1. Introduction

The idea that biological natural kinds, particularly a species such as dogs, have intrinsic underlying natures is intuitively appealing. It has been shown to be widespread even among children (Keil Reference Kitcher1989). It was endorsed by a great philosopher, Aristotle. Under the influence of the logical positivists, Popper (Reference Quine1950), Quine (Reference Ruse and Ereshefsky1960), and others, it fell from philosophical favor in the twentieth century until revived by Hilary Putnam (Reference Rosenberg1975), Saul Kripke (Reference Mayr1980), and David Wiggins (Reference Wiggins1980). Many philosophers probably now take the view for granted. If so, they are right out of touch with biologists and, especially, philosophers of biology. For, the consensus among philosophers of biology, and a widespread view among biologists, is that this sort of “Aristotelian essentialism” is deeply wrong, reflecting “typological” thinking instead of the recommended “population” thinking (Sober Reference Stanford1980, 247–48). This essentialism is thought to arise from a naive and uninformed view of biology, indeed to be incompatible with Darwinism.Footnote ¹ This view is nicely presented and argued for in a paper by Samir Okasha (Reference Putnam2002). I shall take that as my main text. I shall defend intrinsic biological essentialism. I think that the children are right and the philosophers of biology wrong.Footnote ²

I start by saying something about essentialism in general and about the essentialism I shall defend in particular. A property P is an essential property of being an F if and only if (iff) anything is an F partly in virtue of having P. A property P is the essence of being an F iff anything is an F in virtue of having P. The essence of being F is the sum of its essential properties. Essences can be fully intrinsic; for example, the essence of being gold is having atomic number 79. Essences can be partly intrinsic and partly extrinsic and relational.Footnote ³ For example, the essence of being a pencil is partly being an instrument for writing, which an object has in virtue of its relation to human intentions, and partly having the sort of physical constitution that distinguishes it from a pen, which an object has intrinsically. Finally, essences can be fully relational and extrinsic; being Australian is probably an example because it seems that anything—Rupert Murdoch, Phar Lap (a horse), the Sydney Opera House, a bottle of Penfolds Grange, the expression “no worries mate,” and so on—can have the property provided it stands in the right relation to Australia.Footnote ⁴

The doctrine I want to defend, which I shall call Intrinsic Biological Essentialism, abbreviated sometimes to Essentialism, is that Linnaean taxa have essences that are, at least partly, intrinsic underlying properties. This calls for some clarification and comment.

i. By “Linnaean taxa” I mean kinds that are thought to fall under the biological categories in the Linnaean hierarchy: kingdoms, phyla, classes, orders, families, genera, species, and even subspecies (varieties).Footnote ⁵ I do not mean kinds like those of predators or parasites. And I do not mean the categories themselves. Essentialism is a thesis about what it is for an organism to be, say, a dog not a cat, not about what it is for, say, dogs to be a species not a genus. (This distinction is vital in dealing with what is thought to be the central objection to essentialism; see sec. 4.) The focus of my discussion will be on species, but, I emphasize, Essentialism covers kinds that fall under all the categories.

ii. I include the qualification “at least partly” because I shall not take issue with the consensus that a species is partly a historical entity.Footnote ⁶

iii. In sexual organisms the intrinsic underlying properties in question are to be found among the properties of zygotes; in asexual ones, among those of propagules and the like.Footnote ⁷ For most organisms the essential intrinsic properties are probably largely, although not entirely, genetic. Sometimes those properties may not be genetic at all but in “the architecture of chromosomes,” “developmental programs,” or whatever (Kitcher Reference Kitcher1984/2003, 123).Footnote ⁸ For convenience, I shall often write as if the essential intrinsic properties were simply genetic, but I emphasize that my Essentialism is not committed to this.

iv. Intrinsic Biological Essentialism would certainly be opposed by the consensus because of its commitment to intrinsic essences. But the consensus should not be opposed to biological essentialism in general because, as I am understanding essentialism, the consensus is that species have essences but these are extrinsic or relational. And Kim Sterelny and Paul Griffiths, in their excellent introduction to the philosophy of biology, Sex and Death, are explicitly not opposed to this sort of essentialism: “the essential properties that make a particular organism a platypus … are historical or relational” (Reference Webster and Goodwin1999, 186). Of course, the very term ‘essentialism’ has become so distasteful to biologists because of its association with Aristotelian metaphysics that a biologist would doubtless be reluctant to admit to any sort of essentialism. But the essentialism I have defined need not come with those Aristotelian trappings. Many philosophers would be similarly reluctant because the term ‘essentialism’ strikes them as quaintly old-fashioned, scholastic, even unscientific. But such reluctance would be a merely verbal matter. The issue of essentialism would remain even if the term ‘essentialism’ were dropped. It is the issue of in virtue of what, an organism is a member of a certain Linnaean taxon; the issue of what makes an organism a member of that taxon; the issue of the very nature of the taxon. I stick with ‘essentialism’ because it is the term that philosophers of biology use for the doctrine that they want to reject and I want to promote. Those who are offended by the term should replace it with one of the other ways of characterizing the issue.

v. There is some controversy over whether species are natural kinds or individuals. Michael Ghiselin (Reference Griffiths1974) and David Hull (Reference Keil1978) seem to see individualism as an antidote to essentialism.Footnote ⁹ But, as Okasha points out, “the issues about essentialism … do not depend on which view of the ontological status of species we favour” (Reference Putnam2002, 193–94; see also Kitcher Reference Kitts and Kitts1989, 137–40).Footnote ¹⁰ Thus, if a species is an individual rather than a kind, our essentialism issue for species becomes that of saying in virtue of what, organisms are parts of a certain species, for example, the species Canis familiaris. And the consensus answer should be that it is entirely in virtue of the organisms’ historical or relational properties, entirely “because they are part of the genealogical nexus” (Hull Reference Keil1978, 309), whereas my Essentialism's answer would be that it is, at least partly, in virtue of the organisms’ intrinsic underlying properties.Footnote ¹¹ Indeed the essentialism issue can be posed “nominalistically” in a way that is noncommittal on the ontological status of species: In virtue of what is an organism, say, a Canis familiaris? And that is how I do pose the issue in section 4. I mean to be neutral on the ontological issue but for convenience will mostly talk of species as if they were kinds.

vi. Essentialism is primarily concerned with the natures of the actual groups identified by the folk and biologists for explanatory purposes. This “descriptive” issue needs to be distinguished from the “normative” issue of the natures of the groups that we should identify for explanatory purposes. Clearly, we might not be doing what we should be doing. However, I shall write as if we are. If we are not, then my Essentialism should be taken to cover the groups that we should be identifying for explanatory purposes as well.

In section 2, I give evidence that the consensus really is opposed to Intrinsic Biological Essentialism. In section 3, I argue for the doctrine: explanations in biology demand that there be essential intrinsic underlying properties. In section 4, I summarize my responses elsewhere to objections (Devitt Reference Devitt2008). On the one hand, the consensus is that the standard relational “species concepts” make Essentialism untenable. On the other hand, some general features of Darwinianism, associated with variation and change, are thought to undermine Essentialism. I think that these objections fail.

2. Evidence of the Consensus

I have claimed that the consensus among philosophers of biology is that doctrines like Intrinsic Biological Essentialism are wrong. Among those philosophers, the claim hardly needs support because the consensus is so established. Still, among philosophers in general, the claim does need support because, influenced by Kripke and Putnam, many find the claim incredible and so think I must be struggling with a straw man. The epigraphs to this paper, drawn from the works of Elliott Sober and Paul Griffiths, two leading philosophers of biology, are some evidence that I am not. Here is some more.

The consensus starts by denying that members of a species share a distinctive set of genetic properties. Thus, according to Okasha, “virtually all philosophers of biology agree that … it simply is not true that the groups of organisms that working biologists treat as con-specific share a set of common morphological, physiological or genetic traits which set them off from other species” (Reference Putnam2002, 196).

Clearly, if the members of a species do not share a distinctive set of genetic properties, then those properties could not be essential properties of that species. Indeed, Okasha claims that “biologists and philosophers of biology typically regard essentialism about species as incompatible with modern Darwinian theory” (Reference Putnam2002, 191). And John Dupré claims that “it is widely recognized that Darwin's theory of evolution rendered untenable the classical essentialist conception of species” (Reference Dupré and Wilson1999, 3). Alex Rosenberg says that “the proponents of contemporary species definitions are all agreed that species have no essence” (Reference Sober and Ereshefsky1985, 203). Mohan Matthen claims that “species … are associated with no nonrelational real essence” (Reference Mayr and Ereshefsky1998, 115). Sober expresses this consensus as follows: “biologists do not think that species are defined in terms of phenotypic or genetic similarities”; tigers are “not defined by a set of traits” (Reference Sterelny and Wilson1993, 148). Sterelny and Griffiths put the point bluntly: “no intrinsic genotypic or phenotypic property is essential to being a member of a species” (Reference Webster and Goodwin1999, 186). Ghiselin puts it even more bluntly: “That John Doe has a particular set of genes is about as relevant to his being a specimen of Homo sapiens L. as it is to his working for the manufacturers of Brand X” (Reference Griffiths1974, 283).

Finally, if the essence of a species is not in the least intrinsic, then it must be entirely relational. I have already quoted Sterelny and Griffiths's claim to this effect about the platypus. And they think that nearly everyone agrees with them: there is “close to a consensus in thinking that species are identified by their histories” (Reference Webster and Goodwin1999, 8). Their view is endorsed by Okasha (Reference Putnam2002, 202). Sober declares that tigers are “historical entities” (Reference Sterelny and Wilson1993, 148). “Two organisms are conspecific in virtue of their historical connection to each other, not in virtue of their similarity” (150). Similarly, Marc Ereshefsky speaks for “Darwin, the founders of the Modern Synthesis, and most cladists” (Reference Ghiselin and Ereshefsky2001, 209). Finally, Hull claims, “If species are interpreted as historical entities, then particular organisms belong in a particular species because they are part of that genealogical nexus, not because they possess any essential traits. No species has an essence in this sense” (Reference Keil1978, 313). Ruth Millikan says much the same (Reference Okasha2000, 19).

The consensus is broad, but some are not part of it. Thus David B. Kitts and David J. Kitts (Reference Kripke1979) urge an intrinsic essentialism like mine. According to Richard Boyd (Reference Boyd and Wilson1999) and Robert Wilson (Reference Wilson and Wilson1999), species are “homeostatic cluster kinds,” and I take it that they think that they have at least partly intrinsic essences.Footnote ¹² And Philip Kitcher has this to say: “I want to remain agnostic on the issue of whether any species taxon has a nontrivial essence” (Reference Kitcher1984/2003, 132 n. 16).

I take the opposition to Intrinsic Biological Essentialism to be established. It is now time to argue for the doctrine.

3. An Argument for Intrinsic Biological Essentialism

I shall offer two reasons for believing Essentialism. The first is superficial but still, it seems to me, indicative of where the truth lies. Such essential properties seem to be part of what “genome projects” are discovering. The projects seem to be throwing light on the very nature of certain species. Thus the New York Times recently reported that researchers hope “to discover, from a three-way comparison of chimp, human and Neanderthal DNA, which genes have made humans human” (Wade Reference Wiggins2006, A14).Footnote ¹³ Philosophers of biology disparage this common view (Sterelny and Griffiths Reference Webster and Goodwin1999, 7; Okasha Reference Putnam2002, 197), but the view is certainly appealing.

The second reason is deep and shows why the view is appealing. We group organisms together under what seem, at least, to be the names of species or other taxa and make generalizations about the morphology, physiology, and behavior of the members of these groups: about what they look like, about what they eat, about where they live, about what they prey on and are prey to, about their signals, about their mating habits, and so on. These generalizations are the stuff of popular nature programs and are to be found throughout the writings of biologists and philosophers of biology. For example, we are told that ivy plants grow toward the sunlight (Sober Reference Sterelny and Wilson1993, 6); that polar bears have white fur; that Indian rhinoceri have one horn and African rhinoceri have two (21); that Hawaiian Drosophila “routinely form interspecific hybrids in the wild” (156); that the Australasian bittern is superbly camouflaged (Sterelny and Griffiths Reference Webster and Goodwin1999, 32); that “Major Mitchell cockatoos occasionally hybridize with galahs” (189); that “Australian trees … are not just drought-proof; they are fireproof as well” (203); that “magnetotactic bacteria … come equipped with little compasses called magnetosomes, which they use to navigate away from oxygen-rich surface water because oxygen is toxic to them” (209).

Generalizations of this kind demand an explanation. Why are they so? Why, for example, is there this difference between the Indian and African rhinos? Such questions could, of course, be seeking an explanation of the evolutionary history that led to the generalization being true. Set that aside for a moment. The questions could also be seeking an explanation of what makes the generalization true. Regardless of the history of its coming to be true, in virtue of what is it now true? What are the mechanisms? The truth of these generalizations cannot be brute facts about the world and so must be explained. Explanations will make some appeal to the environment, but they cannot appeal only to that.Footnote ¹⁴ There has to be something about the very nature of the group—a group that appears to be a species or taxon of some other sort—that, given its environment, determines the truth of the generalization. That something is an intrinsic underlying, probably largely genetic, property that is part of the essence of the group. Indeed, what else could it be?Footnote ¹⁵ Some intrinsic underlying property of each Indian rhino causes it, in its environment, to grow just one horn. A different such property of each African rhino causes it, in its environment, to grow two horns. The intrinsic difference explains the physiological difference. If we put together each intrinsic underlying property that similarly explains a similar generalization about a species, then we have the intrinsic part of its essence.Footnote ¹⁶

The generalizations we have been discussing reflect the fact that it is informative to know that an organism is a member of a certain species or other taxon: these classifications are “information stores” (Sterelny and Griffiths Reference Webster and Goodwin1999, 195). But being a member of a certain taxon is more than informative; it is explanatory. Matthen points out that “many biologists seem committed to the idea that something is striped because it is a tiger” (Reference Mayr and Ereshefsky1998, 115). And so they should be: the fact that an individual organism is a tiger, an Indian rhino, an ivy plant, or whatever explains a whole lot about its morphology, physiology, and behavior. At first sight, the explanation of the animal's stripes may seem rather superficial, but it is not really. For, when biologists group organisms together under some name on the basis of observed similarities, they do so partly on the assumption that those similarities are to be explained by some intrinsic underlying nature of the group. It seems clear to me that this is their practice, whatever they say about essentialism.Footnote ¹⁷ So the apparently superficial explanation points to the deep fact that there is something intrinsic, probably unknown, partly in virtue of which the animal is a tiger and that causes it to be striped. That something is an essential intrinsic property. The sum of those properties, together perhaps with some historical ones, constitutes the essence of a tiger. Sober rightly insists that the essence of a species must explain why its members are the way they are. It must be “a causal mechanism that acts on each member of the species, making it the kind of thing that it is” (Reference Stanford1980, 250). That is exactly what this (partly) intrinsic essence is.

I distinguished two sorts of explanation that might be sought in asking why members of a species have a certain property. In doing so, I am following in the footsteps of Ernst Mayr (Reference Mayr1961). He regards an explanation of the mechanisms within members of a species that make a generalization true (regardless of the history) as concerned with “proximate” causation and part of “functional biology.” In contrast, an explanation of the evolutionary history that led to the mechanism being present in the members of a species is concerned with “ultimate” causation and is part of “evolutionary biology.” The use of “ultimate” to describe the latter explanation seems like a gratuitous putdown of the former. Kitcher's evenhanded description in response to Mayr's is prima facie more appropriate: “there are indeed two kinds of biological investigation that can be carried out relatively independently of one another, neither of which has priority over the other” (Reference Kitcher1984/2003, 121). Adopting Kitcher's terminology, I shall call the former sort of explanation “structural” and the latter “historical.” Structural explanations, as he says, seek to “explain the properties of organisms by means of underlying structures and mechanisms.” He gives a nice example: “A biologist may be concerned to understand how, in a particular group of bivalve mollusks, the hinge always comes to a particular form. The explanation that is sought will describe the developmental process of hinge formation, tracing the final morphology to a sequence of tissue or cellular interactions, perhaps even identifying the stages in ontogeny at which different genes are expressed” (Reference Kitcher1984/2003, 121). He goes on to claim that “explanations of this type abound in biology: think of the mechanical accounts of normal (and abnormal) meiosis, of respiration and digestion, of details of physiological functioning in all kinds of plants and animals” (121). And my point is that, at bottom, structural explanations will advert to essential intrinsic, probably largely genetic, properties.Footnote ¹⁸ It is because the bivalve mollusks have a certain intrinsic underlying nature that the hinge takes that form. That is the deep explanation.Footnote ¹⁹

Sober claims that “evolutionary theory has removed the need for providing species with constituent definitions” and hence with intrinsic essences (Reference Stanford1980, 255). I suspect that this sort of focus on evolution, hence on historical rather that structural explanations, has misled biologists and philosophers of biology about essentialism (see, e.g., Matthen Reference Mayr and Ereshefsky1998, 117–21; Griffiths Reference Griffiths1999, 219–22; Millikan Reference Okasha2000, 18–20). This having been said, I suspect that even historical explanations demand a partly intrinsic essence; that, for example, the explanation of how polar bears came to be white will ultimately depend on essential intrinsic properties of polar bears and of their grizzly ancestors. But I shall not attempt to argue this.

That is the core of my positive argument for Intrinsic Biological Essentialism. We might sum it up: structural explanations in biology demand that kinds have essential intrinsic properties. That is my first main point in defense of Essentialism. I have further developed the case for Essentialism elsewhere by responding to objections (Devitt Reference Devitt2008, 355–78). I shall briefly summarize those responses in the space remaining to me.

4. Summary of Responses to Objections

The central objection to Intrinsic Biological Essentialism is that according to nearly all current “species concepts”—particularly, the biological species concept, the ecological niche concept, and the phylogenetic-cladistic concept—species are relational. Okasha expresses this consensus as follows: “On all modern species concepts (except the phenetic), the property in virtue of which a particular organism belongs to one species rather than another is a relational rather than an intrinsic property of the organism” (Reference Putnam2002, 201).

My response to this objection rests on a crucial distinction made by Mayr (Reference Mishler and Wilson1982, 253–54). It is the distinction between two problems, a taxon problem and a category problem:

This distinction yields two ways of understanding the question “What is a species?” The question could be asking about the nature of any group that happens to be a species or it could be asking about what it is to be a species.

My second main point in defense of Essentialism is that the relational species concepts are primarily concerned with the category question, whereas Essentialism is concerned with the taxon question. Not only do the species concepts straightforwardly yield answers to the category question, that seems to be what they are intended to do. And, contrary to the consensus (aside from Okasha [Reference Putnam2002], see, e.g., Sterelny and Griffiths [Reference Webster and Goodwin1999] and Wilson [Reference Wilson and Wilson1999]), the species concepts do not answer the taxon question, nor are they even associated with relational answers that are close to being explanatorily adequate.Footnote ²⁰ Indeed they can be happily wedded to Essentialism's (partly) nonrelational answer.Footnote ²¹

Where has the consensus gone wrong? The obvious answer is that the error has come from somehow conflating Mayr's distinction between taxon and category problems.Footnote ²² But then why have they been conflated? My tentative diagnosis is that the conflation stems from a mistaken answer to the conspecificity problem, the problem of saying in virtue of what, organisms are in the same species. People rightly think that a relational nonintrinsic answer to this problem entails a relational nonintrinsic, hence anti-Essentialist, answer to the taxon question. And it is tempting to suppose that the species concepts imply such an answer to the conspecificity problem. But this supposition is mistaken. Furthermore, Essentialism's partly intrinsic answer to the taxon question requires a partly intrinsic answer to the conspecificity problem.

Finally, I argue that some general features of Darwinism that have been thought to count against essentialism (see, e.g., Hull Reference Hull and Ereshefsky1965; Sober Reference Stanford1980, Reference Sterelny and Wilson1993; Ruse Reference Sober1987; Ereshefsky Reference Ereshefsky and Ereshefsky1992a, Reference Ereshefsky1992b; Sterelny and Griffiths Reference Webster and Goodwin1999; Wilson Reference Wilson and Wilson1999; Griffiths Reference Hull and Ereshefsky2002; Okasha Reference Putnam2002) do not undermine Intrinsic Biological Essentialism. Variation within a species can be seen to be compatible with Essentialism once one realizes that an intrinsic essence does not have to be “neat and tidy”—my third main point in defense of Essentialism—and that Essentialism is not wedded to the Aristotelian Natural State Model. Essentialism can accept the gradual change of one species into another. Still, there are some concerns raised by the lack of sharp boundaries between species. First, Essentialism must accept a certain indeterminacy about species. But this is no worry because this indeterminacy has to be accepted whatever one's (Darwinian) view of species and of essentialism; biology faces a sorites problem. That is my fourth main point in defense of Essentialism. Next, Essentialism is compatible with there being a certain amount of arbitrariness in choosing which groups of organisms to name as species. But this choice, however arbitrary, must not be confused with a choice we do not have: the choice to make things members of a group we have named.Footnote ²³ The earlier-mentioned indeterminacy enables Essentialism to deal with the problem of “monsters”—organisms lacking what might plausibly be proposed as the intrinsic essences of their parents’ species—and to maintain the lawlike status of biological generalizations despite apparent exceptions.

The responses I have just summarized deal with the objections to Intrinsic Biological Essentialism that I have found in the literature. Perhaps there are other objections that would be more effective. Perhaps it can be shown that my argument in favor of Essentialism—the argument from explanation—is inadequate. Given the strength and longevity of the consensus against such an essentialist doctrine, it seems reasonable to predict this. Still, it remains to be seen whether it is so. At the very least I hope to have shown that the case for the consensus needs to be made a great deal better than it has been.

If my arguments are good, the consensus relational view about particular species is quite mistaken. And many claims that biologists make day in and day out about the living world require species to have natures that they do not have according to this consensus.