Classifying Life, Reconstructing History and Teaching Diversity: Philosophical Issues in the Teaching of Biological Systematics and Biodiversity

Abstract

Classification is a central endeavor in every scientific field of work. Classification in biology, however, is distinct from classification in other fields of science in a number of ways. Thus, understanding how biological classification works is an important element in understanding the nature of biological science. In the present paper, I discuss a number of philosophical issues that are characteristic for classification in biological science, paying special attention to questions related to science education. My aims are (1) to provide science educators and others concerned with the teaching of biology with an accessible overview of the philosophy of biological classification and (2) to show how knowledge of the philosophy of classification can play an important role in science teaching.

1 Introduction

Classifications are ubiquitous. In everyday life, we classify things into kinds all the time because allocating what we encounter to kinds helps us cope with the diversity that surrounds us. It is useful to distinguish between vegetables and fruit, between poisonous mushrooms and edible mushrooms, between category 2 toxic substances and category 3 toxic substances, between naturally occurring fauna and invasive species, etc., because these distinctions help us handle parts of plants and fungi in the marketplace or in the kitchen, draw up transport regulations for chemicals, or decide between different nature management options. Classifications thus play an important role in everyday practices, because classifications make the diversity of things manageable for us.

Within the more specific context of natural science, too, classification serves to make diversity manageable, to the extent that without suitable classificatory systems scientific investigations seem impossible. For example, lacking a suitable classification of the things under study into kinds, it seems very difficult—if not impossible—to achieve general knowledge that can be applied in various situations. Not many people would be interested in studying, say, fruit flies if the knowledge that is obtained applies only to those particular fruit flies that have been studied in this particular laboratory at that particular time. Later investigators of fruit flies want to be able to build upon the knowledge about fruit flies that was achieved by their predecessors. Thus, what scientists are looking for is knowledge that applies to all, or at the very least most, organisms of the species Drosophila melanogaster—future ones as well as present and past ones. What science (and, it may be presumed, humanity in general) is after is knowledge that, once it has been obtained by studying a few samples, can be extrapolated to the other members of the same kind. In this way, in science classificatory practices are closely connected to one of the core aims of science, i.e., of obtaining generally valid knowledge.

This is, however, not merely a matter of making science possible by making diversity manageable. In science, classification often also constitutes an aim in itself. According to some philosophers of science, an important aim of science is to provide us with an inventory of the various kinds of “things” (material objects, processes, phenomena, effects, properties, behaviors, functional parts of systems, systems’ states, organizational structures, …) that make up the furniture of the world, as well as the typical properties that the members of these kinds exhibit and the laws of nature that hold about them (e.g., Armstrong 1983, p. 3).

Classification in biology is, however, different from classification in other fields of science in a number of ways. Indeed, it may well be that each field of science has its own idiosyncrasies when it comes to how the things under study are classified into kinds. Thus, understanding how classification works in a particular science constitutes an important element in understanding the nature of that science and, accordingly, has a direct bearing on teaching the nature of science in the classroom. One of the things that make biology unique is the way in which biologists classify the entities they study. So, understanding how biology works—how it explains phenomena in the living world and what the knowledge it produces might mean for societal and practical issues—consists to a considerable extent in understanding how biological classification works.

My aims in the present paper are (1) to provide science educators and others concerned with the teaching of biology with an accessible overview of the various philosophical issues that arise with respect to biological classification and (2) to show how knowledge of the philosophy of classification can play an important role in science teaching. A complicating factor in this context is the fact that classification in biology is a very diverse enterprise. Biology consists of a considerable number of specialized fields, each of which have their own specific subject matter (ranging from the very small to the very large and including genes, gene networks, cells, organelles, organismal parts and structures, organisms, populations, metapopulations, biomes and ecosystems), their own specific aims and their own specific methodologies. Often they use their own ways of classifying their subject matter. Moreover, biology is a vibrant area of science in which besides the traditionally established fields (such as evolutionary biology, ecology, genetics, behavioral biology and biological systematics) new fields continue to come into existence that sometimes develop their own classifications for their own specific purposes (think of such fields as genomics, proteomics, metabolomics, systems biology or synthetic biology). Biologists, then, classify a broad diversity of things both in a variety of ways and in a variety of investigative contexts. This makes it an impossible task to cover the entire domain of biological classification within the confines of a single paper.

In what follows I shall discuss some of (what I take to be) the most important philosophical themes related to classification in biology, with a view on connecting these to topics from the realm of biology education. I shall begin in Sect. 2 by reviewing two important features of biological classification, namely the role of common descent and shared function as fundamental criteria in the construction of biological classifications. These two factors are of central importance in biology, but they are much less important in most other sciences (although, to be sure, functional classifications are found elsewhere too). Accordingly, they constitute a crucial part of what makes biological classification different in nature from classifications in other domains of science.

In Sect. 3, I continue by discussing some of the principal features of the classifications that are being produced in the field of biology that is especially dedicated to classifying life: systematics, that part of biology that is entirely focused on reconstructing evolutionary history and classifying life in relation to that history. While in most other domains of science classification is deeply interwoven with other activities, such as conducting experiments or building theories, and there are no fields that focus on classification, in biology classification to some extent leads a life of its own. To be sure, there is much classification going on in other fields of biology besides systematics. But the classifications produced in systematics occupy a special position within the whole of biology: when triggered by the term ‘biological classification’, most people tend to think of the Linnaean system of species of organisms that cluster into genera, families, orders and so on, such that one might say that systematics is the classificatory part of biological science.¹ Therefore, understanding how classification in contemporary biological systematics works is an important element of understanding biological classification in general and, consequently, of understanding the nature of biological science.

In Sect. 4, I turn to what is perhaps the most distinctive philosophical aspect of biological classification: the ongoing debate on nature of biological taxa, in particular species. What has come to be known as “the species problem” has occupied both biologists and philosophers of biology for more than one and a half centuries. But despite the enormous amount of work that has already been invested in addressing the issue and recurring claims that the issue was now close to being resolved once and for all (for example, De Queiroz 1998, 1999), there still is no widely accepted solution and opinions on the nature of biological taxa continue to diverge as before. Biological classification cannot be understood without understanding the nature of the species problem, as the problem has its origins precisely in the fact that classifying organisms into species is something completely different from classifying atoms into elements or classifying elementary particles into kinds. It is safe to say that if species had been kinds of a similar sort as the kinds that feature in other sciences, there would not have been a species problem. The species problem will thus occupy a considerable part of this paper.

In Sect. 5, I will close by summarizing what in my view are the main issues that should be addressed when teaching about biological classification and the diversity of life that it represents.

2 Classification in Biology is Different

2.1 What Makes Biological Classification Unique

According to Ernst Mayr, one of the most prominent biologists of the twentieth century, biological classifications serve two distinct, albeit closely related, kinds of functions: what Mayr called practical functions and scientific/metaphysical functions in comparative studies (Mayr 1981, pp. 511, 515; 1982, pp. 148–149; 1997, pp. 125–127; Ereshefsky 2001, pp. 170–171). Essentially, this is the dichotomy that was already mentioned above between classification as a tool for and an aim of doing science.

The practical function of classifications consists in serving as information storage and retrieval systems for the particular fields of work in which they are used. Classifications facilitate scientific research by providing information about individual entities once these have been identified as members of particular classes in the relevant classification. As Mayr observed, “If I have identified a fruit fly as an individual of Drosophila melanogaster on the basis of bristle pattern and the proportions of face and eye, I can “predict” numerous structural and behavioral characteristics which I will find if I study other aspects of this individual.” (Mayr 1961, pp. 1504–1505). This practical role is one that classifications play throughout all the sciences, as philosophers of science acknowledge: “The main function of classification is to convey information. If a chemist tells you that something is a metal, that tells you a lot about its likely behavior.” (Okasha 2002, p. 103). In this respect classifications in biological science do not differ from classifications found in the other sciences.

But in the context of natural science there is more to classification than information storage and retrieval for various practical purposes. As was already pointed out in Sect. 1, an important aim of natural science is to provide us with an inventory of the various kinds of things that exist in the world “out there” and of their various characteristic traits and behaviors. Classifications in natural science aren’t merely useful storage-and-retrieval systems, they are also claims about nature. Mayr emphasized this when he held that biological classifications have the status of scientific theories (Mayr 1968, p. 546).² Accordingly, Mayr held that ordering systems that only serve the function of information storage and retrieval should not be thought of as being classifications at all because they do not fulfill both functions of classification. For instance, he argued that the cladograms produced by cladistics, one of the major schools in systematics, should not be thought of as classifications (Mayr 1982, p. 230; 1997, pp. 143–146; Mayr and Bock 2002, p. 190; see Sect. 3 below for discussion).³ Natural science is supposed to tell us what there is in the world that we live in and scientific classifications are supposed to contribute to this aim by representing relevant aspects of the natural state of affairs. Thus, above and beyond serving as practical tools for doing science (information storage and retrieval devices), biological classifications are hypotheses or theories about the state of affairs in the natural world.⁴

It is with respect to this second function of classification that biological classification is different. Since the eighteenth century investigators studying the living world have sought for a “natural system” of classification of organisms, that is, a system of classification that best reflected the naturally existing order in the living world.⁵ This search is the central theme in the contemporary field of biological systematics (or, systematic biology), a field of work that can be traced back to the early beginnings of the study of the living world in Greek Antiquity, in particular to the works of Aristotle (Kearney 2007, p. 212).⁶ Indeed, the idea of a natural system connects to a traditional theme in Western philosophy, namely the distinction between natural classifications, which are supposed to represent divisions that exist in nature, and artificial or conventional classifications, which are entirely based on human considerations and do not represent such natural divisions (e.g., Daly 1998; Kornblith 1999; Bird and Tobin 2008, Sec. 1.1). For a long time it was thought that the natural order in the living realm was a linear hierarchy of organismal forms, ranging from the very simple to the very complex and advanced, called the scala naturae or Great Chain of Being (Lovejoy 1936; Mayr 1982, pp. 148–149; Nee 2005). The antecedents of the idea can be traced back to Aristotle’s History of Animals ⁷:

Nature proceeds little by little from things lifeless to animal life […] In plants there is a continuous scale of ascent toward the animal. In the sea there are certain objects concerning one would be at a loss to determine whether they are animal or plant […] And so throughout the entire animal scale there is a graduated differentiation in the amount of life and the capacity for motion. (quoted in Grene and Depew 2004, p. 14).

Although as a fundamental assumption about the nature of the order in the living world the idea of a linear scala naturae remained influential as a topic of discussion until well into the late eighteenth and early nineteenth century (Lovejoy 1936; Grene and Depew 2004, pp. 161, 324; Kearney 2007, p. 212), diverging views on the natural order in the living world emerged as biology developed further.

Most importantly, many biologists came to think that the natural order that the natural system should represent consisted in the plan of the Christian god’s creation. This was a view that was held by Carolus Linnaeus, the eighteenth century founding father of biological systematics, as well as many biologists throughout the nineteenth century (Grene and Depew 2004, pp. 213–214). For example, in various locations Linnaeus expressed it as follows: “There are as many species as the Infinite Being produced diverse forms in the beginning” (Linnaeus, quoted in Wilkins 2009b, p. 41) and

Species are as many as there were diverse and constant forms produced by the Infinite Being; which forms according to the appointed laws of generation, produced more individuals but always like themselves. Therefore there are as many species as there are diverse forms or structures occurring today. (Linnaeus, quoted in Wilkins 2009b, p. 41).

Similarly, more than a century later biologist Louis Agassiz held that “Natural History must in good time become the analysis of the thoughts of the Creator of the Universe, as manifested in the animal and vegetable kingdoms, as well as in the inorganic world” (Agassiz [1859] 2004, p. 137) and “all we can really do it, at best, to offer imperfect translations into human language of the profound thoughts, the innumerable relations, the unfathomable meaning of the plan actually manifested in the natural objects themselves” (Agassiz [1859] 2004, p. 187).

This was the tradition of thought against the background of which Darwin proposed his theory of common descent. As Darwin noted in an often-quoted passage from the Origin of Species:

Naturalists try to arrange the species, genera, and families in each class, on what is called the Natural System. But what is meant by this system? Some authors look at it merely as a scheme for arranging together those living objects which are most alike […] But many naturalists think that something more is meant by the Natural System; they believe that it reveals the plan of the Creator (Darwin 1859, p. 413).

But he immediately objected to this view and proposed an alternative:

… but unless it be specified whether order in time or space, or what else is meant by the plan of the Creator, it seems to me that nothing is thus added to our knowledge. […] I believe that something more is included; and that propinquity of descent, – the only known cause of the similarity of organic beings, – is the bond, hidden as it is by various degrees of modification, which is partially revealed to us by our classifications (Darwin 1859, pp. 413–414).

Biologists quickly came to accept Darwin’s view that a natural system of classification should reflect the common descent of living organisms.⁸ In contemporary biology, the view that the classification of organisms should in some way or another reflect the common ancestry of all of life on earth still lies at the basis of biological systematics.

What makes biological classification different from classifications in other sciences, then, is that biology studies a realm in which the natural order of things (according to currently accepted biological theory) is of a different kind from the natural order in the realms studied by physics, chemistry and other sciences. In both the physical and the chemical realms, the natural order is based on similarity: things of the same kind share a number of kind-specific properties. All and only electrons, for example, possess a particular combination of mass, electric charge and spin that distinguishes them from other kinds of elementary particles in the Standard Model. Similarly, to give a timeworn example, all and only gold atoms possess a particular number of nuclear protons, which distinguishes them from the other elements in the Periodic Table. Electrons and gold atoms thus constitute kinds based on having a limited number of the same properties: the fact that all members of the kind ‘electron’ share a set of traits is the reason for grouping them together into the particular kind. In addition, the kind ‘electron’ is eternal in the sense that electrons may exist at any place and at any time in the universe at which the conditions for their existence obtain. In biology, in contrast, common descent rather than sameness in a number of traits is the “hidden bond” underlying the natural order of organisms in the sense that common descent explains sameness of kind. Also, species of organisms are not eternal in the sense that kinds of elementary particles are, as organisms of the same species can only exist as members of the same genealogical nexus. Once a species has become extinct, it cannot at a later time have members again. By consequence, “[t]he ordering system is a phylogenetic tree, a hierarchical system that groups taxa according to relative recency of common ancestry” (Kearney 2007, p. 213). Accordingly, biological systematics today is practiced in the form of phylogenetic systematics, and “the task of the systematist can be seen as the knitting together of species via evidence of common ancestry into a phylogenetic tree” (Kearney 2007, p. 213).⁹

In the next section, I shall take a closer look at the foundations of phylogenetic systematics. But before turning to phylogenetic systematics, two further aspects of biological classification should be noted.

2.2 The Role of Similarity in Biological Classification

The first point to note is that biologists do use similarities between organisms with respect to relevant morphological and genetic traits in building their classifications in taxonomic practice. In addition, biologists often use the congruence of classifications based on different sets of traits pertaining to the same organism groups as evidence for the correctness of their classifications (Nickels and Nelson 2005, p. 286). However, these similarities do not constitute the ultimate grounds of the classification of organisms into species and of species into higher taxa, but rather serve as the tools with which these classifications are built. The grounds of classification is hypothesized commonness of descent. This is an important issue in the context of teaching about biological classification that sometimes remains underemphasized. According to Nickels and Nelson, for example,

The core tasks in teaching biological classification are to help students understand three things. The first is that organisms exist in discrete, hierarchically nested groups. The second is that these groups are defined non-arbitrarily by congruence across characters. The third is that these congruent characters reflect, and are strong evidence of, shared common ancestry and, thus, of evolution. […] It is exactly this congruence among diverse data sets that makes biological classification unique and, further, ties biological classification directly to evolution (Nickels and Nelson 2005, p. 287).

I agree with Nickels and Nelson that it is crucial to point out to students that character congruence is evidence of common ancestry. However, I disagree that one should suggest that taxa are defined by congruence across characters or that congruence across different data sets is what makes biological classification unique and constitutes “most distinctive feature of biological classifications” (Nickels and Nelson 2005, p. 287). What is unique about biological classifications of organisms is not the fact that different data sets often lead to the same or at least very similar classifications. Rather, it is the underlying reason why this is so, namely common descent. Congruence of classifications built on different data sets is not unique to biology and the idea that such congruence can be taken as an indication for the correctness (or naturalness) of a classification is widespread.¹⁰ In other sciences, however, the causes underlying this congruence are different. Therefore, suggesting to students that congruence of characters defines taxa and is the distinctive feature of biological classification can easily lead them to miss the main point: that common descent is what makes biological classification unique.

2.3 Functional Classification

A second point that should be noted is that its being based upon common descent is not the only feature of biological classification that makes it different from classifications in other sciences. In the preceding discussion, ‘biological classification’ was used to refer to the classification of organisms into species and higher taxa. But of course there is more to biological classification than allocating organisms to species, identifying the proper locations of species in the phylogenetic Tree of Life and classifying species into nested higher taxa. Besides organisms and species, biologists also classify parts of organisms, genes, gene networks, behaviors, biomes, etc. In many cases, the relevant classifications are into partly or wholly functional kinds, that is, whether or not biological entities have the same function is among the criteria used for grouping them together into the same kind or not.

One classificatory context in which functions play a role is constituted by so-called bio-ontologies. In particular in new, data-intensive fields of investigation such as genomics, the practical aim of information storage and retrieval is increasingly being realized by means of formalized classificatory systems, bio-ontologies, the most prominent of which is the Gene Ontology. In the Gene Ontology genome sequence data, polypeptide sequence data and data about the functions of particular DNA-segments from various model organisms are integrated in such a way that these data are easily accessible to researchers from various fields of work (Bodenreider and Stevens 2006; Jensen and Bork 2010; Leonelli 2010). A bio-ontology is “a set of vocabulary terms whose meanings and relations with other terms are explicitly stated and which are used to annotate data” (Rhee et al. 2006, p. 345). As such, bio-ontologies primarily serve as storage and retrieval tools, i.e., as databases. However, bio-ontologies are widely perceived as being more than mere storage-and-retrieval systems, namely classifications that express knowledge about what kinds of things there are in reality (Bodenreider and Stevens 2006, p. 257). That is, bio-ontologies are conceived of as theories about the world (Leonelli 2010). The rise of formalized bio-ontologies is a new and important development in biological science—a development that might change the nature of biological classification and hence might change a part of the nature of biological science. As it is a very recent development, however, philosophers of science have only just begun to analyze its implications and so far it remains unclear whether—and if so, to which degree—bio-ontologies will render biological classification more distinct from or rather more similar to classifications in other disciplines.¹¹

In the context of functional classifications in biology an important issue to consider is the lively philosophical debate on how, exactly, biological functions are to be conceived of. At present, four or five principal notions of biological function stand at the focus of the debate (for overviews, see Wouters 2003, 2005). A resolution of the debate has not been achieved and, moreover, it remains unclear whether all usages of the term ‘function’ in biological science can be subsumed under a single function concept, or multiple notions of function are required to cover all the usages of the term in biological science. Clearly, this debate has consequences for the nature of functional classifications in biology. For example, on one prominent notion of biological function, the notion of function as selected effect (e.g., Williams 1966; Millikan 1989; Neander 1991), the functions of biological entities are directly related to their descent, such that classifications on the basis of function will tend to coincide with classifications based on common descent (cf. Griffiths 2006). However, on other notions of function that prominently feature in biology, such as the notion of function as the causal role of a part within a larger system (e.g., Cummins 1975), this link between function and descent is absent, such that there is no reason to expect that a conjunction of function-based and descent-based classifications will generally occur.

Because of the importance of functional classifications in biology, in teaching contexts attention should be given to the fact that many biological classifications involve a criterion of function (in contrast to classifications in physics and chemistry, for example), as well as to the specifically biological debate on what, exactly, biological functions are. These aspects of biological classification in my view also constitute crucial elements in the understanding of the nature of biological classification, next to the notion of common descent.

However, I think that for several reasons it is also important not to overemphasize the use of functional classifications in biology. For one thing, the use of functional classifications is not unique to biology: functional classifications are found elsewhere too, for example in psychology, cognitive science, the social sciences or the engineering sciences. In addition, clearly not all classifications in biology involve the functions of the classified entities—most importantly of course the classification of organisms into species and higher taxa. And, lastly, many functional kinds in biology aren’t exclusively functional kinds, that is, the functions of the classified entities constitute one but not the only criterion for classifying. An important example of the latter point is the classification of genes into kinds: genes are identified as members of particular kinds on the basis of their functions in the production of proteins in combination with the species in which they perform their function and the copying lineage in which they stand (for detailed analysis of this case, see Reydon 2009). In sum, classifying by function is an important aspect of biological classification, but what makes classification in biology crucially different from classification in other sciences is that biological classifications are based on similarities in traits (functional or otherwise) that can be explained by common descent.

3 Phylogenetic Systematics

3.1 The Foundations of Phylogenetic Systematics

As mentioned above, the classification of organisms into species and higher taxa is a part of the field of biological systematics, which today takes the form of phylogenetic systematics. But systematics encompasses more than just classification. Another important aim of systematics is the reconstruction of the actual history of evolution of life on earth, that is, phylogeny reconstruction (Wiley 1981, pp. 6, 15; Kearney 2007, p. 218; Baum and Offner 2008, p. 225). However, opinions differ among systematists regarding the relative importance of these two aims. For example, one school of thought in systematics (pattern cladism) sees classification as the only aim of systematics (e.g., Nelson and Platnick 1981; for discussion, see Ereshefsky 2001, pp. 75–79). Others emphasize the reconstruction of evolutionary history as the most important aim of biological systematics and some even hold that conceiving of systematics as a classificatory enterprise is mistaken and the reconstruction of evolutionary history is the sole aim of systematics (O’Hara 1991,1997, p. 327; Mayden 1992, p. xvii).¹² Most present-day systematists, however, think of phylogenetic systematics as pursuing both aims. Here, I shall follow this line of thought and shall ignore the different views that have been advanced of how, exactly, systematics, classification and taxonomy are related to each other.¹³

Phylogenetic systematics takes into account two of the core principles of Darwin’s theory of evolution: the principle of common descent and the principle of gradual evolutionary change.¹⁴ But these two principles entail two related problems in classificatory practice, because both descent and gradual change refer to processes that lack discrete transitions. For one, there is the question whether biological classifications should be based only on hypothesized common descent, i.e., whether common descent (as hypothesized on the basis of observed similarities) should be the single criterion for grouping species into higher taxa, or additional criteria should be taken into account. In addition, if evolution proceeds gradually, there will be no clear-cut boundaries between species, as ancestral species gradually change into descendant ones. Suppose that at some point in time a population splits off from the main population. Even though the two populations might eventually go their own evolutionary ways, for quite some time the organisms in the two populations will continue to resemble one another morphologically and genetically and be able to produce fertile offspring. If the two populations were to merge at a later time, biologists would quite likely see them as two populations within the same species that have been separated for a while. If they don’t merge and indeed drift apart evolutionarily, they will eventually be thought of as distinct species. Moreover, if we decide to place the speciation event at the exact point in time at which the populations separated, we’ll have a situation in which parent organisms belong to a different species as their immediate offspring. But if we choose a later point in time, we’ll have a founder population of a new species that itself still belongs to the ancestral species, while later offspring suddenly belongs to the descendant species (and again, we have parents belonging to different species as their offspring). So, where exactly do the boundaries between species lie? Francis Galton, the famous English polymath and Charles Darwin’s cousin, nicely expressed this particularity of biological classification in the following way:

Natural groups have nuclei but no outlines; they blend on every side with other systems whose nuclei have alien characters. A naturalist must construct his picture of nature on the same principle that an engraver in mezzotint proceeds on his plate, beginning with the principal lights as so many different points of departure, and working outwards from each of them until the intervening spaces are covered (Galton 1874, pp. 2–3).¹⁵

In what follows, I shall consider in more detail how phylogenetic systematics deals with these issues.

Two features of phylogenetic systematics as a classificatory enterprise should be noted. First, phylogenetic systematics is not directly concerned with classifying organisms into species. Rather, what phylogenetic systematists do is to classify species and other taxa into higher taxa on the basis of hypotheses about how recent their latest common ancestor was in their evolutionary history. As Mayr emphasized with respect to biological classification in general, “classification deals with populations (species), while individual organisms are merely assigned to species (that is, are identified) […] evolutionary classifications operate through the grouping of related taxa into higher taxa” (1982, p. 238; see also 147; Mayr 1968, p. 546).

A given species is grouped together with the species with which it shares its most recent common ancestor, that group is grouped together with the species with which it shares its most recent common ancestor, and so on. From such procedures, tree-like diagrams are obtained like the ones shown in Fig. 1. In Fig. 1, taxa D and E are grouped together into a higher taxon because they share a more recent, in this case unknown ancestor (X) with each other than with any of the other taxa in the tree; taxa F, G and H are grouped together into a higher taxon because they share a more recent ancestor with each other than with any of the other taxa in the tree; etc. If the taxa are species, one could, for example, recognize the group F, G and H as a genus within which species F and G share a more recent common ancestor not shared with C, and the group D, E, F, G and H as a family encompassing two genera, D and E and F, G and H. There might, however, also be good reasons to place the various groups in the diagrams at higher or lower ranks in the Linnaean hierarchy than the ranks suggested here. Note that the different shapes of the diagrams (a cladogram on the left and a phylogenetic tree on the right) do not necessarily imply differences in the information contained in them, such as how far back in evolutionary history the most recent common ancestor of taxa D and E lies or at which taxonomic level the various groups should be placed.¹⁶ In Fig. 1, the cladogram and the phylogenetic tree just show which taxa should be grouped together with which on the basis of observed similarities (although phylogenetic trees are usually taken to also reflect recency of common ancestry, for cladograms this is not usually the case).

Fig. 1

Cladogram (a) and phylogenetic tree (b) showing the degree of relatedness of eight taxa A–H

Second, in phylogenetic systematics the most important factor in classification is not the shared traits of organisms, but evolutionary relationships that are explained in terms of hypothesized shared descent (Mayr 1968, p. 546; Bock 1974, pp. 377–378; Okasha 2002, pp. 107–108; cf. Sec. 2.2 above). Similarities function as indications of common descent and as such constitute useful tools for classification—especially since commonness of descent is hardly ever known directly, but must be inferred from relevant similarities between groups of organisms. As Darwin put it: “where there has been close descent in common, there will certainly be close resemblance or affinity. […] we have to make out community of descent by resemblances of any kind. Therefore we choose those characters which, as far as we can judge, are the least likely to have been modified in relation to the conditions of life to which each species has been recently exposed.” (Darwin 1859, p. 425). Thus, the basis of classification is common descent, which is indicated by trait resemblances.

3.2 Cladistics Versus Evolutionary Systematics

The procedure for grouping species into higher taxa that was described above is, however, not uncontroversial. In contemporary biological systematics there exist several schools of thought, which differ with respect to the criteria they adopt for grouping species into higher taxa (Ereshefsky 2001, pp. 51–60, 66–79; 2007, 2008; Grene and Depew 2004, pp. 303–306; Kearney 2007, pp. 214ff.). All agree that a central criterion should be how recent the most recent common ancestor of two groups is (Bock 1974, p. 379; Mayr 1997, pp. 136–137; Mayr and Bock 2002, p. 190). From the viewpoint of cladistics, today perhaps the most prominent school of thought in systematics, recency of common ancestry is the only criterion for grouping species and higher taxa into higher taxa.¹⁷ This is expressed in the concept of monophyly: a monophyletic group is a group of species that includes an ancestral species and all of its descendants (Wiley 1981, pp. 76, 84–85; Okasha 2002, p. 108; Ereshefsky 2008, p. 108). Thus, in Fig. 1, D, E and X constitute a monophyletic group. Cladists recognize only monophyletic groups of species as natural taxa in their classifications.

From the viewpoint of evolutionary systematics or evolutionary taxonomy (which was championed by Mayr), in contrast, whether or not important evolutionary novelties have occurred in particular branches of the Tree of Life is adopted as a criterion too, next to common descent (Mayr 1968, pp. 547–548; 1981, p. 513; 1982, pp. 233–235; 1997, p. 137; Bock 1974, pp. 377, 381; Wiley 1981, pp. 241–242). As Mayr put it, “the traditional methodology of classifying […] consists in attempting to represent in the classification not only the branching of phyletic lines but also their subsequent divergence” (Mayr 1982, p. 233). Here, with ‘divergence’ Mayr means evolutionary divergence consisting in the occurrence of novel characters in lineages. Mayr (1981, p. 513; 1997, p. 137) traced this principle back to Darwin, who asserted that

… all true classification is genealogical […]. I believe that the arrangement of groups within each class […] must be strictly genealogical in order to be natural; but that the amount of difference in the several branches or groups, though allied in the same degree in blood to their common progenitor, may differ greatly, being due to the different degrees of modification which they have undergone, and this is expressed by the forms being ranked under different genera, families, sections, or orders (Darwin 1859, p. 420; original emphasis).

This quotation shows why it is important to endorse evolutionary divergence as a criterion for biological classification in addition to commonness of descent: the branching off of lineages from one another and the evolutionary divergence of lineages are not absolutely correlated but in part proceed independently of one another (Mayr 1997, p. 138). Thus, a classification that only represents the order in which taxa have branched off from ancestral taxa represents only part of evolutionary history: it doesn’t represent the evolutionary change of the various taxa.

A well-known example of the difference of opinion between cladists and evolutionary systematists concerns the classification of crocodiles and alligators (together in the order Crocodilia), turtles (order Chelonia), snakes and lizards (together in the order Squamata) and birds (Mayr 1982, p. 235; Ershefsky 2001, pp. 54–55; 2008, p. 109; Okasha 2002, pp. 109–111; Grene and Depew 2004, p. 304; Kearney 2007, p. 214). Birds and crocodiles/alligators share a more recent common ancestor with each other than each of these groups share with the turtles, snakes or lizards. According to cladists, therefore, birds and crocodiles/alligators together constitute a taxon. Evolutionary systematists, however, place emphasis on the fact that birds are quite different from crocodiles, alligators, turtles, snakes and lizards in a number of characters, including the possession of feathers and the capability of flight. After the birds had branched off from the crocodile/alligator group, a lot of evolutionary change has taken place in the bird lineage, a fact which should be represented in the Tree of Life, evolutionary systematists argue. Thus evolutionary systematists recognize two separate classes, the Reptilia (encompassing the Crocodilia, the Chelonia and the Squamata, but not the birds) and the class Aves, the birds (see Fig. 2). Evolutionary systematists thus make a “cut” in the Tree of Life between the birds and the reptiles that according to cladists should not be made. The fact that birds and crocodiles/alligators share a recent common ancestor is of course recognized by evolutionary systematists, but judged to be less important for the purposes of systematics than the fact that a major evolutionary novelty has occurred in the birds after the earliest birds had gone their own evolutionary way. Cladists only recognize monophyletic groups based on common ancestry and thus do not recognize Aves and Reptilia as separate classes. According to cladists, any group that includes crocodiles and alligators must also include the birds.

Fig. 2

The cladogram showing the evolutionary relationship between birds, crocodiles, snakes, etc. can be turned into a classification in several ways. Evolutionary systematists a recognize the classes Reptilia and Aves, which cladists, b do not recognize. Note that this cladogram is not exhaustive (cladograms and phylogenetic trees typically aren’t): not shown are, for instance, the mammals (Mammalia), which would occupy a branch between the turtles and the lizards/snakes, leading to additional difficulties with the class Reptilia as a non-monophyletic group

A similar case concerns the classification of humans among the apes, which include chimpanzees (Pan troglodytes), bonobos (Pan paniscus), gorillas (genus Gorilla) and orangutans (genus Pongo). Chimpanzees, bonobos and gorillas share a more recent common ancestor with humans than they do with orangutans, so from a cladistic point of view chimpanzees, bonobos, gorillas and humans should be grouped together in a monophyletic taxon that encompasses the genera Pan, Gorilla and Homo but excludes the genus Pongo. This group of three genera, in turn, is grouped together with the genus Pongo into a monophyletic taxon (Hominidae). But in this case, too, evolutionary systematists hold that humans constitute a separate taxon (Hominidae, excluding chimpanzees, bonobos, gorillas and orangutans) from the other great apes (Pongidae, encompassing the genera Pan, Gorilla and Pongo, but excluding humans) because of the “extraordinary change in the hominid line since it split off from the pongid line” (Mayr 1968, p. 548; Grene and Depew 2004, p. 304). One of these evolutionary novelties that separates humans from apes is the Broca region in the brain, which is associated with the capability of speech (Mayr 1981, p. 514). This is indeed a trait of the utmost importance when comparing humans with chimpanzees, gorillas, etc., such that it seems that evolutionary systematists have a powerful argument on their side in the debates. The debates are, however, still far from a final resolution.

The bone of contention in the opposition between cladists and evolutionary systematists lies in differences of opinion about what classification is and what the aims of biological systematics are. For many cladists, as was mentioned at the beginning of Sect. 3.1, systematics should restrict itself to classification, i.e., allocating species to higher taxa. Reconstructing evolutionary history and representing the reconstructed history in the classification of organisms and species cannot be an aim of systematics. Accordingly, some cladists (pattern cladists) even hold the view that assumptions about how evolution occurs—or even the assumption that it occurs at all—do not lie at the basis of cladistics, such that cladistic analyses and cladogram construction can be done without making any assumptions whatsoever about the process of evolution (Brower 2000; for discussion, see Grene and Depew 2004, p. 298). Evolutionary systematists, however, emphasize that grouping organisms and species by genealogy alone does not yet give us a classification, because the amount of evolutionary change should be reflected in a classification too (Mayr 1981, p. 510; 1997, p. 137; Mayr and Bock 2002, p. 184). For evolutionary systematists, thus, a classification has the additional purpose of adequately reflecting evolutionary history. Thus, evolutionary systematists have a different opinion than cladists both about what classification is (at least within the domain of biology) and about what the aims of systematics are. Mayr expressed this difference of opinion thus: “The great deficiency […] of cladistics is the failure to reflect adequately the past evolutionary history of taxa.” (Mayr 1981, p. 515). As such evolutionary systematics takes a middle position between the two extremes mentioned at the beginning of Sect. 3.1.

3.3 Relevant Issues for Science Teaching

Within the context of science teaching, the debate between cladists and evolutionary systematists and the two concrete cases discussed above (regarding the classification of birds among the reptiles and of humans among the apes) can be used to illustrate several aspects of how biology works, as well as of the nature of science in general.

First, the cases provide an illustration of the differences of opinion that scientists may have with respect to the purposes of classification in science. Is a classification only an information storage-and-retrieval system that enables scientists to obtain knowledge about entities once they have been identified as members of particular kinds? Or is a classification more than this—is it also a system that tells us something about the state of affairs in nature itself, in the case of biology for instance about how the evolution of life on earth has actually proceeded? Given that cladists and evolutionary systematists regularly produce different, incompatible classifications of the same groups of organisms, as the two cases of birds and humans illustrated, one might ask students which reasons they can come up with to prefer one or the other way of classifying over the alternative, thus introducing them to some of the considerations that scientists are faced with in their everyday research practice.

Second, the discussion between cladists and evolutionary systematists on the best way to classify species and higher taxa shows how in a well-established field of science different schools of thought can be present that each interpret the available data in their own way, leading to hypotheses and theories that conflict with one another. Because it pertains to a conflict that exists within one particular field of science, this case illustrates that not only can there be large differences of opinion about the aims of particular elements of scientific work between very different sciences (say, physics and sociology), but also within the same field of work such differences of opinion can and commonly do exist. It also shows that the philosophical assumptions underlying scientific work as well as scientific knowledge as the product of science is open for debate, criticism and revision.¹⁸

Depending on which perspective one assumes, the cladistic or the evolutionary systematic, birds are reptiles and humans are apes, or birds constitute a separate group from the reptiles and humans a separate group from the apes. This is not a matter of which perspective gets the facts about nature aright (the trees are the same on both perspectives)—rather, it is a matter of which perspective has a more acceptable philosophical underpinning, as well as of which classificatory terminology one employs (widely used everyday terms from folk biology, terms from similarity-based taxonomy, etc.). While it is clear that the debate between cladists and evolutionary systematists is of major importance for biology, it is also clear that the debate will not be settled by scientific investigation. The issue is a philosophical one, pertaining to which aims can be set for the classificatory aspect of scientific work and which knowledge can be contained in particular scientific classifications. As such, the debate can be used to introduce students to a key philosophical aspect of the nature of science.

Two additional points to be taken into account when teaching about phylogenetic systematics concern the nature of the trees and cladograms that are produced in phylogenetic analyses and the nature of their constituent basic units, i.e., species.

O’Hara (1997, p. 327) observed that although biologists no longer understand cladograms and phylogenetic trees as representing an evolution of simpler forms toward more complex forms (as was the idea underlying the scala naturae), many students and members of the general public still do. Thus, O’Hara argued, “Just as beginning students in geography need to be taught how to read maps, so beginning students in biology should be taught how to read trees and to understand what trees communicate.” (O’Hara 1997, p. 327). Indeed, empirical studies have shown that students often have profound difficulties with interpreting cladograms and phylogenetic trees and understanding what information they convey (Baum et al. 2005; Novick and Catley 2006: Meir et al. 2007; Baum and Offner 2008; Gregory 2008). For example, students often incorrectly interpret the longest line in a diagonal diagram (e.g., the line on the right in Fig. 1a, ending with taxon H) as being the supporting structure, i.e., the main line in evolution that leads to a distinct endpoint and from which the other lines (leading to taxa A–G) branch off as “side alleys” (Novick and Catley 2006; Meir et al. 2007; Gregory 2008). Teachers thus need to point out that such an interpretation is mistaken and that the only information conveyed by cladograms and phylogenetic trees is how closely related (in terms of recency of shared ancestry) the groups included in the tree are.

The question of the nature of species is a central problem in the context of biological systematics, but also a problem that profoundly affects the whole of biology. As was remarked above, systematics is not directly concerned with classifying organisms into species. Species are taken as given basic units, identified by field and museum taxonomists and represented by their member organisms and their characters, and the concern of systematics is to classify them into higher taxa and these, in turn, into even higher taxa. Nevertheless, the ongoing controversies on the nature of species affect both the everyday practice of systematists (see, for example, Wheeler and Meier 2000; Kearney 2007, pp. 222, 225), as well as of virtually all other subdisciplines of biology. It is to this issue that I turn next.

4 The Species Problem

4.1 The Roots of the Species Problem

At first sight, the various species of organisms that feature in biology textbooks, research articles and natural history museum exhibits—Escherichia coli, Tuber melanosporum, Arabidopsis thaliana, Drosophila pseudoobscura, Homo sapiens and their ilk—appear to be for biological science what the chemical elements listed in the Periodic Table are for chemistry and the kinds of elementary particles in the Standard Model for fundamental physics: the fundamental kinds of things that constitute part of the basic furniture of the world and are studied in these respective domains of science.

For example, at the heart of chemistry is the study of the properties and composition of substances and of how substances change into one another in chemical reactions.¹⁹ In their studies of these subjects, chemists employ a large variety of kinds with different degrees of inclusivity, ranging from kinds of compounds and more inclusive kinds such as acid and base, via kinds of molecules to kinds of simple substances and kinds of atoms. On the fundamental levels of chemical classification are the elements (H, He, Li, Be, …), the isotopes (³He, ⁴He, …) and, more inclusive, kinds like alkali metal, metal, halogen, etc. Amidst the various kinds of chemical entities that feature in chemical classification the elements seem to occupy a privileged position. More inclusive kinds are usually specified in terms of the constituent elements, as is obvious in the structural formulae that are used to denote both single molecules and macroscopic substances, but can also be seen from the definitions of compound substance in terms of their constituent molecules, which in turn are specified in terms of their constituent atoms.²⁰ Moreover, the typical properties of the members of more inclusive kinds are often explained in terms of the typical properties of their constituent member entities by referring to the properties of the constituent atoms and the positions that these occupy in the spatial structure of the molecule or substance.

In biology, a largely similar situation seems to obtain. In their studies of phenomena in the living world, biologists employ a variety of kinds with different degrees of inclusivity, ranging from kinds of ecosystems, biomes and local biotic communities (e.g., chaparral, temperate deciduous forest, tundra, polar community, snowy-forest community, etc.)²¹ via kinds of ecological roles (primary producer, trophic specialist, detrivore, predator, etc.) and the various levels of taxa in systematic biology—from kingdoms or domains at the most inclusive level, via phyla or (in older terminology for plants and fungi) divisions, classes, orders, families, and genera to species, subspecies and varieties—,²² to kinds of organismal and cellular parts (the various kinds of organs and parts, and of cell organelles) and kinds of genes (antennapedia, engrailed, eyeless, Pax6, DFNA1, etc.). But species appear to occupy a special position among the various biological kinds in a number of aspects.

For one, the typical properties of kinds of biological entities at higher levels of organization, such as kinds of biomes, are explained in part by referring to the typical properties and behaviors of the species of organisms that constitute them. In addition, kinds on lower levels of organization, such as kinds of organs, kinds of organismal parts and kinds of genes, are usually specified with respect to the species of organisms in which they occur. For example, biologists do not simply study how genes of the kind diaphanous function in the context of organismal development. Rather, they study diaphanous in fruit flies or diaphanous in humans and recognize D. melanogaster diaphanous and H. sapiens diaphanous—a.k.a. DFNA1, or diaphanous homolog 1 (Drosophila), or DIAPH1—as distinct subkinds of the kind diaphanous, that is, as kinds that constitute homologs with respect to one another within the overarching kind diaphanous (Reydon 2009, pp. 421–423). Similarly, biologists do not simply group organismal parts and structures that share a common ancestral part or structure into the same kind, but distinguish between the parts or structures of different species as kinds that constitute homologues of one another. For example, biologists don’t just investigate the various kinds of bones in vertebrate limbs, such as the ulna (elbow bone), as the basic kinds of bones about which new knowledge needs to be obtained, but rather examine human ulnae, Beluga whale ulnae or fruit bat ulnae, thereby connecting the kinds of organismal parts as homologous kinds to the species in which these parts occur (Reydon 2009).

Thus, species of organisms occupy an epistemologically special position among the biological kinds, even though it would be overstating it to say that species are the single most important kinds of biology. Species are where most (although probably not all) biological knowledge comes together in much the same way as in chemistry the elements are where much of chemical knowledge coalesces. As Mayr (2004, p. 171) remarked, “[m]ost research in evolutionary biology, ecology, behavioral biology, and almost any other branch of biology deals with species.” Similarly, Cracraft (2000, p. 6) pointed out that “time-honored questions in evolutionary biology—from describing patterns of geographic variation and modes of speciation, to biogeographic analysis and the genetics of speciation, or to virtually any comparison one might make—will depend for their answer on how a biologist looks at species.”

But species also occupy an ontologically special position: while there remains much discussion about the reality of taxa and many biologists tend toward the view that higher taxa, such as genera, families and phyla, do not constitute real divisions in nature, most biologists continue to view species as real entities in nature (e.g., Claridge 2010; but see Mishler 1999, 2010). In the philosophy of science this is reflected in the fact that philosophers of science often mention biological species and chemical elements in the same breath as paradigmatic examples of natural kinds, i.e., kinds of entities that exist in nature independently from human classificatory practices. That there are such natural kinds is a deeply rooted view in Western philosophy that can in principle be traced back to the works of Plato and Aristotle.²³ According to this tradition of thought, there is a natural order in the world that we humans can try to recognize, that we can try to represent to a higher or lesser degree in the classifications of things we construct, and that we can try to obtain knowledge about by means of our scientific investigations. As Bird and Tobin (2008; original emphasis) put it:

To say that a kind is natural is to say that it corresponds to a grouping or ordering that does not depend on humans. We tend to assume that science is successful in revealing these kinds; […] when all goes well the classifications and taxonomies employed by science correspond to the real kinds in nature. The existence of these real and independent kinds of things is held to justify our scientific inferences and practices.

Thus, some classifications succeed in picking out kinds that really exist in nature and accordingly the kinds that appear in these classifications occupy an ontologically special position among kinds in general. In addition, the reality of the kinds accounts for the special epistemological position that these kinds occupy: they are where the knowledge of a particular domain of investigation comes together, because they constitute that part of reality that this domain investigates.

On this picture of how scientific classification relates to the world, different sciences address different parts of the natural order of things at different levels of organization, ranging from elementary particles on (at least for the moment) the most fundamental level of organization all the way up to planets, stars, galaxies, etc. on the highest levels of organization. Thus, it is a brute fact about the world (or so it seems at least) that there are numerous sorts of atoms in the world that naturally fall into kinds like Helium and Carbon, independently of whether or not there are humans who actually group the various atoms together into kinds (even though humans are required to give the kind a name, of course.) Similarly, it seems a brute fact that there are Drosophila melanogaster fruit flies, Lares argentatus sea gulls and myriad other kinds of organisms “out there” in the world, independently of whether there are humans who actually group them together into different species of organism and call these species ‘Drosophila melanogaster’, ‘Lares argentatus’, and so on. Accordingly, biological species are commonly listed as paradigmatic natural kinds, along with the chemical elements and the kinds of elementary particles in the Standard Model (e.g., Boyd et al. 1991, p. 778; Currie 1996; Daly 1998; Koslicki 2008).

On a closer look, however, it becomes difficult to endorse the view that biological species are natural kinds and that they function in biological science in a similar way as the chemical elements in chemistry or the kinds of elementary particles in particle physics.

First of all, there is the issue of variation. Natural kinds are traditionally associated with essences: all member entities of a particular natural kind are the same in particular respects, that is, they all share a particular set of essential properties, the possession of which is characteristic for the members of the kind. According to the traditional view of natural kinds, all and only the member entities of a particular kind possess the set of essential properties associated with this kind, and possession of these properties is both necessary and sufficient for being counted as a member of the kind. However, for species of organisms there are no such essential sets of properties. To the contrary: variation between the organisms of a species is a crucial feature of the living world. As evolutionary theory explains, species are subject to open-ended evolutionary change such that in principle every trait can be lost at some stage of a species’ evolution and novel traits can come into being. Thus, upon consideration of a species over its entire period of existence there are no traits that are necessarily exhibited by all and only the member organisms of a species. Also, evolutionary theory tells us that for evolution to occur, there must be variation between the members of a species at any one time: natural selection needs trait variation to operate on. The required variation can in principle occur with respect to any trait that an organism possesses, so when considered at a particular time too it will be unlikely that there are traits that are exhibited by all and only the members of one particular species. In addition, for those traits that are shared by all member organisms of a species throughout its lifetime, without exception, it is very likely that these traits are conserved over time periods that extend far beyond the lifetimes of the individual species and thus will be exhibited also by members of ancestral and descendant species of the species in question (Reydon 2006; Ereshefsky 2010a, Sec. 2.1). Such traits, then, aren’t unique to the members of one species.

Related to this issue is the fact that there are no biological laws of nature that refer to species. Often the special position of natural kinds is associated with the role of laws of nature in science: natural kinds are often conceived of as those kinds of things that laws of nature are about. This is what makes natural kinds both epistemologically and ontologically special: science searches for the laws of nature that govern the world and, consequently, the kinds that feature in these laws take up a special position among the rest of the kinds used in the sciences.²⁴ However, biologists have so far not uncovered any laws of nature about any species and, moreover, there is a general consensus among philosophers of biology that there are no, or at most a handful, proper laws of nature in biological science (e.g., Rosenberg 1994; Beatty 1995; Waters 1998).

Another difficulty pertains to the extensions of species in space and time. As products of evolutionary processes and entities that undergo evolutionary change during their time of existence, species are limited in space and time. Every species comes into being at a particular time and location as a descendant of an ancestral species, it undergoes evolutionary change, and it becomes extinct at some later time. Kinds, however, are supposed to be universal in the sense that a kind does not come into being and members of a kind can exist anywhere in the universe where the conditions are right. Even though at early stages of the universe there were no gold atoms, for example, the kind gold is eternal and gold atoms may exist at any place and at any time. Gold atoms do not come into being as descendants of other gold atoms. Drosophila melanogaster, in contrast is not eternal: there could not have been any D. melanogaster organisms at times before the speciation event in which D. melanogaster branched off from its ancestor species and there cannot be any D. melanogaster organisms after the species has gone extinct. And D. melanogaster fruit flies come into being as the offspring of other D. melanogaster fruit field: if one finds a fruit fly that looks in all respects like a D. melanogaster, but is not related to other fruit flies that are D. melanogaster organisms, the former fly just isn’t a D. melanogaster! Thus, species are spatiotemporally limited and internally continuous entities. As such, they do not fit the traditional picture of natural kinds as the ontologically and epistemologically special kinds of things that the sciences focus on.

Due to these difficulties with respect to conceiving of species as natural kinds, the suggestion has been made that species in fact aren’t natural kinds of organisms at all. Instead, it was suggested, species should be conceived of as concrete entities of which organisms are the constituent parts.²⁵ On this view of what species actually are, which has come to be known as the “species are individuals” thesis, species are things of a fundamentally different sort than the chemical elements or the kinds of elementary particles. They are not groups, kinds or classes of things, but rather are themselves concrete, individual things, material entities that exist in space and time and the constituent parts of which are living organisms.

4.2 The Species Problem as a Complex of Questions

Despite the fact that quickly after it was suggested this view of the nature of biological species has come to be widely endorsed by both biologists and philosophers of biology and the fact that it fits well with how species feature in the context of systematics (see Sect. 3), it has not been able to resolve the controversies on the nature of species once and for all. Thus, what has come to be known as “the species problem” continues to be a subject of debate in the biological and philosophical literature. As is typically the case for philosophical problems, different authors conceive of the species problem in different ways.²⁶ Grene and Depew, for example, thought of it as the general question, “What are species?” (2004, p. 291). Stamos (2003, p. 1) formulated it as the question, “Are biological species real, and, if real, what is the nature of their reality?” Hey (2001, p. 4) described it as being constituted of “the related uncertainties of how to define the word ‘species’ and how to identify actual species in nature.” For Wilkins (2009a, p. 173), “[i]t is the problem of defining what rank it might be that species achieve when they become species.” According to Okasha (2002, p. 106), “there is the problem of how to sort organisms into species, known as the ‘species problem’. […] Secondly, there is the problem of how to arrange a group of species into higher taxa […].” And according to Richards,

This then is the species problem: there are multiple, inconsistent ways to divide biodiversity into species on the basis of multiple, conflicting species concepts, without any obvious way of resolving the conflict. […] Lurking behind the species problem are two philosophical worries. On the first realism worry, we might […] doubt that species are real things in nature. […] We might also worry about pluralism. If species are real, is there some one kind of species thing, or are there multiple kinds of species things? (Richards 2010, pp. 5, 10–11; original italics)

In addition to these different questions that are subsumed under the species problem, Mayr (1982, pp. 253–254; 1996, pp. 267–268; 2004, p. 174) distinguished between the species taxon problem and the species category problem (see also Wiley 1981, p. 21; Panchen 1992, p. 338; Ereshefsky 2001, p. 80; Richards 2010, p. 15). The species taxon problem is the question how to delimit species taxa and how to allocate individual organisms to species. The species category problem is the question how the term ‘species’ is to be defined, that is, what makes particular taxa into species rather than taxa on another level of the Linnaean hierarchy and whether there is any single characteristic that all species have in common that distinguishes them from other groupings of organisms.

Thus, the species problem is not a single problem but a complex of questions that encompasses at least the following questions:

• 1. What are species? What is the nature of those things biologists call species and denote with binomials such as ‘Drosophila melanogaster’ or ‘Lares argentatus’?

• 2. In particular, are species real entities (whatever their precise nature might be), or are they artificial groupings of organisms constructed by us for particular purposes?

• 3. What, if anything, distinguishes species (as entities and as a rank in the Linnaean hierarchy) from other biological taxa (and ranks)? What makes a particular group of organisms into a species, rather than a genus, a family, a subspecies, a variety, etc.? That is, is there a single answer to question (1)?

• 4. On what basis are individual organisms allocated to species? What is it that makes a given organism a member of one species rather than another?

On each of these questions there is a considerable amount of disagreement among biologists and philosophers of biology. For none of these questions is a definitive resolution within sight.

An aspect of the discussion in which these disagreements are most visible is the manifold of so-called species concepts, i.e., definitions of the scientific term ‘species’, that are available in the literature.²⁷ A species concept, such as the Biological Species Concept (BSC), the Evolutionary Species Concept (ESC) or the Phylogenetic Species Concept (PSC), provides an answer to the question, What is a species?, and as such constitutes a potential basis for answering the other three abovementioned questions. For example, according the BSC, the species concept that was advanced in the 1940s by Ernst Mayr and today still is the widely used textbook definition of ‘species’, species are “groups of actually or potentially interbreeding natural populations, which are reproductively isolated from other such groups” (Mayr 1942, p. 120). This definition answers question (1) above and indirectly answers questions (2)—species are real, as they are systems of actually or potentially interbreeding populations—and (4)—organisms belong to the same species if they are able to produce fertile offspring. However, it doesn’t provide an answer to question (3). Other species concepts answer these questions differently. In the contemporary biological literature around 30 distinct species concepts are in use (for overviews, see Wiley 1981, pp. 21–37; Panchen 1992, pp. 337–338; Mayden 1997; Wheeler and Meier 2000; Ereshefsky 2001, pp. 80–92; Wilkins 2009b, Sect. 11), but the actual number of definitions that have been advanced throughout the history of biology is an order of magnitude higher (for lists of more than 100 definitions, see Lherminer and Solignac 2000; Wilkins 2009b).

4.3 Relevant Issues for Science Teaching

When it comes to the teaching of biology, however, I believe that what is important for students to understand is not what all the different species concepts used in different fields of work and proposed by different authors say in detail. After all, definitions in science are being modified all the time, and old species concepts are abandoned and new ones introduced.²⁸ Rather, it is important to understand why there is a diversity of concepts in the first place and what this diversity of concepts means for biological practice. In my view, this suggests that the following core issues from the philosophy of science might fruitfully be addressed in teaching contexts: essentialism, pluralism and realism.

The issue of essentialism is relevant when it comes to dealing with widespread misconceptions of students about biological kinds. As was mentioned in Sect. 4.1, biological species are often seen as natural kinds of organisms which are all the same in some respect, in the same way as chemical elements are kinds of atoms which are all the same in a particular aspect of their material composition. This way of thinking about kinds is what philosophers refer to as essentialism: the idea that members of a kind all share the same essence, which gives them their identity as things of that particular kind and which is responsible for observable similarities that the kind’s members share. Simply out, a particular organism is a tiger because it possesses or instantiates the unknown tiger essence and it looks like other tigers because the shared tiger essence causes them to have highly similar properties. Studies have shown that students often conceive of biological species as natural kinds of organisms that all share some (generally unknown) essential property that organisms of other species do not possess (e.g. Shulman 2006). Psychological research suggests that such views are innate, as essentialist thinking seems to be a reasoning heuristic used by children and adults from a variety of cultural contexts (e.g., Gelman and Hirschfeld 1999; Gelman 2003). Thus, teachers should be aware of the widespread presence of essentialist thinking among their students and of the conflict between essentialist thinking and the biological understanding of the nature of species as is discussed above: students may think in terms of tiger essences, but biology tells us that there is no such thing as a tiger essence! If essentialist thinking is indeed an innate predisposition, as psychological research suggests, it will be present in every new generation of students and can thus constitute a considerable barrier for students’ attempts to understand central aspects of the nature of biological science (Van Dijk and Reydon 2010, p. 663).

The issue of pluralism comes into play when introducing students to the way biologists deal with the species problem in research practice. The present diversity of species concepts is largely due to the fact that so far no single concept has been found applicable throughout all of biological science, to all organismal groups from microbes to higher animals and in all contexts of biological investigation. As a consequence, biologists have felt the need to come up with their own species concepts, applicable to the particular organisms they study and in the particular investigative context in which they work. Accordingly, many philosophers of biology today endorse a pluralist stance with respect to species concepts, according to which biological science needs a plurality of species concepts and different areas of biology classify organisms into species on the basis of their own particular species concept which doesn’t necessarily apply anywhere outside its own area of work (for various views of pluralism, see Kitcher 1984; Dupré 1993, 1999; Wilkins 2003; Grene and Depew 2004, p. 302; Ereshefsky 2001, pp. 129–195; 2008, pp. 104–107; 2010a, Sec. 3; Richards 2010, pp. 115–124). Pluralism with respect to species concepts receives additional support from the increasing acknowledgement among philosophers of science that there is no good reason to assume that one single classificatory system would be able to capture the actual diversity of things in the world—given that the world may well be a more complex place than was imagined at first, a plurality of classificatory systems might be required to deal with the actual diversity of things (Dupré 1993).

In this context the species problem can be used to show students a crucial aspect of the nature of science: that science isn’t a homogeneous enterprise. When examining science at a particular point in time, it becomes clear that not only do different sciences (physics, chemistry, biology, …) work in different ways, but also do the various fields within one science, in this case biology, work in quite different ways. Instead of suggesting to students that science is a unified domain that is characterized by certain properties (a particular scientific method, for example), it should be made clear that the whole of science consists of a large variety of fields, each with its own aims, methods and ways of classifying the subject matter under study. Within biology, this diversity of fields is reflected by among other things the fact that one scientific term may have multiple meanings, both at different stages of the development of a science and at the same time. The available species concepts are, after all, different definitions of the term ‘species’ that represent diverging views of what a species is supposed to be and on what basis organisms belong to their species.

In addition, the species problem also can be used to show how the meaning of a scientific term can change and even multiply through time as science progresses. In my view, the species problem is fundamentally a problem of meaning change: what happens is that at a particular stage of the history of a field of investigation a term is introduced with a particular meaning, later developments cause the meaning of the term to change while some people in the field retain the old meaning and others adopt the new meaning, and much later we find a situation in which the term is used in a number of different meanings simultaneously. Such cases are not uncommon in the sciences and have long been a topic of attention for both scientists and philosophers of science. In 1855, for example, Claude Bernard (the founder of experimental physiology) already pointed to the occurrence of meaning changes of scientific terms:

When we create a word to characterize a phenomenon, we then agree in general on the idea that we wish to express and the precise meaning we are giving to it; but with the later progress of science the meaning of the word changes for some people, while for others the word remains in the language with its original meaning. The result is often such discord that men using the same word express very different ideas (Bernard 1957 [1927], p. 188)

Since the influential works by historian/philosopher of science Thomas Kuhn and philosopher of science Paul Feyerabend from the 1960s, meaning change is science is an important theme in philosophy of science (for an overview, see Oberheim and Hoyningen-Huene 2010).

Often new meanings will displace older meanings of a scientific term, for example when a new meaning corrects mistakes that were included in the older meaning. In such cases, the discord that exists when different scientists use the same term in different meanings will be of a temporary nature. However, there may also be cases in which the different meanings pick out different, equally interesting aspects of nature. In such cases there is no reason to abandon one meaning in favor of the other and the term in question will continue to be used with multiple meanings. Elsewhere (Reydon 2004, 2005), I have argued that this is the case with respect to the term ‘species’ and that at present the term is used in biology as a homonym with four distinct meanings.²⁹ While for everyday practice in biological science this situation may be confusing, in teaching contexts in my view it constitutes a good illustration of how messy—and interesting!—real science can be.

Finally, the issue of realism (or better, scientific realism), a central topic in the philosophy of science, surfaces in the context of teaching students about what biological classifications say about the world. It is the question whether our best scientific theories provide an adequate picture of reality and whether the terms featuring in these theories refer to real aspects of the world “out there” (e.g., Devitt 2008, pp. 224–226; Boyd 2010; DeWitt 2010, pp. 21–22). In biology, the issue of realism arises with respect to the various categories in the Tree of Life (i.e., species taxa and taxa above and below the species level), but also with respect to, for example, the concepts of race and variety. Many biologists today tend to conceive of taxa above and below the species level as not being real. Thus, while the grouping of species into higher taxa reflects real relations of shared most recent ancestry, the ranking of these higher taxa as genera, families, phyla, etc. is merely a matter of convenience (Grene and Depew 2004, pp. 299–300). That is, recognizing a monophyletic group of species as reflecting the real fact that these species share a most recent common ancestor, which they don’t share with other species not belonging in the particular group, is unproblematic, but ranking this group as an order rather than a family is just a matter of choice.³⁰

This view of taxa above and below the species level notwithstanding, because of their special position among the other taxa in the Linnaean hierarchy many biologists consider species as uniquely real divisions in nature, that is, as the basic kinds of organisms that exist in the world (e.g., Panchen 1992, p. 333). However, not all biologists and philosophers of biology hold this view of species and a number of authors have argued that the species category should not be thought of as a real category of kinds (e.g., Mishler 1999, 2010; Ereshefsky 2010b). As in the case of higher and lower taxa, this view entails that although there are real divisions between groups of organisms in nature, these divisions do not delimit species: “there are such real entities deeply nested inside each other, with no one level fundamental or unique. Species are real, but not in a unique and special way.” (Mishler 2010, p. 119). As species, then, the taxa usually thought of as species aren’t real—we could also have thought of them as genera or varieties, for example.

Indeed, Darwin himself seemed to have held a view of the species level as constituting an artificial status for groups of organisms. In a much-quoted passage from the Origin of Species, for example, Darwin asserted:

… that I look at the term species, as one arbitrarily given for the sake of convenience to a set of individuals closely resembling each other, and that it does not essentially differ from the term variety, which is given to less distinct and more fluctuating forms. The term variety, again, in comparison with mere individual differences, is also applied arbitrarily, and for mere convenience sake (Darwin 1859, p. 52)

Elsewhere in the same chapter of the Origin, Darwin stated not to be concerned with defining the notions of ‘species’ and ‘variety’ (Darwin 1859, p. 44) and to see the experience of naturalists who are intimately familiar with the various properties of the organisms they study as the only guide to follow when deciding whether a particular organismal form should be ranked as a distinct species or a mere variety of a species (Darwin 1859, p. 47; see also Ereshefsky 2010b, pp. 406–410). Thus, on Darwin’s view—as well as the view of a number of contemporary biologists and philosophers of biology—the ranking of (real!) groups of organisms as species is merely a matter of convenience and personal opinion: if an experienced naturalist finds a particular form to constitute a distinct species, so be it; if (s)he denies it the status of species, then that is the way it is.

In this respect, understanding the nature of the species problem contributes to understanding a central aspect of the nature of biological science. For one, it is not self-evident that scientific classifications tell us what kinds of things there are in the world “out there”. That is, realism about kinds of organisms is not unproblematic. Moreover, biologists and philosophers of biology disagree about exactly what it is that the classification of organisms into species and higher taxa does tell us. These issues arise with respect to classification in all the sciences, but the specific form they take here is unique to biological classification. Discussing these issues with students, thus, will introduce them to important features of the nature of biological science.

5 Concluding Remarks: What to Teach About Classification and Biodiversity

In the preceding sections, I have argued that biological classification is different from classifications found in other fields of science and that, therefore, understanding how biological classification works constitutes an important element in efforts to teach and understand the nature of biological science.

Classification in biology distinguishes itself from classification in other sciences in at least two ways: (1) many classifications used in the biological sciences are phylogenetic in nature, that is, they are based on the common descent of organisms; and (2) biological classifications often are in part functional classifications, that is, biological entities are often grouped together on the basis of shared function. To be sure, these are not features that apply to all biological classifications or exclusively to biological classifications. Some classifications in biological science, most importantly the classification of organisms into species and higher taxa, do not have a functional component. In addition, the use of function-based classifications is not unique to biology, but is found in other sciences too (such as chemistry, psychology and cognitive science). Moreover, biological classifications typically aren’t based on one single criterion, such as common descent of shared function, but rest on multiple criteria. Still, functionality and phylogeny constitute central elements in biological classifications and, therefore, should stand at the focus of teaching how classification in biology works (even though I have largely ignored functional classifications in this paper).

In addition to the centrality of phylogeny and functionality in biological classifications, the nature of the basic biological kinds, species, constitutes a crucial element in the understanding of the natural of biological classification. The species problem occupies a unique position within the philosophy of science. For one, it has been one of the main driving forces behind the development of the philosophy of biology as an independent subfield of the philosophy of science (Haber et al. 2010, p. 184). But more importantly, there are no compatible problems in the philosophies of the other sciences: there is, for example, no “particle kinds problem” in the philosophy of physics and no “element problem” in the philosophy of chemistry (although philosophers of chemistry increasingly consider the foundations of chemical classification and the question whether chemical kinds are natural kinds; e.g., Vihalemm 2003; Scerri 2005; Hendry 2006).

Notwithstanding the fact that the species problem remains unresolved, much of biological science seems to work quite well without having an ultimative, all-purpose definition of the term ‘species’. This may suggest that the species problem is a philosophers’ problem without many consequences either for scientific practice or for societal or educational issues. This would, however, amount to an underestimation of what is at stake. In scientific practice, the species problem manifests itself in the form of differences of opinion on which classifications of particular groups of organisms are the correct ones. Such differences of opinion may, however, have profound consequences for the real world, for instance in the context of biodiversity conservation. Often, conservation efforts are aimed at the conservation of particular species of organisms. However, given the different views of what species are and on the basis of which criteria organisms are to be allocated to species, how do we know what to conserve?³¹

Note that on the view that the species level does not occupy a special position in the Linnaean hierarchy and that the ranking of taxa into species, genera, etc. is a matter of convenience, the species problem should not affect conservation issues. Thus, according to Mishler (2010, p. 118), “biodiversity isn’t species—biodiversity is the whole tree of life, not just the arbitrary place at which species are named. […] Species are not comparable between lineages in any manner, just an arbitrary cut-off somewhere along a branch in the tree of life. […] Biodiversity is a much richer tapestry of lineages and clades.” Mishler suggested that instead of counting species, phylogenetic measures of biodiversity should be used, that measure biodiversity in terms of the number of nodes in the tree or the length of the branches in the tree between two taxa.

In a recent book on the notion of biodiversity, philosophers of biology James Maclaurin and Kim Sterelny agreed that if the grouping of organisms into species is merely a matter of convenience and personal opinion, without reflecting any aspect of the state of affairs “out there” in nature, measuring biodiversity in terms of species and undertaking action to conserve species is unwarranted (Maclaurin and Sterelny 2008, p. 27). But they argued that:

… there is a sensible motivation that warrants fixation on species. Species are empirically accessible. Phenomenological species are observable, identifiable, and reidentifiable aspects of the biological world. Moreover, phenomenological species correspond, in many cases, to evolutionarily significant lineages in the tree of life. So this approach to biodiversity does capture something real, despite the complexities of the species problem (Maclaurin and Sterelny 2008, p. 29).

Thus, Maclaurin and Sterelny favored a phenomenological approach to resolving the issue, instead of relying on measures based on phylogenetic theory. Deciding between these approaches, however, involves resolving part of the species problem: is there a special level of order in the living world that can be considered to encompass the fundamental kinds of organisms? This presents an additional reason to pay special attention to the species problem when teaching about biological diversity.

In addition, the considerations presented in this paper in my view have implications for the position that teaching about biological classification should occupy within the teaching about biology in general. Most importantly, they speak against addressing the topic of classification and introducing students to the various kingdoms or domains of the living world, discussing exemplary phyla, classes, orders, families and species from a morphological perspective, early on in biology teaching.³² The idea of evolution needs to be addressed first, before classification comes into play. Discussing classification before evolution runs the risk of presenting the Tree of Life as if it were a classificatory system of the same nature as, say, the Periodic Table. But, as I have argued, the Tree of Life represents evolutionary history and relatedness of descent, rendering it a fundamentally different classificatory system than the Periodic Table. This means that students will not be able to fully appreciate the particular nature of the classification of organisms into species and higher taxa if they are not first introduced to the fundamentals of evolutionary thinking.

The same holds for functional classifications in biology, at least on some of the main notions of biological function available in the literature that conceive of biological functions in terms of selection history (e.g., Williams 1966; Millikan 1989; Neander 1991; see Sect. 2.3). The very nature of biological classification—both for classifications based on common descent and for such function-based classifications—thus provides an argument for giving the teaching of evolution a central position in the teaching of biology (see also Van Dijk and Kattmann 2009). There is a well-worn quotation from the famous geneticist Theodosius Dobzhansky: “nothing in biology makes sense except in the light of evolution” (Dobzhansky 1964, p. 449; 1973, p. 125). This is certainly true for much of (albeit perhaps not all of) biological classification and, therefore, should be realized by everyone involved in the teaching of biology.