Descartes’ famous rule states that everything we apprehend lucidly (clearly and distinctly) is true. If the rule never fails, then clearly, intuiting the Cogito and the existence of an omni-God entails the truth of both.[1] But on what basis does Descartes conclude that the rule is true? (I will henceforth refer to the ‘clarity and distinctness entails truth’ rule as ‘CDT’.) There does not seem to be a basis over and above intuition itself, for it is in part from the conclusion that there is an omni-God that Descartes seems to conclude CDT. But unless we presuppose CDT, it seems we can’t trust that the omni-God thesis (or any thesis, for that matter) is true. This point is crucial to what known as the circularity objection.

The objection insists that Descartes put forth the following set of arguments (where (A1) and (A2) are assumptions):

Another problem closely intertwined with the circularity objection is alluded to by Descartes himself.

It is tempting to conclude from this that Descartes is willing to swallow the consequences of the circularity objection, and conclude that the things we intuit may be false. But if we conclude this, we inherit the burden of explaining away Descartes’ insistence on the following:

According to Rubin (1965), it would not be wise to charge Descartes with circularity. The reason: the assumption that there is an omni-God plays a causal, but not an evidential role in the formulation of CDT (p. 207). CDT plays the same role when we conclude from it that there is an omni-God. So the justification of these basic beliefs, formed as a consequence of pure intuition, is not given via derivation (nor via justificatory appeal to any explicit argument), but in virtue of the fact that it cannot be shown to be false, and that one cannot help but accept it as true if one understands it (given our psychological makeup—the further assumption is that we all generally have the same psychological makeup [CSM 1, p. 111]). It is important to note that although arguments are used, they do not play a justificatory role on Rubin’s view.

I agree that the link between CDT and the omni-God thesis is not as it prima facie appears to be, but the causal relation between the compulsion to assent and the truths assented to does not imply the absence of justification. In order to show this, it will be necessary to consider the status that reason, derivation, and basic beliefs have in the Cartesian system.

2.1 Justification

Descartes’ method assumes that the mental operations of intuition and deduction "are the simplest of all and quite basic". This explains why his method "cannot go so far as to teach us how to perform" these mental operations. "If our intellect were not already able to perform them, it would not comprehend any of the rules of the method, however easy they might be" (CSM 1, p. 16). And of course, it is plausible to think that our intellect does, in fact, lead us to an understanding of Descartes’ rules.

So it seems that Descartes did not attempt to justify the reliability of the faculty of reasoning from self-evident principles to further self-evident conclusions. It is not only wrongheaded and useless, but also harmful to analyze the tool of inference itself. To avoid corrupting good sense, we must practice the rules inherent in "the kind of logic which teaches us to direct our reason with a view to discovering the truths about which we are ignorant" (CSM 1, p. 86). But my primary concern for the moment is that Descartes repudiates any technique attempting to define or further elucidate anything that is already understood: this includes the conclusion that reason is truth-conducive, that there is an omni-God, and that the Cogito is true. This point goes for all basic, fundamental beliefs, for "[a]ll the clear and distinct notions which are in us" are already understood and require no definition or further elucidation (ibid.).[6] Basic beliefs are principles from which he deduces other truths. They are principles "known for all time and indeed accepted as true and indubitable by everyone," which indicates that they are clear and self-evident; this, in turn, entails that they are true (CSM 1, p. 184).

Descartes aimed to take advantage of the deductive principles of (what we may now refer to simply as) mathematics without adopting its obfuscating defects. The defects arise out of unneeded postulates and laws that only serve justify errors which, given the superfluous presuppositions, cannot be identified as errors on pain of contradiction. The beneficial results of the method: derivation of further self-evident conclusions.

Not only is it clear that the intellect, when used properly,[7] does not lead us to falsity, but there is a contradiction is supposing that it does (CSM 2, p. 260). Here, the role of the CDT rule becomes evident. There seems to be no pretense, on Descartes’ part, of explicitly justifying the rule, belief in basic, self-evident truths, or reason: to demand an explicit justificatory argument for the reliability of reason is ipso facto to doubt its truth-conduciveness regarding the apprehension of the basics. The resulting contradiction: thinking P to be false (by doubting reason) while assenting to P (because P is self-evident and assenting to P is unavoidable). The fact that this is a psychological contradiction will turn out to be crucial for purposes of understanding his technique. More on this shortly.[8]

In sum, it seems that the rules by which we arrive at derived truths are also too fundamental in nature to require—or even allow for—justification via explicit argument.[9] What remains to be seen is whether Descartes would repudiate attempts at any kind of justification.

Insofar as the intellect has been given to us by our omni-God, it cannot compel the will to assent to falsity (which is "under the guise of truth").[10] This much must be taken for granted, so to speak. This is why he states that his method consists of

Of course, the circularity objection works only if the belief in CDT is compelled by a justificatory argument that has CDT as its conclusion. Circularity thus seems to be avoidable via the irresistibility thesis: no kind of epistemic justification is necessary for an agent to be warranted in believing CDT and the omni-God thesis. In what follows, I address some possible objections to and consequences of this view in order to motivate my own.

Two objections to the irresistibility thesis are suggested by the following:

It seems that unless he can convince himself that there is a veracious God, he is not warranted in concluding that any other self-evident belief is true; this despite the fact that we have a compulsion to believe. Descartes shows a high regard for justification.

If the omni-God belief is necessary for justification of all our beliefs, then it is necessary for justifying CDT. Warranted belief in CDT, the circularity objection states, relies on the veracious God for its justification. This raises the objection once again, leaving us with a dilemma: either Descartes makes the elementary error of arguing in a circle, or he disregards the need of justification, bypassing it all together via psychology. Neither option is attractive. The first has it that Descartes commits an elementary blunder, and the second that he was a psychologistic rhetorician.

The solution may lie in the irresistibility thesis. However, if Descartes thinks the (clear and distinct [15]) belief in a veracious God is necessary for justification, it seems the solution is mistaken. For this entails that Descartes does think some species of justification is required for warranted belief in any primitive truth [16] (other than belief in a veracious God, of course; this seems, prima facie, to be the foundational belief).

Let us contrast, for elucidatory purposes, derivation for justificatory purposes with its counterpart (call it simple derivation for lack of a more attractive term), the latter of which simply causes a "spontaneous impulse" to assent.[17] Simple derivation, as I shall characterize it here, does not endow beliefs with justification.

The syntactic form of a simple derivation is indistinguishable from that of a justificatory one. The difference lies in their intended purposes and uses. For instance, when Descartes goes through the motions of argumentation in the Meditations for truths that are already self-evident (to him), he does so as a way of leading the reader to perceive that which he himself has already perceived lucidly. In this way, he causes the reader to form clear and distinct perceptions by taking him through every step in the process leading to the belief that there is a veracious omni-God. If the method invoked is that of simple derivation, it does not play the justificatory role that brings about the circularity objection.

Of course, Descartes is interested in deriving certain truths, all of which become self-evident as a consequence of abiding by his simple rules. So if we follow the Cartesian method flawlessly, we should begin with the self-evident, and end with further self-evident conclusions. The question thus arises: what distinguishes beliefs resulting from justificatory and simple derivation? It cannot be their degree of lucidity, for they are all equally lucid once derived. If the suggestion is correct that derivation plays no justificatory role for self-evident truths, what prevents us from thinking that none of the arguments Descartes offers are justificatory? If none are justificatory, it seems we have a reductio of the *irresistibility thesis.

     
        3.1.1 Externalist Justification

We saw that the *irresistibility thesis, which denies that Descartes thought justification was necessary, is problematic because he appeals to arguments involving the omni-God thesis. These arguments do seem to play justificatory roles. Yet if they do, CDT is justified via appeal to God, and vice versa. To see why Descartes can have both justification and an argument that is not viciously circular, we must distinguish two drastically different approaches to epistemic justification.

Internalist positions regarding epistemic justification require that one be able to explicitly provide reasons, or even a justificatory argument, for the truth of a belief. For an internalist who also happens to be a coherentist, for instance the perceptual belief, there is an apple before me, can and must be justified with an argument that appeals to the reliability of perception, or something very much like that.[18] But such an argument, the externalist will insist, is not required for justification of perceptual beliefs (or some other species of basic beliefs).[19] What then, is the role played by simple derivation for the externalist? The answer is that simple derivation, just like any explicit justificatory reason- giving procedure, is not necessary for justification.

On the one hand, Descartes offers an account of the nature of a reliable belief-forming process: reason. He seems to justify his claim that reason is reliable by appeal to a veracious God, providing an explicit argument for the reliability of reason. This is what brings about the circularity objection. For when we form a belief about the omni-God, or any belief, for that matter, we use reason to arrive at that belief. If reason is unreliable, then the belief may be false. But in order to justify belief in the reliability of reason, a veracious God premise is needed. How then, can it be claimed that for Descartes, explicit arguments for the reliability of reason are not needed for justification?

If we distinguish the Cartesian account of the reliability of reason from the conditions for epistemic justification, we see that there is no inconsistency in insisting that Descartes is an externalist about justification. Quite the contrary: only when we recognize that these are two distinct types of claims can we understand what Descartes meant when he replied to what Mersennes believed to be a counterexample. Mersennes tries to put forth a reductio of Descartes’ view by cleverly showing that if clear and distinct knowledge of the existence of God is necessary for accepting the Christian religion, and if the only way to avoid sinning is to assent only to what is clearly and distinctly perceived, then the atheist actually sins in accepting Christianity, for he has no such clear and distinct knowledge (CSM 2, p. 90). Descartes turns the tables and claims that the reason why infidels are sinners is that they do not make use of the natural light, or reason (ibid.). The consequence of Mersenne’s example is one Descartes gladly accepts. The point is that the infidel does not sin in virtue of the fact that she assents to a conclusion derived from some argument that lacks an explicit appeal to the omni-God, but because she did not come to her conclusion via the reliable facilty of reason. If Descartes had insisted that she sins because she made no explicit appeal to the omni-God premise, it would be plausible to conclude that his conditions for justification (for belief in the Christian religion) are externalist. But this is not why she sins. If Christian doctrines are obscure to her, and she "[embraces] them by fallacious arguments", she does not thereby sin: she sins for a related, but different reason: the agent sins "by not using [her] reason correctly" (CSM 2, p. 105). If she does not allow her reliable process to carry out its natural function, then there is a sense in which she sins.[20] She thereby lacks justification.

Descartes’ mention of fallacious arguments is misleading, and seems to suggest the requirement that an agent must explicitly invoke justifying arguments. But the above considerations make another reading possible. Namely,

1. If some agent, S, refuses to use her reliable belief-forming faculty (reason), then S fails to see clearly that there is a veracious God;
2. That S fails to see clearly that there is a veracious God entails that she has not used reason correctly;
3. If reason has not been used correctly, the resulting belief cannot be a justified one for S.

On this reading, Descartes puts forth the following externalist condition for justification:

(J) Belief P is justified if and only if P is the result of a reliable belief-forming process.

(J) is consistent with the claim that a merely causal relation between a belief and the reliable faculty of reason is necessary and sufficient for justification.[21] No explicit access to justificatory reasons are necessary. But an obvious objection arises which poses the greatest threat to this externalist interpretation.

It seems indisputable that Descartes requires, for justification of any belief, explicit justificatory access (if only dispositional) to the claim that there is a veracious God. For it appears that this is the claim in virtue of which a belief is rendered indubitable. Let us see.

The passage suggests a necessary condition for knowledge.

(C) If P is (psychologically) dubitable for some agent S, then S does not know that P.

If the only way to make S’s belief indubitable is to realize that there is a veracious God, then it seems that the omni-God thesis is an internalist justificatory requirement:

S concludes that K

However, it is possible that realizing that there is a veracious God plays a causal role in making K psychologically indubitable. (R) does not entail that an explicit justificatory premise is necessary for S’s justification in the relevant conclusion, even though one such a requirement would be consistent with the (R). Perhaps then, it is not directly via the realization that there is a veracious God that justification is bestowed upon beliefs, but via the correct use of the reliable faculty of reason; for reason is used correctly only if we invoke the omni-God thesis. So (J) may be necessary and sufficient for justification. But (J) is not an internalist condition. So an externalist reading is possible.

Support for this view is given by Descartes’ insistence that the faculty of reason is a "faculty for recognizing the truth and distinguishing it from falsehood", and that this "is clear merely from the fact that we have within us the idea of truth and falsehood". Here, Descartes immediately concludes that "reason must tend towards the truth" when used correctly (CSM 2, p. 103; my italics). The externalist position I attribute to him renders this consistent with his further claim that awareness of a veracious God is necessary for justification, for it is only necessary in that it makes possible the use of our reliable (and indeed infallible) faculty of reason.

I don’t think any other reading of Descartes can explain why he is entitled to state that the basis of all "human certainty can be founded" on the fact that immediate assent to intuited truths is irresistible—a psychological fact. For "as soon as we think we correctly perceive something, we are spontaneously convinced that it is true". And "if this conviction is so firm that it is impossible for us ever to have any reason for doubting what we [intuit], then there are no further questions for us to ask: we have everything we could reasonably want" (CSM 2, p. 103; my italics). Halting inquiry into the reliability of reason at the very point when we come to understand this psychological fact is entailed by externalism, and itself entails an externalist position; no further justificatory questions can be asked, and no further justificatory conditions can be offered once we realize that reason is as reliable as it is psychologically compelling. The realization that there is a veracious God only serves as a necessary condition for making sure that we are using the right cognitive faculty. If there were some other way to make sure we are using reason correctly, the omni-God thesis would not be necessary but sufficient. The condition that does all the epistemic justificatory work is (J). And (J) only demands the right causal link between a belief and the faculty by which it is formed.

3.2 Skepticism, Externalism, Naturalism

If the externalist thesis is right, what was the purpose of the evil demon in the Meditations? Surely the deceiver was meant to be taken seriously. If we are able to entertain the deceiver hypothesis, it seems we can only escape him by appealing to explicit reasons for rejecting him. But like any good externalist, I think Descartes does not attempt a direct refutation of the skeptic. Frankfurt’s (1965) thesis offers an explanation.

Frankfurt’s account has it that mistrusting reason is unwise on the Cartesian view because Descartes shows that "there is no good reason for doubting the trustworthiness of intuition" (p. 154). In his argument, Frankfurt makes explicit Descartes’ appeal to God: belief in Him is sufficient for guaranteeing this certainty. But on my view, this belief is only necessary for purposes of employing reason correctly. Employing reason correctly is, in turn, necessary and sufficient for being justified in assenting to the beliefs it produces. And I think my claim is more compatible with Descartes’ view.

Frankfurt concludes from the following passage that a "person is justified [in his belief] if he knows that God exists" (p. 149):

But Frankfurt’s claim is that knowledge of a veracious God is sufficient for justified belief, while Descartes insists here only that it is necessary. Descartes’ claim is that remembering that we intuited something is sufficient only if we know of the omni-God. My own interpretation has it that appeal to God is necessary for the correct use of reason; this seems more in accord with Descartes’ claim. I add that the omni-God thesis is necessary only in that it makes the use of reason sufficient and necessary for justification. And this coheres nicely with Descartes’ emphasis on the truth-conducive faculty of reason.

This brings us back to Descartes’ strategy against skepticism, and the circularity objection. Surely Descartes did not think that he was refuting skepticism by assuming that there is an omni-God. If we could build a non-fallacious argument on this assumption, we would be well on our way to a refutation of skepticism. But it cannot be so assumed. Yet Descartes was not deluded when he told Bourdin that it was by way of argument that he became "the first philosopher ever to overturn the doubt of the skeptic" (CSM 2, p. 376). For as an externalist, he is entitled to hold that it is in virtue of psychological facts that our beliefs are justified: this does not require having introspective (or conscious) access to the justificatory (cognitive) processes.

Access to an unwavering belief that there is a veracious God is not that which confers justification: such access is only necessary for the proper functioning of reason. Belief in a veracious God, as we saw, can be the result of fallacious reasoning. If it is the result of fallacious reasoning, then reason has not been employed correctly. It is for this reason that the resulting beliefs fail to be justified. The real necessary (and sufficient) justificatory condition is a reliabilist, psychological, causal one. And this point is only strengthened by the fact that for Descartes, even belief in the omni-God is the result of a psychological compulsion to assent. For this reason, failure to believe in the omni-God is important and error-conducive because it indicates that reason has not been employed properly.

In sum, the skeptic asks a question that externalism does not allow: namely, "how do we know that reason and your method are truth-conducive?". Descartes’ reply would simply be, "we cannot but think that they are". Unless we put forth independent reasons for refuting externalism regarding justification, Descartes’ response cannot be seen to conflict with his interest in epistemic justification. So although his appeal to the omni-God is at best indirectly justificatory for what it indicates about reason, the circularity objection fails because the real justificatory condition is externalist, and thus a causal one. We can thus construe his argument as follows:

(A1) There is an omni-God

(P1) My mind, and hence my faculty of reason, is the product of this omni-God

(P2) Omni-God is no deceiver

(C) So all that I clearly and distinctly apprehend is true (A2)

(A2) If I clearly and distinctly apprehend something, then it is true

(P1*) I clearly and distinctly apprehend that there is an omni-God

(C*) So there is an omni-God (A1)

Recall that (A1) and (A2) are assumptions, and this is the reason that the circularity objection is compelling. For (C*) and (C) allegedly justify the assumptions, respectively. The circle seems vicious because the conclusions seem to depend on justificatory appeal to each other. But for the externalist reliabilist, justification of the two conclusions (and thus the assumptions) is causal: it is not supposed to rely on a prior understanding of what it is to be justified in accepting the claims. This is the appeal of externalism, and a causal account of justification: it is not viciously circular. And of course, this was exactly Goldman’s (1979) general strategy in his attempt at a solution to the Gettier problem.

Right causal connections include (but aren't restricted to):

For Goldman, to ask what justifies us in thinking that perception (or memory, or inference) is reliable is to ask a question that must be addressed not by epistemology, but empirical psychology (Goldman [1979]; [1999]). If Descartes was also inclined to take an externalist, and indeed a naturalized approach to furthering our understanding of reliable belief-forming mechanisms, this would nicely explain his diligent interest in empirical psychology, or physiology of the brain. But more interestingly, it explains why he does not obviously fall prey to the elementary circularity objection.

Perhaps then, when Descartes insists that it is wrongheaded and even harmful to investigate the faculty of reason (CSM 1, p. 86), he means epistemological, not empirical investigation. This interpretation renders consistent the facts that he both appears to attempt some kind of validation of reason, and insists that we ought to leave well enough alone: it is only internalist approach to justification that he is frowning upon. His validation is causal, and perhaps even naturalistic.