Are species theoretical objects?

[Note: this is a paper that has sat in my drawer for a while now. I am posting it to follow from my last post on the theological origins of species. If species are not ranks in biology, what are they?]

It is often claimed that species are the units of evolution, but this is not defined or clearly explained. In this paper I will argue that species are phenomenal objects that stand in need of explanation, but that they are not objects required by any theory of biology. I further define, or rather describe, species as the genealogical cluster of various lineages at the genetic, haplotype, genomic, organismic, and population level, in keeping with my previous discussions.

The nature of “species”

Species: A term which everybody thinks they understand, but which nobody agrees upon, to denote the “basic units” of groups of biological organisms.

In his response to Darwin’s Origin of Species, Louis Agassiz wrote

If species do not exist at all, as the supporters of the transmutation theory maintain, how can they vary? and if individuals alone exist, how can the differences which may be observed among them prove the variability of species? (Agassiz 1860)

Darwin responded rather grumpily in a letter to Agassiz’ one-time colleague, Asa Gray

I am surprised that Agassiz did not succeed in writing something better. How absurd that logical quibble “if species do not exist how can they vary”? As if anyone doubted their temporary existence? (August 11 1860, Darwin 1888; Quoted in Gayon 1996: 229; see also the discussion in Ghiselin 1969: chapter 4)

If Agassiz is responding to anyone, it is the French tradition that derived from Buffon, in which species are unreal but individuals are not, and Lamarck’s subsequent view that transmutation rendered species merely concepts of convenience.

Many scientists and philosophers have said something like “species are the units of evolution” or “biodiversity”; it is even the title of a well-known book of essays on the subject (Claridge, Dawah, and Wilson 1997). What does this even mean? So far as I can tell, nobody has really fleshed this out. In this essay I intend to argue that species are salient phenomenal objects rather than objects of any biological theory, let alone of evolutionary theory.

In addressing the impact of evolution upon the concept of species we must first ask what are the units of evolution? Such questions of ontology depends a lot on what theory is being employed. When talking about population genetics, the basic units, of course, are the allele and the locus (Griffiths and Stotz 2007). When talking about development, then the unit is the organism, as it also is when you are talking about ecological interactions, although “species” is used here as the term for a class of ecologically exchangeable organisms; that is, organisms that play the same role in the local ecosystem (Wilkins 2007a, 2007b). Although organisms are pretty well all different (which is the point of population genetics), for the purpose of trophic webs (the food webs of ecology), conspecifics are treated as being interchangeable elemental units.

Then there are the larger units of evolution: populations (Godfrey-Smith 2009), and the particular revision of that concept, the “deme” (Winsor 2000). A deme is basically the population that can interbreed – the term in the equations of population genetics is Ne, the effective, or reproductive, number of individuals. However, nonbreeders also play a role in many species in contributing to the fitness of their kin, by helping raise them, or finding food, so the ontology here depends solely upon what the research issues are. While “population” is itself somewhat fuzzy (Gannett 2003) – the more migration there is between two (sexual) populations the more they start to look like a single population – it is a theoretical object (Millstein 2006).

But species? I am aware of no theory that requires them. Having made that outrageous claim, I had better explain what I mean before I am attacked by roving bands of disaffected taxonomists. Yes, ecologists and conservation biologists use species, but what they are really doing is using field guides as a surrogate for the ecological roles played in an ecosystem by individuals of the species who are more or less normal – the “wild type”. Likewise, medical and biological researchers do the same thing with their model organisms. Mus musculus, or the common mouse, is used as a study organism because it is assumed that each individual member of that species shares the same properties (developmental cycles, phenotypes). But in practice they use “strains” that are specially bred to see the effects of gene knockouts, for example. The “objects” here are the genetic strains and the organisms (Ankeny 2000).

Systematists “use” species because they describe and name them, but the explanations given of species being species are manifold. The notion of a “gene pool” or “metapopulation” is the foundation of one such explanation (de Queiroz 2005, 2007). But the theories used, the explanations, are not theories of species; they are theories of gene exchange, reproduction, fitness, adaptation, and so on. Species are being explained; they do no work in explaining. One possible exception is the work species do in “species selection” theories (Stidd and Wade 1995; Jablonski 2008; Rice 1995; Grantham 1995; Gould and Lloyd 1999; Lloyd and Gould 1993), but it is arguable whether these are actually theories as such, and equally arguable whether the properties that are “theoretical”, which play a role in causal explanations, are those of the species, populations within the species, or of the individuals or kin groups. If species selection is taken to mean that species whose members have a particular property (like eurytopy) tend to speciate more often, then “species” in this sense is merely a mass noun (Grandy 2008).

So, there are two ways we might go if species aren’t theoretical objects. One is that we may deny that species exist, and a lot of people do this, and have done since the Origin and even before. I call them species deniers, because they deny that species exist, although the usual term is species conventionalists, or nominalists (both philosophically and historically misleading terms). A version of species denial is to replace the term species with some “neutral term”. Deme was one of these, but as Polly Winsor has shown, it got subverted by population geneticists for the meaning given above. Other examples include Operational Taxonomic Units (Sneath and Sokal 1973; Sokal and Sneath 1963), Least Inclusive Taxonomic Units (Pleijel and Rouse 2000), Evolutionarily Significant Groups (Hey et al. 2003; Hey 2001), and so on. In each case, the term species came or is coming back into use.

Why is species so durable? The answer comes from not taking the term and concept as a theoretical term. Species is a useful term because species are real phenomena [1]. That is, they are things observed that call for explanation, they are explicanda. The theories of biology explain why there are species, although not all the same theories apply for all species, of course (Wilkins 2003). Biology is not that neat. Some species are explained the way the textbooks say – through the acquisition of reproductive isolating mechanisms formed in geographical isolation. Some aren’t. There are species formed by hybridisation [2], by sexual selection, and of course asexual or mostly asexual species that are formed, as I argue elsewhere (2007a) by adaptation to niches.

If there is no general theoretical account of species, why do we have this category? Well, it might be because we tend to name things that look similar to us. This is what species deniers think: it’s all about us and our cognitive dispositions, not the things themselves. I don’t agree with this. I think there are some general features of species that license us calling them all species. So here’s the claim: Species are salient phenomenal objects. They are salient not because of our perceptual tendencies alone but because they do exist. They’re a bit like mountains. Each particular mountain is caused by definite processes, but every mountain is not caused by the same processes. We identify mountains, because they’re there. We explain them with theories of tectonics, vulcanism, or even (if they are dunes) wind. In a similar fashion, species are clusters of genomes, phenotypes, and organismic lineages. We explain them because they need explaining. A species is (roughly) where the lineages of genes, genomes, parent-child relationships, haplotypes, and ecological roles all tend to coincide – in bacterial systematics this is referred to as a polyphasic approach (Vandamme et al. 1996; Vos). Not all of these need coincide in every case, but so long as most of them do, they are species, and we must give an account of them. And we can and do.

Phenomenal objects

A Scientific American essay on the definition of species (Zimmer 2008) has the following subhead:

The debate over species definition is far from over and is more than a mere academic spat. Proper classification is essential for designating the endangered list.

This is perhaps the most crucial practical aspect of the species concept debate, but it isn’t the most theoretically interesting. Biology, like most sciences, has a need for units of measurement, and like most sciences those units need to be grounded in the real world. So what species – the “rank” of biology that is agreed on most sides as the most or only natural one in the Linnaean hierarchy – are determines many measures of biology in fields from genetics to ecology. If, as a significant number of specialists think, the rank is a mere convention (Mishler 1999), then those measures become arbitrary and meaningless.

Therefore we need to consider what sort of “unit” a species might be. I can think of three alternatives. The first is that species are, in fact, simply a matter of convention, which is to say, something that makes things convenient for us in communication [3], just as John Locke said in the Essay (Bk III, chap. V, §9; although that was about logical species, not biological species). Instead, say researchers like Paris polychaete specialist Frederick Pleijel and Rutgers geneticist Jody Hey, we need to replace the notion species with something like a “least inclusive taxonomic unit” (LITU, Pleijel) or “evolutionary group” (Hey). There are other replacement concepts in the offing. And the so-called “phylogenetic species concept” is not really a concept of species, at least in one of the versions under that name, so much as something very like a LITU that gets called a species (Agapow et al. 2004; Wheeler and Platnick 2000).

The second alternative is that species is a term that plays a theoretical role in biology, and this seems intuitively right: we talk about species as the units of evolution, so they are supposed to be required by evolutionary biology, and likewise in ecology, species are the unit that is crucial in defining the biodiversity of a region or ecosystem. But if species are theoretical objects, we ought to find them as a consequence of theory, not as a “unit” that we feed into theoretical or operational processes, and so far as one can tell, this is not the case. Population genetics and evolutionary theory have populations, haplotypes, alleles, trophic nodes, niches and so on, but what they do not have are species. In every case where species are used in theory, they are primitives, or stand as surrogate terms for the other things mentioned. Theory does not define species.

This might be challenged by adherents of Mayr’s biological species concept, or one of the derivative or related conceptions – a species is a protected gene pool, as Mayr said (Mayr 1970: 13). This is certainly the view of Coyne and Orr in their Speciation book of a few years ago (Coyne and Orr 2004). But as Zimmer points out, the vast bulk of life would not be in species if that were the case, and anyway, species were well described and identified long before genetics was developed, some two centuries before. So they must at least be things that can be observed in the absence of theory. Of course, some species are harder to identify than others, requiring techniques that are recent, but that still doesn’t make species theoretical objects.

This brings us to the third alternative: species aren’t theoretical objects at all; they are objects that have phenomenal salience [4]. That is, we do not define species, we see them. Consider an analogous case: mountains. Mountains are hard to define, and they have a multitude of geological causes, ranging from uplift, subduction, vulcanism, differential erosion, and so forth. “Mountain” is not a theoretical object of geology – subduction zones, tectonic plates, and volcanoes are. A mountain is just something you see, although there are no necessary sets of properties (or heights) that mountains have to have, and it is often vague when differentiating between them. A mountain calls for an explanation, and the explanation relies on theory, but equally so do mesas, land bridges, and caves.

So the suggested answer to the question: what is a species? is that a species is something one sees when one realizes that two organisms are in the relevant manner the same. They are natural objects, not mere conveniences, but they are not derived from explanations, but rather they call for them.

Speciation modes

To elaborate on this claim, let’s consider the modes of speciation that are called into account for the existence of species. Here is a list revised from Sergey Gavrilets (Gavrilets 2004; Wilkins 2007b):

Vicariant – divergent selection and stochastic factors like drift after division of a population by extrinsic factors such as geographical changes;

Peripatric – a small subpopulation, mostly isolated, at the extreme of the parent range. The idea is that it will have both a non-standard sampling of alleles, and also be subjected to divergent selection pressures in extreme environments (for that species);

Centrifugal – central populations that carry a sample of many alleles that become isolated through, say, “island” formation (such as the mountain “islands” in the Amazon);

Punctuated equilibrium – the appearance of relatively rapid speciation and subsequent stasis as the population reaches equilibria of alleles. In my opinion, this is inappropriately included here, for it is a “pattern” rather than a “process” (or “event”) of speciation, and as such can be caused by any of the other scenarios/modes;

Chromosomal speciation – the rearrangement of chromosomes, either by duplication or insertion, fission, fusion or inversion;

Hybridization – the fusion of two genetic lineages, usually from distinct species, including allopolyploidy. In allopolyploidy the genetic complement of two species is paired up by a loss of secondary division, giving a symmetrical set of chromosomes;

Reinforcement – once hybrids are of lowered fitness for whatever reason, selection will tend to reinforce separation of the gene pools (for example, a hybrid rock and grass dwelling lizard might be less able to survive in either environment as well as the “pure” lines);

Competitive – this is Darwin’s scenario. Members of a species that are in strong selection for a limited resource may result in specialized forms that are thus in less competition with the ancestral forms that make use of some other resource;

Clinal/ecotonal – Gavrilets calls it “speciation along environmental gradients”, where limited migration and selection leads to aggregation of forms at the terminal ends of the cline;

Host shift – this is the case of the Rhagoletis fruit flies mentioned above, that Stuart Berlocher (Berlocher 1999, 2000; Berlocher and Feder 2002) and colleagues have studied. Host fidelity replaces geographic isolation;

Runaway sexual selection – this is secondary selection by mate choice of polygenic traits (Lande 1981).

In that paper, I locate each of these modes as a unique coordinate or region in a three-dimensional space, the axes of which are:

Gene flow – the rate of migration between populations, or the amount of genetic material exchanged in a mating event, from 0% to 50% – the Haldane parameter m (Haldane 1930);

Selection – the degree to which selection is endogenous to the organisms, such as climatic selection, say on the Soay sheep (Coulson et al. 2001) or the finches of the Galápagos Islands, or is intrinsic to the organisms (such as mate choice or immunological compatibility); and

Stochasticity – which is basically whether selection is directional or stabilizing, or whether the rate of change of the genetic constitution of the population is due to genetic drift and other stochastic “forces”.

Now, suppose we have a species of flowering plant. It buds off a new species due to climatic and pollinator adaptation, with no exchange of genes between populations. In binary terms, its coordinate would be <0,1,1>. That is the explanation of that species in terms of how its speciation occurred. If it is in sympatry (shares a geographical range) with its parental species, it must have adaptations that prevent competitive exclusion, and so remaining a species, rather than merging back into the parental metapopulation, is due to selection for those niche adaptations. This, it need not be said, will not be true of species that occupy other regions of “speciation space”. The plant has its own explanation, and its own phenomenal reality. We know it is a different species simply because introgression and hybridization are inhibited, and competitive exclusion does not occur because parent and child species occupy different points of the fitness landscape. But is this sort of species (which may have formed in isolation from the parental species) required by any theory of speciation or evolution, or biology in general? And generalizing, is any type of species so required? The answer is, no. Simply put, if it merged back into the parental species, it would not be identified as a species itself, but a variety (and maybe a fairly ephemeral one at that).

What we have is a post hoc phenomenon: being a species at t. If a species is distinguished by its genetic and ecological properties, however acquired, then it is a species. If it isn’t, then the issue doesn’t arise. This is a bit like weak anthropomorphism – any universe that lacks humans isn’t going to be discussed by them, but you cannot make the inference that there is something special about human-occupied universes simply because we are here talking about them. Similarly species – if a species is separate from other species (in its own unique and contingent manner) then it is a phenomenon that demands explanation, but if it doesn’t, the theories of population genetics, developmental biology, ecological interactions and the like all continue on apace. Post hoc explanation is not an illicit move in science; it can’t be, because one of the major roles of a scientific theory is to explain what is observed. But that doesn’t imply that the phenomena are theoretically significant. Just that they interest us enough for them to call for an explanation.

Theory-dependence and derivation

Traditionally, something was a theoretical object, that is, an object that was only theoretical, if it was something that the theory required or employed but which was not empirically ascertainable. Examples were “electron” c1920, “gene” prior to 1952, and perhaps still “Higgs boson” for reasons that I do not understand. But this is a positivist sense of theory – a formal system in which objects are either verifiable or not. Whether or not one is now a logical empiricist instead of a logical positivist, objects are much more nuanced than that.

There is a school of thought that treats scientific ontology, the set of objects that one thinks exists in a domain, as basically the bounded variables of the best theory of that domain: i.e., Quine’s “to be is to be the value of a variable”, and the subsequent development of that view (Quine 1948). In this case, species would be theoretical objects if they were such variables of a theory. But as they aren’t, we need to establish what sort of ontological status they, and other phenomenal non-theoretic objects, may have.

Consider planetary orbits. They were observed and debated for a very long time before Newton proposed a general physics that accounted for them (and made predictions about them). But in so doing, Newton demoted these orbits from theoretically important objects to special cases of larger and more universal physics. “Planetary orbit” is a special kind of astrophysical dynamics, one which aperiodic comets, entire star systems, and even entire galaxies all obey. Even if no orbits actually existed (and we can perhaps envisage this in some universe) under this physics, the movements of objects would be still covered by Newtonian dynamics.

Likewise species. They obey, and when they occur are post hoc explained by, the biology of populations, interbreeding, selection, drift, and so on. But they are not themselves theoretical objects, any more than planetary orbits are in physics. [I know – the “nebular hypothesis” of planetary formation requires that bodies in the disk of accretion form more massive objects by gravitational attraction, but if there were no such disks, they needn’t.] Species occur, and are explicable in a multiplicity of ways, but they do not follow formally from any theory of biology.

This paper’s general characterization of species is that they are the nexus of the coalescence of genes, haplotypes, parent-child lineages and so on, at or about the same level; they are polyphasic. In abstract terms, species are these coalescences that are distinct from other such coalescences, something I have called the synapomorphic conception of species (Wilkins 2003), and each and every one has a general set of properties and modes of speciation, and a unique set of these that only they have (the synapomorphies, or shared characters, which are causally active in maintaining separation) [5]. Because each species is a unique historical event, that makes the modality as I have called it, of each species, something that they will tend to share only with those taxa they are closely related to, just as liberal democracies resemble each other because they tend to be derived from a common source.

As a result, the being of a species, its modality, is something that is as much an evolved trait as having a vertebral column or a nuclear membrane. The causal process whereby a species evolved and is maintained is going to be something that depends upon shared ancestral traits such as developmental machinery, genetic sequences, and ecological resources. While these may be very similar between related species, they cannot be expected to be the same in each case – not all Rhagoletis species will speciate by host race transfer, for instance – and so each species will have a unique set of causes. No theory will capture all and only these causes (not every sexual species will be caused by allopatric isolation, and not every asexual species will be caused by a single niche adaptation) except at a level of generality that is so vague as to exclude explanation in terms of mechanisms. That is why I say species aren’t theoretical objects, but are phenomenal ones. As a take-home exercise to the reader, try to imagine under which conditions organisms like ours wouldn’t form species at all.

Theoretical objects

There have been several proposals for what makes an object “theoretical” [6]. To begin with, calling an object theoretical is not to cast doubt upon its reality. That is the old conflation of vernacular uses of the word “theory” (and possibly also cultural uses) with the scientific sense. To avoid that confusion let me note that to call something a “theory” in science is to give it the highest possible status as a concept or explanation.

But within philosophy, there seem to be a few major views on the matter. So far as I have encountered them, let me list and explicate them. The first is that of Quine. Something exists just to the extent that our best theory of a given domain requires them. On this view, species are simply not theoretical, and indeed do not exist, because if I am right that no theory of biology requires species, then they are never the value of a bound variable in any model of biology.

There is another, similar, but not so restrictive view: that of “Ramseyfication”. On this account, what a theory requires is based on a formalisation – a “Ramsey sentence” (Psillos 2000; Psillos 2006; also called a “Carnap-Ramsey sentence” or a “Ramsey-Lewis sentence”, see Koslow 2006) – of the theory. Objects are held to exist so long as they are represented either by primitive terms (values of variables, or constants) of the theory or combinations or derivations of those. A primitive here might be something empirical, so that species might be primitives of biological theory, but are not themselves explained by it. I think this is not the case with species, because in every such case of which I am aware, one can replace “the species X y” used with something like “a local population of X y” or “organisms that behave in such a way, which is typical of X y” for functional accounts such as ecological ones. In other words the species X y is replaceable with objects that the theory actually employs. The Ramsey approach, recently also called the “Canberra Plan” (Braddon-Mitchell and Nola 2009; Jackson 1998), treats these objects as non-objects. Sometimes this is played out as “Structural Realism” in which a theory as a structure is true (Psillos 2000, 1999; Psillos 2006), but the objects it poses which are “unobservable” may or may not be real, so long as the theory is empirically adequate in other ways. This is irrelevant here.

So, my question is this: what makes an object theoretical and are there other roles objects and their representation play in science? For my view to work, it must be that there are objects that are described by the theory, which in the domain of that theory have a certain coherence or unity as objects. Mechanisms, like geology’s tectonic drift, are obviously theoretical in that sense. But mountains are more difficult. Mountains are real things, but the category as a whole lacks theoretical coherence. That is, a mountain has no theoretical place qua “mountain”, but as a particular mountain, say, Mount St Helens or the Matterhorn, it calls for explanation [7]. I call these phenomenal objects. Like sand dunes they are real things – if you have to map them, travel around them, or climb over them, they are as real as anything can be [8], but the choice of demarcation between peaks can be conventional or even just something that perception hands to us on a plate. Nothing in theory demands that this particular mountain exists, nor even that there are mountains. On a planet with no tectonics, after a reasonable period, there may be no mountains.

Species are like that. They are phenomenal objects, real facts about the world, which we perceive rather than define. Of course, this makes them relative in a way to the rules and capacities of perception. If we have poor vision, we might not “perceive” mountains until we had telescopic surveyor’s sights. Once we have that technology, though (which, note, doesn’t rely upon the theories of geology) we do see mountains. Similarly, we may need to use all kinds of assay techniques to see species, but when we have them they are seen.

An example that is instructive is the discovery by Murray Littlejohn and his advisor of the different species that had previously been called Rana pipiens, the “leopard frog” of the southern United States (Littlejohn and Oldham 1968). The leopard frog is a widespread species, and Littlejohn was using a new piece of equipment designed for speech therapy – the sonograph – to graph the mating calls of these frogs. He discovered that there were a number – up to six (although new species have since been identified: Platz 1993) – distinct mating calls. Since mating calls in amphibians are highly species-particular, Littlejohn proposed that this was in fact a species complex, in which morphology and ecology were indistinguishable, but that mating was restricted within the mating call groups, and these were species. Subsequent work proved this to be the case. The differentiation was always there, but you needed the right assay technique.

This is not species being “constructed” or any other bad “postmodern” nonsense. While the concept we have of those species is being constructed (and reconstructed as new evidence comes in), the concept refers to, or denotes, realities, and those realities are either of classes of things that are theoretical, such as populations, haplotypes, genes, developmental sequences or cycles, and so on, or of things that are not required by the theory. Such are species. When we construct a concept, we are learning about the things we describe. It’s like finding that Everest has a hitherto-hidden peak that is even higher. Our concept of Everest changes, but the thing itself was already as it is.

The implication here is that theories are not all that is going on in science. Objects in a domain exist prior to the demarcation of the domain, or are not objects we infer from the theory – traditional philosophy of science (and by implication of language) has ignored three quarters of what is going on in science. Classification and passive observation still occur. So that’s why I say species aren’t theoretical objects, but are phenomenal ones.

So, what are species?

First of all we must take issue with the entire way the debate has been framed over the past 150 years or so and assert: There is only one species concept (Wilkins 2011). That is to say, there is only one concept that we are all trying to define in many ways, according to both our preferred theories of how species come into being and maintain themselves over evolutionary time, and what happens to be the general case for the particular group of organisms we have in our minds when we attempt our definitions. The former case is what we might call theoretical conceptions of species, where a “conception” is a definition of the word and concept of species. The latter are the prototypical conceptions of species. If you work in, say, fishes, then your conception of species has to deal with the usual facts about fishes (Rosen 1978). If you are a fern botanist, then those organisms set up your prototype (Wagner 1983). And the debate over what species are has been driven by differing prototypes as much as by different theories of speciation.

Elsewhere (Wilkins 2009), I list some 26 conceptions [9] of species in the modern (post-Synthesis) literature, which cites the various proponents and their original publication. I am going to focus now on the few basic ideas that underlie nearly all of these. The first concept is based on reproductive isolation.

Since the Synthesis of genetics and Darwinian evolution was formed, the ruling notion of species generation (speciation) was based on the criterion of sexual populations that are isolated from each other, so that they evolve in divergent ways, leading to populations that, when they meet, if they do, in the same range, they no longer tend to interbreed, and their gene pools are now distinct over evolutionary time scales.

The conception of species that the Synthesis adopted as a result of this genetic-evolutionary view is sometimes misleadingly called the biological species concept (or BSC). It is called this because it was contrasted to the practices of museum taxonomists, who identified species based on differences in the morphology of captured or collected specimens. This was held to be a sterile methodology where the data was more in the heads of the taxonomists than in the real world. Hence, the BSC was biological, while the museum approach was conventional (due to the conveniences of the taxonomists). But the leading idea of the BSC is not that things live, or that they are in messy populations, although that is part of it, but rather that these populations are reproductively isolated from each other. So I prefer to call this conception the Reproductive Isolation Species Conception (RISC), or “isolationist” conception for short. There are several versions of it, but the basic idea – that something inhibits interbreeding when they meet – is common to them all.

Criticisms of the RISC began early. For as start, it was observed that there was a disconnection between the theoretical justification for the RISC, and the ways in which taxonomists who adopted it did their taxonomy. To be sure that you have a RISC taxon, you really need to do breeding experiments to be sure. Many quite diverse morphs in, say, butterflies, that were identified as distinct species in the 19th century, turned out to be different genders of the same species. “Aha!” said the isolationists, “This is a failure of the morphologists.” But when similar cases occurred and were found to be different genders before the Synthesis, these so-called “morphologists” had no problem with making them the same species on that ground. It was understood that form was only a guide to the underlying biological reality, not an end in itself. Worse, the isolationists themselves used morphology to identify their species. Breeding experiments, even when they are technically possible, take enormous time and resources, which nearly all the time we don’t have. So while theoretically isolationists are basing their work on reproductive isolation, practically they are doing just what their supposed mistaken predecessors did. This might lead us to think that the older workers weren’t so silly after all. RISCs include genetic clustering accounts, “lock-and-key” mating system accounts, and so on (Coyne and Orr 2004).

The second of our broad classes of conceptions of species is based on ecological isolation, and is often called the Ecological Species Conception. This goes back in one form or another to Linnaeus. However, it got currency in modern times when another Swedish botanist named Göte Turesson did some studies during the 1920s of plant morphologies in different ecological conditions (Turesson 1922, 1925, 1929). Turesson coined the term ecotype to describe these differing morphologies. He distinguished between ecotypes and ecospecies, which were populations prevented by adaptation to a particular ecological niche from interbreeding. In the 1970s, Leigh Van Valen offered a new version, based on the fact that American oaks will freely interbreed, but that the ecological types remain constant (Van Valen 1976). So, he concluded, in these cases, the “species” is effectively maintained by the ecological niche. Similar cases are common in plants, and less so among animals. Bacteria and other single celled organisms which do not often exchange genes may be entirely maintained by this. Lacking sex, they cannot be RISC species, and Turesson coined another term for them, agamospecies (meaning, sexless species). In animals asexual reproduction has evolved from sexual species several times, and are called parthenogens (”virgin origins”), while in plants, it is much more common and they are called apomicts (”Apart from mixing”).

The third kind of species conception is known variously as Morphological, Typological or Essentialist, but all these are misleading. Sometimes it is called the Linnaean Conception, because it is supposed to be the default view before genetics and evolution were discovered, and hence the view of Linnaean taxonomy. This is a bit unfair – Linnaeus never clearly defined a species concept [11], and the standard view at the time was that of John Ray, in which a species was twofold – a form, which is reproduced. This morphological conception was never isolated from normal reproduction by parents. Moreover, Linnaean and Rayesque species were not defined by essences either, but that’s for another time. The important thing was that it was the overall organization of the organisms that defined them as a species, so long as it was reproduced. Ray’s own “definition” was “the distinguishing features that perpetuate themselves in propagation from seed”  (italics added; quoted in Mayr 1982: 256).

Ray’s definition was designed to cover plants, but he explicitly extended it to animals, and it was the first time any biologist had ever given a purely biological definition of “species”. It was not based on Aristotle or any logical system, but observation. This earlier definition remained the standard view at the time Darwin began his work, via the authority of Baron Cuvier. A variant of this is sometimes called the phenetic view of species – based on the “overall similarity” of the reproduced form, phenetics (from the Greek ???????, “manifest”) was an attempt to delimit species without using any theoretical methods or concepts. It failed, because there are too many ways to measure similarity, and they don’t all coincide, not even often.

A fourth general class of species conception is one based on the convenience of biological work, including mutual communication between specialists. It is called (wrongly) species nominalism, and more accurately species conventionalism. It is the view that, as Locke had said, species are made for communication, and nothing else. Darwin once wrote ironically to a friend that he had at last found a definition of species from a taxonomist: “Any form that a taxonomist has given a name to!” Of course, Darwin didn’t believe that about species. For him they were real but temporary things, and he believed there was no special rank or level in biology that was unique to species. Contrary to common opinion since the turn of the 20th Century (and earlier, vide Agassiz) Darwin was not a conventionalist, but evolutionary thinking made it harder to be exact about species.

This leads us to our final conception: based on evolutionary history, it has two main versions: the phylogenetic species conceptions based on cladistics, and the so-called evolutionary species concepts, which are often a mixture of the RISC, the ecological species conception, and phylogenetic accounts of reconstructed history. The former are often more like the RISC, because they rely on there being separation of lineages over large time as defined by their sharing or not evolved traits, and this implies genetic isolation. The latter do not rely on RISC, but only that after the fact the lineages remain distinct for whatever reason (thus admitting ecotypes and ecospecies).

These conceptions are process-based, and are equally as non-operational as the RISC, but cladistics at least has a large number of mathematical and analytic techniques for drawing up their cladograms. The problem is that, without some way of saying what the level of separation is for species, cladistics can divide lineages up to a very small level (such as haplotype groups), leading to “taxonomic inflation” (Isaac, Mallet, and Mace 2004; Padial and De la Riva 2006). Phylogenetic species can be as much as nine or ten times in number compared to the ordinary (“Linnaean”) kind. The debate rages through the modern systematics community.

So, after all that, what is a species? I think, and this is very much an eccentric view, that there is no single thing that species are (Wilkins 2010). Species modality depends on the group of organisms under investigation – there may be a mode of being a RISC species in birds, for example, where the sex that has the sex determining chromosome are the females, not the males as in mammals, which is different from being a RISC species in another group such as monotremes (Grutzner et al. 2004; Rens et al. 2004; Veyrunes et al. 2008; Warren et al. 2008). Any universal concept of species has to range over the entire evolutionary tree, but the modes of being a species will depend on what ways they have evolved to remain distinct from each other. Hence, none of the particular conceptions are sufficient or necessary to cover being a species in all organisms. However, that only tells us what species sometimes are. It doesn’t tell us why these different things should even be called “species”. For example, RISC proponents will often say that asexual organisms (agamospecies) aren’t really species at all; because they lack the defining properties of species which is, of course, reproductive isolation (Fisher 1930: 135; Dobzhansky 1951: 275; Mayr 1942: 122). So we should call them something else – agamospecies, quasispecies, pseudospecies, paraspecies, etc. This has an unwanted consequence – the bulk of life doesn’t exist in species, but only those few clades that happened to evolve sex do. Instead we should say that all organisms come in “kinds”, some of which are sexual kinds. Others come in genetic bundles or are clustered for ecological reasons, and many are a mixture.

So here is a “definition” of the word “species”: A species is any lineage of organisms that is distinct from other lineages because of differences in some shared biological property. It has to be a lineage, to distinguish biological species (but not just RISC species) from species of chemical compounds, minerals, and diseases. However, while all species are lineages, not all lineages are species, not even the monophyletic ones. It has to be a causal definition, because formal approaches such as the one Agassiz was fond of do no explanatory work (in short, the formalist definition merely restates that there are differences). And it has to be based on biological properties, because nonbiological properties like range or geography are not enough to include or exclude populations and organisms from a species. All the various conceptions of the concept try to give the differences in shared biological properties some detail – differences in sexual reproductive mechanisms, differences in genetic structure, differences in ecological niche adaptation, and so on. And when we look at them that way, it becomes clear why none of them are sufficient or necessary for all species: the mechanisms that keep lineages distinct evolved uniquely in every case, and so generalizations only cover some, not all, of life.

A species is a lineage (that is, an ancestor-descendant ensemble of populations over time). If it’s distinct as a lineage, then it’s a species. Of course, not all lineages are species – gene, haplotype and population lineages exist, for example – so the point at which lineages coalesce into different kinds of species is not something that we can define abstractly. Instead, it is a phenomenon that we observe, and seek to explain with one of the 26 or so conceptions in each case.


Taking this approach makes sense of several facts about biological science. It explains why we recognized species well in advance of there being anything remotely like a theoretical explanation of them, from the sixteenth century onwards. John Ray formally defined biological species for the first time in 1686, but his view was implicit in the work of natural historians going back to Aristotle and Theophrastus. Genetic and developmental accounts of species did not arise until around 1900.

It explains why when replacement terms are proposed for species, they tend to settle on the same sorts of phenomena, and eventually species makes a comeback. It also explains why it is that when autochthonous peoples employ organisms economically, say by hunting or raising them, they recognize the same sorts we do for scientific reasons (Atran 1985). These things are phenomenally salient if you have to interact with them.

But most of all it explains something about science, and I’d like to briefly sketch what I think are the implications of phenomenal objects in the ontology of a science. In the traditional view of science, observation is theory-dependent and objects are theoretical as I have described above. I am proposing that some objects are not theory-dependent. In doing so I can explain why it is that so much of biology is what Rutherford sneeringly called “stamp collecting”. Before you can begin to formulate theories, you have to gather together the objects under explanation and organize that information into a taxonomy, otherwise it is not even clear what the domain of the theory is. The traditional view of science of the twentieth century ignored classificatory activities as uninteresting; I am suggesting it is one of the crucial and essential aspects of a science. This has been hidden to some extent by focusing on theory-dependence.

One might object that of course these objects are theoretical: to observe them is to identify a difference by measurement, and that implies an assay or methodological protocol. This is usually true, although species and mountains do not need much if any theoretical ancillary assumptions. But the point is that they do not need the theory under investigation in order to be phenomenal objects. That is, if they are theory-dependent, they are dependent on theories outside the domain in question. Moreover, they are often tokens of a class of phenomenal objects that call for explanation in those theories as well (consider optical theories, or genomic clusters in genetic theories).

Since the dependence here is a general kind (such as for optics), the theory-dependence is benign. With respect to our theory T, there is no special dependence on which the observations are being made, so the phenomena are T-independent. This doesn’t mean there is such a thing as completely naive observation – nobody ever starts from total naivety or from a tabula rasa. Even observers in the mountains of Papua New Guinea are informed by prior ideas and experience. But we can say they observe species, and do not thereby need to define them.

To summarize, the following claims are being made here: 1. A species is something that forms phenomenal, salient, lineages of populations of organisms and genes; 2. A species can have a particular mode based on evolved biological properties; 3. The species conception applied in each case depends on whether that species meets the conditions for that conception; and 4. Each species is a phenomenon that calls for a conception and an explanation. So we don’t need to have a monistic or singular definition of species, because species are things to be explained, they are explicanda, not an a priori category or rank into which every biological organism must be fitted.



1. This is not to deny that species is also maintained by conventional and social practices. If entire volumes are dedicated to describing species, anyone who wishes to be taken seriously in that field has to refer to those described objects.

2. For example Arnold 1992, Barrington, et al. 1989, Bergman and Beehner 2004, Birkhead and Balen 2007, Cortes-Ortiz, et al. 2007, Detwiler 2003, Muir, et al. 2000, Greig, et al. 2002, Jolly, et al. 1997, Knobloch 1959, Lotsy 1916, Mallet 2007, 2008, Arnold and Meyer 2006, Rieseberg and Willis 2007, Salzburger, et al. 2002, Dowling and Secor 1997, Wagner 1983.

3. This was the view of John Maynard Smith as well (Maynard Smith 1958).

4. Gal Kober (in a talk in July 2009) suggested that species are a fourth alternative: units of classification. This is consistent with my third alternative unless one thinks, as Kober does, that classification is a theoretical operation.

5. This may seem like essentialism, but the point is merely that if they did not have these properties we would not even notice them as species.

6. See Brittan 1986, Ladyman 1998, French and Ladyman 2003.

7. Neil Thomason informs me that H. P. Grice in a seminar criticized Quine’s view on bounded objects by remarking “Quine thinks we can’t count the mountains in the Rockies”. However, he never published this so far as I can find, and he may have been referring to vague objects or Sorites rather than a theory–phenomena distinction about objects.

8. Echoing Hacking’s comment about electrons, that if you can spray them, they are real (Hacking 1983: 23). To avoid unnecessary metaphysical concerns, I will say they are real enough – that is, as real as anything else in biology.

9. A conception is a variant definition of a concept. This in turn is distinct from the various formulations of conceptions of the concept. For example Lherminer and Solignac 2000 give several hundred formulations, but mostly these are of a much fewer number of conceptions.

10. The historical account of this terminology, which derives from Ernst Mayr, is given in my book (Wilkins 2009).

11. Although he did use an interbreeding criterion (Müller-Wille and Orel 2007), and so may be charitably interpreted as holding a RISC.


Thanks to Gal Kober, Marc Ereshefsky, Quentin Wheeler and Brent Mishler for pushing me on this. I must also acknowledge all the specialists with whom I have had long, often frustrating for us both, arguments on this topic leading to my 2009 book and the PhD it was based upon. In particular I must thank David Williams of the NHM, Malte Ebach, Gareth Nelson and Brent Mishler, none of whom can be fairly thought to agree with me. Ingo Brigandt first made the claim that species are not theoretical objects (Brigandt 2003), to my knowledge.


Agapow, P. M., O. R. Bininda-Emonds, K. A. Crandall, J. L. Gittleman, G. M. Mace, J. C. Marshall, and A. Purvis. 2004. The impact of species concept on biodiversity studies. Q Rev Biol 79 (2):161-179.

Agassiz, Louis. 1860. [Review of] On the Origin of species. American Journal of Science and Arts (Ser. 2) 30:142-154.

Ankeny, Rachel A. 2000. Fashioning Descriptive Models in Biology: Of Worms and Wiring Diagrams. Philosophy of Science 67 (Supplement. Proceedings of the 1998 Biennial Meetings of the Philosophy of Science Association. Part II: Symposia Papers):S260-S272.

Arnold, Michael L. 1992. Natural Hybridization as an Evolutionary Process. Annual Review of Ecology and Systematics 23:237-261.

Arnold, Michael L., and Axel Meyer. 2006. Natural hybridization in primates: One evolutionary mechanism. Zoology 109 (4):261-276.

Atran, Scott. 1985. The early history of the species concept: an anthropological reading. In Histoire du Concept D’Espece dans les Sciences de la Vie. Paris: Fondation Singer-Polignac:1-36.

Barrington, D. S., C. H. Haufler, and C. R. Werth. 1989. Hybridization, reticulation, and species concepts in the ferns. American Fern Journal 79 (2):55-64.

Bergman, Thore J., and Jacinta C. Beehner. 2004. Social System of a Hybrid Baboon Group (Papio anubis × P. hamadryas). International Journal of Primatology 25 (6):1313-1330.

Berlocher, Stewart H. 1999. Host race or species? Allozyme characterization of the ‘flowering dogwood fly’, a member of the Rhagoletis pomonella complex. Heredity 83 (Pt 6):652-662.

———. 2000. Radiation and divergence in the Rhagoletis pomonella species group: inferences from allozymes. Evolution Int J Org Evolution 54 (2):543-557.

Berlocher, Stewart H., and J. L. Feder. 2002. Sympatric speciation in phytophagous insects: moving beyond controversy? Annu Rev Entomol 47:773-815.

Birkhead, T. R., and S. Van Balen. 2007. Unidirectional hybridization in birds: an historical review of bullfinch (Pyrrhula pyrrhula) hybrids. Archives of Natural History 34 (1):20-29.

Braddon-Mitchell, David, and Robert Nola. 2009. Conceptual analysis and philosophical naturalism. Cambridge, MA: MIT Press.

Brigandt, Ingo. 2003. Species pluralism does not imply species eliminativism. Philosophy of Science 70 (5):1305-1316.

Brittan, Gordon G. . 1986. Towards a Theory of Theoretical Objects. PSA: Proceedings of the Biennial Meeting of the Philosophy of Science Association 1986:384-393.

Claridge, Michael F., H. A. Dawah, and M. R. Wilson. 1997. Species: the units of biodiversity. London; New York: Chapman & Hall.

Cortes-Ortiz, Liliana, Thomas F. Duda, Domingo Canales-Espinosa, Francisco Garcia-Orduna, Ernesto Rodriguez-Luna, and Eldredge Bermingham. 2007. Hybridization in Large-Bodied New World Primates. Genetics 176 (4):2421-2425.

Coulson, T., E. A. Catchpole, S. D. Albon, B. J. T. Morgan, J. M. Pemberton, T. H. Clutton-Brock, M. J. Crawley, and B. T. Grenfell. 2001. Age, Sex, Density, Winter Weather, and Population Crashes in Soay Sheep. Science 292 (5521):1528-1531.

Coyne, Jerry A., and H. Allen Orr. 2004. Speciation. Sunderland, Mass.: Sinauer Associates.

Darwin, Charles Robert. 1888. The life and letters of Charles Darwin: including an autobiographical chapter. Edited by F.Darwin. Lond.: Murray.

de Queiroz, Kevin. 2005. Different species problems and their resolution. BioEssays 27 (12):1263-1269.

———. 2007. Species Concepts and Species Delimitation. Systematic Biology 56 (6):879-886.

Detwiler, Kate M. 2003. Hybridization between Red-tailed Monkeys (Cercopithecus ascanius) and Blue Monkeys (C. mitis) in East African Forests. In The Guenons: Diversity and Adaptation in African Monkeys:79-97.

Dobzhansky, Theodosius. 1951. Genetics and the origin of species. 3rd rev. ed. New York: Columbia University Press.

Dowling, Thomas E., and Carol L. Secor. 1997. The role of hybridization and introgression in the diversification of animals. Annu. Rev. Ecol. Syst. 28:593–619.

Fisher, Ronald Aylmer. 1930. The genetical theory of natural selection. Oxford UK: Clarendon Press,  (rev. ed. Dover, New York, 1958).

French, Steven, and James Ladyman. 2003. Remodelling Structural Realism: Quantum Physics and the Metaphysics of Structure. Synthese 136-141 (1):31-56.

Gannett, Lisa. 2003. Making populations: Bounding genes in space and in time. Philosophy of Science 70 (5):989-1001.

Gavrilets, Sergey. 2004. Fitness landscapes and the origin of speciesMonographs in population biology; v. 41. Princeton, N.J.; Oxford, England: Princeton University Press.

Gayon, Jean. 1996. The individuality of the species: A Darwinian theory? – from Buffon to Ghiselin, and back to Darwin. Biology and Philosophy 11:215-244.

Ghiselin, Michael T. 1969. The triumph of the Darwinian method. Berkeley: University of California Press.

Godfrey-Smith, Peter. 2009. Darwinian populations and natural selection. Oxford: Oxford University Press.

Gould, Stephen Jay, and Elisabeth A. Lloyd. 1999. Individuality and adaptation across levels of selection: how shall we name and generalize the unit of Darwinism? Proc Natl Acad Sci U S A 96 (21):11904-11909.

Grandy, Richard. 2008. Sortals. In The Stanford Encyclopedia of Philosophy, edited by E. N. Zalta.

Grantham, Todd. 1995. Hierarchical approaches to macroevolution – recent work on species selection and the Effect Hypothesis. Annual Review of Ecology and Systematics 26:301–321.

Greig, D., E. J. Louis, R. H. Borts, and M. Travisano. 2002. Hybrid speciation in experimental populations of yeast. Science 298 (5599):1773-1775.

Griffiths, Paul  E, and Karola Stotz. 2007. Gene. In Cambridge Companion to Philosophy of Biology, edited by M. Ruse and D. Hull. Cambridge: Cambridge University Press:85-102.

Griffiths, Paul  E, and John S. Wilkins. In Press. When do evolutionary explanations of belief debunk belief? In Darwin in the 21st Century: Nature, Humanity, and God, edited by P. R. Sloan. Notre Dame, IN: Notre Dame University Press.

Grutzner, Frank, Willem Rens, Enkhjargal Tsend-Ayush, Nisrine El-Mogharbel, Patricia C. M. O’Brien, Russell C. Jones, Malcolm A. Ferguson-Smith, and Jennifer A. Marshall Graves. 2004. In the platypus a meiotic chain of ten sex chromosomes shares genes with the bird Z and mammal X chromosomes. Nature 432 (7019):913-917.

Hacking, Ian. 1983. Representing and intervening: introductory topics in the philosophy of natural science. Cambridge UK: Cambridge University Press.

Haldane, J. B. S. 1930. A mathematical theory of natural and artificial selection (Part VI. Isolation). Proc. Cambr. Philos. Soc 26:220-230.

Hey, Jody. 2001. The mind of the species problem. Trends in Ecology & Evolution 16 (7):326-329.

Hey, Jody, Robin S. Waples, Michael L. Arnold, Roger K. Butlin, and Richard G. Harrison. 2003. Understanding and confronting species uncertainty in biology and conservation. Trends in Ecology & Evolution 18 (11):597-603.

Isaac, N. J. B., J. Mallet, and G. M. Mace. 2004. Taxonomic inflation: its influence on macroecology and conservation. Trends in Ecology & Evolution 19 (9):464-469.

Jablonski, David. 2008. Species Selection: Theory and Data. Annual Review of Ecology, Evolution, and Systematics 39 (1):501–524.

Jackson, Frank. 1998. From metaphysics to ethics: a defence of conceptual analysis. Oxford; New York: Clarendon Press.

Jolly, Clifford J., Tamsin Woolley-Barker, Shimelis Beyene, Todd R. Disotell, and Jane E. Phillips-Conroy. 1997. Intergeneric Hybrid Baboons. International Journal of Primatology 18 (4):597-627.

Knobloch, Irving W. 1959. A Preliminary Estimate of the Importance of Hybridization in Speciation. Bulletin of the Torrey Botanical Club 86 (5):296-299.

Koslow, Arnold. 2006. The Representational Inadequacy of Ramsey Sentences. Theoria 72 (2):100–125.

Ladyman, James. 1998. What is structural realism? Studies In History and Philosophy of Science Part A 29 (3):409-424.

Lande, Russell. 1981. Models of speciation by sexual selection on polygenic traits. Proc Natl Acad Sci U S A 78 (6):3721–3725.

Lherminer, Philippe, and Michel Solignac. 2000. L’espèce: définitions d’auters. Sciences de la vie:153-165.

Littlejohn, Murray J., and R. S. Oldham. 1968. Rana pipiens complex: mating call structure and taxonomy. Science 162:1003-1005.

Lloyd, Elisabeth Anne, and Stephen Jay Gould. 1993. Species selection on variability. Proc Natl Acad Sci U S A 90(2):595-599.

Lotsy, J. P. 1916. Evolution by means of hybridization. The Hague: Martinus Nijhoff.

Mallet, James. 2007. Hybrid speciation. Nature 446 (7133):279-283.

———. 2008. Hybridization, ecological races and the nature of species: empirical evidence for the ease of speciation. Philosophical Transactions of the Royal Society B: Biological Sciences 363 (1506):2971-2986.

Maynard Smith, John. 1958. The theory of evolution. Harmondsworth, UK: Penguin Books.

Mayr, Ernst. 1942. Systematics and the origin of species from the viewpoint of a zoologist. New York: Columbia University Press.

———. 1970. Populations, species, and evolution: an abridgment of Animal species and evolution. Cambridge MA: The Belknap Press of Harvard University Press.

———. 1982. The growth of biological thought: diversity, evolution, and inheritance. Cambridge MA: The Belknap Press of Harvard University Press.

Millstein, Roberta L. 2006. Natural Selection As a Population-Level Causal Process. British Journal for the Philosophy of Science 57 (4):627-653.

Mishler, Brent D. 1999. Getting rid of species? In Species, New interdisciplinary essays, edited by R. A. Wilson. Cambridge, MA: Bradford/MIT Press:307-315.

Muir, Graham, Colin Fleming, and Christian Schlötterer. 2000. Species status of hybridizing oaks. Nature 405:1016.

Müller-Wille, Staffan, and Vitezslav Orel. 2007. From Linnaean Species to Mendelian Factors: Elements of Hybridism, 1751-1870. Annals of Science 64 (2):171-215.

Padial, José, and Ignacio De la Riva. 2006. Taxonomic Inflation and the Stability of Species Lists: The Perils of Ostrich’s Behavior. Systematic Biology 55 (5):859-867.

Platz, J. E. 1993. Rana subaquavocalis, a Remarkable New Species of Leopard Frog (Rana pipiens Complex) from Southeastern Arizona That Calls under Water. Journal of Herpetology 27 (2):154-162.

Pleijel, Frederik, and G. W. Rouse. 2000. Least-inclusive taxonomic unit: a new taxonomic concept for biology. Proceedings of the Royal Society of London – Series B: Biological Sciences 267 (1443):627-630.

Psillos, Stathis. 1999. Scientific Realism: How Science Tracks Truth. London, New York: Routledge.

———. 2000. Carnap, the Ramsey-Sentence and Realistic Empiricism. Erkenntnis 52 (2):253.

Psillos, Stathos. 2006. Ramsey’s Ramsey-sentences. In Cambridge and Vienna: Frank P Ramsey and the Vienna Circle, edited by M. C. Galavotti. New York: Springer International:67-90.

Quine, Willard Van Orman. 1948. On What There Is. Review of Metaphysics 2 (5):21-38.

Rens, Willem, Frank Grützner, Patricia C. M. O’Brien, Helen Fairclough, Jennifer A. M. Graves, and Malcolm A. Ferguson-Smith. 2004. Resolution and evolution of the duck-billed platypus karyotype with an X1Y1X2Y2X3Y3X4Y4X5Y5 male sex chromosome constitution. Proceedings of the National Academy of Sciences of the United States of America 101 (46):16257-16261.

Rice, Sean H. 1995. A genetical theory of species selection. Journal of Theoretical Biology 177 (3):237-245.

Rieseberg, Loren H., and John H. Willis. 2007. Plant Speciation. Science 317 (5840):910-914.

Rosen, Donn E. 1978. Vicariant patterns and historical explanation in biogeography. Systematic Zoology 27:159-188.

Salzburger, W., S. Baric, and C. Sturmbauer. 2002. Speciation via introgressive hybridization in East African cichlids? Mol Ecol 11 (3):619-625.

Sneath, P. H. A., and Robert R. Sokal. 1973. Numerical taxonomy: the principles and practice of numerical classificationA Series of books in biology. San Francisco: W. H. Freeman.

Sokal, Robert R., and P. H. A. Sneath. 1963. Principles of numerical taxonomyA Series of books in biology. San Francisco,: W. H. Freeman.

Stidd, Benton M., and David L. Wade. 1995. Is species selection dependent upon emergent characters? Biology and Philosophy 10:55-76.

Turesson, Göte. 1922. The species and variety as ecological units. Hereditas 3:10-113.

———. 1925. The plant species in relation to habitat and climate. Hereditas 6:147-236.

———. 1929. Zur natur und begrenzung der artenheiten. Hereditas 12:323-334.

Van Valen, L. 1976. Ecological species, multispecies, and oaks. Taxon 25:233-239.

Vandamme, P., B. Pot, M. Gillis, P. de Vos, K. Kersters, and J. Swings. 1996. Polyphasic taxonomy, a consensus approach to bacterial systematics. Microbiol Rev. 60 (2):407–438.

Veyrunes, Frédéric, Paul D. Waters, Pat Miethke, Willem Rens, Daniel McMillan, Amber E. Alsop, Frank Grützner, Janine E. Deakin, Camilla M. Whittington, Kyriena Schatzkamer, Colin L. Kremitzki, Tina Graves, Malcolm A. Ferguson-Smith, Wes Warren, and Jennifer A. Marshall Graves. 2008. Bird-like sex chromosomes of platypus imply recent origin of mammal sex chromosomes. Genome Research 18 (6):965-973.

Vos, Michiel. A species concept for bacteria based on adaptive divergence. Trends in Microbiology In Press, Corrected Proof.

Wagner, Warren H. 1983. Reticulistics: The recognition of hybrids and their role in cladistics and classification. In Advances in cladistics, edited by N. I. Platnick and V. A. Funk. New York: Columbia Univ. Press:63-79.

Warren, Wesley C., LaDeana W. Hillier, Jennifer A. Marshall Graves, Ewan Birney, Chris P. Ponting, Frank Grützner, Katherine Belov, Webb Miller, Laura Clarke, Asif T. Chinwalla, Shiaw-Pyng Yang, Andreas Heger, Devin P. Locke, Pat Miethke, Paul D. Waters, Frédéric Veyrunes, Lucinda Fulton, Bob Fulton, Tina Graves, John Wallis, Xose S. Puente, Carlos López-Otín, Gonzalo R. Ordóñez, Evan E. Eichler, Lin Chen, Ze Cheng, Janine E. Deakin, Amber Alsop, Katherine Thompson, Patrick Kirby, Anthony T. Papenfuss, Matthew J. Wakefield, Tsviya Olender, Doron Lancet, Gavin A. Huttley, Arian F. A. Smit, Andrew Pask, Peter Temple-Smith, Mark A. Batzer, Jerilyn A. Walker, Miriam K. Konkel, Robert S. Harris, Camilla M. Whittington, Emily S. W. Wong, Neil J. Gemmell, Emmanuel Buschiazzo, Iris M. Vargas Jentzsch, Angelika Merkel, Juergen Schmitz, Anja Zemann, Gennady Churakov, Jan Ole Kriegs, Juergen Brosius, Elizabeth P. Murchison, Ravi Sachidanandam, Carly Smith, Gregory J. Hannon, Enkhjargal Tsend-Ayush, Daniel McMillan, Rosalind Attenborough, Willem Rens, Malcolm Ferguson-Smith, Christophe M. Lefèvre, Julie A. Sharp, Kevin R. Nicholas, David A. Ray, Michael Kube, Richard Reinhardt, Thomas H. Pringle, James Taylor, Russell C. Jones, Brett Nixon, Jean-Louis Dacheux, Hitoshi Niwa, Yoko Sekita, Xiaoqiu Huang, Alexander Stark, Pouya Kheradpour, Manolis Kellis, Paul Flicek, Yuan Chen, Caleb Webber, Ross Hardison, Joanne Nelson, Kym Hallsworth-Pepin, Kim Delehaunty, Chris Markovic, Pat Minx, Yucheng Feng, Colin Kremitzki, Makedonka Mitreva, Jarret Glasscock, Todd Wylie, Patricia Wohldmann, Prathapan Thiru, Michael N. Nhan, Craig S. Pohl, Scott M. Smith, Shunfeng Hou, Marilyn B. Renfree, Elaine R. Mardis, and Richard K. Wilson. 2008. Genome analysis of the platypus reveals unique signatures of evolution. Nature 453 (7192):175-183.

Wheeler, Quentin D., and Norman I. Platnick. 2000. The phylogenetic species concept (sensu Wheeler and Platnick). In Species concepts and phylogenetic theory: A debate, edited by Q. D. Wheeler and R. Meier. New York: Columbia University Press:55-69.

Wilkins, John S. 2003. How to be a chaste species pluralist-realist: The origins of species modes and the Synapomorphic Species Concept. Biology and Philosophy 18:621-638.

———. 2007a. The Concept and Causes of Microbial Species. Studies in History and Philosophy of the Life Sciences 28 (3):389-408.

———. 2007b. The dimensions, modes and definitions of species and speciation. Biology and Philosophy 22 (2):247 – 266.

———. 2009. Species: a history of the ideaSpecies and Systematics. Berkeley: University of California Press.

———. 2010. What is a species? Essences and generation. Theory in Biosciences 129:141–148.

———. 2011. Philosophically speaking, how many species concepts are there? Zootaxa 2765:58–60.

Wilkins, John S., and Paul E. Griffiths. In Press. Evolutionary debunking arguments in three domains: Fact, value, and religion. In A New Science of Religion, edited by J. Maclaurin and G. Dawes. Chicago: University of Chicago Press.

Winsor, Mary Pickard. 2000. Species, demes, and the Omega Taxonomy: Gilmour and The New Systematics. Biology and Philosophy 15 (3):349-388.

Zimmer, Carl. 2008. What is a Species? Scientific American (June):72-79.

19 thoughts on “Are species theoretical objects?

  1. You distinguish between:

    * This is what species deniers think: it’s all about us and our cognitive dispositions, not the things themselves.

    * Species are salient phenomenal objects.

    I’m not convinced that those are distinct.

    A little later, you say:

    If, as a significant number of specialists think, the rank is a mere convention, then those measures become arbitrary and meaningless.

    I don’t think that follows. Arbitrariness might follow, but not meaninglessness.

    Looking at this, in terms of my study of cognition I see a species, most importantly, as a category. In fact, I see it as a rather useful category. And I see categorization as the foundation of knowledge. So I’m not sure why all of the concern about deciding what a species is.


    1. The need for salience and “it’s all about us and our dispositions” are not incompatible:

      If we only relied upon our perceptual dispositions, then any species claim would be a claim about our perceptual dispositions under such and so conditions. But if we use salient properties, such as (for instance) karyotype, or mating call in the ultrasonic range, neither of which are perceptually salient to us, then we can identify species well and species claims are not about us (solely). In effect we become one instrument among many.

      And I will stick with my claim that if there is no rank of species, there is no meaning to biodiversity in terms of species. Of course the sentence “there are 59 species living here” has meaning; it means that the speaker counts certain things as species. But it has no meaning as a measure of biodiversity.


      1. Perhaps this is all miscommunication.

        If we only relied upon our perceptual dispositions, then any species claim would be a claim about our perceptual dispositions under such and so conditions.

        You previously mentioned “cognitive dispositions.” Now you are changing that to “perceptual dispositions.” That’s quite a change. My comment was about cognitive dispositions, not about perceptual dispositions.

        And I will stick with my claim that if there is no rank of species, there is no meaning to biodiversity in terms of species.

        Your previous remark was with respect to “rank is a mere convention”. Now you are changing that to “there is no rank”. Again, that is quite a change.

        Whether we measure length in inches, or in centimetres, is a matter of arbitrary convention. Change from one to the other certainly changes the numbers. But nobody suggests that length is meaningless.

        For myself, I don’t see anything thing “mere” about conventions. I see them as important, and I see our pragmatic practice of creating useful conventions as an important part of what scientists do.


        1. It may be a miscommunication. For myself, I think of cognitive dispositions as including perceptual dispositions. Perception is a part of cognition. If you think otherwise, then of course my statement is incorrect, but that is about how we divide these things up verbally.

          If rank is a mere convention, then there is no rank [out there apart from our mental representations]. Do you think otherwise?


          1. Yes, perceptual is part of cognition. But your earlier reply specifically mentioned examples (mating call in ultrasonic region) which is unlikely as a perceptual disposition. Yet it seems quite likely as a cognitive disposition among scientists studying an organism with such a mating call.

            If rank is a mere convention, then there is no rank [out there apart from our mental representations].

            That makes no sense, unless we mean different things by “convention”. If species are conventional, then we should presume that those scientific conventions include criteria to be used to classify organisms into the particular conventional ranks. An alien from Andromeda, if following the same criteria, would classify into the same ranks. Yet that alien would not have access to our mental representations.


            1. I think we are talking at cross purposes Neil. One can use things cognitively that are not percepts, like ultrasonic measurements, yes, but likewise our percepts do inform and are a part of our cognising. If you think otherwise, I don’t know what to say.

              As to conventions, again we may mean different things. I am basing my understanding on a read through of Lewis’ Conventions a while back. Consider correctly driving on the left side. Yes, if we all did the same things we’d all be driving on the left, but there is no fact of the matter which is best, left or right. In the same way, we may all choose to classify using the same conventions, but there need be no fact of the matter tracked in virtue of it being a conventional classification. If all we are doing is following conventions, then the ranks or categories so constructed are flatus vocus. There is nothing “out there” that is being tracked.


  2. I enjoyed this post tremendously, thank you.

    Ever since I discovered Hume’s Fork I have used it to spear many debates (probably too many, of course). It seems to me that there are resonances with the ‘world of ideas’ (species) and the ‘world of facts’ (groups of individuals sharing *specifc* characteristics) in your definition. Which leads on to how you bridge the is/ought divide on this subject…

    As for the question about how organisms like ours couldn’t form a species… perhaps if you were the last human, or the last passenger pigeon, or the last dodo then you are no longer part of a species, only the relict of one? How many grains of sand does it take to make a ‘heap’?


  3. I enjoyed this essay tremendously. You clarify and put into words vague ideas I myself have had as a practicing taxonomist; and make the meaning of what I do more clear to me, thank you.
    A tiny pedantic comment is that there is no such thing as a rhagoletid, Rhagoletis is a tephritid. There is no universally accepted plural for genus names; I would have written “Rhagoletis flies” for your particular example. Worth picking up on if this is to be published (which I think it should be).


    1. Thanks for the kind word, Gunnar. That is what a philosopher of science likes to hear; that a scientist has been influenced by what they do somewhat…

      I know that Rhagoletis was a tephritid, but thought there’d be some generic noun. I’ll fix the “print” version.


  4. ‘…“definition” of the word “species”: A species is any lineage of organisms that is distinct from other lineages because of differences in some shared biological property.’

    I too enjoyed reading this very much, but this bit (as I know you recognize) doesn’t or can’t specify what size or type difference is relevant. But it seems to me that species is not less vague than population, which we definitely need.

    PS I found “coalescent” slightly jarring because of its technical connotations.


    1. This is a complex thesis, ironically. It can mean one of three things:

      1. Species are metaphysical individuals, which is to say they are located in space and time in a contiguous fashion;

      2. Species are functional systems that act as an individual; or

      3. Species are phenomenal objects, apprehended in observation.

      If S-A-I is conceived in terms of 1, then it is surely correct. There are problems with it (“Lazarus” species, for example), but that doesn’t detract from the major claim. Trouble is, nobody seems to have held the alternative view: that species are universals (there always exists the possibility or form of Homo sapiens, for example; one exception to this is Louis Agassiz).

      If the S-A-I is conceived, as Ghiselin does, in terms of 2, then it is clearly false. Many species have discontiguous “parts” (populations, for instance) for long periods – in no way can they be said to act as a single functional system.

      If the S-A-I is conceived in terms of 3, then it is also false if offered as a universal claim and true occasionally. That is, many species are “cryptic” to us – we can only identify them in terms of such things as chromosomal structure, genetic relationships or traits (like mating calls) that are not salient to us.


  5. Given 2, didn’t a species immediately following speciation need to be a functional unit? As time progresses and dispersal increases or habitat is fragmented, the species becomes less cohesive, but it is either the same species or multiple species, no? In a sense, you could lose a limb say – equivalent to a population going extinct – or lose signaling contact with the limb after a stroke. If the nerves were reconnected functioning could be restored as could gene flow between isolated populations, no? Yet again, asexual species could be ecologically connected by occupying the same niche range – functional group similarity. Just some quick thoughts.


  6. I spent a career studying speciation in flowering plants, and so always defended the reality of species, lest the study of speciation become akin to discerning the number of angels that can dance on the head of a pin. (I even said to a student once in my mechanisms of speciation course, “If species aren’t real, why are you taking this course?”)

    But what I found in my research was that I was learning interesting things about genetics, and chromosomes, and secondary chemistry, and breeding and incompatibility systems, and ecology, and biogeography. But I never once learned anything generally useful about species.

    You have lucidly explained why, and I’m deeply grateful. You may not fully realize it, but (if people pay attention) you’ve set the study of speciation free. The need to tie in to some theory of species is no longer a hindrance. And the study of speciation has always been a solid research program for investigating many areas of comparative biology.


  7. I spent my professional years as a fish taxonomist and systematist. I see species pretty much as you do. I also think that the variety of living things is such that we don’t need a procrustean definition of species, or other levels of taxonomic discrimination. Rather the students of each group describe their relationships as best they understand them.


    1. Again, what every philosopher of biology wants to hear. I find the usual discussions of natural kinds rather divorced from actual scientific practice. But philosophy is getting better, I think.


Leave a Reply