Gewirth answers the problem of the Cartesian Circle via the notion of psychological certainty. Cleaned up a bit, the best version of his view runs as follows. We first arrive at a clear and distinct perception that God exists and is no deceiver by inferring it from premises that are themselves clearly and distinctly perceived. We then define the epistemic notion of certainty in terms of there being no proposition that is a reason for doubting. Next, revising Gewirth a bit, we understand a reason for doubt so that X is a reason for doubt only if it is false that its negation is clearly and distinctly perceived. Then we note that the only reason for doubting clear and distinct perceptions is that God is a deceiver. From these premises it follows that all clear and distinct perceptions are certain in the epistemic sense.
Jim objects to this argument in the Cartesian context by noting that this epistemic sense of certainty is not Descartes’ notion of metaphysical certainty.
Descartes’ notion requires that metaphysical certainty be factive and the above notion isn’t. But notice that for those of us who don’t think epistemic certainty needs to be factive, this difficulty disappears. Given such a noncartesian position, what should we say about the Gewirth argument?
One thing to note is that the last premise is false, the premise that claims that the only reason for doubting C&D perceptions is that God is a deceiver. There are other reasons that entail that God is a deceiver, if we think about the skeptical hypotheses in question (and interpret the theistic hypothesis conditionally as the claim that if there is a God of the sort Descartes considers, he is a deceiver). We thus need a closure principle of a rather troubling sort, or we must hold that only those who go through the Cartesian therapy are capable of epistemic certainty.
Van Cleve objects to the argument in another way. He objects that the standard for grounds for doubt is too low. Instead, he claims, you have to be certain that, e.g., the Demon Hypothesis is false in order for it not to be a reason to doubt.
Two points here. I think Jim is right that the Gewirth standard for grounds for doubt is too high. But Jim’s seems to me too low. What is important is that a ground for doubt is an epistemic possibility for you. To get to Jim’s account, you have to understand the notion of epistemic possibility in terms of not being certain of its denial. Jim defines epistemic possibility in this way, but as Huemer’s earlier post on epistemic possibility reveals, such an account is mistaken.