Group selection (was The Purpose of Life)

From: Smigrodzki, Rafal (
Date: Thu Jan 31 2002 - 16:26:24 MST

From: Eliezer S. Yudkowsky [] wrote:

What I mean is that you *may* be able to say:

"Evolution predicts that there should be an equilibrium between mutation
rate ensuring sufficient mutation among a group of offspring that some
will be able to adapt to environmental changes, and mutation prevention to
maximize the overall health of those offspring." It would be pushing it,
but you might be able to prove it. It's been demonstrated to be true at
least once, for the special case of the immune system, where the locks
have to change every generation.

But you can't say "maximum variation for potential evolution" or "adequate
health of the population" without inherently postulating group selection,
and postulating group selection is almost always a mistake.

### Mutation rate is one of the physiological parameters which evolve, just
like body temperature, or serum osmolality, in response to a large number of
influences. Some of these parameters act primarily on the individual level.
For example, the tradeoff between the amount of resources invested into DNA
proofreading vs. other functions (e.g. maintenance of large stores of
glycogen in the liver) has effects on short-term survival (enhanced by a lot
of glycogen for running) vs. long-term individual survival (enhanced by low
mutation rate). So, group selection is not necessary to produce a setpoint
for mutation rate (this is what you wanted to say, but then as far as I
understand it, Alden didn't imply the necessity of group selection, either).

Yet, on the other hand, there are some processes acting on whole populations
- for example, during trypanosomal infection, there is a cyclic interaction
between the host immune system, which produces antibodies to the parasite's
coat proteins, and the generation of new coat epitopes by a hypermutation
mechanism. While turning this mechanism down in a particular individual
parasite isn't likely to impact its survival in the short run (2-3 weeks,
during which it can freely proliferate and even outcompete the hypermutable
ones, if only by chance), once the full immune response is on, the whole
population of the low-mutation parasites is wiped out, and only the
offspring of the few mutated ones survive. So, in the longer run, the
parasites are maintained at a (presumably) optimal mutation rate.

Then there are processes involving social groups of animals (such as
wolfpacks, human hunter-gatherer groups), where social cohesion and the
behaviors necessary to maintain it have a huge impact on survival of all the
members of the social unit - should it split into inappropriately small
groups as a result of intragroup aggression (a group-related event), all of
the individuals will suffer. (This example doesn't relate to the regulation
of mutation rate but it does support group selection).

More at It appears that
the theory of group selection is having a comeback.


This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 13:37:37 MST