The nothing machine

Evolutionary Genetics

十月 27, 2000

Richard Lewontin is one of the most respected and most ignored geneticists of his generation, in particular by the majority of his research colleagues who have dedicated this volume of essays to him. And it is not Lewontin the Marxist dialectical biologist, the scourge of cheap reductionist justification by which the biotech industry cajoles biology into grandiose genome projects and the like, but Lewontin the evolutionary geneticist who can read his way through a gene sequence and juggle the maths with the best of them.

In general, there are two disparate approaches to the nature of biological organisms and the reasons they are capable of evolving. On the one hand, there are the mathematical modellers of theoretical biology who would claim that they are doing the best they can in constructing an edifice by which we can understand the ways of the natural world. On the other, there is Lewontin (and several like-minded biologists) who insist that the models are based on simplifying assumptions that fail to get under the skin of the complex beast they strive to circumscribe.

Lewontin's sarcasm wrapped this dilemma in the following amusing metaphor in his classic book, The Genetics of the Evolutionary Process : "For many years population genetics was an immensely rich and powerful theory with virtually no suitable facts on which to operate. It was like a complex and exquisite machine, designed to process a raw material that no one had succeeded in mining. For the most part the machine was left to the engineers, forever tinkering, forever making improvements, in anticipation of the day when it would be called upon to carry out full production.

"Quite suddenly the situation has changed. The mother-lode has been tapped and facts in profusion have been poured into the hoppers of this theory machine. And from the other end has issued - nothing. The entire relationship between the theory and the facts needs to be considered."

What Lewontin meant in 1974, and continues to emphasise in his own contribution to this book, is that the living, breathing, reproducing individual organism (the phenotype), and not the gene, is the unit of all that happens in ecology and evolution. Hence, until we understand the specific, complex dynamics of the genetic and environmental interactions by which a unique phenotype emerges from its unique genotype, then we can never get a handle on the central mystery of life.

There is, as yet, an unbridgeable gap between what Lewontin calls "genotype space" and "phenotype space". Despite the far-reaching recent discoveries of the ways genes go about their business in the development of phenotypes, Lewontin rightly says that this gap cannot be bridged unless "environmental contingency of variation in development is considered". What is at stake here is the concept of contingency and the fight to return biology to the historical one-off process it is, not yielding to grand unifying theories and global assumptions. If human history has managed to ditch Marxism as a theoretical framework, then biological history has to ditch the search for the "universals" so beloved of "evolutionary psychologists" (nature) and "culturalists" (nurture). The irony of universals within a discipline whose raison d'etre is about genetic differences is nicely pointed out by John Maynard Smith.

As always, Lewontin finds the appropriate withering words: "The fitness of a single locus ripped from its interactive context is about as relevant to real problems of evolutionary genetics as the study of psychology of individuals isolated from their social context is to an understanding of man's socio-political evolution. In both cases context and interaction are not simply second-order effects to be superimposed on a primary monadic analysis. Context and interaction are of the essence."

Bizarrely, this last sentence is recognised by the editors as of sufficient fundamental importance to place it prominently on the dust cover, but the contributions fail to take up what should have been the main leitmotiv. Apart from some traditional nods to epistasis (two or more genes - one function) and pleiotropy (one gene - two or more functions), plus four essays on phenotypic evolution, there seems to be no serious recognition of Lewontin's central concern.

Be that as it may, for Lewontin "concentration on molecular genetic variation has depauperised the problematics of population genetics by marginalising studies of the phenotype". Perhaps to ensure that we do not become too swayed by this, co-editor Costas Krimbas gets in his introductory pre-emptive strike: "the impossibility of solving new questions created by the new sequence data might turn the field of evolutionary theory to a consideration of evolution only at the phenotypic level. But this will significantly depauperate the evolutionary theory." So, the fight is on between the dedicators and the dedicatee from the outset.

I am not sure why Krimbas is so pessimistic about the possibility of satisfactorily answering questions raised by DNA sequence variation. The book consists of some of the leading players in sequence analysis who consider themselves to be doing just that, albeit within the rather limited need to resolve the age-old existentialist angst between Darwinian pan-selectionists and neutralists. The former considers most variation to be the result of selection for past adaptive necessities. The latter considers most variation to be the result of stochastic fluctuations of neutral (non-adaptive) variants. Both camps happily ignore developing phenotypes and construct ever more sophisticated mathematical models, statistical tests and computer simulations to discriminate between the two. Indeed, there is a point, once neutrality is shown to be inoperative, at which all that moves can be explained by selection. Selection becomes such a powerful manipulative tool that, as Peter Medawar wrote, it is too good for its own good. Today we can pick and mix our way through balancing, directional, diversifying, background, density-dependent, frequency-dependent, lineage-dependent, fluctuating, sex-dependent, stabilising and many more modes of selection to arrive at the place we intend to arrive at. As Frank Kermode said about literary criticism: "We are all members of a secret society, of which the principal ritual is to speak well of Shakespeare." Substitute Darwin for Shakespeare and there we have it; not that Darwin would have been pleased by what goes by way of evidence in the maths and statistics. It is all very jolly, but all rather meaningless if we want to consider "context and interaction" in its fullest sense.

The best sections contain those essays that in essence ignore the selectionist-neutralist debate and get on with the business of unravelling what is going on outside the computer screen. The section on the origin of species is particularly instructive. Ever since Darwin, speciation has suffered from a surplus of theories and a paucity of facts not made any the easier by arguments over the definitions of species and the methods used to arrange them in trees of relationships. Judith Masters goes so far as to say that species concepts cannot be refuted by the discovery of counter-examples and therefore must be classified as "metaphysical statements". I would agree; although it might be easier to have said that there is only one tree of life and that the events occurring at each branch point were unique to themselves: that is the way the cookie crumbled. Setting this aside, Jerry Coyne and Allen Orr offer their expected pugnacious account of recent discoveries of "speciation" genes, as such. They start with the brazen question "Why are there species?", and end with the statement that "the distinctive feature of the genetics of speciation is epistasis". This might sound more like a concluding whimper than a bang, but the statement is loaded.

What Coyne and Orr are emphasising is the normality of a gene's activities in individuals of the same species, in contrast to its abnormality (leading to inviability and sterility) in hybrid individuals containing the genomes of two species. Normality and abnormality are achieved, not by the naked presence of single genes (no selfish gene naivety for these geneticists) but by the interactive (epistatic) context in which such genes find themselves. This is as close as the contributors get to Lewontin's insistence on "context and interaction".

So far so good, but the overall sadness is that all the essays (including that of Lewontin) are woefully thin on most of the key parameters that define modern biology. These revolve around the phenomena of modularity, redundancy and turnover in species genomes. The first two reveal that evolutionary innovation is achieved through new combinatorial permutations of promiscuous modules of genetic information, in particular in the regions of DNA that control the on-off switching of genes; and that new combinations can be stabilised as a result of the buffering effects of redundant parts. Turnover is all about a variety of ubiquitous and frequently operative DNA mechanisms, such as the ability of modules to jump from one controlling region to another or to convert one module to another, which ensure that selection and drift are not the only two horses in town for the long-term spread of genetic novelties.

Were the theoreticians to familiarise themselves with modern developments and embrace them in their models, we would have an intellectually exhilarating future. As this book stands, it comprehensively summarises, in the main, a world gone by.

Gabriel Dover is professor of evolutionary genetics, University of Leicester.

Evolutionary Genetics: From Molecules to Morphology

Editor - Rama S. Singh and Costas B. Krimbas
ISBN - 0 521 57123 5
Publisher - Cambridge University Press
Price - £60.00
Pages - 702

请先注册再继续

为何要注册?

  • 注册是免费的,而且十分便捷
  • 注册成功后,您每月可免费阅读3篇文章
  • 订阅我们的邮件
注册
Please 登录 or 注册 to read this article.