Skip to content

Thoughts on classification

Last updated on 23 Mar 2023

Species concepts and definitions in science

As Robbie Burns once noted, the world is full of things. Philosophers and scientists alike ask whether that also means the world is full of kinds of things. One of the oldest philosophical questions in the western tradition is known by the somewhat opaque title, “The One and the Many”, but it can be better expressed as the problem of natural kinds. Do they exist or are they our minds seeking order? As one scientist, Theodosius Dobzhansky, posed it, is this part of the order of the world, or of an order-loving mind?

In biology, this is known as the Species Problem. Sometimes it asks “what makes a collection of organisms (things) a species (a kind of thing)?” Sometimes it asks “why do we think species are natural things?” Sometimes it’s the philosophy, and sometimes it’s the methodology.

Why does this even matter, except to scientists? The answer is not straightforward. “Species” is sometimes called the “unit of evolution”, but it is also used as a unit of conservation, medical research, ethics, and politics. And yet, there are numerous scientists and philosophers who deny that species are a real thing. That is to say, they reject the notion there is a kind of thing that should be called species, but not that individual species (like Homo sapiens) can exist.

Now, there are two main approaches to this problem. One is that species just are natural kinds, which leads to one set of issues. The other is that species are merely conceptual or labels, which leads to another set of issues. For example, in the past few years there have been arguments made that species should have civil rights, but if they are just concepts or conventional names, they cannot. Only real things, like persons, can have rights (this is complex, because corporations have rights in the United States and similar societies, but most philosophers of law treat this as a “useful fiction”).

But if species are natural kinds – that is, if they are facts of nature not of scientific practice alone – how do we tell when a supposed species really is a species? For this we need some sort of definition, or a list, at least, of the properties a group of organisms has to have in order to even be a species. This view, which is known as “essentialism”, seeks the essential characteristics of species, and the issue is that nothing which has been proposed is close to being universally accepted. Not fertility (hybrids are common and fertile), not behaviour, nor genetics, nor body form, nor adaptations.

So biology is on the horns of a dilemma: what we want and need the term or category “species” to do for us either doesn’t work, or it works but we do not know how.

Now philosophers are not scientists (unless one happens to have a double degree) and so they shouldn’t be telling scientists how to do their job, but neither are scientists philosophers (in general), and there is a host of problems associated with many scientific terms, even such well established ones as “element”, “gene”, “organism” and so on. This is where philosophy does have something to say. In fact, it almost has too much to say, because the broad approaches I mentioned above are split down to a host of differing approaches.

But the point of a philosophical discussion of science is not to solve problems, but to clarify them. For example, can species have rights even if they are not real? That isn’t a scientific question. Do we need to conserve species if they are just conventional names? Neither is that. Science answers questions, while philosophy provides many alternative solutions that cannot be settled with more facts. Philosophy is a necessary part of doing science, to be sure, and many scientists will engage in it, but philosophy of science isn’t science.

Many scientists have denigrated philosophical approaches to their science, at least since the second world war (before the ultra specialisation and big research programs that came out of the war, scientists were able to acquire a philosophical education as well; vide Einstein). But if scientists stop doing philosophy, science would be the worse for it. Every field of intellectual reflection and effort does some kind of philosophy. Just asking “how do we know this?” is a philosophical question.

Why do we classify?

One question that does not arise often enough is why science needs to classify at all. The physicist Ernst Rutherford said, several times, “in science there is either physics or stamp collecting”. So much for biology, geology, astronomy, social sciences, and anything else that is not immediately reducible to physics. Rutherford’s comment is at best naive and at worst special pleading in favour of (you guessed it) his field of science. But many scientific writers have argued that we must classify before we can start to make sense of a domain of science, because we need generalisations to store in our memory and to discuss those areas of knowledge.

Thomas Henry Huxley once wrote:

By the classification of any series of objects, is meant the actual, or ideal, arrangement together of those which are like and the separation of those which are unlike; the purpose of this arrangement being to facilitate the operations of the mind in clearly conceiving and retaining in the memory, the characters, of the objects in question. [1868]

This is a practical reason: we generalise about things because we cannot recall the details of too many individuals. Today we talk about regression curves or trends to the same effect. Once you get a lot of variation, you need to find a more recognisable and comprehensible pattern. Species summarise one (often in palaeontology there is only one specimen) or many different variant organisms, and the taxonomic description, and choice of holotype (the standard specimen that is kept for reference), allow us to recognise other organisms of that species.

This makes classification an epistemic process, but there has always been a strain of thought that classification is only worth something when the things being grouped together really are closely related. For example, although Huxley, J. S. Mill, and many others held that similarity was the grouping criterion, this is something that depends very much of what seems similar to the taxonomist, and that is a subjective property, not one that sits in the world outside the taxonomist’s head. As I like to say, such taxonomies tell you more about the classifier than the things being classified.

But we do not end with this sort of subjective conceptualism. Even the most idiosyncratic taxonomy involves something tangible and concrete: organisms. Whether or not the taxa are real (in whatever sense we take objects to be real – I prefer the “good enough for government work” criterion), the organisms, or specimens, are. So a subjective taxonomy not only tells the observer about the taxonomist, but also about how the taxonomist interacts with the sensible biological world. While this doesn’t resolve Kantian noumena-phenomena issues, it is worth keeping in mind when considering scientific behaviour in classification.

How do we keep our classifications natural?

One way we group living things is by the use of identical parts. But this is not the same as saying that the parts are identical. The identity here is in the homologies, which is to say, the parts (which can be organs, bones, chemical processes, or even genes) are the same parts even if they are not really that similar in form or function. For example, all vertebrate hearts are homologous no matter how they are shaped (number of chambers, etc), but an earthworm’s functional hearts are not homologous with a vertebrate’s. But the genes underlying the development of the earthworm’s blood vessels that function as pumps may be homologous with some of the genes involved in vertebrate heart development.

So that is what anchors biological classification, but what about other sciences? Is there an underlying principle apart from similarity that would work, in, say, physics or chemistry? I believe there is: and it goes under the late Aristotelian name of essence: the what-it-is-to-be something. But mere similarity won’t work. Consider gallium and mercury: they are both liquid at human-bearable temperatures. Yet they are distinct elements, and we know this because they have difference essential properties. Unlike traditional essentialism, however, their essences vary as isotopes, and in principle they can even be transmuted (although I do not know with what) by making the atomic number of one element the same as another’s (in this case from atomic numbers of 31 to 80 or vice versa).

The choice of atomic number, though, is not arbitrary, even though it was not even known or hinted at when these elements were first named. The theory, here, came after the discoveries, by way of a systematisation of chemical knowledge, leading to the final version of Mendele’ev’s table we know now as the periodic table (see one of Eric Scerri’s works listed below or on PhilPapers). The point is that whether by informed guesswork or by some underlying assumptions, both chemists and biologists happened upon criteria that were reliable enough in identifying real patterns in the data of their science. Homologs and atomic number are both essences in the broader Aristotelian sense. More than just definienda they are the what-it-is-to-be those kinds of things.

In summary, stamp collecting is a lot more vital (sorry!) to science than Rutherford knew. And even today, physicists are seeking the essences of particles, although they call them fundamental particles (or fields; I don’t know that literature well enough). Essences, although not the kind that Aquinas through to Mill bequeathed us, are still required. Caveat: this doesn’t mean we need a scientific essentialism. Everyone uses essences sometimes; but the notion that the basis of science is essentialism is an inversion of the truth. Not everything has an essentialist metaphysics. More on that if it occurs to me.


Huxley, Thomas Henry. 1868. An Introduction to the Classification of Animals. London: Churchill. Available on

Scerri, Eric. 2012. “What Is an Element? What Is the Periodic Table? And What Does Quantum Mechanics Contribute to the Question?” Foundations of Chemistry 14: 69–81.

Scerri, Eric R. 2007. The Periodic Table: Its Story and Its Significance. New York: Oxford University Press.

———. 2008. “The Past and Future of the Periodic Table.” American Scientist 96 (1): 52–58.

Wilkins, John S. 2013. “Essentialism in Biology.” In Philosophy of Biology: A Companion for Educators, edited by Kostas Kampourakis, 395–419. Dordrecht: Springer.

———. 2023. Understanding Species. Understanding Life. Cambridge: Cambridge University Press.

Wilkins, John S., and Malte C. Ebach. 2013. The Nature of Classification: Kinds and Relationships in the Natural Sciences. London: Palgrave Macmillan.