Essential Membership Author(s): Joseph LaPorte Source: Philosophy of Science, Vol. 64, No. 1, (Mar., 1997), pp. 96-112 Published by: The University of Chicago Press on behalf of the Philosophy of Science Association Stable URL: http://www.jstor.org/stable/188371 Accessed: 02/05/2008 20:37 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=ucpress. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We enable the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact
[email protected].
http://www.jstor.org
Essential Membership* Joseph LaPortetl Amherst Departmentof Philosophy,Universityof Massachusetts,
In this paper I take issue with the doctrine that organismsbelong of their very essence to the natural kinds (or biological taxa, if these are not kinds) to which they belong. This view holds that any human essentiallybelongs to the species Homo sapiens,any feline essentiallybelongs to the cat family, and so on. I survey the various competing views in biological systematics.These offer different explanations for what it is that makes a member of one species, family, etc. a member of that taxon. Unfortunately, none of them offers an explanationthat is compatiblewith the essentialismin question.
1. There is an ancient philosophical doctrine according to which you and I essentially belong to the human species and, similarly,my pet cats could not have been non-felines. Contemporaryphilosophersare wont to state that doctrine as follows: individualbiological organisms essentially belong to the natural kinds to which they belong. (Such a renderingmay or may not be problematic-see below, Section 2.) This doctrine is not to be confused with the position that biological natural kinds have microstructuralessences. The two essentialist claims are quite different, though they are sometimes run together (see Forbes 1985, 193). The claim that organismsessentiallybelong to their respectivenatural kinds may be relatedin interestingrespectsto the claim that those *ReceivedSeptember1995;revisedJanuary 1997. tSend reprint requests to the author, Department of Philosophy, Bartlett Hall, The Universityof Massachusettsat Amherst, Box 30525, AmherstMA 01003-0525, USA. $For helpful comments on previousdrafts of this paper and/orprofitablediscussionof issues in it, I thank Eddie Abrams, Bruce Aune, Phillip Bricker,Fred Feldman, Lucy O'Brien, and audiences at the University of Massachusetts,Amherst, the 1996 MidSouth Philosophy Conference at the University of Memphis, and the 1996 Eastern Division Meeting of the APA in Atlanta. Specialthanks to the commentatorsfor these talks: Ben Bradley,Timothy Huson, and Marc Lange, respectively.I also thank anonymous refereesfor this journal. Philosophy of Science, 64 (March 1997) pp. 96-112. 0031-8248/97/6401-0005$2.00 Copyright 1997 by the Philosophy of Science Association. All rights reserved.
96
ESSENTIAL MEMBERSHIP
97
natural kinds have microstructuralessences, but the differencesbetween the two positions can be made clearby consideringvariousways natural kinds could turn out to be delimited. If (and this is an if) all natural kinds should turn out to be necessarilycharacterizedby particular microstructuralcharacteristics,then it would be true that natural kinds have microstructuralessences. But it would not follow that organismsbelong to their respectivekinds essentially.The latterwould have to be admittedonly if everymemberof any naturalkind possessed its kind's definingstructurein every possible world in which that member exists. For example, suppose that it were true that natural kinds had microstructuralessences, and that Homo sapiensturned out to be a natural kind characterizedby some DNA structureX. In that case, Darwin and all other humans would have X in virtue of being human. But this might be true even should there be possible worlds in which Darwin exists (as a non-human)while failing to be characterizedby X. So if it could be established that kinds have essences, it would not follow that individualmembersessentiallybelong to those kinds. Take a very different scenario: suppose that, say, morphological characteristicsshould define natural kinds. In that case, it would be false that natural kinds have microstructuralessences. But it might be true, nevertheless,that the membersof each naturalkind are members of the kind essentially.This would be the case if each memberof each natural kind were to possess the defining (morphological)characteristics of the kind in every possible world in which the memberexists. So the two essentialistclaims are distinct. The position that natural kinds are defined by microstructuralessences has sufferedfrom a variety of serious attacks launched by philosophers of science (see, e.g., Zemach 1976; Mellor 1977; Kitcher 1984a, b; Dupre 1981, 1989;Aune 1994;and LaPorte 1996;Mayr and other scientistshave made related criticisms).The claim about the essentialityof membershiphas, by contrast,largelyescapedthe attention of philosophersof science.This is interesting,since the doctrineis popular, and furthermoreit has been used to support other substantial philosophical claims. (Cook in his 1980, for example, has urged that the rigidity of the term 'cat' rests on particularcats being cats in all possible worlds containing them, and so on for like terms.) This neglect by philosophers of science is unfortunate. For the essentialismin question is, like the issue of the ontological status of species, such that while it is at core a metaphysicalclaim, it can nevertheless not be properlyevaluatedapart from biological considerations.In the present paper, essentialismabout membershipwill be examinedin the light of biological systematics.It shall be seen that essentialistshave tended to be rathernaive on scientificmattersrelevantto their claims.
98
JOSEPHLAPORTE
When the biological informationthat is neededto evaluatethe doctrine is supplied, I hope to show, it becomes apparent that the doctrine is quite implausible.' 2. The doctrineunder considerationcan be put as follows: TEM (for Theory of Essential Membership):If, in any possible world, an organism belongs (or does not belong) to a particular naturalkind, then thereis no possible world in which the organism ever fails to belong (or not to belong) to that kind. It should be observed that TEM gives an answer to two distinct questions: first, whether a thing could have failed, from its inception, to belong to a natural kind it belongs to, and second, whether it could cease to belong to its naturalkind, turninginto somethingof a different kind. TEM says no, no. The second answer is the more controversial, in part because of sorites problems. It can seem possible for a change of kind to occur in gentle stages (see Carter 1986). This consideration and others put TEM's answerto the second question on less firmfooting than its answer to the first. Thus, Kripke (1980, 114-115: see especially the lengthyn. 57) distanceshimselffrom the second issue while suggestingan essentialistline on the first.Nonetheless,explicitendorsement of the principle'sanswer to the second question has been given by various philosophers, including Wiggins (1980, 169-170) and Doepke (1992), for whom no member "of a natural kind ... could
become a memberof a differentkind" instead (p. 89). For example,he says, "no human person could become a member of another species" (ibid.). Now, TEM may need revision. Although, as is observed above, essentialistsdo attemptto expresstheirposition in termsof naturalkinds, it is clear that they intend their theory to apply to taxa: species, and so on. Sometimes this intended application to taxa is made explicit. Thus, N. Cocchiarella,for whom "an individualcanbelong to a natural kind only if being of that kind is essential to it" (1976, 205, original emphasis)explicitlyunderstands"naturalkinds to include the various genera and species of plants and animals," (ibid., 203; my emphasis) along with lifeless substanceslike water. Other times, essentialists'examples provide the clear signal that their theory is intended to apply to taxa: thus, the essentialist claim is applied to species (e.g., Homo sapiens;Doepke 1992), families (e.g., the cat family; see Cook 1980), and so on. If biological taxa are indeed natural kinds, as essentialists 1. Assuming, anyway, that biologists' conclusions are to be endorsed,somethingmost philosopherswill accept.
ESSENTIAL MEMBERSHIP
99
apparentlyassumewhen they appealto principleslike TEM, then TEM should fairly accurately convey the essentialist idea it is intended to convey. But if taxa are not kinds, then TEM, being a claim about natural kinds, is not an adequate renderingof the essentialistidea behind it, the latter being rooted in intuitions about the essentiality of humanhood, cathood, oakhood, and so on. If taxa are not natural kinds, then the essentialist view should be somethinglike: TEM': If, in any possible world, an organismbelongs (or does not belong) to a particularbiological taxon, then there is no possible world in which the organismever fails to belong (or not to belong) to that taxon. I happen to think that taxa are kinds. Hence, I think TEM captures the essentialistintuitions it is supposed to generalize.However, it has often been argued that species are individualsratherthan kinds (Ghiselin 1974; Hull 1976, 1978, 1980, 1987; but cf. Kitcher 1984b, 1987), and sometimes, too, that higher taxa are historical entities, if not individuals, and therefore that they must be non-kinds (see Ereshefsky 1991 and referencestherein). To try to counterthese opposing choruseswould requiremore space than I can afford to devote here. Since I take taxa to be kinds, and thus find no reason not to attributeTEM to essentialists,TEM is the doctrine I shall criticallyevaluate, under the assumptionthat taxa are indeed kinds (even if they are historicallyconnected;cf. Kitcher 1984b, 314). If I am wrong in supposing that taxa are kinds, then TEM is no doubt inadequate,and TEM' provides a much better renderingof the relevant essentialism. So those who maintain that taxa are not kinds are welcome to attributeTEM' to the essentialist,and to read the following argumentsas attacks on that claim. All of the argumentsI raise against TEM serve quite as apparentlyto undermineTEM'. 3. TEM has had its critics. J. Ballie (1990), for example, takes on Wiggins (1980), by appealing to examples from science fiction: it seems broadly logically possible that a man could turn into a giant fly, Ballie says, or a dog into a blob of living organic tissue. But the arguments of the present paper will be quite unlike Ballie's and the many others like them. Ballie argues for the possibility of massive changes in an organism'sfunction and design;the intuition that such changesare not possible lies at the very core of essentialist thinking. I will not offer such wholesaleresistanceto essentialists'intuitions.Rather,I shall suggest that essentialism fails because membershipin natural kinds can
100
JOSEPHLAPORTE
hinge on features that would seem to be accidentalon just about anyone's account.
This method of attack is surprisinglynew. The beginningsof a gesture in the directionI head has been providedrecentlyby M. A. Khalidi (1993). Khalidi has taken stages of an individual organism'sdevelopment to constitute natural kinds: hence, the natural kindspupa, larva, and imago. He suggests that membershipin these natural kinds is not essentialto an individual,and thereforethat TEM is false. Membership in these kinds is not essentialbecause individualsmay instantiatethese kinds without doing so in all possible worlds;indeed,many developing individualswill instantiate and then cease to instantiateeach of these kinds in the actual world (ibid., 112). I find Khalidi's case to be interesting. He may be right. My own present thesis can only be strengthenedif he is right. Still, it should be pointed out that his examples are not ideal for making his claim. Essentialists'examples of naturalkinds from biology (human,cat, and so on) tend to fit into the Linnaean categories (species, etc.) reasonably neatly. Examples as differentfrom these as pupa, etc., invite the reply that the newly purportedexamplesjust are not naturalkinds. The fact that pupa, etc. are temporal stages that organismspass in and out of in the natural course of life might allow for their principledexclusion from the natural kind camp. Our grasp of the concept naturalkind is by way of example, and typical examples that have been used as a handle for natural kindhood simply do not involve this temporalelement. Thus, Wiggins (1980, 24-27, 64, Chaps. 3 and 4) and othershave not counted exampleslike Khalidi's to be natural kinds.2 This possible objectionto Khalidimay or may not succeed.Whether it does or not is irrelevantto the fate of my arguments.But it is desirable to clearly avoid such an objection in the argumentsthat follow. The problem for Khalidi arises on account of essentialistsnot recognizing his examples as ones that were ever intended to apply to their theory. To avoid this problem, it will be important to evaluate the essentialistdoctrine in the light of the sorts of examplesthat essentialists themselves offer by way of illustratingtheir theory's application: species and higher taxa. That way, it will be ensuredthat the examples consideredin the argumentsare the very entities that are supposed to conform to the doctrinein question. So in what follows, I will consider 2. It is also questionablewhether Khalidi's examplesshould be recognizedas genuine taxa (as Hennig 1966, 6 suggests), and hence whether they undermineMET'. If they are taxa, the essentialistmight still accept MET' after addingto it the qualificationthat the relevant taxa must be of a category at least as high as, say, the species level. This qualificationwould not create trouble for any of the essentialist'sfavorite examples: humans, oaks, and so on.
ESSENTIAL MEMBERSHIP
101
examples that are found in essentialists'own accounts. In the present section, I will consider whether organisms essentially belong to their respective species. In the section that follows, I will go on to inquire whether higher taxa, like oak, cat, and mammal,are such that one is essentiallya member or not. In both cases I shall conclude that essentialism is untenable. I argue in the present section that organismsdo not essentiallybelong to the species to which they belong. To show this, I do not want to assume any particulartheory about what determinesthe boundaries of species.This complicatesmatters,sincebiologists do not agreeabout what species are, about what it is that makes the membersof a species members of the species. It is not feasible or necessaryto review every last one of the many approachesto the speciesquestion that are on the market (see the great variety in Ereshefsky's 1992b alone). But it is necessaryto run through the theoriesthat are both promisingand very prominent in the literature.These tend to fall into three camps: the interbreedingapproach, the ecological approach, and the cladisticapproach. Which, if any, of these approaches succeeds is immaterialto my project.I only wish to show that each of these approaches,and any which incorporatesone or more of them, such as any typical form of speciespluralism(e.g., Dupre 1981;Kitcher 1984a, 1984b, 1987;Mishler and Brandon 1987;Ereshefsky1992a;Stanford 1995) sacrificesthe essentialityof species membership. Consider, first, the interbreedingapproach. This is most famously exemplifiedin Mayr's biological species concept (BSC). According to the BSC, species are "groupsof interbreedingnaturalpopulationsthat are reproductivelyisolated from other such groups" (Mayr 1969, 26; 1970, 12; 1976, 518). The BSC is closely related to Paterson's (1985) recognitionspeciesconcept, on which a speciesis a group of organisms with a shared mate recognition system. According to the second, ecological approach,endorsedby Van Valen (1976) and Andersson(1990), a species is characterizedas a lineage with a unique adaptive zone, or ecological niche. Both the interbreedingapproach and the ecological approachseem to distinguishspecies on the basis of featuresthat are only contingently possessed by organisms. Hence, species membershipor non-membership would not seem to be essential if either account is right. Consider a large population of organisms,from which a small population splinters off and takes up a new ecological niche, adaptingto a new way of life. Organisms of the two branches cease to be recognizable to one another as mates, and thus become reproductivelyisolated. Both of the above approacheswould consider the two branchesto be distinct species. Organismsof the smallerbranchdo not belong to species"A,"
102
JOSEPH LAPORTE
whose membersconstitutethe largerbranch.Yet this could be a plainly contingent matter. Had the membersof the little branchnot taken on a new niche, or had there not been reproductiveisolation (or failureof mate recognition) between them and the members of A, they would belong to species A, given the species concepts in question. And surely it is possible that membersof the side branchshould have remainedin the original niche. Taking on a new niche might have involved such alterationsas a change in diet or in predators.But clearlythe members of an isolate like the above are not essentially such that their lineage has the diet or predatorsit has. Likewise, it is certainly possible that the side branch should have failed to become reproductivelyisolated from members of the main branch. Reproductiveisolation between lineages need not be essential to organismsin those lineages.Considerthe reproductiveisolation that obtains between two lineages of floweringplants on account of insect pollinators' preventingcrossing between the lineages. It is hardly essential to any individual plant that its lineage is not crossed with another lineage because of local insects' behavior. Likewiseit is possible that organismsof the two branchesshould have continuedto recognize one another as mates. Being such that one recognizesthese as opposed to those as mates would not seem to be an essential property. For example, many frogs use distinct calls to recognize mates. But it is unlikely indeed that any frog should essentially have the call it has, especially since the acoustic propertiesof its call depend on variables like the temperatureof the surroundingenvironment.Since organisms in the isolate of our above example might not have taken on a new niche/becomereproductivelyisolated/ceasedto recognizemembersof the stock branch as mates, its membersmight have been membersof the original species A, according to the ecological and interbreeding accounts. In the same way organismsof the smallerbranchconstitutea unique species, call it "B," but this is a contingent matter for the above accounts of species; things could have gone in such a way that these organisms were not members of B (being members of A instead). Hence, on the interbreedingand ecological accounts of species,organisms can be contingentlyincludedin or excludedfrom a species. Now let us turn to cladism. According to this growing school of systematics,founded by Willi Hennig, genealogy is all that counts for classification. On this view, an individual species (see Figure 1) is a lineage of organisms between two speciation events, or between one speciation event and one extinction event (see Ridley 1989 for further details). Now, cladism might seem at first to offer some hope to the essen-
103
ESSENTIAL MEMBERSHIP
tialist. If whetherone belongs to this species or that, depends on one's place in the family tree, then it might be a matter of necessity:afterall, Kripke has argued (1980) that organisms are essentially products of their genetic parents. One's place in one's lineage might, therefore,be essential to an organism,ratherthan contingent. In that case, granted cladism, one's place in one's own species might seem to be essential too, since it is, for cladism, entirely determined by one's place in a genealogicaltree. But alas! Cladism is bound to disappoint the essentialist.Observe that accordingto cladism,a speciesgoes extinctwheneverit sendsforth a new side species. This is so even if the lineage undergoesno change after sending out the side branch, so that earliermembersare indistinguishablefrom later ones. Thus, in figure 1, speciesA goes extinctupon the arrival of C, even though the lineage including A and B does not evolve at all. Various critics have charged that this is a defect of cladism, since whether a species, such as A, goes extinct depends on mattersentirely removed from the organismscomprisingit, namely, what happens to potential side branches like C. Kitcher (1989, 200-202) has brought these critics' point forcefully home by remarkingthat whether or not a species (like A) goes extinct at a particular time depends upon
t7
-
3
E3
t4
83
..........
82
t2
C
DC
A
Figure 1. Cladistic Species. From W. Hennig's Phylogenetic Systematics. Copyright 1966 by the Board of Trustees of the University of Illinois. Used with permission of the University of Illinois Press.
104
JOSEPHLAPORTE
whether a side isolate threateningto branch off at that time is wiped out by a cataclysmbefore achievingreproductiveisolation. I am not in the businesshere of criticizingcladism, so I will not join the discussion about whether the above implications undermine it; rather, my aim is to point out that, be they a credit or a liability for cladism, those implications do set cladism at odds with essentialism. Members of some new species, such as B above, are not essentially membersof that species. They would be membersof the ancestralspecies instead, if only a side branch (C, in this case) had never attained reproductiveisolation, perhaps having been destroyed by some cataclysm first. Whethermembers of B belong to one species or the other is dependenton a contingent event entirelyexternalto those members. It is clear, then, that if the cladistic account of species is good, placement in a particularspecies is not essential to an individual.3 Indeed, it is doubtful that any account according to which species are historicalentities (lineages),resultsin organismsessentiallybelonging to their respectivespecies.This is best shown by way of an example from Kitcher (1989, 202ff.), though Kitcher uses the example to make a very differentpoint than the one I wish to make, one that has nothing to do with essentialism. Kitcher imagines (see Figure 2) an evolving population, which divides at t into equal halves. By t' the brancheshave divergedsufficiently to constitute separatespecies, whateverthe criterionused to determine that: reproductiveisolation, or somethingelse. The divergencestops at t'. At no time is the distance between the ancestrallineage and either branch sufficientfor speciation;speciation occurs only because of the distance attained between the two branches. One is not sure what to say in such a case. S1 and S2 are stages of differentspecies. It might be said that Soand SI are conspecific,as are So and S2, though this is odd, given that S, and S2 are not conspecific.
Perhapswe can live with the oddity of Sobelonging to two species.But if so, then it cannot be said that membershipin any given speciesis an essential property. Members of So might not have belonged to both species they are members of. They might have belonged only to one species. They wouldhave, had one branch been wiped out before speciation could occur. Another option is to declare the ancestralspecies to be extinct at t. This, too, ill accordswith essentialism.Membersof each branchwould 3. It should be noted that not all of Hennig'sfollowers endorsehis claim that a species must go extinct when it splits; nevertheless,less strict cladistic concepts are subjectto argumentssimilarto the argumentsgiven above (see Kitcher 1989, 200-202, on Wiley 1981), as well as the more generalargumentgiven below.
ESSENTIAL MEMBERSHIP
105
ttime
s 2-
?'
t
so
Figure 2. Speciation on a historical account. From P. Kitcher's "Some Puzzles About Species", in M. Ruse (ed.), What the Philosophy of Biology Is: Essays Dedicated to David Hull. Copyright 1989 by Kluwer Academic Publishers. Used with permission of Kluwer Academic Publishers.
not necessarilybelong to a species other than the ancestralspecies. A cataclysm could have made things otherwise: for had either branch been preventedfrom evolving until t', there would not have been sufficient distance between the branchesfor speciation to have occurred. Kitcher'sthird solution involving pluralismmay be the best one available, but clearly it provides no lifeline for essentialism.It looks as if the matter cannot be resolved in a way that allows membershipin the species to be essential.
106
JOSEPHLAPORTE
Essentialismis, then, in some trouble. The three prominentspecies concepts discussedabove, the biological, ecological, and cladisticconcepts, are all hostile toward it. Of course, I have only consideredlandmark biological views about how species are to be delimited.Though these views are ill-disposed toward essentialism, there is always the possibility that a friendlierview will come into prominenceand outcompete the others. But, of course, if essentialismis forced to bank upon this hope, it is undergirdedby little support indeed. Further,any friendlierview that might come along would apparently have to be non-historical;4if the above considerationsareright,any historical speciesconcept will allow for membersto contingentlybelong to a species.But at present,the standardviews, at least, are historical.(Indeed, accordingto one author(Ereshefsky1992a,687),allcurrentlyproposed monistic views are historical, though this is probably an overstatement.) For the above reasons, essentialism finds itself to be embarrassinglyat odds with mainstreamcurrentsystematics. 4. Species are apparentlynot natural kinds whose members are necessarilymembersof them. But, it might be wondered,how about higher taxa? Could individualmammalsbe such that they are mammalsin all possible worlds?The same might be wondered for birds, oaks, and so on. Perhaps the essentialist model works for this vast and important class of taxa, even if it fails for species. But, in fact, an examinationof higher taxonomic placementreveals that here, too, essentialism is hard pressed for a toehold. There are three main competing modern schools of classification:phenetic taxonomy, evolutionarytaxonomy, and cladism. Again, my aim is not to argue for any one school, nor to try to bring out the various strengths and weaknessesof the schools: rather,my aim is to show that none of the schools is consonant with TEM, and hence to show that TEM is out of touch with biological systematics. Take, first, phenetic taxonomy, whose outstanding proponents are Sneath and Sokal (1973). This method of taxonomy eschews consideration of evolutionary descent, and other non-observablecriteria,in classifyingthe biological world. It takes into account the variouscharacteristicsobservedto belong to organisms,and gives these characteristics equal weight. Then, it makes use of computersto providegroupings by overall similarity. But if our taxa are defined phenetically, essentialism about kind 4. As a pluralist,Kitcher (1984b) arguesthat non-historicalconcepts should take their place beside coexistinghistoricalones. But the essentialistis committedto the exclusive reign of the non-historicalconcept. If Kitcher'spluralisticview succeeds,thenmembers of at least some species (the historicalones) would be so contingently.
ESSENTIAL MEMBERSHIP
107
membershiploses. For the pheneticistprides herself on not giving any special weight to characterstraditionally viewed as essential. So the various charactersconsidered,such as presenceof hair, hair color, hair length, skin color, body length, arm length, ear size, immunoglobulin concentration, etc., are often what the essentialistunhesitatinglycalls accidental.No one supposesthat an individual,Socratesto take an old standby, is essentiallysnub-nosed, or hirsute, or small-eared,or of his particularcolor, or height, etc. Yet just these sorts of characteristics place, by phenetic principles,a population of organismsinto one taxonomic camp (of any rank) ratherthan another. So pheneticismdoes little for the essentialist. When discussingpheneticism,it is naturalto discussa school known as pattern(or transformed)cladismin the same breath.Patterncladists have been called peculiarphenetic taxonomists (Ridley 1986, 91), and this subsumptionis apt for the purposes of the present investigation. While there are some important differencesbetween pattern cladism and standard pheneticism, they cross with essentialismfor much the same reason. Pattern cladism, like cladism proper, classifiesorganismsaccording to shared derived characters.But with ordinary cladism (hereinjust called 'cladism' without the qualifying adjective), derived characters are understood to indicate phylogeny, and are used as clues to classify by phylogeny. Pattern cladists believe the evolutionaryframeworkto be otiose in systematics.Therefore,while they use cladistictechniques, they do so without assumingtheir classificationsto be phylogenetic. Now, pattern cladists classify by characters,just as ordinarypheneticists do. But the former redefine 'character'in order to allow a characterto define a group even after it has been modified or lost. In that way, patterncladismsharplydifferentiatesbetweenabsentand lost attributes.So for the patterncladist, possession of fins can defineVertebrata,even though in many vertebrates,the fins of theirfish ancestors have been modified or lost (Nelson and Platnick, 1981, 328; Platnick, 1982). The relationship between essentialism and pattern cladism is not quite as simple as that between essentialismand straightforwardpheneticism. For let us suppose, plausiblyenough, that the patterncladist could definethe tetrapodsas having four limbs (in virtue of a common four limbed ancestor). Some (e.g., Beatty 1982, 31) have maintained that in that case, pattern cladists are committed to holding that tetrapods could cease to be tetrapodsby losing their limbs. But this is mistaken: for pattern cladism, the creatureswould still be "recognizable as having the original plesiomorphic character (state)" (Nelson and Platnick 1981).Thus, the patterncladistcan count snakesas tetrapods. Nevertheless,pattern cladism is at odds with essentialism.For it is
108
JOSEPH LAPORTE
apparentthat the derivedcharacteror charactersdefininga taxon need not be essentially possessed by individual organisms. This is not because the lineage containing the organisms might have lost the character(s), but because the organisms comprising it might never have possessed the character(s)to begin with: hence, they might never have comprisedtheir unique taxon. The derived charactersthat define taxa need not be, and in many cases would seem patently not to be, essential to organisms. Among such characters might be included possession of fins, possession of limbs, having shortenedteeth, having internalremnantsof limb bones (whales),or eye-resemblingorgans (cave fish), and so on. None of these attributeswould seem to be essential to one who possesses them. And it may be just one of these inessential traits, or some conjunction of them, that defines a taxon for pattern cladism. Thus, for patterncladism, the members of particulartaxa might not have been. It looks as if pheneticism, whether standard or pattern-cladisticstyle, disagrees with essentialism. So phenetic classificationdoes little for the essentialist.Can evolutionary taxonomy do better?It seems unlikely. Evolutionary taxonomy, championed by Mayr, Linsley, and Usinger (1953), Simpson (1961), and Mayr (1969), uses both phenetic and phylogenetic information in classification. That is, evolutionary taxonomy considers the various observable characteristicsof organisms as it organizes them into groups, but it also considers evolutionary ancestry. In the course of using phenetic principles,the evolutionarytaxonomist inheritsthose featuresof pheneticismthat make the latterincompatible with TEM. So long as phylogenyis respectedeitherway a given split between taxa is made, evolutionary taxonomy can, like pheneticism, allow individuals existing simultaneously,or at different times, to fall into separatecamps on the basis of plainly accidentalcharacters. A simple example will suffice to illustrate the point. Birds originally evolved from the reptiles. But, of course, evolutionary change is not instant. The firstpioneers along the line leading to birdswereprobably reptilian in all respects, though perhaps with feather-likescales. The evolutionary taxonomist must, as Mayr (1976, 456-457) confesses, draw a line, on the basis of phenetic principles, somewherebetween present-daybirds, with all of their avian characters,and the ancestors that first displayed one of these characters.So at some point along the line, some newly acquired character (flight, perhaps) will distinguish the first birds from the birds' last ancestors. But surely there is no guarantee, nor even a likelihood, that the characteristicthat separatesthe one taxon from the other is essential
ESSENTIAL MEMBERSHIP
109
to those who possess it. Supposeit is flight,for example.No essentialist will want to say that the very identityof an individualrestson its ability to fly. Whetheran individualtakes to the air is not essentialto its being that very individual.But in that case the problem for the essentialistis plain: the last non-bird ancestorsof the birds were not essentiallynonflyers, and thus they were not essentially non-birds. And the earliest birds, though they flew, might not have flown, and hence they might not have been birds.SEvolutionarytaxonomy does not yield taxa that are such that nothing can belong to them accidentally. Perhapsas the first two taxonomic methods have failed due to phenetic factors, the remainingmethod, cladism, which is entirelyfree of pheneticism,will count taxa as being delimitedin such a way that organisms will belong or fail to belong to these taxa essentially.Cladists hold that evolutionary branchingconstitutes the origin of new higher taxa, just as it does in the case of species. We have seen the cladisticspeciesconcept disappointthe essentialist. Cladism applied to higher level taxonomy will do likewise. Cladism recognizesonly taxa that are "monophyletic,"which is to say (for cladists; others use the term differently)taxa that include all descendants of a common ancestral species. For cladism, a "stem" species, which gives rise to differentsuccessorspecies that togethercomprisea single taxon, is itself part of the taxon (Hennig 1966, 71-72). For example, some particularspeciesof Archaeopteryxgave rise to the many existing birds. Hennig (ibid.) counts that speciesas the stem speciesof the birds, and hence counts it itself as a species of bird. But if cladism is right, it is a contingent matter that any individual member of the relevantArchaeopteryxspecies is a bird (see Figure 3). For consider:it could have happened that the Archaeopteryxspecies also budded forth a non-bird taxon at any time before the first bird branchwas sprouted. Suppose, for example, that this non-birdbranch should have led to a taxon having, say, the same numberof organisms, variation,and rank as the Aves, but that was comprisedof salamanderlike creatures.In that case, those organismsthat exist prior to the nonbird offshoot (t, in the figure),and that, as things actuallyare,comprise the stem for the Aves, would fail to count as birds. Otherwisethe Aves taxon would not include all ancestors of the stem (since it would not 5. This example assumes that evolution proceeds graduallyenough that the dividing line between ancestraland derivedtaxon marks no stark, uniformcontrast.Even proponents of the theory of punctuatedequilibriagrantthis, since they do typicallyaccept some instances of Darwiniangradualevolution (e.g., Gould and Eldredge 1977, 125129), and since, furthermore,even by the rapid evolution they defend, speciation requires thousands of generationsto occur:it is rapid only comparedwith the longevity of species (Eldredge1985, 128;Jones [1981] 1982, 168).
110
JOSEPH LAPORTE
time
\
\^
B\ irds, given the exist-
enceof non-birdtaxonN
Birds,inthe absenceofN
l/ stem
.- -"'"
N
'
Figure 3. A hypothetical reconstruction of the Aves taxon up to a time t' as it is, for cladism, and as it might have been.
include the non-bird taxon that branchesoff), and this would violate the cladistic restrictionthat all taxa be monophyletic. So a contingent event's occurrence or non-occurrencedeterminesmembershipin the Aves: whether the side branch does or does not bud off decides the boundariesof the taxon. Indeed, it is a contingent matter even whetherany of the organisms that could belong to a taxon do so, if cladism is right. Again, this is because of the cladists' insistencethat taxa include all ancestorsof the stem. This restriction bars the cladist from recognizing the fish: the would-be fish do not share a common ancestral species exclusively; tetrapods, such as elephants, are also descendantsof the closest common ancestor of the would-be fish. If cladism holds the key to taxonomy, then the fish do not comprise a scientificallyrecognizednatural kind. Their superficialsimilaritiessimply misled us into believingthem to comprise a natural group. Accordingly, individual salmon, lungfishes, coelacanths, etc., fail to belong to a common, exclusive taxon. The taxon they were thought to comprise, Pisces, turns out to be non-existent. Nevertheless,this is quite an accident. It couldhave turned out that these creaturesdid belong to the fish taxon: it all hinges on the coming into existence of non-fish descendants of the fish (or would-be fish), such as the elephant. Suppose the fish nevergave rise to such creatures. No branchinglineage leading to non-fish ever got started, or if it did,
ESSENTIAL MEMBERSHIP
111
it was quickly terminated.This could have happened.And had it happened, individual salmon, lungfishes, coelacanths, and so on would have belonged to the fish taxon, which would have existed according to cladisticprinciples.Therefore,it is a contingentmatterwhetherany individualsalmon or lungfishbelongs to a fish taxon. As it happens,it belongs to no such taxon. But it could have. 5. In conclusion, the essentialistposition summarizedin TEM (or perhaps TEM') seems wrongheaded.It is tempting to suppose that nothing could belong to any species but its own, and that likewise a thing essentiallybelongs to its own family, and so on. Despite the naturalness of this supposition, it does not measure up to currentbiological systematics. Biological taxa are not distinguishedby essentialproperties. If metaphysicsis to keep pace with biology, essentialismwill have to be jettisoned. REFERENCES Andersson, L. (1990), "The Driving Force: Species Concepts and Ecology", Taxon 39: 375-382. Aune, B. (1994), "Determinate Meaning and Analytic Truth," in G. Debrock and M. Hulswit (ed.), Living Doubt. Netherlands: Kluwer Academic Publishers; pp. 55-65. Ballie, J. (1990), "Identity, Survival, and Sortal Concepts", The Philosophical Quarterly40: 183-194. Beatty, J. (1982), "Classes and Cladists", Systematic Zoology 31: 25-34. Carter, W. R. (1986), "Mapping Semantic Paths: Is Essentialism Relevant?", in P. French, T. E. Uehling, and H. K. Wettstein (ed.), Midwest Studies in Philosophy XI. Minneapolis: University of Minnesota Press, pp. 53-73. Cocchiarella, N. (1976), "On the Logic of Natural Kinds", Philosophy of Science 43: 202-222. Cook, M. (1980), "If'Cat' is a Rigid Designator, What Does it Designate?", Philosophical Studies 37: 61-64. Doepke, F. (1992), "Identity and Natural Kinds", The Philosophical Quarterly42: 89-94. Dupr6, J. (1981), "Natural Kinds and Biological Taxa", The Philosophical Review 90: 66-90. . (1989), "Wilkerson on Natural Kinds", Philosophy 64: 248-251. Eldredge, N. (1985), UnfinishedSynthesis. Biological Hierarchies and Modern Evolutionary Thought. Oxford: Oxford University Press. Ereshefsky, M. (1991), "Species, Higher Taxa, and the Units of Evolution", Philosophy of Science 58: 84-101. . (1992a), "Eliminative Pluralism", Philosophy of Science 59: 671-690. . (1992b), The Units of Evolution: Essays on the Nature of Species. Cambridge, MA: MIT Press. Forbes, G. (1985), The Metaphysics of Modality. New York: Oxford University Press. Ghiselin, M. (1974), "A Radical Solution to the Species Problem", Systematic Zoology 23: 536-544. Gould, S. J. and N. Eldredge (1977), "Punctuated Equilibria: the Tempo and Mode of Evolution Reconsidered", Paleobiology 3: 115-151. Hennig, W. (1966), Phylogenetic Systematics. Urbana, IL: University of Illinois Press. Hull, D. (1976), "Are Species Really Individuals?", Systematic Zoology 25: 174-191. . (1978), "A Matter of Individuality", Philosophy of Science 45: 335-360. . (1980), "Individuality and Selection", Annual Review of Ecology and Systematics 11: 311-332.
112
JOSEPH LAPORTE
. (1987), "Genealogical Actors in Ecological Roles", Biology and Philosophy 2: 168-184. Jones, J. S. ([1981] 1982), "An Uncensored Page of Fossil History", in J. M. Smith (ed.), Evolution Now. A CenturyAfter Darwin. San Francisco: W. H. Freeman and Co., pp. 167-170. Khalidi, M. A. (1993), "Carving Nature at the Joints", Philosophy of Science 60: 100-113. Kitcher, P. (1984a), "Against the Monism of the Moment: A Reply to Elliott Sober", Philosophy of Science 51: 616-630. . (1984b), "Species", Philosophy of Science 51: 308-333. . (1987), "Ghostly Whispers: Mayr, Ghiselin and the 'Philosophers' on the Ontological Status of Species", Biology and Philosophy 2: 184-192. . (1989), "Some Puzzles About Species", in M. Ruse (ed.), What the Philosophy of Biology Is: Essays Dedicated to David Hull. Boston: Kluwer Academic Publishers, pp. 183-208. Kripke, S. (1980), Naming and Necessity. Cambridge, MA: Harvard University Press. LaPorte, J. (1996), "Chemical Kind Term Reference and the Discovery of Essence", Nous 30: 112-132. Mayr, E. (1969), Principles of Systematic Zoology. New York: McGraw-Hill. - . (1970), Populations, Species, and Evolution: An Abridgmentof Animal Species and Evolution. Cambridge, MA: Harvard University Press. . (1976), Evolution and the Diversity of Life: Selected Essays. Cambridge, MA: Harvard University Press. Mayr, E., E. G. Linsley, and R. L. Usinger (1953), Methods and Principles of Systematic Zoology. New York: McGraw-Hill. Mellor, D. H. (1977), "Natural Kinds", British Journalfor the Philosophy of Science 28: 299-312. Mishler, B. and R. Brandon (1987), "Individuality, Pluralism, and the Phylogenetic Species Concept", Biology and Philosophy 2: 397-414. Nelson, G. and N. Platnick (1981), Systematics and Biogeography: Cladistics and Vicariance. New York: Columbia University Press. Paterson, H. (1985), "The Recognition Concept of Species", in E. Vrba (ed.), Species and Speciation. Pretoria: Transvaal Museum, pp. 21-29. Platnick, N. (1982), "Defining Characters and Evolutionary Groups", Systematic Zoology 31: 282-4. Ridley, M. (1986), Evolution and Classification: The Reformationof Cladism. London: Longman. . (1989), "The Cladistic Solution to the Species Problem", Biology and Philosophy 4: 1-16. Simpson, G.G. (1961), Principles of Animal Taxonomy. New York: Columbia University Press. Sneath, P. and R. Sokal (1973), Numerical Taxonomy: The Principles and Practice of Numerical Classification. San Francisco: W. H. Freeman and Co. Stanford, P. K. (1995), "For Pluralism and Against Realism About Species", Philosophy of Science 62: 70-91. Van Valen, L. (1976), "Ecological Species, Multispecies, and Oaks", Taxon 25: 233-239. Wiley, E. 0. (1981), Phylogenetics. New York: Wiley. Wiggins, D. (1980), Sameness and Substance. Cambridge, MA: Harvard University Press. Zemach, E. M. (1976), "Putnam's Theory on the Reference of Substance Terms", The Journal of Philosophy 73: 116-127.