the scala naturae and the continuity of kinds

The Scala Naturae and the Continuity of KindsAuthor(s): Herbert GrangerSource: Phronesis, Vol. 30, No. 2 (1985), pp. 181-200Published by: BRILLStable URL: http://www.jstor.org/stable/4182226 .

Accessed: 14/09/2013 11:10

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Phronesis.

http://www.jstor.org

This content downloaded from 132.174.255.116 on Sat, 14 Sep 2013 11:10:41 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=bap

http://www.jstor.org/stable/4182226?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


The Scala Naturae and the Continuity of Kinds

HERBERT GRANGER

Aristotle's theories of essentialism and classification of his logic and metaphysics demand that substantial kinds be completely distinct or discrete kinds. In his biology Aristotle adopts as a classificatory scheme the scala naturae, which, according to some of his expounders, is incompatible with the theories of essentialism and classification of the logic and metaphysics. According to their interpretation, kinds in the scal form a continuous series in which they overlap one another, and thus it is impossible for them to be completely distinct from one another or to be discrete kinds. Arthur 0. Lovejoy argues for this interpretation of the scala in The Great Chain of Being,' and several other more recent scholars adopt or argue for much the same interpretation.2 I shall argue that this interpretation is wrong. I shall first consider what the "continuity of kinds" might mean, and what sort of threat continuity might level against Aristotle's

The Great Chain of Being: A Study of the History of an Idea (Cambridge Mass., 1936): 55-58. Lovejoy holds that Aristotle adopts a continuity of overlapping substantial kinds for the scala naturae and that his biology provides the evidence for this continuity in the form of kinds that overlap or participate in distinct substantial kinds, and which recent scholars have come to call "dualizers" (see p. 188 and n. 12). "Nature refuses to conform to our cravings for clear lines of demarcation; she loves twilight zones, where forms abide which, if they are to be classified at all, must be assigned to two classes at once" (p. 56). Lovejoy also recognizes that a continuum of overlapping substantial kinds conflicts with Aristotle's theories of essentialism and classification of his logic and metaphysics. He does not, however, seem to think that Aristotle gives up these theories when he adopts the continuum of the scala, but rather seems to think that Aristotle continues to hold them together with a belief in the continuum (pp. 57-58). 2 Anthony Preus maintains that the continuity of kinds in the scala is a continuity of overlapping kinds, and he argues that, as a consequence, the biology does not conform to a version of essentialism in which kinds are discrete, which is the version I attribute to Aristotle's logic and metaphysics, and which Preus characterizes as "Noah's Ark Essen- tialism": "Eidos as Norm in Aristotle's Biology," Nature and System 1, 1979, 80-82. The basic features of Preus's argument for continuity coincide with those of Lovejoy's, except

Phronesis 1985. Vol. XXX/2 (Accepted December 1984) 181



that Preus introduces the issue of hybrids. Preus holds that hybrids are evidence for a "fuzziness of the edges of the species-concept in Aristotle" (p. 81), which presumably he takes as evidence for an overlap between the kinds whose members give rise to the hybrids. Hybrids, however, would introduce such an overlap only if Aristotle holds that a criterion for membership in a kind is reproductive isolation from all other kinds, and there is some evidence to suggest that Aristotle does hold to such a criterion (e.g., GA 715b2-4). But it is doubtful whether Aristotle believes that hybrids form genuine substantial kinds on their own. In GA 738b27-34 Aristotle holds the view that hybrids after several generations revert to the form of the female that gave rise initially to the hybrid, which is a view Preus also notes (p. 82). The impermanence of hybrid kinds suggests that

they do not form genuine substantial kinds independent of the kinds of their parents. In his criticism of Preus's view that Aristotle does not hold the version of essentialism traditionally attributed to him, William Jacobs refers to the reversion of hybrids, and argues plausibly that they are deformed instances of the kinds of their parents (cf. Met. 1033b29-33): "Preus on Aristotle's eide," Nature and System 3, 1981, 115-118. For comments on Jacobs' criticisms, see Preus's "Reply to Jacobs," Nature and System 3, 1981, 119-121. Some of David M. Balme's remarks in his inaugural lecture to Queen Mary College strongly suggest that he holds that a contuity of kinds brought on by dualizers arises in the biology and that it is incompatible with Aristotle's views on essentialism and classification found in his logic and metaphysics: Aristotle's Use of Teleological Explan- ation (Inaugural Lecture at Queen Mary College, Univ. of London, 1965): 13, 16 and 17; cf. Balme's remarks about the scala in his Aristotle's De Partibus Animalium I and De Generatione Animalium I (Oxford, 1972): 97 (Hereafter cited as Arist. 's De Part.). Yet, in "Aristotle's Use of Differentiae in Zoology," Balme seems to hold that dualizers do not participate in kinds that are distinct substantial kinds: reprinted in Articles on Aristotle: 1. Science, eds. J. Barnes, M. Schofield and R. Sorabji (London, 1975): 189-190 (hereafter cited as "Dif."). Also, in the same piece Balme holds that Aristotle does not abandon in his biology the system of classification by genus and species of the logic and metaphysics, only that he ignores it in, for example, the History of Animals, because he is involved in preliminary investigations necessary to the establishment of such a system of classification (pp. 188 and 192). Balme holds the same view about classification in the biology in his "Genos and Eidos in Aristotle's Biology," Classical Quarterly 12, 1962, 98. Stephen Clark holds that many kinds of the scala naturae form a continuity of overlapping kinds, which is due to kinds that participate in kinds adjacent in the series, and which conflicts with a scheme of classification in which kinds are discrete: Aristotle's Man: Speculations upon Aristotelian Anthropology (Oxford, 1975), 31-32, 36, and 43. Jonathan Barnes seems to believe that the kinds of the scala naturae form a continuum of overlapping kinds, although he does not say that it is due to kinds that participate in distinct substantial kinds or that such kinds or the continuum threaten the versions of essentialism and classification in which kinds are discrete: Aristotle (Oxford, 1982) 62-63. G.E.R. Lloyd holds that kinds participating in distinct kinds, or "dualizers," present Aristotle with difficulties in his classificatory scheme of mutually exclusive kinds, and that they force him to speak about a continuity of kinds in the scala naturae. Nevertheless, Lloyd suggests, Aristotle continues to believe that substantial kinds are mutually ex-

clusive, and, rather than complicate his classificatory scheme through a revision that accommodates dualizers, he tolerates them as "anomalies": Science, Folklore and

Ideology: Studies in the Life Sciences in Ancient Greece (Cambridge, 1983), 50-52 (hereafter cited as Sci.).

182



theories of essentialism and classification. Then, I shall make fully explicit the argument for Aristotle's adoption of a form of continuity for the scala that endangers his views on essentialism and classification found in his logic and metaphysics. My exposition of the argument basically depends on Lovejoy's argument because it is the most complete. It encompasses, however, the important points put forward by those defending the same interpretation as Lovejoy's because on the fundamental issues Lovejoy's argument and theirs do not differ significantly. The argument falls basically into two parts: (A) an interpretation of the use of'"continuity" in the two passages in the biology in which Aristotle holds that kinds are continuous; (B) Aristotle's evidence for continuity, which takes the form of certain kinds, which have come to be called "dualizers," and which allegedly belong to different substantial kinds. I shall take up an examination of each part of the argument. I shall argue that they are inadequate, and defend the view that the form of continuity Aristotle adopts for the scala naturae does not undermine the theories of essentialism and classification of the logic and metaphysics.

I

There are three senses of the "continuity of kinds" pertinent to an examination of the scala naturae.

(I) Kinds form a continuous series because they form a dense series. Given any two kinds, A and B, a third kind C can always be distinguished between them in the series. In principle, such a series would form an infinite series of kinds, although each kind would be discrete or distinct from every other kind in the series. I shall call this form of continuity "dense continuity," and it is probably the type we would today take a continuous series of kinds to form. In the Physics Aristotle does offer one account of "continuity" that might, if applied to kinds, give rise to a dense continuity of kinds. In Physics 232b24-25 he says that he means by "continuous" what is "divisible into items always divisible," which is a property he ascribes to magnitudes, times and motions. Kinds as infinitely divisible would most likely mean that there is always another kind between any two given kinds, however close they might seem to be in the series, and thus such kinds would form a dense series. Aristotle would not, however, apply this sense of "continuity" to kinds because it would mean that kinds form an infinite series. Aristotle believes that the natural world is intelligible and knowable. But an actual infinite series of kinds would not be knowable,

183



and thus Aristotle must reject it out of hand.3 Consequently, dense continuity cannot be the continuity of the scala naturae, and I shall dismiss it from further consideration.

(2) Kinds form a continuous series because they form a series of overlapping kinds. Two continuous kinds, A and B, overlap because they have members in common. A and B do not mark off exclusively A's and B's but mark off as well things that are both A and B. Because A and B have members in common, there is a progression from the members of A to those of B that is continuous or uninterrupted by a sharp distinction between A and B. Since no clear-cut interruption divides sharply A and B from one another, by dividing sharply their members from one another, A and B are not discrete or possessed of distinct borders with respect to one another in the series. I shall call this kind of continuity "strong continuity," and it is the sort defended by those who take the continuity in the scala naturae to be a serious threat to Aristotle's views on essentialism and classification found in his logic and metaphysics.

(3) Kinds form a continuous series because they form a graduated series that progresses only by small steps. In such a series the kinds do not share any members, and there are only slight differences between adjacent kinds in the series. In a series composed of the kinds A, B, C, and so forth, A and B have no members in common and differ from one another only slightly, with respect to some set of attributes; B and C have no members in common and differ from one another only slightly, with respect to some set of attributes, which differs only slightly from the set that distinguishes A and B, and so on through the series, in which progress continues to proceed only gradually. In such a series, especially if there is a large number of kinds, difficulties could arise in distinguishing between kinds adjacent in the series; they may even appear to blend into one another and thus may appear to form a strong continuity. But each kind is still discrete or distinct from every other kind in the series because they have no members in common. Consequently, the difficulty of their identification and dif- ferentiation is merely a practical problem, although it may be one that is never in fact resolved. This kind of continuity I shall call "weak continuity," and it is the sort I shall attribute to the scala naturae of Aristotle.

3 See, e.g., GA 715bl4-16 for a reference in Aristotle's biology to Nature's avoidance of the infinite, and for the unknowability of an infinite series, see, e.g., APo. 82b37-83a 1; for a brief discussion of this issue and for additional references, see Joan Kung, "Aristotle on Thises, Suches and the Third Man Argument," Phronesis 26, 1981, 225 and notes 48 and 49.

184



Whatever may ultimately be the full account of Aristotle's theories of essentialism and classification in his logic and metaphysics, it is at least clear that they forbid strong continuity, but not weak or dense continuity.4 In Aristotle's accounts of essentialism and classification in his logic and metaphysics, substantial kinds are mutually exclusive or discrete, but under the terms of strong continuity kinds are compatible because they overlap. Kinds, however, remain discrete in weak, as well as in dense, continuity. For Aristotle substances possess a finite number of essential attributes that place them in their substantial kinds, their genera and species, which are incompatible with all other substantial kinds, except those species or genera superordinate and subordinate to them. A definition of a substance's species, its basic substantial kind, provides a definition of its essence or basic nature, since it reveals what it is. The genera of its species are also substantial kinds of the substance, and, although less basic, since they encompass the other species that belong to them, a definition of them too yields information concerning the nature of the substance.5 Since for Aristotle a substance has only one nature or essence, it can belong to only one species and the genera of that species: "'. . . for we count as more than one thing ... of which the species is not one, or of which the definition is not one."6 Therefore, a substance's substantial kind, its species and genera, must be incompatible with all other substantial kinds. If not, then a substance would belong to more than one species or to different genera not subordinate to one another, and thus it would have more than one basic nature or essence.7 Accordingly, in classification by genus and species, substantial kinds must be incompatible with all other substantial kinds other than their genera and species. If not, then a substance could be classified as belonging to more than one species or to different genera not subordinate to one another (cf. Top. 140a27-29). Therefore, if two kinds are compatible, and they are not subordinate to one another as species to genus, then one or both of them are not substantial

I For the traditional interpretation of Aristotle's theories of essentialism and classification, with which my interpretation conforms, see, e.g.: Eduard Zeller, Aristotle and the Earlier Peripatetics, (London, 1897), 213-217; W. D. Ross, Aristotle (London, 1923), 37-38; J. H. Randall, Aristotle (New York, 1960), 174-176; Balme, "Diff.", pp. 183-185; Marorie Grene, A Portrait of Aristotle (Chicago, 1963), 87; 231-232; G. E. R. Lloyd, Aristotle: The Growth and Structure of His Thought (Cambridge, 1968), 48 (hereafter cited as Arist.); Sci., pp. 50-51; Ernst Mayr, Principles of Systematic Zoology (New York, 1969), 66. 5 Cat. 2b7-14,29-34; cf., e.g., Top. 10Ib38,139a26-27,141b22-28,APo. 82b37-83al. 6 Met. 1016b9-1 1; all of the translations are mine. Cf. Met. 1052a29-34. 7 Cf. Met. 1037b24-27, Top. 125b37-39; cf. also APr. 96a31-34, Met. 1030a6-14, b7-11.

185



kinds. Accordingly, the defenders of strong continuity are right to argue that if the scala naturae introduces strong continuity, then the continuity of kinds in Aristotle's biology overthrows the theories of essentialism and classification of the logic and metaphysics. If, however, it introduces merely weak continuity, then these theories are not threatened by the continuity of kinds. I turn now to the relevant material in Aristotle's biology, and a full development of the case for strong continuity.

II

In several passages in the biology Aristotle organizes animals and sometimes plants into a hierarchically arranged series, which has come to be known as the scala naturae.8 The criteria he uses to determine its ranks vary: posture; the degree of perfection of the offspring at birth; the degree of life and motion.9 In only two of these passages, however, does Aristotle hold that the kinds that make up the series form a "continuity" (uuvEXeLa), and they are the two passages the advocates of strong continuity appeal to in defense of their position.

Parts ofAnimals IV.5.68la12-15 Nature changes continuously from inanimate things to animals through animate things that are not animals, in such a way that it seems that one [kind] differs from another altogether very little because of their proximity to one another.

History of Animals VIII. 1.588b4- 16 Thus nature changes little by little from the inanimate to the animate; so that because of the continuity their border escapes notice, as well as to which side a middle [genus] belongs. For after the genus of inanimate things the genus of plants is first. And these differ from one another according to the more they appear to participate in life. For the genus as a whole, when compared to inanimate things, appears almost as if it were alive, but when compared to the genus of animals it is lifeless. And the change from these to animals is continuous, as has been said before. For one would be at a loss to determine whether some forms in the sea are animals or plants; for they live attached, and many of these perish if detached; for instance, the pinna lives attached and the solen when detached is not able to live.

In both of these passages Aristotle organizes animals, plants and inanimate things into a series. The passage from HA, together with a remark occurring a few lines later at 588b21-23, show that in HA VIII.l he also organizes

I Plato is perhaps the first to suggest that animals are arranged hierarchically, Timaeus 90E-92C; see Friedrich Solmsen, "Antecedents of Aristotle's Psychology and Scale of Beings," American Journal of Philology 76, 1955, 148-164. 9 Posture: PA 686a24-687a5, On the Progression ofAnimals 5, On Youth and OldAge; On Life and Death 467b26-468al3, On Respiration 13, cf. HA 494a26-bl; perfection of the offspring: GA 732bl5-733bl6; life and motion: HA 588b4-23, cf. PA 681al0-bl3.

186



animals and plants on a scale according to the degree of life and motion they possess. The more life plants possess the closer they are to animals and the higher they are on the scale. Animals are ranked not only by the degree of life they possess but also by their degree of motion: presumably, the more life and motion they possess the higher they are on the scale. In both the passage from PA and HA kinds in the series are "continuous," but in neither of them does Aristotle explicitly explain what he means by "continuous."

Lovejoy appeals for his interpretation of "continuity" to the definition of "continuity" of which Aristotle offers parallel accounts in the Metaphysics and Physics, and upon which a plausible case for strong continuity can be based.10 In Physics 227all- 17 Aristotle defines the "continuous" in these terms:

... I call things continuous, when the limits of each thing by which they touch become the same and one, and, just as the name signifies, they are held together. This is not possible when the extremities are two. From this definition it is evident that continuity is in those things that naturally become something one on account of their union. And in whatever way the thing holding them together becomes one, so also the whole will be one, for example, either by a bolt or by glue or by contact or by organic unity."

Here Aristotle explains "continuity" in terms of spatial objects or the surfaces of such objects. For example, two surfaces become continuous when their borders become identical and, when this identity arises, which may be achieved in various ways, through organic unification or even a bolt, the surfaces become one surface, which apparently is then one uninterrupted or continuous surface. Presumably, Lovejoy believes that Aristotle models the continuity he applies to kinds on spatial continuity, and that, as a consequence, the continuity of kinds is strong continuity. On this interpretation of continuity, if two kinds are continuous, then, as in the case of continuous surfaces, the distinction between their borders disappears. They do not delineate sharply their members, so that their members belong exclusively to one or the other kind. Instead, the kinds have members in common, and thus they overlap one another. Hence no clear- 10 Lovejoy, P. 55. Lovejoy merely quotes the passage from the Metaphysics on continuity, and assumes its relevance for a discussion of the continuity of the scala He does not actually elaborate a defense of spatial continuity as Aristotle's model for the continuity of kinds. In his discussion of continuity, Stephen Clark also appeals to the same definition of continuity, although its version in the Physics (p. 29). Of the defenders of strong continuity, Lovejoy and Clark are the only ones to suggest that Aristotle bases the continuity of the scala on his account of spatial continuity, and they are also the only ones to suggest a basis for Aristotle's conception of the continuity of the scala. 1 The corresponding passage in the Met. is 1069a5-8.

187



cut distinction marks off the members of one kind from those of the other, just as in a continuous surface there is no interruption in the surface to distinguish one part of it from another. Without such a distinction to mark the difference between the kinds, they are not discrete or in possession of distinct borders with respect to one another within the series. Rather, the borders between two continuous kinds are identical or merge because the kinds have members in common. Their identity brings about a unity of the two kinds, which allows for an uninterrupted or continuous progression from the members of one kind to those of the other, and which is analogous to the unity arising between two continuous surfaces, over which move- ment is continuous or uninterrupted by a break in their surface.

The interpretation of the scala in terms of strong continuity takes the above passage from HA VIII. 1 to maintain that the distinction is not a sharp one between animate and inanimate kinds, but one dulled by the introduction of the plant. It is an intermediate kind that participates in both kinds - compared to animals it is inanimate, but compared to inanimate bodies it is animate - and thus it allows the borders of animate and inanimate to merge, because its members are common to both kinds, and allows these kinds, which as putative substantial kinds ought to be mutually exclusive, to be compatible and continuous. Similarly, on this interpretation, no sharp distinction distinguishes animals from plants. There are sea creatures between animals and plants which are both, the famous zoophytes, which provide the bulk of Aristotle's examples in his discussions of continuity in the scala.

The defenders of strong continuity maintain that Aristotle's evidence for the continuity in nature is provided by the kinds that participate in different kinds, allegedly substantial kinds, which in terms of the essentialism of the logic and metaphysics are mutually exclusive (cf. also PA 68 1a25-28, 681 a36-b2). Aristotle expresses the character of these kinds as participants in different kinds through the verb sa~T&porepitrmv. This word has several related meanings: for example, "to play a double game" or "stand neutral"; "to halt between two opinions"; "'to equivocate"; "to admit a double sense." A. L. Peck tries to capture Aristotle's use of it in biological contexts through the word "dualize," which has come to be the standard translation.'2 Consonant with the account of continuity as modeled on

12 1 am especially indebted to A. L. Peck's discussions of dualizers, and the expressions, "terrestrial" and "aquatic," in the introduction to his translation of the first three books of the History of Animals in the Loeb series: lxxiii-lxxv; Ixxvii-lxxxix. For the uses of i?(LO rEpmpeLv before Aristotle and in contexts in Aristotle outside the biology, see Lloyd, Sci., pp. 44,48,49, and n. 172 on p.44.

188



spatial continuity, the "dualizer" provides then the "bolt" that unifies the kinds to which it belongs, and brings about their strong continuity. The zoophyte, for example, dualizes between animal and plant because its members possess characteristics only an animal and a plant would seem to possess. Thus plant and animal are not after all mutually exclusive and discrete kinds, as one would expect of substantial kinds on the interpretation of the logic and metaphysics, but compatible and continuous in the strong sense, because the zoophyte participates in both in significant, presumably essential, respects. PA IV.5, in which one of the two passages on the continuity of kinds occurs, provides a good example of the zoophyte as a dualizer. Knides or acalephae, because they "dualize" between plants and animals, are not testaceans and fall outside the "defined genera." Like animals, they can live detached and fall upon their food, perceive objects they come in contact with, and use the roughness of their bodies for defense. On the other hand, like plants, they are incomplete and thus will attach themselves to rocks; and they have no residue (68 1a36-b8).

Aristotle does not limit his discussions of dualizers to his expositions of the scala naturae, nor does he limit his examples to zoophytes and the kingdom of plants. Dualizers arise in virtually every treatise in the biology, and they are found between many animal kinds. The latter part of PA IV. 13, which provides the longest account of dualizers outside HA VIII.2, discusses the cetacean, seal, bat and ostrich (697al5-b26). Cetaceans, animals having a blowhole, such as the whale and the dolphin, dualize because "in a way" they are both terrestrial and aquatic: terrestrial in that they breathe, aquatic in that they do not have feet and take their nourishment from the sea. The seal, like the cetacean, dualizes between the terrestrial and the aquatic, whereas the bat between the terrestrial and the flier or the aerial: they participate "in both" the groups they dualize between and "in neither." The ostrich dualizes between the quadruped and the bird in several respects: for example, as a bird, it is feathered and is two-footed; as a quadruped, it is large, which is why it cannot fly, and thus it is cloven-hoofed. Aristotle often mentions the dualizing nature of the seal, which arises in several areas. Here in PA IV. 13, he holds that if one considers it as an aquatic, it has feet, but as a terrestrial, it has fins. Also, like a fish, it has sharp and interlocking teeth; hence in HA 11. 1, because of its teeth, it "overlaps" (Qa\oaX&'rTeLv) the genus of fishes (50la2 1-23). Also, in HA VI. 12 it is terrestrial because it breathes and because it sleeps and gives birth on land; whereas it is aquatic because it spends most of its time in water where it gets its food (566b27-567a 14). Moreover, in HA II. 17 it has a forked tongue, like serpents and lizards (508a27).

189



One could continue for some time to recount dualizers and the many different ways in which they dualize. Swine, for instance, dualize in that some species are solid-hoofed and others are split-hoofed (HA 499b 12-14). Man dualizes with respect to the number of offspring: one, few and many (GA 722a37-b3). Man, like some other animals, also dualizes by being both gregarious and solitary (HA 488a 1-7). Some fish dualize by inhabiting both deep and shallow water (HA 598a 14-15). Selachians, externally viviparous fish, dualize with respect to the location of the uterus (HA 51 la24-28). The water-newt has gills yet is footed (Resp. 746a5-6, HA 589b27-29). Other dualizers include the elephant (PA 659a2-8), primate (PA 689b32-35, HA 11.8-9), serpent (HA 505b5-7), hermit crab (HA 529b24), hare (GA 774b2-4) and probably any animal, which, although terrestrial or aerial, pursues its food in water, such as the otter, beaver, shearwater and so forth, but, in addition to the seal, certainly the sea-turtle, crocodile, hippopotamus, tortoise and frog (cf. HA 487a 19-23, 589a 16-3 1). Clearly, animals dualize not only with respect to anatomical or morphological features, but also with respect to all the other features Aristotle designates as important, those that concern life, activity and character (e.g., HA 487al0-I 1).

Dualizers are not then an isolated exception. They arise throughout the animal kingdom, and the whole genus of plants even seems to dualize between animate and inanimate kinds. If dualizers, as the defenders of strong continuity maintain, participate in distinct substantial kinds, then they would provide a problem for the theories of essentialism and classification of the logic and metaphysics, and they would also provide the best support for the view that kinds form a strong continuity in the scala naturae. For if some animals and plants can possess attributes that ordinarily would seem to be definitory or essential in nature, which would place those animals and plants into different, prima facie mutually exclusive, kinds, then the kinds they fall in would lose any claim to a sharp distinction from one another. Consequently, they would lose any claim on being substantial kinds, at least on the interpretation of the logic and metaphysics, and also a strong continuity would arise between them. And if those animals and plants could never be placed in mutually exclusive kinds, no matter how one attempts to classify them, then no substantial kinds could exist for them. Hence they could not have discrete kinds that could be placed in definite niches in a classificatory scheme according to classification by genus and species.

Therefore, it should be clearly noted, dualizers are a problem for Aristotle's theories of essentialism and classification of his logic and metaphysics only if, under every interpretation of their nature, they always

190



fail to fall into only one substantial kind, but fall into what prima facie are two or more non-subordinate substantial kinds, because of certain, apparently essential, properties they possess. The seal's possession of a property it shares with the shark is not itself troublesome, only that it shares with it what seems to be an essential property, which belongs exclusively to fishes, which thus calls for the assignment of the seal to the kind fish, and which therefore results in its belonging to what would appear to be two mutually exclusive substantial kinds, the mammal and the fish. In many cases, however, dualizers would not be problematic: the dualization of man between gregarious and solitary behavior would not be a serious problem because one could probably devise a definition of man that avoids these properties. Yet other cases certainly seem troublesome. Terrestrial and aquatic, along with aerial, occur throughout Aristotle's works as important classifications,13 and thus an animal that dualizes between these apparently incompatible substantial kinds would seem to provide a serious problem for the theories of essentialism and classification of the logic and metaphysics. Presumably, advocates of strong continuity hold, when Aristotle's biological investigations led him to recognize numerous dualizers of this sort, he was compelled to abandon, or at least to ignore, the essentialism of his logic and metaphysics and to adopt, as a classificatory scheme for his biology, a scale of continuous kinds in which kinds overlap one another.

III

The argument for strong continuity consists of two parts: (A) an interpretation of the use of "continuity" in the two passages in which Aristotle speaks about the continuity of kinds, in which "continuity" is taken to be modeled on the definition of continuity from the Metaphysics and Physics, which applies to spatial objects; (B) the existence of dualizing kinds. On examination neither part supports the case for strong continuity.

13 One or another or all of these classificatory terms show up in the important discussions concerning genus and differentia in divisions and definitions: Top. VI.6,APo. II.5 and 13, Met. VII. 12, X.9, PA I.2-3. D. M. Balme suggests that in Academic classification they probably serve as afundamennum divisionis: Arist. 's De Part., p. 129. In GA 732b15-27 Aristotle discredits them because they do not correspond to the methods of reproduction, yet he continues to appeal to them throughout the biology: cf. G. E. R. Lloyd, "The Development of Aristotle's Theory of the Classification of Animals," Phronesis 6, 1%6, 75 and n. 4 on p. 75; Sci., pp. 50-51. For some references to them in the biology, see, e.g.: HA 504b 13, 536b32, 542a25, PA 668b33.

191



(A)

In the two passages in which he speaks about a continuity of kinds, PA IV.5./681al2-15 and HA VIII.l./588b4-16, Aristotle does not explain what he means by "continuity." The context in each passage, however, does shed light on its meaning, and in neither passage does the context suggest a strong continuity of kinds, but rather only a weak continuity. In the passage from PA kinds differ only slightly from one another, because of their proximity to one another in the series, which suggests that kinds form a graduated series that progresses only by small steps, and not a series in which kinds overlap one another.14 Similarly, in the passage from HA, nature changes "little by little" from inanimate to animate kinds,'5 which also suggests the weak continuity of a graduated series of small steps. In the passage from HA Aristotle also says that the "border" between inanimate and animate kinds "escapes notice" because of the continuity and also that it "escapes notice" into which of the two the kind in the "middle" -

namely, plants - falls. These remarks too suggest weak rather than strong continuity. Because there are only slight differences between kinds in the series, it is difficult to determine when inanimate kinds end and animate kinds begin. Hence their border "escapes notice," and it is difficult to determine in which kind plants, which are similar to both animals and inanimate things, belong. But if Aristotle had in mind strong continuity he would instead say that the borders of inanimate and animate kinds are identical and that the kind in the middle belongs to both of these kinds. This is actually how Lovejoy takes this passage, as his translation reveals: "... their continuity renders the boundary between them indistinguish- able; and there is a middle kind that belongs to both orders."''6 But Lovejoy's translation is wrong, and if it prompts his interpretation of

14 Jacobs makes much the same point in his criticism of Preus, p. 116. 15 In my translation of the passage from HA, I have taken the first use of ~Ca, which is at 588b4, to mean "animate," and its other uses, singular and plural, at 588blO, 11 and 13, to mean "animal" or "animals." This interpretation makes the best sense of the passage, and the four translations I have checked of this passage basically support my interpretation. All four translations differ in the phrasing they pick to render the first use of ~Cxi, but agree in translating its other forms as "animal" or "animals." Lovejoy translates its first use as "animate" (p. 55); G. E. R. Lloyd renders it as "living creatures" (Arist., p. 88); P. Louis in the Bude translation takes it to mean "etres doues de vie." D'Arcy W. Thompson in the Oxford translation renders the first use as "animal life," which, I suppose, might mean the same thing as "animate." At any event, even if the first use of gCxp should be taken to mean "animal," it would not seriously affect my argument. 16 Lovejoy, p. 56.

192



continuity, he is ill served by it. The Greek is clear and could not justify his translation: L)(Yo? TFrTAVV iX tXOVO('V?LVTO LEOOpLOV rVT:rV XtLT p0a0ov EOT?Tp@PV

UYTiV.17 In addition, in his expansion on his remark about the middle kind plants, Aristotle only draws out the similarities plants possess with both animals and inanimate things, when plants are contrasted with each of them, and does not hold that plants belong to both of these kinds. A continuity also arises between animals and plants. But Aristotle does not say that zoophytes, such as the pinna and solen, which are the evidence for the continuity, are both animals and plants, which they would be if the continuity were the strong variety. Rather, he says that "one would be at a loss to determine" (&rarop'aF?Ltev &V TLS) whether zoophytes were plants or animals. This comment could mean that Aristotle has in mind merely weak continuity, in which kinds differ only slightly. Thus, on this interpretation, the difficulty arises of determining at what point kinds differ from one another and to which kind a given kind belongs, such as the zoophyte. There is a practical difficulty in fixing the substantial kind of the zoophyte, because there are reasons for its assignment to distinct substantial kinds, but theoretically it must belong to only one of them.18 On the interpretation of the scala in terms of weak continuity, dualizers such as the zoophyte still provide the evidence for the continuity of kinds, but instead of strong continuity, they evince only weak continuity. For they are responsible for the slight differences between kinds because of their resemblance to mutually exclusive substantial kinds, and thus they are responsible for the graduated series that progresses through small steps, which kinds form in the scala naturae.

The two passages from PA and HA provide the only places in Aristotle's works in which he holds that kinds are continuous. They fail on examination to yield evidence for strong continuity, and instead support the case for weak continuity. Since these passages clearly argue for weak continuity, it does not then seem legitimate to appeal to the definition of continuity in the Metaphysics and Physics that yields an interpretation of the continuity of kinds as a strong continuity. Besides, Aristotle actually

17 588b5-6. With the exception of Lovejoy's, the translations I refer to in n. 15 corrobo- rate my rendition of this passage: ". . . in such a way that we are unable, in the continuous sequence, to determine the boundary line between them or to say on which side an intermediate kind falls" (Lloyd); ". . . in such a way that it is impossible to determine the exact line of demarcation, nor on which side thereof an intermediate form should lie" (Thompson); ". . . si bien que cette continuite empeehe d'apercevoir la frontiere qui les separe, et qu'on ne sait auquel des deux groupes appartient la forme intermediaire" (Louis). 18 Cf. Lloyd,Sci.,p. 51.

193



applies that definition to the continuity found among spatial objects rather than kinds. There are no other passages in or outside the biology that involve Aristotle's use of "continuity," which I know of, which would contribute anything further to the issue of the continuity of kinds.

(B)

The two passages from PA and HA on the continuity of kinds indicate that Aristotle's use of "continuity" to describe kinds in the scala supports only the case for weak continuity, which poses no threat to the theories of essentialism and classification of the logic and metaphysics, and the passage from HA indicates as well that Aristotle does not appeal to dualizers in a way to suggest that they are evidence for strong continuity. Chapter Two of HA VIII, which immediately follows the chapter from HA on the continuity of kinds, also indicates that Aristotle's views on dualization do not demand a scale of substantial kinds that form a strong continuity, and consequently do not require the abandonment of his theories of essentialism and classification of his logic and metaphysics. Chapter Two of HA VIII is a long chapter, a large portion of which is devoted to the categorizing expressions, "terrestrial" and "aquatic," and also to the nature of dualizing animals.'9 It offers an account of "terrestrial" and "aquatic" that clearly suggests a way to explain how dualizers need not lead to a strong continuity of substantial kinds. It is difficult to review the discussion of HA VIII.2 because it appears to be made up of material written at different times,20 but the basic point it pursues is clear enough: that classificatory expressions like "terrestrial" and "aquatic" are

19 The chapter runs from 589alO to 592a28, and much of it concerns the habits and diets of various sea creatures. The material relevant to this paper is found at the beginning, in 589alO through 590a16. 20 In HA VIII.2 there are two contiguous discussions of the ambiguity of classificatory expressions: 589a10-31 and 589a3l-b29. The first one concerns the ambiguity of both "terrestrial" and "aquatic," and it closely parallels a discussion of the ambiguity of these expressions found in HA 487al4-b7. The second one concerns the ambiguity of "aquatic" alone. These two discussions cover much the same ground and are certainly compatible. Moreover, they do not seem to acknowledge one another or to depend on one another. There is also a minor inconsistency between them concerning animals that take in water for the sake of cooling but pursue their food on the land or in the air: cf. 589a22-23 and b26-29; see also note 22. These considerations suggest that they were written in- dependently of one another. Because they are compatible and support the same basic position, a later editor, perhaps even Aristotle, probably placed them together in HA VIII.2. In my discussion I do not distinguish between these two discussions of ambiguity.

194



ambiguous, and thus once one recognizes their ambiguity and makes explicit their various senses, there is no reason to believe that there are any kinds that threaten to fall into distinct substantial kinds.

In HA VIII.2 Aristotle argues that "terrestrial" and "aquatic" are ambiguous expressions and that animals may be divided into terrestrial and aquatic in three different ways. One distinction concerns the taking in of air or water. Some animals are terrestrial or aquatic because they take in air or water for the sake of cooling the blood, which is Aristotle's explanation of the purpose of respiration (Resp. 10). These animals, such as man and fish, have lungs or gills.21 Aquatics with respect to taking in water include, however, another distinct group. Cetaceans breathe and thus are terrestrials, but they are also aquatic because in the process of taking up their food they take in water, which they discharge through their blowhole, provided for that purpose. Therefore, aquatics with respect to taking in water come in two forms: those that take in water to cool the blood; those that take in water incidental to feeding. The latter, exemplified by the dolphin, is an important case, which I shall shortly return to. Some animals, although they take in neither air nor water, are terrestrial or aquatic because their constitutions demand, for the sake of cooling, that they live on land, where the air cools them, or in the water, where the water cools them. Insects, such as wasps and bees, exemplify these terrestrials; sea anemones and shellfish are examples of these aquatics. The last of the three distinctions concerns nourishment. Some animals are terrestrial or aquatic because they pursue their food and make their habitat on land or in water. Examples of these come exclusively from dualizers. For example, the seal, which takes in air and breeds on land, is a terrestrial, but because it spends most of its time in water for the sake of its nourishment it is an aquatic with respect to nourishment.22 Indeed, interestingly enough, Aristotle says in this context that dualizers like the seal "alone" dualize, "for one ought to set them down as both terrestrial and aquatic" (589a20-22). Of course, as Aristotle notes, an animal might be terrestrial or aquatic in all three respects outlined here, or in only some of them (590al 1-16).

21 The discussion of the ambiguity of "terrestrial" and "aquatic" in HA 1.1, which I mention in note 20, provides these examples, as well as the examples of terrestrials and aquatics whose constitutions are cooled by the medium in which they live. Aristotle does not provide examples of these animals in HA VIII.2. 22 At HA 589a22-23, Aristotle denies, as he also does in HA 487b2-3, that there is any animal that takes in water yet is terrestrial or aerial because it spends time on land and in the air for the sake of its nourishment. Later in HA VIII.2, however, at 589b26-29 he says that there is only one such case known, the "so-called cordylus" or water-newt.

195



Through offering different senses for "terrestrial" and "aquatic," Aristotle is in a position to offer an explanation of dualizers that allows them to be compatible with the essentialism of his logic and metaphysics, and does not demand a strong continuity of kinds in any classificatory scale. Since "terrestrial" and "aquatic" are ambiguous, the terrestrial-aquatic need not appear to dualize between incompatible categories. The seal is terrestrial because it takes in air for the sake of cooling; "terrestrial" in this sense is incompatible with "aquatic" when it signifies "taking in water for the sake of cooling." But the seal is not aquatic in that sense; rather it is aquatic in important, perhaps even in essential, respecs, but not in in- in water for the sake of its nourishment. Hence it is both terrestrial and aquatic in important, perhaps even in essential respects, but not in incompatible respects. Instead of designating the seal by the expressions, "terrestrial" and "aquatic," and thus leaving the impression that it belongs to incompatible substantial kinds, which one must then reinterpret as compatible and continuous, Aristotle should designate it by the phrases, "terrestrial with respect to taking in air for the sake of cooling" and "aquatic with respect to nourishment," which are explicitly and obviously compatible and which, consequently, cannot both express substantial kinds. Aristotle could then go on to designate the fish by the phrases, "aquatic with respect to taking in water for the sake of cooling" and "aquatic with respect to nourishment." Although the seal and the fish are both aquatic with respect to nourishment, Aristotle could nonetheless neatly distinguish them by the phrases, "terrestrial with respect to taking in air for the sake of cooling" and "aquatic with respect to taking in water for the sake of cooling." These incompatible phrases Aristotle could then use to designate mutually exclusive substantial kinds.

Therefore, the seal and dualizers like it, which at first glance seem to belong to incompatible substantial kinds, are not in fact troublesome. Because of Aristotle's appeal to ambiguity, the mystery of their dualizing disappears. The categories they dualize between only appear to be incompatible because Aristotle uses the same expressions to signify different, although related, ways in which an animal might be terrestrial or aquatic. Accordingly, one can after all sharply distinguish the seal from other kinds, such as the fish, since they are not aquatic in exactly the same sorts of ways. Thus no appeal to a strong continuity between the mammal and the fish seems necessary just because of the existence of the terrestrial-aquatic. Of course, if Aristotle continues to use "terrestrial" and "aquatic" simpliciter as important classificatory expressions, then dualizers like the seal would continue to appear to fall into incompatible substantial kinds and appear to

196



entail the strong continuity of those kinds. But because of their ambiguity, they are useless as classificatory expressions. They should give way to distinct expressions for the distinct ways in which an animal may be terrestrial or aquatic, so that they may no longer obscure the discreteness of kinds.23 Indeed, Aristode ought to give them up, if he is to remain faithful to his explicit requirement that substantial kinds, genera and species, be "synonymous" with their subjects, where synonymous things have the same name and definition. For in order to meet this requirement the expressions for substantial kinds must be unambiguous.24 Accordingly, if these considerations concerning the ambiguity of expressions for kinds are correct, and if the seal and the terrestrial-aquatics like it are in fact, as I noted, the only genuine dualizers according to Aristotle, then dualization would not prove to be a problem for the theories of essentialism and classification of the logic and metaphysics.

Furthermore, I think that in HA VIII.2 Aristotle recognizes that through an appeal to ambiguity he could avoid the appearance of placing kinds in incompatible categories. This recognition on his part becomes apparent in a passage devoted to one dualizer in particular, the dolphin. In that passage Aristotle notes that one cannot say that the dolphin, and cetaceans generally, are merely terrestrial or merely aquatic, if "terrestrial" means "taking in air" and "aquatic" means "taking in water." For the dolphin partakes of both the terrestrial and aquatic nature: it takes in air through its lungs, and in pursuit of its food it takes in water, which it discharges through its blowhole. Here Aristotle treats "terrestrial" and "aquatic" as unambiguous and as opposing expressions; nonetheless the dolphin falls in both the kinds they signify. The dolphin provides then, one should think, the sort of dualizer that leads defenders of strong continuity to the position that substantial kinds are in fact continuous in the strong sense, rather than

23 In his note to HA 589a II, in his translation of HA in the Oxford series, D'Arcy W. Thompson maintains that "A[ristotleJ is hampered throughout this discussion by his habitual reluctance to coin a new nomenclature or to abandon a popular classification embodied in current speech; he clings to the terms ae6v and lvv8pov after they are shown to be unfit to connote the sought-for distinctions." In holding that Aristotle is reluctant to give up the classifications of ordinary language, Thompson no doubt has in mind such passages as PA 643blO-12, in which Aristotle appeals to the distinctions made by "the many" concerning animal kinds to furnish him with his genera (cf. APo. 98al3). HA VIII.2 reveals then how strong a grip ordinary language, including the language of the educated, has on Aristotle's thought; for he will continue to appeal to it in the development of a discussion even when he recognizes that on that occasion it presents him with considerable difficulties. 24 Cat. 3a33-b9, Top. 123a27-29, 127b5-7, 154a17-18; NE 1096a24-28.

197



incompatible and discrete: it possesses what appear to be essential attributes that place it in substantial kinds ostensibly mutually exclusive. But Aristotle offers a different interpretation, for he goes on to say:

Now it is unreasonable to assign these animals to both divisions, since they are opposed, but it seems that one must supplement the definition of "aquatic" (589b12-14; cf. GA 732b15-27, PA 642b18-20).

And Aristotle continues his discussion by showing how "aquatic," in ways now familiar, is ambiguous, so that the dolphin can without an in- compatibility be both a terrestrial and an aquatic animal.

With respect to the issues of continuity and dualization, the above quoted passage and the accompanying discussion of the dolphin are especially significant. For Aristotle explicitly rejects as unreasonable the assignment of a kind to genuinely opposed divisions, as we should expect any right-minded person to do. And where it appears that an animal, such as the terrestrial-aquatic, must belong to opposing categories, he does not say that these categories are therefore continuous, and thus for that reason are not really opposed. Rather, his remarks clearly reveal that he believes that the opposition is only an apparent one that will disappear, when the definitions of the expressions signifying the divisions are supplemented, when, in other words, the expressions are shown to be ambiguous, and in just what way ambiguous.25 Therefore, the issue that substantial kinds form a strong continuity because of alleged kinds that do not fall neatly under one or the other of two substantial kinds disappears. Hence there would be nothing extraordinary about dualizers. They would merely appear to partake in opposing natures, simply because of the ambiguity of the classificatory expressions. Or their dualizing with respect to genuinely opposed categories would arise in unproblematic areas, where, for instance, the characteristics in question are not essential or always possessed by all the members of the dualizing group. For example, only some swine are solid-hoofed, whereas others are split-hoofed; not every man, or at least not every man at every time, is both gregarious and solitary. One can mark off sharp borders between kinds, at least in the important cases, which concern substantial kinds, because one can make explicit the distinct senses conveyed by the same classificatory expression when it applies to incompatible or different groups. Hence, the theories of essentialism and classification of the logic and metaphysics are possible, at least insofar as

25 Balme notes this point about ambiguity, but fails to appreciate its important bearing on the continuity of kinds in the scala: "Dif.", pp. 189-190; see my comments on Balme in n. 2.

198



one can fix the borders between substantial kinds. Moreover, even if Aristotle cannot make unproblematic all the troublesome dualizers by an appeal to ambiguity, he nonetheless makes it clear in HA VIII.2 that he is unhappy with dualizers that appear to fall into incompatible divisions, and that an appeal to ambiguity, rather than continuity, is at least one way to prevent this from appearing to happen to some dualizers.

On several counts Aristotle's appeal to ambiguity to solve the problem of dualizers should not come as a surprise. It is one of his most important maneuvers for dealing with troublesome topics,26 and this consideration alone should dispel any initial doubt about the authenticity of the argument of HA VIII.2. More important, E qEorpitLv, which in following Peck I translate by "dualize," suggests that kinds that participate in distinct kinds do so through an ambiguity. For, as I noted in the above, it could mean "equivocate." A kind that participates in distinct kinds because the expressions for those kinds are ambiguous might be said, metaphorically speaking, to "equivocate" with respect to the expressions for those kinds. Some of his discussions of dualizers outside HA VIII.2 also suggest that Aristotle believes that some dualizers only dualize through an ambiguity. In the one extended discussion of dualizers outside HA VIII.2, the latter part of PA IV. 13, which I have already discussed, Aristotle says of cetaceans that they dualize between terrestrial and aquatic "in a sense" (TpO'rov TLVa, 697a29), which means that they dualize only in a qualified sense, perhaps only because of the ambiguity of "terrestrial" and "aquatic." In the same discussion the seal and the bat belong to both and to neither of the kinds they dualize between: 0p(poTEpWv TE ILETEXOVOL XaL

OV'?ET'pV (697b4). Elsewhere in PA the same description applies to the dualization of the ape between biped and quadruped (689b32-35). This odd characterization of dualizers might mean that, for instance, in one sense the seal is a terrestrial and not an aquatic, but in another sense it is an aquatic and not a terrestrial: hence it is both and neither, but in different respects. If this interpretation is justified, then Aristotle seems to think of the seal and animals like it as dualizing only through an ambiguity. Finally, in other areas both in and outside the biology, Aristotle clearly shows that he has no qualms about treating categorizing expressions as ambiguous. For example, at HA 487a 14-b7 Aristotle discusses most of the different senses of "terrestrial" and "aquatic" in much the same fashion in which he discusses them in HA VIII.2. Also, at HA 487bl8-20 he holds that birds and bees are "winged" or "aerial" in different ways (cf. PA 682b 18-21). In

26 E.g., Met. 992bl8-24, VII.4, 1043a29-b4, 1087a10-25, NE 1096al8-29, Phy. 233a21-26.

199



addition, in PA 1.3 "biped" and "blooded" differ in meaning when they apply to different sorts of animals (643a3-5; cf. 647b36). Outside the biology in the Topics and probably in the Categories, an expression for a differentia must be ambiguous if it applies to different genera, of which one is not subordinate to the other: "sharp" is equivocal when it functions as an expression for the differentiae of note and solid (Top. 107b 19-26, Cat. lbl6-20; cf. Top. 144bl2-30). In the Topics "life" is ambiguous: it means one thing when it applies to animals and another when it applies to plants (148a29-31). The ambiguity of classificatory expressions, including some very important ones, seems, then, often to arise in Aristotle.

PA IV.5 and HA VIII. I and 2 provide significant evidence in favor of the view that in his biology Aristotle continues to adhere to the discreteness of substantial kinds, countenanced by his views on essentialism and classification formulated in his logic and metaphysics, and that he does not abandon the discreteness of substantial kinds for a continuum of overlapping substantial kinds when he recognizes dualizing kinds and adopts the continuity of the scala naturae.27

Arizona State University

27 For helpful suggestions for improving this paper, I am indebted to Frank Lewis, Jonathan Barnes, James Lennox who commented on a version I presented to the 1983 Western Division Meetings of the American Philosophical Association, under the title, "On Dualizing Kinds and the Continuity of Kinds in Aristotle's Biology," and Stephen Leighton for his comments on a verison I presented to the 1982 Workshop in Greek Philosophy at the University of Texas at Austin.

200



the scala naturae and the continuity of kinds

Documents