Re: "The Selfish Gene"

From: Stevan Harnad (
Date: Sun Nov 02 1997 - 22:04:07 GMT

On Thu, 16 Oct 1997, emma jane fletcher wrote:

> Atoms did not 'exist' in the sense that they had conscious
> experience of 'life', they just were: solitary units devoid
> of any experience, although presumably they had some
> continuance of being. [Are quarks and other such constituents
> of atoms immortal, or are they subject to decay?].

Good question: I'm not sure (because I'm not a physicist). According to
the conservation laws of physics, energy does not increase or decrease.
it just changes form -- and matter in the form of atoms and quarks is
among the shapes it takes. It's also true (the 2nd law of
thermodynamics) that everything is "decaying" -- if by decaying you
mean moving from a state of order to one of disorder.

But if what prompted the question was that you were wondering why we
don't speak of the selfish photon or any other invariant that is
preserved across different physical interaction, I think it's because
we know living things have minds, so we know that there really is an
"entity" there; hence it is ironic or counterintuitive to learn that the
entity is just a survival machine, "driven" by mindless, selfish genes!

> should the
> molecule that produced the duplicate molecule decay, the
> molecule type would still be preserved in the form of the
> 'younger' duplicate molecule. [This is assuming that
> environmental factors remained stable]. In this way
> successful configurations of building blocks may be seen to
> have achieved immortality.

If what you mean by "immortality" is continuing to exist, unaltered. But
of course what we mean by mortality is: Capable of dying, which we
reserve for living things (for the reason I mentioned above, I think),
rather than applying it to all "stable" things. (The sand in the Sahara
may be ever so stable, but we don't think of it as "immortal."

> A gene then, is in effect, " long-lived" stable replicator,
> which through the process of " self replication" ensure the
> survival of the gene form into the next generation. As such,
> a single gene IN ITSELF is not immortal. Immortality is
> achieved through a continuous process of exact replication
> through successive generations. I.e. the gene must produce
> a faithful reproduction of itself before its "death", and the
> younger replica must produce a faithful reproduction of
> itself before its death and so on.

It's probably wrong to think of the gene as alive, or immortal.
It is the organism and perhaps autonomous parts of itself that are
alive, and they certainly are not immortal. The cell that reproduces by
splitting (mitosis) is not immortal either, for then which of the two
clones is it, living on endlessly? For the same reason, meiosis is not a
ticket to immortality, otherwise we could again ask: what lives on,
before and after the fertilisation?

Unlike immortality, replication is fairly straightforward: What is being
replicated is the code and the (capacity for) a certain structure.

> A gene may be shown to be "bad" if it is consistently
> unsuccessful; that is copies of the original gene always
> happen to be in a vehicle which fails to succeed. It is
> unlikely that a would be successful gene will always happen
> to share a body with unsuccessful genes. In the same way, it
> is improbable that a replica that always happens to be in a
> body that perishes is simply unlucky. It is much more
> probable that the gene is a bad gene. Ultimately a gene is
> bad because it is unsuccessful i.e. it perishes.

I had trouble following this: Do we, in addition to the language of
success and failure at replicating, add the notions of "bad" and "good"?

> What makes a Gene Good?
> The gene which happens to be good at cooperating with other
> genes in the gene pool will be more likely to survive and
> duplicate [gene pool = the long term environment of the
> gene]. As such, genes tend to be favoured by natural
> selection if they tend to work well in their environment.
> This includes making use of other genes in the gene pool.
> Such genes are more likely to survive. Ultimately a good gene
> is "good" because it is successful: since nothing succeeds
> like success.

So it sounds like "successful" and "unsuccessful" are enough; no need
to bring in good or bad (especially at this subpersonal level where it
makes no sense to say "who" is good or bad, or even "for whom" they are
good or bad).

> Dawkins speculates that "tiny influences on survival
> probability can have a major impact on evolution: this is
> because of the enormous time available for such influences to
> make themselves felt". Is Dawkins speaking about
> environmental influences shaping evolution rather than
> genetic influences affecting survival probability?

I think he means both: A small advantage can be vastly magnified across
many generations.

> transported in a vehicle may span decades). If the replicated
> gene is successful it is selected (since only successful
> genes survive and so enter the next generation).

I think you're multiplying entities needlessly: success IS successful
replication. Selection IS having been successful.

> Surely then,
> uncertainty about whether a gene will be selected, and hence
> uncertainty about what effect it will have on survival and
> evolution, could not possibly span across generations. Even
> recessive genes are selected to be carried in their vehicles.

That's true: But being diploid (mother and father's germ line) means a
lot of genes can be carried for the ride as fellow-travellers, neither
helping nor hindering survival, just being there. Yet their advantages
(or disadvantages) may not be felt until much later, should the
environment change to make them positively advantageous or

This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:08 GMT