Rocking the foundations of Cartesian knowledge: critical notice of Janet Broughton, Descartes's Method of Doubt.
A careful treatment of the topics of this book has been long overdue. Broughton's ideas are innovative, engaging, and clearly developed at every stage. The wide-ranging issues addressed remind the reader of why Descartes's thought is of continuing philosophical interest. Throughout, her interpretation is sensitive to the exegetical concerns of scholars. This rich book deserves the attention of every serious student of Descartes.
The present paper is a critical study of part 2--the more ambitious part of this highly ambitious book. I begin with an overview of Broughton's account. Sections 2-4 contain my analysis.
1. Broughton's Account
On Broughton's interpretation, Descartes builds up indefeasible knowledge from an epistemically privileged class of truths. These favored truths constitute a subset of clear and distinct truths. (2) Whereas the-certainty of ordinary clear and distinct perception is defeasible by the deceiving God hypothesis, the certainty of the privileged truths is utterly immune to doubt. Which truths belong to the favored set?
Among those favored ideas are "I exist," "I am doubting," "I desire to know more," "I seem to see a light," "I have an idea of God," and "Something cannot come from nothing." But mathematical ideas are not among the favored few, despite their clarity and distinctness; not even the idea that two plus three equals five is indubitable. (181)
Broughton calls the privileged truths "indubitable" (and occasionally "absolute certainties"). This fosters some confusion. For in her technical sense, indubitability is not understood in the usual way: it concerns not "the power of the human mind to enter into a state of doubtfulness about a proposition," but instead the condition whereby "it is impossible both that the proposition be false and that I be doubting whether it is truer (100). To avoid confusion, I refer to as "super-indubitable" these privileged truths that figure importantly in Broughton's interpretation. On her account, the nature of the super-indubitables explains their priority in Descartes's project. Their certainty is not only utterly indefeasible, it arises from engaging the method of doubt.
Among the strengths of Broughton's treatment is her in-depth analysis of super-indubitability. In itself, the idea of targeting some privileged truths is not new--the cogito has long been admired and discussed for its unique certainty. On three counts, however, Broughton's development of the "privileged truths" theme is both new and interesting. First, she purports to explain the super-indubitability of the cogito without presupposing that the mind is certain about its own thinking. Second, she purports to explain why the cogito's super-indubitability extends to other truths. Third, she purports to explain how engaging in methodic doubt is integral to super-indubitability.
One of the startling implications of the account is its radical makeover of the Cartesian mind. On standard accounts, Descartes's epistemological turn--a sea change in the history of philosophy--is marked by an inside-out approach to philosophical inquiry, an approach owed to the priority of thought. The I think defines the starting point of philosophical investigation--its proper Archimedean point--because of the unshakable certainty it affords. Broughton's revisionist account reflects her sense that the priority-of-thought doctrine widely attributed to Descartes--a doctrine that assumes a first-person realm of privileged access and certainty--is philosophically suspect. Since her Descartes does not hold the suspect doctrine, standard accounts err in presupposing that "for Descartes certainty about 'I exist' is derived from certainty about 'I think'" (110). On her account, super-indubitability is derived not from any starting point of privileged certainty, but from its unique relation to methodic doubt:
The antiskeptical force of Descartes's arguments does not arise from a starting point about which he has certainty. The meditator does not need to start out by claiming he is certain he is raising doubts. Rather, Descartes's arguments get their antiskeptical force by showing the meditator that there are some propositions he cannot rationally doubt, because the possibility of doubting them depends upon granting their truth. (193-94)
Her Descartes aims "to establish the absolute certainty [super-indubitability] of some of his beliefs by showing that their truth is a condition of his using the method of doubt" (98).
How, according to Broughton, is Descartes's project supposed to establish this? Her novel answer is bound up with what she calls "dependence arguments"--the proofs of which super-indubitable truths are conclusions. Part of the novelty arises from the anti-foundationist implications: the truth of the super-indubitables--the most certain of truths--is not self-evident, but inferential. Another part of the novelty is in forcing obstinate skeptics into submission: dependence arguments purport to show that accepting the super-indubitable truths is a necessary condition of doubting them.
A claimed virtue of the interpretation is that the dependence arguments underwrite exactly the propositions Descartes needs to demonstrate the divine guarantee of clear and distinct perception. En route to that guarantee, Descartes needs somehow to establish a variety of matters, including: that he exists, the Third Meditation causal principles, various logical principles, and even the principle of sufficient reason. To avoid circularity, the certainty of these matters cannot depend on the eventual divine guarantee they help ground. Other truths, including those of mathematics, are not needed for the divine guarantee. On Broughton's understanding, all this requires a bifurcation of truths according to whether their certainty depends on the guarantee. Prima facie, however, the details of the bifurcation appear arbitrary. As Broughton notes, it's puzzling why Descartes would maintain that the Third Meditation causal principles "cannot be doubted, when he is prepared to doubt whether two plus three equals five" (162). Her interpretation purports to make sense of this. On the super-indubitable side of the bifurcation, the truths are provable by dependence arguments--that is, doubting them leads one to accept them. The truths on the other side of the bifurcation are not provable by dependence argument. Knowledge of these truths depends on the divine guarantee. Broughton explains:
[B]y understanding the method of doubt in the way I have been proposing, we will see that Descartes has a good rationale for treating some but not all of his clear and distinct ideas as absolutely certain in the early stages of his inquiry. On the reading I am proposing, each of the premises for the argument that God exists and is not a deceiver is among the clear and distinct ideas that have a special status: they are all claims that are conditions of using First Meditation doubt. This means that the meditator can be absolutely certain they are true even before he has ruled out the skeptical scenarios. Once the meditator sees that these absolutely indubitable claims together imply that God exists and is not a deceiver, he is in a position to be absolutely certain that God exists and is not a deceiver, for God's existence is a condition of his doubt too. And by achieving absolute certainty, that God exits, the meditator can also achieve absolute certainty that all of his clear and distinct ideas are true. (181)
The account promises big payoffs. Among them is to dissolve the Cartesian Circle--itself raising the most intractable interpretative problems surrounding the Meditations. On the canonical formulation of the circle, Descartes's argument unfolds as a circle defined by two arcs:
Arc 1: The conclusion that a non-deceiving God exists is founded on the veracity of clear and distinct perception. Arc 2: The veracity of clear and distinct perception is founded on the conclusion that a non-deceiving God exists.
Broughton's account breaks the circle at the first arc. For on her account, the conclusion that a non-deceiving God exists is founded not on the veracity of mere clarity and distinctness, but on super-indubitability. Descartes is thus acquitted of the charge of circularity. Importantly, the acquittal does not come at the cost of arbitrariness. Whereas other interpretations appealing to a class of privileged truths have Descartes exempting such truths from doubt in arbitrary ways, Broughton's Descartes exempts them on principled grounds---their super-indubitability.
As I have tried to make clear there is much to admire in this highly innovative interpretation. But is it the correct interpretation? Notwithstanding its many strengths. I think the account is mistaken. On close inspection of the relevant philosophical, doctrinal, and textual considerations, I believe the weaknesses of the account overshadow its strengths. In what follows, I shall argue my case--focusing on what I take to be wrong in the account. Section 2 is devoted to the Cartesian Circle. In section 3, I discuss her account of how dependence arguments underwrite the cogito (and other super-indubitable truths). Section 4 focuses on important texts bearing on her account of the super-indubitables.
2. Resolving versus Relocating the Cartesian Circle
In its canonical form, the Cartesian Circle has Descartes's project impaled on one horn of a skeptical dilemma tracing back to ancient Pyrrhonian skepticism. Accordingly, efforts to found knowledge inevitably succumb to either circularity or infinite regress. (3) Descartes's meditator founds his conclusions about God on the veracity of clear and distinct perception; the veracity of clear and distinct perception is itself founded on ... On what? His inquiry might have traced a regressive path, but, in its canonical rendering, it arcs back completing a circuit (see Arc 2 above). As we've seen, Broughton's account breaks the first arc of the circle, thus acquitting Descartes of the canonical circularity. But how does Broughton's formulation of Descartes's knowledge-founding-path fare with regard to the Pyrrhonian dilemma? Her meditator founds his conclusions about God on the certainty of super-indubitable propositions; the certainty of super-indubitable propositions is itself founded on ... On what? It is with Broughton's answer to this question--her appeal to dependence arguments--that I am at present concerned. My concern will be not with the dependence character of these arguments (this concern arises in section 3), but with their character as arguments. For according to a foundationist tradition dating back at least to Aristotle, the only acceptable resolution of the Pyrrhonian dilemma is to found knowledge on basic truths--that is, truths themselves not founded on further arguments. (4) Such issues come to the fore, because on Broughton's interpretation the super-indubitables are not basic truths. As I'll argue, Broughton's interpretation serves merely to relocate the juncture at which Descartes's project succumbs to the Pyrrhonian dilemma.
On what, then, is the certainty of super-indubitable propositions founded? Broughton's answer: On the steps of dependence arguments. The Pyrrhonian skeptic now asks, On what are the steps of the dependence arguments founded? Since the steps of the dependence arguments must themselves be perceived with certainty, it would appear that only two kinds of answers are open to Broughton: these argumentative steps are themselves either (i) super-indubitable, or (ii) merely clear and distinct. At this juncture, Broughton's Descartes confronts the horns of dilemma. Option (i) initiates an infinite regress--for, on this option, the steps of each successive dependence argument will be founded on yet further dependence arguments, and so on, ad infinitum. Option (ii) results in circularity--a circle defined by the following four arcs:
Arc 1: The steps of dependence arguments are founded on the veracity of clear and distinct perception. Arc 2: The veracity of clear and distinct perception is founded on the conclusion that a non-deceiving God exists. Arc 3: The conclusion that a non-deceiving God exists is founded on the certainty of super-indubitable propositions. Arc 4: The certainty of super-indubitable propositions is founded on the steps of dependence arguments.
Arc 1 is the very option, (ii), under consideration. Arcs 2-4 are theses entailed by Broughton's interpretation--theses serving to trace a circuit. On this rendering, the circularity is more subtle than with the canonical formulation. It is circular, nonetheless. Note too, given option (ii), that the steps of dependence arguments are vulnerable to the deceiving God hypothesis--only the super-indubitables are immune, on Broughton's account.
It thus appears that Broughton's interpretation manages not so much to resolve the underlying Pyrrhonian dilemma as to relocate it. For on her account, it remains the case that Descartes's project succumbs to either circularity or infinite regress. One might try to avoid the horns of the dilemma, rejecting both (i) and (ii) in favor of some third option. But this resort serves merely to redirect the dilemmic questioning toward the ground of this third option. (5)
The difficulties surrounding the Pyrrhonian dilemma help underscore why foundationist interpretations have flourished--why, on standard interpretations, Descartes aims at founding knowledge on non-inferential ground. Even so, the standard accounts must contend with the canonical formulation of the Cartesian Circle. To reiterate, the apparent circularity arises in that the groundbreaking truth about God appears itself to be founded on premises underwritten by the divine guarantee. Though this is not the place to detail my own account, I contend that, in part, Descartes thinks he avoids both horns of the dilemma, because, in the final analysis, the truth about God is not founded on any premises--at all. Instead, this divine truth emerges as self-evident, though it is discovered via defeasible, pre-theistic premises--a contention finding strong textual support (so I argue elsewhere). (6)
Granting my objection that Broughton's account merely relocates the Cartesian Circle, one might plausibly hold that her account nonetheless is the correct interpretation. Accordingly, the locus at which Descartes succumbs to the Pyrrhonian dilemma is correctly represented not by the canonical formulation, but the four-arc circle above--thus attributing to him a more subtle error. Section 3, and especially section 4, address more closely the doctrinal and textual evidence bearing on Broughton's interpretation.
3. How Dependence Arguments Underwrite Super-Indubitability
My present aim is to examine the dependence argument schema on which super-indubitable truths are founded. The examination will serve also to develop more fully Broughton's highly original views about the cogito.
Broughton maintains that the usual 'cogito'-talk in the secondary literature (and which I have been using) is misleading. It emphasizes unduly the epistemic importance of thinking in such remarks as, "if I convinced myself of something then I certainly existed" (2:16-17, AT 7:25). (This emphasis explains why 'cogito' is the de facto shorthand for 'cogito ergo sum'; following Broughton, I hereafter adopt 'I exist' as the shorthand.) According to Broughton, Descartes "endorses a claim that licenses an inference from 'I convince myself' to 'I exist"; he denies that the convincing (the thinking) is epistemically prior--the meditator's "certainty that he exists is not licensed by prior certainty that he is convincing himself of something" (116). As already noted, her Descartes does not presuppose the philosophically suspect, priority-of-thought doctrine. The dependence arguments provide "a different way of understanding how Descartes arrives at certainty about 'I exist'" (110).
How are dependence arguments supposed to underwrite the super-indubitable certainty of the I exist? Let's begin with the standard kind of argument that, as Broughton maintains, has no force against a skeptic who doesn't accept the priority-of-thought-doctrine:
(A) I am doubting something.
(B) I exist.
Broughton calls this sort of argument a "transcendental argument":
A transcendental argument has as its starting point a claim that concerns people's experiences or their cognitive states or capacities. The argument then aims to show that the truth of some other statement is a necessary condition of the starting point. (187)
On her account, Descartes accepts the validity of the transcendental argument for I exist (cf. 110). He thinks it lacks anti-skeptical force, however, because of uncertainty about its premise: even relative to skeptics for whom the argument is sound, if the skeptic is uncertain about (A) he will be able to sustain his doubts about (B). What is needed is, somehow, to derive the (B)-claim without presupposing certainty about the (A)-claim. The dependence argument strategy, is tailor-made for such a scenario. Broughton explains:
One way to tighten the grip of the argument on the skeptic, then, would be to show that success in doubting the B-claims would depend in some way upon granting the A-claims. So one good antiskeptical strategy for a transcendental arguer to pursue is to show that the skeptic could doubt the B-claim only if he granted the truth of the A-claim. Then the transcendental argument from (A) to (B) would show that once the skeptic has granted the A-claim, he must concede the truth of the B-claim. (189)
Broughton takes this strategy to be realized in the following form of argumentation--the dependence schema (189):
1. If I raise a doubt whether (B), I must grant that (A) is true.
2. But if (A), then (B).
3. So if I raise a doubt whether (B), I must grant that (B) is true.
Line 2 is the conditionalized form of the transcendental arguments on which dependence arguments build. Though such transcendental arguments lack anti-skeptical force, the broader dependence arguments are supposed to mark an improvement in this regard. Broughton contends that the structure of dependence arguments brings the skeptic to "see that be cannot rationally doubt the B-claim" (189). Applying the schema to the I exist yields the following dependence argument (117):
1. If I use a consideration to doubt whether I exist, I must grant that I am doubting something.
2. If I am doubting something, then I exist.
3. If I use a consideration to doubt whether I exist, then I must grant that "I exist" is true.
Reflection on this argument is supposed to reveal the futility in doubting one's own existence. Importantly, it is supposed to reveal this without presupposing the suspect priority-of-thought doctrine. (7)
The ingenuity of the dependence argument strategy is one of the many indications of the subtlety of Broughton's interpretation. The strategy of forcing the obstinate skeptic into submission is Cartesian at its core. And as usual, Broughton's discussion is sensitive to a host of texts. Even so, I want to argue that the strategy does not after all succeed, nor is it in Descartes's employ.
My central objection is that dependence arguments do not provide the advertised improvement over the transcendental arguments on which they build. Consider again the above dependence argument for I exist. How exactly is this an improvement over the original transcendental argument (or line 2), in relation to a skeptic who's uncertain about his doubting? Clearly, the claimed improvement does not come in the form of an added premise the antecedent of which is certain--any skeptic uncertain of the antecedent of line 2 will be uncertain of the antecedent of line 1. In what form does the improvement come? The claimed problem with the original transcendental argument is that it establishes no more than conditional existence, while leaving the skeptical meditator uncertain of the condition. In this regard, the dependence argument offers no improvement. The dependence argument establishes that if I am doubting my existence, then I exist. But essentially the same result is already explicit in line 2. It would seem that the dependence argument serves only to add a layer of conditionality. Granted, there are minor differences between the two conditionals--lines 2 and 3. These differences, however, favor the unadorned version in line 2, not the embellished version in line 3. For any skeptic who accepts the condition of line 3 (that I am using a consideration to doubt whether I exist) will therefore accept the condition of line 2 (that I am doubting something); but not vice versa. How, then, does the conclusion of the dependence argument improve on line 2?
Broughton notes that "even a completely sound and valid" transcendental argument won't convince the skeptic who's uncertain of his thinking. But here again, the corresponding dependence argument offers no improvement. This is well illustrated by highlighting what Broughton works hard to achieve--the dispensability of the first-person perspective. This dispensability may be clarified by substituting the variable, x, for occurrences of 'I':
1'. If x uses a consideration to doubt whether x exists, x must grant that x is doubting something.
2'. If x is doubting something, then x exists.
3'. If x uses a consideration to doubt whether x exists, then x must grant that "x exists" is true.
This argument is sound just in case the original is. That its soundness is indeed preserved where x stands for a non-existent unicorn helps clarify that the argument does not establish that "x exists" is true. What, then, could reflection on the dependence argument reveal to a skeptical meditator who's uncertain about his own thinking? For all this argument shows, the meditator might himself be non-existent. Of course, there is the following distinction between the skeptical meditator and a non-existent unicorn: unlike the meditator, a fictional being is not actually doubting or thinking--it cannot see that its existence is presupposed by its own doubts. On Broughton's interpretation, however, this is a distinction without any epistemic difference: for her skeptical meditator is quite uncertain of his own doubting and thinking.
It would therefore seem that, for all its ingenuity, the dependence argument strategy fails to deliver the advertised improvement. Nor does it help that, in agreeing to engage the method of doubt, the skeptic does thereby assume what is uncertain--namely, that he is doubting something. For this assumption is of equal value relative to the original transcendental argument. And in a quest for indefeasible knowledge, uncertain assumptions have no value.
Aside from whether dependence arguments do the claimed philosophical work, we want to know whether Descartes employs them. The relevant texts are numerous, and Broughton's treatment of them is extensive and impressive. Space does not permit such a careful treatment here. I will, however, call some textual attention to three regards in which her account is at odds with the more standard accounts. First, standard accounts typically take the I exist to be exemplary not of super-indubitability, but of clarity and distinctness. Second, they typically maintain the priority-of-thought doctrine, whereby the I exist is derived from epistemically prior certainty about I think. Third, they typically maintain that the I exist is apprehended self-evidently. Noteworthy texts speak to each issue.
A Third Meditation passage, purports to characterize the ground of the I exist. In context, the meditator is attempting to extend its certainty to a general rule--a criterion for "being certain about anything" (2:24, AT 7:35). Reflecting on its ground, the meditator says: "In this first item of knowledge there is simply [nihil aliud] a clear and distinct perception of what I am asserting" (ibid., my italics). On this basis, the meditator tentatively concludes:
So I now seem to be able to lay it down as a general rule that whatever I perceive very clearly and distinctly is true. (2:24, AT 7:35)
This is hardly the general rule one would expect given Broughton's interpretation. For, on her account, what is impressive about the I exist is not simply its clarity and distinctness, but its super-indubitability--itself underwritten by a dependence argument.
The exchange between Descartes and Gassendi produces an exceptionally clear statement of the priority-of-thought doctrine that Broughton opposes. Gassendi writes:
[Y]ou conclude that this proposition, I am, I exist, is true whenever it is put forward by you or conceived in your mind. But I do not see that you needed all this apparatus, when on other grounds you were certain, and it was true, that you existed. You could have made the same inference from any one of your other actions, since it is known by the natural light that whatever acts exists. (2:180, AT 7:258-59)
Gassendi's remarks are noteworthy because, like Broughton, he plays down the first-person perspective. Descartes replies:
When you say that I "could have made the same inference from any one of my other actions" you are far from the truth, since I am not wholly certain of any of my actions, with the sole exception of thought. (2:244, AT 7:352)
Broughton acknowledges this passage, admitting that "Descartes himself says that certainty about 'I exist' derives from certainty about 'I think'" (111). She contends that the passage is misleading, that in his frustration in dealing with Gassendi, Descartes is careless--he "conflates the dependence argument for the indubitability of 'I exist' with a quite different argument" (142-43). More generally, Broughton does "not think we can look outside the Second Meditation for much confirmation" of her reading (141). She thus focuses her primary efforts on trying to show that it is her account, not that which prioritizes the I think, that provides "the best interpretation of the famous passage in the Second Meditation" (112).
Broughton's focused treatment of the relevant Second Meditation passage is carefully argued and thorough (cf. her chapter 7). In my judgment, she shows that that passage can be read along the lines she suggests. As she shows, there is inferential structure in the reflection leading up to the declaration of I exist. I would argue, however, that it does not follow from the fact that the I exist is first discovered with the help of inferences that the inferences endure as its epistemic ground. A young girl might discover that 2 + 3 = 5 via inferential reflection, though, in time, the ground of her certainty may emerge in the form of serf-evidence. (8) Descartes says, of truths that people discover via argument, that they may "eventually be just as self-evident to them as the fact that the number two is even or that three is odd" (2:115, AT 7:164-65). My own sense is that a great deal of the "argumentation" of the Meditations is intended to help the meditator to arrive at a self-evident grasp of various truths--notably, the truths concerning God. (9) In response to the second objectors, whom Descartes takes to misconstrue the I exist as resting on inference, Descartes writes:
When someone says "I am thinking, therefore I am, or I exist," he does not deduce existence from thought by means of a syllogism, but recognizes it as something self-evident by a simple intuition of the mind. (2:100, AT 7:140)
4. Super-Indubitability and the Texts
My primary aim in this final section is a clarification of important texts bearing on Broughton's account of the super-indubitables. I'll focus on three interpretative theses--each entailed by her account of super-indubitability:
A. The most certain of truths (the super-indubitables) are utterly immune to hyperbolic doubt.
B. Knowledge (that is, Descartes's brand of fully-indefeasible knowledge) of the most certain of truths does not depend on knowledge of a non-deceiving God.
C. Propositions of math are not included among the most certain of truths.
I shall argue that there is a compelling textual case against all three theses. Beat in mind that some version of A and B (versions stripped of any allusion to super-indubitability) is widely presupposed in the secondary literature. Consequently, my textual argument against these theses is not limited to Broughton's interpretation.
A preliminary remark is in order. Given the magnitude and complexity of the texts bearing on A, B, and C, it would be all too easy to find problem texts for her interpretation. I say too easy, because every interpretation must confront such texts. Citing a few cases hardly makes for a telling refutation, given that the interpreter can cite supporting texts of her own. (Part of what is impressive in Broughton's account is the vast number of problem texts for which she offers a plausible reading.) A telling refutation would require an overwhelming body of textual evidence.
I shall build my case on what I take to be an overwhelming body of textual evidence--overwhelming in both quantitative and qualitative terms. These texts--I'll refer to them as "master texts"--are not only numerous and mutually reinforcing; their clear, detailed exposition arises in pivotal contexts. They provide Descartes's most detailed exposition of the effects of hyperbolic doubt on our epistemically best perceptions. Their collective weight renders them his most considered and measured statements on the subject.
There are four master texts. I have added text markers that cross-reference with the three theses above (along with further theses I'll address below)--the (a)-texts bear on thesis A, the (b)-texts on thesis B, and so on:
But what about when I was considering something very simple and straightforward [(c)] in arithmetic or geometry, for example that two and three added together make five, and so on? Did I not see at least these things clearly enough to affirm their truth? Indeed, the only reason for [(e)] my later judgement that they were open to doubt was that it occurred to me that perhaps [(f)] some God could have given me a nature such that I was deceived [(a)] even in matters which seemed most evident [manifestissima]. And whenever my preconceived belief in the supreme power of God comes to mind, I cannot but admit that it would be easy for him, if he so desired, to bring it about that I go wrong [(a)] even in those matters which I think I see utterly clearly [evidentissime intueri] with my mind's eye. Yet [(d)] when I turn to the things themselves which I think I perceive very dearly, I am so convinced by them that I spontaneously declare: let whoever can do so deceive me, he will never bring it about that I am nothing, so long as I continue to think I am something; or make it true at some future time that I have never existed, since it is now true that I exist; or bring it about that two and three added together are more or less than five, or anything of this kind in which I see a manifest contradiction. And since I have no cause to think that there is a deceiving God, and I do not yet even know ['or sure whether there is a God at all, any reason for doubt which depends simply on this supposition is a very slight and, so to speak, metaphysical one. But in order to remove even this slight reason for doubt, as soon as the opportunity arises I must examine whether there is a God, and, if there is. whether he can be a deceiver. For [(b)] if I do not know this, it seems that I can never be quite certain about anything else, (Third Meditation; 2:25, AT 7:35-36) Admittedly [(d)] my nature is such that so long as I perceive something very clearly and distinctly I cannot but believe it to be true. But my nature is also such that 1 cannot fix my mental vision continually on the same thing, so as to keep perceiving it clearly; and often the memory of a previously made judgement may come back, [(e)] when I am no longer attending to the arguments which led me to make it. And so other arguments can now occur to me which might easily undermine my opinion, [(b)] if I were unaware of God; and I should thus never have true and certain knowledge about anything, but only shifting and changeable opinions. [(c)] For example, when I consider the nature of a triangle, [(a)] it appears most evident [evidentissime] to me, steeped as I am in the principles of geometry, that its three angles are equal to two right angles: and [(d)] so long as I attend to the proof, I cannot but believe this to be true. But [(e)] as soon as I turn my mind's eve away from the proof, then in spite of still remembering that 1 perceived it very dearly, I can easily fall in to doubt about its truth, if I am unaware of God. For [(f)] I can convince myself that I have a natural disposition to go wrong from time to time [(a)] in matters which I think I perceive as evidently as can be [evidentissime percipere]. (Fifth Meditation: 2:48, AT 7:69-70) [The mind] finds certain common notions from which it constructs various proofs; and, [(d)] for as long as it attends to them, it is completely convinced of their truth. For example, the mind has within itself ideas of numbers and shapes, and it also has such common notions as: [(c)] If you add equals to equals the results will be equal; from these it is easy to demonstrate that the three angles of a triangle equal two fight angles, and so on. And so [(d)] the mind will be convinced of the truth of this and similar conclusions, so long as it attends to the premisses from which it deduced them. But [(e)] it cannot attend to them all the time; and subsequently, recalling that it is still ignorant as to whether it [(f)] may have been created with the kind of nature that makes it go wrong [(a)] even in matters which appear most evident [evidentissima], the mind sees that it has just cause to doubt such conclusions, and that [(b)] the possession of certain knowledge will not be possible until it has come to know the author of its being. (Principles, 1:197, AT 8a:9-10) In your second objection you [Regius] say: [(c)] "the truth of axioms which are clearly and distinctly understood is self-evident." This too, I agree, is true, during the time they are clearly and distinctly understood; for [(d)] our mind is of such a nature that it cannot help assenting to what it clearly understands. [(e)] But because we often remember conclusions that we have deduced from such premisses without actually attending to the premisses themselves, I say that on such occasions, if we lack knowledge of God, we can imagine that the conclusions are uncertain even though we remember that they were deduced from clear principles: because perhaps [(f)] our nature is such that we go wrong [(a)] even in the most evident matters [evidentissimis]. Consequently, even at the moment when we deduced them from those principles, we did not have know/edge of them, but only a conviction of them. I distinguish the two as follows: there is conviction when there remains some reason which might lead us to doubt, but knowledge is conviction based on a reason so strong that it can never be shaken by any stronger reason. [(b)] Nobody can have the latter unless he also has knowledge of God. (1640 letter; 3:147, AT 3:64-65)
The first two master texts are pivotal in the Meditations. Upon clarifying the nature and scope of hyperbolic doubt, the first master text concludes with the meditator declaring his need to "examine whether there is a G6d, and, if there is, whether he can be a deceiver." This marks the beginning point of an extended argument--running through the Third, Fourth, and Fifth Meditations--aimed at overcoming the deceiving God hypothesis (and equivalent doubts). (10) The second master text marks the end point of this extended argument. It is part of a four-paragraph passage (at the end of the Fifth Meditation) in which Descartes announces victory, while summarizing how his theistic strategy does finally prevail over skepticism, thereby instituting the divine guarantee. The third master text is from Principles 1, and the fourth from a letter to Regius. Though neither comes from the Meditations, both articulate the same themes. That the four master texts tell a clear and consistent story is remarkable. For the first master text is notorious in the literature, because of the seeming difficulty in reconciling its statements with the broader epistemological program. Though many commentators have wanted to discount it, as an anomalous passage at odds with Descartes's considered views, the striking fact is that its statement are reiterated in each master text. (11) What commentators should conclude is that, collectively, these passages clarify central doctrines that any adequate interpretation must explain. (12)
Concerning Broughton's thesis A: The (a)-remarks of each master text systematically employ superlatives, thereby expressing that the deceiving God hypothesis defeats our certainty even in its greatest degree--our most evident perception. (13) Other parallel passages convey the same (cf. AT 7:77, 7:141,8a:6, 8a:16). Taken at face value, the (a)-texts imply that thesis A is false--that there is no privileged class of super-indubitables utterly immune to doubt. (14)
Concerning Broughton's thesis B: This thesis is a corollary of A-the doubt-immunity of the super-indubitables implies (given other details of Descartes's program) that indefeasible knowledge of them does not depend on knowledge of a non-deceiving God. Taking the (b)-texts at face value, however, they clearly imply the negation of thesis B.
Broughton is aware of texts unfavorable to her A and B. She addresses one of them--the (b)-text occurring in the first master text--offering two notes in her defense. First, she notes that
Descartes expresses only tentatively the thought that at this stage of his inquiry he cannot be certain of anything: he says this seems to him (videor) to be true. (182)
Broughton thus reads Descartes as invoking the appearance-reality distinction: though it seems that everything is in doubt, perhaps this is not really the case. But 'seeming' qualifications need not imply a contrast with reality. Indeed, none of the other (b)-texts include the qualification. Broughton's second note is a suggestion of how to read the broader paragraph (that is, the first master text). She suggests that it is intended to convey the following confusion:
Well, it is hard for him to see why some clear and distinct ideas cannot be doubted, while others--equally dear, distinct, and compelling--can be. The sentence in question is an expression of this difficulty, I think. (183)
The passage, however, does not bear out this suggestion. There is no indication that Descartes means to be contrasting two sets of clear and distinct ideas--the super-indubitables versus ordinary cases. To the contrary, the difficulty expressed in the passage concerns the vulnerability to hyperbolic doubt of al/maximally evident matters--barring when we're attending to them (a point to which I'll return). Other clear passages confirm the master texts relative to (a) and (b). The Fifth Meditation reiterates the theme twice (in addition to the second master text):
And what is more, I see that the certainty of all other things depends on this [the certainty of an existing, non-deceiving God], so that without it nothing can ever be perfectly known. (2:48, AT 7:69) Thus I see plainly that the certainty and truth of all knowledge depends uniquely on my awareness of the true God, to such an extent that I was incapable of perfect knowledge about anything else until I became aware of biffs. (2:49, AT 7:71)
The Second Replies clarifies that atheists cannot achieve indefeasible knowledge--not even of the most evident matters:
Now since we are supposing that this individual is an atheist, he cannot be certain that he is not being deceived on matters which seem to him to be very evident [evidentissma] (as I fully explained). And although this doubt may not occur to him, it can still crop up if someone else raises the point or if he looks into the matter himself. (2:101, AT 7:141)
Concerning Broughton's thesis C: On her interpretation, mathematics is not included with our epistemic best (the super-indubitables). When taken at face value, however, the master texts again tell a very different story. Each of the (c)-texts uses mathematics as an illustration of superlative certainty. (15) The contra-C theme is underscored in the first master text, in that arithmetic and the I exist are put on an epistemic par--as being of the same epistemic kind: "let whoever can do so deceive me, he will never bring it about that I am nothing, so long as I continue to think I am something ... or bring it about that two and three added together are more or less than five, or anything of this kind in which I see a manifest contradiction" (my italics).
The master texts thus present a formidable challenge to Broughton's interpretation. It is, of course, possible that the master texts are systematically misleading--that, in fact, Descartes affirms A, B, and C, notwithstanding the master texts. One might even take such textual skepticism to find support from the conventional wisdom surrounding the I exist: accordingly, the I exist is (consistent with A) not undermined by hyperbolic doubt, nor (consistent with B) does knowledge of it depend on God. If, then, the master texts paint a misleading, one-sided textual picture with regard to A and B, perhaps a more balanced look at the broader textual story is more favorable to Broughton's interpretation. This is a plausible rebuttal to the case I've thus far argued. To combat it, I want to turn attention to the relevant texts surrounding the I exist.
It is true that Descartes characterizes the pre-theistic [exist as "knowledge." The meditator refers to his certainty of the I exist as his "first item of knowledge [prima cognitione]" (2:24, AT 7:35). The parallel passage of the Principles adds: "this piece of knowledge [cognitio]--I am thinking, therefore. I exist--is the first and most certain of all to occur to anyone who philosophizes in an orderly way" (1:194, AT 8a:7). Though such statements might seem to imply A and B, they belie the subtleties of Descartes's use of cognitive terms.
Close inspection reveals that Descartes adopts some version of a contextualist stance. That is, he does not restrict the use of 'knowledge' talk to the fully-indefeasible brand of cognition that he seeks. He instead employs such talk, approvingly, for all clear and distinct cognition, (16) even pre-theistic, defeasible cognition. When characterizing the defeasible cognition of atheists, Descartes uses the very same epistemic language--'cognitio'-talk--as he uses to characterize the (pre-theistic) I exist:
The fact that an atheist can be 'clearly aware [clare cognoscere] that the three angles of a triangle are equal to two right angles' is something I do not dispute. (2:101, AT 7:141)
Descartes immediately adds, of the atheist's "awareness" (cognitionem), that it "can be rendered doubtful" by hyperbolic doubt (ibid.). Pretheistic, clear and distinct knowledge (= cognitio) is achievable, even though defeasible. (17) Arguably, this is a desirable outcome. It allows for the expansion of genuine knowledge, say, mathematical knowledge, without insisting that the practitioners of math found their results on theistic metaphysics. This contextualist element bears on Descartes's characterization of the I exist, as an item of "knowledge" (cognitio). Evidently, the characterization implies nothing stronger than that the I exist is clear and distinct, but not (yet) that it is indefeasible.
So, according to Descartes, some cases of knowledge are defeasible, while other cases are not. Does not this indiscriminate use of 'knowledge'-talk encourage confusion? Though it does, in many critical passages Descartes marks the indefeasible brand of knowledge of interest to him by using strengthening adjectives in conjunction with the language of certainty and knowledge: he refers not to mere certainty, but complete certainty; not to mere knowledge, but perfect knowledge; and so on. Consider some examples from key passages concerning indefeasible knowledge (all italics are mine):
For if I do not know this [whether God is a deceiver], it seems that I can never be completely) certain [plane certus] about anything else. (2:25, AT 7:36, trans, altered) so that without it [knowledge of God] nothing can ever be perfectly known [perfecte sciri]. (2:48, AT 7:69) if I were unaware of God; and I should thus never have true and certain knowledge [veram et certam scientiam] about anything. (2:48, AT 7:69) Now, however, I have perceived that God exists, and ... there are no counter-arguments which can be adduced to make me doubt it, but on the contrary I have true and certain knowledge of it [veram et certam de hoc haveo scientiam]. (2:48, AT 7:70) I was incapable of perfect knowledge [perfecte scire] about anything else until I became aware of him [God]. And now it is possible for me to achieve full and certain knowledge [plane nora et certa] of countless matters. (2:49, AT 7:715) For the supposition which we are making here is of a conviction so firm that it is quite incapable of being destroyed; and such a conviction is clearly the same as the most perfect certainty [perfectissima certitudo] . (2:103, AT 7:145) Had I not been looking for greater than ordinary certainty [certitudinem vulgari majorem],... (2:159, AT 7:226)
Broughton acknowledges that there's some textual evidence for a distinction between two brands of certainty/knowledge (183). She denies that any such distinction can be used--as I am using it--to establish the I exist as a counterexample to her theses A and B. For, as she writes, "this would require the meditator somehow to be taking back what he said in the Second Meditation"; adding that there is "no sign that Descartes meant for the meditator to take this reasoning back" (184-85).
Though one can read the Second Meditation texts in this way, I want to argue that they are not best read in this way. To help elucidate why we can regard the I exist as defeasible, but without "taking back" anything said in the Second Meditation, I'll consider three further theses. Unlike A, B, and C, these further theses do not represent core elements of Broughton's interpretation. But whereas A, B, and C, are in tension with the master texts, I'll argue that the further theses are clearly supported in the master texts (again, see the corresponding text markers):
D. While the supreme evidentness of clear and distinct perception is occurrent, the mind--even the pre-theistic mind cannot but assent to the proposition it perceives.
E. The supreme evidentness of pre-theistic, clear and distinct perception is apparent only while such perception is occurrent.
F. The deceiving God hypothesis undermines our most evident propositions not during moments of clear and distinct attention to particular such propositions, but upon redirecting attention to the possibility that our cognitive faculties are defective.
Concerning thesis D: The second and third master texts clearly convey what Descartes elsewhere writes: that independent of a divine guarantee, "the nature of my mind is such that I cannot but assent to these things, at least so long as I clearly perceive them" (2:45, AT 7:65). Given our cognitive wiring, we cannot avoid forming judgments in such utterly convincing cases. The role of the guarantee of clear and distinct perception is not to augment its evidentness,, but to guarantee that such evidentness entails truth.
Concerning thesis E: What this thesis adds, to D, is that the utter convincingness of pre-theistic clarity and distinctness is non-enduring. It lasts only while our perception remains clear and distinct. This is significant, because, as the meditator says (in the third master text), "I cannot fix my mental vision continually on the same thing, so as to keep perceiving it clearly." It is during moments of inattention that hyperbolic doubt can creep in.
Concerning thesis F: The deceiving God hypothesis poses an indirect doubt of evident propositions, in the sense that it is operative only when the mind's attention is redirected from the clarity and distinctness of the proposition: the doubt is inoperative while its clarity and distinctness are occurrent--in accord with D. Toward what is the mind's attention diverted, in moments of doubt? Toward the general ground of their evidentness--their clarity and distinctness. The doubtful mind's attention is on some general rubric referring to such evident propositions: perhaps a deceiving God made me with a nature whereby I am "deceived even in matters which seemed most evident"--whereby "I go wrong even in those matters which I think I see utterly clearly" This indirect character of the doubt is indicated in each master text (cf. also AT 3:64-65, 7:77, 8a:6, 8a:16). The implication is clear: vulnerability to hyperbolic doubt will remain as long as the general veracity of clarity and distinctness is not itself clear and distinct. Establishing this veracity is the very point of Descartes's theistic maneuvers.
The degree of hyperbolic doubt's indirectness is a function of whether the undermined proposition is itself inferential. In the case of self-evident truths, the doubt is fully indirect. That is, in the case of a "very simple and straightforward" matter, as illustrated in the first master text, the proposition resists doubt if I am attending directly to it: because self-evident, direct attention induces assent. In the case of inferential truths (conclusions), the doubt need only be partially indirect; that is, it can be partially direct. I can directly attend to a conclusion and doubt it if I am not also attending to its grounding argument, for in such circumstances my perception of the conclusion will not be clear and distinct. (18) The second, third, and fourth master texts characterize the operation of doubt in these inferential cases. (19)
This understanding of the indirect operation of hyperbolic doubt helps clarify why we can regard the I exist as defeasible, without "taking back" anything said in the Second Meditation, and why, more generally (contra thesis A), there could not be a class of pre-theistic super-indubitables utterly immune to doubt. Nothing need be taken back, because what the Second Meditation asserts implies nothing stronger than thesis D. The meditator notices that hyperbolic doubt is inoperative relative to his occurrent attention to the particular proposition, "I am, I exist"--the proposition is indubitable "whenever it is put forward by me or conceived in my mind" (2:16-17, AT 7:25). This is fully compatible with the thesis that the I exist is defeasible via an indirect doubt as characterized in the first master text; indeed, that very passage reiterates the assertion of the Second Meditation, namely that when we "turn to the things themselves," we succeed in fending off the deceiving God hypothesis. More generally, the indirect character of the doubt means that it is unable to discriminate between the various propositions it undermines. The doubt is an equal opportunity defeater. Its shadow casts over every evident proposition. All clearly and distinctly perceived matters thus share the same susceptibility to doubt, implying that there can be no pre-theistic, indefeasible knowledge. The master texts can indeed be taken at face value--even in the (a)- and (b)-remarks.
Together, theses D, E, and F, help explain one aspect of how Descartes avoids circularity in his broader argument for the divine guarantee. The theses explain the sense in which the meditator knows (= clearly and distinctly perceives) the premises leading to the conclusions about God, even while those steps remain defeasible.
There is, then, a very powerful textual case against Broughton's interpretation. At the core of her interpretation are theses A, B, and C, yet clear statements in numerous, pivotal texts contradict all three. Bear in mind, however, that my argument has centered around a clarification of what the master texts sa3 rather than an examination of how best to interpret them. The best interpretation will not always take every text at face value. It is always open to the interpreter to try to show that the philosopher in question does not mean what she or he says. In cases involving numerous and complex texts as in the present case--the best interpretation, all things considered, may well require a strained reading of seemingly transparent texts. When a straightforward explanation remains elusive, the thoughtful interpreter may need to explain away important texts, perhaps in the name of charity. The consensus emerging from centuries of careful scholarship has it that at least some of the relevant texts (in projects like Broughton's) will require creative interpretative efforts, including strained readings. Doing justice to Broughton's project requires that we place it in a longstanding history of strained efforts to make sense of Descartes's epistemology in a manner sensitive to his undisputed genius. In this context, hers ranks among the more creative and subtle efforts.
I thank Janet Broughton and Alan Nelson for valuable comments on a draft of this paper. I also thank the graduate students in my Descartes Seminar for their lively and helpful discussion of Broughton's book: Britta Berkey, Diana Buecafurni, Charles Hudgins, David Laraway, Marina Lelanuja, Gordon Mower, Joseph Ulatowski, Anna Vaughn, Shelley Ver Steeg, and Carmine Vincenzo.
(1) Princeton: Princeton University Press, 2002. Page citations not associated with any other work refer to this book. References to Descartes's works include two sets of volume and page numbers. The first refers to Philosophical Writings of Descartes, ed. and trans. John Cottingham, Robert Stoothoff, and Dugald Murdoch (Cambridge: Cambridge University Press, 1984). The second, marked by 'AT', refers to Oeuvres de. Descartes, ed. Charles Adam and Paul Tannery (Paris: J. Vrin, 1904).
(2) By 'clear and distinct truths' (or propositions), I mean those capable of being clearly and distinctly perceived--it being perception, not propositions, that are the fundamental bearers of clarity and distinctness.
(3) Cf. Sextus Empiricus, Outlines, of Pyrrhonism, bk. 2, chap. 4.
(4) Cf. Posterior Analytics.
(5) In part 1 of her book, Broughton argues for significant differences in the way Descartes and the ancient skeptics approach issues of doubt (see especially chapters 2 and 5). Granting these differences does not mitigate the dilemmic problem before us. For what is problematic in the Pyrrhonian dilemma does not arise from something uniquely Pyrrhonian--as if presupposing outdated doctrines, h arises from the perfectly general need to avoid circularity and indefinite regress.
(6) The Fifth Meditation meditator claims that though his apprehension of God is discovered by arguments, it emerges as self-evident; he claims to come eventually to see that nothing "is more serf-evident than the fact that the supreme being exists" (2:47, AT 7:69). In the Second Replies, Descartes adds that if his readers will "spend a great deal of time and effort on contemplating the nature of the supremely perfect being," this will help them to "realize that God exists"--a truth that "will eventually be just as self evident to them as the fact that the number two is even or that three is odd" (2:115, AT 7:164-65). I take Descartes to hold that the "arguments" for God function as aids to discovery, rather than as enduring epistemic foundations. In Plato's Meno, the role of the Socratic midwife is to help the slave boy discover the innate truths within him--to be "able to dig out certain truths from his own mind," as Descartes understands the story (3:222-23, AT 8b:167). The midwife assists the slave boy in discavering innate knowledge, but she is not the ground of his knowledge, as if it were founded on authority. While making an analogy to the Socratic assistance provided to the slave boy, Descartes claims that "our knowledge of God is of this sort" (ibid.). On this understanding of the project, knowledge of God is, in the final analysis, grounded not in argument, but in serf-evidence. And once the groundbreaking truth about God is self-evident, the deceiving God hypothesis is rendered unformulable: it "implies a conceptual contradiction-that is, it cannot be conceived" (3:222; AT 8b:60). I argue all of this in my "Circumventing Cartesian Circles" (with Alan Nelson), Nous 33 (1999): 370-404.
(7) Employing dependence arguments does presuppose various logical principles. Broughton tries to use dependence arguments to explain this: "a meditator cannot carry forward any sort of rational inquiry without accepting that, say, p and 'If p, then q' together exclude the falsity of q. The rational character of an inquiry means that someone conducting it must accept basic logical principles such as this. Thus if the meditator is using the method of doubt as a method for rational inquiry, it is a condition of his using the method that he accept these logical principles" (163). But this would seem question-begging. For one must already accept logical principles in order to conclude that using the method of doubt requires one to accept logical principles.
(8) Nor does it follow from the fact that a complex proposition has an imbedded inference--as in "I am thinking therefore I am"--that the ground of the complex proposition is inferential. Modus ponens, an inference, is widely thought to have the starts of a serf-evident axiom.
(9) Cf. note 6.
(10) It is widely accepted that the Third Meditation is integral to Descartes's case for the divine guarantee of clear and distinct perception. Less well understood is the role of subsequent Meditations--Broughton indeed makes clear that her book has "little to say" about the Fourth and Fifth Meditations (xi). In my "The Fourth Meditation" (Philosophy and Phenomenological Research 59 (1999): 559-91), I argue that the Fourth Meditation is integral to the broader argument for the divine guarantee, that, indeed, we should take at face value Descartes's remark that "in the Fourth Meditation it is proved that everything that we clearly and distinctly perceive is true" (Synopsis, 2:11, AT 7:15). In my "Circumventing Cartesian Circles," I argue moreover that the Fifth Meditation is integral to Descartes's case for the divine guarantee (cf. note 6).
(11) Arguably (though I won't argue it here) there are no unequivocal texts (much less a body of them) that contradict the master texts, though there are isolated remarks that can be read in contrary ways.
(12) There are some notable passages that I do not include with the master texts (though they are fully consistent with them). I do not include the First Meditation passage introducing the deceiving God hypothesis. What is of interest about the master texts is what they say about the effects of hyperbolic doubt on our most certain perceptual states. Prior to the Second Meditation, the meditator has not yet learned to perceive anything with clarity and distinctness (cf. AT 7:460). Nor do I include the Second Meditation I exist passage. What is crucial about the master texts is that they clarify the operation of hyperbolic doubt both during moments of attentive consideration to evident propositions, and later moments of inattention. Unlike the master texts, that Second Meditation passage deals only with the former, but not the latter. I return to this theme in connection with theses D, E, and F, below.
(13) In my use of 'evidentness'-talk (and its cognates), I do not exclude the a priori apprehension of truths (in the post-Kantian sense of a priority). As the (a)-texts convey, Descartes uses the language of evidentness in a broad sense to encompass all manner of perceptual grounds.
(14) One might object to my characterization of the (a)-texts by maintaining, with Stephen Menn, that Descartes qualifies his remarks precisely to invoke an appearance-reality gap. According to Menn, the point of hyperbolic doubt (contra what I am arguing) is not that we might be deceived about matters that really are most evident; the point is that "God could deceive us 'even about those things that seemed very manifest', or again, "even in the things I think I see evidently with the eyes of" the mind'" (Descartes and Augustine (Cambridge: Cambridge University Press, 1998), 233). Menn's reading, however, cannot be sustained. First, it is false that Descartes always employs some version of the 'seeming' qualification. In the (a)-portion of the fourth master text, the doubt is simply that "we go wrong even in the most evident matters." Second, Descartes regularly does use some version of the 'seeming' qualification in contexts that do not refer to an appearance-reality gap. For example, on the heels of the ontological argument--a context intended to convey the superlative clarity of his perception of God--Descartes writes: "Although it needed close attention for me to perceive this, I am now just as certain of it as I am of everything else which appears [videtur] most certain [certissimum]" (2:48, AT 7:69). Moreover, in the very paragraph Menn discusses (the first master text), the meditator includes the I exist with the list of items that, as he says, "I think I perceive very clearly." Menn's reading of the 'seeming' qualifications, therefore, has unacceptable consequences--implying that the meditator does not really achieve clear and distinct perception of God, nor of the I exist. We must recognize that Descartes often employs 'seeming'-talk not to convey mere seemings (Menn's view)--as if in contrast with reality--but to emphasize what is actually apparent in our ideas. Descartes's introspective inquiry requires that we carefully distinguish what is perceived clearly--what actually appears before the mind--from what is confusedly judged. In the paragraph immediately preceding the first master text, the meditator reflects on his previous habit of conflating these, with regard to external things: "But what was it about them that I perceived clearly? Just that the ideas, or thoughts, of such things appeared before my mind [menti meae obversari]" (2:24, AT 7:35, my italics). In Principles 1:66, Descartes notes that even sensations "may be clearly perceived provided we take great care in our judgements concerning them to include no more than what is strictly contained in our perception--no more than that of which we have inner awareness" (1:216, AT 8a:32). Thus, in the Second Meditation, upon learning to discriminate what is actually apparent in his ideas from what is not, the meditator is able to say: "Yet I certainly seem [videor] to see, to hear, and to be warmed. This cannot be false" (2:19, AT 7:29, my italics).
(15) The fourth master text (to Regius) is not as clear on this point, though I take the reference to ax/0ms as an allusion to math.
(16) Translators regularly render Descartes's uses of 'cognitio' and 'scientia', and their cognates, into the English "knowledge', and its cognates.
(17) The second objectors are confused on this count. They mistakenly attribute to Descartes the view that one "cannot know [cognoscere] anything clearly and distinctly until [having] achieved clear and certain knowledge of the existence of God" (2:89, AT 7:124-25). In his reply, Descartes denies that the certainty of clarity and distinctness depends on God, a denial immediately followed by his remarks about atheist geometers.
(18) This difference between self-evident and inferential truths explains one of Descartes's difficult replies to the second objectors. Because the I exist is self-evident, it counts as an item of clear and distinct knowledge--even prior to the divine guarantee--whenever we attend to it. Conclusions, however--absent the divine guarantee--can be items of knowledge only if we are attending to their grounding arguments. Confused about this distinction, the second objectors mistakenly pose the following objection concerning the I exist yon cannot "yet clearly and distinctly know that yon are a thinking thing, since, on your own admission, that knowledge depends on the clear knowledge of an existing God; and this you have not yet proved in the passage where you draw the conclusion" (2:89, AT 7:125). The objection wrongly presupposes that the I exist is vulnerable to hyperbolic doubt in the manner of conclusions. Thus, in his reply, Descartes writes: "when I said that we can know nothing for certain until we are aware that God exists, I expressly declared that I was speaking only of knowledge of those conclusions which can be recalled when we are no longer attending to the arguments by means of which we deduced them" (2:100, AT 7:140). The reply continues with Descartes clarifying that the I exist is not inferential, but instead "self-evident."
(19) Note that this partially indirect doubt of a conclusion is parasitic on a fully indirect doubt of its serf-evident premises. So, Descartes's characterization of the indirect doubt for conclusions adds to the characterization for self-evident truths, rather than conflicting with it.
University of Utah