Re: "The Selfish Gene"

From: Stevan Harnad (harnad@cogsci.soton.ac.uk)
Date: Tue Jan 13 1998 - 20:23:09 GMT


> From: "Katherine Lyne" <kml295@soton.ac.uk>
>
> In consideration to the selfish gene theory - how would sociobiology
> explain suicide??

To put this in context: ALL of us commit suicide eventually! We don't
do it consciously or deliberately of course; it's our genes who do it
for us, because we are just their vehicles. The way we all commit
suicide is by aging, hence eventually dying! See Dawkins's discussion
of this form of genetic "planned obsolescence" and its advantages over
living forever in his sections on mortality/immortality.

A good way to look at it is that it is costly to stay alive, so you need
to select FOR a mechanism that pays that cost (by growing, eating,
etc.). If paying the cost advantages your genes, you may evolve the
mechanism to pay it. But once your genes have replicated, you actually
risk getting into competition with them if you stay alive (there may be
only so much food to go around). So you're doing your own genes a
favour if you check out; your genes then "make" you check out simply by
never having evolved a mechanism that keeps you going well beyond a
sufficient number of reproductive years (and child-rearing years, if
you're that kind of species) in the first place. Ask not why my genes
would make me die: ask how long and under what conditions is it useful
for my genes to keep their vehicle alive!

Apart from that special case of self-destruction (mortality), the
answer to your question depends on which species you are asking about.
Since it is very unlikely that animals other than us have a concept of
"death" or "mortality," that means that even when they DO do something
that hastens their death, they are not committing suicide in our sense:
By ceasing to struggle and going into shock when overpowered by a
predator, for example, an animal is not so much committing suicide as
failing to have evolved a positive inclination to continue useless
struggling under such conditions in the first place: The the Blind
Watchmaker can select for considerable struggle against odds, but not
against infinite or even overwhelming odds. (Indeed, against a global
flash-flood, for example, it COULD not select the tendency to struggle
on and on, because no one would survive, so there would be no basis for
selective advantage; moreover, one-off environmental disasters are by
definition not the everyday events that the Blind Watchmaker works
with; so eventually you give up and drown, not because you committed
suicide, but because your genes' capacity to code for heroic struggle
is limited.)

Something like this is also true about self-endangering or
self-destructive behaviours such as the fighting rather than fleeing of
soldier ants when their colony is threatened by enemies: This is not
conscious suicide either: just an extreme altruistic behaviour, fighting
to the death for the sake of the shared selfish genes of those who
survive; in other words, "inclusive fitness."

Even in the alleged "population control" adaptations, like the lemming
"collective suicide" (if it really exists) that Dawkins discusses, it
is again not conscious suicide, but merely a behaviour that happens to
result in their losing their lives under certain adverse conditions of
crowding or climate or whatever. Think of this as part of the extra
flexibility that genes gain in how they handle various different
possible environments: instead of coding in one rigid possibility, they
"use" a space/time "cue" (such as crowding) to flip the switch on two
(or more) behavioural strategies: If there is plenty of space and time,
be fruitful and multiply; if crowded, don't (it's no use, and it just
makes it LESS likely that any genes at all will survive); instead, let
yourself be borne by the crowd (even if it hurtles over the brink).

The lemming case is only a bit tricky, because it seems to require a
group-selection explanation (the "suicide" of most helps the few who
do happen to survive), which Dawkins thinks is impossible, because
there is no "group" vehicle other that of inclusive fitness (we'll
discuss this again in Debates next semester). Try to think of what might
make it a good individual strategy in the same way moving together is a
good camouflage strategy for zebras even though it means not hiding
individually. (You see how it's easy to tell just-so stories! but at
least it gets you thinking of suicide the right way around, which is
rarely that of deliberately trying to die.)

Now with people the story is different: We can want to die consciously,
and can try to make it happen deliberately. Now I think you'll agree
that wanting to die when one has unbearable, incurable pain is just an
extension of the general strategy of trying to escape pain, so it does
not need a special mechanism. The adaptive value of avoiding or
escaping pain in general is obvious.

The same could be said of depression: There are many speculations about
its "adaptive value". Perhaps in moderation it prevents a lower-ranking
individual from needlessly risking its life to keep trying to claw its
way to the top through combat in which the odds are against it. If
feeling "down" protects you from risks in the normal case, then, like
everything (weight, agility, attractiveness) there will be a
bell-shaped curve of variance, and there will be some individuals at
the extremes, whose reaction to even minor disappointment is
catastrophic and suicidal. Or maybe for some people the mood metre is
stuck on low, again not for its own special evolutionary reasons, but
because adaptive mechanisms, such as perhaps moderate reactive
depression, can go wrong.

So I doubt that we have any specific genetic inclination toward
conscious, deliberate suicide, any more than we have the specific
genetic inclination to run a 4-minute mile. But suicide and 4-minute
miles are among the athletic and cognitive options that are open to a
species whose athletic and cognitive resources are the size and shape of
ours!

[If you think any of this sounded like a just-so story, say so! That is
one of the liabilities of this kind of theorising, and it should be
looked at critically.]

> I can understand that if it is more beneficial to your offspring to be
> dead for resources sake, then suicide would make sense.

It's all about successfully replicating genes. If your genes make it
into the next generation vehicle, then the continuing needs of your own
vehicle, once it can no longer help along your progeny, are in competition
with your progeny. Apart from that, there are Dawkins's general
arguments against immortality.

> But what about childless suicide?? And mass suicide like Jim Jones in
> the 70's or David Koresh? Isn't this an example of cognitive
> mechanisms working against the selfish gene???

You left out the EEA! A surviving postpubertal human adult is normally
the parent of several offspring. A male has no way to know whether he
is or isn't a parent, so he is free to off himself when he feels very
down. A female who is childless in adult reproductive years is probably
infertile, and there is no special provision by the Blind Watchmaker
for miracle cures of infertility later in life, so a childless woman is
as ready as a man to off herself -- indeed, even moreso, because a
man's promiscuous reproductive strategy is adaptive longer in the life
cycle than a woman's, especially a childless woman's. From this we
would predict more suicides by women than by men (true) and more perhaps
by older childless women (not sure). Childhood and adolescent suicide
can only be explained by the depression mechanism (or some other
mechanism) going wrong. (Post-partum depression, the one some women feel
just after giving birth, which results in a few suicides, is also best
explained as a normally adaptive hormonal mechanism -- depression
slowing one down after a major trauma? -- gone wrong.) Mass suicide
because of some ideal or some delusion is just part of the general scope
of the general cognitive abilities we have inherited. It's the same
thing that lets us decide to go to the moon: it wasn't specifically
selected for!

Is cult mass suicide "against" the selfish gene? Well, anything that
thwarts replication is against the selfish gene, in a way, but we do
that all the time, even when we make an impromptu decision to eat a
sandwich instead of mating with an available partner! That's just
variance that the selfish gene is "buffered against." Those who start
to prefer sandwiches TOO often, if it is for genetic reasons, may not be
able to pass on that unfortunate tendency, but that's just the selfish
gene again story again.

In a way, it's impossible to go "against" the gene: Even if Koresh had
persuaded or forced every human on earth to do himself in, the outcome
would be no more "against" the selfish gene than a natural disaster that
wiped us all out would be. In general, one-offs are not what biology or
evolutionary theory is about.

Perhaps what you mean is something like difference between the EEA and
today: Sugar was rare and good then, common and bad now, so the gene
coding our sweet tooth is acting against its interests: Not even that
is right. Genes get selected blindly because they help us survive and
reproduce. If the environment changes, some formerly beneficial traits
may now be a handicap. If they are enough of a handicap, they will be
selected against in the usual way. Specific GENETIC tendencies toward
forming suicide cults would have the same effect: they would be
selected against, in the usual way (but such specific genetic tendencies
do not exist). Cultural tendencies toward suicide cannot be selected
against genetically except by selecting against the whole general
cognitive capacity that allows us to have culture in the first place,
and that seems too radical a turning back of the evolutionary clock for
the Blind Watchmaker to tumble into.

A lifetime renunciation of reproduction can be described as going
against our genes' selfish interests, I suppose. By that token, every
(genetically caused) impulse we resist goes against our genes, but
in that sense we go against our genes quite a lot!



This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:09 GMT