PHILOSOPHICAL PSYCHONAUTICS What Can We Learn from Lucid Dreaming - IDEAS AND DISCOVERIES - The Ego Tunnel - Thomas Metzinger

The Ego Tunnel: The Science of the Mind and the Myth of the Self - Thomas Metzinger (2009)


Chapter 5. PHILOSOPHICAL PSYCHONAUTICS What Can We Learn from Lucid Dreaming?

During the night of May 6, 1986, I became consciously aware that I was sleeping and also spiraling out of my physical body, in the typical manner described by Swiss biochemist Ernst Waelti (see chapter 3). Here is my “case study”:

Standing in front of my bed, I immediately realized that, for the first time in two years, I had entered the OBE state again. The clarity, the same electrified sense of lightness in my double body, made me excited and extremely happy, and I immediately began to experiment. I moved toward the closed glass door of the second-floor balcony in my parents’ house. I touched the door, gently pushing it until I penetrated it and slid out onto the balcony. I flew down into the garden and landed on the lawn, where I moved around in the dim moonlight and looked at things. Again, the overall experience was crystal clear.

When I became afraid of not being able to sustain the condition much longer, I flew back up, somehow returned to my physical body, and awoke with a mixture of great pride and joy. I had not managed to make any verifiable observations, but I had had another OBE, in a clear, cognitively lucid way, fully controlled and without any intermediary blackouts. I sat up, wanting to take notes as long as everything was still fresh, but couldn’t find a pencil.

I jumped out of bed and went over to my sister (who slept in the same room), woke her up, and told her, with great excitement, that I had just managed to do it again, that I had just been down in the garden, bouncing around on the lawn a minute ago. My sister looked at her alarm clock and said, “Man, it’s quarter to three! Why did you have to wake me up? Can’t this wait until breakfast? Turn out the light and leave me alone!” She turned over and went back to sleep. I was a bit disappointed at this lack of interest.

I also noticed that while fumbling with the alarm clock, she had accidentally set it off. It was beeping away and I hoped it hadn’t wakened anybody else. Too late! I could hear someone approaching.

At that moment, I woke up. I was not upstairs in my parents’ house in Frankfurt but in my basement room, in the house I shared with four friends about thirty-five kilometers away. It was not quarter to three at night; the sun was shining and I had obviously been taking a short afternoon nap. For more than five minutes, I sat on the edge of my bed almost frozen, not daring to move. I was unsure how real this situation was. I did not understand what had just happened to me. I didn’t dare move, because I was afraid I might wake up again, into yet another ultrarealistic environment.

In dream research, this is a well-known phenomenon called false awakening. Did I really have an out-of-body experience? Or did I only have a lucid dream of an out-of-body experience? Can one slide from an OBE into an ordinary dream via a false awakening? Are all OBEs forms of lucid dreaming in the first place? To wake up twice in a row is something that can shatter many of the theoretical intuitions you have about consciousness—for instance, that the vividness, the coherence, and the crispness of a conscious experience are evidence that you are really in touch with reality. Apparently, what we call “waking up” is something that can happen to you at any point in phenomenological time. This is a highly relevant empirical fact for philosophical epistemology. Do you recall from chapter 2 the discussion about the evolution of human consciousness and how the distinction between things that only appear to us and objective fact became an element of our lived reality? Now we can see what it means that the appearance/reality distinction emerged only on the level of appearance: False awakenings demonstrate that consciousness is never more than the appearance of a world. There is no certainty involved, not even about the state, the general category of conscious experience in which you find yourself. So, how do you know that you actually woke up this morning? Couldn’t it be that everything you have ever experienced was only a dream?1

Dreams are conscious because they create the appearance of a world, but, as noted in chapter 2, they are offline states—global states of conscious experience in which the Ego is decoupled from sensory input and unable to generate overt motor behavior. The dream tunnel not only contains the appearance of a world but also (in most cases) creates a fully embodied, spatially extended self moving around in a spatially extended environment. The virtual self thus born is an exclusively internal phenomenon in an even stronger sense than that of the waking self: It is immersed in a dense mesh of causal relations, all of which are internal to the brain. Dreamers are self-aware, but functionally they are not situated . Dreams are subjective states in that there is a phenomenal self; however, the perspective from which this conscious self perceives the world is very different—and much more unstable—than it is during wakefulness.

Have you ever noticed that you cannot control your attentional focus in your dreams? High-level attention is typically missing. Accordingly, the dream-self generated inside the Ego Tunnel when you are sleeping lacks the specific phenomenal quality I described in the preceding chapter as attentional agency, the conscious experience of directing the beam of your inner flashlight deliberately and selectively at various objects. But attentional agency is not just the ability to “zoom in” on certain things or point your mind at particular features of your world-model; it also entails the sense of ownership—ownership of the selection process preceding the shift in attention. Both aspects are missing in dreams. In a way, you are like an infant or a severely intoxicated person. The dream Ego is much weaker than the waking Ego.

If one penetrates deeper into the specific phenomenology created by the dreaming Ego, one discovers a considerable weakness of will and severe distortions of the thought process. In ordinary dreams, you sometimes cannot experience yourself as any sort of agent at all. It is difficult, for example, to make a decision and follow through with it. But even if you manage that, you are typically unable to ascribe agency to yourself. The dreaming self is a confused thinker, severely disoriented with regard to places, times, and people’s identities. Short-term memory is greatly impaired and unreliable. Also, only rarely does the dream self have such sensory experiences as pain, temperature, smell, or taste. Even more interesting is the extreme instability of the first-person perspective: Attention, thinking, and willing are highly unstable and exist only intermittently, yet the ordinary dreaming Ego does not really care about this, or even notice it. The dream self is like the anosognostic patient, who lacks insight into a deficit following brain injury.

At the same time, the dream self creates intense emotional experiences—some aspects of the self are clearly stronger in the dream tunnel than in the tunnel of waking consciousness. Anyone who has ever had a nightmare knows how intense the feeling of panic can become during dreams. In the dream state, the emotional self-model can be characterized by unusually intense degrees of feeling, though this is not true for all emotions; for example, fear, elation, and anger predominate over sadness, shame, and guilt.2

Occasionally, the dream tunnel enables the Ego to access information about itself that is unavailable during the waking state. Whereas short-term memory is commonly impaired, long-term memory can be greatly enhanced. For instance, it is possible to relive childhood episodes vividly—memories that would never have been accessible during wakefulness. We tend to forget these afterward, because most of us have weak dream recall. But as long as the dream lasts, we have access to state-specific forms of self-knowledge.

Blind people are sometimes able to see in dreams. Helen Keller, who turned blind and deaf at the age of nineteen months, emphasized the importance of these occasional visual experiences: “Blot out dreams, and the blind lose one of their chief comforts; for in the visions of sleep they behold their belief in the seeing mind and their expectation of light beyond the blank, narrow night justified.”3 In one study, congenitally blind subjects produced dream drawings that judges were unable to distinguish from drawings of sighted subjects, and as EEG correlates between were sufficiently similar, this strongly suggests that they can see in their dreams—but do they?4 It is also interesting to note that Keller’s dream tunnel contained the phenomenal qualities associated with smell and taste, which most of us experience only rarely in the dream state. It seems as if her dream tunnel became richer because her waking tunnel had lost some of its qualitative dimensions.

The dream tunnel shows to what extent conscious experience is a virtual reality. It internally simulates a behavioral space, a space of possibilities in which you can act. It simulates real-life sense impressions. As discussed in chapter 3, this is exactly what modern designers of virtual realities are trying to achieve (indeed, one of the best scientific journals on virtual-reality technology is titled Presence). It is precisely this sense of presence and full immersion that our biological ancestors achieved long ago. The resultant Ego, however, has created a more robust sense of presence for dreaming and for waking life as well. If it had not done so, we probably would not be trying to create virtual realities today, nor would we research the ability of the human brain to achieve this miracle within itself.

Even though dreams are behavioral spaces, they are not causally coupled to the real behavioral space of the dreaming human organism. Dreamers are not bodily agents; their behavior is internal, simulated behavior. The inhibition of the spinal motor neurons prevents bodily behavior from being generated during dream sleep—that is, REM (rapid-eye-movement) sleep. This is how the dream Ego is separated from the physical body. When this motor inhibition fails, as it does in a disorder known as RBD (for “REM-sleep behavior disorder”), internal dream behavior is acted out in the waking world. Typically found in men over sixty, RBD is associated with a loss of the muscle atonia that typically accompanies REM sleep. Patients suffering from RBD are forced to act out dramatic and often violent dreams. They will shout or grunt. They may attempt to strangle their bed partners, set fire to their beds, jump out of windows, even fire a gun.5 Later they will recall little or nothing of this physical activity—unless they fall out of bed or bump into furniture or injure themselves or someone else and wake up. But they can usually recall the dreams themselves, which typically involve such physical activities as fighting, running, chasing or being chased, and attacking or being attacked. These patients also seem to experience violent and aggressive dream content more frequently than healthy subjects do. Obviously, this is a dangerous condition that can lead to self-inflicted injuries and serious sleep deprivation. What we can learn from it is how the dream body, in normal circumstances, is decoupled from the physical body. Normally, dreamers are not bodily agents, and all their behavior is purely internal, simulated behavior. But when motor inhibition fails, as it does in RBD, internal dream behavior is enacted by the physical body.

The most interesting feature of ordinary dreams leads to some deeper philosophical considerations about the nature of consciousness. The dream tunnel is generated in a very special configuration: During REM sleep, as noted, there is an output blockade, responsible for the paralysis of the sleeper, and there is an input blockade, which prevents (at least to a degree) sensory signals in the sleeper’s environment from penetrating conscious experience. At the same time, chaotic internal signals are generated by what are known as PGO waves. They are electrical bursts of neural activity named for the brain areas involved (the pons, the lateral geniculate nucleus in the hypothalamus, and the occipital primary visual cortex) and are closely related not only to eye movements but also to the processing of visual information.6

As the brain tries to understand and interpret this chaotic internal pattern of signals, it starts telling itself a fairy tale, with the dream ego playing the leading role. The interesting point is that the dream Ego does not know that it is dreaming. It does not realize the signals it is turning into an internal narrative are self-generated stimuli—in philosophical jargon, this feature of the dream state is a “metacognitive deficit The dream Ego is delusional, lacking insight into the nature of the state it is itself generating.


The natural question to ask is if there are any dreams with additional insight, dreams in which the dream self-model has become so strong and rich that it allows us to understand what is happening. Can one consciously enjoy one’s own internal virtual reality? Is it possible to dream without the metacognitive deficit? The answer is yes. You can have dreams in which you are not only aware of the fact that you are dreaming but also possess a complete memory both of your dream life and your waking life, as well as the phenomenal property of agency on the levels of attention, thought, and behavior. Such dreams are called lucid dreams. They are highly interesting—not so much for the sheer fun of the drama but because they open new ways of investigating the phenomenon of conscious experience. In particular, they help us understand how the various layers of the self-model are constructed and woven into the dream tunnel.

Dutch psychiatrist Frederik van Eeden, who coined the phrase “lucid dreaming,” reported the following experience to the Society for Psychical Research in 1913:

In January, 1898 … I was able to repeat the observation… . I dreamt that I was lying in the garden before the windows of my study, and saw the eyes of my dog through the glass pane. I was lying on my chest and observing the dog very keenly. At the same time, however, I knew with perfect certainty that I was dreaming and lying on my back in my bed. And then I resolved to wake up slowly and carefully and observe how my sensation of lying on my chest would change to the sensation of lying on my back. And so I did, slowly and deliberately, and the transition—which I have since undergone many times—is most wonderful. It is like the feeling of slipping from one body into another, and there is distinctly a double recollection of the two bodies. I remembered what I felt in my dream, lying on my chest; but returning into the day-life, I remembered also that my physical body had been quietly lying on its back all the while. This observation of a double memory I have had many times since. It is so indubitable that it leads almost unavoidably to the conception of a dream-body.7

Van Eeden’s “dream-body” is the self-model in the dream state. Lucid dreams are fascinating because our naive realism—our unawareness of living our lives in an Ego Tunnel—is temporarily suspended. They are therefore a promising route of research for solving what I termed the Reality Problem in our tour of the tunnel in chapter 2. A lucid dream is a global simulation of a world in which we suddenly become aware that it is indeed just a simulation. It is a tunnel whose inhabitant begins to realize that he or she actually operates in a tunnel all the time.

Hugh G. Callaway, a British experimenter in out-of-body experiences who published under the pseudonym Oliver Fox, recorded the following classic episode, occurring in 1902, when he was a young science student at the Harley Institute at Southampton:

I dreamed that I was standing on the pavement outside my home… . I was about to enter the house when, on glancing casually at [the pavement] stones, my attention became riveted by a passing strange phenomenon, so extraordinary that I could not believe my eyes—they had seemingly all changed their position in the night, and the long sides were parallel to the curb! Then the solution flashed upon me: though this glorious summer morning seemed as real as real could be, I was dreaming! With the realization of this fact, the quality of the dream changed in a manner very difficult to convey to one who has not had the experience. Instantly, the vividness of life increased a hundredfold. Never had the sea and sky and trees shone with such glamorous beauty; even the commonplace houses seemed alive and mystically beautiful. Never had I felt so absolutely well, so clear-brained, so inexpressibly free! The sensation was exquisite beyond words; but it lasted only a few minutes and I awoke.8

Maybe you’ve had a lucid dream yourself; the phenomenon is not rare. If not, you can try a number of different induction techniques. For instance, you can adopt the habit of performing “reality checks” several times a day. Each reality check should last at least a minute. It consists in carefully inspecting your current inner and outer environment for any indications that this might not be ordinary waking reality. Here is a checklist that readers interested in exploring the dream tunnel can use as a guideline.

✵ Is all the furniture arranged the way it normally is?

✵ Are the paving stones, the tiles, or the design of the carpet on the floor arranged in the same pattern as usual?

✵ Do objects or persons suddenly appear and disappear, or do they change their identity?

✵ Do you know who you are and where you are?

✵ Do you remember what day of the week it is and when you last woke up?

✵ Are there any gaps in your short-term memory of recent events?

✵ Does your visual attention shift the way it usually does?

✵ Are you engaging in any unusual physical activities, like flying?

✵ Are you constantly trying to remember something you know is of great importance but can’t remember what it is?

✵ Does your current situation have a metaphoric or symbolic character, or do you have the feeling of being close to an important discovery?

If you perform reality checks of this type several times a day, you have a good chance of eventually becoming a lucid dreamer. By pure habit, you will one day perform a reality check in a dream—and if you are lucky, you will correctly realize that you are dreaming.9

Other methods of inducing lucid dreams are even more efficient. Try setting an alarm clock early in the morning and carefully writing down the events of your last dream. Get up, move around for a short period of time, and then go back to bed. While you are falling asleep, try to rehearse the last sequence of dream events in as much detail as you can. You may find that you can consciously reenter the dream and stay lucid throughout it.10

As an intrepid philosophical psychonaut, I have of course tried to build devices to do this kind of exploring, involving headphones and tape loops whispering, “Watch out—this is a dream!” at thirty-minute intervals all night long. I also bought an expensive lucid-dreaming device called a Nova Dreamer, which looks a bit like the eye masks you sometimes see people wearing on long-distance plane flights. The Nova Dreamer is activated when your rapid eye movements signal the start of a dream. After a couple of minutes, it begins submitting mild subliminal visual stimuli, and you can perceive these soft, red, ring-shaped flashes of light through your closed eyelids. They are meant to alert you to the fact that you are dreaming; however, they are more likely to be integrated into your dream story. Here’s one of my own dreams thus invaded:

I am an astronaut. I have been waiting for this moment for years. My friend and I are lying on our backs in the Space Shuttle, awaiting takeoff with a mixture of anxiety and great excitement. Deep below our backs, we can feel the rumbling and rattling of the ignition give way to a thundering roar. Then red lights start flashing everywhere on the control panel. Suddenly, every possible alarm system is activated. Someone says, “Something must have gone terribly wrong!” We feel the spaceship slowly tilting to the side and losing its vertical position while the roar at our backs gets louder and louder.

Unfortunately, all I ever got out of my expensive lucid-dreaming device was terrible nightmares—with an interesting twist. In Germany, the flashing lights of police cars are blue. So what I got from this device was American nightmares, with American police cars hunting me down and cornering me, flashing red lights and all. Every two years or so, I give my Nova Dreamer another try; lately, it has had a different effect on me. I wake up in the morning and the device is gone. If I go looking for it, I find that it has been hurled across the bedroom by some stranger. Apparently there is someone inside me who does not want to be a philosophical psychonaut or a serious practitioner of first-person phenomenological research at all—someone who just wants to sleep.

So what, exactly, is a lucid dream? In a lucid dream, the dreamer knows that she is currently experiencing a dream and is able to ascribe this property to herself. If we opt for a strong definition, another condition is that she also has access to memories of her previous dream and waking life. Autobiographical memory is fully intact. The dreamer has full access not only to past conscious experiences in waking life and in ordinary dreams but also to previously experienced lucid dreams. The overall level of mental clarity and cognitive insight is at least as high as it is during normal waking states. A further defining characteristic is that, according to subjective experience, all five senses function just as well as they do during the waking state. Finally, and perhaps most important, the property of agency is fully realized in the lucid dream. Phenomenologically, the lucid dreamer knows about her freedom of will. Not only can she direct the focus of attention wherever she likes, but she can also actually do whatever she wants—fly, walk through walls, or engage in conversations with dream figures. The subject of a lucid dream is not a passive victim lost in a sequence of bizarre episodes but rather is a full-blown agent, capable of selecting from a variety of possible actions.

Full control of one’s attention is an important feature distinguishing lucid dreams from ordinary dreams. Insight into one’s freedom to act is also an important criterion of lucid dreams (but is it an insight?). During what are sometimes called pre-lucid dreams, we frequently become aware that none of this is real, that this must be a dream, but we remain passive observers. With the onset of full lucidity, the dreamer often turns from a passive observer into an agent—someone who takes charge, moves around, explores and experiments, who deliberately starts to interact with the dream world and shape it.

My favorite experiment in lucid-dream research was conducted by Stanford University psychophysiologist Stephen LaBerge and his colleagues more than a quarter of a century ago.11 It exploited the interesting fact that our conscious self-model is firmly anchored in the brain in a fascinating way: There is a direct and reliable relationship between the gaze shifts reported by lucid dreamers and the eye movements observed in their sleeping bodies. In the sleep lab, these eye movements can be recorded using a polygraph. The fact that the movements of the dream-eyeballs in the dream-body are directly correlated with the movements of the physical eyeballs in one’s physical body was used by LaBerge in a particularly ingenious experiment. Veteran subjects deliberately indicated the initiation of a lucid dream with specific ocular signals determined before the experiments—that is, by rapidly moving their eyes up and down. Two such eye shifts would inform the experimenters of the onset of a lucid dream; four signaled awakening. The polygraphic analysis revealed that the onset of lucidity is typically correlated with the first two minutes of an REM phase, or with short intervals of waking consciousness during an REM phase, or with heightened phasic REM activity (characterized by bursts of eye movements and sometimes by motor twitches and widespread synchronized activity in specific thalamocortical networks).12 Put simply, lucidity seems to occur when there is a brief and sudden increase of the general cortical level of arousal: All nerve cells become more active, the result being the sudden availability of more “computational power,” or capacity for information processing. With regard to the dream itself, lucidity seems to lead to increased vividness, heightened fear or stress, the discovery of contradictions within the dream world, and, of course, the subjective experience of becoming aware of a “dreamlike” or “unreal” quality of reality.

I like these experiments because they are a rare example of trans-tunnel communication. When the lucid dreamer in the sleep lab emits eye signals by deliberately moving his or her dream-eyes up and down and scientists in the waking world read these signals off their instruments, a multiuser link between the dream tunnel and the waking tunnel is established. Because the gaze shifts performed by the dream-body are functionally linked to those of the physical body, and because the lucid dreamer is aware of this fact, a bridge connecting the two tunnels is built. In this experimental setup, information from one type of conscious reality tunnel can be transmitted to another type, one created by the brains of other human beings.

We need more good empirical research on lucid dreams. It is plausible to assume that lucidity depends on the degree to which the prefrontal cortex, where the organizing of cognitive and social behaviors takes place and the so-called executive functions are located, can form a stable functional link with other brain regions that generate the conscious dream self. The prefrontal cortex is thought to arrange thoughts and actions in accordance with internal goals. It also has to do with differentiation among conflicting thoughts, planning, assessing future consequences of current activities, predicting outcomes, generating expectations, and the like.

Allan Hobson, a psychiatrist and dream researcher at the Massachusetts Mental Health Center and author of The Dreaming Brain, has speculated that for lucidity to occur, “the normally deactivated dorsolateral prefrontal cortex (DLPFC) must be reactivated, but not so strongly as to suppress the pontolimbic signals to it.”13 This part of the brain may allow us to refer to ourselves by engaging in reflective thought. In the lucid-dream tunnel, this leads to the reestablishment of executive control and the reemergence of a full-blown agent. If Hobson is right, the moment we consciously think, “My God, I’m dreaming!” may be the moment the self-model of the dream state becomes hooked up to the prefrontal cortex, making proper reflexive self-consciousness possible again and reestablishing cognitive agency.

Here are some questions for future research: What precisely happens to the conscious self during the transition from an ordinary dream to a lucid dream? What are the fine-grained functional differences between the dream self-model and the lucid self-model? Could there also be something like “lucid waking”? And what, exactly, happens during a false awakening?

As we have seen, false awakenings can happen to all of us. This brings up another classical philosophical problem, the issue of solipsism (from Latin: solus, alone, and ipse, self). How, exactly, can I refute the skeptical hypothesis that my mind is the only thing that I know to exist?

How can I exclude the possibility that the external world—and other conscious minds in particular—cannot be known and might not exist at all? Finally, here is a little thought experiment in applied tunnel epistemology, introduced and illustrated by a lucid dream reported by the late German dream researcher Paul Tholey:

I briefly looked back. The person following me did not look like an ordinary human being; he was as tall as a giant and reminded me of Rübezahl [a mountain spirit in German legend]. Now it was fully clear to me that I was undergoing a dream, and with a great sense of relief I continued running away. Then it suddenly occurred to me that I did not have to escape but was capable of doing something else. I remembered my plan of talking to other persons during the dream. So I stopped running, turned around, and allowed the pursuer to approach me. Then I asked him what it actually was that he wanted. His answer was: “How am I supposed to know?! After all, this is your dream and, moreover, you studied psychology and not me.”14

Imagine that while in the dream tunnel, you suddenly become lucid and find yourself at a major interdisciplinary conference, where dream scientists and dream philosophers are discussing the nature of consciousness:

While they’re standing around during the coffee break, one of them claims that you do not really exist, because you’re just a dream figure in your own lucid-dream tunnel, a mere possibility. Amused, you reply, “No, you are all figures in my dream—just figments of my imagination.” This response is greeted with laughter, and you notice, too, that colleagues at other tables are grinning and turning their heads in your direction. “All of this is happening in my brain!” you insist. “I own the hardware, and you are all just simulated dream characters in a simulated environment, processed and created by my central nervous system. It would be easy for me…” Here, more laughter interrupts you—roars of laughter. A young PhD student arrogantly starts explaining the basic assumptions about the nature of reality shared by this particular scientific community: No such things as brains or physical objects ever existed. The contents of consciousness are all there is. So all phenomenal selves are equal. There is no such thing as an individual “tunnel” in which one self-model represents the true subject of experience and all other person-models are just dream figures.

The strange philosophical concept this dream community of scientists has developed as their background assumption is known as eliminative phenomenalism. As the slightly overambitious PhD student explains: “Eliminative phenomenalism is the thesis that physics and the neuroscientific image of man constitute a radically false theory, a theory so fundamentally defective that both the principles and the ontology of that theory will eventually be displaced, rather than smoothly reduced, by a completed science of pure consciousness.” All reality, accordingly, is phenomenal reality. The only way you can drop out of this reality is by making the grandiose (but fundamentally false) assumption that there actually is an outside world and that you are the subject—that is, the experiencer—of this phenomenal reality, that there actually is a consciousness tunnel (a wormhole, as they ironically call it), and that it is your own tunnel. By entertaining this belief, however, you would suddenly become unreal and turn into something even less than a mere dream figure yourself: a possible person—exactly what your opponent claimed at the beginning of the discussion.

“Listen, guys,” you say, in a slightly irritated voice, “I can demonstrate to you that this is my consciousness tunnel, because I can end this state, and your very existence, at any point in time. A well-known technique for terminating lucid dreams is to hold one’s hands up in front of one’s eyes and fix one’s visual attention on them. If I do this, it will interrupt the rapid eye movements in my physical body and thus end the dream state in my physical brain. I will wake up in the Waking Tunnel. You will simply cease to exist. Do you want me to show you?” You note that your tone of voice sounds triumphant, but you also note that the amusement in the eyes of the other scientists and philosophers has changed to pity. The arrogant PhD student blurts out again: “But don’t you see that simply falling back into what you call ‘waking up’ doesn’t prove anything to anybody? You must demonstrate the truth of your ontological assumptions to this scientific community, on this level of reality. You cannot decide the question by simply degrading yourself to a virtual person and disappearing from our level. By waking up, you will learn nothing new. And you cannot prove anything at all—certainly not to us, but not to yourself, either. If you want to humiliate yourself by vanishing into your waking wormhole, then just go ahead! But the serious pursuit of consciousness research and of philosophical theory of science is something entirely different!”

How would you react? If I had not made the right decision at this point, I might never have finished this book. But enough tunnel epistemology for now.



Allan Hobson is professor of psychiatry at Harvard Medical School, where he founded the Laboratory of Neurophysiology in order to study the brain basis of dreaming.

Working with Dr. Robert McCarley, Hobson developed the reciprocal interaction model, according to which REM (rapid eye movement) sleep is generated by cholinergic brainstem mechanisms, and the activation-synthesis theory, which views dreaming as the result of automatic brain activation and the synthesis of chaotic internal signals during sleep. In the course of his extensive experiments with human sleep laboratory data, Hobson invented the Nightcap method of home-based sleep recording and, with Robert Stickgold, used this method to characterize conscious states around the clock. Hobson and Stickgold also developed a new approach to the study of sleep effects on learning.

Recently, Hobson has integrated his own ideas and findings with new data coming from PET and lesion studies of human sleep in a general model of state-dependent aspects of consciousness. Called AIM, the new model maps three dimensions—activation (A), input-output gating (I), and chemical modulation (M)—to define a state space through which the brain-mind travels in a recurrent loop as we wake, sleep, and dream.

Hobson is the author of many books, including The Dreaming Brain (1989), Sleep (1995), Consciousness (1999), Dreaming as Delirium: How the Brain Goes Out of Its Mind (1999), The Dream Drugstore (2001), Out of Its Mind: Psychiatry in Crisis (2001), and 13 Dreams Freud Never Had: The New Mind Science (2004).

Metzinger: What exactly is special about consciousness in the dream state, compared to consciousness in wakefulness and non-REM sleep?

Hobson: Dream consciousness is more intense, more single-minded, more elaborate, and more bizarre than consciousness in waking. Hence, it can reasonably be viewed as the most autocreative state of the brain-mind. It is also the most psychosis-like state of normal consciousness. Because its neurobiology is so well known, its study offers us a unique scientific opportunity to understand ourselves better in both health and disease.

Metzinger: And what exactly is the relationship between REM sleep and dreaming?

Hobson: The correlation is quantitative, not qualitative. Dreamlike mental activity is also correlated with sleep onset (stage I) and with late-night sleep (stage II), but at all times of the night or day, the correlation is highest in REM. As for the actual relationship, my hypothesis is that dreaming is our subjective awareness of our brain activation in any state of sleep. Activation is highest in REM sleep. So is dreaming. I think that dreaming and REM sleep are our subjective and objective references to the same fundamental process of the brain-mind. I am a monist, through and through. How about you?

Metzinger: Sure—I have always liked philosophers like Spinoza, Bertrand Russell, or Herbert Feigl, who were neutral monists and thought that the distinction between physical and psychological states is actually quite superficial and rather uninteresting. For us philosophers, the more important problem, of course, is what precisely “through and through” means. But right now, you are the one that has to answer difficult questions! So, how do you explain the relationship between chaotic dream content, generated by the brainstem, and the more nonrandom and seemingly meaningful aspects of dreaming?

Hobson: Be careful, Thomas, or you will fall into the “either/or” trap that has swallowed up so many of our distinguished colleagues. The answer is “both/and.” REM sleep is generated by the brainstem, while dreaming is the subjective experience of the brainstem’s activation of the forebrain in REM sleep. The REM generation process has many chaotic features, which the forebrain tries its best to integrate into a coherent story. But the forebrain is also in a different state than it is in waking, which makes its job more difficult. The forebrain does the best it can under difficult circumstances. Whether you think it does well or not so well depends on whether you think the cup is half empty or half full. Both are true.

Metzinger: Which parts of the human brain are absolutely necessary for dreaming? Without which parts is it impossible to dream?

Hobson: The second question has empirical evidence contributing to an answer, but the first question is much more interesting and much more complex. Unfortunately, it cannot be answered scientifically.

Take the second question first. The neuropsychologist Mark Solms asked some three hundred stroke patients whether they had noticed any change in their dreaming after their strokes. Patients reported a complete cessation of dreaming if their stroke damaged either the parietal operculum or the deep frontal white matter. These claims were particularly interesting, because these same brain regions were selectively activated in PET studies of REM sleep. Another finding of interest is the report of dream cessation after prefrontal lobotomy, which Solms discovered in the literature of the 1940s and 1950s.

On their face, these findings suggest that dreaming depends upon the brain’s capacity to integrate emotional and sensory data when activated offline. But of course this doesn’t answer the first question at all. Many other brain regions are likely to be equally essential to dreaming. For example, the visual system must be involved—and, indeed, Solms’s patients reported the loss of visual imagery in their dreams if their strokes affected the occipital cortex. Presumably, the loss of dreaming is an example of what Norman Geschwind called disconnection syndrome. In other words, the damaged areas are cerebral crossroads which, when damaged, prevent other parts of the brain from interacting properly. The important role of the brainstem is unlikely to be revealed by this technique, since lesions large enough to impede dreaming are likely to be fatal or lead to unresponsive vegetative states.

There are several problems with this approach to dream science. The first is that the answer to question two does not answer question one. It is possible, for example, to imagine that Broca’s and Wernicke’s areas may be quite important to the confabulatory quality of dreaming, but this possibility cannot be tested if the patient has lost his ability to give dream reports! Furthermore, it is important to point out that all of Solms’s data deal with dream reporting, which cannot be equated with dreaming. In fact, most of us have little or no memory of our dreams.

In Solms’s studies and in the earlier works of [Cristiano] Violani and [F.] Dorrichi and of [M. J.] Farah and [M. S.] Greenberg, which reached similar conclusions about the parietal operculum, there was no effort made to record the patients’ sleep or to wake them up to elicit dream reports. These controls are important and yet to be performed. Solms and others are to be congratulated for opening up the neuropsychological study of dreaming. We look forward to learning more from this approach. For the present, all we can say is that dreaming depends on the selective activation and deactivation of many brain regions, including those which, when damaged, lead to the failure to report dreams.

Metzinger: What do you think was most likely the evolutionary function of dreaming, and when did it first develop?

Hobson: Regarding evolution and the functional advantages of having a brain that can dream, I have both conservative and speculative views. The conservative position is that there is no evidence that dreaming itself serves any purpose whatsoever. That is to say, neither the conscious awareness of dreams while they are occurring nor recall of such awareness on awakening from sleep is likely or has been shown to be useful. I think we must take seriously Owen Flanagan’s suggestion that dreams are the spandrels of sleep.15

At its most extreme, the argument says that dream consciousness is an epiphenomenon, which humans and other animals can do just as well without. The most cogent reason for thinking this may be true is the all-but-complete amnesia that we have for our dreams. If dream recall were adaptive, surely we would have more of it! But taking this position about dreaming as a conscious experience does not negate a healthy, speculative interest in the functional significance of having a brain that can self-activate in sleep. Such a brain could be doing many things. These include the already known enhancement of motor learning, the regulation of dietary and thermal calories, and the improvement in the immune functions. I don’t have to be aware of those functions, even if they are essential to my survival and my reproductive success.

Here we are at the nub of a number of critical philosophical questions, including the common confusion of brain activity and awareness. Our conscious awareness during waking is an obvious adaptive advantage, but our conscious awareness during sleep may not be. It may even be an adaptive advantage not to remember dream content. Allowing for some therapists’ assertion that dreaming is the royal road to the unconscious, it is still possible to ask, “Who wants to go there?” Those who do are free to try, but I myself see no adaptive advantage to dream recall and dream interpretation, even though I myself indulge, with great pleasure, in both sports.

My own special theory is that dreaming is a highly distinctive form of conscious awareness that can be used to better understand the brain activity that leads to consciousness, whether it be in waking or in sleep. As Gerald Edelman and Giulio Tononi have pointed out, it is the vast thalamocortical system that must be activated to produce consciousness. In waking and in sleep, this system is activated by the brainstem, but the chemical modulation accompanying the activation is very different in the two states. The contributions of other structures, like the limbic system and the modulatory systems of the brainstem, are very significant in that they “color” consciousness as well as activate it.

Humans and most other mammals have brains that can self-activate in sleep, when environmental conditions such as cold and darkness do not favor waking behavior, and it is this capacity, not the awareness of it, that is significant for evolutionary success.

Metzinger: What is known today about the phylogenetic manifestation of the sleep-wake cycle? How did it come about in our ancestors? And how is that manifestation related to consciousness?

Hobson: The answer is that a lot is known! Without going into elaborate details, it can be safely said that the fully developed sleep-wake cycle, with alternative phases of NREM and REM sleep, is an adaptation reserved to homeothermic animals—namely, mammals and birds that regulate their body temperature. What is the adaptive link between homeothermia and sleep? Again, the answer is simple. Keeping brain temperature constant despite enormous fluctuations of environmental temperature guarantees reliable brain function in a wide variety of environmental contexts. In other words, temperature control and brain function are tightly linked, and sleep secures that link.

With respect to the consciousness angle, I follow Edelman, who refers to primary consciousness—that is, perception, emotion, and memory—and secondary consciousness, which is awareness of awareness and the ability to describe it. Secondary consciousness, which depends upon language and other sophisticated abstractions, is exclusively human. Primary consciousness is widespread among mammals and could even be present in some submammalian species. Unfortunately, these assertions can never be more than intelligent guesses, because no subhuman animal can communicate its subjective experience verbally. Animal-rights activists, like right-to-life agitators, are quite right in claiming that many subhuman and immature animals are, to a limited but significant degree, conscious. If we are to take their lives or cause them pain, we had better have a strong moral reason for doing so. And we do. It is the reduction of human suffering. I am an unapologetic human supremacist. Just as I take animal and vegetable life to survive, I take it to promote the quality of that life.

Metzinger: Could we build a machine that dreams but never wakes? Are there animals that dream but do not enjoy waking consciousness?

Hobson: Again, it’s easier to answer the second part of the question. Given the limitations to scientific knowledge I’ve emphasized, the answer is no. If an animal can activate its brain in sleep, it has that capacity in waking also. So it stands to reason—but it is only reason—that animals that have (necessarily limited) dream consciousness also enjoy consciousness in waking. As for the first part, a dream machine can already be designed, but there is a state-of-the-art limitation that cripples the program. That limitation is the problem of generating linguistic statements from a biographical database. The last time I consulted with language specialist Roger Shanks, he told me that this crucial piece was still missing from the AI (artificial intelligence) puzzle. Activating perception and emotion modules poses no problem, and making them responsive to or independent of input and output can be done, as John Antrobus of CUNY has already shown. Any dream machine that would now be designed would likely have a wake-state mode of operation, because we’re interested in the similarities and differences between those two states and how they are generated. But it’s theoretically possible to develop a dream-only machine.

The fact that—as far as we know—evolution has not yet produced any dream-only animals suggests a deep meaningful and functional link between the waking and dreaming states of consciousness and brain activity. It is possible to argue, as I’ve already pointed out, that the brain is activated offline to benefit the brain online and vice versa, without postulating a causal link between the conscious awareness of the two states.

Metzinger: Cultural as opposed to biological evolution certainly gives a place to dream content, but whether that place is truly adaptive is still questionable.

Hobson: Many cultures have accorded prophetic meaning to dreams. The widely shared view of all such prophets is that dreaming is a message, in code, from important external or internal agents and needs decoding. Such decoding is seen by the practicing cultures as not only valid but also determinant of important conscious personal and political decisions. The dream sorcerers helped kings decide whether or not to go to war. Should modern psychoanalysts help individuals decide, say, whether to pursue a relationship further or not, based on the patients’ dreams?

One problem with this approach is the religious belief that there is some hidden truth that only dreams can reveal. Thus, one mystery, dreaming, is used to explain another, decision making. There is no evidence that this belief is justified. As Adolf Grünbaum has shown in his discussion of the Tally Argument, customer satisfaction cannot be used as a scientific warrant for the truth-value of a prophetic assertion—or a dream-interpretation scheme.

It might well be that dreaming reveals one’s cognitive repertoires in dealing with emotion, but that is not particularly difficult to discern in waking. The stronger claim, by psychoanalysis, that dream interpretation reveals hiddenlinks between cognition and emotion, has no scientific proof whatsoever.

Metzinger: I am particularly interested in the transition between ordinary dreams and lucid dreams. What are the necessary and sufficient conditions in the brain for lucidity to occur? What exactly is the role of the dorsolateral prefrontal cortex?

Hobson: The occasional awareness that one is in fact dreaming is an extremely informative detail of modern dream science. The fact that such insight can be cultivated thickens the plot considerably. Taken together, the data suggest that the conscious state accompanying brain activation in sleep is both plastic and causal. It is plastic because self-reflective awareness occasionally does arise spontaneously, and because with practice its incidence—and its power—can be increased. It is causal because lucidity can be amplified to command scene changes in dreams and even to command awakening, the better to remember, and enjoy, occasional dream-plot control. My position about lucidity is that it is real, it is powerful, and it is informative.

With respect to the third point, we already know, thanks to Stephen LaBerge, that sleep lucidity occurs in REM sleep, and we can predict that during lucid REM sleep dreaming, the dorsolateral prefrontal cortex, or DLPFC, which is selectively deactivated, may become reactivated so that the ponto-thalamical show of dreams comes under conscious control. I believe that this hypothesis, which is testable, contains the answer to many fundamental neurobiological and philosophical questions, such as the relationship of brain activity to consciousness and the causality of consciousness—free will.

If, as I predict, the DLPFC does reactivate during lucid dreaming while the ponto-thalamocortical dream show continues, then Daniel Dennett’s despised Cartesian theater does exist. One part of the brain—the seat of the executive ego—wakes up and watches, or even directs, the dream show thrown up on the consciousness screen by the activation of the pons, thalamus, cortex, and limbic system. Eat your heart out, Daniel Dennett!

The evanescence and fragility of the lucid dream state testify to its unlikelihood and its nonadaptive nature. The lucid dream also demands the special attention that all such revelatory rarities deserve. Unfortunately, it is unlikely to get that attention. The reason is that the experiments will be difficult to perform and expensive to underwrite. This would be a barrier to many more trivial exercises in cognitive neuroscience, but lucid dreaming has a bad name because (a) many scientists still do not believe it is real, (b) many do not trust LaBerge’s data about its occurrence in REM sleep, and (c) many will not go near the lucid-dream problem, because they fear being labeled as cranks or nuts! You, Thomas Metzinger, should easily be able to understand this fear.

Metzinger: Well, I certainly know what you’re talking about. The right strategy would be not to declare such areas taboo but to invade them with open-minded, unbiased scientific rationality. The problem in the background, of course, is that if we want to be realistic, we also have to admit that the newly emerging field of consciousness research is not populated by philosophical saints interested in the pursuit of self-knowledge as such. It is driven at least as strongly by what I sometimes call the Teflon-coated Darwin machines of Academia—brute individual career interests. Scientists, of course, are self-sustaining, risk-avoiding Ego Machines as well. Sad to say, the field of lucid-dream research is not moving well at this time.

Hobson: In my opinion—which is not widely shared, even by Thomas Metzinger—we need to work on a science of subjectivity. In order to be able to utilize first-person data, we need to be both cautious and versatile. Reports of conscious experience must be collected from many individuals in many states. These reports must be rigorously quantified, and the states with which they are associated must be objectified. The brain states must be more fully characterized using a full panoply of techniques, including PET and MRI in humans, cellular and molecular probes in animals, behavioral tests in humans, and more.

Who will do all these things? As far as I know, I’m the only person in the world even to have tried. I say this with all due modesty and even sincere self-deprecation. I am proud of my accomplishments, but I can easily understand the criticism of my work as a fool’s errand. At the root, the approach I advocate is in the service of the emergentist hypothesis of great scientists like Roger Sperry and great philosophers like William James. Such thinkers are few and far between.

More common, and far more handsomely rewarded, are those who dig deep discovering molecular widgets within and between nerve cells. Such discoveries are truly wondrous, but they will never lead to an understanding of conscious experience. Interestingly, even such well-known colleagues as Sigmund Freud worked within this ill-fated reductionist paradigm. Here I use the term “reductionist” in its popular sense, implying eliminative materialism.

Metzinger: Why are you interested in philosophy? What contributions from the humanities are you looking for?

Hobson: I’m interested in philosophy because I believe it is a foundational discipline—along with psychology and physiology—of cognitive neuroscience as it tries to figure out how to study consciousness. I myself try to “do” philosophy, but I need help. That’s why I turn to people like you, Owen Flanagan, and even David Chalmers. In general, I get positive responses from philosophers. They’re genuinely interested in my efforts and they generously share their insights with me. You’re no exception.

Regarding the second part of your question, I want philosophers and other humanists to realize that the scientific study of brain-mind states constitutes one of the greatest challenges and opportunities to better understand ourselves that has ever been presented to us in our long intellectual history. There is room for many disciplines in this effort, which is as simple and broad-based as it is ambitious. To bring more coworkers up to speed is my own private goal. We need all the help we can get. I even believe that brain-mind science is one of the humanities.

Metzinger: So, today, is there still any point in psychoanalysis, or is it just a lot of hot air? What do you think of Solms’s arguments?

Hobson: Sigmund Freud was fifty-percent right and a hundred-percent wrong! So is Mark Solms, but for different reasons. Freud was right to be interested in dreams and what dreaming could tell us about the human psyche, and especially its emotional aspects. His dream theory is now obsolete, but its errors are still being promoted by such psychoanalysts as Mark Solms.

Here’s a checklist of Freudian hypotheses and the corresponding alternatives offered by modern neurobiology:

1. Instigation of Dreaming.

Freud: release of unconscious wishes.

Neurobiology: brain activation in sleep.

2. Characteristics of Dreaming.

a. Bizarreness.

Freud: disguise and censorship of unconscious wishes.

Neurobiology: chaotic, bottom-up activation processes.

b. Strong emotion.

Freud: Can’t explain that one!

Neurobiology: selective activation of limbic lobe.

c. Amnesia.

Freud: repression.

Neurobiology: aminergic demodulation.

d. Hallucinations.

Freud: regression to the sensory side.

Neurobiology: activation of REMs and PGO waves.

e. Delusion, loss of self-reflective awareness.

Freud: ego dissolution.

Neurobiology: selective deactivation of the dorsolateral prefrontal cortex.

3. Function of Dreaming.

Freud: guardian of sleep.

Neurobiology: epiphenomenon, but REM sleep essential to life via enhancement of thermoregulatory and immune functions.

As we say in America, “Ya pays your money and ya takes your choice!” I choose neurobiology. How about you? As for Solms, he is nothing but a very smart psychoanalyst who wants to save Freud from the ashbin. His arguments, based on his important neuropsychological work, are weak. He has given up on disguise/censorship, but wants to resuscitate wish fulfillment. While it is true that dreams often do represent our desires, they are rarely truly unconscious, and dreams also represent our fears, a fact Freud could never explain. So what is left after Solms has given up disguise/censorship and only weakly defined wish fulfillment? Not much!

Solms attacks my activation-synthesis hypothesis of dreaming because of the observed dissociation between REM sleep and dreaming. As I have repeatedly pointed out, the correlation between REM sleep and dreaming is quantitative, not qualitative. The brain begins to shift from waking to REM sleep as soon as sleep begins. This means that the probability of dreaming begins to rise at sleep onset, to persist even in deep NREM sleep, when brain activation is still at eighty percent of waking levels, and to increase to its peak in REM sleep.

Why, then, do I say that Freud and Solms are fifty-percent right? Because dreams are not entirely nonsensical. They do make salient interrogations between emotions and cognition. Hence, they are worth reporting, discussing, and even interpreting, in terms of what they tell us about our emotions and how they influence our thoughts and our behavior. But they do this directly and openly, not via the symbolic transformation of forbidden wishes from the unconscious.

The good news is that you don’t have to pay—or even leave your house—if you want to use dreams to explore your emotional life. You need only to pay attention, keep a journal, and reflect on the messages from your emotional brain, the limbic lobe. If you’re a scientist, like me, you can do much more. You can use dreams and dreaming to build a new theory of consciousness.