Tuesday, May 27, 2008

Making Rudyard Kipling Proud

We wish to question a deeply engrained habit of thinking among students of evolution. We call it the adaptationist programme, or the Panglossian paradigm.
S.J. Gould & R.C. Lewontin (1979) p. 584
A typical just-so story has two components. First, it postulates the existence of an allele "for" some trait in the absence of evidence that the gene(s) actually exist (or even that such genes are possible). Second, it postulates that the allele "for" the trait was selected in the past so that now it has become fixed in the population. The attractiveness of most just-so stories lies in the creation of an elaborate, but plausible, adaptive advantage for the postulated allele.

The field of evolutionary psychology seems to have been largely taken over by those who can create the most elaborate just-so stories to "explain" modern society. For example, the avoidance of incest in most (but not all) societies is due to fixation of an anti-incest gene in our ancestors [Another Boring Just-so Story]. As with most just-so stories, there is no evidence for the existence of multiple alleles of a gene where one allele confers incest avoidance while the other allele confers acceptance of incestuous relationships. (The problem becomes even more difficult if it's a trait due to multiple alleles at different loci.)

There's a trendy extension of just-so storytelling that looks superficially like evidence. It's the creation of a computer program to simulate one's just-so story. Naturally, these programs always work as expected since that's the nature of a just-so story. You have a postulated beneficial allele with a postulated selective advantage and, presto!, the allele becomes fixed in your simulated population. It doesn't prove a thing. If your program doesn't work as expected, then all you have to do is fiddle with the selective advantage (s) until it does.

This year's fad in just-so stories is the religion gene. Here's one of the latest from NewScientist, which should know better [Religion is a product of evolution, software suggests]. The article reviews the speculations of James Dow, an Emeritus Professor of evolutionary anthropology at Oakland University in Michigan.
To simplify matters, Dow picked a defining trait of religion: the desire to proclaim religious information to others, such as a belief in the afterlife. He assumed that this trait was genetic.

The model assumes, in other words, that a small number of people have a genetic predisposition to communicate unverifiable information to others. They passed on that trait to their children, but they also interacted with people who didn't spread unreal information.

The model looks at the reproductive success of the two sorts of people – those who pass on real information, and those who pass on unreal information.

Under most scenarios, "believers in the unreal" went extinct. But when Dow included the assumption that non-believers would be attracted to religious people because of some clear, but arbitrary, signal, religion flourished.

"Somehow the communicators of unreal information are attracting others to communicate real information to them," Dow says, speculating that perhaps the non-believers are touched by the faith of the religious.
Make no mistake. This is bad science. It does not meet any of the criteria of good science.

From time to time we challenge the veracity of press releases so it's always wise to check the source to see if the views of the author have been misrepresented. In this case, the original paper is online at The Jounral of Artificial Societies and Social Simulation website [Is Religion an Evolutionary Adaptation?]. Here's the abstract. You can read the article and decide for yourself whether you think this is a worthwhile contribution to the literature on evolution.
Religious people talk about things that cannot be seen, stories that cannot be verified, and beings and forces beyond the ordinary. Perhaps their gods are truly at work, or perhaps in human nature there is an impulse to proclaim religious knowledge. If so, it would have to have arisen by natural selection. It is hard to imagine how natural selection could have produced such an impulse. There is a debate among evolutionary scientists about whether or not there is any adaptive advantage to religion at all (Bulbulia 2004a; Atran and Norenzayan 2004). Some believe that it has no adaptive value itself and that it is just a hodge podge of of behaviors that have evolved because they are adaptive in other non-religious contexts. The agent-based simulation described in this article shows that a central unifying feature of religion, a belief in an unverifiable world, could have evolved along side of verifiable knowledge. The simulation makes use of an agent-based communication model with two types of information: verifiable information (real information) about a real world and unverifiable information (unreal information) about about an imaginary world. It examines the conditions necessary for the communication of unreal information to have evolved along side the communication of real information. It offers support for the theory that religion is an adaptive complex and it disputes the theory that religion is a byproduct of unrelated adaptive processes.
How many of you think that this work supports the just-so story and refutes other possibilities?


  1. From the title of your post, I thought you were going to write about this blog I came across a few days ago:

  2. My supervisor teaches an entire 4th year undergraduate biology class about the subject of 'just-so' stories and how they relate to the study of human origins (I prefer Jared Diamond's term for these: paleopoetry).

    What I find frustrating about many of these types of studies is that they make no attempt to refute the hypothesis of neutral evolution in any of their models (or any other type of indirect selection based model such as linkage with another trait). I'd also prefer a more satisfactory way of ruling out a cultural basis rather than genes as a negative hypothesis.

    It's completely inconceivable that we have specific genes for every single type of behaviour that can be given a name. If behaviours are obviously the product of multiple genes, then selection for certain types of behaviour (e.g., the desire for sex, dichotomization of 'in-group' and 'out-group', etc.) must affect other types of behaviour. Thus modeling the evolution of behaviours in isolation is probably as effective (and potentially misleading) as modeling the evolution of individual codons without taking into account the surrounding genome.

  3. Computer simulations aren't necessarily bad science. They can provide rigorous evidence of what outcomes are POSSIBLE or NECESSARY, for a given set of assumptions or observations. In this role simulations can be very valuable in confirming or debunking proposed explanations. For example, if someone did find genetic variation for the kind of trait Dow assumes, a simulation would help us identify how evolutionary forces could or could not act on it.

    Problems arise when we forget, as the author of the New Scientist article seems to have done, that the inputs are assumptions, not facts.

  4. I don't think the description "just so stories" is very good, in the sense that many things in biology sound like just so stories, even when they are fully documented. So, the result is that some things get called unfairly just so stories that do not deserve it (creationists use this mistake to their advantage, by the way); also, some fictional adaptationist elements can be sometimes interspersed with actually quite interesting data. I read them "sieving" the adaptationism out.

    "Panglossian", while requiring a little bacgkround on Voltaire, conveys what the criticism to adaptationism is about.

    An interestig thing is that within adaptationists that often get it wrong, you will find that many will declare themselves 1) Rationalist 2) Empirical. Such people can be prone to think adaptationism is in some trivial and direct manner "just the way it is", a fact as undeniable as stones will fall. No alternatives!

    Creationists are another case of those who think their religious picture is simply... "the way it is"

  5. If I understand correctly, these authors:
    1. constructed a computer simulation of the spread (to fixation) of an allele in a population
    2. constructed a selection gradient in silico that gave increased reproductive success to one allele
    3. ran their simulation with that selection gradient.

    The result: the beneficial allele went to fixation.

    What am I missing?

  6. why would you need to run a simulation to understand that? Genes favored by selection spread in the population. On other news, dog bites man.

  7. On the other hand, you gotta love the idea that religious expression is dependent upon biologic evolution.

    Creationists heads would explode over that one.

  8. There is a market. They get notes in the news and press on things like "why having a crooked nose is maladaptive" or "why do men prefer blondes".

    That's the only reason why this kind of crap floats; it gets a scientific appearance by invoking a "darwinian hypotheses", and deals with a topic people get all excited about. (Notice too that many theists love this kind of hypothesis: belief is a natural outcome)

    The press eats this up, you know, like those studies of the kind "avoiding hiccups makes you live longer".