Publication Date



Open access

Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PHD)


Philosophy (Arts and Sciences)

Date of Defense


First Committee Member

Elijah Chudnoff

Second Committee Member

Amie L. Thomasson

Third Committee Member

Colin McGinn

Fourth Committee Member

Robert Hopkins


My dissertation concerns sensory imagining: experiences like imagining a colorful parrot, or imagining its squawk. Just as with perception, there are three sorts of question to be asked about imagination. First, what is it like, phenomenologically, to imagine? Second, what can we do by imagining? Third, what is the essential nature of the experience? There are many theories about perception's nature, which explain its phenomenology and capacities. By contrast, there is some work on imagination's phenomenology, and a lot of work about its putative capacities, but very few theories of its nature. In the dissertation, I give a theory of imagination's nature, taking as explananda facts about its phenomenology and some of its capacities. The most striking facts about imagination's phenomenology concern its quasi-perceptual nature: ways in which it is like and unlike perception. Imagining is like perceiving, in that it is sensory, perspectival, and presentational: objects seem to appear to one. But it is unlike perceiving, in that the objects of imagination seem to be dependent on the subject, and in that investigation of the objects will not reveal anything more about their nature. These observations are sufficient to rule out three simple views of imagination. According to denialism, there is not really any such thing as sensory imagining. According to the perceptual model, imagining is very much like perceiving, differing only in vivacity. According to the pictorial model, imagining is like seeing an internal picture. None of these models adequately accounts for the phenomenology of imagining. For example, unlike when we perceive, our visual field need not be replete when we imagine. And unlike when we see a picture, we can't attend to both the object and the vehicle of a mental image. The challenge is to improve on these models. To do so, I first examine the sensory nature of imagination. I argue that the Humean intuition that you can first experience a color by imagining it amounts to the claim that you can secure de re reference to sensible properties by imagining them. This in turn means that dependency views, like M.G.F. Martin's, on which imagining constitutively depends on perceiving, are false. I also argue that this de re capacity of imagination cannot be explained by content views, on which to imagine is to stand in an attitude-like relation to a propositional content. I argue that a relational view of experience of sensible properties is the best way to explain this capacity. Since imagining involves a relation to sensible properties, it might also involve a relation to an extraordinary object which has sensible properties. I consider three such proposals regarding imagination: sense-data, Meinongian, and sensible profile views. I argue that none of these proposals is adequate. Each recapitulates problems with the simple perceptual and pictorial models. I argue that instead of a relational view, we should adopt a view on which sensible properties are structured by intentional content intrinsic to imaginative experiences. Starting from A.D. Smith's account of the nature of perceptual intentionality, I argue that the subjectivity of imagined objects can be explained by adopting Sartre's notion of positing, and the idea that sensible properties are predicated of posited objects. This intentional view faces two problems. The first is that it makes what might seem a dubious appeal to intentional objects. The second is that it seems to be a theory of how we imagine appearances, rather than objects. I argue that the first problem is not pressing; talk of intentional objects is not ontologically pernicious. The second problem is more serious. To address it, I argue that all the views of imagination thus far discussed are variants of the additive view, on which to imagine is to apply conceptual content to an ambiguous mental image. However, additive views are bound to fail as views of imagination. The root of their problems is the idea that imagining involves multiple intentional experiences which are somehow amalgamated. I propose instead an approach to imagination which draws on Gestalt psychology. Imaginative experiences consist of a range of parts, which are seamlessly synthesized into a whole experience. The whole experiences are intentional, but the parts are not; they inherit their intentionality from the whole. It is not the case that one generates an image and then applies intentional content to it; rather, the image, the intentionality, and the rest of the content come combined. I argue that this approach not only helps to explain how imaginative experiences can be about particulars, but also promises to be fruitful in explaining features of imagination, in part via the Gestalt idea of Prägnanz.


Imagination; sensory properties; intentional content; Sartre; Gestalt theory