Abstract
The purpose of the present paper, the first of a set of four dealing with different, though closely interconnected, problems, is to prove that all realistic solutions advanced in order to settle the age-old problem whether “species” are real entities existing in Nature or are abstractions, though based on the observation of populations actually existing in Nature, may be falsified. The author first examines the logics of the concept of species and shows that a “species” is the kind of intellectual tool that medieval logicians called a “universal”. Evolution actually occurs in natural populations and we may speak of the “evolution of species” only per translatum. The author thence proves that all the realistic concepts of “species” fail to account for a number of actual situations occurring in nature and must therefore be dismissed. Finally, the author points also to the difficulty of arriving also at a precise definition of the concept of “population”.
Introduction
My work on evolutionary problems concerning various phyla and geological periods led me to consider and doubt the validity of some commonly followed theories and methods in evolutionary and especially phylogenetic studies. So, from time to time (Simonetta Citation1983, Citation1986, Citation1988, Citation1992, Citation1993, Citation1995, Citation1996, Citation1999a,b, Citation2000, Citation2007), I have challenged realist concepts of species, cladistic theories and practices, the possibility of a reliable “molecular clock” and the supposed structure of the earliest metazoans. I had honestly hoped for open, perhaps hot, debates, as I obviously realise that colleagues who have assumed that “species” are real natural entities, and/or have used cladistics as the basic tool for their phylogenetic and/or taxonomic work, will be most reluctant to give up cherished assumptions or the paraphernalia of cladistics and try to find new, alternative ways to tackle problems that they deemed to have settled. Again: my doubts on the molecular chronologies and phylogenies are certainly obnoxious to the scholars that endeavoured to use molecular evidence for dating the steps of animal evolution or the relationships among the major clades, etc.
A bit more palatable may be my ideas on the organisation of the earliest metazoans, but I doubt it.
Thus, as I said, I had hoped for a hot, perhaps angry, reception of my arguments and a really fruitful and clarifying debate. This did not follow as, apparently, on one side most cladists or scholars of molecular phylogenetics either do not read their critics’ papers, or consider that it is not worth the trouble to answer their criticisms in detail. Other colleagues may have been reluctant to discuss general and apparently common-sense ideas supported by first-class biologists in the light of problems of general pure logics and debates dating from hundreds of years ago.
As age has not damped my naughty trends, I have thought to fire once more all my guns into some entrenched opinions with the hope of producing some damage. Thus, I prepared a set of four papers that cover different main topics. However, as I hope to show, the conclusions of each one seriously impinge on the others and, when considered together, they require some re-thinking of some basic concepts and common assumptions in the field of evolutionary studies.
The purpose of this first essay is to prove that the real objects of evolutionary studies, what evolves in nature, are not “species”, these being abstract concepts, but single, individual populations, though in quite a few instances what in nature works as a population cannot fit in any possible concept of “species”.
In my previous papers, I have touched on various aspects of the nature of taxa, obviously including the so-called “species problem”, as this clearly impinges on any discussion of the principles and methods towards an understanding of the mechanisms of evolution, of phylogenetic reconstruction and of systematics.
Indeed, I consider that the basic aim of any study on evolution is either to contribute to the understanding of the mechanisms which determine the evolution of living organisms or to clarify the genealogies of some or all living beings, their mutual relationships and historical interactions, and the how and why of the changes in the biosphere since the appearance of the first living beings.
Therefore, I have thought that it might be useful, as a preliminary to a renewed discussion on some methods of phylogenetic reconstruction, to submit a fresh criticism of the concept of “species”:
Are “species” real entities in nature and are they the real natural units of evolution or, “species” being indeed an intellectual instrument in the classic nominalistic tradition, should we rather consider single populations to be the natural and real evolving entities?
As a premise to any further discussion on evolutionary systematics, I shall prove that the age-old problem of the basic nature of collective groups of beings and objects and of their relationship with the names and concepts by which they are “known” is subsumed, whether consciously or, more often, unconsciously, in any discussion on the nature of “species” in biology, and the various concepts of “species” advocated by different scholars have been quite relevant in the development of the various evolutionary models proposed. Darwin himself was conscious of its implications and, though he used the word “species” even in the title of his masterpiece, was a declared nominalist – that is, he held that “species” are concepts, abstractions that we devise from our experience of the real populations that we may study and that they are necessary for purposes of communication.
As we shall see, the core of the problem is still the argument about “universals”, a debate that goes back to ancient Greece and was systematically debated by medieval logicians. This problem, as far as the biologist is concerned, is quite relevant both for the reconstruction of phylogenies and for the formal classifications stemming from them.
The biologist’s immediate concern is to decide whether taxa or, better, at least “species” really exist in nature and therefore evolve, or whether taxa including “species” are convenient, not to say necessary, intellectual devices, but that what evolve in nature are single populations. As I said this is, indeed, a classic general problem and its solution requires a more general historical and empirical approach than that from the narrow outlook of the biologist and, therefore, it is necessary to provide a brief outline of the story and relevance of the debate on the “universals”, as these “things” were called by medieval scholars, or “classes” in later logics.
As the core of the problem has clearly been debated almost since the birth of philosophy, at least some hallmarks of the ancient developments of the ideas on the problem are, I think, unavoidable. This, however, involves a problem with “references”: the ideas of many ancient thinkers, whose writings are lost, are known to us from their quotations by later authors and in other instances, especially for medieval writers, are almost buried in more or less old editions of difficult access. As this makes it almost impossible to provide references according the standards of modern scientific journals, the reader is thence referred to a few basic treatises, where, if interested, he may find the necessary guidance to a more complete bibliography (Russel, 1946; Cohen & Nagel 1963; Copleston 1972).
As the ultimate target of this section is a proper understanding of the true nature of a basic intellectual tool in biology and especially in evolutionary studies and, as a consequence, in systematics, let us preface our argument with a brief consideration of the purposes of phylogenetic reconstructions and of taxonomic classifications.
The purposes of the two endeavours – phylogenetic reconstruction and fitting the evidence into a practical taxonomic frame – are clearly different, though interrelated. The purpose of taxonomy is just to allow the scholars to group (I am tempted to say: to pack) the known organisms into nested assemblages, so that a mutual understanding of the characters and limits of the objects of our verbal or written statements may eventually be possible.
Dichotomous classifications of organisms had occasionally been used by scholars previous to Lamarck, but dichotomous keys devised for identification purposes were first introduced by Lamarck (Citation1779), and are perfectly suitable for the identification of the specimens belonging to already-known taxa and usually also for the recognition of new ones. They do not necessarily consider any biologically relevant character, but rather focus on such features that are easy to check on the specimens and are as unambiguous as possible. Likewise it is quite possible, by an appropriate choice of a set of characters, to devise an entirely conventional taxonomy of nested groups of taxa, which may be quite suitable for any practical purpose.
However, since the general acceptance of evolution, no one has doubted that it was desirable to use a formal taxonomy mirroring as far as practically possible such phyletic relationships among organisms as it was possible to ascertain. Unfortunately, as we shall see, the debate about phylogenies often forgets that what actually offer their variability to selection to operate are the individual members of single populations, not such collective entities as Canis lupus, mammals or Coleoptera, which count dozens, thousands or millions of quite different populations, each of them with its own variability and subject to entirely different selective pressures.
I shall therefore try to prove the following proposition:
Proposition 1: Species do not exist nor have ever existed in Nature. They are concepts closely akin to Platonic ideas: taxa (not necessarily the same as the formal taxa of the International Codes) and among them “species” are common names, that is, they are “universals” in the sense of the classical debate between “realists” and “nominalists”. As a corollary, I shall argue that although, for practical reasons, debates on phylogenies must use all the acknowledged taxonomic categories, including species, the very concept of “species” is irrelevant to evolutionary theory.
As I have already stressed, this is the old, debated, “species problem”: do “species” exist in Nature or are they concepts, an intellectual tool that we use to communicate among ourselves? This being the core of the problem, from its solution stem a number of subordinate questions, and we shall mention some of them and their impact.
As a premise to the proper consideration of the “species problem” in evolutionary biology, we must consider a much more general problem.
The subject of the next few paragraphs should be familiar to undergraduates from their high-school curriculum in philosophy. However, this topic is commonly both taught in a cursory way and obscured by a language plagued by technical terms of Latin origin used by Medieval scholars which seldom plainly convey their meaning. Finally, in colleges, the eclectic philosophies of the Roman imperial age are but seldom given adequate attention, while the work of the comparatively late Hellenistic thinkers was extremely relevant for the later development of scientific thought.
As shown in the following paragraphs, these considerations are plainly relevant not only to establish a basic distinction between what evolves in Nature and formal taxonomy, but it is possible to prove that what in both common and technical parlance is called “species” is a purely intellectual tool.
Though further on I shall prove that none of the “realistic” concepts of “species” proposed is logically and empirically tenable, I shall first consider the problem from the most general standpoint.
The problem of the logical connection between the observations possible on the individual objects, whatever their nature, and the comprehensive categories by which they are named came naturally to the Greeks just because of their language. Indeed also in ordinary parlance ancient Greeks made a distinction: names and adjectives were flexed to tell their function in the sentence (subject, object, etc.) but also the article was used. In fact, if the name was prefixed by the article, this meant that that name was being used with reference to one or more specific individuals/specimens of the class meant by the name, so “the dog” would mean one specific dog: Xenophon’s dog, a certain black dog, and so on, while if the name was used without prefixing it by the article that meant that a generality was meant, as, for example, “dog is the friend of man” does not refer to any specific dog or group of dogs, but to all past, present and future dogs.
Aristotle, besides having been the first to investigate systematically the fundamentals of logics, also in his De anima, De memoria et reminiscentia and in the Parva naturalia pioneered the study of psychology and if we consider together the Organon and his thought on psychology, it appears that he was well aware of a problem or rather of a complex of problems which may be summarized by the question: “What is science?”.
Aristotle thinks that logics, that is, the core of the Organon, is just a tool necessary to attain knowledge, but, in essence, it deals only with the verification of such sentences that may be said on given subjects.
Aristotle agrees with Socrates and Plato that science deals only with general propositions, those that we call “principles, scientific hypotheses or laws”, while any statement which may be made concerning only given individuals or single events, though it may well be either true or false, cannot be the object of science.
Aristotle also assumes as basic the “principle of no self-contradiction”. The Stagirite, indeed, maintains that only such a proposition or group of propositions which do not imply an internal contradiction or that are not mutually contradictory can be the object of science (that is, that are not absurdae, to use the term employed by medieval scholars, as they developed the argument into the “theorem of Pseudo-Scotus”: ex absurdis sequitur quodlibet = “from contradictory premises one can derive any consequence he likes”, or “no certain conclusion is possible”).
The kind of logical analysis necessary to verify the validity of a general statement or concept, or, again in medieval philosophy, a “universal” (such as, for instance, the sentence: “celestial bodies move along curved trajectories”), as seen by Aristotle, finally depends on suitable “definitions” of the terms of the proposition. The definition, according to Aristotle, must be by gender and by specific difference, the “generic quality” being such as to be considerably inclusive (to continue with the previous example, one such as “the Moon is a celestial body”, which puts the Moon into the very comprehensive class of celestial bodies), while the “specific difference” is a quality or attribute (and it may well be a whole group of qualities) that is appropriate and unique to the object of our definition and such as to clearly separate what is included and what is excluded from the object of the definition. To continue with our Moon example, I can say: “The Moon is a celestial body (gender) satellite of the Earth (species)” as the predicate “satellite of the Earth” is both inclusive, as the Moon qualifies for the inclusion within a class “satellites”, and at the same time exclusive, as it rules out the possibility that any celestial body which is not a satellite of the Earth may be the Moon. It is worth remembering that this principle is used by Linnaeus as one of the founding principles of his systematics.
This basic framework, clearly stated by Aristotle, had a fundamental impact on the later development of sciences and is clearly the basis of the “International Codes of Nomenclature” which rule bacterial, botanical and zoological procedures.
Clearly, if we do not agree with these basic tenets, Aristotelian logics will not work. Should we, for instance, agree either with Protagoras that reality is just a contingency corresponding with the sensations, or accept the transient instability advocated by Heraclitus, then science as conceived by Aristotle is impossible, as the same thing may either be or not be according to the observer or the different times of observation (and this may be relevant for the study of evolutionary phenomena).
However, Aristotle was not only a logician, but also an excellent naturalist and a keen observer, and in his “Metaphysica” and “Topici”, he considered the problem of whether we may have “science” of variable matters and especially of those which change with time. He also considered how we can judge hypothetic propositions: suppose that I say “tomorrow a battle will be fought”; may such a proposition be judged true or false?
Moreover, Aristotle maintains that the qualities or “attributes” (those that finally identify the species) do not really exist as entities independent from the subject (the substance, literally what stays under). He notes, “health exists when the man is healthy, the figure of the copper sphere exists just simultaneously with the copper sphere” and a little further on he makes a statement of basic relevance, against Plato who maintained that individual beings and objects were the approximate materialisation of eternal and perfect models (the “ideas”, the archetypes); he writes: “There is no reason to suppose the existence of the eidos (a term that may be translated either as “idea” or “species”), man is born from man”. and, again,
Each mode has no existence of its own, none of them may be detached from the ousia, they appear to exist only because there appears to be a certain being…. And this is the substance, that is that special being to which the various qualities are attached. Good, sitting, are meaningless without this substance, which is that particular being which is known through his various attributes. It is, therefore, evident that the existence of attributes depends on the very existence of substance...
As the Greeks often do, Aristotle, in order to sort this problem, suggests what is basically a verbal solution. He says that we must distinguish between dynamis (literally “potentiality”) and “act”. Potentiality is that which may be, but presently is not, such as for instance the colour “red” in its more general meaning, or, in a more complex example, the eidos horse; the “act” is what really exists: the individual red object or the horse that we may actually see.
Thus, as conceived by Aristotle, even the undefined, the substance, exists by itself only potentially and becomes real only when it joins with the eidos.
Going one step back, we must stress that as eidos may be translated as “form”, “idea” or “species” as it basically means the ensemble of qualities that characterize a given group of beings or objects, and in the relation “genus-species”, the “genus” is a special kind of substance, as the real objects, such as living beings, belong to a sort of chain such as: animal-mammal → mammal-carnivore → carnivore-canid → canid-wolf → wolf–a given wolf in a pack; at each one of these levels the inclusive member of the two terms is the genus while the included is the species. Thus, conceptually, the “substance” becomes more and more real the more it is precisely limited.
As conceived by Aristotle in the framework of his idea of an eternal and unchanging world, this individualisation is not an historical process through a lineage of individuals, but only by the development of the individual itself.
Aristotle had then to face a problem: if he was to agree that universals exist “in act”, that is, in reality, only in the individual objects or beings, he should have admitted to Protagoras’ individualism and concluded that a science of universals is impossible and that science is not possible at all.
He met the challenge with a brilliant argument: as it is possible to conceive the materialisation of beings by a series of progressive steps, then a science of the potential universals is possible as they may be conceived as the factors of the steps in the reification of the actual realities.
The theory of science of Aristotle continued to be basic in all branches of the scientific debate until physicists in the 19th century begun to change the concept of substance, but it is still relevant in our field of investigations.
Moreover, Aristotle had honestly pointed to a series of kinds of propositions, which are not liable to verification by the implementation of his logics (his famous 127 “non-syllogistic propositions”).
These were soon tackled by the scholars of the “second Stoa”.
We do not need to discuss the details of the later Stoics’ logics. For our purposes, it is sufficient to remember that it had its core in the need to maintain that science is about general propositions while the only real things are individual objects. One pupil of Socrates, Antisthenes, once retorted to Plato: “Plato, I see the horses but not Horsyness!”.
The Stoics were, to use the medieval word, “terminists”; they considered the ideas underlying words such as “dog” or “goodliness” as having a special kind of reality, entirely different from that of the actual beings for which they functioned as “symbols”. Thus, it was possible to have general rules which could be valid and to some extent independent of the passing contingency of individuals.
Anyway, the Stoics completely agreed with Aristotle in holding that all knowledge is finally derived from empirical experience.
This was negated by Arcesilaus and his followers, the sceptics. They objected that an empirical observation is not necessarily true (for instance an optic illusion is a sensation, yet it provides false information); thus, they correctly argued, acceptance of a sensation as evidence depends on its rational assessment and this presumes that the observer already has some theoretical standards of evaluation, that is “universals”, which, however, are dependent on the true information obtained by previous experiences and so on ad infinitum.
Let us now see how the problem of the “universals” impinges on the evolutionary debate, and that will take us again back into Antiquity and the Middle Age, but also into the core of some essential parts of the debates that followed the publication of the Origin of species. Let us assume that science does not deal with single occurrences or individuals, but aims to recognize more or less comprehensive “universals”; for our purposes, the “species” and, eventually, the different “species” that we recognize in Nature. Should we go back to Plato’s belief of eternal archetypes, the eideia, commonly translated in textbooks with “idea”, and assume that the corresponding objects that we observe are the more or less perfect materialisations of the eideia implemented by the Demiurgus?
Plato’s influence was powerful throughout antiquity and we shall here skip all its manifold mystic and religious influxes; yet, as we shall see, it is still, unfortunately, with us.
During late antiquity the platonic views were revived to some extent by Neoplatonic thinkers. However, Neoplatonism (a religion rather than a philosophy) had, just at the close of Antiquity, three very different and yet very important thinkers. Simplicius (c. 529), Proclus (410–485) and Boethius (480–c.526) and, moreover, Platonism had a notable impact on the thought of Saint Augustine. The ideas of these thinkers and the immense authority of the saint made Platonic influences ever present and pervasive for centuries in the development of Christian thought.
The main purpose of the three philosophers just mentioned was to develop a philosophical synthesis based on a solid historical groundwork. The formal framework of their writings is plainly Aristotelian but, just on the matter of universals, it was strongly tinged by platonic realism.
Simplicius and Proclus belong to the eclectic paganism of the late Academy, while Boethius was a Christian, and, just because he wrote Latin instead of Greek and was a Christian, his influence was especially important for the development of Scholasticism. However, he was a Christian so tinged by Neoplatonism that some scholars have wondered whether he was really a Christian.
As far as the problem of our concern, that of the significance of universals, the fundamental answer by Proclus (in the context of an important discussion of geometrical problems) is the classic thomistic stance, which Saint Thomas and Ockam received as mediated by the Persian Avicenna and Saint Albert the Great; this assumes that universals exist ante rem, before the actual thing, in the mind of God; in re, as a materialisation of the “eidos”, the form, into the material things, and finally they exist post rem, after the real thing, in the mind which thinks about the matter.
To make it simple: a universal exists ante rem, before any corresponding material thing, for instance the idea of dog, in the mind of God even before the world’s creation; it exists in re as it manifests itself in the actual dogs that we see; it exists post rem, after the material thing, as we can compare all our experiences of dogs and thus we can visualize in our minds the Idea of dog. Thus universals do exist but only as “intellectual tools”.
Ockam’s ideas were largely followed in medieval universities, though the crucial text debated was the comment by Boethius on Porphiry’s Isagoge on Aristotle’s “Categories”, where the Roman philosopher clearly discusses whether genera and species are pure mental images or have some sort of empirical reality.
We shall not deal here with the religious implications of the debate, which, however, impinged on the attitudes of the different thinkers.
The different conclusions range from the strong “realism” of thinkers closer to neoplatonism such as the members of the “Florentine Academy” of the late 15th century, but is still present in the German Naturphilosophen of the 19th, to strong “nominalism” such as that of Roscellinus, who held that universals were flatus vocis, mere words, or more moderate positions, closer to Aristotle, such as that of Abelardus who held that as ideae ante rem universals did not really exist, but that they exist in actu only in rebus, that is, only as the material things observed, but may be used by the philosopher as abstractions positae in nudis intellectibus, that is, they exist, but only in the scholar’s mind.
There are, naturally, intermediate positions, rather close to Aristotle, that hold that any attribute or cluster of attributes is both universal and particular; it is a universal as it is common to an indeterminate “many” (such as having all the qualities of a dog) and particular as they are unique when they appear in a given individual.
Should today’s naturalist think a bit on these ancient debates which, also because of their rather obnoxious language, may look irrelevant, he would, indeed, find that they deal with the core of current debates on the general principles and methods of systematics.
Nor are they irrelevant to the broader field of scientific research. Indeed, we pride ourselves that ours is the “age of experimental sciences”, yet any experiment is devised to answer a question and this is necessarily a hypothesis that is plainly a typical “hypothetic universal”; precisely, as the Greeks discovered, we do not make scientific hypotheses concerning single events occurring to single individuals!
Just to take an example: until the 17th century, no one doubted that spontaneous generation could occur. Even a great naturalist like Aristotle, though thinking it rare and concerning only a few groups of organisms, had no doubts about it and no one thought of any experiment to verify whether it was really possible that a living being could be born from non-living matter. When some new evidence allowed Harvey to advance the hypothesis that every organism must be born from another living being, within but a few years, first Redi and then, as the problem appeared more complex, more and more researchers devised the necessary experiments to verify how reproduction really occurred in various organisms.
Science is of “universals” and Plato, Aristotle, Duns Scotus, Saint Thomas and the others were right and their controversies can still teach much to the modern scholar.
We must still consider an aspect of the general problem that is especially relevant to us. An important development of medieval nominalism is named “terminism” (and we have already mentioned it). This was especially developed by Ockam. To Ockam, the only reality is the individual, and the “universal” is a creation of our mind (for instance, single animals or a given population of animals are real, but the species is our abstract creation: it is a symbol). The signum is an abstract idea that allows us to recognize individual beings and substitutes for them as a symbol; it exists naturaliter and has an objective value. Instead, the intentio secunda is the completely abstract universal, that is, a general quality, such as, for instance, the red, or a cluster of qualities, such as, for instance, the general idea of fish which does not directly relate to individual things; this does not exist naturaliter; it is not a natural phenomenon but exists secundum institutionem voluntariam, that is, it is entirely our creation. However, a science is nevertheless possible also of such concepts that are created by intentio secunda (and such is the concept of species) as while only individuals are materially extant, as the abstraction depends on a set of qualities of the individuals, it is a universal which derives from the information provided by the real single, particular beings; thus, as a universal it may be the object of science as it is really related to realities: the “universal”, though it is a creation of mind derived from the experience of the real things, is nevertheless a necessary tool for science.
We must, I think, agree that, like all common names, any collective name we use to identify a given collection of objects, both living and not, be they natural phenomena like stars or waves or man-made such as chairs or poems, are “universals” – abstract ideas in the literal meaning of the word, as, indeed, abstract means “taken out”, that is, the collection of attributes that we consider common to all the individual objects that we have decided to call by that name, independently of their numbers, locations or ages.
And so, we may now go back to the significance of the debate on the universals for the “species problem” in evolutionary biology.
It should be plain that in the following discussion we are not concerned with the “species” of formal taxonomy: these are clearly conventional descriptions, lists of features that may be found to occur and characterise some more or less common specimens, and that are presumed to apply to all members of the population from which the “type-specimen” has been collected, while the “type specimen(s)”, as the mandatory name bearer of the taxon, is/are there for reference in case of doubt. Indeed, the whole framework of the “Codes of Nomenclature” does not consider evolution at all. Names are bestowed on organisms such as they may be described on more or less complete material at a given moment. If a comparison is allowed: a man is born, grows up, gets old and through all these changes yet remains Mr. John Smith; his photograph, say at the age of 30, is very much the equivalent of the formal description of a taxonomic species.
To conclude, the “species” of medieval logicians, including our idea of any community of living beings, is logically a universal, an intellectual tool which simply conveys to the mind a sort of model that to us approximately describes such a more or less comprehensive community of beings, thus allowing for exchanging information between people. Obviously, we can restrict at will the indeterminate collection of beings, present, past and future, that we want to subsume under our “universal”, abstract idea, until we reach the point when our “universal” becomes inclusive of only one “population” of objects that we can see, number, etc., and then it ceases to be a universal and becomes a real thing, a particular. So, the “universals” “dog” or “chair” are concepts that may be restricted to, say, “sledge dogs” or “Chippendale chairs” and finally to a given “pack of dogs” or to the “chair” on which Churchill was sitting when thinking of his “tears and blood” speech, and it is precisely at that point that the words dog and chair change their significance and come to mean real things.
Let us then consider the following proposition:
Proposition 2: It is not possible to conceive of a general concept of “species” such that it may apply to all the populations or groups of populations occurring in Nature; the concept of “species” is a merely conventional device, arbitrarily defined by one or more scientists to fit the purposes of clear argument.
Though Darwin (Citation1859) titled his masterpiece On the origin of species by natural selection, he was a nominalist and he was quite clear on it. Yet since the publication of Darwin’s book, we find in the following debates on all aspects of evolutionary theory and of its practical implementation a conscious or unconscious revival of most of the medieval debate between “realists” and “nominalists”.
Indeed, a number of scholars have made all sorts of attempts to conceive the “species” in a precisely “realistic” view, as a real entity in the natural world; and this has had an impact on many ways of dealing with evolutionary problems. In fact, reading most of the dozens of definitions proposed for the “species concept”, it is clear that to their authors the “species” was a real entity occurring in nature at some more or less limited time and that, eventually, could change into something different which could be recognized as a different “species”. Usually such “entities” which did not fit into the definition were not “species”, though they could be fitted into a taxonomic framework.
As I shall presently show, all these definitions may be faulted both in logic and by factual evolutionary evidence. Basically, when one considers the arguments advanced by the various authors to support their views, one sees that actually they are thinking of the “species” as a meta-population, grouping a smaller or larger number of populations occurring in nature at given times and locations, and which on presumably good evidence may be considered to be rather closely mutually related.
As the single populations are real enough, as we actually see them, contact them in a number of ways, count their members and possibly even feed on them, and the reality of their interrelations may often be tested, the reality of these meta-populations seems obvious, and it is precisely here that the “realists” are at fault, precisely as was argued by the ancient logicians who studied the problem of the “universals”.
Just to clarify the issue, let us begin by taking some simple examples. Let us examine the following sentences:
“This is John’s dog”; “This is a black dog”; “Neapolitan mastiffs are black”; “This is a dog”; “The dog is a carnivore”.
Take the first sentence: “This is John’s dog”. Quite apart from all that we presume to know about John, we assume first that we have recognized in that particular individual “dog” a set of features that we consider as typical of a group of organisms (we do not know how many or where they are) that correspond to our “idea” of dog, and this “idea” is undoubtedly a “universal” as, at least as far as the speaker is concerned, it includes all the past, present and future dogs (and we shall see further on whether it is an “objective” idea as held by Plato, or a subjective one). Moreover, the sentence states that we have identified in that particular dog some unique features that identify him as John’s property.
The second sentence: “This is a black dog”, assumes three “universals”: like the first sentence, it assumes the idea of “dog”, but it also assumes the ideas of “black” and “not black”; these two are “attributes”, that is, “qualities”, but they are “universals” all the same as they apply to sets of objects of unknown numbers, but clearly not unlimited numbers as they are mutually exclusive (this is true in our example: there are instances, such as the series of the even and odd numbers, when two mutually exclusive classes are both infinite).
The third sentence, “Neapolitan mastiffs are black”, is a definition that assumes the existence of a “universal” “Neapolitan mastiffs” and of an attribute, which is also a universal, which works both as “inclusive” – any black dog might be a Neapolitan mastiff (whether it actually is will depends on other attributes) – and exclusive – if it is not black it cannot be a Neapolitan mastiff.
The fourth sentence, “this is a dog”, is the crucial one. Let us suppose that I am strolling around with an Eskimo who had never previously left his country, has never seen a movie and so on. He is quite familiar both with sledge-dogs and with wolves, but has never seen a Chihuahua and, just by chance, we meet with someone walking on a leash one of these miniature dogs. If I say: “Did you see that dog!” my Eskimo friend will almost certainly reply, “Which dog?” Indeed, he “knows” by his personal (and cultural) experience that a dog is a dog and a wolf is a wolf, but that tiny being is quite different from his ideas both of dog and of wolf; hence it cannot be a dog!
And now let us consider the sentence “The dog is a carnivore”. This sentence implies a number of plainly conventional “universals”. We have seen that the first term in the sentence, “dog”, does not refer to any real dog but to an abstract generality, the “universal dog”, the platonic idea of “dogginess”; but by the second term, “carnivore”, we are meaning that it is a member of a collection of “ideas” that we think fits together by a group of chosen features, into a comprehensive category that we have agreed to call carnivores (though some of them may only marginally feed on meat, while others, purely or mainly flesh-eating mammals, as they do not share of all the features that we have incorporated in our definition of Carnivora, we place in other classes). Quite clearly, if we agree that the “dog” of our sentence is the “universal dog”, it is impossible that a comprehensive group, made of universals such as the “universal wolf”, the “universal fox” and others, and excluding other “universals” such as the “universal rat”, the “universal starfish” and so on, can be anything different from a “universal” itself, that is, just an idea.
Indeed practically no one has ever doubted that all taxonomic categories above the “species” level are purely abstract concepts, while advocates of the “realistic” concept of species think that species and the lower categories of the system are real natural objects. This has some curious results; to consider just some examples, Ghiselin (Citation1997) and other advocates of the theory that “species” are collective individuals have argued that as “collective individuals”, a topic that I shall consider further on, species may and do evolve while, should we consider species as “classes” as admittedly are families, orders, etc., as it is assumed that a “class” cannot evolve, we could not conceive of the “evolution of species”. This has been maintained by several scholars with apparently sound logical arguments, though, if agreed, it would plainly imply that sentences such as “Birds have evolved from archosaurian reptiles” or phylogenetic diagrams showing given higher taxa branching off other taxa would not be permissible. Indeed, if you assume that supra-specific categories are classes and that classes cannot evolve, this conclusion is inescapable, though plainly wrong. However, the real meaning of such sentences or diagrams is: “during the evolutionary history of some populations that, because of their affinities we conventionally group as, for instance, reptiles, one or perhaps more among these populations, by the acquisition of some characters and, possibly, by the loss of others, became so different from their parent stock that we consider them as the originators of the new class”, and this is a perfectly correct and logical statement describing a real event. Obviously a strict cladist will in any case reject this last statement as it implies the acceptance of paraphyletic taxa. As I disagree, this problem will be discussed in the next paper of this set.
Students of semiotics have produced and still produce an enormous literature, which is well rooted in classical and medieval debates (for an example which also provides a rich bibliography, see Eco Citation1997). However, as far as the evolutionary biologist is concerned, the essential conclusion (as we have previously argued) is that there is no possible doubt that all the words, when denoting a comprehensive set of phenomena-objects with undefined numbers, ranges or temporal limits, are abstractions, that is, “universals” that supposedly mirror in our minds the empirical experiences that we may obtain from a limited sample of present or past natural populations, and which, by the way, will in any case logically qualify as a “fuzzy set”.
We must, indeed, be always aware that this is an absolutely general problem: whether we speak of dogs, earthworms, mountains or chairs, when we speak of them in a general way, as the ensemble of all the chairs, past, present and future, or, in any case, if I speak of chairs meaning not a specific group of chairs, but something that would include any kind of chairs in indefinite numbers, locations, etc., we are speaking of a “universal”, while if we specify somehow one particular chair or a precisely limited group of chairs (such as, for instance “the chairs in my dining-room”), then we are talking of a “particular” which is (or was or shall be) quite real (we may indeed speak of the chair of Shakespeare or of one we have just ordered from the carpenter). However, our sentence will always imply as a reference a universal “chair”. As we shall see, this is one of the main criticisms that apply to the thesis that biological species are “individuals”, a thesis accepted by many biologists and that, as I said, I shall discuss further on.
More on the “species” in biology
Let us now consider the main different “concepts of species” that have been advocated. Obviously, for each of them, some or many slight varieties have been advocated by this or that author.
Let us begin with antiquity: to both Aristotle and Theophrastus, the “species” was a rather indefinite concept and a definitely variable thing. For them, obviously, it was the overall appearance that was significant and this changes with age, sex and, as stressed by Theophrastus, as far as plants were concerned, with environmental conditions. Theophrastus, indeed, believed that a plant grown on soil and in a climate different from those of its parent could become a different species. Aristotle, on the other hand, thought that new species could arise by hybridisation of pre-existing ones (both of them by modern standards were simultaneously right and wrong and wisely refrained from trying a definition of “species” as the Greek eidos can, indeed, be freely translated as “idea” and “species”).
These views were still common in the 18th century and, indeed, we now know several instances among plants and quite a few among animals of perfectly stabilized hybrids (not to mention the all too common genetic exchanges among bacteria).
It is plain that many scholars do not make a distinction between the supposed “real biological species” that is or was living at some given time in some given place and the taxonomic species which is the ensemble of specimens, both living and dead, that fulfil the requirements set by the “diagnosis” of the taxon. The distinction is, indeed, somewhat a subtle one: given that the essence of the definition of the so-called “biological concept of species” is that the “species” is a “gene pool”, while the “species” of taxonomy is defined by its “diagnosis” and clearly does not make any reference to reproduction, as, indeed, taxonomy commonly awards the same status to “species” which reproduce by parthenogenesis, agamic reproduction, etc., so that there is no doubt that the two are conceptually different. However, a number of biologists imply in their work that the “taxonomic species” they mention are also “biological species” and that, for instance, intersterile populations characterized only by different complements of Robertsonian fusions or by some deletions should be recognized as separate “species”.
With the development of systematics and the need to classify a rapidly increasing number of organisms by definitions conforming with the Aristotelian principles, there naturally stemmed the so-called “typological concept of species” based on the “description”, that is, the listing of such characters which allow for the identification of any given specimen as pertaining to a given species (or, if it does not match any previous description, to describe a new species). By such an approach it is obvious that the “species”, like the other recognized taxonomic categories, are classes and as such they are, by any logic, abstract “universal” ideas.
The next development, which was the necessary consequence of the need to agree on uniform rules for the proposal and employment of the names proposed or that might be proposed in the future, was the establishment of the “International Codes of Nomenclature” and, later, of the types system, so that every name proposed for a given group of organisms (but often single specimens) is permanently linked with a given description and to one or more “type specimens” functioning as a reference, and this makes for difficulties as in many instances “sibling species” can be identified only by characters (chromosomal, molecular etc.) which are impossible to verify on the type specimens.
It is quite evident that this concept of species is merely a conventional system, liable to all sorts of mistakes, and, indeed, there are hundreds or perhaps thousands of instances when different names have been bestowed on the male and females, or to different developmental stages or even simply somewhat different specimens of the same “species”. The purposes of the “codes” and of their underlying concepts are purely practical: an attempt to avoid an inordinate proliferation of names, and some criteria to decide when two organisms should be called by the same name.
In practice and in spite of criticisms, the “typological” or “Linnean” concept of species is the most commonly used as it is practically the only one which can be implemented for all museum specimens that, in fact, make up the vast majority of known “species”.
Obviously this is an atemporal system and a subjective one in spite of all attempts to make it “objective”. It is atemporal as the name and concept that it is supposed to convey clearly refer to a supposed “universal” that the proposing scholar extrapolates from a limited and possibly fragmentary sample having a given age and collected at some definite locality and, for just the same reasons, it is subjective as the whole of systematic literature proves, with all the synonymies and the splits and changes of rank that have been advocated for thousands of nominal species. However, it is perfectly suitable for discussing evolutionary problems, just as the names and biographies of the individual members of a family can be used to establish a genealogy or to write a family history.
When we analyze any discussion of the evidence relevant for understanding evolution, be it of a single or a set of morphological, physiological or behavioural features concerning a certain group of organisms, we immediately perceive that the evidence itself is something which we find occurring in a certain number of individuals which we assume to be representative of some populations occurring in Nature. Though we shall often speak or write of the features under discussion as concerning this and that “species”, this is because we have the “feeling” that what we see in the limited sample of specimens from the population(s) that we are actually studying is something which is/was actually occurring in all the specimens that we group under the definitions of the taxon studied, while, given the mechanisms generally assumed to be at work during any evolutionary process, at the same time we must presume that such changes as we may observe were or are actually occurring in nature in only a fraction of the whole ensemble – a fraction that, at the beginning of the process, concerned a small fraction of the parent population and that, natural selection helping, spread or shall spread to the whole population so that, once that the process will be completed, the modified population will be a “new” taxon.
Anyway, biologists were still challenged by the problem of deciding what are the entities that collectively change during evolution and possibly (I would rather say: just as a habit), calling them “species”.
Several proposals have been advanced, always with the assumption that “species” are “real” entities. But, as they may be conveniently grouped in just a few main groups, I shall discuss here only the “core” of each one of these.
The so-called “biological concept of species”, which enjoys an ancient tradition, has for long been a popular one. Basically, it holds that any set of organisms that may be considered as a “gene pool” should be considered as a good species; a gene pool being a population or a meta-population formed entirely by individuals that may, at least potentially, breed with any member of the opposite sex (obviously, as there are a number of hermaphrodite taxa, it is more proper to say “cross-fertilize”) of the same population, producing potentially fertile offspring, so that every gene of each member of the population may, at least potentially, spread through the successive generations to any member of the later generations.
In fact it is quite common (but there are a number of exceptions) that in a population with bi-parental reproduction and meiosis each gene may pass, to some extent randomly (but seldom completely so), to the next generation and so on, and thus it may increase or decrease in frequency within the population or its sub-sections as a result of many factors, among which selection for maximum fitness has a predominant, though not exclusive, impact. Indeed, we should not forget, though this makes things more complex, that any given character or group of characters is probably affected either at the same time or in succession by many different selective pressures. For instance, it might be supposed that a given feature, when it first appeared, provided a slight sexual advantage, but that, having reached some degree of development, it became a significant asset in predator avoidance and therefore that its later development was so closely dependent on the secondarily acquired function as to entirely mask to our eyes the real original selective advantage which determined the early phases of its development.
Bock (Citation2004) has made an interesting attempt to refine the “biological concept” while providing a realistic concept of “species”. As this is quite recent and has the unquestionable merit of providing a very clear distinction between what he calls “the species concept”, the “category species” and the “taxonomic species”, and that I still disagree with his evaluation of the “species concept” which in his quoted paper appears to be identical with the “population concept”, in order not to duplicate the same arguments, I shall discuss Bock’s ideas at the close of the following criticisms of the “biological species concept”.
As a preliminary, I shall list a number of problems that challenge all the definitions thus far provided for a “biological concept” of species.
Mayr and many others assume that the concept of “species” applies only to bi-sexually reproducing organisms and that species must be characterized by genetic isolation, reproductive isolation and ecological isolation. Quite apart from the fact that such a concept, being plainly an atemporal one, cannot apply in evolutionary studies (and to this I shall come back further on), it is obvious that it does not apply to any purely parthenogenetic or self-fertilizing population, but we know several examples (for instance among the Phasmida and some American and Caucasian lizards) of “species” that have some populations that are entirely parthenogenetic, while others have a bi-parental reproduction. A telling case of transition is that of the moth Dahlica triquetrella (=Solenobia triquetrella) in the Alps: during the last glaciation, apparently some small populations survived in nanutaks and, when glaciers retreated, the now-free areas have been colonised by a parthenogenetic population which is now moving uphill following the retreat of the glaciers, while lower down the “normal” biparental population is spreading, so that it gradually it substitutes for the parthenogenetic one.
Moreover, there are a number of insects with a larval stage of some years and an adult life of but a few days (such as mayflies, cicadas, etc.); in such animals adults are, indeed, present every year, but there cannot be any gene flow between one year’s generation and that of the previous and next one, so that, if we assume the “gene pool” requirement as essential, by definition, as the nominal “species” is a complex of entirely separate lineages, each one of them should be considered as a separate “biological species”. Such situations may be further complicated by geographical and ecological factors; for instance, in the case of Melolontha melolontha, while all the southern populations have a period of 3 years and are thus entirely isolated from one another, some northern populations have a period of 4 years, so that every 12 years, two lineages reach adulthood simultaneously and thus may breed together.
Again: the gene flow follows quite different patterns in animals, such as many insects which are fertilized only once in their life, so that the gene flow is strictly unidirectional besides the already mentioned non-overlap of generations, and long-living organisms where fertilisation may follow different criss-cross patterns.
A very intriguing condition obtains in some insects, especially among some groups of wasps (particularly typical are species of the genus Trichogramma), where in nature we find only self-reproducing females, but it has been found that antibiotic treatment or simple heating is followed by the appearance of males, which, however, in many instances are sterile! Equally embarrassing is what happens in several plants, which depend on insect pollination and which in the wild never cross as they strictly depend on different pollinators, but easily hybridise and produce fertile hybrids under experimental conditions.
Obviously there are a number of “species” which actually have all the proprieties assumed by the “biological concept”, but, quite apart from the many instances of well-known situations where bi-sexual reproduction, parthenogenesis, and asexual reproduction alternate, or occur only in given sections of the population, etc., there are an enormous number, even entire sub-phyla, which unquestionably consist of “populations” that have evolved and evolve just like any bi-parental population. Just to take the first example coming to mind of a well-known historical change in a species: the homopteran grape-wine parasite Viteus vitifoliae (=Phylloxera vastatrix), that in its original home in America has a complex cycle, alternating bi-sexual and parthenogenetic reproductions, phases parasitic on roots and phases parasitic on leaves, and in comparatively recent times has evolved into the European race which is strictly parthenogenetic and lives only on roots! Much the same holds for the other supposed essential characters of “species”. And we can go on with examples damning the “biological concept of species”. In fact, while in plants evolution by hybridisation is quite widespread, at least in some groups, also among animals we know a fair number of stabilized hybrids in nature; some scholars even think that up to 10% of the living species of amphibians may be of hybrid origin. Anyway, hybrids with limited fertility or where only one sex of the hybrids is fertile occur, for instance, in insects (as, to cite one, in the moth genus Lymantria); and what about those insects, for instance, where there are entirely parthenogenetic species, which, however, will not reproduce unless they copulate with the males of another species? Usually, the parthenogenetic “species” is a tetraploid and the required male is that of the diploid “species” from which the parthenogenetic tetraploid derived, and in such instances the parthenogenetic egg just needs to be punctured and thus activated by the sperm.
Again: the “biological concept” does not fit properly for the many organisms, such as not a few plants or animals, where self-fertilisation is an option (such as the members of the family Vaginulidae among Molluscs, some tape-worms, etc.). And what about such organisms as the Rotifera where only members of one small family (Seisonidae) are anphigonic and the Monogononta periodically produce males, but otherwise a large component of the phylum (Bdelloidea) is entirely composed of parthenogenetic species? Finally, it is not rare that potentially inter-fertile organisms are separated by geographical barriers which prevent hybridisation between individuals of different populations, and should we consider them as distinct species or, should they breed true under captive conditions (and there are several examples, for instance with tropical aquarium fishes), should they rank as conspecific?
As for ecological isolation, this is commonly flexible: precise ecological requirements that reduce the interactions between populations (niche separation) are usual, but these are often quite different for the same “species” according the ecosystem used; whether there is competition or not, etc., often seasonally changing or concerning only certain periods of the life cycle of the organisms, is variously modulated in many symbioses, so that it is far from a clear-cut principle.
Moreover, the “biological concept” cannot apply to the many thousands of “species” known only by some museum specimen, and to all the fossils, all instances where, in fact, we rely not only for distinguishing or merging formal taxa on the subjective judgement of individual scholars but, as the core itself of our study is the identification of the links of descent between the specimens available, that is their “phyletic lineage”, we cannot make any use of the “biological concept of species”.
A number of authors have variously tried to refine the “biological concept of species” and, while maintaining that the “concept of species” applies only to bi-sexually reproducing organisms, have advocated, dealing with non-bi-sexually reproducing organisms, the employment of other terms alternative to “species”. All the examples quoted in the previous paragraphs show that this is clearly an entirely artificial attitude as it implies that among some quite closely related groups of organisms some are “species” and some are something else. Indeed, it would be easy to draw a long list of organisms where, in the wild, some populations reproduce by cross-fertilization between hermaphroditic individuals or by a male-female parentage, while others reproduce either by parthenogenesis or by purely vegetative reproduction, and not a few of the populations which do not reproduce by cross-fertilization are clearly derived from a still-living cross-fertilizing stock and fit into the ecosystem just as their cross-fertilizing relatives do.
As I have already mentioned, a recent attempt to refine the “species concept” on the basic tenets of the “biological concept” has been made by Bock (Citation2004), who maintains that what evolves is the “species”, but that this is something distinct from the phyletic lineage, as this is the result of the transformation of the species through time, while the species should be conceived as the instant, temporary ensemble of the individuals sharing the aforesaid basic qualities to rank as “species”. So the common, traceable lineage through time of any population or meta-population is the phyletic lineage, while the “species” should be conceived just as that section of the lineage that we may observe and that is/was living at a given moment in the history of the lineage. Indeed, if this concept of species is agreed, Bock’s “species” is real as, at least theoretically, it is made up of a number of individuals that could be seen, counted, etc. However, Bock’s idea of species is precisely the same as that of “population” in its traditional meaning, and anyway, it is not clear to me how, in Bock’s concept of the phyletic lineage, could be recognized different phyletic lineages as presumably all of them, working backwards in time, will merge into the presumed common ancestor of all living beings.
If we agree that Bock’s concept of “species” means what we usually call “population”, there are still some problems with one aspect of Bock’s idea: either we accept Bock’s implicit identity of “species” and “population” (but then the “species concept” is a superfluous duplicate), or the various “biological concepts of species” are irrelevant in evolutionary studies as these necessarily do not consider Bock’s “species”, but are concerned with the study of “lineages” as, just as Bock has stressed, in his view, the “species” is a closed system with regard to contemporary similar systems, but necessarily merging with its immediate predecessor.
Actually, Bock himself writes:
Cross-sections through a phyletic lineage represent species at the different times of these cross-sections. But regardless of how similar or dissimilar the organisms are at these different cross-sections, one cannot say whether they represent the same or different species if one assumes that evolutionary change is gradual. It is a non-question to ask whether these different time slices of a phyletic lineage represent the same species or different species: in theoretical considerations, limits cannot be placed on a species taxon and its successor. Such boundaries would imply an evolutionary change “between species taxa” different from evolutionary change “within the species taxon,” a distinction which has never been demonstrated. Therefore, if species cannot be delimited along a phyletic lineage, it is not possible to speak of the age of a species. All existing species are of equal age, or in other terms, all species are ageless. Species boundaries are real only in horizontal comparisons, which are between different lineages, and do not exist in vertical comparisons (within a single phyletic lineage). Hence, one cannot make a distinction between evolution within a species and evolution beyond the species boundary (transpecific evolution). Evolution along a phyletic lineage never results in a new species and hence never passes a species limit regardless of the amount of phyletic evolutionary change that has taken place. (Bock Citation2004)
Quite apart from this, the fact is that abrupt changes, though rare (but not so rare, for instance in some families of the Coleoptera), do occur, for instance, when a fertile tetraploid mutant appears, or when, in sympatric “speciation” there is an abrupt ecological separation; as it has been described in several instances in different insects when a small section of a population suddenly adopts as larval food a different plant.
I must avow that I fail to see, when we consider the whole of living organisms and their evolution, how on Earth should we try to restrict the concept of species to a realistic view of something that occurs only in a more or less common number of instances and that may well apply to only some members of a group of closely related organisms, while for any supposed characterising character we have examples of transitional conditions and transition is anyway the basic requirement of evolution.
As we saw, quite correctly, some authors such as for example Bock (Citation2004), distinguish between a restrictive “biological concept of species”, the “taxonomic concept” and the “species category”, but when we are concerned with discussing phylogenies and evolutionary systematics, the “biological concept” is practically useless, besides the crucial fact that the very concept of evolution assumes the historical transition from one “species” to another with a continuity in the transmission of much of the gene pool.
Indeed, to the difficulties arising from the many varied examples of living beings where the concept of gene pool applies only in some limited way, we must add that the “biological concept” as the concept of “breeding true” clearly applies to just a few generations; thus, it is basically an atemporal one, just like the typological concept. Indeed, when we deal with the problem from an evolutionary standpoint, that is, considering considerable time spans, clearly, just as pointed out by Bock, in principle we shall never be able to fix the limits of any taxon in terms of the “biological species concept”, and “species” should rather be conceived as “fuzzy sets” of more or less distinct populations. Obviously, as we have already briefly mentioned, when this is considered from an evolutionary standpoint, that is, when we consider instead of contemporary organisms a sequence through time, the concept of “potentially” breeding true, this will appear probable for populations that are both similar and belonging to generations close in time, but we shall never know when the differences, either genetic, morphological or other, between organisms of the same lineage became sufficient to bar even potential cross-fertilisation.
Indeed, in a number of instances, the barrier preventing interbreeding is morphological or behavioural or, among plants, by their dependence on different pollinators, while artificial fertilisation may be successful.
That this may be a problem even for contemporary organisms is proved in the wild by several examples of “clinal” variability. In different instances of “species” ranging around desert or arctic regions, there are various characters typifying each sub-population, but these are joined by areas where the same characters merge in a gradual transition, and yet there is an area or more where the extreme variants overlap and do not cross. In some instances the impossibility of cross-fertilisation is the consequence of sufficient genetic or chromosomal differentiation, but in others it depends on morphological differences, differences in sound production or other differences which prevent mutual recognition of individuals. Similar instances are known not only for linear clines but also for checkerboard distributions of differentiated populations (= subspecies). It is self-evident that during chronological evolution “chrono-clines” must have been the rule, even if we take for granted the evolutionary model of “punctuated equilibria”.
Given all the above listed objections, we must conclude that the “biological concept of species” may apply only for the distinction of contemporary organisms and even for them to a limited extent and, therefore, it is practically useless in evolutionary studies. Obviously it cannot be denied that there are examples which perfectly embody the “biological concept of species”, that is, instances when population and “species” may be taken to coincide, but this is just when we consider small, strictly endemic population-species, such as when we consider the endemic populations of small islands, caves and the like.
An “evolutionary” concept of species has been advocated by some famous palaeontologists and botanists (e.g. Simpson Citation1951; Grant Citation1983). They suggest that the “species” is the ensemble of individual beings who belong to the same phyletic lineage and occupy the same ecological “niche”.
This is a strange proposal as it assumes that we have a perfect knowledge of a number of factors that we usually either ignore, or almost so. How can we know, even for living animals, to say nothing of fossils, whether two organisms occupy the same ecological niche, unless we have made a thorough study of all aspects of their biology (not to consider the many instances where the same population in the various subareas of its distribution uses different resources, id est occupies different niches according the overall composition of the local biocenosis and even the same individual may occupy different niches according the season, its developmental stage or, in migratory species, its temporary habitat)! And what about animals such as some insects whose population is in the same area subdivided into different strains whose larvae are strictly linked to different plants, or such as Cuckoos (Cuculus canorus) whose different females are linked for reproduction by a complex of imprinting and genetics to different parasitised species?
Moreover, the supposed belonging to the same phyletic lineage is commonly the subjective judgement of individual scholars and is precisely what we try to establish in any study of evolutionary systematics.
Indeed, the most recent advances of genetics, molecular biology and morphology prove that even the more sophisticated methods are not always foolproof when trying to establish whether two specimens belong to the same phyletic lineage.
Another principle suggested for the identification of “true species” is “recognition”, that is, the fact that in nature individuals recognize as sex partners some individuals and not others. This principle simply fails for lack of evidence in a lot of instances and will not work with almost all the organisms which have “external fertilisation” and simply spawn their gametes in the environment, as then are the gametes themselves who recognize their proper partner; it cannot apply to all organisms that do not reproduce by bi-parental reproduction. Finally we know a fair number of instances of animal taxa consisting of parthenogenetic females (commonly tetraploid) which, however, require mating with males of the closest diploid gamic population. In such cases, the aploid sperm enters the diploid egg and so activates the cleavage, but is promptly destroyed; clearly the animals fail to recognize the “specific differences” of their partner.
Equally untenable is the principle of “cohesion”. This has sometimes been practically identified with “gene pool” (e.g. by Templeton Citation1989, but also in a broader sense by other scholars who have maintained that we should consider as a species a system of organisms showing some regular and precise features of “internal cohesion”; by the way, by this principle all instances of obligatory symbionts or parasites should be held to belong to the same species as their lucky or unlucky host), a principle which, in order to be implemented, would demand such a complete knowledge of all aspects of the biology of each population considered such as it is, which is rarely attainable. Moreover, it obviously cannot be used for fossils, for populations which are not contemporary and that commonly were not living in the same locality.
Finally, whether we follow “gradualist” models or we agree with a mechanism of “punctuated equilibria”, we must in any case expect in any evolution from one “species” to another, that there will be a more or less long transitional phase.
We must now consider the cladistic concept, which is often called “phylogenetic” (a bad usage as literally it would have the same meaning as that of “evolutionary concept”).
Though the core of Hennig’s (Citation1950) hypothesis is that each “character” of any species occurs either in its “plesiomorph” or in the “apomorph” state, this being the condition “derived” from the “plesiomorph” state, so that by mapping the distribution of the plesiomorph and apomorph states in the set of species considered there will appear groupings of species linked together by synapomorphies (synplesiomorphies being deemed of no significance) and it is possible to build a cladogram of the phyletic connections of the various taxa studied (properly speaking Hennigian cladograms are phylogenies of characters, not of taxa, while orthodox cladists emphasize the distinction, in practice almost all scholars consider their cladograms to show the phyletic connections of taxa. Moreover, many Hennigian cladists call a “species” the ensemble of individuals which belong to one lineage and that live between one dichotomous branching and the next one.
As we shall see in the next paper, when properly discussing cladistic models, as this concept is rooted in a theoretic framework for the study of evolution that may be squarely falsified, it is thus no better than the others.
Let us now turn to the debate on the ontological status of “species” and precisely whether the species should be considered to be a “class” or an “individual”. In practice, this is just an updating of the old debate on the nature of “universals”. Actually, Ghiselin (Citation1997) and the many scholars who have concurred with him hold that each “species” is an “individual” and basically maintain that species really exist in nature and evolve; more precisely they are what really evolves. “Classes”, as they are by nature definitions, are clearly abstract concepts and, therefore, cannot evolve or, rather, as also the concepts associated with any word can change in time, thus the changes in the definition might well be entirely independent from the actual changes that the “species” has made through time (a terminist can point out that the universal “motorcar” in 1890 described something considerably different from the idea it conveys nowadays, though, should we meet with one of these old contraptions, we would probably recognize it as a motorcar).
Thus, the “species” is thought of as a sort of collective individual, a sort of what, in law, is a “legal person”. There is no question that collective individuals exist: a family and a football team are collective entities with a cohesive behaviour, so that in their activities through time, though their composition changes and they adapt to environmental opportunities, they may still each be considered as a unit. And such “beings” do indeed exist in nature; the populations of a beehive or of an anthill are very much this sort of thing.
This is, to put it as a rough approximation, the essence of the thesis that a “species” is an individual. While not a few scholars subscribe to this theory, about as many definitely oppose it.
My opinion is that the debate depends on a misunderstanding.
As previously said, to the advocates of the “species-individual”, the species as a collective individual is a sort of “legal person”, which is like a company where shareholders or administrators may change, or a football team where the individual players continually change, yet both the company and the team persist and “evolve”. Yet the advocates of the “species as individuals” fail to perceive the fact that, just like any company on the market or a football team in a league, what they are considering is the individual populations, not the species.
This is clearly proved by different situations that, in fact, have often occurred in nature. Let us just consider such well-studied instances as the evolution of the island populations of such archipelagos as Hawaii or the Galapagos: it has been possible to show how island-hopping by small groups of colonists from some other island, having settled into a new habitat which offered a range of potential ecological niches, has led them to radiate into new sets of species which the taxonomist could eventually arrange even into different genera! Though we shall never know, it is possible that sometimes such a process may have been started by a single pair of individuals: a single micro-population evolving into a whole range of other populations, which we call “species”.
Coming back to our argument, naturally it would be easy to list a number of collective organisms that function as single individuals (colonials, many social symbionts) and as many that, though acting independently or even competing in some functions, still act as a unit, for instance, in territorial activities. Yet such collective or semi-collective units are not species, but populations (though also there are great difficulties in providing a good definition of “population”) and it is on single populations that environmental selective pressures apply and thus populations evolve.
At this stage of our discussion it should be clear that we must agree that the “species” is a “universal” to be considered as a legitimate abstraction from our empirical experience of a number of different individuals; that is, we use the names of each taxon in a “terministic” context. It then follows that in discussing evolutionary problems it is perfectly legitimate to use the term species, the names of single taxa and even such secondary abstractions as “prey-species”, “index-species”, etc., provided that it is clear that what really act in the evolutionary process are the mutual interrelations between the members of each local population and those between the populations of each biocenose, and finally, the interactions between the various members of the biotic community and the environmental abiotic factors and that any such population is a sort of “fuzzy set” as, at least in a sufficiently long temporal lapse, it will change into something which we would deem a different sort of thing.
Indeed, we should always remember that those that evolve or possibly become extinct are not species, but populations, which may well include more than one taxonomic species. For instance, the extinction of a given population necessarily results in the extinction of its obligatory parasites and strictly specific predators, and of some plants if they are their only pollinators, etc.
Moreover, it will be a common occurrence that while certain populations belonging to a given taxonomic species evolve, possibly along diverging paths, others, quite possibly even genetically closely related with some fast-evolving ones, remain perfectly stable. Indeed when we consider populations widespread over considerable areas, we must expect that two conditions may occur:
That the population will be unevenly distributed over its range (indeed no one has ever met with a widespread population whose members have the same density over the whole range of the population, as it is considered impossible that over the whole range of a widespread population the environmental conditions will be exactly the same). But when we say that local resources will be more or less suitable, we just say that different sets of selective pressures (or should we say constraints?) drive adaptive selection at each location towards more or less diverging paths. Moreover, as most if not all environmental factors may change either slightly or even considerably, periodically or occasionally through time, though there will be long-term tendencies in selection, the actual impact of the various environmental factors will be necessarily to some extent erratic.
That widespread populations are much more subject to splitting than strictly localized ones. Indeed, as we have stressed, widespread populations not only will necessarily tend to differentiate into diverging sub-populations (their divergence depending on the degree of heterogeneity of the occupied range), but they are also more subject to fragmentation as it its more probable that barriers splitting the population may develop, but also that some catastrophe, such as an epidemic, the appearance of a new predator or the like, may wipe out the very thin populations that lived in the less-favourable environments of the population’s range and thus open the ideal scenario for a typical allopatric process of “speciation”.
When we consider localized populations the picture is considerably different. Unless the range of the population is strictly limited by an effective barrier, such as for instance the sea, a localized population, even when locally abundant, has its range limited by one or more specific environmental requirements or by its inability to meet the challenge of the habitats outside its range. This granted, it seems likely that in a first period an unbranching ortho-evolution will prevail, to be followed by a period when stabilizing selection will prevail. If and when either the local habitat changes, and then the chances are more for the extinction of the population concerned, though a new period of directional evolution is possible, or, if one or more of the limiting factors that prevented the spread of the population vanish from the surrounding areas, the chances are for a spread of the population and the development of the conditions for “parapatric” evolution.
Clearly, these scenarios do not precisely apply to marine populations where the high mobility either of planktonic larvae or of nektonic fishes, cetaceans etc., dictates different conditions (and I feel that the main factors to be considered are the fluctuation of the water temperature and its control on the abundance of micro-plankton with the consequent chain reactions on the total biocenose, while temperature itself is often locally controlled by the morphology of the shores, currents and a global extremely stable abyssal habitat).
Given these premises, let us, indeed, again see by additional arguments why we can not escape the double conclusion that, as I said, (1) what evolves are populations and (2) that evolving populations may well be made up by more than one “species” (and on this, see further on).
The truth of the statement that what evolves are populations should be so obvious in the light of the arguments proposed in the previous pages that it should not deserve any further arguing. However, it may be worth calling attention to commonly known facts and to their general implications that will validate this statement beyond possible argument.
In fact, objective consideration shows that both experimental work and field observation always deal – and it could not be otherwise – with populations, whether they are artificial, such as experimental, captive-bred populations, or when the investigations concern natural populations in the wild, where we study the distribution and changes in some features, be they morphological, molecular or other; and, if we want to make a precise statement, we study a smaller or larger group of specimens that we deem significantly representative of a population, this being an ensemble of individuals that interact among themselves. In a more restricted sense, we call a population the ensemble of similar individuals that occupy and exploit certain resources of a more or less precisely defined ecological niche. Such an ensemble may actually occupy different niches in the same general area, just as happens with seasonal changes or when developmental stages or the adults of different sexes actually use different habitat resources, but, in any case, the interactions between members of this complex of individuals are so closely “mandatorily” related that failing the regular functioning of the network the “population” will perish, must migrate or must change either morphologically, physiology or some other relevant way – that is, must evolve new adaptations.
Indeed, we know of a large number of instances when, be it in the wild or under domestication, a given group of organisms, either as a consequence of isolation from their parent stock or, having colonized a different environment or established a new ecological niche, have developed there a new set of adaptive traits (morphological, physiological, reproductive etc.), that is they have “evolved”. It is the familiar pattern of the textbook’s allopatric, parapatric, etc. “speciation process”. However, when we think of the actual mechanism of the evolutionary process, we see that what has happened is that one small or large group of individuals, perhaps a single one, stemming from a larger population and quite possibly in no way distinguishable from the bulk of the remaining original population, has become isolated (relict populations, founder population, etc.) and then, by the combined effects of genetic drift and selective pressures, its progeny has more or less changed = evolved from its parent population.
Indeed, even if we subscribe to entirely saltationist models of evolution, we must see evolution as the change, whether slow or rapid, of some characters in some individuals and in the increase in numbers of such changed individuals until they become, locally at least, so numerous or exclusive of a given habitat as to deserve, in our subjective judgement supported by what we deem sufficient evidence, to qualify as a distinct taxon.
All this has been familiar to biologists engaged in field work, and especially to those concerned with problems of conservation and management, so that many increasingly concern themselves with “evolutionarily significant units” (ESUs) which are defined as any group, be it large or small, of individuals that by its relative or total isolation, be it due to any cause, may potentially show evolution making it more and more divergent from its ancestors (see Conner & Hartl Citation2000; Crandall et al. Citation2000; Fraser & Bernatchez Citation2001).
Corollary
If we really look into the term, “speciation”, this is commonly and improperly used to mean a change in one population or in a section of a larger population such as to result in a sufficiently distinct new population, while (given the conclusions of Proposition 1) it properly means the mental process of a biologist who, on his own assessment of the available evidence, properly describes a new taxon and assumes that it evolved from a somewhat and sometimes precisely identified different one. Indeed, in principle it is possible to describe the evolution from a situation A to B just by a verbal or graphic series of descriptions without any reference to any formal taxonomy. Phylogenetic cladograms, in fact, should merely graphically describe the relationships among the individuals concerned, and the black blobs in our diagrams really represent an indeterminate number of individuals so closely related that the minute dots representing each one of them merge together in the image. The formal placing of each individual (as in many instances what we have is just one specimen or part of it) or population into a formal classification is a secondary and, as we have shown, necessary convention. To mix the two procedures: to think of establishing the best possible phylogeny, and formalizing it into the framework of a formal taxonomy, and consider these as a single endeavour is definitely a logical mistake. It follows that, as any adaptive evolution of any population of living organisms normally impinges simultaneously on a number of features of the evolving population – co-ordinated and simultaneous changes in the morphology, physiology, ecology etc. of the population as a whole – in principle it subsumes the assessment of the evolutionary affinities of the organisms under consideration as well as the whole of all the possible evidence available, while formalizing into a taxonomic framework such phylogenies as we may reconstruct, and necessarily depends on listing characters as if they were separate items.
Proposition 3: There are difficulties with the concept of population and, if symbioses are considered, what in fact evolves are always and only symbiotic complexes.
I shall reiterate first that, just as was Darwin’s idea, and as argued in the previous pages, I regard the concept of “species” in biology as a kind of secondary “universal”, a rather vague way of grouping the host of mental images corresponding to the primary universals: the individual species, which, in turn, are in no way essentially different from any other common name, apart from being characterised by a definition (or diagnosis) such as fitting one or more populations and provided according the “International Codes of Nomenclature” for the purpose of the identification of a given group of organisms, so that people will know which are the organisms (it does not matter whether dead or alive) discussed in a given context, and that taxonomic species “evolve” only insofar as their diagnosis may be changed in time by the addition or deletion of given characters, may be deleted as synonyms, and may be moved up or down in taxonomic rank. Species therefore are abstract universals comprehensive of one or more populations which we “know” by our experience of some, preferably many, specimens. What really evolve in nature, let us stress it again, are populations.
However, to define what a “population” is, is far from simple.
Let us consider as a first example an apparently simple concept, “Florentines”. On consideration, it is easy to see that the term may be used with entirely different meanings: it may be used to mean all the people presently living in Florence, and thus it would cover a number of foreigners (like in a sentence such as “Florentines require so many cubic meters of water per day”); but take the sentence: “80% of Florentines voted in the last ballot”; here the term covers only Italian citizens legally entitled to cast a vote; and, again: “Florentines considered Leonardo the greatest living artist”; its object is here an ensemble of past people! And we can make as many other examples as we want, so that there arises to some extent a circularity of precisely the type envisaged in the discussion of our first proposition.
It is plain that either we begin a statement by a clear definition of the meaning of each term used in the following discussion, or we trust that the argument itself will clarify the meaning of each term. Anyway, it is clear that the same term may cover some basically different concept of “population”.
If now we consider some zoological examples, we meet with a number of difficulties. Suppose that I am discussing the sparrow population of a given town; then there are no problems: I am dealing with a given section of an ensemble that I have defined as “sparrow”. But what do I mean, should I speak of the “population” of an anthill? For many kinds (“species”) of ants, it will be just the same as in the above example of “sparrow”, but what happens should I be discussing some “slave-making” ants, which do not produce “workers” but only “soldiers”, apart from queens and males? With them the vital function of “workers” is provided by individuals of a different “species” captured as larvae or pupae and thence bred in their captors’ nest. However, the anthill’s life functions just as if the workers were “conspecific” with the soldiers, that is, they are by all means a “population”, but a population made up by two different “species”!
How shall we judge, just to take another example, such instances where one or more “queens” of a given species (such as Atta decapitans) enter into an anthill, kill the original queen of the “parasitized” anthill and substitute for her? With time the population will be a “normal” one as all its original members will be dead, but there will be a transitional phase when the workers of the parasitized species care for the eggs and larvae of the queens that have usurped the anthill.
Still sampling in the ant world: what about the many “myrmecophilous” arthropods specialized to live as either symbionts or parasites in anthills?
A last caveat: we must also remember that there are a large number of instances where the interactions between given plants and given animals or between developmental stages of different taxa are so strictly interdependent that neither taxon of the system may survive in the absence of sufficient numbers of the other, so that they exist as a functional unit (and I already mentioned the instances of plants depending for pollination on specific insects, and one may add those whose seeds will not germinate unless by passing through the gut of some mammal or bird, the substances preventing untimely germination having been removed by the action of the gut secretions).
Taken together, all instances of symbioses or of parasitism, and especially when we deal with strictly interdependent symbioses, pose a problem: should the notion of population refer separately to each member-species of the system or should it be taken to mean the interdependent system as a whole, as its members must have co-evolved interacting as a complex with the selective pressures of the ecosystem? This is, I think, a crucial question, as practically all known organisms are in some ways members of more or less complicated symbiotic webs, quite often strictly obligatory ones.
One last warning: in order to circumvent the difficulties posed by populations of mixed taxonomic species and yet save a “realistic” concept of species, several authors, and especially ecologists, deal with such things as “integrated species complexes”. This is just a verbal trick! Indeed, not only the different member-species of many a symbiosis function as an integrated unit even more than many social organisations made of members of a single species, but they must have co-evolved considerably, even to the family level (e.g. the beetle subfamily Paussinae).
Conclusion
It is plain (and it is not necessary to go into the endless variety of the actual interactions that occur in nature to make up interactive systems among individuals that may properly be considered as “populations”) that what offer the basic raw material on which selection and time work are functional units that we call “populations”, but that this term covers such a wide variety of different mutual interrelationships among organisms that it is impossible to provide a definition of the term “population” precisely covering them all.
Corollary
The considerations made in the previous paragraph result in a concept close to that advocated by the “Gaia hypothesis”, that is, that the whole biosphere and the planet itself should be considered and studied as a unit. Theoretically or, if you like, philosophically, this is a thesis that has considerable appeal, but the scholar is not at all at fault when studying separately the various members of a biocenose or if he tries to trace the lineage of some of its members: an opera or a painting are whole works of art, yet the conductor preparing its performance trains the musicians and singers on the separate acts, single scenes and, quite often, just a few notes, and the art expert may well discuss separately the background landscape or the dressing of a figure; this is perfectly valid – one may say necessary – provided that the conductor or the art expert always remembers that he is working on a detail of something that is in fact a whole.
References
- Bock WJ. 2004. Species: The concept, category and taxon. Journal of Zoological Systematics and Evolutionary Research 42:178–190. doi:10.1111/j.1439-0469.2004.00276.x.
- Conner J, Hartl D. 2000. A primer of ecological genetics. Sunderland, MA: Sinauer. 304 pp.
- Crandall KA, Bininda-Emonds ORP, Mace GM, Wayne RK. 2000. Considering evolutionary processes in conservation biology. Trends in Ecology and Evolution 15:290–295. doi:10.1016/S0169-5347(00)01876-0.
- Darwin CR. 1859. The origin of species by means of natural selection or the preservation of favoured races in the struggle for life. London: John Murray.
- Eco U. 1997. Kant e l’ornitorinco. Milano: Bompiani. ix + 454 pp.
- Fraser D, Bernatchez L. 2001. Adaptive evolutionary conservation: Towards a unified concept for defining conservation units. Molecular Ecology 10:2741–2752. doi:10.1046/j.1365-294X.2001.t01-1-01411.x.
- Ghiselin MT. 1997. Metaphysics and the origin of species. New York: State University of New York Press. ix + 377 pp.
- Grant V. 1983. The synthetic theory strikes back. Biologisches Zentralblatt 102:149–158.
- Hennig W. 1950. Grundzüge einer phylogenetischen Systematik. Berlin: Deutscher Zentralverlag.
- Lamarck JB. (de Monet de Lamarck). 1779. Flore Françoise, ou description de toutes les plantes qui croissent naturellement en France, disopsée selun une nouvelle méthode d’analyse, 3 vols. Paris: Imprimerie Royale.
- Linnaeus. 1758. Systema naturae per regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. X ed. Holmiae: Laurentii Salvii. iv + 824 pp.
- Simonetta A. 1983. The myth of objective taxonomy and cladism. Much ado about nothing. Atti della Societa Toscana di Scienza Naturali, Memorie Serie B 89(1982):175–186.
- Simonetta AM. 1986. Logic, taxinomy, taxa and reality. Contributi faunisici ed ecologici. Camerino 2:1–35.
- Simonetta AM. 1988. Logica, tassinomia e realtà. In: Ghiara G. e coll, editor. Il problema biologico della specie. Collana U.Z.I. - Problemi di biologia e storia della Natura 1. Modena: Mucchi. pp. 59–78.
- Simonetta A. 1992. Problems of systematics: Part 1. A critical evaluation of the ≪species problem≫ and its significance in evolutionary biology. Bolletino di zoologia 59:447–463. doi:10.1080/11250009209386707.
- Simonetta A. 1993. Problems of systematics: Part 2. Theory and practice in phylogenetic studies and in systematics. Bolletino di zoologia 60:323–334. doi:10.1080/11250009309355832.
- Simonetta A. 1995. Some remarks on the influence of historical bias in our approach to systematics and the so called ≪species problem≫. Bolletino di zoologia 62:37–44. doi:10.1080/11250009509356049.
- Simonetta A. 1996. Systematics: Is a historical perspective useful to understanding modern debates on systematics and are we really equipped for sound evolutionary systematics? Memorie della Società Italiana di Scienze Naturali e del Museo Civico di Storia Naturale in Milano 27:109–114.
- Simonetta AM. 1999a. Is parsimony a justified criterion in the assessment of possible phylogenetic reconstructions? Italian Journal of Zoology 66:159–169. doi:10.1080/11250009909356251.
- Simonetta A. 1999b. Una critica del pensiero dicotomico in sistematica. Systema Naturae 2:55–64.
- Simonetta A. 2000. Homologies, analogies, synapomorphies and homoplasies: Problems and considerations. Rivista di Biologia/Biology Forum 93:45–56.
- Simonetta AM. 2007. A reliable absolute molecular clock, is it possible? In “The Rise and Fall of the Vendian (Ediacaran) Biota. Origin of the Modern Biosphere. Transactions of the International Conference on the IGCP Project 493. Moscow: GEOS. pp. 64–66.
- Simpson GG. 1951. The species concept. Evolution 5:285–298. doi:10.2307/2405675.
- Templeton AR. 1989. The meaning of species and speciation: A genetic perspective. In: Otte D, Endler JA, editors. Speciation and its consequences. Sunderland: Sinauer Associates. Chapter 1, pp. 3–27.
Classical and Medieval authors
The Classical and Medieval authors quoted pose a problem for their inclusion in the references: the opinions of many of them, for instance those of Antisthenes and Protagoras, are known only as fragments quoted by other, later authors such as Cicero, Boethius, etc.; otherwise, the first printed editions of the works of the classical authors like Plato or Aristotle have been more or less criticized and these works are now quoted after some more modern, better editing. Similar difficulties occur with the quotation of Medieval authors. Therefore, rather than quoting the following authors by their individual works, when known, I give here three references to general treatises, which may be consulted for guidance in tracing the original statements of the following authors:
Abelardus Petrus; Albert the Great (St.) (Albert of Böllstadt); Antisthenes; Aquinas, Thomas (St.); Aristotle; Avicenna (Ibn Sina); Boethius, Manlius Severinus: Heraclitus; Hobbes, Thomas; Ockam, William; Plato; Porfirius; Proclus; Protagoras; Roscellinus; Simplicius. See:
Cohen MR, Nagel E. 1963. An introduction to logic and scientific method. London: Routledge & Kegan. 467 pp.
Copleston FC. 1972. A history of medieval philosophy. London: Methuen. 399 pp.
Russel B. 1946. History of Western Philosophy. London: Allen & Unwin. 916 pp.