Rob Hopkins is a professor in NYU's philosophy department. Many years ago he published a book on pictures (Picture, Image & Experience CUP 1998). Its final chapter discussed what distinguishes the various forms of sensory imagining from one another. That left dangling the question how imagining itself relates to other mental states: especially perception, but also thought, belief and affect. It's taken a long time, but The Profile of Imagining (OUP 2024) at last offers some answers.
This week at The Junkyard, we’re hosting a symposium on Rob Hopkins’ recent book: The Profile of Imagining (Oxford University Press, 2024). Today we begin with an introduction from Rob. Commentaries and replies will follow Tuesday through Thursday.
* * *
What is sensory imagining? What is it to visualize rain falling on a city street, to summon the way the downpour suppresses the noise of the traffic, to imagine the distinctive odour coming off sun-warmed tarmac when wet? Most answers, whether from philosophers or psychologists, centre on the idea that sensory imagining is like sensing (perceiving), approximating both its phenomenology and its functional role. In The Profile of Imagining, I take a different line. The differences between imagining and perceiving, both phenomenological and functional, are far more striking than the similarities. Overall, sensory imagining is closer to thinking than perceiving. True, imagining something, unlike thinking about it, may capture what it would be like to perceive it. But capturing phenomenology need not involve reproducing it; and need not result in a state with a functional profile akin to that of the target. The book sets out an alternative explanation of this and other features that imagining displays.
The heart of the view lies in two ideas. One is that imagining is connected to agency in ways that perception is not. Action, as I understand it, is essentially a matter of exerting control. Since imagining is not always under our control, it is not always an action. It does, however, always originate in that part of the self that exercises control. It is, in my terms, spontaneous. (It is also experienced as such, at least when clear-sighted; that is, when we are aware of it as imagining.) To defend this view, I examine imagining under its two key aspects: as a power (the imagination) and as those occurrences in which that power is exercised (imaginings).
The other central idea is that imagining is manifestly representational. Visualizing a castle is not only a way of representing a castle; it is (when clear-sighted) given to us as such. Here, I argue, lies another contrast with perceiving. Perceptual states may, as many these days believe, be representations; but that is not how we experience them to be. In perception, the objects perceived, and the properties we perceive them as bearing, are given as constituents of one’s mental state of perceiving them. Direct realists may not be right about the metaphysics of perception, but their claims are, I argue, true to its phenomenology.
So imagining is representing things to oneself, representing that is both spontaneous and manifest. What form does this representational activity take?
In answer, I appeal to the idea of a sensory profile, an idea to be found in the work of Husserl and other Phenomenologists, and more recently in that of Alva Nöe. In perception, the appearance of things shifts with our relations to them. A street looks different from one angle as opposed to another, sounds different if heard through a dampening medium such as the pane of a window, and smells different if one has a cold. We perceive these perspectival appearances, but also the non-perspectival properties that underlie them. Those properties are, I argue, present to us as patterns in the shifting appearances; each having a complex sensory profile that, given a relation in which we stand to the property, determines what appearance it will present. We are presented with the appearances (the perspectival properties thus grasped are given as constituting our experiential state), and we form expectations about how those appearances will shift with our changing relations to the object.
Sensory imagining also exploits sensory profiles. It does so, however, by deploying those resources in a very different framework. Instead of being presented with appearances, we spontaneously and manifestly represent them to ourselves. (We do so in an explicit form I call conjuring.) And instead of forming expectations about how those appearances will shift with our relations to the objects in question, we undertake commitments on that score. We commit to conjuring the appropriate appearances, for any imagined change in our relations to the object imagined. That is what it is to sensorily imagine a non-perspectival property of some object, such as its shape, the sound it makes, or the smell it emits. In sum, while visually perceiving something as a cube involves being presented with specific perspectival appearances while expecting those appearances to alter in specific ways as the cube shifts in relation to us, so visualizing a cube involves conjuring those perspectival appearances while undertaking to conjure specific others as we imagine the cube shifting its relations to us.
Sensory imagining thus exploits structures found in perceiving. Its doing so is a key part of how it is able to capture perception’s phenomenology. But those structures are exploited in very different ways: in undertaking commitments, rather than expectations; and with respect to perspectival properties that are not presented, but merely the contents of spontaneous, and manifest, representings. This has various other consequences, most of which should be familiar. While perception is given as a guide to how things are, imagining is not. While perception’s objects are experienced as independent of our minds, imagining’s are not. (And this, even if what we are imagining is in fact, and is known by us to be, something so independent.) Perhaps most interestingly, while the things we perceive are experienced as having natures richer than any we at any moment grasp perceptually, the objects of imagining are not given as richer than we currently represent them as being. Everything (including every aspect of every thing) we imagine, we imagine consciously. I call this claim No Unconscious Imagining.
That claim needs refining. As it stands, it may seem to conflict with the fact that we may use sensory imagining to learn. Suppose I am confronted with one of Shepard’s famous pair of figures (see figure C). I am shown the image on the left, which is then removed. Then I am shown the image on the right and asked whether the object in it is congruent with the object in the left hand picture. Can’t I answer by visualizing the original object, and rotating it? Rotation thus appears to reveal aspects of the object I visualized of which, when I initially visualized it, I was not conscious. Doesn’t this show that No Unconscious Imagining is false?
The matter is subtle. As I interpret the case, it does not threaten the idea that if I imagine a certain content, I am necessarily conscious of that content. What it does show is that the contents I imagine may have consequences, and I may not be conscious of all of those, simply in virtue of imagining the content itself. The details here of course need filling in, and I devote a chapter (ch.7) to doing just that. But if, as I believe, this approach is right, we stand to our imaginings much as we do to our thoughts. To think a thought content is to be aware that that is what one thinks; the consequences of that thought are quite another matter. Working out what those consequences are may require careful application, just as does working out whether one imagined object is congruent with another object. In perception, in contrast, what we perceive always, and plainly, involves features that outstrip our grasp completely. What lies on the hidden side of a seen object, or beyond our present tactual or auditory reach, is something on which we have no perceptual purchase whatsoever. We cannot work out how the hidden parts are by manipulating the original content. The only way forward is to bring more of the thing into our perceptual reach at all—for instance, by walking round the thing to see what lies behind.
If one contrast between imagining and perceiving lies in our epistemic relations to their objects, others lie in how the two relate to affect. Perceiving something is the canonical way to engage with it aesthetically. Is imagining it at least a decent substitute in that respect? Contrary to what may seem obvious, and to what some influential theory suggests, I argue that it is not. Aesthetic response, rooted as it is in pleasure, has a natural course of development that in the case of imagined objects is blocked. Nor is aesthetic response the only affective state imagining is unable to elicit. Various affects require their objects also to be experienced as their causes. In clear-sighted imagining, that alignment fails: picturing something disgusting, such as a slug, offers one an object for the putative affect (disgust), but not one able to cause the feeling, or even to seem to do so. The visualized slug plainly lacks the reality required for it to be the cause of any response on one’s part. These affects thus stand in functional relations to imagining radically different from those in which they stand to perceiving. If matters have seemed otherwise, that is due to various considerations, such as our ability to learn about the affective character of things by imagining them, and the fact that affective imagining oftens involves the same bodily changes as are involved in real affective reactions. If imagining doesn’t provoke real affect, how can this be? I have answers, but to find out what they are, the interested reader will, I’m afraid, have to turn to the book.