Where's the Species

8/2/2019 Where's the Species

1/8


2/8

MARC ERESHEFSKYcan be used for individuating such taxa: "(1) one or more diagnosticcharacters (of any kind as long as they are heritable), and (2) reproductivecohesion" (1987, pp. 341-342). Of these criteria, only the first is neces-sary; the second is useful "because the diagnostic characters of a speciesmay be restricted to a single sex, ontogenetic stage or morph" (1987, p.342). Thus species are groups of genealogically connected organismswhich share at least one diagnostic character. Furthermore, that charactermay be, bu t need not be, the ability to interbreed, or following Paterson(1985), the ability to recognize each other as mates.If we take Cracraft's description of species literally, then it implausiblyfollows that every group of genealogically connected organisms whichshare a single distinctive heritable trait is a species. For example, accord-ing to Cracraft's definition, every time Dobzhansky and his co-workerssuccessfully introduced a new and heritable mutation into an interbreedinggroup of Drosophila, they created a new species. And, according toCracraft's definition, the group of organisms Homo sapiens consists of atleast thirty different species (corresponding to the thirty races of Homosapiens Mayr speaks of (1970, p. 396)) and probably many many more.While there may be reasons for recognizing these groups as some sort oftaxonomic units, it seems implausible to assert that each is a distinctspecies.Furthermore, from the description given above, we see that a species is"defined as an irreducible cluster of organisms, within which there is aparental pattern of ancestry and descent, and which is diagnosably distinctfrom other such clusters" (Cracraft 1987, p. 341; my italics). Earlier in hispaper Cracraft writes, "Given a definition of species, irreducibility impliesthat entities meeting this definition cannot be subdivided into other,smaller entities which themselves play the role of species within theorieshaving them as their entities" (1987, pp. 333-334). If, as Cracraft asserts,the role of species in evolutionary theory is to be the most basal groups ofgenealogically connected organisms which share at least one distinctivetrait, and a species cannot be subdivided into smaller such groups, then itfollows that only groups of genealogically connected organisms whichshare a single distinctive trait are species. So, according to Cracraft'sspecies concept and his notion of irreducibility, the group Homo sapiens,for example, cannot be a species because it can be subdivided into smallergenealogical clusters of organisms which share a distinctive trait. Thesame consideration applies to almost any other group of organisms whichbiologists usually take to be a species, or even a subspecies.On the one hand, Cracraft's description of species asserts, quiteimplausibly, that taxa below the level of race are species. On the otherhand, it excludes most if no t all groups of organisms which are usuallytaken to be species from being species.Cracraft may counter that my portrayal of what species are by his

90


3/8

WHERE'S THE SPECIES?Phylogenetic Species Concept is not what he means by species. This maybe the case since I suspect that Cracraft thinks that species are what manybiologists now take to be subspecies (see Cracraft 1983, p. 173). Never-theless, his description does not capture this. If that is what Cracraft takesspecies to be, then he needs to revise his description.

Another criticism I have of Cracraft's Phylogenetic Species Conceptconcerns what his concept does not do. His definition is entirely a claimabout what phylogenetic groups or patterns of organisms constitutespecies; it gives no information about the processes or causes whichproduce and maintain such groupings. Nevertheless, Cracraft writes,

One definition is preferable to another because it allows us to understand nature better,that is, within the context of theories and hypotheses about biological patterns andprocesses, that definition lends more coherence and explanatory extensibility to thoseconjectures (1987, p. 331).

Since Cracraft's definition of species makes no reference to an y processeswhich make species what they are, I do not see how his definition can lenditself to any "explanatory extensibility" concerning species phenomena.To make this point clearer, consider the Biological Species Concept

of Mayr (1970, p. 12): "Species are groups of interbreeding naturalpopulations that are reproductively isolated from other such groups."Undoubtedly this definition of species has its problems, but at least itidentifies processes which might help explain certain species phenomena.For instance, the processes of gene flow and reproductive isolation, if trulyeffective, can be cited to explain why species maintain their unity. Andciting the lack of gene flow between the subpopulations of a species canprovide an account of how speciation occurs. I agree with Cracraft (1987,p. 339) that a major problem with the Biological Species Concept isthat the processes it specifies are simply not found in many groups oforganisms thought to be species by many biologists. But at least thatdefinition posits some processes which if present and truly effective couldhelp explain certain species phenomena.

Cracraft's Phylogenetic Species Concept, on the other hand, is only apattern claim about species; according to that concept, species are thesmallest diagnosable clusters of genealogically connected organisms. Un-doubtedly stipulating that species are such clusters does imply whichgroups of organisms are and are not species. But this does not helpexplain, for example, how species maintain their unity or how speciationoccurs. In other words, Cracraft's Phylogenetic Species Concept providesno insight into why species are species; or more specifically, it provides noinsight into how species maintain and lose the unity which we perhapssomewhat vaguely associate with species. Providing some such insight intothe nature of species seems to be an important desideratum of an y speciesconcept. Because Cracraft's definition fails to provide any such insight, it

91


4/8

MARC ERESHEFSKYfails Cracraft's own criterion for a preferable species concept - it lacks"explanatory extensibility" (1987, p. 331).My final point concerning Cracraft's Phylogenetic Species Concept isno t so much a criticism of the concept, bu t a criticism of how Cracraftargues for it. After reviewing the Biological Species Concept, Cracraftintroduces his Phylogenetic Species Concept as follows: "the previousdiscussion suggests a deceptively simple solution: equate species withevolutionary units" (1987, p. 341). This would be a simple solution if wehad a clear idea of what an evolutionary unit is. But, as Cracraft is wellaware (1983, p. 163), there is a wide range of opinions on what groups ofbiological entities are evolutionary units. For example, Cracraft andEldredge (1980, pp. 89-90) and Ghiselin (1987, p. 137) maintain thatevolutionary units must be reproductive units.l Hull (1980, pp. 327-329)and Rosenberg (1985, p. 139) hold that evolutionary units are lineages:groups of organisms, and perhaps other biological entities, connected byparent-offspring relations. Mary Williams (1985, pp. 584-585) seems tothink that evolutionary units are groups of genealogically connectedorganisms which are exposed to the same selection forces. And Bunge(1981, p. 284) asserts that evolutionary units are "biopopulations." Giventhe diversity of opinions among biologists and philosophers as to what anevolutionary unit is, equating species with evolutionary units is not a"deceptively simple solution" to the species problem. It is just replacingone problem with another problem.Let us turn to the Phylogenetic Species Concept suggested by Mishlerand Brandon (1987). This concept was first introduced in Mishler andDonoghue (1982) and is summarized as follows:

A species is the least inclusive taxon recognized in a classification, into which organ-isms are grouped because of degree of monophyly (usually, but not restricted to, thepresence of synapomorphies), that is ranked as a species because it is the smallest"important" lineage deemed worthy of formal recognition, where "important" refers tothe action of those processes that are dominant in producing and maintaining lineagesin a particular case (1987, p. 46).This definition places two requirements on species. First, species must bemonophyletic groups or lineages of organisms. Second, such lineages mustbe important in the sense that they are produced and maintained by suchprocesses as (but not limited to) actual interbreeding, selective constraintsand developmental canalization (Mishler and Brandon 1987, p. 37).Mishler and Brandon, following Mishler and Donoghue (1982), call thefirst requirement a "grouping" criterion, and the second a "ranking"criterion. They maintain that each criterion is necessary, and together theyare sufficient, for a group of organisms to be a species. I have commentsto make concerning both these criteria.Suppose we have a group of organisms and we know that they form a

92


5/8

WHERE'S THE SPECIES?monophyletic group. The group thus passes the "grouping" requirement ofMishler and Brandon's definition of species. What, according to thisdefinition, would make that group a species? From the above description,we see that it must be an "important" lineage. But what makes it an"important" lineage? According to the above description, it is thoseprocesses that are dominant in producing and maintaining such lineages.Thus an "important" lineage is a lineage which is produced and maintainedby those processes which produce and maintain "important" lineages. Isubmit that the "ranking" criterion of the concept being promoted byMishler and Brandon is circular; and because it is circular it does no t giveany way to distinguish species from 6ther monophyletic taxa. In otherwords, the Phylogenetic Species Concept of Mishler and Brandon doesnot tell us what species are, other than that they are monophyleticlineages.Mishler and Brandon might reply that their concept gives more infor-mation than that about species because it refers to those processes whichmay cause a monophyletic group to be a species. This is true, but thisinformation still does no t tell us what a species or an "important" lineageis. Here is an analogy: Suppose you do not know what chairs are and youwant to learn; I tell you that a chair is something which can be made witha saw and hammer, with a metal cutter, or by weaving reeds together. Ieven demonstrate how each process can cause the existence of a chair bybuilding chairs with each method. Do you now know what a chair is? I donot think so. Being told that each one of these processes, among others,can cause the'existence of a chair does not tell you what a chair is. Youneed to know what it is about these resultant objects, caused by variousprocesses, which makes them chairs. For instance, you need to know thatthey are devices on which people can sit, or some such description ofchairs.According to Mishler and Brandon's "ranking" criterion, species aremonophyletic lineages produced and maintained by gene flow, selectiveconstraints, developmental canalization, or some other process whichproduces and maintains species. Each one of these processes may beimportant in causing and maintaining the existence of particular species.But still, we do no t know why the lineages caused and maintained by suchprocesses are species. In other words, we do not know what it is aboutthese lineages which makes them species. Since Mishler and Brandon's"ranking" criterion does not tell us why such lineages are species, otherthan that these lineages are produced and maintained by those processeswhich cause lineages to be species, it leaves us in the dark as to whatspecies are.

What about their "grouping" criterion? According to Mishler andBrandon's species concept, all species must be monophyletic. But, asMishler and Brandon point out, the cladistic definition of monophyly

93


6/8

MARC ERESHEFSKY(the one they support) is no t usually applied at the species level (1987,pp. 46-47). To remedy this, they suggest the following definition ofmonophyly:

A monophyletic taxon is a group that contains all and only descendents of a commonancestor, originating in a single event (1987, p. 47).Is it plausible to assume that all species satisfy this definition of mono-phyly?Suppose A is a distinctive taxon because of a cluster of uniquesynapomorphies which originated at event x (see figure 1). A is a large,'healthy' species within which there is much interbreeding. Its membersare exposed to a common selection regime and they contain similarhomeostatic genotypes. Some years after event x, A gives rise to aperipheral isolate, B. The members of B interbreed, are exposed to theirown selection regime and sooner or later give rise to their own unique setof synapomorphies at event y. A, meanwhile, is unaffected by the estab-lishment of B. According to Mishler and Brandon's Phylogenetic SpeciesConcept, B is a species. I have referred to A as a species, and I think thatthis is fairly uncontroversial.2 The problem for Mishler and Brandon'sspecies concept is that A does not satisfy their definition of monophyly: Adoes no t contain all and only the descendants of event x; some of thedescendants of event x are contained in B.

Mishler and Brandon could reply that A is no t a species. But this seemsimplausible, for they would have to assert that there are no cases ofspecies which continue to exist after some of their members give rise to anew and different species. Nevertheless, suppose they follow Hennig(1966) and assert that no parental species survives a speciation event. Inother words, suppose that A is not one but two species, one originating atevent x and the other at event y. This maneuver, however, does not escapethe present problem because the first species in A fails to satisfy Mishlerand Brandon's definition of monophyly: the first species in A does notcontain all and only the descendants of event x; some of the descendantsA

B

Figure 1. A phylogenetic tree, where clusters of synapomorphies are shown as cross-bars.See text for discussion.

94


7/8

WHERE'S THE SPECIES?of event x are in B and some are in the later species of A. Regardless ofwhether one adopts Hennig's approach or not, there seem to be obviouscases of species which do not satisfy Mishler and Brandon's definition ofmonophyly. There are thus species which do no t satisfy the "grouping"criterion of their Phylogenetic Species Concept. As a result of this,Mishler and Brandon's species concept does no t plausibly apply to allspecies.Both Cracraft (1987) and Mishler and Brandon (1987) include adiscussion of the problems facing the Biological Species Concept. For themost part, each of their Phylogenetic Species Concepts avoids theseproblems. But what do these concepts have to offer in place of theBiological Species Concept? On the one hand, Cracraft's concept im-plausibly asserts that only the most basal groups of organisms, thosegenealogically connected and which share at least one distinctive heritabletrait, are species. Furthermore, Cracraft's concept is merely a patternclaim about species, providing no further insight into the nature of species.On the other hand, Mishler and Brandon's concept implausibly assertsthat all species are monophyletic. And even if we assumed that all specieswere monophyletic, their concept does not in a non-circular fashiondistinguish species from other monophyletic taxa.Some may find the Phylogenetic Species Concepts a step in the rightdirection because these concepts avoid some of the problems of theBiological Species Concept. I agree. But both definitions require revisionand supplementation if they are to be fully adequate. Adopting a popularAmerican expression, I conclude that both definitions need to moreadequately address the question in the title of this paper.

NOTES* The author thanks Joel Cracraft, Brent Mishler and David Hull for commenting onearlier drafts of this paper, and Elliott Sober for his constant help and encouragement.' For Cracraft and Eldredge (1980, p. 89), reproductive units are groups of organisms"among which there is at least occasional sexual reproduction." Whereas, for Ghiselin,reproductive units are groups of organisms among which there is "reproductive competi-tion" between their organisms (1974, p. 538).2 One author who disagrees with this is Hennig (1966). See text below for a considerationof Hennig's approach.

REFERENCESBunge, M.: 1981, 'Biopopulations, Not Biospecies, Are Species and Evolve', Behavioral

andBrainSciences 4, 284-285.Cracraft, J. : 1983, 'Species Concepts and Speciation Analysis', Current Ornithology 1,159-187.

95


8/8

MARC ERESHEFSKYCracraft, J.: 1987, 'Species Concepts and the Ontology of Evolution', Biology andPhiloso-

phy 2, 329-346.Eldredge, N. and Cracraft, J. : 1980, Phylogenetic Patterns and the Evolutionary Process,Columbia University Press, New York.Ghiselin, M.: 1974, 'A Radical Solution to the Species Problem', Systematic Zoology 23 ,536-544.Ghiselin, M.: 1987, 'Species Concepts, Individuality, and Objectivity', Biology and Philoso-phy 2, 127-143.Hennig, W.: 1966, PhylogeneticSystematics, University of Chicago Press, Urbana.Hull, D.: 1980, 'Individuality and Selection', Annual Review of Ecology and Systematics11,311-332.Mayr, E.: 1970, Populations, Species, and Evolution, Harvard University Press, Cam-bridge.Mishler, B. and Donoghue, M.: 1982, 'Species Concepts: A Case for Pluralism', SystematicZoology 31,491-503.

Mishler, B. and Brandon, R.: 1987, 'Individuality, Pluralism, and the Phylogenetic SpeciesConcept', BiologyandPhilosophy2, 37-54.Paterson, H.: 1985, 'The Recognition Concept of Species', TransvaalMuseum Monograph4, 21-29.Rosenberg, A.: 1985, The Structure of Biological Science, Cambridge University Press,New York.Simpson, G.: 1961, The Principlesof Animal Taxonomy, Columbia University Press, NewYork.van Valen, L.: 1976, 'Ecological Species, Multispecies, and Oaks', Taxon 25,233-239.Williams, M.: 1985, 'Species are Individuals: Theoretical Foundations for the Claim',Philosophyof Science 52, 578-590.

96

Where's the Species

Documents