Re: Dissecting the Aquatic Ape: Bipedalism

Leonard Timmons (ltimmons@mindspring.com)
Thu, 08 Aug 1996 10:10:31 -0400

HARRY R. ERWIN wrote:
>
> Leonard Timmons (ltimmons@mindspring.com) wrote:
> : I have noted that among most species, when males fight they try to determine
> : one another's ability to inflict damage without actually doing so. I have
> : heard of the survival advantages to this strategy and thus the reason for
> : its evolution. However, when I look at us and our ancestors, I see animals
> : that are just about completely unarmed. We cannot really inflict damage on
> : one another of the same magnitude as one horned animal goring another. I
> : don't know why we lost our incisors, but the prevalance of alliances among our
> : ancestors may have eliminated even these (I guess).
>
> This game is chaotic, by the way, if the strategies are learned. It
> converges rapidly to an ESS if learning is not important.

Since you mention the chaotic nature of this game, I assume that there
are
computer simulations of these strategies that show this chaotic
behavior.
Is this correct?

>
> : One of the things that I noted in Ms. Goodall's book _In the Shadow of Man_,
,
^^^^^^^^^^^^^^^^^^^^^^
This attribution may be wrong. I think the this was in her larger more
extensive book. It has been so long since I read it and I can't find it
now.

> : is the difficulty that a group of chimps had in killing another individual
> : even when he was alone and outnumbered. The attacks were brutal, inefficient,
> : and prolonged. I remember thinking to myself, that I didn't remember reading
> : anywhere about the evolutionary consequences of two animals who fight, but
> : one is always killed and the other always unharmed (or only trivally so).
> : It seemed to me that the evolutionary consequences of such a situation would
> : be significant. In fact, I thought that such a situation must be somehow
> : unstable -- and that is why I had never heard it discussed.
>
> See above.
>
> : To me, in order for this situation to exist, the two animals would have to
> : be nearly completely unable to harm one another. This seemed possible for
> : our ancestors in light of the accounts of the attacks above. However, at
> : some point our ancestors would have discovered how to quickly and efficiently
> : kill a rival and almost always remain unharmed. In my mind this would
> : cause a rapid spurt of evolution until the situation was corrected.
>
> Fitness is defined in relative terms. If the interpersonal behavior
> within a specific group became deadly, the group would disappear.

I can't imagine that this practice would have a disastrous impact on the
group. After killing a few rivals, the strangulation male would be the
alpha male. He would then need only to kill his rivals as they
appeared.
He would almost never have to harm a relative and a relative would never
have to harm him. But in any event, at first, this technique would be
very
difficult for a protohumanoid to perform. The success rate would be
low.
But the rewards of success would be quite high. If a strangulation
attempt
failed, no one is worse for the wear. The enemies are still enemies.

In the ESS studies that you did, did you consider the slow adoption of a
strategy, or did your models begin with strategies that are executed
perfectly by the participants? Since chaotic behavior often results
from
systems that are overdriven, a new strategy which is suddenly executed
perfectly by the participants might lead to a chaotic solution, but at
some low level of proficiency, the solutions might be marginally stable.

> : So I asked myself, "In what way could one protohumanoid learn to kill another
> : protohumanoid and be spared in the vast majority of cases?" My answer was
> : that this could be done by strangulation. This is a learned behavior and
> : can probably be taught to chimps today.

Does anyone want to comment on the ethics of teaching an isolated group
of
chimps this method of dealing with their rivals. One of the reasons
that I
like this theory is that unlike many others, it is testable.

-leonard