Thanks to Eric Brown, Victor Caston, Brad Inwood, Fred Feldman, Calvin Normore, Mark Rollins, Margaret Scharle, Byron Stoyles, Paul Studtmann, and Martin Tweedale for discussion; the participants of my graduate seminar on the De Anima at the University of Alberta; and the auditors at papers delivered at the American Philosophical Association Pacific Division Meeting, the Canadian Philosophical Association Annual Congress, and the American Philological Association Annual Meeting. I gratefully acknowledge the support of the Social Sciences and Humanities Research Council of Canada Standard Research Grant #410–2008–0431.

1 See Kosman, L. A., “Perceiving that We Perceive: On the Soul III, 2,” Philosophical Review 1969: 499–519Google Scholar; Osborne, C., “Aristotle, De Anima 3.2: How Do We Perceive that We See and Hear?” Classical Quarterly 33 (1983): 401–11CrossRefGoogle Scholar; Caston, V., “Aristotle on Consciousness,” Mind 111 (2002): 751–815CrossRefGoogle Scholar.

2 All translations of De An. are from R. D. Hicks, Aristotle De Anima with Translation, Introduction and Notes (1907; reprint London: Cambridge University Press, 1965), except where noted.

3 This is in apparent conflict with his claim elsewhere that all of the senses “are accompanied by a common power, in virtue whereof a person perceives that he sees or hears (for, assuredly, it is not by sight that one sees that he sees … for there is one sensory function, and the controlling sense organ is one, though differing as a faculty of perception in relation to each genus, e.g., sound or colour)” (On Sleep, 455a 13 ff., emphasis added). In this passage, Aristotle aims to explain how it is that all of the senses go idle simultaneously when one falls asleep. The claim that higher-order perception is the function of a common-sense faculty, rather than the function of any of the special- sense faculties, is central to his explanation of this phenomenon. There’s a difficulty of interpretation, then, in reconciling De An. 425b12–25 and On Sleep 455a13ff., which I will ignore. My concern in the body of the paper is with the question, What is the activity which constitutes higher-order perception?, and not with the question, What faculty performs this activity?

4 Hamlyn, D. W., Aristotle’s De Anima Books II, III, Translated with Introduction and Notes (Oxford: Clarendon Press, 1968), 122Google Scholar.

5 Kosman, op. cit., 508.

6 One might hold, contra Aristotle, that we can perceive an item unawares. The distinction I am drawing here between perceiving awares and perceiving unawares is similar (modulo the comments above on misreading Aristotle as advocating a contemporary notion of consciousness) to those drawn—for example by N. Block, “On a Confusion about the Function of Consciousness” in Behavioral and Brain Sciences 18 (1995): 227–47, and M. Tye in Ten Problems of Consciousness (Boston: MIT Press, 1995)—between perceptual consciousness with access-consciousness and perceptual consciousness without access-consciousness. Contemporary higher-order theorists of consciousness draw on similar examples of nonconscious perceptual states. See Armstrong, D., A Materialist Theory of the Mind. (New York: Routledge, 1968)Google Scholar for the case of absent-minded driving, and P. Carruthers, Language, Thought and Consciousness (Cambridge: Cambridge University Press, 1996) for the case of blindsight. Such cases might be thought to lend support to higher-order theorists of consciousness. Aristotle is committed to denying that, strictly speaking, one can perceive unawares. For a discussion of how Aristotle might have handled blindsight cases, see Caston op. cit.

7 See, for example, Rosenthal, D., “Thinking that One Thinks,” in Davies, Martin and Humphries, Glyn W. (eds.), Consciousness (Oxford: Blackwell, 1993)Google Scholar, and P. Carruthers, Phenomenal Consciousness: a Naturalistic Theory. (Cambridge: Cambridge University Press, 2000).

8 Caston, op. cit.

9 See, for example, U. Kriegel, “Consciousness as Intransitive Self-Consciousness: Two Views and an Argument,” in Canadian Journal of Philosophy 33 (2003): 103–32. Kriegel holds a weakened self-representational view that while conscious states are not identical to the states that represent them, they nonetheless entertain constitutive, non-contingent relations to them.

10 Johansen, Thomas, “In Defense of Inner Sense: Aristotle on Perceiving that One Sees,” in Proceedings of the Boston Area Colloquium in Ancient Philosophy, Vol. 21, ed. Cleary, John and Gurtler, Gary (Boston: Brill, 2005), 244–5Google Scholar. Johansen characterizes his aim as a defence of the ascription to Aristotle of an inner sense theory, but his main concern is to argue that perceiving that we see cannot be a necessary condition of any ordinary perception but is instead a special kind of perception. So, although the label “inner sense theory” is often used to identify a theory of consciousness, Johansen holds that perceiving that we see is not Aristotle’s account of consciousness but rather is a special case of perception. Johansen’s characterization of his opponents’ view is also potentially misleading for similar reasons. Caston, for example, takes perceiving that we see to be a higher-order theory of consciousness, although he holds that the first-order state and the second-order state are one and the same token state. That is to say, the view which Caston ascribes to Aristotle is what is usually called an inner sense theory of consciousness. So it would be more in keeping with current terminology to say that Johansen denies, and not defends, the ascription to Aristotle of an inner sense theory.

11 Kosman makes a similar point in his “Commentary on Johansen,” in Proceedings of the Boston Area Colloquium in Ancient Philosophy, Vol. 21, ed. John Cleary and Gary Gurtler (Boston: Brill, 2005), 280–2. Johansen argues that perceiving that we see cannot be a necessary condition of any ordinary perception but is instead a special kind of perception on the grounds that first- and second-order perceptions are different in complexity. In the first-order perception, one sees colour or what has colour. In the second-order perception, one perceives also that one is seeing colour. Johansen holds that the difference here is not well characterized by claiming that first-order perceptions are non-propositional while second-order perceptions have propositional content. For Aristotle, Johansen claims, can express the content of a first-order perception by a proposition, writing at 418a15, for example, “that [there is] color.” Moreover, first-order perceptions can involve accidental or common sensibles and so have content describable only with propositions such as that expressed by “this white thing is moving towards me.” Rather, first- and second-order perceptions are different in complexity in the sense that “the second order perception conveys not just the content of the first order perception but also the information that one is perceiving that content” (op. cit., p. 258). However, it is precisely the point of contention between Johansen and authors such as Kosman whether the first-order seeing in expressions like “perceiving that one sees” is a conscious perception of colour or what has colour.

12 412a10 suggests calling this stage first actuality since it is the actualization of first potential; 417a21ff. suggests calling this stage second potentiality since second actuality is its actualization.

13 Hicks’s translation misleadingly contributes to the appearance of the distinction as being between two faculties. Hicks translates: “[A]nd to the one intellect (ho men toioutos nous), which answers to this description because it becomes all things, corresponds the other (ho de) because it makes all things.” However, instead of taking toioutos as adjectively modifying nous, we might just as naturally take it as a substantive. Then the passage reads: “[T]here is some aspect by which intellect becomes all things and another aspect by which (the one and the same) intellect makes all things.” Since the Greek supports either reading, this passage fails to provide evidence against the claim that the active intellect is not a separate faculty. In his commentary on this passage, Hicks, op. cit., 500, notes the possibility of two readings and their significance for the interpretation of the distinction. However, Hicks analyses the possibilities, both philological and philosophical, somewhat differently than as I have in the body of the essay. Hicks takes toioutos to be predicative and standing for both “passive” with ho men and “active” with ho de. He here translates the passage as: “[T]he one intellect is passive, like matter, in that it becomes all objects, the other intellect is active, like the efficient cause, in that it makes all objects.” Hicks continues: “If toioutos were attribute and not predicate, estin must mean ‘there exists’ and the sense must be ‘passive intellect exists in so far as it becomes all objects, active intellect, in so far as it makes all objects.’ Those who press this interpretation deny that A[ristotle] ever really taught the existence of two distinct intellects in the sense in which the art which constructs is distinct from the material which it works upon: they content that A[ristotle]’s one intellect is sometimes passive, sometimes active, as it is sometimes theorētikos, sometimes praktikos.” I have offered a somewhat different reading of the Greek and a different claim about the unity of the intellect. In particular, I claim neither that the distinction is temporal nor that it is related to the distinction of theoretical and practical reasoning.

14 I don’t claim any originality for this thesis. For antecedents, see Brentano, F., “Nous Poietikos: Survey of Earlier Interpretations,” originally published in Die Psychologie des Aristoteles (Mainz: Franz Kirchheim, 1867)Google Scholar, translated in and quoted from Essays on Aristotle’s De Anima, ed. Nussbaum, and Rorty, (Oxford: Clarendon Press, 1992), 322Google Scholar; and Kosman, L. A., “What Does the Maker Mind Make?” in Essays on Aristotle’s De Anima, ed. Nussbaum, and Rorty, (Oxford: Clarendon Press, 1992), 355Google Scholar. This paper’s contribution is to defend this line of interpretation through several new considerations: consideration of the analogy Aristotle often draws between perception and thought; considerations of the local argument of De Anima surrounding relevant passages such as 425b12–25; and considerations of issues within these passages.

15 See his “Aristotle on the Mind’s Self-Motion” in Self-Motion: From Aristotle to Newton, ed. Gill, M. and Lennox, J. (Princeton: Princeton University Press, 1994), 81–116Google Scholar. For other considerations against surge theory, see Wedin, op. cit.

16 For discussion, see Caston, V., “Why Aristotle Needs Imagination,” Phronesis 41 (1996): 20–55CrossRefGoogle Scholar.

17 For a recent advocation of this latter view, see Caston, V., “Aristotle’s Two Intellects: A Modest Proposal,” Phronesis 44 (1999): 199–227CrossRefGoogle Scholar.

18 See, for example, Morrison, D., “Separation in Aristotle’s Metaphysics” in Oxford Studies in Ancient Philosophy, vol. 3 (Oxford: Oxford University Press, 1985),125–58Google Scholar; Spellman, L., Substance and Separation in Aristotle (Cambridge: Cambridge University Press, 1995)CrossRefGoogle Scholar; and my “Aristotle on Ontological Dependence,” Phronesis 53 (2008): 65–92.

19 An anonymous reviewer for this journal objects that the active intellect is indeed dependent on the body since it presupposes the possession of concepts by the passive intellect, and concept acquisition depends on the body. However, even if the possession of a body is a precondition for the activity of the human active intellect, and so arguably the human active intellect cannot exist without a body, it does not follow that the active intellect is dependent on the body in the relevant sense of dependence. In general, not all necessary conditions are causes or asymmetric dependency relata.

20 Caston, op. cit.

21 Caston, op. cit.

22 Caston, op. cit., 201.

23 Caston, op. cit., 202.

24 This is not the place for a detailed discussion of Caston’s careful and innovative interpretation. It will perhaps suffice for my present purposes to show that some of the textual considerations which Caston brings to bear on the issue do not immediately refute my position. First, Caston holds that the taxonomic character of the discussion at 430a10–17 suggests that subsequent division of the rational soul into a passive and an active intellect is a division between two natural kinds of the soul and not two faculties or divisions within the human soul. The argument at 430a10–17, as Caston notes, rests on the premise that it is because natural kinds exhibit certain characteristics that rational souls do as well. However, this premise suggests only that the intellect is a natural kind, and not that the active and passive intellects are themselves distinct natural kinds. So while I agree with Caston that the active and passive intellects are not two distinct faculties, I disagree that they are not divisions within the rational soul. Second, Caston notes that Aristotle characterizes the active intellect and the divine intellect similarly. Both the active and the divine intellects are described as separate (430a17 and 1073a4, respectively), impassible (430a18 and 1073a11), unmixed or without matter (430a18 and 1074a33–34), actuality (430a18 and 1072a25–26; 1072b27–28), the same as its object (430a20 and 1075a1–5), prior in time to a capacity (430a21 and 1072b25), uninterruptedly or eternally thinking (430a22 and 1075a10), and solely what it is or just its essence (430a22–23 and 1075a1–5). But the active and divine are also contrasted in ways that suggest they are distinct: for example, the active intellect is called “more honourable” (430a18) while the divine intellect is called “most honourable” (1074a26; 1072a35–b1; 1072b28); the active intellect is the necessary condition for all thought (430a25) while the divine intellect is the necessary condition for everything (1072b13–14; 1075b24–26). So, although Aristotle characterizes the active intellect in ways that draw a surely intentional comparison with the divine intellect, some of these characterizations are striking in their differences. At best, we can conclude that the active intellect is like the divine intellect.

25 See for example APo. 2.19 (100b15) and EN 6.6 (1141a7–8).