1. Introduction

What is the “essence,” “nature,” or “identity” of a biological taxon? The commonsense answer, particularly about species, is that the essence consists, at least partly, in intrinsic, underlying, and mostly genetic properties. Let us call this answer Intrinsic Biological Essentialism (IBE). It has been urged by the influential philosophers Hilary Putnam (Reference Putnam1975), Saul Kripke (Reference Kripke1980), and David Wiggins (Reference Wiggins1980). Nonetheless the consensus in biology, especially in the philosophy of biology, is that the answer is quite wrong, smacking of “Aristotelian essentialism” and reflecting a naive and uninformed view of biology that is incompatible with Darwinism. Okasha captures the consensus well: “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 Okasha2002, 196). On this matter, according to Leslie, “there is a degree of consensus among philosophers of biology (and indeed biologists) that is almost unprecedented in philosophy at large” (Reference Leslie2013, 132).

Clearly, if a species has an essence at all and that essence is not intrinsic, then it must be relational. The consensus is indeed that the essence is relational: for an organism to be a member of a certain species, it must have a certain history. As Sterelny and Griffiths put it, there is “close to a consensus in thinking that species are identified by their histories” (Reference Sterelny and Griffiths1999, 8).

In “Resurrecting Biological Essentialism” (Reference Devitt2008), I went along with the consensus in accepting that there is a historical component to the essence of a taxon, and I have since argued for this (Reference Devitt2018a). However, “Resurrecting” went sharply against the consensus, particularly over species, in arguing that there is also an underlying intrinsic component: I argued for IBE. This argument has received detailed and interesting criticisms from five knowledgeable philosophers: Matthew Barker (Reference Barker2010), Marc Ereshefsky (Reference Ereshefsky2010), Tim Lewens (Reference Lewens2012), Sarah-Jane Leslie (Reference Leslie2013), and Matthew Slater (Reference Slater2013). These authors have made many more criticisms than I can answer here. I have chosen to make detailed responses only where that advances the positive case for IBE rather than simply defends the case already made.Footnote ¹

2. Clarifications

We need to clarify two matters. First, what are the taxa in question? They are those that are thought to fall under the biological categories in the Linnaean hierarchy of kingdoms, phyla, classes, orders, families, genera, species, and even subspecies. I say “thought to fall” because I sympathize with the doubts of some about this hierarchy (see, e.g., Ereshefsky Reference Ereshefsky and Wilson1999; Mishler Reference Mishler and Wilson1999). Even if these doubts are appropriate, it is quite clear which groups are thought to fall under the categories. Those are the taxa that concern our essentialism issue. So the concern is not with some other biological groups, for example, predators, parasites, females (Devitt Reference Devitt2008, 346).Footnote ²

There is an important further point: our working assumption should be that these taxa are very largely explanatory. Biologists are obviously striving for an explanatory classification, and we should surely assume that they have mostly achieved this. We know from the history of taxonomy that mistakes have been made, and they are surely still being made. Still, we should assume that our taxonomy is very largely good and explanatory. Thus, we should assume that Canis and Canis familiaris are explanatory notions, whether or not it is appropriate to take them as being about a genus and a species, respectively.

Second, taxon essentialism is a property or kind essentialism. A property P is an essential property of being an F if and only if anything is an F partly in virtue of having P. A property P is the essence of being an F if and only if 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; 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 Murdock, 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.

Some may resist any talk of “essence,” thinking that the term smacks of Aristotelian metaphysics and has a scholastic air. But the term does not matter. What I am picking out with that term—also ‘nature’ and ‘identity’—is the property in virtue of which an object is member of a certain kind; the property that constitutes its being a member or makes it a member. Those who find my terms for this property distasteful should choose another (Devitt Reference Devitt2008, 347–48). Some may think that there is no such property. I think that they are very wrong, for a reason I will give in section 5.

To say that kinds have essences (natures) is not to say, of course, that these essences are always worth investigating. Whether they are depends on how explanatory the kinds are. So the essence of being Australian is of little interest; that of biological taxa, of great interest; that of biological categories, of uncertain interest.Footnote ³

3. Summary of the Argument for IBE

Why believe in IBE? My answer came in two parts (Devitt Reference Devitt2008, 351–55), which I summarize here.

The first part of my argument for IBE was that biological generalizations about the morphology, physiology, and behavior of a taxon require explanations that advert to an intrinsic underlying, probably largely genetic, property that is part of the essence of the taxon. If we put together each such underlying property that similarly explains a generalization about a taxon, then we have the intrinsic part of its essence.

The explanations in question here are “structural” ones about the underlying states in members of a taxon that, along with the environment, make the generalizations true. In urging this argument, I emphasize a crucial distinction, made by Mayr (Reference Mayr1961) and renamed by Kitcher (Reference Kitcher1984/2003), between these structural explanations and “historical” ones about how members of the taxon evolved to have such states.Footnote ⁴ My argument is that structural explanations demand (partly) intrinsic essences: “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” (Devitt Reference Devitt2008, 352). And that something cannot be a historical/relational property. I was particularly harsh on the idea that the explanatory property might be that of standing in an appropriate relation to designated individuals, perhaps a “type specimen” (Hull Reference Hull and Ereshefsky1978/1992, 311–12) or, “say, Brigham Young” (Ruse Reference Ruse and Ereshefsky1987/1992, 344). I called this idea “explanatorily hopeless” (Reference Devitt2008, 363).

The second part of my argument is related to the first. Matthen points out that biologists think “that something is striped because it is a tiger” (Reference Matthen1998, 115; thus exemplifying the just-made point that biological classifications are explanatory). I put it like this: “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” (Reference Devitt2008, 352). Why does it? Because for an organism to be, say, a tiger, is for it to have the intrinsic underlying nature of tigers, and that nature, in combination with the environment, causes its phenotypic features, its morphology, physiology, and behavior: the same underlying properties that make the organism a tiger cause it to be striped. That is why being a tiger is explanatory.

In sum, the intrinsic nature of a taxon explains both the truth of generalizations about its members and why being in the taxon is explanatory.

An important methodological point: Leslie claims plausibly that the traditional argument for IBE “makes critical use of intuitions” (Reference Leslie2013, 109). As can be seen, my argument does not. It makes critical use of biological explanations.

What does the consensus have against IBE? “Resurrecting” identified two main objections. Okasha expresses one 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 Okasha2002, 201). I argued (Reference Devitt2008, 356–63, 366–70) that this objection is seriously mistaken, conflating another distinction due to Mayr, that between “the category problem” and “the taxon problem” (Reference Mayr1982, 253–54). Mayr introduced the distinction for species, but we can generalize it. Then, the category problem is concerned with what it is for a taxon to be in any particular category, for example, genus. The taxon problem is concerned with what it is for an organism to be in any particular taxon, for example, Canis. The “species concepts” that Okasha is alluding to are theories that address the category problem for species: they say what it is for a taxon to be a species (rather than, say, a subspecies or genus). These concepts provide a constraint on an answer to the taxon problem for any taxon that is a species but fall far short of providing an answer even for those taxa. And the concepts throw no light at all on the problem for taxa that are not species. In contrast to these concepts, IBE is an answer to the taxon problem and has nothing to say about the category problem. So IBE and the concepts are answering different questions. The consensus, as expressed by Okasha, is simply wrong.

Mayr’s distinction is well established and yet, as I demonstrated (Reference Devitt2008, 363–66), its significance has been standardly overlooked in discussing taxon essentialism. And it seems to me that my critics Barker (Reference Barker2010, 76–77, 89), Ereshefsky (Reference Ereshefsky2010, 680–81), Lewens (Reference Lewens2012, 752), and Leslie (Reference Leslie2013, 138–42) continue this practice in mistakenly resting their case for an entirely relational view of a species’ essence on the species concepts. My argument for this must await another time (Devitt, Reference Devittforthcoming).

Why has the significance of this distinction been standardly overlooked in discussing essentialism? My tentative diagnosis was that it has been overlooked because of the appealing, but mistaken, idea that the species concepts imply relational answers to the conspecificity problem, the problem of what makes two organisms conspecific. For, if that idea were right, IBE could not be (Devitt Reference Devitt2008, 363–66). In response, Barker claims that this idea about conspecificity is “common knowledge” among “species aficionados” (Reference Barker2010, 76). The idea is certainly present in my critics, Ereshefsky (Reference Ereshefsky2010, 681), Leslie (Reference Leslie2013, 140–41), and perhaps Lewens (Reference Lewens2012, 752). So, my diagnosis is no longer tentative. The substantive issue is, of course, whether the aficionados’ idea is right. Is their “common knowledge” knowledge at all? My argument was that it is not. Barker disagrees (Reference Barker2010). My response to Barker must also await another time (Devitt, Reference Devittforthcoming).

The other main objection to IBE is that it is at odds with the variation and change that are central to Darwinism. I responded to this at some length (Reference Devitt2008, 370–78). I will say a lot more here.

Discussions of essentialism in biology are dominated by talk of species. This dominance is unfortunate in two ways. First, it leads to the almost total neglect of taxon problems for nonspecies, so answers are not comprehensive. For example, taxon problems for nonspecies go totally undiscussed by four of my critics.Footnote ⁵ Second, the domination muddles the essentialism issue, which arises equally for taxa of all categories, with the vexed issue, addressed by the species concepts, of what it is for a taxon to be in the species category rather than some other category. As noted, the essentialism issue concerns the property an organism must have to be a member of some taxon that has been identified and named by biologists and that is standardly thought to have a certain place in the Linnaean hierarchy. Whether that taxon has that place, whether it counts as a subspecies, species, genus, or whatever, is an issue that is quite independent of taxon essentialism.

How are the intrinsic essences of taxa in a hierarchy related? Where taxon T is in a lower category than taxon T* and falls within T*, the essence of T must include the essence of T* but not vice versa. Thus, everything that has the essence of Canis familiaris has the essence of Canis but not vice versa; some Canis are foxes.

I aim now to develop the case for IBE by responding to two sorts of criticisms. First, in section 4, I respond to criticisms that IBE is at odds with certain biological variations. Then, in section 5, I respond to criticisms of my answer in “Resurrecting” to the charge that IBE’s talk of intrinsic essences is an uncalled for metaphysical addition to biology. I accept the criticism and offer another answer.

4. Variation

Leslie finds my argument from structural explanation to IBE “intuitively appealing, but not … ultimately successful” (Reference Leslie2013, 133). Focusing on variation, she makes a number of good points about underlying states and phenotypic features. She thinks that these points count against IBE. Indeed, facts about variation are often thought to count against intrinsic essentialism: the idea “that variation among organisms is the crucial stuff of changing life and of life’s progress” is thought to be “devastating to essentialism” (Hey Reference Hey2001, 62). In contrast, I have argued that IBE accommodates variation nicely (Reference Devitt2008, 370–78). Leslie’s variation objections are interestingly different from the ones I have discussed and I will address them in some detail. I think that IBE can handle them too.

First, Leslie considers my example of the rhinoceros and comments: “Devitt is, of course, indisputably correct that each particular African rhino has some intrinsic features that, in combination with the environment, are causally responsible for that individual’s having horns. This does not entail, however, that those very same intrinsic features are also responsible for other African rhinos’ having horns. Whether this is so is a substantive empirical hypothesis, not one whose truth can be intuited in advance” (Reference Leslie2013, 134). Leslie supports this with the case of jade. As is well known, two different chemical compounds, jadeite and nepthrite, have been lumped together as jade. So no common intrinsic chemical structure explains the similar observable features of all samples of jade.

Leslie is right, of course, that the fact that some intrinsic feature causes one rhino to have horns does not entail that that feature causes other rhinos to have horns. But the argument for IBE does not rest on any such entailment. It rests on what, as a matter of fact, explains the (nonaccidental) fact that rhinos have horns. The claim that some one intrinsic feature of rhinos is responsible for this is, as Leslie says, “a substantive empirical hypothesis.” But the modal force of the hypothesis arises from an “inference to the best explanation” not an entailment. The argument for IBE rests on fallible empirical hypotheses of just this sort. I claimed that such hypotheses, implicitly embraced even by little children (Keil Reference Keil1989), are implicit in the practice of taxonomists (Devitt Reference Devitt2008, 352–53). In particular, I now claim, these hypotheses are supported by the contemporary role of genetic analysis in reclassification.Footnote ⁶

Take tigers (Panthera tigris), for example. A paper begins with the following claim about what were thought to be the several subspecies of tiger: “available molecular evidence suggests that extant tigers are extremely similar genetically” (Cracraft et al. Reference Cracraft, Feinstein, Vaughn and Helm-Bychowski1998, 139). Nonetheless, using DNA sequencing techniques, the paper argues, on the basis of genetic differences, for a reclassification: the Sumatran tiger is not a subspecies of Panthera tigris but a distinct species (148).

Similarly, a paper on African elephants argues against “the consensus that all belong to the single species Loxodonta africana” largely on the basis of a “deep genetic division between the forest and savannah populations” (Roca et al. Reference Roca, Georgiadis, Pecon-Slattery and O’Brien2001, 1473–74). The paper concludes that these populations form two species, the former, Loxodonta cyclotis, and the latter, Loxodonta africana (1476).

“Cryptic species” provide many examples of taxonomic reclassification on the discovery that phenotypic properties that were thought to be caused by the one underlying nature are actually caused by two: “Cryptic species are defined as ‘two or more distinct but morphologically similar species that were classified as a single species’ (Pfenninger and Schwenk, Reference Pfenninger and Schwenk2007). Because of this morphological similarity, most cryptic species were initially considered to be a single species until genetic data and rigorous scrutiny of phenotypic characters indicated otherwise. Cryptic species are relatively common across a wide range of taxa and habitats” (Andrews et al. Reference Andrews, Williams, Fernandez-Silva, Newmang, Copus, Wakefield, Randall and Bowen2016, 361).

What we see is a pattern of tying a taxon to an underlying genetic structure, a structure that causes its phenotypic properties, and of reclassifying whenever it is discovered that there is significant genetic difference among two populations within a taxon. And this is just what we should expect because the underlying structure makes the taxon explanatory (secs. 2 and 3).Footnote ⁷

Second, Leslie claims that “a member of one species … may have more genetically in common with a member of another species … than with a member of its own species” (Reference Leslie2013, 133). The argument for IBE requires that members of a species share a genetic structure that largely constitutes the nature of the species. It does not require that the rest of the genetic structures of these members be the same. So IBE is quite compatible with what Leslie claims. What matters to IBE is not an overall similarity among taxon members, established by “counting genes,” but a similarity in one particular part of the underlying structure, the part that causes (in its “normal” environment) the distinctive phenotypic features of the species.

Leslie’s third objection moves from genetic variation to phenotypic variation: “conspecificity is compatible with a great deal of variation in phenotype at a time, and even more dramatically over time” (Reference Leslie2013, 134). The argument for IBE hypothesizes an underlying nature that causes the phenotypic properties that are shared among conspecifics. Conspecifics can of course share those phenotypic properties while differing in many others. This is very obvious when we think of male and female conspecifics, which often differ dramatically. It is even more obvious when we think of the higher taxa (remember, IBE applies to them too); thus, the primate taxon (an order, in Linnaean categories) includes lemurs as well as humans, which differ greatly in phenotype. The phenotype variations noted by Leslie do not undermine IBE.

Those variations are caused by intrinsic underlying differences within a taxon. But, of course, phenotypic variations can also be caused by environmental differences. Consider the standard example of air temperature determining the sex of fetuses in some turtles and reptiles, and the impact of temperature, soil, and so on, on the properties of plants is obvious. From the perspective of IBE, there is no puzzle about this. The one underlying nature of a taxon causes its members to have certain features in one environment, other features in another.

This raises an important general point: the underlying nature of a taxon is just as explanatory of the atypical phenotypic features of a member produced by an “abnormal” environment as of the typical features of members in their “normal” environment. The nature plays its causal role across all environments. (And, probably, no phenotypic feature is essential.) Finally, the generalizations about a taxon that its intrinsic essence is supposed to explain, like that zebras are striped, are implicitly restricted to the taxon’s “normal” environment.Footnote ⁸

Slater sees the causal role of the environment as generating a problem in taking “genetic structure” as the essence of a species, just as we take chemical structure to be the essence of a chemical kind: “there is an important contrast between … chemical structure and an organism’s ‘genetic structure.’ … An organism’s ‘genetic structure’ … does not alone determine, or even strictly suffice to explain, facts about the organism’s outward character or behavior in the absence of information about its actual environment” (Reference Slater2013, 44). But there is no significant contrast. Change the environment of a chemical kind enough, and its outward character or behavior may change; thus, zinc ceases to be malleable below 300 degrees Fahrenheit.

Fourth, Leslie points to variation in the causes of phenotypic features: “In general, within a given species, individuals who share a common phenotypic feature need not share intrinsic microstructural bases that gave rise to the feature” (Reference Leslie2013, 136). A feature that is “genetically induced” in some individuals might be “environmentally induced” in others (137). And so it might.

Consider environmentally induced phenotypic features first. We have just given some examples. Leslie provides a nice further one, the Himalayan rabbit, a breed of the Common Rabbit (Oryctolagus cuniculus): “Himalayan rabbits, when raised in moderate temperatures, have white body fur with black tails, noses, and ears; if they are raised in cold temperatures, however, they develop wholly black fur” (Reference Leslie2013, 137). From the perspective of IBE, there is no puzzle about this. The one underlying nature of this breed causes its members to have certain features in moderate temperatures and certain other features in cold temperatures.

Now consider genetically induced phenotypic features. Leslie points out that the black fur of a rabbit can arise not only from the environment, as with the Himalayan rabbit, but “relatively straightforwardly from a given rabbit’s genetic makeup—that is, as an inherited trait that manifests itself across various environments” (Reference Leslie2013, 137). But note that such a rabbit’s black fur does not arise from the part of its genetic makeup that is essential to its being a rabbit (else rabbits would typically be black). So, the underlying nature of the taxon rabbit does not determine that fur color. So this phenomenon has no bearing on IBE.

In sum, the underlying nature of a taxon causes different phenotypic properties in different environments. Different genetic makeups among organisms that share the common underlying nature of a taxon can lead to different phenotypic properties. Indeed, the latter point is obvious when we remember that both lemurs and humans are in the primate taxon.

Finally, Leslie’s most interesting objection to IBE is to be found in her vivid discussion of cases demonstrating that the path to a particular phenotypic property is complicated and varied:

We have just discussed the impact of the environment on the developmental pathway to a phenotypic property in species members and need say no more. But what about the impact of the genes?

Leslie is emphasizing that the role of genes is complicated, involving “a variety of genes.” But it is no part of the argument for IBE that the paths from genes to phenotypic traits are simple. Indeed, although for convenience I sometimes talk simply of a genetic essence, I am noncommittal on precisely what underlying properties constitute the essence of a taxon and hence cause its phenotypic properties: “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. 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)” (Devitt Reference Devitt2008, 347).

Indeed, it would be foolhardy for IBE to have commitments on these empirical biological matters, and it has none. We already know enough, of course, to reject “the crude idea that there is, say, ‘a tiger gene’” (Devitt Reference Devitt2008, 371). Okasha thinks “that species are distinguished by clusters of covarying [chromosomal and genetic] properties” (Reference Okasha2002, 197). So perhaps, I suggest (Reference Devitt2008, 371), that is where we should look for the underlying essence. But perhaps not. Consider these interesting claims by Walsh in the course of arguing for intrinsic essences:

The moral of this is that the wise philosophical essentialist leaves the details of essences to scientists; biological essences are to be discovered by biologists.

This point generalizes. Leslie criticizes Kripke-Putnam essentialism not only about biological kinds but also about chemical kinds (Reference Leslie2013, 142–58). Her criticism draws heavily on Needham’s long-running campaign against Kripke-Putnam essentialism about water (see Needham [Reference Needham2011] and the earlier papers cited). This campaign, based on an illuminating presentation of the scientific facts about water, includes criticism of the Kripke-Putnam talk of “H2O” as the essence of water. But it is central to the positions of Putnam (Reference Putnam1975, 224–25) and Kripke (Reference Kripke1980, 119–29) that we should look to empirical science, not to philosophers like themselves to fulfill the supremely important task of discovering the essence of “natural kinds” like water. Needham shows convincingly that their talk of the essence of water being “H2O” is quite inadequate, but this talk should be seen as nothing more than a philosopher’s hand wave toward the scientific facts.

I conclude that a complicated role for genes in causing phenotypic traits is not a problem for IBE.

Leslie is also emphasizing something that may seem more problematic for IBE: the genetic component involved in the developmental pathway to a phenotypic property can vary among the members of a species. Thus, commenting on the cause of human female genitalia, Leslie says: “For each individual infant, there will be a genetic component to the explanation (as well as an environmental component), but this genetic component need not be the same for each infant” (Reference Leslie2013, 136). She makes similar remarks about the three toes on the hind feet of guinea pigs: “having three toes on the hind feet is a characteristic property of guinea pigs (Cavia porcellus). Possession of this phenotypic property is due to a flexible interaction effect between a number of factors, both genetic and nongenetic—that is, there is a generous range of pathways, all of which lead to having three hind toes. Individual guinea pigs can differ significantly from each other with respect to these factors and yet each have three toes” (137). See also her discussion of the number of eggs produced by female salamanders (137–38). The previous untroubling point was that many genes are involved in causing a phenotypic property. The current point is that the ones involved can vary within a taxon: the posited essence of a taxon can cause a phenotypic property in more than one way. Walsh is illuminating on this. He continues the above passage on regulatory gene networks as follows: “These networks, or circuits, are characterised by flexibility, redundancy and robustness (Salazar Ciudad et al. [Reference Salazar Ciudad, Sole and Newman2001]; Szathmary [Reference Szathmary2002]). All of these features function to preserve the capacity of a gene network to produce its normal output in the face of perturbations and variations of initial conditions. For example, gene regulatory networks show a remarkable capacity to compensate for the ‘knockout’ of elements of the network. If a gene is removed, or ‘knocked-out’, of a regulatory network the network typically compensates and finds alternative ways of producing its characteristic output (Greenspan [Reference Greenspan2001])” (Walsh Reference Walsh2006, 436–37).

Walsh does not see this variation in the effective genetic component as a problem for intrinsic essentialism. Quite the contrary. He takes the essence of a species to be “the causal capacities of its developmental systems” that realize its “phenotypic plasticity” (Walsh Reference Walsh2006, 441). And that plasticity consists in just the sort of “flexibility, redundancy and robustness” illustrated by Leslie’s guinea pig. So, why does Leslie think that the variation is a problem?Footnote ¹⁰

Leslie objects to the “disjunctive” nature of the underlying cause of a phenotypic property, its being either this genetic component or that one. Using the example of jade again, Leslie argues that IBE cannot allow this:

Lewens makes a similar point, claiming that there might be various microstructural causes of stripiness in tigers and continuing: “Of course, we will be able to identify a characteristic ‘pattern’ of genotypic properties across the species, simply by enumerating whichever genes actually cause the instances of species-typical phenotypic properties, but since this must be the case however unruly these underlying genes are, the sense of essence we salvage here is trivial” (Reference Lewens2012, 753).

This is interesting. But the wrong moral to draw from it is that the one underlying nature cannot yield disjunctive explanations. The disjunctive nature of jade does explain its observable properties. It just does not explain them well enough for the purposes of science. Kinds like jade, with extremely disjunctive underlying essences, are not scientifically acceptable because alluding to them does not provide adequate explanations; they do not “carve nature at its joints.” From the perspective of IBE, that is why a cryptic species is not a biologically acceptable species. And, for similar reasons, Sumatran tigers are no longer classified as a subspecies of tigers and forest elephants are no longer placed in the same species as savannah elephants. But none of this shows that the explanatory needs of biology cannot tolerate any disjunction in the way an underlying nature causes a taxon’s phenotypic properties. Indeed, the moral of Leslie’s cases of female genitalia, the guinea pig, and the salamander, and of Walsh’s discussion, is precisely that biology embraces some disjunctions: the underlying nature may provide several different causal paths to a phenotypic property. Nonetheless, the argument for an intrinsic essence presented in section 3 holds: it is because of the underlying largely genetic essence of guinea pigs that they have three toes on their hind feet, and it is because a certain animal is a guinea pig, hence has the underlying nature of a guinea pig, that it has three toes on its hind feet. These are biologically sound explanations. But similar essentialist explanations of a phenotypic property of a cryptic species are not acceptable. And that is why cryptic species are not species.

I conclude that the disjunctive explanations that Leslie and Lewens are pointing to do not undermine IBE but rather throw an important light on a species’ intrinsic explanatory essence. The explanations demonstrate that the essence of tigers is not only not “a tiger gene,” it is a long way from a tiger gene. So, what sort of disjunctive explanation is scientifically acceptable in biology? What is the difference between the acceptable explanation of the guinea pig and the unacceptable one of a cryptic species? This is the sort of biological question on which, as just noted, it would be wise for a philosopher to be noncommittal. But, presumably, the acceptable disjunctive explanations, the ones that do not demand reclassification, must be ones arising from a certain underlying unity in what is flexibly responsible for phenotypic properties.

5. The Added Metaphysical Claim

Ereshefsky is quite unconvinced by my argument for IBE. He notes that the structural explanation of why zebras have stripes adverts to intrinsic properties while insisting that the historical explanation must advert to genealogical relations (Ereshefsky Reference Ereshefsky2010, 680). So far, then, we are in agreement. Disagreement begins with the move I make from the structural explanation to IBE, to “the added metaphysical claim that the character cited in the explanans is essential to membership in a taxon” (680). This disagreement is reminiscent of a nice question put to me by Peter Godfrey-Smith. I expressed the question like this: “It is of course the case that the truth of any … generalization [about the phenotypic properties of a taxon] must be explained by an intrinsic, probably largely genetic, property, but why does that property have to be an essential property of the kind in question?” (Devitt Reference Devitt2008, 354). I attempted an answer to this question by appealing to the law-like nature of the generalizations and subjunctive conditionals (377–78). Both Lewens (Reference Lewens2012, 755–56) and Slater (Reference Slater2013, 51–53) do a good job of showing that this answer is inadequate. So I hereby abandon it.

Implicit in my argument for IBE was another answer that I strangely overlooked. A clue to this answer was given in section 4 above: “The nature plays its causal role across all environments.” Any organism in taxon T has certain phenotypic properties because it is in T; it is because an animal is a zebra that it is striped. So the property that makes something a zebra, whether we call that property an “essence,” “nature,” or whatever, must cause that zebra, in its environment, to have stripes; the essence of zebras must explain the place that certain organisms have in the causal nexus just because they are zebras.Footnote ¹¹ Sober is getting at this with the following demand: an essence “must be explanatory. … A species essence will be a causal mechanism that acts on each member of the species, making it the kind of thing that it is” (Reference Sober and Ereshefsky1980/1992, 250). The essence of T is whatever property, as a matter of actual fact, plays that causal role in an environment. This is the basis for the “added metaphysical claim.”

So Lewens is right in supposing that I do not “wish to say that species have intrinsic essences just so long as there are clusters of properties common and peculiar to them” (Reference Lewens2012, 755). But he is wrong to wonder whether my “essentialism amounts to nothing more than the unobjectionable assertion that … intrinsic properties of organisms are relevant to determining their species” (756). Intrinsic essences play a fundamental causal role.

6. Conclusion

My argument for IBE in “Resurrecting,” summarized in section 3 above, was that the intrinsic nature of a taxon explains both the truth of generalizations about its members and why being in the taxon is explanatory. This argument came in for interesting criticisms from Barker (Reference Barker2010), Ereshefsky (Reference Ereshefsky2010), Lewens (Reference Lewens2012), Leslie (Reference Leslie2013), and Slater (Reference Slater2013). My main aim in this article has been to give a detailed response to criticisms where that response advances the positive case for IBE.

My first detailed response, in section 4, rejects the charge that IBE is at odds with certain biological variations. Thus, according to IBE, the underlying nature of a taxon causes different phenotypic properties in different environments, different genetic makeups among organisms that share a taxon’s underlying nature can lead to different phenotypic properties, and the many genes involved in causing a phenotypic property can vary within a taxon.

My second detailed response was to the charge that IBE’s talk of intrinsic essences is an uncalled for metaphysical addition to biology. The response to this charge in “Resurrecting” has been shown to be inadequate. I now claim that the essence or nature of a taxon simply is whatever underlying state causes members of that taxon, in their environment, to have their phenotypic properties; the essence explains the place that those organisms have in the causal nexus just because they are members of the taxon.

It is a striking feature of the consensus, I argue elsewhere (Reference Devitt2018a), that it has failed to produce a historical/relational answer to the taxon problem that is complete, plausible, and comprehensive. And, I would argue, none of my critics have rectified this situation. The absence of such a historical alternative to IBE is part of the case for IBE. That case increases if I am right that there is indeed a historical component to a taxon’s essence but it is one that demands an intrinsic component. For, the relevant history of a taxon is of organisms of a certain intrinsic kind evolving into organisms of a certain other intrinsic kind, until we reach the taxon in question (Devitt Reference Devitt2018a).Footnote ¹²