Beyond the Selfish Gene: Against Genetic Reductionism By Brian Simpson

In my pile of scientific photocopies I have Peter Godfrey-Smith, “Genes Do Not Encode Information for Phenotypic Traits,” in C. Hitchcock (ed.), Contemporary Debates in Philosophy of Science, (Blackwell, 2004), pp. 275-289. By “phenotypic traits,” it is understood the structure, size, shape, colour and behaviour patterns of an organism. These are taken by the likes of sociobiologists such as Richard Dawkins, to be caused by genes, in the sense that genes have a informational role in the production of the phenotype of the whole organism.

First, in terms of Shannon’s information theory, information exists where a reduction of uncertainty at one place correlates with a reduction of uncertainty at another (p. 277) In this sense, the presence of a gene g1 could be correlated with some phenotypical trait P1,  but, the situation can be turned around, looking at correlations between environmental conditions and phenotypical traits, so that the phenotype could be seen as carrying information about the genes. So, this does not deliver the conclusions the genetic reductionist requires.

Genes do not produce some coded message that specifies phenotypic traits. Genes make protein molecules, and regulate other genes. There are complexities in involved here, which we can omit, What is coded for “is the primary structure of a protein molecule; not even higher-level protein structure is coded for.” It is a long way from that to the genetic determinist claim that “selfish genes” are some sort of central directing agency, since the regulation of gene protein synthesis also depends upon other activities in the organism.

The critique of genetic reductionism/determinism was extended by the school of process structuralists, inspired by the rational morphologists who preceded Darwin. This position holds that organism development needs to be explained by “generative laws” and not by the mechanisms of natural selection and random genetic mutations. See: G. Webster and B. Goodwin, “The Origin of Species: A Structuralist approach,” Journal of Social and Biological Structures, vol. 5, 1982, pp. 15-47.

Others have proclaimed that the paradigm of genetic determinism is in crisis failing to actually show how complex behaviour actually arise from genetic causes: R. C. Strohman, “The Coming Kuhnian Revolution in Biology,” Nature Biotechnology, vol. 15, 1997, pp. 194-200.

“Conflicts within genetic determinism Genetic determinism is in trouble for the following reasons. Anomalous findings, the first signs of a Kuhnian revolt, are showing up almost weekly in major journals of molecular and cell biology. Among the disturbing findings coming from the Human Genome Project and elsewhere are that genome complexity found in humans and mice, for example, is not correlated with the differences of form and function found between them'. Sequence comparisons between different species are not always informative; nevertheless, there are many examples in which no, or little, correlation exists between genetic and morphological complexity. Humans and mice have the same estimated number of expressed genes (exclusive of so-called junk DNA), and yet they are radically different creatures. Older findings have revealed the similarity ( 98%+) between human and chimpanzee DNA, and yet these two organisms manage to construct very different results from their nearly identical genes. Still older evidence dealing with "sibling species" shows us organisms that appear to be identical under the most stringent anatomical observation, and yet are found to be entirely different when examined at the level of their genes and of their proteins. These organisms offer the reverse of the human-chimp problem, because they extract from extremely different genomes identical phenotypic end points. Molecular biologists often counter these findings with the observation that mice and humans are not really that much different after all-it depends on how closely one looks. True, in mice and humans, most molecules are identical; even at cellular and physiological levels of observation, there are enormous similarities. And that is precisely the problem. The human organism in development distills from identical genomes over 200 different behaviors in the form of 200+ cell types, and from similar genomes it distills 2 species of remarkably different overall behavior. The question is, what are the constraints-the boundary conditions-the developmental rules that serve to organize these similarities into different constructions. We cannot assume, as genetic determinism does, that constraints and developmental rules are all in the form of genetic programs. Most biologists, when pushed, do agree that there is not enough information in any genome capable of mapping out the details by which morphological structures arise in organisms'. There is a striking lack of correspondence between genetic and evolutionary change. Neo-Darwinian theory predicts a steady, slow continuous, accumulation of mutations (microevolution) that produces a progressive change in morphology leading to new species, genera, and so on (macroevolution). But macroevolution now appears to be full of discontinuities (punctuated evolution), so we have a mismatch of some importance. That is, the fossil record shows mostly stasis, or lack of change, in a species for many millions of years; there is no evidence there for gradual change even though, in theory, there must be a gradual accumulation of mutations at the micro level. Molecular and standard evolutionary biologists argue, for a variety of reasons, that this really presents no problems and that effects such as genetic drift in small populations can produce new species gradually without being noticed in the fossil record. But there are serious internal debates going on over this issue and generally on the issue of mutation rates. At the moment, we can only conclude that the neo-Darwinian model of evolution, based on a gradual accumulation of point mutations, appears to be incomplete at best. It is not, of course, that one doubts that evolution has occurred. The theory is in trouble because it insists on locating the driving force solely in random mutations. An alternative theory of evolution that emphasizes the importance of nonrandom (epigenetic) changes during development could explain the problems now being encountered by evolutionary theory. Developmental change as a source for the creation of new form equal to the "creativity" of a random natural selection is a scientific possibility of great merit. But, mostly because it is tinged with the element of "nonrandornness;' it appears to favor a religious creationism rather than the materialistic process that it actually is; it is therefore left mostly unexplored. There is a growing recognition that information for function may not be located solely in genomic databases'. That is, it is becoming clear that sequence information in DNA, by itself, contains insufficient information for determining how gene products (proteins) interact to produce a mechanism of any kind. The reason is that the multicomponent complexes constructed from many proteins are themselves machines with rules of their own; rules not written in DNA … New approaches designed to get around this insufficiency use evolving technologies to target and change specific genes in specific stages of development so that, in this assembly line model of function, the details of a wrongly assembled machine can be identified. These approaches make dubious assumptions concerning similarity of genes and gene complexes in different species--between mice, fruit flies, and humans, for example, that are related to unique functions. But time and time again we find that genes associated with diseases of mice have no such association with those genes in humans; the reverse is also true.”

The evidence against genetic reductionism and determinism continues to grow, but still the paradigm continues. For how much longer?