Aristotle on Genera, Species, and "the More and the Less" JAMES G. LENNOX Department o f History and Philosophy o f Science University o f Pittsburgh l~'ttsburgh, Pennsylvania 15260

Aristotle is taken, by many of the most distinguished evolutionary biologists, to be the fountainhead of a typological theory of species that is absolutely inconsistent with evolutionary biology. The foUowing remark of Ernst Mayr is typical: "The concepts of unchanging essences and of complete discontinuities between every eidos (type) and all others make genuine evolutionary thinking well-nigh impossible. I agree with those who claim that the essentialist philosophies of Plato and Aristotle are incompatible with evolutionary thinking." 1 D'Arcy Thompson, on the other hand, in On Growth and F o r m , has argued that the idea of using quantitative techniques to help understand morphological relationships took root in his mind during his work on Aristotle's biology: "Our inquiry lies, in short, just within the limits which Aristotle laid down when, in defining a genus, he showed that (apart from those superficial characters, such as colour, which he called 'accidents') the essential differences between one 'species' and another are merely differences of proportion, or relative magnitude, or as he phrased it, of 'excess and defect.'" 2 A theory that holds that species of a genus differ only in the relative magnitudes of their structures is very different, and is usually taken to be incompatible with, a theory that claims "complete discontinuities between every eidos (type) and all others." Can Aristotle have held 1. Ernst Mayr, Populaffons, Species, and Evolution (Cambridge, Mass.: Harvard University Press, 1963), p. 4. Similarly Theodosius Dobzhanskycomments in Genetics o f the Evolutionary Process (New York: Columbia University Press, 1970), p. 351: "Platonic and Aristotelian doctrines of archetypes or essences have strongly influenced medieval as well as some modern philosophies"; and G. G. Simpson says in Principles o f Animal Taxonomy (New York: Columbia University Press, 1961), p. 46: "Typology stems from Plato and his sources and came into taxonomy along with Aristotelian, Neo-Platonic,Scholastic and Thomist philosophy and logic." Cf. David Hull, "The Metaphysics of Evolution", Brit. J. Hist. Sci., 3, no. 12 (1967), 309-337; and M. T. Ghiselin, The Triumph o f the Darwinian Method (Berkeley: University of California Press, 1969), pp. 50-52. 2. D'Arcy W. Thompson, On Growth andForm, abridged ed. by J. T. Bonnet (Cambridge: Cambridge University Press, 1971), p. 274. Journal o f the History o f Biology, vol. 13, no. 2 (Fall 1980), pp. 321- 346. 0022-5010/8010132-0321 $02.60. Copyright © 1980 by D. Reidel Publishing Co., Dordrecht, Holland, and Boston, U.S.A.

JAMES G. LENNOX both these views? I will argue that he can, and did. The crucial conceptual move he had to make was to apply his technical notions of "excess and defect" or "the more and less" to biological differentiae, that is, to see biological differentiae as ontologicaUy on par with qualities varying along a c o n t i n u u m (e.g., temperature, color, tone, or texture). Such a move would conflict with the sort of typology traditionally ascribed to Aristotle by biologists and philosophers. After showing a possible route Aristotle might have taken 3 in arriving at this application of the concept of "the more and the less" to biological differentiae, I will indicate a means of preserving the "reality" of species that relies, n o t on fixed and eternal "essences," but on the teleological requirements of the life of each species, the individuals of which are systems of interacting, functioning organs adapted to a specific environment.4 3. Space limitations preclude my here entering into the debate about the chronological relationship between the Metaphysics and the various biological treatises. I disagree with the dominant view of D'Arcy W. Thompson's preface to his translation of the Historia Animalium for Oxford; H. D. P. Lee, "Place Names and the Date of Aristotle's Biological Works", Class Quart. 42 (1948), 61-67; John Herman Randall, Jr., Aristotle (New York: Columbia University Press, 1960), chap. 11; H. D. Hantz, TheBiologicalMotivation in Aristotle (New York: published privately, 1939); and Marjorie Grene, A Portrait o f Aristotle (Chicago: University of Chicago Press, 1963), Bibliographical Note, pp. 254-255; all of whom maintain that the most mature aspects of Aristotle's Metaphysics must be based on his biological research. My own position (which can be found in my unpublished dissertation, "The Interaction between Aristotle's Metaphysics and Hist Biological Works, Univeristy of Toronto, 1978, chap. 1) agrees with Geoffrey Lloyd's assessment that while "the particular answers he gave to those questions show that he paid special attention to the conditions which apply in the natural sphere, these major [metaphysical] doctrines do not presuppose his detailed investigations in natural science, but provide, rather, the framework within which they were carried out" (Aristotle: The Growth and Structure o f his Thought [Cambridge: Cambridge University Press, 1968], p. 91); see also W. D. Ross's Comments in his Introduction to the Scribner's edition of Selections from Aristotle's work (New York, 1927), pp. ix-x. Most dating procedures implicitly employ a question-begging evaluative premise about what is "most important," or "most mature," or "most Aristotelian" in the corpus. 4. I have followed the Oxford English translations of the Greek texts with these exceptions: for De Partibus Animalium I, I have used the Balme translation (Aristotle's De Partibus Animalium I and De Generatione Animalium I) [Oxford: Clarendon Press, 1972]; for De Generatione Animalium and Historia Animalium~ A. L. Peck's translations for the Loeb Classical Library (Cambridge, University Press, 1963, 1965). Where I believe the translations are questionable, I have included the Greek, or placed angled brackets around the English expressions not mirrored in the Greek. In the notes I have used the following 322

Aristotle on Genera, Species, and "the More and the Less" Being of a certain species and being of a certain genus are two ways in which numerically distinct entities can be one, according to Aristotle. Thus Aristotle's m a n y discussions of u n i t y s in his works lead into the issue of how species of a genus are differentiated and related. An individual entity is numerically one due to its being either naturally (a tree or a cat), artificaUy (a statue or table), or accidentally (a heap of sand or a pool of water) continuous. 6 Two or more entities that are numerically separate may, nonetheless, be one in form (gv guSe0. By this Aristotle most characteristically means that each of them is structurally and functionally identical, though the organization and function are embodied in discontinuous segments of material. 7 Organisms of different species may yet be one generically (gv 7~ve0. Aristotle offers two different accounts of this unity. In the Metaphysics the genus is often described as the matter or substratum for differentiation into members of its species. In Parts o f Animals and History o f Animals the members of different species of the same genus are, in addition, said to differ by "'excess and defect" or "the more and the less."8 (For completeness, it should be noted that organisms in possession of functionally identical structures are said to be one by analogy -- Kat" dvako"fiav).9

abbreviated forms for Aristotle's works: Cat. (Categories), De An. (De Anima), De Incessu (De Incessu Animalium), DPA (De Partibus Animalium), DGA (De Generatione Anirnalium), HA (Historia Animalium), Met. (Metaphysica), Pol. (Politicus). The pagination is that of the Bekker edition of the Opera. DK refers to H. Diels and W. Kranz, Die Fragmente der Vorsokratiker, 6th ed. (Berlin 1951). 5. On distinguishing between the various sortsof unity, see Met. V, 6,1016b32 ft.;Met. V, 9, 1018a7 ft.;Met. X, 1, 1052a15 ff.;DPA 1,4, 644a15-23, 644bl16;HA I, 1,486a15;HA I, 2,488b31-33. 6. Met. X, 1, 1052a15-30. There is in addition a ranking of numerical unity based on the assumption that inherent unity is more deserving of the title than that imposed by an external agency. In this ranking natural entities are, of course, one in a superior sense to artificial productions; and unit-heaps - one only by accident of place - are even further down the scale of numerical identity. 7. Met. VII, 8, 1034al-7. 8. The expressions rb #(~hhov ~a'~~rTov and [rlrepox~ Kd~ ~hhtt~¢¢ are generally interchangeable in Aristotle, though on some occasions one or the other seems to possess greater generality in Aristotle's mind. At Physics I, 4, 187a 15-20, and Met. VIII, 2, 1042b32-36, the former expression seems to be the more general, while the opening lines of DPA IV, 12, suggest that the latter expression is so. 9. Two examples of analogues relationships mentioned by Aristotle are, lungs: air-dwellers :: gills : water-dwellers (DPA I, 5, 645b8); bone : land dwellers :: spine or cartilage : water-dwellers (DPA II, 8, 653b32 ff. and Post. An. II, 14, 323

JAMES G. LENNOX N o w h e r e outside the biology does Aristotle use the c o n c e p t s o f " t h e m o r e and the less" or "excess and d e f i c i e n c y " to express the nature o f the relationships b e t w e e n species o f a genus - t h o u g h these c o n c e p t s play a central role in his understanding o f perceptible qualities and qualitative change. 1° In Parts o f Animals and History o f Animals we are apparently witnessing a n e w application o f these c o n c e p t s to explicate the relationship b e t w e e n the species o f a genus, n I will argue in w h a t follows t h a t the application o f the c o n c e p t s o f "excess and d e f e c t " or " m o r e and less" to differentiae is an e x t e n s i o n o f the "genus-as-matter/species-as-differentia" a c c o u n t o f the genusspecies relation f o u n d in the Metaphysics. I f the case can be m a d e , it has i m p o r t a n t c o n s e q u e n c e s for c o n t e m p o r a r y claims that a quantitative and e v o l u t i o n a r y a p p r o a c h to biological classification is i n c o m p a t i b l e w i t h a b e l i e f in essentialism. Cdnversely, I h o p e it will be clear h o w unfair are those biologists w h o see in Aristotle a w a t e r e d d o w n Platonic typologist. To Aristotle's u n q u e s t i o n e d status in the histories o f biological observation, I h o p e we can add his positive c o n t r i b u t i o n to the theoretical understanding o f " t h e species q u e s t i o n . "

98a20-23). Aristotle's concept of analogy is as close to contemporary biology's as it could be, given its nonevolutionary perspective, as the following quotation makes clear: "when the element of an ultimate ancestral structure in common is lacking . . . the structures developed for the function in question are not homologous but only analogous. Insect wings, as compared with the wings of flying vertebrates, are a good example" (G. G. Simpson, The Meaning of Evolution [New York: Bantam Books, 1971], p. 163). 10. I will discuss Metaphysics VIII, 2, later in this paper. 'Aristotle never explicitly applies what he says there to organisms (which I take it are paradigmatic Aristotelian substances), but he does imply that an "analogous" account could be given; and makes one suggestive comment on how organs are differentiated. 11. I should here indicate that the Aristotelian concepts of 3,~uo¢ and e~8o¢ do not coincide with our notions of genus and species. Most organisms now referred to as species would not be so designated by Aristotle, and most of his greatest 7~v'0 (Fish, Bird, Cetacea, Testacea, Crustacea, Cephalopods, Insects, Viviparous Quadrupeds, and Oviparous Quadrapeds) appear at roughly our level of classes. I will be using "genus" (as Aristotle tends to) to refer to such groups and "species" to refer to distinguishable types within these groupings. While I cannot argue the point here, I do not find quite the chaos in Aristotle's biological use of ~vo¢ and e~8o¢ that Balme does; see "F~vo¢ and E~o¢ in Aristotle's Biology," Class. Quart., xli (1962), 81 ff. The clue to sorting this issue out is Met. V, 28 (on ~/~vo¢), where Aristotle claims that one legitimate meaning of this term is in reference to genetically related series, not in reference to the underlying subject of differentiation. (1024b6-9). The reference class of the term so used would include "species." 324

Aristotle on Genera, Species, and "the More and the Less" This aspect o f Aristotle's thought has not been the focus o f much attention, but one man, D'Arcy W. Thompson, spent a great deal o f time thinking about it. In two separate discussions Thompson offers two very different accounts o f "'the more and the less," and looking briefly at Thompson's work will serve as a useful introduction to the issues that concern us here.12 In the later chapters of On Growth and Fo rm , Thompson is attempting a daring application o f transformation theory to the problem o f understanding the relationships between biological forms. He argues that the evolutionary and morphological relationships between different species may be better grasped by transcribing a representation o f a given species onto a Cartesian coordinate system, which may then be transformed according to pre-established equations. This would display certain species as systematic deformations of others, and indeed could be used to suggest hypotheses about evolutionary relationships and lineages. Thompson goes some way toward establishing the value o f the method, but the importance o f his work has only recently been grasped. Thompson credits Aristotle with first making progress toward understanding species relationships in this way. "It is precisely this difference of relative magnitudes, this Aristotelian 'excess and defect,' which our coordinate method is especially adapted to analyze, and to reveal and demonstrate as the main cause of what (again in the Aristotelian sense) we term 'specific' differences." 13 If this is Aristotle's doctrine, it would seem he is committed to differences between species being a matter o f quantitatively measurable degree, leaving open the possibility o f species "shading into" one another. 14 To those familiar with the realist essentialism o f Aristotle's 12. Thompson, On Growth and Form, abridged ed., 1971. This edition is an abridgment of Thompson's 1942 revision of the 1917 original, and thus contains a footnote (p. 274 n. 1) to "Excess and Defect: Or the Little More and the Little Less," Mind, 38 (1928), 43 ff. This does not alter the fact that the two accounts are essentially polar opposites. 13. Ibid., p. 274. That the relationships between qualifies which differ by "excess and defect" can be expressed quantitatively is dearly implied at De Sensu 3, 439b19-440a6, esp. 439b28-33. On his passage and its implications, see R. Sorabji, "Aristotle, Mathematics, and Colour,'" Class. Quart., 22 (1972), 293-308. 14. In backing up this claim one might be inclined to invoke Aristotle's various remarks about "dualizing" creatures (e.g., seals and bats) at DPA 681b5, 697a15 ff., DGA 731b8, etc.; see Stephen Clark, Aristotle's Man: Speculations upon Aristotelian Anthropology (Oxford: Oxford University Press, 1975), pp. 31-32. These claims are invariably about species which cannot be easily classified into 325

JAMES G. LENNOX

Metaphysics, this sounds highly un-Aristotelian. The account most characteristic o f that work is o f species radically demarcated b y contrary differentiae, is It is important at this point to take an initial l o o k at the texts in question in the History and Parts o f Animals, to verify Thompson's interpretation o f Aristotle's words.

Parts o f Animals I, 4,644a16-21 F o r all kinds that differ b y degree (~aO, Onepox¢w) and b y the more and the less have been linked under one genus, while all those that are analogous have been separated. I mean that bird differs from bird by the more or by degree (for one is longfeathered, another is s h o r t - f e a t h e r e d ) . . .

Parts o f Animals I, 4, 644b8-15 The genera have been marked o f f mainly b y the shapes (ox(l#aot) o f the parts and o f the whole body, wherever they bear a similarity, as the birds do when compared among themselves and the fishes, cephalopods and testaceans. F o r their parts differ not on the basis o f analogous likeness, as bone in man is to spine in fish, b u t rather by bodily affections such as largeness and smallness, softness and hardness, smoothness and roughness, and such - in general b y the more and the less.

History o f Animals I, 1,486a21-24 In other cases, they are, it is true, identical, but they differ in respect o f excess and defect: this applies to those whose genus is

one "greatest kind" or another, however. Aristotle does not normally refer to a "variety" as being difficult to classify as a member of one species rather than another. He does countenance some hybridization (see DGA II, 7, 746a29746bl 3), but (see DGA II, 4,738b26-36) claims the result, after an indeterminate number of generations, will be an offspring like the original female in form (rat& rb O(Tku¢¢7vpop~v). Thus neither of these ideas speaks directly to the issue of the relationship between species of a genus, though they are further evidence against viewing Aristotle as a rigid typologist. 15. E.g., Met. VII, 12, 1037b18-20, 1038a5-25, andMet. X, 9, and 10. There is at least one passage where Aristotle states quite clearly that members of different species do not differ by the more and the less, Pol. I, 5, 1259b36: "Nor can [the ruler and the ruled] differ by the more and the less, for ruler and ruled differ in form (e~o¢), but the more and the less do not." Cf. Physics IV, 9, 21765. 326

Aristotle on Genera, Species, and "the More and the Less" the same; and b y genus I mean, for example, bird and fish: each o f these exhibits difference with respect to genus, and o f course there are numerous species b o t h o f fishes and o f birds. According to this account o f the relationship between species o f a genus, there are a set o f general differentiae (KaO6~oo 6ta~opaL) 16 c o m m o n to all the individual birds or fishes. The differences between species are a matter o f roughly quantifiable differences in these common differentiae. The differences among species o f birds might involve, for example, different lengths o f feather, beak, or leg, differences in the hardness or softness o f the beak, in the oiliness o f the feather, or in the amount o f skin between the toes. 17 Very little is said about the central concept o f the metaphysical account o f form, the final, specific differentia peculiar to that one species alone. Yet it is hard to believe that these two accounts are n o t compatible, for in the previous chapter o f Parts o f Animals I, this very notion o f specifically different individuals possessing the proper and peculiar differentiae o f their species plays a central role. 18 It will, I think, be a lengthy task to show the compatibility o f these two accounts, a task we m a y begin b y looking at a second a t t e m p t by Thompson to come to grips with this issue. In a paper responding to A. E. Taylor's " F o r m s and Numbers: A Study in Platonic Metaphysics," 19 Thompson is primarily concerned to show that "the more and the less" plays a role in understanding the theory attributed to Plato by Aristotle, that "the number is generated from the one and the indefinite dyad. ''z° This argument need not trouble us, b u t 16. Here I follow Balme's translation of DPA I, 3, 624b25. 17. For specific applications of this notion of differentiation by the more and the less, I recommendDPA IV, 11 and 12, and my discussion below. 18. DPA I, 3, 643a1: "it is not possible for some form of being which is single and indivisible to belong to things different in form, but it will always be different"; 643a7-9: "But if the kinds of animals are indivisible, and so are the differentiae, and no differentia is common, then the differentiae will be equal to the indivisible kinds of animals." The stress throughout chaps.2-3 is on the unity of each species, its distinctness from all others, and on the one-to-one correspondence between complete differentiation and speciation. 19. A. E. Taylor, "Forms and Numbers: A Study in Platonic Metaphysics," Mind, 35 (1927), 419-440, and 36 (1927), 12-33. 20. Met. XIII, 7, 1081a15. Thompson argues that dpttglz6~ had a technical sense in Greek mathematics which it carries in this passage, referring to irrationals (p. 43). I can f'md no such meaning mentioned in Heath's Greek Mathematics (New York: Dover Books, 1963), and Thompson cites no texts to support this claim. At any rate, this is not at issue for the purposes of my discussion. 327

JAMES G. LENNOX Thompson concludes his highly speculative paper with an attempt to read the biological passages we have just been looking at in a highly Platonic mold. But now that we understand the phrase "excess and defect" better than ever before, we m a y read its technical meaning into the Aristotelian passage, and t~md the whole significance vastly altered and improved. F o r I take it now that Aristotle was thinking, m o r e Platonico, o f all the fowls o f the air as mere visible forms or e ~ , mere imperfect representations o f or approximations to, their p r o t o t y p e the ideal Bird . . . we may, as it were, survey the whole m o t l e y troop o f feathered things, only to fred each one o f them falling short o f perfection, deficient here, redundant there: all with their inevitable earthly flaws. Then, b e y o n d them all we begin to see dimly a bird such as never was on sea or land, with blemish, either o f excess or defect: it is the ideal Bird, the 7mp&SetTl~a b' ~u O O p ( w ~ ~v6x etraz. 21

However much we m a y admire the prose style o f this concluding passage, the conclusion itself will n o t do as an interpretation o f Aristotle. 22 Aristotle is quite clear throughout his works that the standard o f perfection for an o~aia is the form o f that oOam, a standard that is the normal and natural end o f biological development. 23 Aristotle's work is unified by a persistent attack on the notion o f real "universals" over and above the individuals that e m b o d y specific f o r m ) 4 Further, Aristotle's nature, as he so often reminds us, "does nothing by chance ''2s the biological world is notably characterized by the fact that species are n o t "deficient here, redundant there." That, for Aristotle, is the wonder o f the natural world. Aristotle consistently argues that there is no bird that "never was on sea or land."

21. Thompson, "Excess and Defect," p. 55. 22. This has not prevented translators such as A. L. Peck from referring to Thompson's paper as a standard discussion of the subject (Aristotle, Parts o f Animals [Cambridge, Mass.: Harvard University Press, 1968], Introduction, p. 20. n. b). 23. Met. V, 16, 1021bl5ff., Met. VIII, 4, 1044a33-1044b2; Met. IX, 8, 1050al5-19;DeAn. II, 1,412a20 ft. 24. Met. VII, 12, 1038a5, andMet. VII, 13, 1038b10 ft. 25. See De lncessu 704b15-18; DPA IV, 12, 694a15; and further see Name Hermann Bonitz, Index Aristotelicus, vol. 15 of Aristotelius Opera (Berlin: De Gruyter et Socias, 1961), 836b28-46. 328

Aristotle on Genera, Species, and "the More and the Less" Clearly, this will n o t do as an interpretation o f the characterization o f the relation Aristotle believes to hold among species o f a genus. I propose to l o o k to the rest o f the Aristotelian corpus for clues to the exact meaning o f the passages that concern us, to characterize what I see as the underlying metaphysics and embryology o f Aristotle's a t t e m p t to distinguish species o f a genus on the basis o f "more and less" differences, and finally to try to discover how he viewed this theory in relation to his realistic account o f form. The abstract concepts o f "excess and defect" or "the more and the less" are n o t found in any extant fragments o f the pre-Socratics, 26 But Aristotle notes in the Physics 27 that virtually all previous natural philosophy made use o f different "species" o f excess and defect as principles b y means o f which the landerlying material o f the cosmos was (or is) differentiated - for example, into the h o t and cold, the dark and light, the dense and rare, or the great and small. Aristotle's use o f these concepts in the Categories seems to be rooted in Plato's Philebus (23c-25a), where such variable qualities as h o t t e r and colder, stronger and weaker, are said to have within them "the more and the less" (~b #&kk6v ~at r~rrov). Plato, then, had already used this phrase to denote qualities that differ along a continuum. In the Categories, "the more and the less" plays a central role in explicating the categories o f o0tr/a (I shall read "substance"), quantity and quality. Of substance, Aristotle writes, "Substance, it seems, does not admit o f a more and a l e s s . . . F o r example, if the substance is a man, it will n o t be more a man or less a man either than itself or than another m a n . " 2s This passage states d e a r l y that the merdbers o f a species, qua species members, cannot differ b y the more and the less. This is, o f course, consistent with the Parts o f Animals, for it is members o f different species whose characteristics are there said to differ only in degree. 29 Nonetheless, if the members o f one species do n o t differ by 26. Empedocles' "painter" fragment (DK 31B, frag. 23) notes that the pigments are harmoniously mixed ~& I~eV 7three ~ h a 8"~hhttoow, but the cosmological significance of this passage is unclear. DGA II, 6, 743b20-25, may be a conscious echo of this fragment. Cf. Anaxagoras, DK 59, B6, for a use of "the great and small" as differentiae. 27. Physics 1,4,187a15-20;Physics 1,6, 189b8-10. 28. Cat. 3b33 ff. But see note 31 below. 29. The claim that is made here goes entirely unjustified: and on the basis of DGA IV, 3, one might want to raise the issue of the status of "monstrous creations", which seem to be treated as less fully human than perfect developments. 329

JAMES G. LENNOX the more and the less, can different species o f a genus do so? Can such an account be offered at the interspecies level without assuming that the individuals within a species also differ by the more and the less? One is inclined, even without considering the biological evidence, to say "No." 3o If "red" did not vary continuously and evenly along the color spectrum, it could not shade into "orange" in such a way that we could say "red" and "orange" differ only by the more and the less. Thus it may seem as if Aristotle has here a theory o f substance that is inconsistent with the passages already quoted from his biological works. A means o f escape is provided for Aristotle, if he is willing to argue that there is a radical difference between the essential differentiae o f substances and their "accidental" qualifications. 31 He might then argue that the latter attributes (e.g., the color, texture, or hardness o f a hawk's beak) are continuously distributed within a species, while the hawk's beak, qua beak, is identical in each species member and different in kind from the beak of any other species o f bird. As we shall see in a moment, the Categories and Physics suggest that at some stage in his career Aristotle contemplated such a move. The first indication that Aristotle wished to restrict the application of differences in degree to qualities is found a bit later in the Categories: "Qualifications admit o f a more and a less; for one thing is called more or less pale than another. ''32 The constraints on this claim are that "unless both [things] admit o f the definition o f what is under discussion, neither will be called more that than the other. ''33 To use his example, both things must be within a given range on a dark/pale continuum for the concept "pale" to apply, but the concept applies to a range o f degrees o f paleness, not to one visible property that is discontinuous with others in the same class. In the Categories, then, Aristotle restricts the concepts o f "excess and deficiency" and "the more and the less" to the description o f perceivable variations o f a given quality. Thus it is not surprising to f'md that, in the Physics, Aristotle characterizes changes of quality - i.e., alterations - as changes '%vithin the same genus, but with respect to the 30. The modern, evolutionary, answer is also no. What makes it possible for the eventual isolation of a new species to occur is a certain degree of constant variation (and variability) within the populations of ancestor species. See Dobzhansky, Genetics of the Evolutionary Process, pp. 30, 311-313. 31. In fact, in the Met. (VIII, 3, 1044a10 ff.), Aristotle allows that obcria ttez& ~ ¢ iJ~¢ will "!admit the more or less." 32. Cat. 10626-28. 33. Cat. l l a l l ff. 330

Aristotle on Genera, Species, and "the More and the Less" more and the less." 34 Earlier, in Physics IV (217b5), he had noted that variations on a density/rarity continuum do not require an atomistic explanation, "for it is not by leaving a gap (or "by being discontinuous" rob ~uz~elnew) that there is the less or more." Again, we see these concepts applied to variations along a continuum. It might be thought that a glimmer of the doctrine of the Parts of Animals I, is present here, for the more and the less are said to apply to changes "within the same genus." However, "genus" here clearly refers to "kinds of quality." Aristotle has just characterized alternations as changes of quality. Now he is explaining that alterations are changes along such continua as pale/dark or hot/cold (the genera being 'color', or 'temperature'). To support this analysis, consider the following passage from On the Senses: " I f then the objects which, though in different genera, are called corresponding (e.g., I call sweet and white corresponding, though different in genus) . . . , it would be even less possible to perceive these simultaneously than things belonging to the same class." as Thus there may be analogies between items on the color continuum and on the taste continuum, while items on the one or the other are in the same "genus." Nonetheless, while Aristotle's discussion is restricted to continuous variations within a sensory mode, the formal analogy between this model and that used in Parts of Animals for grasping relationships between species and genera is striking. In comparing members of different genera of sensory quality, we may note analogies - e.g., between sweetness and whiteness, a6 Within a genus, say color, we can distinguish species (red, blue, yellow), while noting that these represent localizable segments of a continuum, the differences between them being of "excess and defect." The realization of this analogy may have been crucial in Aristotle's decision to apply the theory of "more and less" to species differentiation. Aristotle goes on to make a distinction between two types of quality in the Physics, and seems to imply that the account he has given of "alteration" is only applicable to changes of "affective qualities": "Motion in respect of quality let us call alteration, a general designation that is used to include both contraries; and by quality I do not mean a -

34. PhysicsV, 2, 226b3-4. 35. De Sensu 5,448a13-16. 36. Cf. Aristotle's remarks about the ambiguity of such terms as "sharp," "flat," "clear," or "obscure," due to their having application to more than one variety of perception (Topics I, 15,106a9-36). 331

JAMES G. LENNOX p r o p e r t y o f substance (for a differentia, t o o , is called a q u a l i t y ) " . 37 This distinction is, o f course, a central plank in Aristotle's essentialist program. Entities m a y change their color, or t e m p e r a t u r e , or even size, w i t h o u t b e c o m i n g different entities. But there is a set o f " q u a l i t i e s " differentiae - t h a t are "in the obo/a o f a t h i n g , " and these constitute its " f o r m . " 38

Such a distinction between "essential" and "accidental" properties, combined with the explicit restriction of the notions of "excess and defect" and "the more and the less" to the "accidental" 0.e. nondifferentiating) side of the distinction,points in the direction of Aristotle the typologist. Scientific knowledge is of the differentiating(formal) properties of things. To take the line so far sketched is to deny the value of "the more and the less" in any science that studies natural substances - biology being the obvious case in point. Ill the works considered up to this point, it is clear that differences of "the more and the less" are restricted to what Aristotle most often calls bodily or sensible affection.39 A step toward applying this concept to differentiae is taken in Metaphysics VIII. In Chapter 2 of that book, Aristotle says that "there are as many sorts of being as there are ways of being differentiated,"a° and, indeed, that differentiation is the source and cause of a thing'sbeing what it is.41 Here he treats "sensible affection" as a type of differentiation: "and others [differ] by the affections proper to sensible things, e.g.,hardness and softness,density and rarity, dryness and wetness; and some things by some of these qualities, others by them all,and in general some by excess and some by defect.''42 Later he argues that "we must grasp then, the kinds of

37. Physics V, 2,226a26-29. 38. This distinction is made clearly in Aristotle's discussion of the concept of quality in chap. 14 of Met. V. 39. Aristotle has various expressions for such nonessential qualities, e.g., r& rog¢ aoa#art*ro[¢ *rdO~ (DPA I, 4,644b14), r& rCov alaO~r~v *rdO~ (Met. VIII, 2, 1042b22), r& ndO~ T~v a~op.d,cov (DPA II, 1, 646a20). DGA V is entirely devoted to discussing the variations within a species of these qualities, which are taken to be allowable variations of the differentiae of a given species. It is of importance, as we shall see, that Aristotle's primary criterion for distinguishing these qualities from those essential to a species is whether or not they are for the sake of anything; they are not distinguished on morphological or taxonomic grounds. 40. Met. VIII, 2, 1042b30-34. 41. Met. VIII, 2, 1043a3-8, 1042b33. 42. Met. VIII, 2, 1042b21-25. 332

Aristotle on Genera, Species, and "the More and the Less" differentiae (for these will be the principles of the being of things) e.g., the things characterized by the more and the less." 43 Throughout this chapter Aristotle quite consciously (and irritatingly) avoids using differentiae of substances as his examples. We hear that ice is water differentiated by freezing, that a threshold is wood differentiated by its position, or that dinner is food differentiated by the time it is served. He explains that none of these cases is an o0o~z, but the differentiation of an o0o~z is analogous to these examples (1043a5). There is, however, one remark about the differentation of a hand or foot in this chapter: "And the being of some things will be def'med by all these (qualities), because some (parts of them) are mixed, others are blended, others are bound together, others are solidified and others use the other differentiae; e.g. the hand and the foot (require such complex differentiae). ''a4 Organic parts, while not themselves substances, are the type of things that serve to differentiate one species from another. 4s Here they are said to differ from merely sensible affections in that they are organized by ~ complex set of differentiations. Each of these singularly gives rise to a difference of "excess or deficiency" within a specific continuum; but a biological structure is a unified sum of such activities. There is much more to be said about the genesis of such structures, and it will soon need to be said. For now it is crucial to observe Aristotle's methodological point: if organic structures are simply complexes of organized material, they must have texture, color, size, density, and so on. This opens the way for the application of "the more and the less" to differentiae, and to quantitative comparisons between the organs of two species of a genus. I wish now to consider one characteristic way in which Aristotle analyzes genus-species relationships in the Metaphysics, as providing the theoretical underpinnings for seeing species within a genus as a more or less continuous series. This theoretical underpinning has formal (not empirical) similarities to the modern systematicist's way of viewing the relationship, thus making Thompson's original insight deeper and more profound than he may have realized. In no text that I am aware of does Aristotle describe a "genus" as a class or set composed of N members. Rather, he most usually

43. Met. VIII, 2, 1042b32-34. 44. Met. VIII, 2, 1042b28-31. 45. Cf. Met. VII, 16, 1034a5-8;DPA I, 4,644b9. 333

JAMES G. LENNOX characterizes the genus as the substratum (lmoreOaevov) or matter (bT~r~) for differentiation.46 As Richard Rorty has pointed out, the primary motivation for the introduction o f this identification of a taxonomic with a metaphysical category seems to be Aristotle's concern for ensuring the material/ formal unity of his primary substances - and his refusal to be tempted by either idealism or materialism.47 However, I should like to raise a second (related) problem for which viewing genera as material for differentation is an answer - the problem of how we are able to identify members of a genus among individual species members. EMo¢ in Aristotle's ontology (notoriously) does double duty as the structural/functional aspect of individuals and as the basis of our Identifying these individuals as species members. Regardless of the difficulties of this duplicity,4s it allows no hint of arbitrariness in the process of identifying species members. But what of classification above the species level? a9 When Aristotle abandoned the viewpoint, expressed in the Categories, that both species

46. The central texts for the theory that genus is matter are Met. V, 6, 1016a27; Met. V, 28, 1024b8; Met. VII, 12, 1038a6-7;Met. VIII, 6, 1045a35; Met. X, 3, 1054b30;Met. X, 8, 1058a23-24; and DPA I, 3, 643a24. 47. The concept of genus as matter has been the subject of a recent exchange between Richard Rorty ("Genus as Matter: A Reading of Metaphysics Z-H," in Exegesis and Arguments, ed. E. N. Lee, A. P. D. Mo~elatos, R. M. Rorty, Phronesis supplementary vol. 1 [Assen: Vangorcum, 1973], pp. 393-420; "Matter as Goo: Comments on Grene's paper," Syntheses, 28 [ 1874], 71-77) and Marjorie Grene ("Is Genus to Species as Matter to Form? Aristotle and Taxonomy," Synthese, 28 [1974], 51-69). In my account I have attempted to retain the Rorty reading for natural substances (my primary concern) while widening the equation of genus and matter to include those counterexamples raised by Grene to Rorty's reading. 48. They are many. For a good summary see J. H. Lesher, "Aristotle on Form, Substance, and Universals: A Dilemma," Phronesis, 17 (1972), 169-179. 49. The debate goes on, of course. In recent years taxonomists and systematists have debated the "reality" of taxa above the species level, arguing that only at the species level is there an actual organizational bond (potential interbreeding) among individuals. See Mayr, Populations, Species, and Evolution, pp. 12-20; G. G. Simpson, Principles o f Animal Taxonomy (New York: Columbia University Press, 1961), chap. 6; Ghiselin, The Triumph o f the Darwinian Method, chap. 4, J. R. Gregg, The Language o f Taxonomy (New York: Columbia University Press, 1954). A useful survey of the issues is found in the papers by Mayr, Pratt, and Hull in Topics in the Philosophy o f Biology, ed. Marjofie Grene and Everett Mendelsohn,Boston Studies in the Philosophy o f Science, vol. 27 (Boston: Reidel, 1976). 334

Aristotle on Genera, Species, and "the More and the Less" and genera were "secondary substances," he both elevated e~8o¢ to primary substance and argued that the genus is no substance at all. s° Does this lead to the conclusion that classification above the species level is not based on an actual unity among certain species, but is a matter o f taxonomic convention? The "genus-as-matter" theory seems to be an attempt to explain the unity of individuals classified genericaUy. Aristotle introduces the doctrine in three contexts: (1) in his discussions o f levels of unity in Metaphysics V, 6 and 8, and X, 3; (2) in his discussion o f division as mirroring the differentiation o f a genus into species in Metaphysics VII, 12, and X, 8; and (3) in his analysis o f the meaning of the genus concept in Metaphysics V, 28. This last discussion is summarized as follows: "Genus, then, is used in all these ways, (1) in reference to the continuous generation o f the same kind, (2) in reference to the first mover which is o f the same kind as the things it moves, (3) as matter (d~¢ b%r/); for that to which the differentia or quality belongs is the substratum, which we call matter, sl When he picks up on this model in Metaphysics X, Aristotle is careful to specify which sense o f genus he intends: "and the genus is the matter o f that which is called the genus, not in the sense in which we speak o f the genus or family o f the Heracleidae, but in that in which the genus is an element in the thing's nature." s2 "Genus" does not represent an abstract category, but refers to a shared or c o m m o n aspect o f the nature o f individuals, serving as the substratum of differentiation. How is this position to be reconciled with Aristotle's anti-Hatonic claim that the genus "doesn't exist apart from the species qua species o f a genus"? s3 Interestingly, it is precisely here that Aristotle once more introduces the genus-as-matter model: the line just quoted goes on "or if it exists but only as matter (for the voice is genus and matter, but its differentiae make (~otobow) the species (e'er/), i.e., the letters, out o f it. ''s4 It is not absolutely clear whether Aristotle is discussing differentiation as a natural or logical process here - and elsewhere - but I will soon argue that it is both, 50. Cfi Met. VII, 13, 1038b10-18. 51. Met. V, 28, i024b6-9. The ftrst two meanings are the standard Greek usage from Homer on, and the phylogenetic content of this meaning is implicit in Aristotle's use of the term in his biology. This goes a long way toward explaining the apparently "rough and ready" use of -r~vo¢ and e~8o¢ (see note 11 above). The referents of meanings 1 and 2 will sometimes be species members. 52. Met. X, 8, 1058a23-25. 53. Met. VII, 12, 1038a6. 54. Met. VII, 12, 1038a7-8. 335

JAMES G. LENNOX

with priority going to the former as the basis for a correct logical division. The attempt to reconcile the natural basis of generic classifications with the view that, simply as genera, such classes do not exist, leads to a paradoxical status for genera. For by genus I mean that one identical thing which is predicated of both and is differentiated in no merely accidental way, whether (conceived> as matter or otherwise. For not only must the common (nature) attach to different things, e.g., not only must both be

animals, but this very animality must also be different for each (e.g., in the one case equinity (rb ~Tr~rov),in the other humanity (rb dvOpo~rov), and so this common (nature) is specifically different for each from what it is for the other, ss The genus is both one and the same across species boundaries, andyet, because it exists only as the underlying basis of fully differentiated species members, it is different qua aspect of horse and aspect of human. This paradoxical, Janus-faced character of the genus underlies Aristotle's discussion of the proper methodology for taxonomic divisions in Parts of Animals I, 3, as the following passage indicates: "But if a single and individual kind (e[6o¢) of being cannot belong to things differing in species, but will always be differentiated, as bird is from man (for their two-footedness is other and different) then if they [two species] are blooded the blood is different. ''s6 Here Aristotle notes that, while we may predicate trans-specific properties (two-footedness, blood) of more than one species, they will nonetheless be different in each species - or else, the distinction between the species is spurious. It is time to ask what is to be made of the concept of "generic material." The evidence is dearly against viewing generic concepts as class concepts - in some sense, the genus-term in a definition picks out a shared substratum that is differentiated differently from one species to the next. I recommend viewing the genus as a unified system of

shared characteristics (e.g., winged, two-legged, beaked, feathered, oviparous), all of which are possessed by each of the members of at least two species. This system of characteristics is the "material" for a variety of unified systems of specific differentiae in three ways: (1) 55. Met. X, 8, 1057b37-1058a5. 56. DPA I, 3, 643al-5. Balme translation, but translating e~6o~ in 643a2 as "kind" rather than "species." 336

Aristotle on Genera, Species, and "the More and the Less" each specific differentiae is an (but not the) actualization of a potential possessed by the generic character; (2) every generic character is a necessary condition of each specific character, and each specific character implies the existence of the generic character; and (3) our ability to identify genera indicates that the generic characteristics are still apparent in their differentiated state. A passage in Generation o f Animals suggests a metaphor to aid in understanding the recognizability of the genus as an aspect of fully differentiated species members. Interestingly, it is a description of differentiation during biogenesis: In the early stages the parts are all traced out in outline; later on they get their various colours and softnesses and hardnesses, for all the world as if a painter were at work on them, the painter being nature. Painters, as we know, first of all sketch in the figure of the animal in outline, and after that go on to apply the colours, s7 If we take it that the sketched in parts are the generic attributes (I will support this claim momentarily), then we may take Aristotle's claim to be that their further articulation will not imply the obliteration of the general pattem that has been laid down. That Aristotle does envisage the process of ontogenesis as passing through a generic stage prior to speciation is clear from an earlier passage in his discussion of embryogenesis: "for it is not the fact that when an animal is formed at the same moment a human being, or a horse, or any other particular sort of animal is formed, because the end or completion is last of all." ss Temporally, we can identify a stage where an embryo is clearly an animal before it is clearly any specific sort of animal. In addition, in his discussion of the inheritance of characteristics, Aristotle makes a comment about the more general characteristics being implied by the more particular: "'all particular individuals are accompanied by (dxoXooOe~) this characteristic [being human] : since "human being' is general, whereas Socrates who is the father, or the mother whoever she may be, are particular individuals." s9 The Greek verb 6xo?,ooOeZv is a semitechnical expression of Aristotle's for that which follows either causally or logically from something else. 6° Thus the above translation tends to hide the significance of the 57. 58. 59. 60.

DGA II, 6, 743b20-25. DGA II, 3, 736b2-5. DGA IV, 3, 768b13-15. See Bonitz, Index Aristotelicus, 26a25-61. 337

JAMES G. LENNOX passage. It is a claim about the causal (and logical) relationship between characteristics at two different levels of classification and two different stages of development. The model of the genus as a unified system of trans-specific characteristics that serve as the material for the fully articulated or differentiated characteristics of two or more species may have gained support from Aristotle's understanding of ontogenesis. At the same time, it appears to underlie his discussion of division in Parts of Animals I, 2-3. I have already noted that he approaches the relationship between ffmal differentiae and indivisible species by noting how generic characters such as "two-footed" or "blooded" must be more fully articulated at the species level. His own vision of taxonomic division is as follows: "the general differentiae must have species for otherwise what would make it a general differentiae and not a particular one? Some differentiae certainly are general and have species, for example featheredness: the one feather is unsplit, the other is split. ''61 But no actual species is lust feathered. Every feather is either split or unsplit (if this is an exhaustive list of types). Each general differentia is further differentiated within each species, the differentiation forming a hierarchical series (e.g., footed/four-footed/hoofed-four-footed). But to "capture" the nature of a real species we must follow the whole set of general (generic) differentiae simultaneously: If man were merely a thing with toes, this method would have shown it to be his one differentia. But since in fact he is not, he must necessarily have many differentiae not under one division. Yet more than one cannot belong to the same object under one dichotomy, but one dichotomy must end with one at a time. So it is impossible to obtain any of the particular animals by dichotomous division. 62 Aristotle's zoological #eTiora 7~vrl are distinguished from one another not by one trait that is necessary and sufficient for membership in the kind, but by an organized set of general traits. Birds are oviparous, feathered, beaked, two-legged, and winged. Fish are cold-blooded, gilled, scaled, aquatic, oviparous, and so on. Every species of bird or fish will possess each generic property, differentiated in a unique way. Aristotle may have arrived at this idea by noticing that it is the 61. DPA I, 3,642b25-29. 62. DPA I, 3,644a6-10. 338

Aristotle on Genera, Species, and "the More and the Less" generic characteristics to which our taxonomic expressions most naturally refer. "One should try to take the animals by kinds in the way already shown by the popular distinction between bird kind and fish kind. Each of these has been marked off by many differentiae, not dichotomously. 63 This is why one should divide off the kind straight away by many differentiae." 64 Typically, he offers an explanation for this alleged fact about natural languages: the specification or differentiation of these generic properties is a matter of quantitative changes in a pattern laid down at the generic level. The organic structures of different species of bird differ only by "the more and the less". The application of this notion to differentiae to explain how species are related generically, only a promissory note in Metaphysics VIII, 2, is achieved in the discussion of division in Parts o f Animals. Here I might fiote an analogy with evolutionary accounts of species relations. It is not unusual to see the precursors of an adaptive radiation referred to as having a more "generalized" form than their descendants. 6s The homological identity of such descendant forms can be seen as due to their all being "variations on the anscestor theme." Aristotle's picture of the genus as the shared substratum of its differentiated species bears more than a superficial similarity to the evolutionary view though, I hasten to add, his model is ontogenetic, not phylogenetic. That the model of the genus as matter for differentiation is (not too far) in the background in Aristotle's account of division is suggested by a comment which David Balme has recognized as an integral part of Aristotle's theory of division: "It is the differentia in the matter that is the species. For just as there is no part of an animal without matter, so there is none that is only the matter. ''66 Balme comments, "Aristotle may be referring to the physical matter or to the logical genus. But it makes no difference here (does it anywhere?). Each is one way of considering that which is potentially X. ''67 Much later, in a chapter of -

63. DPA I, 3,643b10-13. 64. DPA I, 3,643a23-24. 65. Onthe notion of phylogenesisastending to progressfrom more generalized to more specialized forms, see the discussions in Simpson, Principles of Animal Taxonomy, pp. 93-101; Bernhard Rensch, Evolution above the Species Level (New York: Columbia University Press, 1959), chap. 6; W. E. LeGros Clark, The Fossil Evidence for Human Evolution (Chicago: University of Chicago Press, 1964), pp. 41-47; and el. Darwin'sguarded discussion in On the Origin of Species (pp. 120-132 of the Collier edition). 66. DPA I, 3,643a24-27. 67. Balme, 643a24n, p. 114. 339

JAMES G. LENNOX

Parts of Animals IV I will shortly examine in detail, the manner in which the generic characteristics (the general pattern or sketch, to revert to an earlier metaphor) set limits on speciation is explained in terms of a generically common material: "Furthermore, though these birds [waders] are not great fliers, they are composed of the same materials (~K r~)¢ ~ ' cu)r~)¢ bT~r?¢)as the rest, and thus nutriment which in others goes to produce the tail feathers, in these is used up on the legs and makes them grow longer." 68 The material is shifted so as to shorten or "lessen" the tail feathers and increase the length of the leg - quantitative differences. The process described could be represented, as Thompson recognized, by deforming an abstract "mean bird" placed on a coordinate grid. It is interesting to note that Aristotle could offer a parallel explanation for the differences among the length and shape of the beaks of the finches of the Galapagos Islands; again, the explanation would be ontogenetic, not evolutionary. I have argued for the hypothesis that Aristotle's account of the differences between species of a genus as being mere differences of degree arose from the application of the concepts of "the more and the less" to understanding the differentiation of a generic substratum (or material), and thus to understanding the generation of an 000/a. Interestingly, in Metaphysics VIII, 3, in a passage comparing substances with numbers, Aristotle reiterates a common feature of theirs he had pointed out in the Categories: "as number does not admit of the more and the less, neither does substances"; but he quickly qualifies himself: "not substance as form (ehSo¢), but if any substance does, it is only the substance which involves matter (/aer& r~)¢ b%r/¢).''69 Natural substances, organisms in particular, are differentiated material things. But we have seen that the shared material is the basis of generic kinship. It is because of this c o m m o n substratum that the differentiae can be represented as differing in degree. Has Aristotle collapsed the distinction between differentiae and bodily affections (i.e., between essential and accidental properties) insisted on in the Categories, Physics, and in certain passages of the Metaphysics? At best, it seems that the general differentiae of the genus represent real discontinuities in nature, while variations from species to species are merely accidental "variations on a theme." I propose to answer the question of how the integrity of species is 68. DPA IV, 12,694b15-20. 69. Met. VIII, 3, 1044a10-12. 340

Aristotle on Genera, Species, and "the More and the Less" maintained in the face of their differentiae being characterized as varying along a generic continuum by introducing an element purposely omitted until now - teleology. Of all places, as Aristotle insists, it is in biology that we must grasp that for the sake of which each thing exists. 7° Differences that, viewed structurally, add up to nothing but differences of degree from one species to the next, will be, when viewed as adaptations to the ~5t0¢/~t'o¢ of each species, seen in a different light. To bring this point home, we need only see how Aristotle deals with the variations within a genus in Parts o f Animals IV. Aristotle begins discussing the "genus" of birds just as we would now expect: "In the genus of birds the differences relative to each other are in the excess .and defect of their parts, i.e., according to the more and the less." 71 Aristotle organizes his discussion not species by species, but generic differentia by generic differentia. Species (our genera, in most cases) are viewed as possessing each differentia in the appropriate degree for their special life-style. From the point of view of taxonomy, the chapter could be represented by the following model. Bird (W, X, Y, Z) Ducks

Cranes

Hawks

Sparrows

WalXb2YcaZd a Wa4XbaYb2Za 1 Wb2Xb4Ye6Zd 8 Wa4Xa4Yb6Zb 6 (The capital letters stand for generic differentiae, while the subscripts and superscripts represent the different relative degrees to which these differentiae possess certain perceptible qualities from one species to the next.) Showing that the relationships between the various species of birds discussed is one of "excess and defect of their parts" is only a small part of the intention of these chapters, however. By far the central question on Aristotle's mind throughout the discussion is to what end do these variations exist. Why is the hawk's beak hooked and why are its talons sharp and curved? Why are a duck's toes united by webbing, and why is its beak wide and flat? For what purpose are the legs and toes of cranes long and thin, and is there some relationship between these facts and cranes' relative dearth of tail feathers and wings? These questions draw attention to the fact that the variations within 70. DPA I, 1,639b15, 20, 641a2, 641a28, 641b11-12.

71. DPA IV, 12,692b3-5. 341

JAMES G. LENNOX a genus are dearly related to the requirements of adaptation to different environments. Viewed in abstraction from their specific environments, the variations between the beaks, wings, feathers, or legs of different birds appear to be nothing but variations in degree of the common differentiae of the genus of birds. But when they are viewed as adaptations to a peculiar mode of life, it becomes clear that each variation is the proper and peculiar one for a given species' way of life; no other differentiation would serve the needs of that species. A second feature of living things emerges when we focus on the environmental context of each species, and that is the causal unity of the differentiae of each species. Aristotle's explanations of the "excesses and deficiencies" are in terms of the shifting of the "ornithic" material here or there so that the adaptive requirements of each species may be met in the best possible way. It is an axiom of Aristotle's biology, borne out by every detail of the living world, that "nature produces nothing by chance, but always the best for the being of each kind of animal from what is possible. ''~2 For example, "Ducks have webbed feet because it is better on account of their way of life, in order that, living in water where wings are useless, they have feet useful for swimming." 73 Thus it is that every difference of beak, foot, wing, leg, feather, and bodily size and shape is explained in this chapter. The material used for lengthening legs in one species produces webbing between the toes in another. Taken individually and out of context, the organs of one species differ from another only in whether they have more or less length or breadth, hardness or softness, curvature or coarseness. Yet it is also true that each species possesses an organized set of adaptations, the best possible organization of material for its specific/3(o¢. Aristotle's account of animal generation also reflects his insistence that an organism is not a mere collection of unrelated parts, and that its development is not simply an unrelated series of alterations. Generation is o n e change with an underlying X63,o¢.74 The entire development, and all the changes which comprise it, are for the sake of a specific kind of animal - that is, for a specific mode of life. On t h e S o u l ' s application of the matter/form distinction to living things 7s makes it clear that ~oXr~,the e~8o¢ of a natural organic body, is 72. 73. 74. 75. 342

De lncessu 704b15-18, and note 25, above. DPA IV, 12,694b6-8. DGA II, 1,734b21-735a4,DeAn. II, 4,416a17. D e A n . II, 1,412a28-413a10.

Aristotle on Genera, Species, and "the More and the Less" to be understood as the inherent capacity o f each species of animal to live its own pr6per life. 76 This capacity is manifested in the successful functioning o f each creature in its peculiar environment. The " f o r m " of living things is their proper activity, and the generic material is differentiated into a species member for the sake of this. What ensures the integrity and "reality" o f Aristotelian species - in spite o f the fact that their organs differ structurally only in degree along a variety of continua - is that the members o f each species are individuals adapted to a specific manner o f life. They could not survive if a different mode of functioning were required o f them. Some evidence that Aristotle adopted this teleological method o f maintaining the distinction between "essential" and "accidental" characteristics comes from Generation o f Animals V. It is a discussion of the naOr~l~ara "by which the parts of animals differ" (778a16-17). He takes as essential those characteristics which "belong in the ~byo¢ o f each thing," understood to be those characteristics which either are directly for the sake o f the proper function o f that species, or which are means to the performance o f some function (778b13-14). The reading o f this book is fraught with difficulty, but I think one thing is clear: the "affections by which the parts differ" are n o t all accidental. How are those which are accidental differentiated from those which are for some end? Empirically, you might begin with the tentative assumption that a property universally possessed by a species serves some goal, while one that varies does not (see 778a30-33). But for various reasons y o u will be fooled (variation from one species member to the next, can, in certain niches, be highly advantageous; see 785b16 ff.). Ultimately, Aristotle's answer is: if a property serves a function necessary for that species' life, it is part o f the ~6yo¢ o f that species; if it does not it is accidental. Eyes with a specific functional capacity are essential for a specific animal's existence; "some eyes contain too much fluid, some too little to suit the right movement, others contain just the right amount. ''77 How do such differences differs? "Eyes intermediate between these two extremes differ merely by 'the more and the less.'"Ts Thus teleologically essential characteristics may differ from one species to the next by a slight, quantifiable

76. De An. II, 2, 413a30; and this approach i s echoed in the opening chapter of DPA. "If, then, the form of living things is soul or some aspect of soul.., then the natural scientist ought to speak of and understand the soul" (641a18-22). 77. DGA V, 1,779b26-27. 78. DGA V, 1,779b33-34. 343

JAMES G. LENNOX degree - but the difference is nonetheless essential (and notice how untypological that concept has become). 79 The limits imposed on D. W. Thompson's On Growth and Form by its author placed limits as well on his ability to understand the teleological aspect o f Aristotle's doctrine o f "the more and the less." The method Thompson recommends for the study o f formal relationships among species o f a genus seems to be a mathematical extension o f Aristotle's. Yet Aristotle would have thought Thompson's work had failed to ask, let alone answer, the crucial question: why is a given species possessed of these differentiae in just this degree and conformation? As Thompson says, "My sole purpose is to correlate with mathematical statement and physical law certain o f the simpler outward phenomena of organic growth and structure or form." 80 The account of the differences bdtween species of a genus in terms of the more and the less is descriptively accurate. However, noting the subtle relationships between those differences and the life required o f each species draws attention to the unity o f the differentiae of a species, and to the absolute isolation o f one species from others in the same genus. Differentia by differentia, compared out o f context, the species o f a genus differ only by "excess and defect," or "the more and the less." But when the balanced unity o f those differentiae in each species is studied, and when it becomes clear that the possession o f just this degree of length of beak is needed for the members of this species to get their food, Aristotle insists that the forms o f a genus are indeed different in form, and not merely in degree. Throughout most of Aristotle's corpus we fred a careful distinction between essential properties of substances and their accidents. He sometimes characterizes this distinction as one between the differentiae of a substance, which amount to its specific form, and the bodily affections of a substance, which are mere accidents, capable of altering or changing without the substance doing so. Generation is distinguished from mere alteration on grounds that the former is the organized construction of a form, the creation o f a substance, while the latter is only 79. Much of what I say here is, I believe, supported by David Balme's unpublished paper "Forms and Species in Aristotle's Biology," read at the 1976 Princeton Colloquium on Aristotle's biology. I have benefited a great deal from Balme's comments on an earlier draft of this paper. 80. On Growth and Form, p. 9. On Thompson's attitude toward natural selection as the cause of adaptation, see the comments of J. T. Bonnet in his introduction, p. x. 344

Aristotle on Genera, Species, and "the More and the Less" a change in the quality of a substance. "The more and the less," in this account, denotes the character of qualitative differences and changes alone, and the relationship among those qualities - they range along a continuum, differing only in quantifiable degree. At some point in his philosophical career, Aristotle took the step of applying the concept of "the more and the less" explicitly to the differentiae of the species of a genus. This has the apparent effect of making genera "real" and discontinuous, while species range along a continuum for each of their features, suggesting an ontology quite different from that normally associated with the Metaphysics, and in some respects more "Platonic." I have suggested a way of reconciling these two accounts of generic unity. The fact that the application of "the more and the less" to species differentiae seems to follow quite naturally from the genus-asmatter doctrine, and that this doctrine plays a significant role in the biology, indicates that the two views of generic identity can be reconciled. Species integrity is provided in the biology, perhaps more explicitly than in the Metaphysics, by the teleological (i.e., adaptive) unity of the specific differentiae. The unity of a species turns not on its having organs absolutely incomparable with other species (for this would destroy generic unity), but on the X67o¢, or form, that its differentiae display as a whole, a form defmed by the functions demanded of that species by its environment. The results of this study have two significant consequences. First, biologists, and historians and philosophers of biology, should reevaluate Aristotle's place in, and contribution to, the history of taxonomy and systematics. Aristotle does hold a brand of essentialism - there are certain functions, and consequent details of organization, that are necessary for the "life-style" of each species of a genus that determine the exact conformation of each generic character within that species. But this essentialism is not "Platonic" or inimical to evolutionary biology. On the contrary, with its stress on the way in which the requirements of adaptation determine the precise differences between species within a genus, it is a spiritual ancestor of the evolutionary approach. Second, Aristotle should be seen as a source of riches for students of the history of biology, particularfly on the interaction between questions traditionally assigned to "metaphysics" (no longer the pejorative term it once was) and questions of biological methodology. Parts of Animals I, is the first attempt at a philosophy of biology, and to date one of the best. 345

JAMES G. LENNOX

Acknowledgments This is a much revised version of a paper originally read to a departmental colloquium in the Department of History and Philosophy of Science, University of Pittsburgh, in March 1977. I would like to thank the participants on that occasion, and especially Myles Burnyeat, for their comments. In addition, I would like to thank David Balme, Allan Gotthelf, Alexander Nehamas, and this journal's anonymous referee for their many helpful comments and criticisms.

346