2 Aristotle's main discussion of sense perception is in De Anima IL5-12 and III 1-2 and De Sensu, along with other discussions in the Parva Naturalia. There are also important references to perception in the Metaphysics, the Physics, Generation of Animals, Nicomachean Ethics, and other treatises. These works seem to be in basic agreement on the issues discussed here, in spite of some differences of detail. 3 In translating I use 'perception' and related English terms for aïerOl1erl!> and related Greek terms, when the context indicates that there is perception of an external object; and 'sensation' and related terms when there is no external object, as in the case of an after image (cf. An. III 2, 425b25). I follow standard English usage in
using 'sense' for the five senses: sight, hearing etc.; and 'sense organ' for eye ear, etc.
4 Aristotle uses the example of the white thing which happens to be the son of Diares. 5 See An. III 3. 427b 1 I -1 Z 6 , 430b29; Sen. 4. 442b8-10. This claim of infallibility is qualified at An. III 3. 428b18ff "or admits the least possible amount of falsehood." Errors occur when the sense organ has been affected in some way: e.g. "Sick persons find everything they taste bitter, because, when they taste, their tongues are overflowing with bitter moisture" (II 10, 422b8-10). Moreover; "when we have looked for a long time at one color, e.g. at white or green, that to which we next transfer our gaze appears to be of the same color" (De lnsomniis 2, 459bull-13).
6 Kahn 1992, 364-5 and 367-8 points out the importance of distinguishing the topics which Aristotle discusses under the heading ofai'er917erl'i'. I will briefly mention the common perceptible objects later on, but I will not attempt to include accidental perceptible objects, because it is not germane to the main topic of this paper. I plan to take up the perception of common and accidental perceptible objects, along with 4>avraaia (imagination or appearance), in a future essay. 7 He says, "Perception seems [or is held to be] to be a sort of alteration (dAAoKoo-tyTt?)" (II 5, 416b33-4). There is a back reference to 4, 415b24, which is also guarded: "Sensation is held to be a sort of alteration." In both instances he uses bond, 'seems' or 'is held to be'; and &XXOIWO-99Tty, 'a sort of alteration.' (Cf. Insomn. 2, 459b4.) As with the English 'a sort of,' Tig can have an indefinite or qualified sense, e.g. "he's a sort of friend." It can also indicate a specific kind of change, cf. Phys.III 1, 201bl3.
8 Contrast An. I 3. 406al2-13, which enumerates four kinds of change distinguishing growth and diminution as separate kinds. 9 Aristotle actually has two words for 'change' in general, /«Ta/3o\7J and Kivrlms. which he often uses interchangeably (cf. Phys. III 1, 201bl2; IV 10, 218b20). But /ACTa,6oX� has a broader application: e.g. a process with an unspecified starting-point, such as coming to be red from having no color, is a �.era,8o��j but not a Kivrlms (V 5, 229b10-14). English translators customarily use 'movement' or 'motion' for Kivrlms, in order to distinguish it from �era�8o�rj, which is rendered as 'change.' Though understandable, this can lead to misinterpretation because the English terms 'movement' or 'motion' suggest locomotion. Aristotle calls locomotion 'the primary change', VII 2, 243all, and often uses forms of Kivnms in some cases where he makes clear he is talking about locomotion (e.g. An. I 3, 406bl). In order to avoid confusion, I translate Kivrys as 'change' unless locomotion is clearly intended.
10 He further notes that there is a difference between being acted on (7TåCT)((IV) or moved (Ktv4E!a-Oat), and being active (evepyeiv), for movement (Kivrlms) is a sort of activity (ivipyeta TLT); but this is a distinction which he ignores for the time being. (An. II 5, 417a14-17; cf. Phys. III 3, 201b31, VIII 5. 257b8; cf.Metaphysics K 9, 1066a20.) We shall return to these matters below (see III C). 11 Aristotle shifts here from kvipyaa to £vHÀ£xua and his remark "for we are now speaking without qualification about them" (417a21-2) implies that the two terms are interchangeable in this context. I translate ivipyeta as 'actuality' when it is contrasted with Swajmy, and as 'activity' when it is contrasted with ki'i/17VT(\4yfia as 'actualization.' I do not, however, intend to convey by 'actualization' that an ivrtkiycia is as such some sort of process. t2 Cf. Metaph. I' 5, 1010b32-3. The Greek verb 7r4o-Xctv (aorist, 7Ta8âv) means 'to be acted upon' or 'to be affected.' It is opposed to 7TpåTHLV or 7Tottiv , 'to act' or 'to make.' The derived noun 7Tå8of> is ordinarily translated 'affection', and as 'passion', 'emotion', or 'suffering' in connection with human subjects. It also has a wider sense of 'attribute' (e.g. An. I 1, 402a9, When the a'ier8T/p.a remains in the sensory system and is not directly caused by an external object, it is more appropriate
to translate it as 'sensation' (cf. Insom, 2.460b2-3 and n3 above). Such residual sensations are the source of dream images and memories (cf. Mem. 1. 450a31; Insom. 3. 461a19, b22). 13 Cf. Bynum 1993. 92. The term 'automatically' is used in our sense-i.e. involuntarily, or without conscious thought or volition—not in Aristotle's sense of ro a�T61iaTOV, which corresponds to our notion of chance or spontaneity (cf. Phys. II 6).
14 ran. II 5. 417al2-20. 418a3-6; cf. II 4 416a29-b9. 15 See the useful discussion of Modrak 1987, 24 et passim, who calls this Aristotle's 'Actuality Principle.' 16 See Everson 1997, 20 ff. for a persuasive argument that Aristotle treats color as by itself qca(J' aLT6) the cause of perception, and that he does not view color as simply visible by definition.
17 Aristotle argues that touch is exceptional in that the medium is a part of our bodies, our flesh, through which tactile perceptions occur (An. II 11, 423a15-17). This helps explain why Aristotle talks about sight first and touch last, even though touch is the most common and basic sense. Touch has some characteristics which are apprehended only by comparison to other senses.
18 This passage illustrates Aristotle's practice of interchanging terms which are elsewhere distinguished. Here he attribute a property to sight (oys) and the pupil (Kdprl), which more strictly belongs to the eye (o��a). This is an example of synecdoche, using a part to represent the whole. 19 The composition of the other organs is less clear: the descriptions in An. III 1, 425a3-8. III 13, 435all-bl; and Sen. 2. 438bl6-439a5 seem inconsistent in some respects.
20 I am greatly indebted here to the discussion of Sisko forthcoming, although he presents the material somewhat differently. See also Sisko 1996. 21 Proponents of the literalist interpretation typically claim that "the K6p?7 is in fact the eye-jelly inside the eye" (Sorabji 1974, 72n22; followed by Everson 1997, 80). Even Johansen 1998. 56, a critic of literalism, says that Aristotle defines the Koprl as the transparent liquid in the eye. But the passages typically cited do not decisively support this interpretation: e.g. Sen. 2, 438b8-16 discusses "what is called the pupil" or "the so-called pupil" (rilv Ka?LOVIAiV77VK6pl7V) (cf. Generation of Animals V 1, 780a27). The word K6p7l refers primarily to a girl and by extension to a doll. As Liddell. Scott. Jones 1996 981 remark, it also refers to "the pupil of an eye because a little image appears therein." This derivation of K6p77 was already mentioned by Plato in Alcibiades I 132e7-133a3. The Latin pupula has a similar etymology. See also History of Animals 9. 491b20-1 where the Koprl is surrounded by the dark iris (jncXaf) and this in turn by the white (XCVK6V) of the eye. Here Aristotle is evidently describing what one sees on looking directly into the eye from the front. In some cases Aristotle uses Kdprl for the entire eye by synecdoche (cf. n18 above).
Everson 1997, 10; Sorabji 1974 and 1992; Slakey 1961. � Burnyeat 1992, 21 and 1995; Broadie 1992; Johansen 1998. 24 Burnyeat, 1995, 425. 25 Ibid. Sisko 1998, 342-3 criticizes a possible literalist reply to this objection. 26 Sorabji 1992, 214. Defending the spiritualist interpretation, Johansen 1998, 219-20 argues that Aristotle is not talking about a literal mean but about "a functional mean, " e.g. "the best ability to function as an organ of sight in the circumstances." Blind spots will occur when the organ lacks this ability. The mean of the eye depends on its degree of transparency, and the mean of the tongue on its mean state of moisture. This seems a forced way of understanding the claim ofAn. II 11, 424a2-10 that the sense of touch is sort of mean between the opposite perceptible qualities, i.e. hot and cold, and wet and dry. 27 Sisko 1998, 347. There is also the problem of afterimages and other perceptual residues; cf. n5 above. 28 Bumyeat 1995, 429-30 infers from the fact that the move of a wave or vibration is not locomotion to the conclusion that it is a quasi-change. But this omits the
alternative that vibration is a real change of some sort (e.g. the parts moving successively up and down) other than locomotion of the entire movement. Johansen 1998, 156 follows Bumyeat. 29 "The sense-qualities may affect bodies without perception, for the sense- qualities may affects them so as to make them perceptible too." Johansen 1998, 274-5 (his emphasis). This is also his unpersuasive interpretation of the fact noted by Aristotle that oil and wine are acted upon by nearby objects and take on their smells. As Aristotle observes, "they take on not only the smells of things thrown into them or blended with them, but also of things that are placed or which grow near the vessels containing them" (Insom. 2, 460a28-32; cf. Everson 1997, 132n58). But, since the original source of the smell may have perished it is incredible to suggest that smelly oil differs from unscented oil only in being perceptible. 3° Tumbull 1978, 22 suggests, "The eye takes on a ratio that is formally akin to; or, if you please, identical to a certain ratio in the thing." Cf. Silverman 1989. Shields 1995 also suggests an interpretation along these lines according to which perceptual states are "isomorphic" in some weak sense with their objects. 31 Silverman 1989, 279. Everson's 1997, 96-9 objections against Silverman 1989 on this point are not decisive. Price 1996 also advances persuasive objections against the literalist interpretation, and defends an analogical interpretation akin to Silverman's. See also Bradshaw 1997 and Ward 1988.
� Burnyeat 1992, 26. Broadie 1992, 158 n2 defends Burnyeat's own interpretation against the charge of dualism (nl2). Nonetheless. Burnyeat does defend a sine qua non of the dualist interpretation: perceptual awareness as such involves no physiological change. 33 A thing undergoes a "Cambridge change" when it undergoes a change of relational properties, which may involve no "real change, " i.e. no change of non- relational properties. Burnyeat 1995, 424 suggests that according to Aristotle the medium undergoes only Cambridge but not a real change: "When the sun rises or a lamp is lit, for the air around us this is a mere change of relation. The statement 'When fire comes to be present in the air, the air is illuminated' is just like the statement 'When I move to the left of my desk, the desk comes to be on my right."' Johansen 1998, 136-47 offers an interpretation similar to Burnyeat's but tries to avoid the implication that the medium undergoes a Cambridge change. On his account the medium and the sense organ undergo "phenomenal" change-but this is still not a real change: the medium and sense organ take on the quality of the sense- object only insofar as the sense-object appears to a perceiver.
34See Slakey 1961.
3s Thus. Everson 1997, 101: "Aristotle's point is not that there is no material alteration but that the alteration can be described without reference to the material constitution of the agent of that alteration." � See Scaltsas 1996. 28-9. 3� It is not altogether clear what the problem is with plants.Everson 1997, 87 offers an explanation of plant insentience along literalist lines, which is criticized by Sisko 1998, 338 nlO.
� Editors and interpreters disagree over whether the passage terminates with a question mark or period. The parallelism of at b17 with the art b14 makes a terminal question mark more likely. 39 The conclusion of the passage involves a textual problem whether the word Kai belongs in Aristotle's text at 424b17. As Kosman 1975, 509-11 points out, Kai is an editor's conjecture; no manuscript has Kai and only one manuscript contains the letters at, which may be a scribal error. Ross accepts Kat in the Oxford Classical Text (1956) but not in his subsequent edition (1961). If Kai does not belong, it would seem that either Aristotle himself falls into inconsistency here, or the simple materialist interpretation is mistaken. Slakey 1961 opts for the former, Burnyeat 1992 for the latter. But, even if Kat can be read, problems remain for the simple materialist interpretation-since it implies that perception involves something in addition to material alteration. But, aside from the Kai controversy, the occurrence of Trapot at 424b17 also implies that smelling is something over and above being affected in a material way. See Sorabji 1992, 219-20 andScaltsas 1996, 34. 40 See Bynum 1993. 107: "Aristotle's theory of perception is 'purely physiological'" and "is a notable philosophical achievement that contemporary materialists would do well to investigate." See Charles 1984 for a comprehensive defense of this view of Aristotle's psychology.
41 Recent defenses of functionalism are found in Wedin 1988, Shields 1990, Irwin 1991, and Nussbaum and Putnam 1992. However, the functionalist interpretation does not strictly entail a non-reductionist materialist interpretation: Shields for example defends a supervenience interpretation. Critics of functionalism include Granger 1990, Code and Moravcsik 199Z and Olshewsky 1992. 42 See especially Sorabji 1971 and 1992. a3 Such interpreters often read Kai at An. II 12, 424b17(Sorabji 1971, 70). According to Everson 1997, 129 and 130n55, "what distinguishes cases of mere alteration from the perceptual case is that the second involves the activity of perception. This saves us from the need to emend the text at 424b20 (by excising Kat) in the way in which Burnyeat is forced to (1995, 25, following Kosman 1975)."This is misleading, since Kai is an editor's conjection (see n39 above). In any case a non- reductionist interpretation does not require the reading of KaL: to wit Kosman 1975, 517-18. 44 See Parts of Animals II 10, 656b16-18 and GA II 6. 743b36-744a5. The passages (7TÓPO() from the nose and ears are filled with breath (7rveDlAa) and lead to small blood vessels around the heart, which proceed in turn to the heart. The passages from the eyes also tie into such blood vessels. Sen. 2, 438b3-16 also implies that the passages play some role in the actualization of the transparent contents of the eye,
analogous to a light source. Aristotle infers that the soul is located inside the eye rather than at its surface. � Metaph. Z 10. 1035b25-7 mentions bodily parts "which are controlling (Kvpla) and in which the formula, i.e. the substance, is primarily present, e.g. perhaps the heart or brain." Cf. also De Motu Animalium 10, 703a36-bl. 46 The special senses depend on it, in that when it becomes powerless, for example due to sleep, "all the other sensory organs must be unable to perceive" (Som. 2, 455a33-bl2) 47 Aristotle also maintains that we perceive the common sense objects (movement. shape, time, etc.) by means of a common sense (KolV1¡aïcr&r¡crl�). and he holds that these qualities are not perceived accidentally (koto (n>/ij8e/3eicdy) by this faculty (An. III 1, 425al4-28). It is tempting to identify this "common sense" with the common power (KOIV1¡ 8wa/«s) mentioned inluv. 3, 469alO, because Aristotle speaks of the common sense as belonging to the primary sense organ, which is also involved in imagination or appearance ((PaVTao-&'a): "One must cognize magnitude and motion with the [faculty] by which we cognize time, and the image (�avraa-pa,) is an affection of the common sense (KOmi7 at'o-�o-t?). Thus it is clear that the cognition of these things belongs to the primary perceptual faculty (aiCT87)TlKÓV). But memory
even of the objects of thought is not without an image. So that memory belongs accidentally to the thinking faculty, but by itself (Ka9' aLT6) to the primary sense organ" (Mem. 1, 450a9-14). But the question of whether the common perceptible objects (shape, movement, etc.) are perceived by such a common sense, as opposed to being perceived by more than one special sense, is very controversial. I will not pursue this issue here, because it is not crucial to the argument of this essay, although I plan to take it up in the future. 48 Compare Kahn 1966, 77: "... in so far as Aristotle's view implies a primary localisation of the act of awareness in the central organ, it presents a striking analogy to some contemporary theories of the identity between states of consciousness and neuro-physiological activity. But whereas most modem theories of identity tend to reduce the psychological aspect to the physiological, Aristotle would regard the bodily processes in sensation simply as the immediate instrument or `mattet' whose form and final completion (iVTEXgXfLa) is provided by the act of perceptive awareness." 49 Reading �v at 431a7 with C, although it is missing from several manuscripts. If correct, it refers back to II 5. 417al6 where Aristotle held this distinction in abeyance (see nlO above).
50 Cf. Ackrill 1965, 125: it is "a contrast between activities whose character sets a limit to their continuance and activities whose character sets no limit." 51 Heinaman 1990, 92-9 gives a good discussion of this distinction and its relevance to the relation between perceiving and the alteration of the sense organ. Everson 1997, 255 recognizes the basic point-by application of Leibniz's Law, perceiving is nonidentical with the alteration of the sense organ-but he curiously makes no reference to An. III 7, 431a4-8. Ackrill 1965, 140-1 contends that the distinction between evipyeia and Kíll71crL�. at An. III 7 431a4-8 is different from that in Metaphysics O 6 and Nicomachean Ethics X 4. From the fact that An. III 7, 431 a4- 8 refers back to An. IIS. 417al6-17, Ackrill infers that Aristotle has in mind the distinction made elsewhere in An. II 5 between acquiring an ability and exercising that ability (i.e. the distinction between change and quasi-change). However, An. II 5, 417al6-17 itself mentions that the distinction between £1I£PY£la andKL'v?7a-ts, has been made elsewhere. This is evidently a reference to Phys. III 2. 201b31 (see Hicks 1907. 353 and Ross 1961, 235), which as Ackrill 1965, 138-40 himself recognizes, makes basically the same distinction as Metaphysics 0.6 and Nicomachean Ethics X.4.
52 See Shields 1988 and Caston 1993. Cf. Kim 1979 and 1982 on modem supervenience theory. 53 Everson 1997, 11. 54 Everson 1997, 278 offers an Aristotelian formulation of epiphenomenalist supervenience: tV�V<�(<�.)0?FC!�&Vy*C!�o
which Aristotle's sensible qualities determine alterations as he says they do." Assuming that a particular event consists in a property exemplified by a particular substance at a particular time, the supervenience of properties (e.g. being angry) implies the supervenience of events (e.g. Thrasymachus' being angry at time to). It should be noted that the above account invokes the patently non-Aristotelian notion of a "possible world." 55 Everson 1997, 281. It has been objected that this interpretation conflicts with Aristotle's notion of hypothetical necessity, which implies that form necessitates matter rather than the reverse (Phys. II 9 and PA I 1). Everson handles this objection fairly convincingly (pp. 239, 272-9). 56 Ibid., 284. 57 Ibid., 286.
58 This argument occurs in the first book of De Anima where Aristotle is criticizing his predecessors and speaking in a nontechnical way. He allows for the sake of argument that perceiving and other psychological affections are changes (Kivrjo-eis) (I 4, 408b1-11). 59 According to Bynum 1993. 94, Aristotle holds that "the sense-organs are completely passive and highly selective" (his emphasis). As I have argued, however, perceptions have an active as well as a passive aspect for Aristotle. The animal is acted upon by the perceptible object through its sense organs (passive aspect) and it discriminates the object's perceptual form with its perceptual faculty (active aspect). Like Bynum, Everson 1997, 82 glosses An. II 11. 423b27-424a7 as "a sense organ
discriminates its proper objects" (my emphasis). But Aristotle says here, precisely, that the sense (ata-tlno-tg) does the discriminating. Properly speaking it is the percipient organism which discriminates, but it does this in virtue of its sensory power. 60 Ebert 1983 also remarks that Aristotle understands "the active side to sense perception" in terms of discrimination. Victor Caston called this valuable article to my attention in his commentary. My appreciation of the importance of discrimination in Aristotle's theory of perception deepened substantially through discussions with a graduate student, Christopher Pece, who wrote a fine term paper on this subject. 61 Cf. An. III 9, 432al5-16: the discriminative faculty (T6 itp&Ttic6v) is the function of perception as well as cognition (8titvota). 2 In a future paper I plan to discuss the place of perception, together with imagination and thought, in the causation of the movement of animals. � The basic meaning of xptfttf is 'separate' or 'put asunder.' It can also mean 'judge' or 'decide' as in the acts of a jury or assembly (Rhet. II 1, 1377a221-2; Pol.
IIL14, 1285be 15, 1286a26). Aristotle uses the term for moral judgment (EN III 5, 1114b7) and for judgments about unjust acts (IV 8, 1135b26). At EN V $ 1128al2 Aristotle evidently compares discrimination of moral character to the (perceptual) discrimination of bodies by their movements. As Ebert 1983 demonstrates, the term means 'discern' or 'discriminate' in perceptual contexts, rather than 'judge.' Ebert (188) also shows that "the underlying idea ... of singling out and of separating by singling out" can be traced to Homer. reading TLvt at 426b14 with UXT°and Barnes' revised Oxford translation. 65 Note that this passage ascribes perceiving that one sees and that one hears and perceiving the special objects of the different individual senses to the same discriminating faculty. When Aristotle talks about perceiving that one sees, he evidently has in mind the ability to distinguish seeing from hearing and other
perceptual activities. Hence Everson 1997, 137-48 does not seem to be justified in equating the perceiving that one sees with the awareness of a material alteration in the eye. 66 The analogy in this passage between discriminating and stating suggests that perceptual discrimination involves propositional attitudes, i.e. that to perceive is to perceive that something is the case. Similarly, Kpiuem may take objects ('to discriminate x from y') or a proposition ('to discern that p'). This is also suggested by An. III 2, 425b12-14 ("it is through sense that we perceive that we are seeing or hearing") and Som. Z 455al6-18 ("it is not by sight that one sees that one sees, and one discerns and is capable of discerning that sweet things are different from white things not by taste nor by sight, nor by a combination of the two, but by some part common to all the sense organs"). This leads to a difficulty because brutes have perception but no koyos (An. III 3. 428a23-43; cf. Politics I 2. 1253a9-10), which implies that they cannot form true or false thoughts or statements (cf. An. III 3. 427b12-14). Although Aristotle observes that some animals can use sounds as signs, e.g. of pleasure or pain, to one another (Pol. I 2, 1253a10-14), he would not agree that they can form true or false propositions, since this requires the intellect: the ability to form thoughts �O�J.U1Ta) and to combine them into a unity (An. III 6, 430a26-8). I cannot address this difficulty here, except to suggest that the above texts only commit Aristotle to the weak thesis that perception is implicity propositional. For example, when I discriminate the apple's redness, I am warranted in asserting that the apple is red.
67 Note that this premiss is reminiscent of Plato's hypothesis of opposites in Republic IV 436b8-cl. The hypothesis is crucial to Plato's argument that the soul is divided into three parts. I discuss this hypothesis and its logical implications in Miller 1999. � Cf. An. III 7, 431a17-20: "the air acts on the pupil (KÓpr¡) in a certain way and this acts on something else, but the ultimate thing is one, i.e. a single mean, and its being (eiwi) is numerous." The ultimate thing (EO-Xarov) referred to here is evidently the central discriminating faculty. 69 1 discuss this sort of treatment by Aristotle of the point in Miller 1973.
�° See Hicks 1907, 451.
71 Everson 1997 in contrast ascribes a representative theory of perception to Aristotle, according to which "the subject is aware of an icon, which represents an external object, or objects, to him." The supposed icon is the aicr611JJ4 understood as a "mental picture" which represents the external object. (203) For example, instead of perceiving a red apple strictly speaking, I am really aware of an internal picture which represents the external apple. But the evidence marshaled by Everson only shows that the affection in the sense organ is part of the causal chain of perception, not that it is the direct object of perception (e.g. Insom. 461b22-6). The direct object of perception is what initiates the process leading to the perception. See Metaph. 1' 5, lOlOb35-lOl 1a2: "Perception is not of itself but there exists something different and apart from perception, something which is necessarily prior to perception because that which brings about change is by nature prior to that which is made to change." Normal cases of perception must be distinguished from cases where the affection remains, as in a dream or hallucination, and causes a quasi-perceptual experience. In the latter case the affection initiates the process and is the direct object.
72 The emergence theory of mind is a minority view among contemporary philosophers of mind, but it does have defenders, including notably Searle 1992. For a clear exposition of emergence theory see Broad 1923, 61-9 and Kim 1998, 226-32. Basal materialism is evident in Broad's example of the emergent dynamical properties of a chemical compound. Even if one grants that the properties of the compound are "completed determined" by those of its elements, in the sense that a compound has those properties when, and only when, these elements are arranged in certain proportions and relations, it may still be the case that the law connecting the properties of the whole with those of its parts is "an unique and ultimate law" (his emphasis). This feature of Broad's account is put somewhat differently by Newman 1998: "emergent properties are not reducible to material properties. although "emergent properties are explicable in terms of material properties." Again, according to Kim 1998, 230, "the emergentist would have no problem with laws connecting emergent properties and their underlying lower-level bases. Indeed, they are committed to there being such laws: When appropriate `basal conditions' are present, emergent properties must of necessity emerge" (his emphasis). See also Kim 1992. � The term 'epigenesis' derives from the verb È7TLYLYV(crOaL. which is often translated 'supervene' but literally means 'comes to be upon.' I use the terms 'epigenesis' and 'epigenetic' here to distinguish my interpretation from supervenience and emergence interpretations. Although these terms are frequently used by modern biologists, the psychological theory of epigenesis discussed here-that psychic capacities or events are epigenetic states of the body-must not be confused with the embryological theory of epigenesis-that an embryo develops organs and organic systems from a seed containing essentially unorganized material through a succession of processes. The latter theory (defended by William Harvey
and other modem embryologists) was opposed to preformation, the discredited theory that growth is merely an increase in size of an initially fully formed seed. It is commonly held that Aristotle also anticipated the embryological theory, e.g. "Aristotle portrayed his epigenetic view of ontogeny as a series of increasingly higher 'souls'-nutritive, sensitive, and rational-entering the human embryo during its development" (Gould 1977, 15). Although they are distinct there are obvious affinities between epigenesis in Aristotle's embryology and psychology. 74 This is rightly emphasized by Burnyeat 1995, 422-3 and Johansen 1998. 157. But this fact does not require their spiritualist interpretation. Bynum 1993 in contrast exaggerates the extent of "microstructure" in the sense organs.
75 Victor Caston in his commentary to this paper rightly points out the affinity of epigenesis with modem vitalism. Indeed, the final great exponent of neo�italism in the twentieth century, Hans Driesch 1914 claims Aristotle as "the first exponent of a scientific 'vitalism'" (11). Aristotle's philosophy of soul does fall under Driesch's s general rubric of vitalism, i.e. "the theory of the autonomy of the processes of life." Moreover, the neo-Aristotelian term 'entelechy' plays a central role in Driesch's s theory. Driesch remarks, however that this term "though well known in the metaphysical terminology of Aristotle, is not here used in the proper Aristotelian sense" (203). "We shall use this word only as a sign of our admiration for his great genius; his word is to be a mould which we have filled and shall fill with new contents" (1908, vol. I, 144). Driesch states. "Entelechy is an agent sui generis. non- material and non-spatial, but acting 'into' space so to speak" (1914, 204). Further, as an immaterial agent with "a non-spatial nature, " the entelechy is a sort of quasi- substance exercising quasi-causality at certain places in the organism, like Descartes' extensionless soul acting through the pineal gland (1908, vol. 11 259-60). Broad 1923. 612 aptly described this theory as "the biological analogue of Cartesian Dualism in psychology." The interpretation of Aristotle which I call 'epigenesis' should not be confused with Driesch's view. But as Broad also points out vitalism may take different forms: substantial vitalism of the sort espoused by Driesch and emergentist vitalism which takes biological properties to be irreducible to chemical properties. The interpretation I favor here might be called 'epigenetic vitalism.' 76 Including Heinaman 1990, Robinson 1983, and Scaltsas 1996. 77 Caston 1997. 337. See also Newman 1998.
Caston 1997, 353. 79 The case is clearer regarding the nutritive faculty, concerning which I defend an epigenetic interpretation in Miller forthcoming. see especially Code 1987, and Code and Moravcsik 199Z with which my interpretation is intended to be in close agreement. 81 Perhaps spontaneous generation could be explained by Aristotle in terms of emergence. Some living things arise spontaneously out of matter containing residual vital heat. Although they do not have an ensouled parent, they simulate the behavior of living organisms. See Metaph. Z 7, 1032a3Q 9, 1034b4-6. 82 I am grateful to Victor Caston. Lawrence Jost, and an anonymous referee for valuable criticisms of earlier drafts of this essay. I also received helpful comments when I presented it at Boston College and Bowling Green State University. Finally, I am indebted to my colleagues in the Greek Philosophy Reading Group, based at the Ohio State University, which devoted several years to the close study of De Anima.