Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
History of genetic engineering wikipedia , lookup
Inbreeding avoidance wikipedia , lookup
Medical genetics wikipedia , lookup
Genetic drift wikipedia , lookup
Polymorphism (biology) wikipedia , lookup
The Selfish Gene wikipedia , lookup
Koinophilia wikipedia , lookup
Group selection wikipedia , lookup
Population genetics wikipedia , lookup
1 From E.F. Keller, “Language and Ideology in Evolutionary Theory: Reading Cultural Norms into Natural Law” THE PROBLEM OF SEXUAL REPRODUCTION In much of the discourse on reproduction, it is common to speak of the 'reproduction of an organism'—as if reproduction is something an individual organism does; as if an organism makes copies of itself, by itself. Strictly speaking, of course, such language is appropriate only to asexually reproducing populations since, as every biologist knows, sexually reproducing organisms neither produce copies of themselves nor produce other organisms by themselves. It is a striking fact, however, that the language of individual reproduction, including such correlative terms as an individual's offspring and lineage, is used throughout population biology6 to apply indiscriminately to both sexually and asexually reproducing populations. While it would be absurd to suggest that users of such language are actually confused about the nature of reproduction in the organisms they study (e.g. calculations of numbers of offspring per organism are always appropriately adjusted to take the mode of reproduction into account), we might none the less ask, what functions, both positive and negative, does such manifestly peculiar language serve? And what consequences does it have for the shape of the theory in which it is embedded? I want to suggest, first, that this language, far from being inconsequential, provides crucial conceptual support for the individualist programme in evolutionary theory. In particular, my claim is that the starting assumption of this programme—that is, that individual properties are primary—depends on the language of individual reproduction for its basic credibility.7 In addition, I would argue that, just as we saw with the language of competition, the language of individual reproduction, maintained as it is by certain methodological conventions, both blocks the perception of problems in the evolutionary project as presently conducted and, simultaneously, impedes efforts to redress those difficulties that can be identified. The problems posed for evolutionary theory by sexual reproduction and Mendelian genetics are hardly new, and indeed, the basic theory of population genetics originates in the formulation of a particular method (i.e. the Hardy-Weinberg calculus) designed to solve these problems. The Hardy-Weinberg calculus (often referred to as 'bean-bag' genetics) invoked an obviously highly idealized representation of the relation between genes, organisms, and reproduction, but it was one that accomplished a great deal. Most important, it provided a remarkably simple recipe for mediating between individuals and populations—a recipe that apparently succeeded in preserving the individualist focus of the evolutionists' programme. One might even say that it did so, perhaps somewhat paradoxically, by tacitly discounting individual organisms and their troublesome mode of reproduction. With the shift of attention from populations of organisms to well-mixed, effectively infinite, pools of genes, the gap between individual and population closed. Individual organisms, in this picture, could be thought of as mere bags of genes (anticipating Richard Dawkins's 'survival machines' (1976: 21))—the end-product of a reproductive process now reduced to genetic replication plus the random mating of gametes. Effectively bypassed with this representation were all the problems entailed by sexual difference, by the contingencies of mating and fertilization that resulted fr0111 the finitude of actual populations and, simultaneously, all the ambiguities of the term reproduction as applied 2 to organisms that neither make copies of themselves nor reproduce by themselves. In short, the Hardy-Weinberg calculus provided a recipe for dealing with reproduction that left undisturbed— indeed, finally, reinforced—the temptation to think (and to speak) about reproduction as simply an individual process, to the extent, that is, that it was thought or spoken about at all. In the subsequent incorporation of the effects of natural selection into the Hardy-Weinberg model, for most authors in population genetics, the contribution of reproduction to natural selection fell largely by the wayside. True, the basic calculus provided a ready way to incorporate at least part of the reproductive process, namely, the production of gametes; but in practice, the theoretical (and verbal) convention that came to prevail in traditional population genetics was to equate natural selection with differential survival and ignore fertility altogether. In other words, the Hardy-Weinberg calculus seems to have invited not one but two kinds of elision from natural selection—first, of all those complications incurred by sex and the contingency of mating (these, if considered at all, get shunted off under the label of sexual, rather than natural, selection),8 and second, more obliquely, of reproduction in toto. I want to suggest that these two different kinds of elision in fact provided important tacit support for each other. In the first case, the representation of reproduction as gametic production invited confidence in the assumption that, for calculating changes in gene frequency, differential reproduction, or fertility, was like differential survival and hence did not require separate treatment. And in the second case, the technical equation of natural selection with differential survival which prevailed for so many years, in turn, served to deflect attention away from the substantive difficulties invoked in representing reproduction as an individual process. The net effect has been to establish a circle of confidence, first, in the adequacy of the assumption that, despite the mechanics of Mendelianism, the individual remains both the subject and object °f reproduction, and second, in the adequacy of the metonymic collapse of reproduction and survival in discussions of natural Election. The more obvious cost of this circle surely comes from its second part. As a number of authors have recently begun to remind us, the equation between natural selection and differential survival fosters both the theoretical omission and the experimental neglect of a crucial component of natural selection. Perhaps even more serious is the cost in unresolved difficulties that this equation has helped obscure. One such difficulty is the persistence of a chronic confusion between two definitions of individual fitness: one, the (average) net contribution of an individual of a particular genotype to the next generation, and the other, the geometric rate of increase of that particular genotype. The first refers to the contribution an individual makes to reproduction, while the second refers to the rate of production of individuals. In other words, the first definition refers to the role of the individual as subject of reproduction and the second to its role as object. The disparity between the two derives from the basic fact that, for sexually reproducing organisms, the rate at which individuals of a particular genotype are born is a fundamentally different quantity from the rate at which individuals of that genotype give birth—a distinction easily lost in a language that assigns the same term, birth rate, to both processes. Beginning in 1962, a number of authors have attempted to call attention to this confusion (Moran 1962; Charlesworth 1970; Pollak and Kempthorne 1971; Denniston 1978), agreeing that one definition—the contribution a particular genotype makes to the next generation's population—is both conventional and correct, while the other (the rate at which individuals of a 3 particular genotype are born) is not. Despite their efforts, however, the confusion persists.9 In part, this is because their remains a real question as to what 'correct' means in this context or more precisely, as to which definition is better suited to the needs that the concept of fitness is intended to serve—in particular, the need to explain changes in the genotypic composition of populations. Given that need, we want to know not only which genotypes produce more but also the relative rate of increase of a particular genotype over the course of generations. Not surprisingly, conflation of the two definitions of fitness is , particularly likely to occur in attempts to establish a formal connection between the models of population genetics and those of mathematical ecology. Because all the standard models for population growth assume asexual reproduction, the two formalisms actually refer to two completely different kinds of populations: one of gametic pools and the other of asexually reproducing organisms. , In attempting to reconcile these two theories, such a conflation is in fact required to finesse the logical gap between them. A more adequate reconciliation of the two formalisms requires the introduction of both the dynamics of sexual reproduction into mathematical ecology and a compatible representation of those dynamics into population genetics. Counter intuitively, it is probably the second—the inclusion (in population genetics models) of fertility as a property of the mating type—that calls for the more substantive conceptual shifts. Over the last twenty years, we have witnessed the emergence of a considerable literature devoted to the analysis of fertility selection—leading at least some authors to the conclusion that 'the classical concept of individual fitness is insufficient to account for the action of natural selection' (Christiansen 1983: 75). The basic point is that when fertility selection is included in natural selection, the fitness of a genotype, like the fitness of a gene (as argued by Sober and Lewontin 1982), is seen to depend on the context in which it finds itself. Now, however, the context is one determined by the genotype of the mating partner rather than by the complementary allele. A casual reading of the literature on fertility selection might suggest that the mating pair would be a more appropriate unit of selection than the individual, but the fact is that mating pairs do not reproduce themselves any more than do individual genotypes. As E. Pollak has pointed out, 'even if a superior mating produces offspring with a potential for entering a superior mating, the realization of this potential is dependent upon the structure of the population' (1978: 389). In other words, in computing the contribution of either a genotype or a mating pair to the next generation's population (of genotypes or mating pairs), it is necessary to take account of the contingency of mating: such a factor, measuring the probability that any particular organism will actually mate, incurs a frequency dependence that reflects the dependence of mating on the genotypic composition of the entire population. Very briefly, the inclusion of a full account of reproduction in evolutionary theory necessitates the conclusion that natural selection operates simultaneously on many levels (gene, organism, mating pair, and group), not just under special circumstances, as others have argued, but as a rule. For sexually reproducing organisms, fitness in general is not an individual property but a composite of the entire interbreeding population, including, but certainly not determined by, genie, genotypic, and mating pair contributions. By undermining the reproductive autonomy of the individual organism, the advent of sex undermines the possibility of locating the causal efficacy of evolutionary change in individual properties. At least part of the 'causal engine' of natural selection must be seen as distributed throughout the entire population of interbreeding 4 organisms. My point is not merely to argue against the adequacy of the individualist programme in evolutionary theory but—like the point of my earlier remarks about competition—to illustrate a quite general process by which the particular conventions of language employed by a scientific community permit a tacit incorporation of ideology into scientific theory and, simultaneously, protect participants from recognition of such ideological influences. The net effect is to insulate the theoretical structure from substantive critical revision. In discussions of sexual reproduction, the linguistic conventions of individual reproduction—conventions embodying an ideological commitment to the a priori autonomy of the individual—both perpetuate that belief and promote its incorporation into the theory of evolutionary dynamics. At the same time, the conventional equation between natural selection and differential survival has served to protect evolutionary theory from problems introduced by sexual reproduction, thereby lending at least tacit support to the assumption of individual autonomy that gave rise to the language of individual reproduction in the first place. The result—now both of the language of autonomy and the language of competition—is to effectively exclude from the domain of theory those biological phenomena that do not fit (or even worse, threaten to undermine) the ideological commitments that are unspoken yet in language, built into science by the language we use in both constructing and applying our theories. In this way, through our inescapable reliance on language, even the most ardent efforts to rid natural law of cultural norms become subverted, and the machinery of life takes on not so much a life of its own as a life of our own. But then again, what other life could it have? Notes 6. Including both population genetics and mathematical ecology. 7. For example, in the absence of other organisms, the fitness of a sexually reproducing organism is, strictly speaking, zero. 8. Darwin originally introduced the idea of sexual selection—always in clear contradistinction to natural selection—in an effort to take account of at least certain aspects of reproductive selection. ... In my view, the recent interest in sexual selection among sociobiologists is a direct consequence of the final, and complete, abandonment of the individual organism as a theoretical entity. Genetic selection theories, it could be said, complete the shift of attention away from organisms begun by the Hardy-Weinberg calculus. Sexual reproduction is a problem in this discourse only to the extent that individual organisms remain, somewhere, an important (even if shifting) focus on conceptual interest. 9. See Keller (1987) for details. References Axelrod, Robert (1984), The Evolution of Cooperation (New York: Basic Books). Bernstein, H., H. C. Byerly, F. A. Hopf, R. A. Michod, and G. K. Vemulapalli (1983), 'The Darwinian Dynamic', The Quarterly Review of Biology, 58,185-207. Birch, L. C. (1957), 'Meanings of Competition', American Naturalist, 91, 5-18. Boucher, Douglas, (1985), The Biology of Mutualism, (Oxford: Oxford University Press). Charlesworth, B. (1970), 'Selection in Populations with Overlapping Generations, I, The Use of Malthusian Parameters in Population Genetics', Theoretical Population Biology, 1/3,352-70. 5 Christiansen, F. B. (1984), 'The Definition and Measurement of Fitness', in B. Shorrocks, (ed.). Evolutionary Ecology: B. E. S. Symposium 23 (Oxford and Boston: Blackwell Scientific Publications) 65-79. Colinvaux, Paul (1978), Why Big Fierce Animals are Rare (Princeton: Princeton University Press). Dawkins, Richard (1976), The Selfish Gene (Oxford: Oxford University Press). Denniston, C. (1978), 'An Incorrect Definition of Fitness Revisited', Annals of Human Genetics, Land. 42, 77-85. Gause, G. F., and A. A. Witt (1935), 'Behavior of Mixed Populations and the Problem of Natural Selection', American Naturalist, 69, 596-609. Ghiselin, Michael (1974), The Economy of Nature and the Evolution of Sex (Berkeley, Los Angeles, London: University of California Press). Keller, Evelyn Fox (1987), 'Reproduction and the Central Project of Evolutionary Theory', Biology and Philosophy 2, 73-86. —(1988), 'Demarcating Public from Private Values in Evolutionary Discourse', Journal of the History of Biology, 21/2,195-211. Lewontin, Richard (1970), The Units of Selection', Annual Review of Ecology and Systematics, 1-1B. —(1982), 'Organism and Environment', in E. H. C. Plotkin (ed.), Learning, Development, and Culture (New York: John Wiley and Sons). Mayr, Ernst (1963), Animal Species and Evolution (Cambridge: Harvard University Press). Midgley, Mary (1985), Evolution as a. Religion (London and New York Methuen). Monod, Jacques (1972), Chance and Necessity (New York: Random House). Moran, P. A. P. (1962), 'On the Nonexistence of Adaptive Topographies', Annals of Human Genetics, 27, 383-93. Pollak, E. (1978), 'With Selection for Fecundity the Mean Fitness Does Not Necessarily Increase', Genetics, 90, 383-9. —and 0. Kempthorne (1971), 'Malthusian Parameters in Genetic Populations, II, Random Mating Populations in Infinite Habitats', Theoretical Population Biology, 2, 357-90. Sober, E., and R. Lewontin (1982), 'Artifact, Cause, and Genetic Selection', Philosophy of Science, 47,157-80. Tennyson, Alfred, In Memoriam, LV-LVI. Weinberg, Steven (1974), 'Reflections of a Working Scientist', Daedalus,103 (Summer) 33-46. Williams, George (1986), 'Comments', Biology and Philosophy, 1/1' 114-22.