ANTI-ESSENTIALISM: A NATURALIST SUCCESS STORY?
Even before we had the beginnings of a theory of the origin of species, some preDarwinian naturalists (in the older sense) recognized the existence of species and sought to understand their properties and relations. From a certain vantage point, such questions might be thought to involve metaphysical speculation from the get-go. What are species ? Into what ontological category do they fit ? The development of evolutionary biology from Darwin’s era into the early twentieth century did little to settle this question explicitly, at least in the minds of (even naturalistically inclined) philosophers. Quine, for instance, addresses the question of species’ ontological category en
passant in his famous discussion of ontological commitment: “[W]hen we say that some zoological species are cross-fertile, we are committing ourselves to recognizing as entities the several species themselves, abstract though they are” (1953, 13; my emphasis). An obvious way of implementing this basic thought, of course, involves treating species as natural kinds. While Quine himself doubted that natural kinds would play any enduring role in mature sciences (1969), his successors were generally inclined to view natural kinds as key targets of scientific investigation (Kornblith 1993). And to many philosophers, species seemed like paradigmatic natural kinds in precisely this sense— along with such kinds as gold, water, electrons, and so on. Perhaps inspired by the growing sophistication of molecular biological techniques at the time, Kripke (1980) and Putnam (1975a, 1975b) famously attempted to subsume biological species under the essentialist model, suggesting that, like the physical and chemical paradigms, species would also be characterized by underlying micro-structural essences. More robust arguments for this essentialist interpretation followed over the next few decades (see, e.g., Caplan 1981; Wilkerson 1995; Devitt 2008).
But this nascent metaphysics of species—that they are natural kinds characterized by essences—did not remain in favor for long. Several influential evolutionary biologists had already been campaigning against treating species as abstract types in either a Platonic or Aristotelian conception (see, e.g., Mayr 1959, 28-29; Ghiselin 1966, 1974). Biologically sophisticated philosophers gradually took up the thread and pronounced the abject failure of natural kind essentialism as applied to biological species. Elliott Sober wrote in a well-known article that “[e]ssentialism about species is today a dead issue, not because there is no conceivable way to defend it, but because the way in which it was defended by biologists was thoroughly discredited” (1980, 353), and because it was a poor explanation of the patterns of variation we observe. Others focused less on the manner of essentialism’s defense than on the empirical falsity of its core commitments: that for each species, there is some underlying microstructural (presumably genetic) property shared by all and only its members that makes them the sorts of things they are. Increasing understanding of the nature of development and the extent of intraspecific genetic variation rendered these claims significantly less plausible.
So we have here what seems to be a pretty obvious victory for the naturalist. A metaphysics of species is proposed and subsequently falsified by empirical investigation. Of course, the story is a bit more complicated than this brief gloss lets on—but perhaps only a bit. For one, the historical narrative offered by Mayr (and taken up by many philosophers of biology) concerning the development of essentialism has been questioned (Winsor 2003, 2006). For two, Sober’s pronouncement of essentialism’s death might have been, as they say, greatly (or anyway somewhat) exaggerated (Devitt 2008). Essentialism about biological taxa has persisted since the
1980s in two forms that we might call historical and (tendentiously) bullet-biting. The former involved abandoning one of the standard tenets of essentialism: that a natural kind’s essence must be an intrinsic property. Perhaps species kinds could be seen as defined by historical essences—for example, by their position on the tree of life in relation to other taxa (Griffiths 1999; LaPorte 2004; Williams 2011).
Let us consider the bullet-biting strategy for defending essentialism first. One might pursue it in two different ways. First, one could propose simply revising the boundaries of the relevant kinds so that they could be characterized by intrinsic essences (e.g., Wilkerson 1995, 132). Second, one might leave recognized taxa alone and argue that they just must have essences, even if so far unidentified and maybe also even if there is specific reason for doubting that there are such properties. Devitt, for instance, argues from the existence of lawlike generalizations about species (e.g., “All Indian rhinos have a single horn”) that these generalizations must be explained by intrinsic biological essences. Though this gambit leaves biological classification more or less unmolested, it does seem to entail that key pieces of biological practice—the way in which systematists have gone about describing species and their relations—are seriously misguided. Devitt’s argument would, at the very least, imply that biologists have been systematically ignoring a key explanatory burden in biology (cf. Lewens 2012b, 752). This alone might suggest to the naturalistically inclined metaphysician that something must be amiss in Devitt’s arguments (or in the other bullet-biting strategies). Contemporary biological practice, like most cases of scientific practice, is motivated by its record of bearing fruit (or promise for doing so) on matters deemed significant by the community pursuing it. As such, it should not be abandoned (or significantly modified) lightly.
Contrast the historical essence strategy mentioned above. In that case, it was (interpreted as) standard practice of evolutionary biologists to treat species as historically defined kinds so that, as Griffiths puts it, “nothing that does not share the historical origin of [a given] kind can be a member of [that] kind” (1999, 219). Though this approach does less violence to large portions of biological practice—effectively leaving alone cladistically inclined classification schemes—it does apparently foreclose on some apparently live options about how to characterize species. Suppose one was interested in pursuing species concepts based on shared gene clusters (Mallet 1995), interbreeding (Mayr 1963; Coyne and Orr 2004), or phenetic similarity
(Sokal and Sneath 1961; Lewens 2012a). The restriction on sharing a historical origin Griffiths posits would have to be grafted onto these species concepts artificially and without internal scientific motivation. This is analogous to the awkward situation Callender discusses concerning the “strange new laws” some metaphysics are forced to postulate in order to accommodate their metaphysical views about “extended simples” to findings in theoretical physics (2011, 39).
So a comparatively mild naturalistic rebuke seems appropriate here as well. To those with even slightly heterodox or pluralist approaches on the species problem, historical essentialism (biological motivation notwithstanding) overreaches. Of course, this rebuke may not move those antecedently committed to classification strategies that already incorporate Griffiths’ stricture. This raises a question about what degree of scientific ecumenicalism a naturalized metaphysics should allow. But this is a question I must leave aside here. For now, let us suppose that, a few complications aside, the anti-essentialist consensus in the philosophy of biology is a case of due attention to the relevant science foreclosing on otherwise appealing metaphysical approaches. I want to now turn toward two other cases—one positive, one negative—that do not transfer quite as cleanly. We start with the positive: the claim that species are individual objects with organisms as their parts (rather than natural kinds). We will then consider a criticism of an alternative conception of species as natural kinds.
-  Though Devitt offers some brief comments concerning the use of genome sequencing projects in revealing theintrinsic essences of various taxa—suggesting that he is concerned with biological practice—one does not typically observe biologists discussing these projects in terms of the discovery of species’ essence.
-  Devitt’s arguments may of course fail on their own terms, as I have argued they do (2013, §3.3); see alsoBarker 2010.