A post by Dominic Gregory.

One idea that pops up every now and again, in philosophical thinking about perception, is that the imagination plays a part in ordinary vision. Here is a natural starting-point for that idea: we often experience the things that we see as having properties that we don’t literally see to be instantiated.

You can now see a computer screen, for instance. And you experience that computer screen as being a normal three-dimensional object—as having a back, say, and as being such that it will look different from different angles. But you can’t literally see the back of the screen, and you’re not seeing the thing from the various different angles. So how does your visual experience incorporate an awareness of the screen’s back, and of what it looks like from alternative places?

Enter the imagination, perhaps. Maybe our visual awareness of invisible aspects of things is owed to our ability to produce mental visual imagery. The idea that mental visual imagery—if not the ‘imagination’ as such—has a role to play here has recently been endorsed by Nanay. Dummett, Sellars, and Strawson have also advocated views of this type. And other philosophers like Husserl and Siegel have argued that appropriate expectations shape visual experience, where it’s fairly tempting to suppose that those expectations rely, at some level, upon our capacity to produce mental visual images.

There is an obvious worry with this suggestion, though. The kinds of visual episodes that we are considering are entirely ordinary. But we just aren’t conscious of constantly producing visual imagery that then shapes ordinary visual experiences. How can visual imagery be responsible for the relevant aspects of ordinary vision, then, if we aren’t aware of producing suitable accompanying visual mental imagery?

One response at this point would be to shrug, and to ask why anyone would think that mental visual imagery should have to be conscious. But that response needs fleshing out a bit. For what properties does an unconscious mental state need to have, to make it fair to regard it as involving mental visual imagery? Does it need to have a certain sort of neurological basis? Our ordinary conception of mental visual imagery hardly seems to insist upon a specific neurological foundation for mental visual imagery. So why would that be the crucial thing?

There is an alternative option, however. The view that something like mental visual imagery plays a significant role in shaping ordinary visual phenomenology has some attractions. But it’s not clear that those attractions derive from anything more than the potential role played within vision by mental states whose contents are akin to the contents of mental visual images. And there are no obvious worries with the idea that there are unconscious mental states whose contents correspond to the contents of mental visual images, even if mental visual images proper—‘seeings with the mind’s eye’—are essentially conscious.

So here’s a suggestion: instead of appealing directly to mental visual imagery as playing an important role in ordinary vision, we might instead appeal just to unconscious mental states whose contents are like the contents of mental visual images, in that they show things as looking certain ways.

Would anything be lost if we took that route? Are there distinctive functional roles associated with ‘mental visual imagery’, for instance? And are those functional roles such that an appeal to mental visual imagery in relation to visual states has a measure of explanatory power which is lacked by a mere appeal to appropriately positioned states whose contents are continuous with the contents of mental visual images?

Those aren’t rhetorical questions. But I can’t myself think of any functional associations involving mental visual imagery that indicate why appeals to mental visual imagery here should be favoured. Of course, the idea that we might cite states whose contents are of a piece with the contents of mental visual images raises the question just what, if anything, is special about the contents of mental visual images; but I think that question has an answer, one that’s developed in my (2013) book.

Now, it might be suggested that there isn’t any real gap between the suggestion just made and the view that ordinary vision involves mental visual imagery. For it could be claimed that all that is required for a mental state to involve ‘mental visual imagery’ is that it has the sort of content that—as we might put it—mental visual images possess.

That idea seems quite appealing. But I think that, as a reason for thinking that vision essentially involves mental visual imagery, its appeal is significantly reduced by the fact that there is a related, but more intuitively appealing, thesis in the vicinity: namely, the view that for a mental state to involve ‘mental visual imagery’ is for it to be a conscious state with an appropriate sort of content. If that last view were to be right, we could allow that vision involves mental states that are like mental visual images in certain crucial respects—in terms of the nature of their contents—but which—being unconscious—aren’t themselves mental visual images. And that seems like a fairly attractive overall position.