Re: Group selection

From: Keith Henson (
Date: Tue 22 Feb 2005 - 14:02:30 GMT

  • Next message: Keith Henson: "Re: Greeks (was Group selection)"

    At 08:53 AM 22/02/05 +0100, you wrote:

    > >The main argument against group selection is that it does not close the <br>
    > >causal loop, mainly against free loaders taking advantage of
    > others. Now<br>
    > >if the group is *also* your relatives, the loop does close.<br>
    >I agree. If the group is completely inbred and effectively keeps outsiders
    >out, then kin selection and group selection will be the same. The problem
    >is that even a small immigration of egoists or back-mutation can destroy
    >the scheme because the egoists can outcompete the altruists inside the group.

    That depends on other genes such as "kill the slackers." :-)

    In actuality, slackers tend to be bred out of the species because males have to hunt to obtain enough status to be attractive to females.

    Another factor is that relatives are always valuable to your genes. That's why people have children and take care of sibs and to a smaller extent the extended members of your tribe, most of which were (in the EES) related to you. Even if you were adopted into a tribe from a far away one, people have mental mechanisms that get activated so you consider these unrelated people as if they were related.

    In fact, we have *specific* mechanisms, capture-bonding, that bonds newly captured tribe members (usually women) into the tribe. That's the mechanism activated in army basic training.

    >My simulations show that the topology (or geographics) og migration plays
    >a significant role. If immigrants come only from nearby groups and group
    >selection is sufficiently strong then the altruist gene can become fixated
    >in an increasing geographic area. But if immigrants can come from distant
    >groups then the altruism gene tends to win.

    Win or lose?

    >This is one of the problems with earlier models. They simplify their
    >models in order to make them mathematically tractable, and these
    >simplified models can't handle the topology. This is why I have to use
    >similation. Also, it won't work without random genetic drift, which most
    >traditional models ignore.
    >The species that have the strongest altruism, i.e. social insects and
    >naked mole rat, are ferociously keeping foreign intruders out. This is
    >necessary for group selection to work.

    I don't exactly understand your argument here. It seems to me you are arguing for Hamilton's gene based evolution of altruism rather than competition and selection at the group level.

    If so, I agree of course.

    Keith Henson

    =============================================================== This was distributed via the memetics list associated with the Journal of Memetics - Evolutionary Models of Information Transmission For information about the journal and the list (e.g. unsubscribing) see:

    This archive was generated by hypermail 2.1.5 : Tue 22 Feb 2005 - 14:21:21 GMT