In these notes, I abbreviate as follows: AT: œuvres de Descartes, ed. Charles Adam and Paul Tannery (Paris: Cerf 1897-1910) HR: The Philosophical Works of Descartes, ed. Elizabeth Haldane and G.R.T. Ross (Cambridge: Cambridge University Press 1972) K: Descartes: Philosophical Letters, ed. Anthony Kenny (Oxford: Clarendon Press 1970) DDM: Harry Frankfurt, Demons, Dreamers and Madmen (New York: Bobbs-Merrill 1970) CC: Alan Gewirth, The Cartesian Circle,’ Philosophical Review (1941) 368-95

1 In CC. See also Gewirth's The Cartesian Circle Revisited,’ The Journal of Philosophy (1970) 668-85.

2 In DDM. See also Frankfurt's ‘Descartes’ Validation of Reason,’ American Philosophical Quarterly (1965) 149-56.

3 I think the most detailed and convincing discussion of clarity and distinctness appears in Alan Gewirth's ‘Clearness and Distinctness in Descartes,’ Philosophy (1943) 17-36.

4 Principles, I xlvi; HR I 237; AT VIII 22

5 I think what follows clearly characterizes Gewirth's interpretation, but I ought to say why I think it characterizes Frankfurt's. Frankfurt says that the Third Meditation argument for God's existence ‘is an attempt to show that there are no good reasons for believing that reason is unreliable’ (DDM, 175). He considers the objection that perhaps ‘what we clearly and distinctly perceive is sometimes false even if we can have no reasonable grounds for supposing so’ (DDM, 179). He decides, for various reasons, that Descartes thinks this absolute falsity (and absolute truth) are ‘irrelevant to the purposes of inquiry. Descartes’ account makes it clear that the notion of truth that is relevant is a notion of coherence’ (DDM, 179). I think that what I say in this section applies to Frankfurt (interpreting truth as coherence) as well as it does to Gewirth (interpreting truth as correspondence).

6 See DDM, 176 (middle paragraph) and 177 (first full paragraph), and CC. 393-4.

7 HR I 148; AT VII 21

8 HR I 172; AT VII 53-4

9 In section 5, I claim also that the Frankfurt-Gewirth interpretation distorts other passages in the Meditations, but I am postponing that criticism, because I think the distortion is part of what is wrong with the original interpretation on which the problem of circularity arises.

10 This, I take it, is the problem Frankfurt raises (DDM, 179) which the truth-ascoherence interpretation is supposed to solve. I think that this isn't a legitimate problem (in its present form). and in any case I do not see how truth-ascoherence would solve it. One could just as well say, ‘So what if you have no good reason to doubt that all your distinct perceptions cohere; it may all the same be false that they do (will).'

11 Here the problem for Frankfurt's Descartes is (at minimum) that we distinctly perceive reason to doubt that our distinct perceptions do (will) cohere.

12 K 15; AT I 152

13 HR I 149; AT VII 24

14 HR I 159; AT VII 36

15 HR I 160; AT VII 38. On Gewirth's view, this passage seems ‘contradictory to the doubt which has preceded it’ (CC, 392). Gewirth resolves the contradiction by claiming that Descartes was raising metaphysical doubt in the earlier passage but is here noting that some perceptions resist psychological doubt. We need not go through these interpretative manoeuvres if we realize that Descartes earlier says metaphysical doubt only seems to be all-encompassing.

16 Alan Gewirth has argued forcefully in conversation that the natural light is nothing but the general faculty of distinct perception. He points out first that Descartes usually does does contrast sensation (and the beliefs it compels) with distinct perception in general (and the beliefs it compels). Second, he asks why Descartes did not simply contrast distinct perception in general with the supposedly narrower notion of perception by the natural light. Third, he points out that elsewhere (for example, in Part I of the Principles) Descartes plainly uses ‘natural light’ to mean ‘faculty of distinct perception.’ I have no direct reply to these objections. Looking just at the use of ‘natural light’ in the Third Meditation, I find the evidence in favor of my suggestion to be counterbalanced by these objections. But looking at all the texts and issues about doubt and certainty in the first three Meditations, I find my suggestion plausible.

17 HR I 162, 163, 165, 167, 168, 171; AT VII 40, 42, 44, 47, 49, 52

18 HR I 162 ff.; AT VII 40 ff.

19 There is, of course, much to be said about exactly how these premises are supposed to resist skepticism, but that lies outside the scope of this paper. As should be clear by now, I do not think these premises are merely ‘psychologically’ certain. Indeed, I do not find much use for the distinction between (mere) psychological certainty and metaphysical certainty.

20 HR II 34; AT VII 134

21 HR II 34-5; AT VII 135

22 HR I 147; AT VII 20

23 HR I 147; AT VII 21

24 I would like to find a notion of being demon-proof that covers both the causal premises and the premises introduced in the Second Meditation. Perhaps the following will do: p is demon-proof if (a) I distinctly perceive that p is true and (b) in any serious attempt to doubt p, I must presuppose that p is true. But I am not sure how well this captures the special status Descartes gives the cogito, and of course the notion of presupposition involved needs a great deal of elucidation.

I should mention also that I think the notion of causation is also essential to Descartes’ dream argument and that recognizing this helps to explain why Descartes answers the dream argument as he does in the Sixth Meditation. But elaboration of this point lies beyond the scope of this paper.

25 Again, the notion of presupposition that would be at work in an argument like this is far from dear.

26 HR I 183-5; AT VII 69-71

27 HR II 38-9; AT VII 140-1; HR II 114-15; AT VII 245-6

28 HR II 39; AT VII 141

29 HR I 185; AT VII 69-70

30 Roughly, I would want to argue that Descartes starts out by distinguishing between perceptions like two and two make four (distinct but not known by the light of nature) and the cogito (distinct and known by the light of nature). Then he associates that distinction with a distinction between perceptions which can and cannot enter into a systematic science. Finally, he associates that distinction with the distinction between perceptions which can and cannot be remembered without attending to their proofs. Each of these associations is natural enough, I would argue, but the overall effect is misleading.

31 I am grateful to many people for their helpful criticisms of this paper. I want especially to thank Burton Dreben, Alan Gewirth, Israel Scheffler and Sam Scheffler.