The Genish Self

Francois-Rene Rideau
Sun, 9 Jan 2000 00:27:32 +0100

Dear Cybernethicians,
   I've just been reading "The Selfish Gene", by Richard Dawkins,
a great read, in case you have the luck to still be able to discover it.
The theses in this book have become classic, and if you're any interested
in biological evolution, you'll find out that you explicitly or implicitly
know most of what the book says, although not with such a clear-cut and
well-thought out point of view.

To sum up the book, over large time scales, and given enough energy and
entropy, there happens self-replicating patterns or "replicators". These
patterns will by definition spread and occupy as much resources as they can,
competing with each other. This competition over limited resources means
that not every replicator will survive, and therefore naturally leads
selection of whichever replicators find tricks to survive at the expense
of other replicators. Replicators will thus build vehicles both to enhance
their replication process and to defend against each other. They will effect
the world at large so as to replicate better and faster (that's their
"extended phenotype"). They will group into same vehicles, and cooperate
in as much as they share the same fate (which gives individual cells,
individual bodies, individual societies of insects; or much more loosely,
species, etc), or fight in as much as they don't (which gives us rivalry,
predators, parasites, genocides). In all cases, the long-term selection
is over replicators, not over vehicles, which is why evolution may be
interpreted in terms of "selfish genes", and not over "selfish individuals";
limited cooperation between individuals (or within individuals)
can be seen as the effect of "selfishness" at the gene level.

Please pardon me if the above summary is completely confusing or
ununderstandable; the book itself is quite easy to read, well-explained
in good english, with lots of examples and justifications, yet absolutely
nothing technical; it doesn't rush to squeeze 300 pages worth of text
into 200 words. If that means anything, chapter 2 of it was selected by
Hofstadter&Dennett in their great collection of commented texts "The
Mind's I", so it can't be a bad book.

Now, for a few questions and possible answers that the reading suggest:
* can the above concepts help us understand why species have different
 numbers of chromosomes? why most of the genetic material seems to have
 no chemical expression as proteins whatsoever? how evolution can go on
 indefinitely despite a finite reserve of genes, and the limiting fragility of
 viable genetic combinations?
* to the first question, the book itself answers how infestation of bacteria
 that manage to pass along with embryo material will end up with a merge
 of genetic material, because genes thus passed share the exact same outcome
 as "normal" genes shared in usual reproduction, and will thus tend to
 cooperate just as much as other genes, until they eventually merge into
 a same common pool.
* the second question is very partly answered by the book: unexpressed DNA
 is propagated "just because it can", for its own good. But there is more to
 it than that. Why does it accumulate? Since it does spend some resources at
 being copied (however small they might be), what do other genes gain from
 their presence so as to tolerate them? Couldn't they be detected? What are
 the tradeoffs?
* Unexpressed DNA can benefit other genes in a common individual in at least
 two ways, one short-term (that makes them stable), and another one long-term
 (which makes them selectible): on the short run, given a systematic rate of
 normal errors at each copy (and we can suspect that at least some cross-over
 happens systematically during meiosis, with every chromosome pair being
 separated from a random location into same-length subcopies), every gene
 suffers a lesser chance of being cut in half by the process if there is
 more available DNA overall. Of course, in locally stable absence of such
 systematic errors, there could be a locally stable strategy to only copy
 expressible genes (by mixing copy with expression, and discarding buggy
 strands); but such strategy would be very expensive (finding the right
 chemical combination for such verification), and would make the resulting
 species very fragile with respect to mutation, since every single mutation
 would modify a useful, expressed gene, and likely remove it. Gene pools
 adopting such strategy would thus adapt to prevent mutation, and hence
 would be at a gross disadvantage in any evolutionary arms race.
* We can thus see that unexpressed DNA can be seen an evolutionary adaptation
 in presence of systematic mutation, that benefit genes in the pool
 from being the victims of said mutation. This in turn raises the question
 of the presence of systematic mutation. Can a gene for some such thing have
 been selected? Well, it is possible that a gene for systematic mutation be
 beneficial in as much as it encourages quicker genetic adaption in the
 gene pools that adopt it; this would sure incur some detriment to the genes
 that risk mutation, but would benefit the gene that induces mutation. It
 could be said that somehow, the gene for systematic mutation is a parasite
 to the other genes; it exploits existing genes, and forces them to evolve
 at a quicker pace than they would like, for its own good. Much like the gene
 for meiosis, that exploit other non-sexual genes, reducing their immediate
 expectation of replication, for its own advantage.

[ "Faré" | VN: Уng-Vû Bân | Join the TUNES project!  ]
[ FR: François-René Rideau | TUNES is a Useful, Nevertheless Expedient System ]
[ Reflection&Cybernethics  | Project for  a Free Reflective  Computing System ]
Why is there only one Monopolies Commission?
	-- Lord Sutch