Under Darwin’s Cosh? Neo-Aristotelian Thinking in Environmental Ethics

ions, as averages over populations of individuals) are not. Sober, rightly in my view, rejects both halves of this claim. (See Sober, ‘Evolution, Population Thinking, and Essentialism’, op. cit. note 18.) In the present context Mayr’s questionable metaphysical picture need not be a cause for concern. His key insight concerning the explanatory priority of individual variation in population thinking does not depend on the dubious metaphysical window-dressing he supplies, and so may be formulated without it (which is what I have endeavoured to do in the main text). I take it that this is Sober’s view also. 23 R. Lewontin, ‘The Organism as the Subject and Object of Evolution’, Scientia 118 (1983), 63-82. developmental temperature is as high as 30° centigrade. And things get more complicated once we allow, in addition, variations in the genotype, and consider the ensuing pattern of interactions with the relevant environmental factor. For example, Drosophila with a mutation known as Ultrabar always end up with less visual receptors than those with wild genotype. The same is true of Drosophila with a different mutation, Infrabar. However, the two mutant genotypes have opposite relations to temperature, such that the number of receptors possessed by Ultrabar flies decreases with developmental temperature, while the number possessed by Infrabar flies increases. In fact, if we make two plots of the number of light receptors against developmental temperature, one for Ultrabar and one for Infrabar (more on this idea in a moment), the two curves will cross over. How is this developmental space to be conceptualized? Let’s begin with the natural state model. According to the strict interpretation of this model, there will be a unique number of light receptors that constitutes the natural phenotypic outcome for insects of this species, although interfering forces during morphogenesis may well mean that this number is often not realized. (In a more relaxed frame of mind, we might allow that the relevant natural state may be specified in a mildly disjunctive way, such that, for example, the natural state will be realized if the number of light-receptors takes any one from a limited, small range of values. This does not alter the fundamental character of the explanation, so, for ease of exposition, I shall continue to work with the strict interpretation.) Each of the mutation-driven, temperature-driven, or interactive variations in phenotypic form that we identified in the data above needs to be characterized as a deviation from some natural state—the natural phenotype. The most likely candidate for the natural phenotype is a compound eye with 1000 light-receptor cells (or some appropriately relaxed take on that phenotype). However, this is not the only option. There is no requirement in the natural state model that the privileged phenotype be statistically the most common. Now let’s turn to the approach recommended by population Under Darwin’s Cosh? 277 24 My brief analysis of Drosophila morphogenesis that follows is, in essence, the local application of a general theoretical analysis, advanced by Sober, of the different ways in which natural state thinking and population thinking approach development; see his ‘Evolution, Population Thinking, and Essentialism’, op. cit. note 18. In that paper Sober considers, only to reject, a number of different moves designed to reduce the tension between the natural state model and population thinking. biology. The population geneticist will appeal to the concept of a norm of reaction. We’ve just seen this idea at work. A norm of reaction is a curve generated by taking a particular genotype, and plotting changes in a phenotypic trait of interest (in our example, the number of receptors) against an environmental variable (in our example, the developmental temperature). In effect, a norm of reaction shows how an organism of a particular genotype would develop in different environments. So one might conceptualize our fruit-fly developmental space in terms of a set of norms of reaction. This way of thinking enshrines individual variation at the root of biological nature. Each norm of reaction identifies a range of possible developmental outcomes for a particular genotype. Moreover, there is a deep sense in which, in terms of our understanding of the fundamental character of biological systems, each of these outcomes, and each of the outcomes for each of the different possible genotypes, is conceptualized as being on an equal footing. Of course, it may be true to say of the fruit-fly not only (a) that there is a wild genotype, but also (b) that in its ordinary developmental ecology, the temperature is regularly within a small range of values. This might explain why the number of lightreceptor cells in the Drosophila compound eye is usually about 1000. Nevertheless this situation, riddled as it is with statistical and environment-relative contingency, seems to fall short of establishing the dual presence of a uniquely privileged developmental outcome and an associated tendency for the organism in question to realize that outcome—the kind of constrained developmental profile that the natural state model requires. These apparent problems with the natural state model reverberate into environmental ethics. If the third Darwinian contribution identified by Mayr is on the mark, and the base-line of biological nature really is that actual organisms are, at root, no more than points on a vast landscape of phenotypic diversity, rather than enforced offshoots from a path that leads to a preferred speciesspecific destination, then it is hard to give any conceptual weight to the idea that in perturbing the developmental trajectory of an organism, we are preventing it from realizing its natural state. Any philosophical strategy for specifying ethical norms that rests on that idea is thereby undermined; and that includes Taylor’s biocentric individualism. But have we got the base-line right? Our first Michael Wheeler 278 25 As mentioned above, in his paper ‘Philosophical Problems for Environmentalism’ (op. cit. note 18, 233–40), Sober traces certain difficulties facing some environmentalist positions to their implicit adoption of the natural state model. Sober’s target is the very general claim, plausibly flirtation with contemporary developmental biology certainly suggests that we have; but perhaps all is not as it seems. 4. Kick-Starting Aristotelianism I now want to suggest that we have been moving too fast, and that there is, in truth, growing support in contemporary biological science for something which looks very much like an Aristotelian natural state model of organismic development. Self-organization is a phenomenon that is now recognized as being widespread in nature—and that includes human nature. Indeed, it appears that wherever we look (e.g. at chemical reactions, lasers, slime moulds, foraging by ants, flocking behaviour in creatures such as birds, Under Darwin’s Cosh? 279 at work in a number of environmentalist positions, that what is morally reprehensible about an action that frustrates an organism’s endogenous developmental tendency to reach its natural state is that any such action places the organism concerned in an unnatural state. As Sober points out, once development is conceptualized on the population biology model, the idea that any one phenotype is the only natural one is deeply problematic. The worry about neo-Aristotelian environmentalism that I present here clearly reprises Sober’s critique in certain respects, although I have endeavoured to add fuel to the fire by showing in detail exactly how that natural state model underlies the detailed neo-Aristotelian structure of one prominent environmental-ethical framework. More importantly, as we shall see, I think the natural state model lives to fight another day, whereas Sober doesn’t. 26 I am not the only person to have claimed recently that modern biological science is inadvertently rediscovering supposedly discarded Aristotelian concepts and principles. For example, Denis Walsh has been arguing that contemporary evolutionary developmental biology explains why organisms have the particular phenotypes they do (and in particular, the organismal capacities that underlie the evolvability of organismal lineages) by appealing to a reciprocal relation between the goal-directed plasticity of organisms and the causal powers of their underlying developmental systems. According to Walsh, this reciprocal arrangement maps onto, and, in the end, plays the same fundamental explanatory role as, the kind of interactive unity between a biological form and its realizing matter that constitutes an Aristotelian organismal nature. See D. Walsh, ‘Evolutionary Essentialism’, unpublished conference paper given at Teleology, Ancient and Modern, University of Edinburgh, 16–18 August 2004. Although the analysis that follows in this paper exploits different aspects of Aristotelian philosophy of biology and of contemporary developmental biology, it is clearly an overlapping and complementary