Are Dreams Less Real Than Waking Perceptions?

Dreams may be more than mere subjective noise.

Posted Sep 29, 2017

Most philosophers and scientists hold the belief that what we see in dreams is less real than what we see in waking perceptions. For example, two philosophers who have thought most carefully about dreams, Jennifer Windt and Antii Revonuso speak of dreams as non-veridical hallucinations or mere, selective simulations of the waking world; and virtually all other philosophers who have discussed dreams at all in their work also view dreams as hallucinatory and a distorted or degraded form of waking perception.

There is good reason to believe that dreams are a degraded form of a perceptual or mental simulation process. It is likely that ALL perceptions whether waking or dreaming are the result of predictive simulations conducted by the brain in order to anticipate world events/opportunities. Thus, the brain in sleep continues its predictive constructions about anticipated events except it does so in the absence of incoming sensory information. That is because during REM sleep some sensory input is gated or blocked by the brain from entering into its system. In addition, motor output is blocked during REM sleep. Thus, the brain is isolated from incoming sensory information and is prevented from acting on the world during REM sleep. Whatever the brain is processing during REM cannot be information from the external world, so the argument goes, so what we view in dreams cannot be real; cannot be perceptions of the external world.

A very good analytic philosopher Brian Leftow in his 1991 book Time and Eternity produced a sustained argument for the lesser reality of dreams relative to waking perceptions. We can take his discussion of the issue to be more or less representative of standard philosophical thinking on the status of dreams. He claims that we can create a set of rough and ready criteria for judging the degree of reality accorded to our experiences. What is most real will exhibit the following properties:

  • Continuity…is order between episodes of experience. Dreams jump from one scene to another and themes or dream content does not typically carry over from one dream to the next
  • Coherence…is order within episodes. Within an episode of consciousness our experience of waking objects is stable and orderly while in dreams there is great instability and some objects may not obey the laws of physics
  • Explicability…an experience is explicable if and only if its objects’ behavior can be rationally explicable. There are regular stable patterns in the waking world that allows us to develop justified expectations about what will occur next. This is not so with regard to objects or events in dreams.
  • Hardness…is the characteristic of resisting one’s wishes and agency. Leftow concedes that while dream objects are sometimes outside of the dreamer’s personal control and agency we sometimes have superhuman powers in dreams.
  • Inclusiveness…is where one world or experience includes another which is subsumed by it. If world A includes world B and not vice versa B is less real than A. We rarely remember the waking world in dreams but we easily remember a dream upon awaking.
  • Intersubjectivity…is checking perceptions against what others’ see. If my perceptions do not agree with others then I may be deluded. My dream worlds are private to me and cannot be experienced by others.

In fairness to Leftow dreams were not his primary interest in developing these criteria for realness. He simply uses dreams to illustrate his larger thesis that we generally hold some experiences to be more real than others and he provides these criteria in order to explain why we are justified in doing so. Unlike other philosophers and scientists, he at least attempts to justify his intuition that dreams are less real than waking perceptions.

Now although I largely agree with Leftow and the vast number of scientists and philosophers who hold similar opinions regarding the lowly status of dreams, I want to push back a little on this dogma concerning the unreality of dreams.

First of all, it is really just dogma and not established scientific fact. The so-called sensory blockade during REM sleep, to begin with, is not total. True the visual sense is blockaded pretty well but none of the other senses undergo anything like total blockade during REM. In addition, robust evoked potentials can be easily obtained from the sleeping brain during REM indicating substantial sensory processing occurring during REM sleep. The only real difference between predictive simulations occurring during waking consciousness vs those occurring during dreaming is that the simulations use different sensory criteria to guide and correct the simulations: in waking life visual criteria most often guide corrections while in REM sleep some other criteria are used (possible violations of expected temporal patterns).

With regard to Leftow’s reality criteria all that I can say is that like most other philosophers and scientists he is simply wrong on all counts regarding dreams and that is because like most other philosophers and scientists he is simply unacquainted with the experimental investigations into dream content that have occurred since the discovery of REM in the 1950s. For example, with regard to continuity and coherence… Most dreams really do not typically jump from one scene to another. On the contrary careful analyses of the narrative structure of dream episodes demonstrate levels of coherence equal to that of waking narratives. In addition, it has been demonstrated that when you awaken individuals several times a night and ask for dream reports you will find thematic continuity across dreams over the course of the night. In addition, there is great continuity of dream themes across nights as well. Indeed longitudinal studies of dream content have demonstrated beyond any reasonable doubt continuity of dream themes can continue for years. With regard to explicability…because most dream episodes have been found to be typically stable and coherent dream characters act accordingly. They anticipate stability and develop plans within dreams based on past events in dreams and so forth. If for example, the dreamer is being chased by a monster and finds that the monster can be slowed down by throwing obstacles in its path, the dreamer will plan accordingly and do so. For most dreams, therefore, we can develop or explicate reasons for why people, characters do the things they do in the dream scenes. There is an internal logic to dreams. With regard to the hardness criterion, it is clear that objects and other characters can and often do resist the dreamer’s wishes and desires. That dreamer with the monster chasing him does not desire the monster to do that. With regard to inclusiveness, the dream world cannot be considered to be parasitic upon the waking world. Dreams draw upon waking experiences extensively to develop their content but there is also a great novelty in dreams where dreamers sometimes encounter complete strangers and new worlds. It simply cannot be argued that the dream world is contained within the waking world while the waking world is not contained within the dream world. The waking world is often referred to and remembered in dreams.

I saved the intersubjectivity criterion for last as that is the criterion that philosophers and scientists most love to use to dismiss dreams as mere subjective noise and fantasy. The intersubjectivity criterion claims that one thing is more real than another thing if it can be seen and verified by more than one person. My dream worlds are private to me and cannot be experienced by others; therefore it is subjective fantasy—or so the argument goes. To push back a little against this criterion let me first point out that in the waking world intersubjectivity is not an entirely reliable criterion for reality. Two or more persons can believe they see the same thing and be utterly deluded. It is easy to get an experimental subject to “see” an object by simply having an authority figure suggest it is there. Conformist bias is also common and occurs when subjects begin to see objects when the rest of the group claim that they see the object. Crowd delusions are ubiquitous and occur when millions of people see and believe the same false thing (“We can make a killing in the housing markets!”) and so on.

While dreams, like any phenomenal experience, are first-person private experiences, that does not mean that they cannot be shared or that they are any less real than other private experiences. While not everything in a private experience can be publicly verified, many parts of private experiences can be so verified. We increase the confidence we have in dream realities the same way we increase confidence we have in any private experiences: we share them and compare notes about them. When we share dreams we find that there are some commonalities between your dreams and my dreams. Their formal properties are similar and so on. That is enough to build a science of dreams on and should be enough to satisfy the intersubjectivity criterion. We can objectively share private experiences. There is nothing particularly mysterious in this procedure.

Where does all this leave us? Are dreams less real than waking experiences? On the face of it, the answer has to be yes but it is very difficult to say just why dreams are less real. Obviously, dreams are a different order of reality than waking experiences but does this fact make them less real than waking experiences?  In my opinion, we do not yet have a satisfactory answer to that question.