Daniel Gregory (University of Barcelona), "Imagining a Way Out of Dream Skepticism"

Forthcoming in Erkenntnis

By Daniel Gregory

How do you know, at any particular point in time, that you are awake and not dreaming? The problem of dream skepticism is familiar to every philosopher and also surely one of those relatively rare philosophical problems which occurs to people far removed from academic philosophy.

Descartes’s treatment in the Meditations (1641/2006) is the most famous. He satisfied himself—eventually—that two telltale differences between waking experiences and dreaming experiences could be identified. One is that the things which happen in dreams are often bizarre in comparison with the mundane events of most of our waking lives; the other is that dream events are ‘never joined by memory with all the other actions of life’ (p. 50). But even if these things are true—and it is actually not clear that they are—there is a bigger problem for Descartes. It was articulated by Hobbes (1641/2006) in his commentary on the Meditations, but it is obvious enough. Dreams don’t seem bizarre or disconnected from our waking lives while we are dreaming, so you cannot rely on these characteristics as a basis for a judgment that you are not dreaming at any particular time. It will typically not seem to you that your current experiences are either bizarre or disconnected from your waking life, whether you are awake or dreaming.

Ernest Sosa proposed a new solution to the problem of dream skepticism in Volume I of A Virtue Epistemology (2007), the published version of the John Locke Lectures which he delivered at the University of Oxford in 2005. Sosa thinks that we do not actually form beliefs in dreams; we merely imagine that certain propositions are true. This means that, if you judge that you are awake and not dreaming, you cannot be wrong. If you are awake and not dreaming, then you will form a true belief. If you are asleep and dreaming, then you will not form a false belief, because you will not actually form any belief at all. You might imagine that it is true that you are awake, but imagining that something is true clearly does not amount to believing that it is true. And there is nothing at all wrong with imagining that something false is true. This means that you are always justified in judging that you are awake and not dreaming. You might form a true belief, and you are never at risk of forming a false one.

Sosa’s reasoning is ingenious, but it is also more than a touch frustrating, as Jonathan Ichikawa (2008) has argued. It is true that, if you cannot form beliefs in dreams, then you cannot form the false belief that you are awake while you are dreaming. And it is true that, if you form the belief that you are awake while you are awake, then you will form a true belief. But you can now wonder: Have I just judged that I am awake and thereby formed a true belief? Or did I merely imagine judging that it is true that I am awake, because I am in fact dreaming, and thus cannot actually judge and form beliefs at all? It will not feel like progress.

In this paper, I try to offer a new solution to the problem of dream skepticism, albeit one which will only be available to those who hold a particular theory about the nature of dreams. Traditionally, dreams have been thought of as hallucinations. They involve having false perceptual experiences on the basis of which we actually form beliefs. Sosa denies that we actually form beliefs when we dream, but he does not comment on the nature of the sensory experiences which we have when we dream. Are they false perceptual experiences, i.e., experiences of the kind which we have in ordinary instances of seeing and hearing etc., except that they do not actually represent the environment one is in? Or might they be something else? Several philosophers over the last thirty years, and a handful prior to that, have suggested that the sensory experiences in dreams are something else, namely, experiences of mental images.

It is very difficult to say how mental images differ from false percepts, and I do not take this up in the paper. But there certainly does seem to be some difference. On the face of it, imagining is different from hallucinating. Everyone has a pre-theoretic grasp of this.

Why might we think dreaming involves imagining rather than hallucinating? There are several arguments for the view. Here’s one, from McGinn (2004). Dreams usually doesn’t affect us in some of the ways that perceptual experiences do. If you hear your phone ringing while you’re asleep, you’ll probably wake up. If you (seem to) hear your phone ringing in a dream, you probably won’t wake up. But if dreams were hallucinations—perceptual experiences, albeit not veridical ones—then the (apparent) sound of the phone ringing should cause you to wake. For an individual having a hallucination, after all, things seem just as they would if the experience were veridical. By contrast, it makes sense that the (apparent) sound of a phone ringing should not cause you to wake if dreaming is imagining. You can imagine the sound of a gunshot and it won’t cause you to flinch, because imagining just doesn’t cause the same kind of reactions that perceiving (veridically or otherwise) does. (The phone example is from Ichikawa (2009).)

Now, here is a surprising limitation on our capacity for imagination: One cannot simultaneously have two entirely separate imagistic experiences in the same modality (McGinn 2004; Hopkins 2018). Suppose you are sitting in your office and you see in front of you your laptop and desk and the wall and window behind them. In this situation, you can visually imagine an entirely separate pastoral scene (or whatever else). But now suppose you are visually imagining what you see when working in your office: laptop and desk in front of you and the wall and window behind them. You cannot simultaneously have a completely separate visual imagistic experience as of the pastoral scene. You can imagine one scene superimposed on the other. Or you can imagine one scene alongside the other, in something like the way that you can see two separate photographs of different scenes alongside one another. But you cannot simultaneously have two visual imagistic experiences which are separate from one another in the way that you can simultaneously have a visual imagistic experience and a visual perceptual experience which are separate from one another.

If dreaming is imagining, and if it is not possible to simultaneously have separate imagistic experiences in the same modality, then it follows that you will not be able to form a mental image of something which is entirely separate from any dream experience which you are having in the same modality. If you are (rather unfortunately) dreaming of your office, then it will not be possible to form an entirely separate visual image of the pastoral scene. This would require simultaneously having two separate imagistic experiences in the same modality, which is not possible. This means that, if you can form a mental image in any modality which is separate from the sensory experience which you are currently having in the same modality, then you are not dreaming. It is only possible to have two sensory experiences simultaneously if one is a perceptual experience and one is an imagistic experience, and having a perceptual experience requires being awake (at least supposing that dreaming does not involve hallucinating).

In a nutshell, then: If you can imagine something, you’re not dreaming!

Descartes, R. (1641/2006), Meditations, Objections, and Replies, R. Ariew & D. Cress (trans. and eds.), Hackett Publishing Company.

Hobbes, T. (1641/2006), ‘Third set of objections’, in R. Descartes, Meditations, Objections, and Replies, R. Ariew & D. Cress (trans. and eds.), Hackett Publishing Company.

Hopkins, R. (2018), ‘Imagining the past: On the nature of episodic memory’, in. F. Macpherson & F. Dorsch (eds.), Perceptual Imagination and Perceptual Memory, Oxford University Press.

Ichikawa, J. (2008), ‘Skepticism and the imagination model of dreaming’, Philosophical Quarterly, 58: 519-527.

Ichikawa, J. (2009), ‘Dreaming and imagination’, Mind & Language, 24: 103-121.

McGinn, C. (2004), Mindsight: Image, Dream, Meaning, Harvard University Press.

Sosa, E. (2007), A Virtue Epistemology: Apt Belief and Reflective Knowledge, Volume I, Oxford University Press.