Selfish genes (long)
seb1005 at mbfs.bio.cam.ac.uk
Tue Mar 2 10:19:16 EST 1993
formulation, teases out some points, and examines them.
--Sorry if this is posted twice. I've been having newsreader problems.
Greetings All. Sorry for entering into this discussion about
two weeks late, but it's been so interesting thus far that I couldn't
help jumping in. There were a few points made which I wish to
accentuate and/or debate either because they are either subtle issues
to which I was only recently made aware or they are topics about which
I am unsure. I look forward to seeing some thoughtful comments upon
Pleas note that I've abstracted bits and pieces of articles
and have not noted my deletions (since I've deleted most of the
articles). Hopefully, I haven't ended up taking anybody's words too
out of context--if I have it means that I misunderstood what you were
getting at, so please correct me! Well, here goes...
This discussion all began with a note which included:
From: Francis Heylighen <fheyligh at vnet3.vub.ac.be>
This implies that if our present environment, where the probability of
being killed by predators, starvation or diseases before reaching old age
is much smaller than in the original human environment, would continue to
exist for a million year or so, natural selection would promote genes
that would make us live longer.
...it's hard to make statements like that if for no other reason that
our present environment most certainly won't last for a very long
period of time; indeed, simply having genes that made us live longer
would change the environment. However, all other things being equal
the argument does make sense to me so long as we also include the
stipulation that reproductive age also increases. (This would
presumably mean that menopause would come later and that there would
be a decrease in the rate of increase of birth defects with respect to
age of the mother.)
From that point of view, I would even reject the more careful formulation
that mortality "may be useful for genetic evolution". The only reason for
mortality would be that not having it (i.e. maintaining the necessary
apparatus for unlimited self-repair of cells and organisms) would take
away many resources from reproduction, for a very small return in terms
of reproductive fitness. But I would hardly call that "being useful".
The notion of "being useful for genetic evolution" is the problematic
one, I think. Under Dawkins' formulation, genes do not "want" to
evolve. In his formulation, evolution occurs despite the best efforts
of genes to thwart it.
I would argue instead that theoretically, mortality is a *consequence*
of evolution and not an accomplice to it. Consider an organism whose
genes would make it immortal and cause it to continually produce
offspring. For a long period of time, we can imagine the genes (in
Dawkins' sense only--see below) as being very successful: lots of
identical copies are made. However, eventually conditions will change
and the genes which made the organism immortal in the old environment
will fail in the new environment. So, the organism will die.
Alternatively, while our friendly immortal organisms (call it Imms)
are busy forever producing identical offspring, other organisms (Muts)
have been undergoing various amounts of mutagenesis & recombination.
After enough time, we would expect that one of these Muts will
eventually be better adapted to the environment than the Imms and
cause the death of the Imms. (In a sense, the second example is
really a special case of the first, since the development of the
more-adapted Muts represents a change in the environment of the Imms.)
This points out, incidentally, what I believe to be one of the central
flaws with the "Selfish Gene" formulation. That is, it denies the
importance of change in evolution. In particular, if you have a gene
X and a gene Y, both of which are equally suitable at time t, but at
time t+1, gene Y is more suitable than X then one would expect gene Y
to dominate. So far, so good. However, perhaps gene X "knows" that
it is very sensitive to current conditions and therefore has somehow
constituted itself in such a way that it receives mutations on a
higher-than-average rate. In this case it is likely that one of gene
X's progeny, perhaps gene Z, is better adapted at time t+1 than either
gene X or gene Y. Consequently, gene Z flourishes.
Dawkins's formulation would not predict that gene Z would ever come
into existence, because it would not predict that gene X (being
selfish) would ever "consider" mutating itself. In other words, the
"Selfish Gene" formulation denies the existence of "meta-evolution."
(The theory of meta-evolution was introduced, to the best of my
knowledge, by David Layzer as the "beta-gene hypothesis" in a paper in
American Naturalist in 1978.)
...later on, the following note, which got more on the topic of the
specifics of Dawins' books was posted:
From: jonesbb at BELOIT.EDU (Ben Jones)
In his zeal to trash group selection, Dawkins trashed the language of
evolution instead. (And produced a book which is eminently readable for
beginners in evolution, and explains simply and concisely many of the most
complicated concepts in evolutionary theory. The notable exception is
group selection, which he does not understand at all.)
I agree--in producing a book which makes most of evolutionary theory
easy to understand Dawkins, did trash the language of evolution. Or,
more importantly, he deftly modified the language of molecular
Most significantly, he defined a "gene" (note the quotes!) to be the
smallest piece of DNA whose sequence is conserved over a long period
of time. It need not have any function, it need not represent one
cistron, one protein, or one piece of RNA. (This is a rough
description, BTW; I don't have the actual book in front of me.)
HOWEVER, THIS DEFINITION HAS LITTLE CORRELATION WITH THE
MOLECULAR-BIOLOGY NOTION OF THE GENE!
What's even sneakier is that he defines all instances of a given
sequence as all being the same gene. That is, if you and I share in
our DNA some sequence, then we both have the very same single gene.
And that SINGLE gene will (under Dawkins' construction) try to
preserve itself, which--in this case--means preserving both YOU and
ME. One gene tries to preserve two individuals.
So, by defining a "gene" as a group, Dawkins' is able to claim all
apparent altruism is really selfishness by the GROUP of DNA sequences
which are collectively a single "gene."
Incidentally, if we define an "eneg" as Dawkins' gene WITHOUT this
built-in group notion, then the "Selfish Gene" theory turns into the
"Altruistic Eneg" theory, for each Eneg wishes to preserve all other
identical Enegs. A friend of mine has gone through most of the major
points in _The Selfish Gene_ and successfully translated them into the
Altruistic Eneg formulation.
A corollary to this is that I think that Dawkins' book is
fundamentally based on the notion of group selection and that he
understands it VERY well, but doesn't admit this to his reader.
Indeed, he may not admit it to himself!
On the other hand, I am often amazed at the number of evolutionary
scientists who think the book "The Selfish Gene" is good science rather
than a well-written tirade against group selection.
I think that this is a bit harsh. As you stated above, the "Selfish
Gene" formulation is a very good way of explaining the difficult
concepts behind evolution to a wide range of readers. What is
important to note is that the "Selfish Gene" is *only* a formulation,
and certainly not the only one. (It is, I will concede, a more
intuitive one than the Altruistic Eneg. ) Moreover, for the
meta-evolution reason stated above, the "Selfish Gene" formulation may
not even be a completely correct formulation.
A consequence of what I've written above is that the following...
From: jonesbb at BELOIT.EDU (Ben Jones)
It seems it never occurred to Dawkins that *NONE* of the "levels of
selection" need be the "fundamental" one. In reality they all go on at
once, independently of each other. If the conditions at a particular level
of grouping permit, then there is selection. If not, then there isn't.
But since Dawkins feels that there must be a fundamental level, he claims
it must be the gene. His main criterion is "indivisible particulateness"
of heredity, and he claims the gene approaches this ideal. I think he
didn't go far enough. He never considered the exon or the codon. By his
criterion, in fact, the *base pair* is really the fundamental unit, and all
other levels are mere group selection. ;-)
...actually is unfair to Dawkins. This is because within Dawkins'
DEFINITION of a "gene" there is room for the exon, codon, or even
basepair to be the fundamental unit of selection. Under his
definition, all of these (as well as large units of chromosomes) are
"genes"! Pretty sneaky, hm?!
It is just this sleight-of-word which permits Dawkins to get away with
using words like "selfish." (As I note above, a different, and
perhaps better, definition of gene allows us to use the world
"altrusitic.") Is is therefore that I agree wholeheartedly with the
From: xia at hardy.u.washington.edu (Xuhua Xia)
If we need to assign any credit, then Dawkins is certainly the first
to prosent the Selfish Gene as a world view.
I am not particularly certain about my own position. I found Dawkins'
writings fascinating, but I found the writing of his critics
fascinating as well. I believe that the controversy cannot be
resolved by biologists.
...the reason it cannot be resolved by biologists is that the issue is
not biology. Evolution is (IMHO) neither selfish nor altruistic. It
is just a phenomenon which is a consequence of the mechanism of
ordering of biological systems.
One can *formulate* (essentially) correct descriptions of evolution
using whatever adjectives they like, if only definitions are chosen
properly as well. Dawkins came up with one very successful
formulation and he is to be credited for that.
From: xia at hardy.u.washington.edu (Xuhua Xia)
(e.g., Gould in his vicious assault on Dawkins and anyone
who favors his arguments, in a recent NYT Book Review)
Would you mind posting a reference to this review? I would
love to read it.
I look forward to seeing any thoughts people have on the points I've
raised. Please do post to the net rather than to me so that we can
continue the discussion.
Steven E. Brenner | Internet seb1005 at mbfs.bio.cam.ac.uk
Department of Biochemistry | JANET seb1005 at uk.ac.cam.bio.mbfs
University of Cambridge | Laboratory +44 223 333671
Tennis Court Road | Home +44 223 314964
Cambridge CB2 1QW, UK | Lab Fax +44 223 333345
More information about the Mol-evol