
Possible solutions to this problem percolated through evolutionary biologists working in the mid-twentieth century. J.B.S Haldane (pictured above) developed the basis of an approach, which was later formulized by W.D. Hamilton into what John Maynard Smith termed kin selection. The idea is simple enough: if you die (before you have children), your genes die with you. But your relatives share a proportion of your genes, so if you die saving your relatives, your genes can live on. A famous story regarding Haldane says that he was seized with the idea while in a pub, and proceeded to scribble calculations on the back of an old envelope before declaring “I am willing to die for four uncles or eight cousins!”.
More formally, a gene which influenced behaviour towards this kind of fraternal sacrifice could propagate itself in a population if the death of one animal carrying that gene helped the survival of other animals carrying the same gene. In what we can call Haldane’s jest, an uncle has a one in four chance of carrying the same ancestral gene variant as you, while a cousin has a one in eight, hence the numbers quoted. Hamilton’s mathematical version of kin selection had at its heart a simple mathematical inequality, known as Hamilton’s rule, whereby a gene encouraging self-sacrifice can propagate if the ‘benefits’ of this act outweigh the ‘costs’, with the degree of relatedness weighting the extent of the benefit. Hence, if you are carrying a gene variant that promotes self-sacrifice, that gene can become more widespread if you die saving the life of four uncles or eight cousins, any of whom could be carrying that gene (presuming the costs and benefits involved are equivalent).
By apparently solving an otherwise insoluble problem in evolutionary studies, kin selection became enormously popular among evolutionists, and the term inclusive fitness entered into the canon as a description of what natural selection optimizes. The idea here is that rather than counting solely a creature’s offspring when imagining fitness, relatives can also be counted too, when modified by a fractional value representing relatedness. Crudely, therefore, if a hypothetical animal is survived by two children, but also eight nieces and nephews, its inclusive fitness will be proportional to three (two plus eight-times-one-eighth), rather than two. Remember that ‘fitness’ here is only a metaphorical measure of reproductive success, and ‘inclusive fitness’ is similarly only metaphorical, but these abstractions can still produce hypothetical claims when they are found in equations such as Hamilton’s rule. Elliott Sober and David Sloan Wilson explain this point clearly:
Before inclusive fitness came along, it was natural to think about individual selection by imagining that individuals “try” to maximize their Darwinian fitness. Although “trying” can’t be taken literally, the as-if quality of this thought is often heuristically useful; we often can predict which traits will evolve by imagining rational agents who are trying to get what they want... Inclusive fitness seems like a natural generalization of this idea – individuals are “trying” to maximize the representation of their genes in future generations, where it is recognized that your genes can be found in your genetic relatives as well as in your own offspring. The idea then gets broadened further, by taking into account the fact that nonrelatives sometimes have copies of your genes (though here “your genes” means genes that are identical by type, not identical by descent); this means that helping nonrelatives can also be a way to get your genes represented in future generations.
Part of the appeal of the kin selection approach was that it enabled altruistic behaviours to be interpreted as a form of self-interest, because the animal making a sacrifice for its relatives maximizes its own inclusive fitness by helping “its” genes. David Sloan Wilson notes that inclusive fitness “made evolution seem just like economics, in which everything can be explained as a form of utility maximization at the individual level.” Indeed, Hamilton’s rule is only this – an optimality model. It is not really a model that explains how the behaviour in question might evolve at all, and indeed, how it might be practically applied “remains a bit of a mystery”.
Thinking in terms of kin selection has a subtle hidden cost: by envisioning any form of co-operation that evolves by natural selection as a form of genetic self-interest, the idea of ‘self-interest’ becomes “an all-encompassing category”. Philosophers are naturally suspicious of ways of thinking that achieve totality via the way they are defined – the satisfaction a theorist can feel at having devised an apparently universal theorem can mask a severely blinkered perspective. However, the idea that self-interest can serve as an ultimate explanation for behaviour (i.e. egoism) goes wildly beyond Popper’s milestone, and can only reasonably be considered metaphysics. It is for this reason that kin selection, while legitimate science in one sense, also becomes mythic, an imaginative story that guides thought towards specific conclusions, and away from alternative perspectives.
In the case of kin selection, the perspective that became brushed under the carpet was group selection. The evolutionists responsible for developing kin selection were aware that there was a possibility that selection might occur at the level of the group rather than the level of the individual, but the consensus – especially in the light of the formulation of inclusive fitness and kin selection – was that it was too weak a force to have any significance. John Maynard Smith, for instance, was willing to accede the possibility that group selection might occur, but felt that the necessary conditions for it were unlikely to come about in practice. By the late twentieth century, anyone hoping to make convince biologists that selection at the level of the group might be a factor in evolution faced a daunting uphill struggle.
Extracted from the draft manuscript of Myths of Evolution, due from Zero Books in 2012.
Next week: Group Selection Strikes Back