selfish genes: the real (scientific) issue

arlin at arlin at
Mon Mar 1 10:18:32 EST 1993

In article <1mhbtbINN2bu at>, xia at (Xuhua Xia) writes:

> I think it important to draw a difference between Dawkins and other 
> people such as Hamilton, Williams and others. The Selfish Gene is not
> just about genes. I think that the book is promoting a world view, and
> this is also what many of Dawkins' critics perceive. I do not know
> Gould personally, but I believe that he also perceive the Selfish
> Gene as promoting a world view. Such a world view has the potential
> to be exploited by Social Darwinists, just as population genetics was
> once exploited by poeple favouring eugenics.  
> X. Xia

"World view?"  I don't see why Xuhua Xia says that Dawkins is a
socio-political writer-- he certainly isn't explicitly so.  Anyway, as long
as authors have something important to say and they make their biases
clear, I don't criticize them for trying to indoctrinate me with the "world
view" of, for example, a Marxist or a PC liberal-- much to the contrary, I
greatly enjoy reading both Lewontin and Gould!   As far as Xia's
suggestions about the threat of pernicious social Dawkinists: if they
exist, they are no more an indictment of the work of Dawkins, than "social
Darwinists" are an indictment of the work of Darwin.  Fair enough?

More to the point, the perspective of Dawkins is ontological reductionism,
which is *not* some flaky socio-political "world view" but a legitimate
philosophical position, though it may not be correct in the present
context.  And this, of course, is what we should be discussing (instead
of trying to dismiss Dawkins for emotional or socio-political reasons): the
possible correctness of the reductionist view of units of selection.

Surely, Gould's contention (in a recent NYT book review, mentioned last
week) that the inner circle of evolutionary savants has resolved the issue
in favor of the individualist view is tendentious.  Maynard Smith (savant
#1) wrote a negative reply to Gould in a subsequent NYT book review.
George Williams (savant #2) maintains major elements of the reductionist
perspective in his new book, _Natural Selection: Domains, Levels and
Challenges_ (Oxford, 1992).  For instance, we were all taught that in order
for natural selection to occur, one has to have variation affecting
fitness; in order for evolution by natural selection to occur, at least
some of this variation has to involve heritable traits.  This formula more
or less presupposes the legitimacy of a population of individual
"organisms" whose "fitness" might vary.  Williams says it differently:

"For natural selection to occur and be a factor in evolution, replicators
must manifest themselves in interactors, the concrete realities that
confront a biologist" (p. 13)

As soon as I finish reading it, I can supply more information about
Williams's book (however, it seems a bit broad to be useful in the present

Regarding the argument offered earlier by Ben Jones, to the effect
that genes cannot be the unit of selection because they are not acted upon
directly by selection, consider the counter-argument of Sober and Lewontin,
who note that 

"if selection is a causal process, in acting on phenotypes it also acts on
the underlying genotypes.  Whether this is 'direct' may be important, but
it doesn't bear on the question of what is and what is not selected for"
(p. 212 in _Conceptual Issues in Evolutionary Biology_, ed. Elliot Sober, 
1984, MIT press).

This simple logic invalidates the objection about "directness" by noting
that it is *not* the philosophical _sine qua non_ that it has been made
out to be.  Yet, this hasn't forestalled the propagation of the "directness
objection" (even Gould continues to spread this meme, in spite of the
fact that he published two chapters in the same volume with Sober and 
Lewontin, 10 years ago).

Arlin Stoltzfus 

More information about the Mol-evol mailing list