What is the difference between the phenomenology of perception and the phenomenology of imagination? Here are three prima facie possibilities:
- There’s no difference
- There’ a difference of degree
- There’s a difference of kind
The first view is that perceiving my dog Julius and imagining him (in the same setting) have exactly the same phenomenology; the only difference is in the surrounding beliefs: the perception is accompanied by a belief that the dog is really there, or that the dog is being perceived, or something like that. The second view is that perception and imagination have the same kind of phenomenology, but perception has a sharper, or more vivid, or finer-grained version of it. The third view is that the two have a different kind of phenomenology – there’s a qualitative difference between seeing Julius and visualizing Julius.
Hume had the view that the only difference between perception and imagination is a difference in phenomenal vivacity: a perceptual experience of a dog and an imaginative experience of a dog can have the same content, but the former is more vivid than the latter. Nobody really buys this today, because there are all sorts of crucial differences in functional role between the two: perceptual experience feeds into reasoning and action-guidance in a way a parallel imaginative experience does not. But the Humean legacy has persisted for the more specific question of the difference between the phenomenologies of perception and imagination. His is a “difference of degree” view, and I have a vague sense that that is the dominant view among analytic philosophers of mind.
In a 2010 paper called ‘Recollection, Perception, Imagination,’ Alex Byrne gives a convincing argument that this “difference of degree” view suffers from a fatal flaw:
… for any episode of visualizing or recalling [a strawberry], it should be in principle possible to create a physical picture of a strawberry such that viewing the picture in certain conditions exactly reproduces the felt quality of visualizing or recalling. And this is what seems wrong [in the “difference of degree” view]: any way of degrading the picture, such as blurring, desaturating, dimming, and so on, just yields another perceptual experience, plainly discernable from visualizing or recalling.
I’m convinced. So what should we go for: there “no difference view” or the “difference of kind” view? I suspect a lot of people don’t like the “no difference view,” but at the same time can’t see exactly what the difference in kind in supposed to be between perceiving Julius and imagining him.
I have a suggestion! It draws again on the idea that there are these subtle attitudinal phenomenal properties that lurk at the deepest structures of our experiences and that are harder to bring out through contrast cases than content-based aspects of phenomenology. When Byrne says that a degraded visual experience is “plainly discernible” from a visualizing, it sounds to me like he means introspectively discernible. You can tell introspectively whether you are seeing a dog or imagining one, even if the contents of the two experiences are indistinguishable. So what does introspection pick up on there? Here’s my suggestion:
(PE) My perceptual experience of Julius presents-as-real Julius
(IE) My imaginative experience of Julius presents-as-unreal Julius
The phenomenology of perception, on this view I’m floating, is not exhausted by the sensible features of what is experienced (the Julian colors and shapes), or even by the “higher-level” properties of what is experienced (being a dog, being Julius). It contains one more phenomenal feature, one which has nothing to do with what is experienced and has only to do with how it is experienced. That is the property of presenting-as-real whatever is presented. That subtle phenomenal property is really and truly a structural property of the experiencing itself.
On this view, when you introspect an imaginative experience of a dog, you can notice a tiny bit more than the phenomenology of being presented with a colorful, shapely canine thing. The extra tiny bit is: you can notice that all that is not just being presented but presented-as-unreal. That is the view of imaginative phenomenology I develop and defend in Ch.6 of my book. One thing I don’t do in the book, though, and am working on now, is to exploit all these subtle attitudinal phenomenal properties to call into question the so-called transparency of experience.
Here’s a very general question that has been niggling at me for some time in relation to this kind of research program. Somebody must have raised this question, but I haven’t seen it presented in the following straightforward (simplistic?) way. (Please tell me who has!)
One of the attractions of the idea that phenomenological properties are special and distinctive comes from the traditional view that some introspectible states have them, while others don’t. This would mean we could use introspection to detect the presence of phenomenological properties, and we could “see” that they were somehow different and special. By contrast, you seem to find a phenomenological difference (whether of degree or in kind) whenever there is an introspectible difference. Inference to the best explanation: there are no distinctive “phenomenological” properties; all that phenomenology amounts to is introspectibility. (Deflationary shades of HOT, also Dennett.)
I put this question to Terry Horgan once, and he acknowledged that he found a phenomenological difference wherever there was an introspectible difference.
I’m actually very sympathetic to the view that there’s no special line to be drawn, within introspectible states, between the phenomenological and the non-phenomenological. But the deflationary conclusion strikes me as the most natural one. Would you disagree? Maybe it doesn’t matter for your present purposes.
Hi Dan,
Yes, I’m with you and Terry on this: every introspectible property is a phenomenal property. The deflationary thesis you propose is a somewhat strengthed version of the converse: every phenomenal property is an introspectible property and (here comes the strengthening) all there is to a property being phenomenal is its being introspectible. It’s an intriguing thought that you can reason by inference to the best explanation from the point you, Terry, and I are in agreement on to this deflationary thesis. Personally, I would *want* to resist it, I suppose by offering a competing explanation that’s better. I know what my competing explanation is, though not immediately how I’d make the case for its superiority.
My competing story is this. What makes a mental state phenomenal, on my view, is that the subject is at least peripherally aware of it. When a state is such that the subject is peripherally aware of it, then it is nomologically possible for the subject to be *focally* aware of it. To be focally aware of a mental state is to introspect it. It follows that when a mental state is phenomenal, then it is nomologically possible for the subject to introspect it – that is, that the state is nomologically introspectible. Nonetheless, the fact that it is nomologically introspectible is not what *makes* it phenomenal. What makes it phenomenal is something deeper that underlies its nomic introspetibility, namely, its being the object of at least peripheral awareness. (This story is told in terms of states rather than properties, but if you construe a phenomenal state as just a phenomenal property-instantiation, then the story can be retold in terms of property instantiations. If you do not construe states as property-instantiations, then that will complicate considerably the telling of the story – but only the telling.)
Hi Uriah,
Interesting post, and I’ll have to think more about it (and read your book!) But for now, I was just hoping I could press you to say a little bit more about the notion of presents-as-unreal. Even before we get to the question of whether something along these lines can adequately capture the difference between imaginative phenomenology and perceptual phenomenology, I’m wondering whether the suggestion adequately captures imaginative phenomenology. While some imaginings might present-as-unreal — as when I imagine something obviously fantastical — I don’t think all imaginings need to present themselves this way. I might imagine my son right now at school as he takes the Spanish test he forgot to study for, and though it doesn’t present itself to me as if it’s happening before my eyes, it also doesn’t present itself to me as unreal. (Compare an imagining of my (non-existent) grandson.) Likewise for all the kinds of cases in which we use imagination for problem-solving purposes. There the fact that my imagination presents-as-real is crucially important to the success of my endeavor.
There is a sense in which imagining is untethered from the reality that’s present before us, but that’s different from the claim that imagining is untethered from reality. So I wonder: If we’re going to go in for this extra phenomenal bit, might it be better to cast it as “presenting-as-not-present” (awkward phrasing aside)?
Thanks!
Hi Amy!
Yeah, this kind of problem exercised me in the book. Here are some approaches to the problem within the framework of appealing to some attitudinal phenomenal properties to capture the distinctive character of imaginative experience.
One approach is to say that when you imagine your son taking his Spanish exam at school, you’re imaginative experience does present-as-unreal your son taking his Spanish exam, but it happens to be accompanied by a belief that your son is in fact taking his Spanish exam. (Compare having a visual experience of a broken teaspoon in a water glass while believing that the teaspoon is not broken.) This approach feels to me a bit heavyhanded, phenonenologically speaking.
A variant on this approach that’s a bit easier to swallow is to say that the experience doesn’t take a stand on the reality of your son, the exam, or the school, but does present-as-unreal their particular composition within the state of affairs imagined. That is, what imagination presents-as-unreal, strictly speaking, is the interrelation among the elements imagined. (Compare imagining rearranging the furniture in the living room, where the items imagined are real, and taken by the imagination to be real, but their imaginary interrelation is what is taken to be unreal.)
Maybe one way to make this approach seem more plausible is to say that when you imagine your son at his desk, the scene is presented in your experience in a visual modality, as though the scene is visible from the assumed spatial standpoint of the person doing the imagining. Maybe you “see” your son from the front of the class and slightly hovering halfway to the classroom ceiling. But in reality the scene does *not* have the spatial relation to you that the imaginative experience presents it as having. So the experience does present-as-unreal the overall scene, construed as one that’s presented as having certain spatial relations. (Don’t know if I made myself clear here.)
A completely different approach is to deny that the state you’re describing in your post is a state of imagining, claiming instead that it’s a state of imaging. Consider that there are two ways not to believe that p. One is to disbelieve that p and one is to entertain that p. Arguably, while belief that p represents-as-true p, the disbelief represents-as-false p while the entertaining merely-represents p (without taking a stand on its truth value). The thought here is that a similar distinction needs to be made between two different ways of having a non-perceptual but still sensory presentation of something: one that presents-as-unreal its object and one that merely-presents its object. A useful pair of words for these two distinct but related phenomena is ‘imagining’ and ‘imaging.’ The experience of one’s son taking the Spanish exam at the school is a case of imaging, not imagining.
An approach similar in spirit but different in letter is to draw the same distinction but allow both types of experience to count as imagining. The view here is that there are simply two very different kinds of imagination, one that presents-as-unreal and one that merely-presents (one that’s a sensory analog of disbelief and one that’s a sensory analog of entertaining, if you will).
Incidentally, this last view is that one I end up with in the book! In the blog post I just presented a first approximation of my view, by way of illustrating the general approach of looking for an attitudinal phenomenal property as the key to understanding a type of phenomenology. The book’s official view is this bifurcated account. One thing I dislike about it is the precisely disunified account of the imagination, which seems to me to involve an underlying commonality in the phenomena; but perhaps from your perspective a disunified account is okay (I think in the book I cite in passing your paper where you argue for this kind of disunity).
Then there’s the approach you suggest above: keep a unified account of imagination but recast the crucial attitudinal phenomenal property as presenting-as-absent/nonpresent rather than presenting-as-unreal. One worry I have with it is that it doesn’t seem to do justice to the phenomenology of things like a voluntary, endogeneous decision to imagine a two-headed flying kangaroo. That kind of experience feels like it’s doing more than saying that no such kangaroo is present in the space I occupy (or the space that’s visible from the location I occupy). It seems to present the object more “dramatically” than that – I want to say it presents-as-nonexistent that kind of object.
None of these approaches is super-satisfying to me. I’m essentially open to being convinced for the superiority of any of them. When writing the book the idea that imaginative experiences present-as-unreal the spatial relations between the presented scene and the person doing the imagining, and in virtue of that always presents-as-unreal the overall state of affairs imagined, hadn’t occurred to me. So I’m enjoying dwelling on this option now.