Locating the dream self in the dream world

We all dream every night, and most of us feel reasonably certain that we know what it is like to dream. But how well do we really know the phenomenology of dreaming? Can we really be certain that in describing our dreams, we are not merely projecting implicit, pretheoretical assumptions or existing theoretical commitments onto the phenomenology of dreaming? And even assuming that we succeed in accurately characterizing our own dreams, can we generalize from our own case to what it is typically like to dream?

A brief consideration of the historical and contemporary philosophical literature on dreaming suggests that intuitive and folk-psychological descriptions of dreaming lead to strikingly different views. The traditional view has long been that dreaming replicates the phenomenology of standard waking perception in all of its detail. This view is implicit in the Cartesian scenario of dream skepticism. In the Meditations, Descartes tells us that even his experience of sitting by the fire, looking at a piece of paper, purposefully extending his hand and moving his head, and even of wondering whether he is now dreaming can (and frequently does) occur in his dreams. As Revonsuo (2006) puts it, “there is nothing in the experience itself, in the actual qualitative character of the experience, that necessarily distinguishes the dream experience from a corresponding perceptual experience in the waking state” (p. 82); “the qualities of dream experience are identical with the qualities of waking experience” (p. 84).

If you feel the pull of this description, you might be surprised that this view, to many, has seemed profoundly counterintuitive. There is a long history of describing dreaming as a form of imagining or even of thinking during sleep. Occasionally, this is combined with the claim that dreaming also feels like waking imagination or daydreaming (McGinn 2004, Ichikawa 2009). Dreams, after all, are in some sense under our control; and because we implicitly author our dreams, we also do not, or so the rival imagination view would have it, truly mistake the persons, objects and events in them for the real thing. In this view, dream imagery lacks the bite of actual percepts—we all know, after all, that pinching yourself in a dream is not really painful (or do we?).

From this disagreement, you might conclude that dream phenomenology is not something that should be conducted in the proverbial armchair; rather, claims about the phenomenology of dreaming should be based on the analysis of large sets of dream reports gathered under controlled conditions, for instance following timed awakenings in the sleep laboratory, and using carefully worded questionnaires (see for instance Domhoff 2003; Kramer 2013). There are also a number of methodological concerns about what are the most accurate methods to score dream reports (Sikka et al. 2014).

However, even what scientific research tells us about the phenomenology of dreaming is subject to change. Eric Schwitzgebel (2002, 2011) has painstakingly documented that there has been a historical swing in opinions from claiming that we predominately dream in color, to claims that we mostly dream in black-and-white (popular in the 1930s-1960s), followed by a return to the claim that we mostly dream in color, which continues to be popular until today. Schwitzgebel suggests that this swing in historical opinions should make us skeptical. We may not know the phenomenology of dreaming nearly as well as we think we do.

Part of this variation, of course, may be an artifact of changing methodological approaches to the study of dreaming. Yet, even with improved methodologies, the difficulties in finding a quick-and-easy way to characterize the phenomenology of dreaming would likely persist. One of the lessons to learn from the parallel histories of thinking about the phenomenology of dreaming and the shifts in historical opinions is that dreams themselves are likely a mixed bag. For instance, one study found that when participants were asked whether they dreamt in color or in grayscale, between 10 and 20% said they had mixed dreams, which were part grayscale, part colored (Murzyn 2008). Dreams also differ in subtle ways from waking experience. When participants were asked to compare photographs with different degrees of color saturation, brightness, and visual clarity to the visual imagery in their dreams, ratings for clarity, for instance, were different for dreams from tonic and phasic REM phases. The authors concluded that “the single most frequently reported category of dream image was one that visually resembles external reality”, adding that “the most frequent departures from a realistic appearance were a loss of color saturation  and  a  loss  of  background  detail” (Rechtschaffen & Buchignani,  1992,  p.150).

These difficulties might alert us to the fact that there is something wrong with the original question of whether dreaming is, phenomenologically speaking, more similar to imagining or perceiving. Perhaps, trying to describe the phenomenology of dreaming by modeling it on that of wake states (whose phenomenal characteristics are themselves, we might add, variable and often poorly understood) is misguided. Perhaps we shouldn’t expect the phenomenology of dreaming to neatly conform to one or the other kind of stereotyped description. Here, my point is only that as soon as one makes any general, overarching claims about the phenomenology of dreaming, one easily slips into an oversimplified view that glosses over real and interesting phenomenological differences between dreams and the wake states they are contrasted with (for a detailed discussion of this problem, see Windt & Noreika 2011).

One also quickly runs into counterexamples. For instance, it has long been known that a majority of dreams involve visual imagery and movement sensations (Hobson et al. 2000; Schwartz 2000); but imageless dreams exist, and visual dream imagery can even be lost selectively following occipito-temporal lesions (called visual anoneira; Solms 1997). Similarly, not all dreams contain movement, and static dreams are sometimes reported from stages 2 and 3 of NREM sleep (Noreika et al. 2009).

So why not simply accept that dreaming is an intrinsically heterogeneous phenomenon, subject to a large degree of inter- and intrasubjective variation? Maybe the best account of dreaming will be messy: maybe there is only a loose family resemblance between different types of experiences that pass under the name of dreaming. This type of cluster account is possible, but at least at the outset, it is not particularly attractive. For the project of remapping the concept of dreaming and determining its boundaries, as well as its relationship to concepts used to describe standard and altered wake states—such as hallucination, imagination, perception, illusion, and belief—the relevant question is whether there is something like a phenomenal core of dreaming, a single and highly invariant phenomenal property that underlies different types of dreams (and thus is compatible with the variability of dreaming), but also helps distinguish dreams from their closest relatives, for instance from hypnagogic hallucinations occurring during sleep onset.

In Dreaming, I suggest that phenomenal selfhood—the experience of being or having a self—is a plausible candidate for describing the phenomenal core of dreaming. Numerous studies have confirmed that the vast majority of dream reports describe the presence of a dream self (e.g. Strauch & Meier 1996). There are interesting differences in self-representation in reports from different sleep stages (Occhionero et al. 2005; McNamara et al. 2007), and there also seems to be a gradual transition during childhood and adolescence from dreams in which the self is a passive observer to dreams in which it is an active participant (Foulkes 1999). Importantly, even in minimal forms of dreaming, in which visual imagery, for instance, has been lost, there is often still a feeling of being present in the dream, or of having a self (Dreaming, chapters 7 & 11).

We have to be careful. In characterizing the dream self, we shouldn’t just assume, from the outset, that the dream self is a phenomenal duplicate of our waking self, a doppelgänger characterized by the same rich phenomenology of being a thinking self and bodily agent that we enjoy in wakefulness (though, come to think of it, it is also less than clear how well we really know, intuitively and without relying on empirical data, the phenomenology of waking, embodied self-experience; see for instance Schwitzgebel 2007). To be sure, the dream self can differ in interesting ways from the waking self. We don’t always act quite like ourselves in dreams (Augustine famously worried about whether he was morally responsible for sins committed in his dreams), and sometimes, the dream character we identify with may be completely different from our waking self. Your dream self might be a younger (or older) version of your waking self; it might have a different job, a different partner, or even a different gender. You might even dream that you are an animal. Such vicarious dreams raise interesting questions about the identity of the dream self (Rosen & Sutton 2013).

At the same time, the distinction between self and non-self dream characters is preserved in most dreams (though self-other distinctions may be more fluid, for instance, in lucid dreams; see Windt et al. 2014). Dreams are intensely social experiences, with the average dream containing 4 non-self dream characters (Kahn et al. 2002). The Social Simulation Theory of dreaming suggests that the abundance of social interaction in dreams may translate into a theory about the evolutionary functions of dreaming (Revonsuo et al. 2015). Self-identification may be more dynamic than in wakefulness: in the course of a single dream, you might successively identify with a number of different dream characters. Still, only one of these, at a given time, is experienced as the self. Self-identification is a highly stable characteristic of conscious experience—so stable, in fact, that pointing this out may seem trivial. Yet, it is fascinating to note that even under highly adverse circumstances—even in our dreams, when we are lying safely in bed, lost in fantastic and often bizarre virtual worlds, behaving nothing like we do in wakefulness and interacting with a host of characters that  are themselves, ultimately, the products of our own making—this basic structural feature of experience is preserved.

For now, I want to propose that the key towards a unified and distinctive theory of dreaming might be to analyze the experienced relationship between the dream self and dream world, the feeling of presence or immersion in dreams. In the next post, I will propose that we can further unpack this notion by concentrating on examples of minimal phenomenal selfhood in dreams. The simplest, most reduced kind of experience that can still be described as involving a self is the experience of spatiotemporal self-location within the dream world. As long as there is still a phenomenal here and a phenomenal now, we still have the feeling of presence; we still locate ourselves, as it were, at the center of the dream world.

In later posts, I will explore how understanding the phenomenal property of spatiotemporal self-location in dreams will also help (re-)locate the concept of dreaming on the map of concepts used to describe standard and altered wake states, such as mind wandering and presence in virtual reality.


Domhoff, G. W. (2003). The scientific study of dreams: Neural networks, cognitive development, and content analysis. Washington, DC, US: American Psychological Association.

Foulkes, D. (1999). Children’s dreaming and the development of consciousness. Cambridge, MA: Harvard University Press.

Hobson, J. A., Pace-Schott, E. F., & Stickgold, R. (2000). Dreaming and the brain: toward a cognitive neuroscience of conscious states. The Behavioral and Brain Sciences, 23(6), 793–842; discussion 904–1121.

Ichikawa, J. (2009). Dreaming and imagination. Mind and Language, 24(1), 103–121.

Kahn, D., Pace-Schott, E. F., & Hobson, J. A. (2002). Emotion and cognition: Feeling and character identification in dreaming. Consciousness and Cognition, 11, 34–50.

Kramer, M. (2013). The Dream Experience: A Systematic Exploration. Routledge.

McGinn, C.  (2004). Mindsight:  Image, dream, meaning. Cambridge, MA: Harvard University Press.

McNamara, P., McLaren, D., & Durso, K. (2007). Representation of the self in REM and NREM dreams. Dreaming, 17(2), 113–126. doi:10.1037/1053-0797.17.2.113.

Murzyn, E. (2008). Do we only dream in colour? A comparison of reported dream colour in younger and older adults with different experiences of black and white media. Consciousness and Cognition, 17, 1228–1237.

Noreika, V., Valli, K., Lahtela, H., & Revonsuo, A. (2009). Early-night serial awakenings as a new paradigm for studies on NREM dreaming. International Journal of Psychophysiology, 74(1), 14–18.

Occhionero, M., Cicogna, P., Natale, V., Esposito, M. J., & Bosinelli, M. (2005). Representation of self in SWS and REM dreams. Sleep and Hypnosis, 7(2), 77–83.

Rechtschaffen, A., & Buchignani, C. (1992). The visual appearance of dreams. In J. S. Antrobus & M. Bertini (Eds.), The neuropsychology of sleep and dreaming (pp.  143–156). Hillsdale, NJ: Erlbaum

Revonsuo, A. (2006). Inner presence: Consciousness as a biological phenomenon. Cambridge, MA: MIT Press.

Revonsuo, A., Tuominen, J., & Valli, K. (2015). The Avatars in the Machine. In T. K. Metzinger & J. M. Windt (Eds.), Open MIND. Frankfurt am Main: MIND Group.

Rosen, M., & Sutton, J. (2013). Self-Representation and Perspectives in Dreams: Perspectives in Dreams. Philosophy Compass, 8(11), 1041–1053. https://doi.org/10.1111/phc3.12082

Schwartz, S. (2000). A historical loop of one hundred years: Similarities between 19th century and contemporary dream research. Dreaming, 10, 55–66.

Schwitzgebel, E. (2002). Why did we think we dreamed in black and white? Studies in History and Philosophy of Science Part A, 33(4), 649–660. https://doi.org/10.1016/S0039-3681(02)00033-X

Schwitzgebel, E. (2007). Do you have constant tactile experience of your feet in your shoes? Or is experience limited to what’s in attention? Journal of Consciousness Studies, 14(3), 5–35

Schwitzgebel, E. (2011). Perplexities of consciousness. Cambridge, Mass: MIT Press.

Sikka, P., Valli, K., Virta, T., & Revonsuo, A. (2014). I know how you felt last night, or do I? Self- and external ratings of emotions in REM sleep dreams. Consciousness and Cognition, 25, 51–66. https://doi.org/10.1016/j.concog.2014.01.011

Solms, M.  (1997). The neuropsychology of dreams: A clinico-anatomical study. Mahwah, NJ: Erlbaum.

Strauch, I., & Meier, B. (1996). In search of dreams: Results of experimental dream research (M. Ebon, Trans.). New York: State University of New York Press.

Windt, J. M., & Noreika, V. (2011). How to integrate dreaming into a general theory of consciousness: A critical review of existing positions and suggestions for future research. Consciousness and Cognition, 20(4), 1091–1107.

Windt, J. M., Harkness, D. L., & Lenggenhager, B. (2014). Tickle me, I think I might be dreaming! Sensory attenuation, self-other distinction, and predictive processing in lucid dreams. Frontiers in Human Neuroscience, 8. https://doi.org/10.3389/fnhum.2014.00717


  1. Thanks for this very useful analysis. It is very helpful to see the empirical literature and philosophical discussion integrated this way. I wonder if Arnold Trehub is reading this, it seems to resonate strongly with his theory.

  2. Jennifer Windt:

    “The simplest, most reduced kind of experience that can still be described as involving a self is the experience of spatiotemporal self-location within the dream world. As long as there is still a phenomenal here and a phenomenal now, we still have the feeling of presence; we still locate ourselves, as it were, at the center of the dream world.”

    Jennifer, you wrote:

    “The simplest, most reduced kind of experience that can still be described as involving a self is the experience of spatiotemporal self-location within the dream world. As long as there is still a phenomenal here and a phenomenal now, we still have the feeling of presence; we still locate ourselves, as it were, at the center of the dream world.”

    I agree with the importance of spatio-temporal self-location within a world as we dream, as well as when we are awake. But my contention is that this essential self-location is actually provided by the volumetric neuronal structure and dynamics of a brain mechanism in which there are neurons at a fixed locus of perspectival origin within this mechanism. I suggest that these cells of perspectival origin constitute our core self (I!), and that our core self is a necessary precursor for the construction of our phenomenal self model (PSM).
    I take your comment about the essential role of self-locationto be in agreement with this view. In dreams, our (PSM) can be dramatically different than when we are awake, but our core self (I!) is unchanged. In out of body experiences, the same happens. See, for example, “Where Am I? Redux”. Here: https://people.umass.edu/trehub/Where%20Am%20I%20-%202.pdf

  3. Jennifer Windt

    Dear Eric and Arnold,

    thanks for pointing this out, I think there is a lot of common ground between these views. One possible difference could be that I perhaps emphasize more strongly the vestibular and also the temporal components of self-location, whereas Arnold seems to emphasize the visual ones; but I very much agree with the general strategy, and also with the comparison of OBEs and full-body illusions. So again, thanks!

    All best,

Comments are closed.

Back to Top
%d bloggers like this: