Traditional Moorean absurdities for belief are sentences like “p, but I don’t believe that p,” or “p, but I believe that it’s not the case that p.” They have certain fundamental features: they can be true, but they are absurd to assert— but only in such first-personal, present-tense forms. Are there any analogous Moorean absurdities for intention? I examine five different kinds of candidate Moorean absurdities for intention. I reject three kinds of candidates that do not share the fundamental features of Moorean absurdities for belief. The remaining two kinds of candidate are promising: they provide sentences involving intention attributions that can be true, but are absurd to assert—but only in first-personal, present-tense forms. Recognizing that these really are Moorean absurdities for intention does not require committing to any particular explanation of the absurdity involved in asserting them—or in asserting the classical Moorean absurdities for belief.

Draft

In ‘underdetermination cases’ of imagination, the content supplied by your mental image underdetermines the total content of your imaginative episode. What settles the rest of the content? I argue that your generation of a mental image as a means to imagining an F makes it the case that the mental image has greater cognitive significance for you than is given by its own intrinsic properties.

Draft

Why do your self-attributions of beliefs and intentions ordinarily constitute authoritative self-knowledge? You can self-attribute a belief or an intention transparently. For instance, you can transparently self-attribute a belief that p by judging that p. You can transparently self-attribute an intention to Φ by deciding to Φ. However, recognizing just this much does not completely explain the epistemology of transparent self-attributions.

Self-attributions of this kind count as authoritative knowledge because they involve a form of practical knowledge. You can intentionally control the kind of attitude you take up in conscious thought, and when you do that, you know what kind of attitude you are taking up in conscious thought. Then, in the context of transparent self-attribution of belief or intention, a judgment that p or a decision to Φ can have a complex identity. A judgment that p can also be a self-attribution of a belief that p, and a decision to Φ can also be a self-attribution of an intention to Φ. To explain how this can be the case I introduce the linked notions of embedded mental action and content plurality.

The view of self-knowledge that emerges also explains why there are contents involving belief attributions that are absurd to assert or to judge even though they can be true. These contents are Moorean absurdities for belief. I argue that there are no corresponding Moorean absurdities for intention, even though you also have transparent self-knowledge of what you intend to do. This points to an important attitudinal distinction between belief and intention: intentions are not beliefs.

The difference between first-personal and third-personal methods of attributing attitudes is subtle. The specialness of the first-personal perspective cannot be explained in terms of epistemic groundlessness, as many have tried to do. You must also make third- personal groundless attributions of belief to understand others’ intentional behavior.

Despite philosophical skepticism on this point, transparent self-knowledge really is valuable, in a special sense. Having complete diachronic transparent self-knowledge involves having no hidden attitudes and having a diachronically unified self of the kind that is required for evaluation in terms of authenticity.

The epistemology of self-knowledge relies crucially on the fact that you can do things in thought. Knowing yourself is something you do because intentional action is indispensable to authoritative, knowledgeable self-attribution of beliefs and intentions.

Final version

Starting in 1905, the painter André Derain began to use bright, clashing hues contiguously to reflect contrasts in luminance, as at the edges of shadows, or between skyline and sky. Derain’s innovative coloration is unusual in at least three ways. First: hue and luminance do not share first-order similarity structure. Second: hue contrast is not the only dimension that reflects luminance contrast in these works, since brightness contrast plays the same role. Third: in these works, the first-order hue of each color sometimes does, and sometimes does not, reflect the hues of the object(s) depicted. Nonetheless, it is tempting to say that hue contrasts depict luminance contrasts in these contexts. If so, this would challenge several established theories of depiction. These include objective resemblance views, experienced resemblance views, isomorphism views, and views that model pictures on maps. I suggest some general lessons.

Handout

Slides