The problem of skepticism is the fundamental epistemological problem Descartes addresses. He introduces three forms of it, each embedded in a possible error-scenario. The first possibility is that, since my sense perception is sometimes misperception, my sensory experience in any given case may not reflect how things are outside my experience. The second possibility is that maybe I am dreaming when I think I am awake. And the third possibility is that maybe I am deceived in all my ideas and beliefs by a powerful demon. The third is the most radical, far-reaching, and potent of the error-scenarios Descartes discusses. Unlike the first two, it threatens knowledge of all kinds. So, if Descartes is to defeat skepticism in a fundamental and comprehensive way, he must eliminate that possible scenario.

There is a wide consensus that, having raised the possibility of massive deception, Descartes is not able to eliminate it. The ironic effect of this is that radical skepticism is in better shape after Descartes than before. The heart of this consensus is that Descartes' principal anti-skeptical argument fails. That is his argument to enlist God to guarantee that his best ideas, the ones which he classifies as being both clear and distinct, are error-proof.

I agree with this consensus. However, for the sake of argument, let's suppose the consensus is wrong. Let us run a thought-experiment in which Descartes successfully proves that God exists and that God is no deceiver. Would he thereby eliminate the possibility of massive deception? No. That's what I want to argue here. If this is right, Descartes' situation is worse than the consensus indicates. For the consensus might be interpreted as leaving open the idea that Descartes' anti-skeptical plan itself – the plan to prove that God exists and that God's non-deceptiveness guarantees clear and distinct ideas to be error-proof – is okay, but that its execution fails. But if I am right, the plan itself is inadequate.

In the course of combatting skepticism, Descartes seems to discover that, at least while he is thinking about being deceived, he could not possibly be deceived about his own existence. For to be wrong about that would mean he couldn't be thinking or doubting in the first place. This is the famous Cogito, ergo sum, I think, therefore I am, idea.

There is room for debate over whether this idea really is beyond any possible doubt. As we will see below, even Descartes himself seems to raise a question about it. But let us not dig into that here. Instead, let us grant for the sake of argument that the Cogito idea is immune to error and doubt.

With this one truth established, Descartes thinks he may have a way to overcome the third of the error-scenarios described, thus a way to win a general victory over radical skepticism. The key to this is his discovering the property in the idea, ‘I think, therefore I am,’ which is immunizing it against the possibility of error. For a parallel, think of a medical researcher analyzing an infection-immune cell to discover the cause of its immunity. Analyzing the Cogito idea with this goal in mind, then, Descartes claims that what makes it error-proof is its clarity and distinctness. If this turns out to be right, he will know what to look for when examining his other ideas and beliefs for possible error. But is this analysis of the Cogito idea itself error-proof? After all, questioning the analysis does not seem to be self-defeating in the way that questioning one's own existence seems to be. Acknowledging this, Descartes seeks a guarantee of the proposed immunizing property. His idea is that God, if God exists, would not permit him to be wrong on this point. So he proposes to prove that God exists. For Descartes, the stakes in all of this are high. As he puts it in the third of his Meditations on First Philosophy, ‘…in order to remove even…[a]…slight reason for doubt…I must examine whether there is a God, and, if there is, whether he can be a deceiver. For if I do not know this, it seems that I can never be quite certain about anything else.’

Let's side-track for a moment to note that there is a dispute about exactly how much is at stake here. In the lines just quoted, Descartes' point seems clear enough: it seems to be that if he does not know that God exists and is no deceiver, then he cannot be certain that any of his ideas are error-proof, even the Cogito idea itself. But there are arguments on the other side too. For present purposes, there is no need for us to examine the issue. For, either way, whether all of Descartes' ideas are at stake in his attempt to prove that God exists, or whether all of them minus a few are at stake, there is a very great deal at stake. End of side track.

In Part IV of The Discourse on Method and in the already-mentioned Third Meditation, Descartes develops the first of two arguments to prove that God exists. It is his version of a cosmological argument. Cosmological arguments are arguments in which an attempt is made to prove the existence of God based on some claim about the world, for instance, that every event has a cause. In his Fifth Meditation, Descartes offers a second argument for the existence of God. This is his version of an ontological argument. Ontological arguments for the existence of God are based on an analysis of the meaning of the concept, ‘God.’ Ignoring the many criticisms of these arguments, let's accept for the sake of argument that Descartes successfully proves that God exists and that God is no deceiver.

To make my case that massive deception would still remain an uneliminated possibility, I'm going to bring in an argument from a different philosophical debate over the existence of God. The debate in question is over the logical consistency or inconsistency of the idea that God exists, on the one hand, and the fact that evil exists, on the other. The argument I'm bringing in is Alvin Plantinga's successful proof that there is no logical inconsistency between the idea that God exists and the fact that evil exists. This proof is part of his free-will defense of theism against arguments designed to show that various facts of evil either conclusively disprove the existence of God or prove its improbability.

Another side-track, to note a second consensus. This time it is the consensus that Plantinga does indeed prove that there is no contradiction in the co-existence of God and evil. I agree with this consensus too. However, so as not to over-estimate Plantinga's success, let us note that it does not mean either that God exists or that evil is not evidence that the existence of God is improbable. Those are separate issues. End of side-track.

The core of Plantinga's proof that there is no contradiction between the propositions ‘God exists’ and ‘Evil exists’ is his description of a possible scenario in which both God and evil exist. In essence, this scenario is made up of the following points, each of which is possibly true, which is to say that none of them has to be false: first possibility, that God intended to bring about a world containing beings who could develop into moral agents, thus a world in which moral goodness could occur; second possibility, that such a world is overall better than any world in which moral goodness could not occur; third possibility, that a world in which the occurrence of moral goodness is possible must contain beings with free will, in the libertarian sense of free will – essentially, libertarian free will means that I possess free will only if, at any time I make a choice, I have before me genuinely open alternatives, such that it is entirely up to me which alternative I choose and that my choice is not caused by anything external to itself; fourth possibility, that, understood in this way, the free choices of moral beings are beyond the power even of omnipotence to cause; fifth possibility, that its being beyond the power of omnipotence to cause the libertarian free choices of moral beings does not cancel divine omnipotence, for this limit on omnipotence is purely logical, and omnipotence does not mean power to do logically impossible things; sixth possibility, that, while God recognizes the possible occurrence of moral evil as a consequence of libertarian free will, God does not intend its occurrence; seventh possibility, that those evils which do not result from bad choices by human beings – natural disasters, various diseases, and so on – also fall into the category of moral evil, since they are the result of the God-granted, libertarian free will of a powerful non-human agent, Satan, say. Plantinga's conclusion is that, as there is no logical inconsistency in this possible scenario, the co-existence of God and evil implies no contradiction.

In thinking about this argument, remember that Plantinga needs only to show that this scenario is a bare logical possibility. His argument does not require that any of the seven points is actually true or even likely. All it needs is that none of them has to be false in the way that, for example, the proposition ‘A bachelor is a married man’ has to be false. So his argument would be unaffected even if some or all of the seven propositions are in fact false, provided that none of them is necessarily false. For instance, the proposition, ‘My car is a Toyota,’ is false, but it is not necessarily false. For it, unlike the ‘bachelor’ proposition, could have been true.

Now let us go back to Descartes. Let us agree that massively deceiving him is a bad thing and that, if he is massively deceived, it is not by a human being. Re-written to reflect those points, the seventh possibility in Plantinga's scenario reads as follows; seventh* possibility, that those evils which do not result from bad choices by human beings – natural disasters, various diseases, massively deceiving Descartes, and so on – also fall into the category of moral evil, since they are the result of the God-granted, libertarian free will of a powerful non-human agent, Satan, say. Adjusted in this way to cover massive deception, Plantinga's scenario is still not self-contradictory. Thus it is logically possible that Descartes could be deceived in any or all of his clear and distinct ideas, even if God exists and is not a deceiver.

Two counterpoints. First, isn't this Satan hypothesis awfully far-fetched? And how could such a remote possibility have such a serious impact on Descartes' argument? I agree about the remoteness, when the context is everyday life. But, in context of Descartes' own deceiving-demon scenario, the Satan hypothesis is clearly relevant. It's cut from the same cloth, after all.

Second, isn't Descartes' own concept of free will, sketched in the Fourth Meditation, a form of compatibilism, not libertarianism? As its name suggests, compatibilism maintains that a choice can be caused by something outside itself and still be a free choice. An example is when the cause is my own character or temperament, even though I do not choose my own temperament. So, this counterpoint goes, if libertarianism is a necessary condition of genuine moral choice in Plantinga's scenario, then perhaps Descartes' being a compatibilist of sorts will enable him to avoid that scenario.

But libertarian free will is not a necessary condition of genuine moral choice in Plantinga's scenario. The possibility of libertarian free will is. And Descartes' having a different theory of free will does not negate that possibility, or even have any bearing on it. Even if libertarianism were known to be false, the possibility that genuine moral choice could occur only on the libertarian conception of free will would stand. Think of the Toyota example again. My car is actually a Subaru, but it is possible that I could have had a Toyota instead.

Finally, in Meditation Three, Descartes acknowledges that ‘there are countless…attributes of God which I cannot in any way grasp, and perhaps cannot even reach in my thought…’ In effect, our thought-experiment takes him up on this acknowledgement. For it sketches the possibility that, if God exists, then, unknown to Descartes, there may be various tradeoffs which God accepts, insofar as good, evil, and free will are concerned, and that the massive deception of Rene Descartes is among them. In sum, then, if my argument is right, Descartes has a deception problem which even proving that God exists and is not a deceiver will not solve.