SJ Gould on genome

Lisa & Ian Murray seamus at accessone.com
Mon Feb 19 10:42:48 PST 2001


Gould's article contains an important fallacy, maybe several. He infers from

the relative paucity of "human" genes that the "central dogma" (one-way

directionality of explanation from DNA through protein) is false, because so

few genes cannot produces od many proteins, and infers from that we must

reject "reductionism," understood as explaining "higher level" phenomena,

e,.g., biological, or maybe even psychological and social phenomena, in

terms of lower level phenomena, e.g., genetic or even biochemical phenomena.

None of these inferences will hold.

******

Well "human complexity" is left undefined by G. Does he mean physiological, biochemical, intracellular, cognitive, sociological? Also unidirectionality is consistent with one-to-one-many, one-to many-to many, one-to many-to one, many-to many-to many, many-to one-to one, one-to many-to many, one-to one-to one and many-to one-to many mappings. Only extremely detailed observation of the temporal dynamics is going to answer this question for each genome.


>>I should remark that years ago I wrote a dissertation defending a

sophisticated reductionsim about the mind (roghtly taht we ould expklain

mental phenomena in physical terms) against abd philosophical objections. I

published several papers out of it, then lost interest, since it seemed to

me that I was saying oveer and over again that either the scientists would

produce reductions, in which case the philosophical objections were

obviously wrong, or they wouldn't, in which case the issue would lose its

urgency. I no longer care very much about reductionism. But I do know a lot

about it. And despite Gould's immense scientific learning, far greater than

mine, what he has presented here is another bad philosophical objection, not

a scientific refutation.

******

Yup


>>OK, the first step: few genes, many proteins. That refutes the idea that

there is a single gene that by itself turns on each protein in every

circumstance. It does not even refute the idea, however, that there is a

single gene that turns on each protein in some circumstance. Nor does it

show that discrete combination of genes do not make each particular proteins

in some or all circumstances. If only two genes are required to turn on or

make each protein, we have already got a huge increase in the number of

explananda in the explanatory base. This means that the central dogma

stands, unless we identify it with the one gene-one protein view. The

paucity of genes does not by itself tell us anything about the explanatory

direction.

********

Welcome to the world of combinatorial explosion :-) Also see my above.

Now, the second move: from the failure of the central dogma to the falsity

of reductionism. This is a bad argument for several reasons. The first is

already indicated; since several genes, or single genese in particular

circumstances, may make single proteins, the lower level phenomena, so

understood, may still explain the high level ones.

********* Well G. doesn't state that the central dogma is = to one-to one-to one models, he says "their" model defines it as such; looks like he's setting up a rhetorical straw man. Who's the "they" since G. doesn't believe in the one-to one-to one monopoly himself it's doubtful many others did. CD is about temporal dynamics, that's all; it's even consistent with models of organisms that don't have Weismann's barrier incorporated into their genomic structure [of which there are many, see Weber and Depew, 1995]

Second, I remark that is is somewhat difficult to imagine how else it could

be: is Gould suggesting that there are genetic explanations for biochemical

phenomena? That, say, the laws of population genetics explain how RNA

encodes information from DNA and makes proteins? How could that be? Maybe he

just means that there are several ways that the biochemsitry of DNA/RNA

means proteins, that, in the jargon of the discussion, the manufacture of

proteins is multiply realized. It's true that multiple realizability is

often taken to be a refutation of reduction, but it isn't. It just shows

that there are several reduction bases.

********

Yup, multiple realizability is ubiquitous in genomes.


>>But even if the "central dogma" is refuted by the results of the human

genome project, and we accept that there are emergent properties that can

affect lower-level ones, evolutionary properties that can affect biochemical

ones somehow, that does not show that reductionism is false, that we cannot

explain higher level phenomena in terms of lower level ones, even

exhaustively. It just shows that explanation can run both ways. It would

take more argument to show that exhaustive explanation of the higher levels

in terms of the lower levels is impossible; that does not follow from the

premise that explanation in the reverse direction is possible.

*******

What kind of emergence? Epistemic or Ontological? Since you say explanation I take it you mean the former? Note how hierarchy becomes a stand in for temporal ordering in your statements. Does explanatory direction map temporal direction and/or can it do so well, if [ontologically] emergent structures may causally effect their operational bases [as in a LISP program to some extent] which makes temporal asymmetries and casual explanations REAL difficult [see Maturana and Verala, L & L, or the work of Michael Conrad]. I don't think CD is refuted unless we identify it exclusively with one-to one-to one maps.


>>I comment that "reduction" is not elimination: no one argues that there are

no proteins just because or if we can esxplain their behavior in biochemical

terms.

I don't see why any of this discussion has any implicatiosn for "bourgeois"

thought one way or the other.

--jks

*********

Well, it furthers Biology's incorporation into "Big Science" and may continue to "fuel" the backlash against the way capitalism is appropriating the biosphere for commodity production and that will have implications. Other than that, the nature-nurture spectrum shifts back towards the latter for the time being.

Ian



More information about the lbo-talk mailing list