1 ‘AT’ refers to Rene Descartes, Oeuvres de Descartes, 11 Volumes, Charles Adam and Paul Tannery, eds. (Paris: J. Vrin 1996); ‘CSM’ refers to Rene Descartes, The Philosophical Writings of Descartes, 2 Volumes, John Cottingham, Robert Stoothoff, and Dugald Murdoch, eds. (Cambridge: Cambridge University Press 1985); ‘CSMK’ refers to Rene Descartes, The Philosophical Writings of Descartes: The Correspondence, John Cottingham, Robert Stoothoff, Dugald Murdoch, and Anthony Kenny, eds. (Cambridge: Cambridge University Press 1991).

2 For instance, see Anthony Kenny, Descartes: A Study of His Philosophy (New York: Random House 1968), 104-5; Robert McRae, ‘Innate Ideas,’ in Cartesian Studies, R.J. Butler, ed. (Oxford: Basil Blackwell 1972), 50-3; Robert Adams, ‘Where Do Our Ideas Come From — Descartes vs. Locke,’ Innate Ideas, Stephen P. Stich, ed. (Berkeley: University of California Press 1975), 76-8; Desmond Clarke, Descartes’ Philosophy of Science (Manchester: Manchester University Press 1982), 50-2; John Cottingham, Descartes (Oxford: Blackwell 1986), 147-8; Bernard Williams, Descartes: The Project of Pure Inquiry (London: Penguin 1987), 133-5; Nicholas Jolley, The Light of the Soul: Theories of Ideas in Leibniz, Malebranche, and Descartes (Oxford: Oxford University Press 1990), 36-42; Margaret Wilson, ‘Descartes on the Origin of Sensation,’ Philosophical Topics 19 (1991) 293-323, at 302-4; Daniel Garber, ‘Descartes and Occasionalism,’ in Causation in Early Modern Philosophy, Steven Nadler, ed. (University Park: Penn State University Press 1993), 22-3 and Descartes’ Metaphysical Physics (Chicago: University of Chicago Press 1992), 365-6, n. 57; Steven Nadler, ‘Descartes and Occasional Causation,’ British Journal for the History of Philosophy 2 (1994) 35-54, at 47-51 and ‘The Occasionalism of Louis de la Forge,’ in Causation in Early Modern Philosophy, Steven Nadler, ed. (University Park: Penn State University Press 1993), 66-8; Stephen Gaukroger, Descartes: An Intellectual Biography (Oxford: Oxford University Press 1995), 408-10; Tad Schmaltz, ‘Descartes on Innate Ideas, Sensation and Scholasticism: The Response to Regius,’ in Oxford Studies in the History of Philosophy, Vol. II: Studies in Seventeenth-Century European Philosophy, M.A. Stewart, ed. (Oxford: Oxford University Press 1997), 34-44 and ‘Sensation, Occasionalism and Descartes’ Causal Principles,’ in Minds, Ideas and Objects: Essays on the Theory of Representation in Modern Philosophy, Philip Cummins and G. Zoeller, eds. (Atascadero, CA: Ridgeview 1991), 42-9; Kenneth Clatterbaugh, The Causation Debate in Early Modern Philosophy: 1637-1739 (New York: Routledge 1999), 38; Marleen Rozemond, ‘Descartes on Mind-Body Interaction: What's the Problem?’ Journal of the History of Philosophy 37 (1999) 435-67, at 449-62; David Scott, ‘Occasionalism and Occasional Causation in Descartes's Philosophy,’ Journal of the History of Philosophy 38 (2000) 503-28, at 516-20; Nancy Kendrick, ‘Why Cartesian Ideas of Sense are Innate,’ The Southern Journal of Philosophy 37 (2000) 413-28, at 414.

3 See Janet Broughton, ‘Adequate Causes and Natural Change in Descartes’ Philosophy,’ in Human Nature and Natural Knowledge, Alan Donagan, Anthony Petrovich, and Michael Wedin, eds. (Dordrecht: D. Reidel 1986), 116-19. Kenneth Winkler is another author who seems inclined to the non-causal interpretation, though his position is more guarded than Broughton. Winkler writes that Descartes ‘is uneasy about the further claim that they [bodies] play a causal role, not because he wants to deny it, but because he is not prepared to specify the sense in which it is true’ (‘Grades of Cartesian Innateness,’ British Journal for the History of Philosophy 1 [1993] 23-44, at 32). John Yolton, in Perceptual Acquaintance from Descartes to Reid (Minneapolis: University of Minnesota Press 1984), 30 and Perception and Reality: A History from Descartes to Kant (Ithaca: Cornell University Press 1996), 193-209; and Daise Radner, ‘Is There a Problem of Cartesian Interaction?’ Journal of the History of Philosophy 23 (1985) 227-31 also argue, though on the basis of very different considerations from one another, that Descartes cannot regard bodies as simply the efficient causes of ideas. But neither of them addresses in any detail the question of universal innateness. See also S.V. Keeling, Descartes, 2nd ed. (London: Oxford University Press 1968), ch. 6.

4 Hobbes: AT 7 186-8; CSM 2 130-32; Gassendi: AT 7 279-88; CSM 2 195-201

5 See the arguments for the innateness of the idea of God in the Mediations (AT 7 51; CSM 2 35), and in the replies to Hobbes (AT 7 188-9; CSM 2 132) and Gassendi (AT 7 364-5; CSM 251-2), as well as the argument for the innateness of the idea of triangularity in the replies to Gassendi (AT 7 381-2; CSM 2 262).

6 Notice that in both of these arguments Descartes exempts certain kinds of judgments (affirmations and negations about the referents of our ideas) from the scope of the universal innateness thesis. In the Comments he distinguishes between ‘innate,’ ‘adventitious,’ and ‘made-up’ ideas as follows: ‘I did however observe that there were certain thoughts which neither came to me from external things nor were determined by my will, but came solely from the power of thinking within me; so I applied the term ‘‘innate’’ to the ideas or notions which are the forms of these thoughts in order to distinguish them from others, which I called ‘‘adventitious’’ or ‘‘made-up’’’ (AT 8B 358; CSM 1 303). Judgments are not adventitious since they do not come from external things, and they are not innate since they depend on the will. Hence in so far as an idea involves a judgment, such as my belief that my sensory ideas refer to external things, it may be considered made-up. For criticism of this sort of account, see Scott (524).

7 This interpretation is also favored by Gaukroger (408-10). Similarly, McRae proposes that in the Comments our sensory ideas are innate only because they depend on an innate faculty of judgment: ‘It is the faculty of judging which is innate to the mind, and if we speak of the conceptions which we form through judgment as innate, it is only because they are formed through that innate faculty, not produced by what comes through the senses’ (53). As Murray Miles, Insight and Inference: Descartes's Founding Principle and Modern Philosophy (Toronto: University of Toronto Press 1999), 503, n. 4, notes, it is hard to reconcile McRae's conception of innateness with Descartes's suggestion in both the Comments and the letter to Mersenne that ideas based on judgments are not innate: ‘there is nothing in our ideas which is not innate to the mind or the faculty of thinking, with the sole exception of those circumstances which relate to experience such as the fact that we judge that this or that idea which we have immediately before our mind refers to a certain thing situated outside us’ (AT 8B 358-9; CSM 1 304); ‘Altogether, I think that all those [ideas] which involve no affirmation or negation are innate in us’ (AT 3 418; CSMK 187). Furthermore, although McRae argues that for Descartes our ideas of the position, distance, size, and shape of bodies involve judgment, Descartes says in the Comments that even our ideas of pains, colors, and sounds are innate.

8 John Locke, Essay Concerning Human Understanding, Peter H. Nidditch, ed. (Oxford: Oxford University Press 1975), 50

9 See also AT 8B 366, AT 3 430, AT 8B 166, AT 4 187-188; CSM 1 309, CSMK 194, CSMK 222, CSMK 248.

12 Thus he explains to Voetius: ‘But notice that all those things whose knowledge is said to be naturally implanted in us are not for that reason expressly known by us; they are merely such that we come to know them by the power of our own intelligence, without any sensory experience’ (AT 8B 166; CSMK 222). Clarke distinguishes between two senses of innateness in Descartes (50-2). Innate1 ideas are ‘irreducible to the type of reality which triggers them in the mind’; Innate2 ideas are ‘independent of experience in so far as the mind can come to have these ideas by reflection on its own intellectual capacities.’ Clarke holds that when Descartes says in the Comments t h a t a l l s e n s o r y i d e a s a r e i n n a t e h e m e a n s innate1: t h e y are triggered by corporeal motions though their contents are not reducible to such motions. I criticize this ‘triggering model’ of the universal innateness thesis at length below. My view is that Descartes's position in the Comments is that all ideas are innate in a sense closer to innate2 since they all ‘come solely from the power of thinking within me.’ But, unlike Clarke, Descartes does not seem to require (at least not in the Comments) that in order for an idea to be innate in this sense it must be the result of reflection. Rather, Descartes says simply that the mind ‘forms’ its ideas on the occasion of sensory stimulation. Miles (ch. 18) develops a detailed ‘reflexive-dispositional’ theory of innateness, which attempts to preserve the best features of McRae's and Clarke's interpretations, and also illuminate Descartes's theory of the mind as a whole. But Miles does not attempt to apply this theory to the Comments thesis of universal innateness. Rather, he thinks that the universal innateness thesis expresses a ‘rather misleading’ conception of innateness (Miles, 293). For another useful discussion of the various ‘grades’ of innateness in Descartes's writings, see Winkler. Winkler ties Descartes's conception of innateness closely to his conception of a person's ‘nature,’ but admits that this conception does not seem to be at work in the arguments for universal innateness (31). Finally, Deborah Boyle, ‘Descartes on Innate Ideas,’ The Modern Schoolman 78 (2000) 35-51, attempts to account for Descartes's disparate uses of the term ‘innate idea’ in terms of his different senses of ‘idea,’ though again she does not discuss the universal innateness thesis in any detail.

11 See AT 6 85, 112-14, and 130-1; CSM 1 153-4 and 165-7.

12 See also Adams, 76-8; Clarke (50), Cottingham (147-8), Schmaltz (‘Descartes on Innate Ideas,’ 40), and Scott (523-4)

13 Descartes's account of the regular or nomological association between brain states and mental states is discussed in greater detail below.

14 In support of the triggering approach, Rozemond points out that there are other works in which Descartes says explicitly that brain states cause ideas (459). But the fact that Descartes elsewhere speaks of the mind as the cause of ideas might be just as well be taken as evidence that he would surely have said the same thing in the Comments if that was what he intended. In any case, these other passages in themselves count as much (or as little) against the view that the contents of ideas are not caused by bodies, which is Rozemond's interpretation of the Comments, as that their occurrence is not triggered by bodies.

15 This point is also stressed by Broughton (118). David Scott has recently insisted that on at least four occasions in the Comments passage ‘Descartes either explicitly or implicitly affirms that motion is transmitted to our mind.’ The four supposed affirmations of transmission into the mind identified by Scott are (a) ‘if we bear well in mind the scope of our senses and what exactly it is that reaches our faculty of thinking by way of them, we must admit’; (b) ‘there is nothing in our ideas which is not innate to the mind, with the sole exception of those circumstances which relate to experience’; (c) ‘We make such a judgment not because these things transmit the ideas to our minds through the sense organs, but because they transmit something’; (d) ‘nothing reaches our mind from external objects through the senses except certain corporeal motions’ (Scott, 518-19). But the most that can be inferred from (a), (c), and (d) is that corporeal motions are transmitted to, or reach, our minds through the sense organs. Yet bodies can certainly transmit motions through the sense organs to the pineal gland without thereby transmitting those motion into the mind, just as a taxi could bring me to the door of a hotel without transporting me inside. To insist that transmission through the sense organs implies transmission into the mind itself is simply to beg the question against non-causal readings of the Comments. Furthermore, the reason Descartes concludes that our sensory ideas are innate is because ‘neither the motions themselves nor the figures arising from them are conceived by us exactly as they occur to us in the sense organs’ (AT 8B 359; CSM 1 3 0 4 ) . T h u s t h e i s s u e o f t h e Comments i s t h e dissimilarity b e t w e e n t h e i d in the mind and the corporeal motions in the sense organs. But if Scott were right that motions are successfully transmitted into the mind, then the issue would be the dissimilarity between these ‘mental motions’ and the ideas that arise. As for passage (b), this is left incomplete in Scott's quotation. What Descartes says is: ‘So much so that there is nothing in our ideas which is not innate to the mind, with the sole exception of those circumstances which relate to experience, such as the fact that we judge that this or that idea which we now have immediately before our mind refers to a thing located outside us’ (AT 8B 358; CSM 1 304; I have italicized the part of the sentence omitted by Scott). Once completed, the relevance of this passage to Scott's case is unclear, since it does not concern the transmission of corporeal motions at all. Rather it says that certain circumstances relating to experience, namely our judgments about the referents of our ideas, are not innate. Perhaps Scott's assumption is that if such judgments are not innate, they must be adventitious and involve a transmission of corporeal motions into the mind. But this certainly does not follow. Judgments are not adventitious since they do not come from without. Rather they depend on the will and fall into the category of ideas that are ‘made-up.’ (See note 6 above.) Furthermore, the relevant judgments concern ideas that are assumed to be already ‘immediately before the mind,’ and therefore do not presuppose any kind of transmission.

16 These two claims are endorsed by Scott (515 and 522), although Scott does not accept Nadler's account of the precise role of bodies as occasional causes. See also McRae (51).

17 Thus, immediately after introducing the distinction by way of the boss and workers example, he applies it to the question of the origin of the idea of God and says that although verbal instruction is sometimes a remote cause of this idea no one can say it is the ‘proximate and efficient’ cause (AT 8B 360; CSM 1 305).

18 Schmaltz (‘Sensation,’ 43-4) argues that Descartes regards bodies as remote, but not accidental, causes of sensory ideas since they, unlike remote and accidental causes, are necessary for the activity of the mind in producing sensory ideas. On this view, bodies are the efficient causes of the activity of the mind which in turn is the efficient cause of the sensory ideas. The problem with this approach is that it fails to explain the manifest difference in the way Descartes describes the role of bodies and the mind in the process leading to sensory ideas: whereas the mind ‘forms’ ideas, bodies merely serve as occasions for something else to act.

19 See also Nadler in ‘The Occasionalism of Louis de la Forge,’ where he says that occasional causes play a ‘necessary role’ in the production of ideas (67).

20 It is a significant weakness of Broughton's defense of the non-causal interpretation of the Comments that she dismisses Descartes as simply being ‘sloppy’ in these later passages (118). This allows her to avoid the whole issue of remote and accidental causation. But the apparent similarity between remote and accidental causation and the model of sensation given in the preceding section of the Comments is too strong to be dismissed so lightly, and has naturally been emphasized by most recent proponents of the causal interpretation.

21 On this view, the remote and accidental causes of ideas are ideas which induce the mind to produce innate ideas at one time rather than another. This account would explain, to some degree, their accidental quality. For it seems reasonable that whether or not a certain mental image or sound will induce a given person to think of God on a given occasion will depend on the previous experiences of that person, on their current mood, on what other ideas arise, and so on

22 As far as I know, the notion of remote and accidental causation is invoked only once in all of Descartes's writings.

23 See also AT 6 114, 130, AT 11 346, AT 7 81, AT 4, 603-4; CSM 1 166, 167, 337, CSM 2 56, CSMK 307. So the corporeal motions in the brain are not ‘accidental’ to the subsequent sensory ideas in the way that the offer of pay is accidental to the work being done. For whereas the work might or might not get done without the orders, there is a constant conjunction bet ween the ideas and the brain motions. I thank Tad Schmaltz for discussion of this point. Cf. Schmaltz, ‘Sensation,’ 43-4.

24 See also AT 7 81, 143-4; CSM 2 56, 102-3.

25 This ‘generic’ sense of cause, as Clatterbaugh (The Causation Debate, 18) calls it, is broader than efficient causation since it does not presuppose a kind of influx or transference, which Nadler regards as part of the standard model of efficient causation. Thus Descartes says that God causes himself not by any ‘positive influence,’ but only by the immensity of his own power (AT 7 109 and AT 7 236-237; CSM 2 79 and CSM 2 165). For a recent analysis of the sense in which God could be his own cause, according to Descartes, see Daniel Flage and Clarence Bonnen, ‘Descartes on Causation,’ The Review of Metaphysics 50 (1997) 841-72. For an instructive overview of the ‘metaphysics of causation’ in Descartes see Clatterbaugh (The Causation Debate, ch. 2).

26 Scott proposes that the link between the occasioning cause and the activity of the mind in sensation is ‘a kind of temporal cue or trigger for the mind's active thinking of ideas latent within itself’ (523). Unfortunately, this notion of a ‘temporal cue or trigger’ is much more obscure than the relation it purports to explain. Scott says that although occasional causes in this sense are not efficient causes — he calls them ‘inefficacious actions’ (520) — they nevertheless depend on a transmission of motion into the mind. But it is hard to see how a body could transmit motion into the mind without being the efficient cause of that motion. Furthermore, how exactly does the motion transmitted into the mind trigger the mind's activity if not by efficient causation? Are we to assume, as the term ‘cue’ suggests, that the mind ‘notices’ that certain motions have been transmitted within? But why do the motions have to be transmitted into the mind in order to be noticed by the mind? More seriously, the model of temporal cueing fails as an account of sensation since it presupposes a kind of inner perception of the cues by the mind. As Descartes himself says in the Optics, we cannot hope to explain sensation by internalizing it — ‘as if there were yet other eyes within our brain with which we could perceive’ (AT 6 130; CSM 1 167).

27 For a discussion of Descartes's views on causation which is much more skeptical about their significance, and even their coherence, see Jonathan Bennett, Learning from Six Philosophers: Descartes, Spinoza, Leibniz, Locke, Berkeley, Hume, 2 Volumes (Oxford: Oxford University Press 2001), ch. 5).

28 Other closely related causal assumptions which he accepts are that nothing can come from nothing (AT 7 135; CSM 2 97); that whatever can do the greater can also do the lesser (AT 4 111-2; CSMK 231); and that the more perfect cannot come from the less perfect (AT 7 41; CSM 2 28). He also holds that everything has a cause (AT 7 112; CSM 2 80) and that causes are not temporally prior to their effects (AT 7 108; CSM 2 78).

29 Mechanistic explanations of natural phenomena are to be preferred over traditional scholastic models for precisely this reason. For although we understand quite well how the size, shape, and motion of one body can produce properties of the same type in another body, yet ‘there is no way of understanding how these same attributes (size, shape and motion) can produce something else whose nature is quite different from their own — like the substantial forms and real qualities which many philosophers suppose to inhere in things’ (AT 8A 322; CSM 1 285).

30 See also AT 7 40-2; CSM 2 28-9.

31 This causal condition is found, in various forms, throughout Descartes's writings: AT 6 34, AT 8 11, AT 7 40-2, 104-5, 135, 161, 165, 168, 366, AT 3 274, 428, 545; CSM 1 128, 198-99; CSM 2 28-9, 76, 97, 114, 116, 118, 252; CSMK 166, 192, 211. For more detailed discussion of the principle, see Kenneth Clatterbaugh, ‘Descartes's Causal Likeness Principle,’ Philosophical Review 89 (1980) 379-402; Eileen O’Neill, ‘Mind- Body Interaction and Metaphysical Consistency: A Defense of Descartes,’ Journal of the History of Philosophy 25 (1987) 227-45; Geoffrey Gorham, ‘Causation and Similarity in Descartes,’ in New Essays on the Rationalists, R. Gennaro and C. Huenemann, eds. (Oxford: Oxford University Press 1999) and ‘Descartes's Dilemma of Eminent Containment,’ Dialogue: Canadian Philosophical Review (forthcoming); and Bennett (ch. 5).

32 The intricate machine example is also presented at AT 7 14, AT 7 105 and AT 7 134-5; CSM 2 10-11, CSM 2 76 and CSM 2 97.

33 Descartes uses the terms ‘reality’ and ‘perfection’ interchangeably to mean specific property. So he also seems to equate ‘degree of reality’ and ‘degree of perfection.’ Thus he says: ‘what is more perfect — that is, contains in itself more reality — cannot arise from what is less perfect’ (AT 7 41; CSM 2 28).

34 So I may eminently contain extension, shape, and motion since these mere modes of bodies while I am a thinking substance (AT 7 41; CSM 2 31). Likewise, if God is the cause of my ideas, he will contain eminently all that belongs to my ideas, since he is an infinite substance while my ideas are merely modes of thought (AT 7 79; CSM 2 55).

35 AT 6 112-4, 130-1; CSM I 165-6, 167. See also his discussion of the difference between our ideas of figures and the actual figures in our environment in the Fifth Set of Replies (AT 7 381-2; CSM 2 262).

36 In none of the passages discussing eminent containment does Descartes consider the possibility of eminent containment by material things. For example, after concluding in the Sixth Meditation that his ideas of bodies must be produced by a substance other than himself, he continues: This substance is either a body, that is a corporeal nature, in which case it will contain formally ‘‘and in fact’’ everything which is to found objectively ‘‘or representatively’’ in the ideas; or else it is God or some creature more noble than a body, in which case it will contain eminently whatever is to be found in the ideas. (AT 7 79; CSM 2 55. See also AT 7 42, 45, 135-136; CSM 2 29, 31, 97) He ignores here the possibility that our ideas of bodies are caused by finite material things which contain the objective reality of those ideas eminently, but not formally. In fact, the passage strongly suggests that bodies could only cause the perfections of my ideas by formally containing them (‘in which case…’). The notion of eminent containment is a notoriously difficult one. For discussion, see O’Neill; Thomas Vinci, Cartesian Truth (Oxford: Oxford University Press 1998), 70-6; David Hausman and Alan Hausman, Descartes's Legacy: Minds and Meaning in Early Modern Philosophy (Toronto: University of Toronto Press 1997), 42-7; and Gorham (‘Descartes's Dilemma’).

37 Other authors who see similarity as crucial to the argument for innateness are Jolley (41), Wilson (304), Hausman and Hausman (52), Rozemond (458-61) and Broughton (116-19). But of these, only Broughton holds that the conclusion of the argument leaves no causal role for bodies in sensation.

38 In her version of the non-causal interpretation, Broughton assumes that these arguments are quite different from the argument for the innateness of sensory ideas, but does not explain what exactly she takes the main differences to be (118-19).

39 It is worth mentioning one other problem, to which Broughton devotes a great deal of attention. She finds there to be a ‘puzzling inconsistency’ between the causal principle and Descartes's account of physical collision (120). If in collision one body is the cause of motion in another, then according to the similarity condition the first body is the source of the motion that comes to be in the second. This seems to imply that motion somehow ‘migrates’ or is transmitted from one body to another. The problem is that in a letter to More Descartes explicitly denies that motion ‘transmigrates’ in physical collision (AT 5 405; CSMK 382). Broughton does not think Descartes was able to resolve this apparent inconsistency between his causal principle and his theory of physical collision. But it seems to me that the problem only arises if one assumes that in collisions bodies are the efficient causes of motion. And this assumption is suspect, since there are good reasons to regard Descartes as an occasionalist about purely physical interaction. In the Principles, for example, he says that the ‘reciprocal impulses and transfers of motion’ involved in collisions are a direct effect of God's continuous creation of the universe (8A 66; CSM 1 243). This conception is consistent with the causal principle since one may assume, as Descartes does, that God may contain motion eminently (AT 7 79; CSM 2 55) and thus continually impart it to bodies. On Descartes's (physical) occasionalism, see Garber (‘Descartes and Occasionalism’), Clatterbaugh (The Causation Debate, ch. 2), and Bennett, 99-101.

40 AT 11 144, AT 9 114, AT 8B 358-60, AT 3 418, AT 9B 64; CSM 1 103, CSM 1 166, CSM 1 304-5, CSMK 187

41 See, for example, AT 6 59, AT 5 222, 347; CSM 1 141, CSMK 358, 375. Broughton assumes that ‘mind-body interaction involves only formal [i.e. not eminent] containment’ (111). But this leaves her with no way of explaining why Descartes's causal principle prevents interaction in the case of sensation (body-on-mind causation) but not in the case of voluntary movement (mind-on-body causation). For minds certainly do not formally contain motion.

42 In the Optics he says: ‘there need be no resemblance (resemblance) between the ideas which the soul conceives and the movements which cause these ideas’ (AT 6 131; CSM 1 167). See also AT 6 85; CSM 1 153 and (in le Monde) AT 11 4; CSM 1 81.

43 In the earlier Discourse (1637) Descartes relies in the causal proof for the existence of God not on the similarity condition per se, but rather on the more general principle that ‘the more perfect cannot result from the less perfect’ (AT 6 34; CSM 1 128).

44 In a recent article, Nancy Kendrick argues, primarily on the basis of passages from the Optics and le Monde, that non-resemblance is not the basis for Descartes's commitment to innate sensory ideas. She holds that his fundamental point in these works is not so much that ideas do not resemble the bodies which cause them, but rather that such resemblance would not explain the origin of our ideas even if it did obtain: ‘Descartes dismisses resemblance in the Optics b e c a u s e i t i s n o n - e x p l a n a t o r y , not because it is inexact’ (418). But her argument conflates the question of the causal origin of our ideas with the question of how they represent the world. For, as we have seen, what his complaints about the need for homunculi and ‘eyes within our brain’ are intended to show is that mere resemblance cannot account for how an idea can be a mental representation. As Kendrick observes, this is his main objection to the scholastic theory of sense perception. But that homunculi are presupposed by a resemblance theory of mental representation does not show that similarity is irrelevant to the issue of mind-body interaction. The former is an epistemic or semantic question, while the latter is metaphysical. And surely one might demand on met aphy sical grounds that the cause of an id ea must have something in common with its effect, even granting that this would not by itself explain intentionality

45 A rather similar account is presented by Hausman and Hausman, though their concern is primarily the semantics of sensory ideas rather than their causal origin. Their view is that innate ideas, rather than bodies, ‘provide us with the material for constructing our sense fields’ (51). (It is not clear whether they think that brain states play a causal role of any kind.) They also argue that the semantic content of these innate ideas depends in turn on ‘exemplars’ which are not themselves contained in the mind. One problem with this approach is that Descartes says in the Comments that innate ideas are those ‘which came solely from the power of thinking within me’ (AT 8B 358; CSM 1 303). So the Hausmans are led to dismiss Descartes's suggestion in the Third and Sixth Meditations that he could be the cause of all his ideas of sense since he could eminently contain the contents of those ideas, as ‘disingenuous’ (46). I discuss the Hausmans’ views in more detail elsewhere (‘Descartes's Dilemma’).

46 For a concise recent statement of this objection to the non-causal approach, see Scott (514-15). The importance of this problem was impressed upon me by Phil Cummins in the discussion period following presentation of an earlier version of this paper.

47 See also AT 7 81, 143-144; CSM 2 56, 102-3.

48 ‘It is much better that it [a certain sensation] should mislead on this occasion than that it should always mislead when the body is in good health’ (AT 7 89; CSM 2 61). See also AT 5 163-4; CSMK 346.

49 Hence, there is no need to suppose, as Broughton does, that the mind has an innate ‘meta-faculty’ which ‘regulates the mind's faculty for having sense ideas so that just the right sense ideas appear at just the right times’ (119).

50 ‘I did however observe that there were certain thoughts which neither came to me from external things nor were determined by my will, but came solely from the power of thinking within me; so applied the term ‘‘innate’’ to the ideas or notions which are the forms of these thoughts in order to distinguish them from others, which I called ‘‘adventitious’’ or ‘‘made-up’’’ (AT 8B 358; CSM 1 303).

51 This tension in Descartes's views about the origin of sensory ideas is evident also in the case of dreams, which he regarded as both involuntary and internally caused. Cf. David Fate Norton, ‘Descartes on Unknown Faculties: An Essential Inconsistency,’ Journal of the History of Philosophy 6 (1968) 245-56, and Wilson (305-6). For an interesting recent analysis of the official proof, which gives special attention to the unknown faculty problem, but which disagrees in several respects with the analysis offered here, see Lex Newman, ‘Descartes on Unknown Faculties and Our Knowledge of the External World,’ The Philosophical Review 103 (1994) 489-531.

52 Tad Schmaltz attempts to resolve this apparent inconsistency between the Comments and the Sixth Meditation by suggesting that Descartes's view in the Comments i s t h at ideas arise in the mind from the combined effects of an active faculty in bodies and a passive faculty (of sensation) in the mind. The latter faculty would presumably not be subject to the Sixth Meditation prohibition against the production of ideas by active internal faculties (Schmaltz, ‘Descartes on Innate Ideas,’ 42-3). The problem with this approach is that the Comments seems to attribute an active role to minds, which ‘form’ and ‘represent to itself’ our ideas, and a passive role to bodies, which merely ‘give the mind occasion’ to form its ideas.

53 In the Principles version of the proof Descartes does not mention the unknown faculty hypothesis (AT 8A 40-1; CSM 1 223).

54 See also the account of the condition of divine deception in the Fourth Meditation (AT 7 54, 62; CSM 2 37-8, 43) and Second Set of Replies (AT 7 144; CSM 2 103). In these accounts what would make God a deceiver is not simply that I should have an instinctive belief that is false, but rather that I should have from God a faculty that would lead me to falsehood, however judiciously I applied it.

55 Descartes says that there is a ‘big difference’ between a belief based on a spontaneous impulse and one based on the natural light: the former are fallible (AT 7 38; CSM 1 26-7).

56 See also Principles II, 3 (AT 8A 41-2; CSM 1 224).

57 The dissimilarity between sensory ideas and external things is noted frequently in the Meditations (AT 7 39, 80, 81, 83; CSM 2 27, 55, 56, 57) and elsewhere (AT 11 4, AT 685, AT 6 130, AT 4 603-4; CSM 1 81, 153, 167, CSMK 307). For Descartes, this lack of similarity is not confined to our ideas of the so-called secondary qualities. He points out that distant bodies do not have the same size or shape as in the sensory ideas they cause (AT 7 39, 83; CSM 2 27, 57). See also his discussion of our perception of size and shape in the Optics (AT 6 140; CSM 1 172) and in the Fifth Set of Replies (AT 7 381-2; CSM 2 262).

58 There are other writings, besides the Comments, in which the inference from efficient causation to similarity (or from dissimilarity to lack of efficient causation) is straightforward: ‘Now God is the cause of me, and I am an effect of him, so it follows that I am like him’ (AT 5 156; CSMK 340); ‘Now any elements in our thought which do not resemble external objects manifestly cannot have originated in external objects’ (AT 7 188; CSM 2 132).

59 See his Preface to the French Edition (AT 9B 1; CSM 1 179).

60 David Scott disputes Garber's contention that we have evidence of a shift away from body-on-mind causation in the French translation. He points out that even in the French translation Descartes assumes, just as in the official proof, that ideas seem to come from an external source: ‘we experience within ourselves that everything we sense comes to us from something other than our thought’ (AT 9B 63, Scott's translation). But from this assumption — which of course is a premise of Descartes's argument, not its conclusion — Scott illicitly infers that ‘Descartes is out to prove in the later version: ‘‘everything comes to us from something other than out thought’’’ (Scott, 510). But if that is what Descartes were out to prove, then Scott fails to explain why his explicit conclusion in the proof is much more circumspect and makes no claim at all about the actual causal origin of our ideas. Descartes's conclusion is: ‘Now, since God does not deceive us, because that is contrary to his nature … we must conclude that there is a certain substance, extended in length, breadth, and depth which now exists in the world, with all those properties which we manifestly know belong to it’ (AT 9B 64).

61 See also the Principles version of the proof: ‘we have a clear and distinct perception of some kind of matter, which is extended in length, breadth and depth’ (AT 8A 40: CSM 1 223).

62 Similarly cautious conclusions are drawn elsewhere. Thus, Principles I I , 3 states: ‘They [sensory perceptions] normally tell us of the benefit or harm that external bodies may do … and do not, except occasionally and accidentally, show us what external bodies are like in themselves. If we bear this in mind we will easily lay aside the preconceived opinions acquired from the senses, and in this connection make use of the intellect alone, carefully attending to the ideas imprinted in it by nature’ (AT 8A 41-2; CSM 1 224). See also in the Sixth Meditation where Descartes says we have ‘no convincing argument’ for supposing that there are things in fire which resemble heat or pain; we can only conclude that there is ‘something in the fire, whatever it may eventually turn out to be, which produces in us the feelings of heat or pain’ (AT 7 83; CSM 2 57).

63 A much more radical revision to the official argument, one which actually abandons dependence on God's veracity, has recently been suggested by Hausman and Hausman. They argue that given Descartes's ex nihilo causal principle (what I call his ‘similarity condition’), the only way to account for the semantic content of our judgments about individual things, such as ‘This is a cube,’ is to posit the existence of a world of real physical things. Thus, ‘Descartes's argument for the existence of such a world is at base a semantic one which depends not on the veracity of God, but on the necessity of the ex nihilo principle’ (58). But, as the Hausmans recognize, strict adherence to the causal principle in this context threatens to prove too much, i.e., that all of our sensory ideas are not only meaningful, but true. For if the content of the judgment ‘This is a cube’ is explained by the existence of physical things which contain the same perfections, then it seems to follow that the judgment is true. Hence, the Hausmans infer that ‘In order to account for the possibility of error, Descartes must abandon the principle when it comes to mind-body interaction’ (63). On this reading, the ex nihilo principle is sufficient to establish the bare existence of physical things, but God's veracity is invoked to account for the truthfulness of some, but only some, of our ideas. Thus, the Hausmans reading of the argument has in common with the reading I have proposed that only the existence of bodies, and not the specific contents of our beliefs about those bodies, is validated. The advantage of my reading is that it does not assume that the causal principle is applied selectively or abandoned. For on my reading, the contents of our ideas comes from the mind, while error arises only from our tendency to assume that our sensory ideas are exactly similar to external things.

64 On this model, sensory ideas are like old memories which are stored in the mind until triggered by events with which they may have no apparent relation. In the case of Cartesian sensation, however, we do not need to assume that the stimulation of our senses actually triggers the innate ideas, since God has assured that they will arise at just the right time. Descartes himself drew attention to the analogy between innate ideas and memories. In the Fifth Meditation he observes that when he reflects on the nature of shape, number, motion, and so on, ‘it seems that I am not so much learning something new as remembering what I knew before’ (AT 7 64; CSM 2 44). And in the May, 1643, letter to Voetius he explicitly compares his account of how innate ideas arise in the mind with Plato's doctrine of anamnesis (AT 8B 167; CSMK 222-3). See also the opening of le Monde, where he entertains the possibility that the mind recollects and represents to us the idea of light each time our eye is affected by the corresponding external motions (AT 11 4; CSM 1 81), and the Conversations with Burman, where he is reported as having said that although there is ‘certainly no relationship’ between the four letters ‘K-I-N-G’ and the notion of power, the intellectual memory allows us to ‘recall what the letters stand for’ (AT 5 150; CSMK 336-7).

65 AT 11 4, AT 6 112, AT 4 604; CSM 1 81, 165, CSMK 307. For a detailed defense of the sign account of Cartesian intentionality see Yolton, Perceptual Acquaintance (ch. 1) and Perception and Reality (ch. 8).

66 For generous comments on earlier versions of this paper I am grateful to Deborah Boyle, David Brokken, Tad Schmaltz, Craig Squires, James White, referees for this journal, and audiences in Minnesota, Iowa, and Newfoundland. Correspondence: Geoffrey Gorham, Department of Philosophy, Macalester College, St. Paul, MN, 55105, USA. Email: gorham@macalester.edu