The Ego Tunnel: The Science of the Mind and the Myth of the Self

Book Review: The Ego Tunnel: The Science of the Mind and the Myth of the Self  

by Thomas Metzinger  (Basic Books 2009)

This ‘cutting edge’ book melds philosophy of consciousness with cognitive neuroscience. Metzinger is a philosopher. Though written for the lay reader, it is hard to follow at times. Quite a lot of concepts, relationships, and terminology are introduced. However, he does re-explain things enough so that the theory makes sense if one persists. He endeavors to cover the issues current in consciousness research and even offers interviews with various researchers and neuroscientists. Some of the subjects are emotions, dreaming and so-called astral phenomena, empathy, free will, AI, and the ethics of cultivating various states of consciousness. The starting point and the overriding theory throughout the book is that the realization that there is no entity that is us, as we normally perceive us – that our reality is a virtual reality based on a model of the self that evolved along with our human form of consciousness.

One of the key concepts, or metaphors, of the book is the Phenomenal Self-Model (PSM). He describes it with the help of a now famous experiment first done at the University of Pittsburgh, called the ‘rubber-hand illusion’. In the experiment the person sticks their hands through holes in a window in such a way that one hand is concealed and one sees an artificial hand in place of it. Repeatedly, including with the author, the experiment shows that when one thinks one is seeing the hand, though actually the virtual hand, one feels it as the hand and reacts as if it were the hand. What one actually feels is what he calls the ‘content’ of the phenomenal self-model. The word “phenomenal” refers to phenomenology which is concerned with the reporting of how things appear to one. For example, color only exists as it is created by our sensory apparatus – other organisms see something different. One might also use the analogy of the “Plato’s Cave” metaphor/parable – where one believes in a reality based only on what one has experienced. When one experiences beyond this, so too does the reality expand.

“The PSM of Homo sapiens is probably one of nature’s best inventions. It is an efficient way to allow a biological organism to consciously conceive of itself (and others) as a whole. Thus it enables the organism to interact with its internal world as well as with the external environment in an intelligent and holistic manner.”

He notes that animals, though having consciousness, have a different PSM than humans. Along with the rubber-hand illusion there are other phenomena that suggest the PSM: out-of body experiences (OBE) and artificial limb syndrome are among them. The body double, body image, or astral body of OBE is also the ‘content’ of the PSM. The ‘ego’ dwells in this body model. We perceive the parts of our body as belonging to us which strengthens this ego-illusion. Thus the ego, as it appears, exists in a ‘tunnel’. Our sense of self, says Metzinger, is “a form of conscious representational content.” The reason it is called a tunnel is that we only perceive a fraction of what is really there as our senses evolved merely to ensure our survival rather than to experience all that is possible to experience.

“Modern neuroscience has demonstrated that the content of our conscious experience is not only an internal construct but also an extremely selective way of representing information.”

Thus the concept of the “Ego Tunnel” is the central metaphor of the book. The PSM is the model. The ego, or “I” or self, is what is in the model. He also calls it a “self-model theory of subjectivity”. A key feature of the model is that it is transparent, so that we do not normally recognize it. It is a brain simulation. It is not reality but an image of reality. This configuration (ego in tunnel) is “evolved representational phenomena, a result of dynamical self-organization on many levels.” The idea of the ego tunnel is based on earlier ideas of the “reality tunnel” found in virtual reality research and promulgated by those he calls non-academic philosophers – namely Robert Anton Wilson and Timothy Leary. The basic idea is that we do not see reality as it is because we filter it (automatically and unconsciously) and so construct our own version, or model of reality.

He also talks about the search for a “neural correlate of consciousness” (NCC) and specifically a global neural correlate of consciousness. The NCC is defined as a “set of neurofunctional properties in your brain sufficient to bring about a conscious experience.” Correlation does not mean causation so discovering these correlates (ie. what happens in the brain during events) does not really explain consciousness. But such discovery can help with the future neuro-technology of inducing states of consciousness.

Consciousness is defined as “the appearance of a world.” Human consciousness developed special features so that we can model our models, be conscious of our consciousness, etc. These are “higher-order” levels of the PSM, he says. In considering the possible effects of knowledge of knowledge (ie. epistemology) he notes the extremes of ‘reductionism’ – reducing everything to one simplistic theory, and ‘mysterianism’ – favoring the impossibility of knowing much about consciousness. He does admit that there are other ways of knowing besides science.

Metzinger describes six problems that need to be solved in formulating a philosophical and neurological theory of consciousness. The first he calls the one-world problem: the unity of consciousness. The ‘one world problem; is described like this: “In order for a world to appear to us, it has to be one world first.” He gives examples of psychiatric disorders – sometimes due to brain lesions – where people can’t recognize what they are seeing, touching, cannot name their limbs, or cannot integrate seeing and hearing. Many of these disorders are versions of agnosia. Some neurologists speak of a “world-binding function” as in the dynamical core hypothesis. The neural properties underlying consciousness – the global NCC - may give us the sense of the “unity of consciousness.” This sense of the unity or coherence of consciousness may be virtual. It may be a result of the “global synchrony” of neurons making it seem as if what is in the cloud of consciousness (the dynamical core) is our subjective experience and what is not within the cloud is not part of our subjective experience. Here is an interesting quote but the context is a bit unclear. Regarding neurological experiments with “mathematical instruments” he notes that “they show us how self-organization in our brains strikes an optimal balance between integration and segregation, …” Consciousness may create the world in the manner of “context loops” where the immediate past conditions the present by creating a context for it. In visual perception, for instance, this may occur in the time frame of micro-seconds. The synchrony of neural responses may allow for the optimal balance between integration and segregation as mentioned above. Neuronal firings are often in synch and/or in rhythm which allows for information coding. Neuroscientists studying meditators have noted clear “global phase-synchrony” on EEG readings of those who are very experienced meditators. It is apparent that meditation can change the deep structure of consciousness. With synchrony, intensity of firing, rate of firing, etc – being the main information coding mechanisms, it can be seen that the unity of consciousness is involved with ‘time’.

The next problem is called – the now problem – subjective experience of time is a necessary part of consciousness. A major function of consciousness according to psychologist Richard Gregory is “flagging the dangerous present,” Our inner model of time is likely an adaptive mechanism. Our sense of presence is knowledge that can help us survive. We experience simultaneity, succession, and duration as aspects of time – our temporal gestalt. “Presence is a necessary condition for conscious experience.” We have a sense of presence but we are not really in touch with the presence because neural information processing takes time. Now is merely an appearance we experience as now. The ego is inseparable from here and now. Our sense of presence is our inner representation of time.

Next is – the reality problem. If the first two problems are solved we have a unified world and a model of the present moment. Our world-models and now-models are transparent, which is why we don’t experience them as representative models. The author talks about higher-order representations. For instance, a first-order process would be visual perception and a second-order process would be attentional processing of visual perception. While transparency may be caused by the speed of information-processing the author suggests that it has more to do with the speed of processing of different order processes relative to each other, ie. speed of visual perception relative to speed of attentional processing of visual perception. The bottom line is that the brain is invisible to itself. He suggests that the reason the model stayed transparent to us is that it would not have been cost-efficient in terms of metabolic price (a concept whereby new ways of knowing/doing cost energy in terms of sugar metabolism) to know that we were knowing, or to make images about images, also known as metarepresentation. He notes that “whatever appears to us – however it is mediated – appears as reality.”  The phenomenal transparency is our “invisible interface to reality.” Situations where the transparency, the transparent walls of the ego tunnel, disappear (in varying degrees), may include dreams, lucid dreams, OBEs, NDEs, optical illusions, visions, hypnagogic hallucinations, psychiatric maladies, and psychedelic drug experiences.

Next is – the ineffability problem – much of the details of consciousness can’t be recalled or communicated. The contents of consciousness can be ineffable in several ways – we may not have language to describe certain things; describing visual things to a blind person is not really effective; our ability to discriminate things like subtle differences in color hues exceeds our ability to form concepts about them. We can detect contrast in our sensory experiences but we generally cannot identify or recognize the same exact event without it being contrasted against the next value. We have limited perceptual memory. There are those that have finer degrees of discrimination to detect nuances – vintners, musicians, perfume designers. Conscious experience is so subtle that the nuances can hardly be communicated. One must experience them for oneself. The author suggests that one day neurobiological concepts may replace inadequate phenomenological concepts but it is yet unclear how this would occur.

Next is – the evolution problem – why and exactly how the human form of consciousness evolved is not known. Surely, it is built on previously evolved forms of animal consciousness, both sensory and emotional. There are many possible avenues that could have been taken to arrive at human consciousness. Many consciousness researchers agree that one of the main functions of consciousness is making information “globally available” to an organism. Attention may see with more detail than memory or conceptual thought – and so has a higher degree of availability. The author sees consciousness as a new kind of organ, a virtual organ. Other virtual organs may include feelings, episodic memory, and immune response. Global availability of information allows us to do many things:

“If you have a conscious, transparent world-model, you can, for the first time, directly compare what is actual with what is only possible, the actual world with simulated possible worlds you’ve designed in your mind.”

This functional property of comparing the actual to the possible, he says, is an evolutionary step upon which everything else rests. Offline simulations can plan online actions so this property allowed us to use episodic memory to plan actions. We became able to utilize the past and simulate the future in order to deal with the “dangerous present.” The transparency of the model, he says, allowed us to see the difference between the “real” world (though technically still a virtual world) and the many possible worlds of simulation.

Finally, there is – the who problem – “consciousness is always bound to an individual first-person perspective.” Subjectivity is vague. Just who is “I”? A severe psychiatric disorder called Cotard’s Syndrome is sometimes characterized by the person claiming that they don’t exist. Some famed mystics have also come to this conclusion based on their experiences of the loss of a sense of self. The sense of unity in mystical experience perhaps points to the illusion of the limited self within the tunnel. Metzinger speculates that the evolution of our immune systems first developed determinations on the cellular level as to what is us and what is not us, ie. the invader. That was perhaps the beginning of the conceptual self/world distinction. Our self-model was likely based on our world-model. Even so, the whole idea of the self and subjectivity is still difficult to comprehend.

An interview with German neurophysiologist Wolf Singer concludes the chapter where he talks about the “powerful self-organizing mechanisms of the brain.” Many of these mechanisms are controlled by “synchronized oscillatory discharges.” Precise synchronization of neuronal events may account for (the apparent?) unity of consciousness. Singer notes that a current big challenge is refining our knowledge of how information is encoded in the brain: rate of firing, amplitudes and duration of discharges, etc. Pattern searching through complex mathematical algorithms is a current method of searching for precise encoding mechanisms in order to search for the NCCs. He also suggests that future collaborations between cognitive neuroscientists and philosophers and others in the humanities will be required as more becomes known.

The body-mind nexus is next examined. Having a body gives us a sense of ownership of it. The interplay between vision, touch, and proprioception can be manipulated, as in the rubber-hand illusion, to alter the self-model. We can utilize our body-image self-model to better use tools, thus integrating vision, touch, and proprioception. The tool becomes part of the bodily self. It seems likely that the evolution of tool use may have co-evolved to some extent with the bodily self-model. Experiments have shown that monkeys, also tool users, also have this sort of bodily self-model.

“… the evolution of language, culture, and abstract thought might have been a process of “exaptation,” of using our body maps for new challenges and purposes.”

OBEs are the next subject. He talks about it as a taboo subject in philosophical circles when he was in school as materialism was in vogue and any talk about soul components was stifled. He suggests that the “soul” as a metaphysical concept may have arisen to from the phenomenological concept of the experienced OBE. He says that the soul is the OBE-PSM. He goes through several accounts of OBEs, several his own, and Susan Blackmore’s interesting theory about them – that they are a conscious simulation of the world from a third-person perspective that include a realistic body representation that we think is real. We notice the body, not as subject, but as object. The phenomena of “autoscopy,” of viewing one’s body from a distance, is very related. There are autoscopic  hallucinatory experiences of four main types: autoscopy, heautoscopy – changing bodies, OBE, and “feeling of a presence.” The phenomenology of OBEs varies, with different methods of separating from the body and speeds of doing so. Some sense themselves in another body while others feel they are bodiless. OBEs were inadvertently induced in an epileptic patient in 2002 by Olaf Blanke and colleagues, while using invasive neuro-imaging to try and find legions. Electrical stimulation of the brain (ESB) has also been found to induce the “feeling of a presence.” Blanke and colleagues note that two pathological conditions are necessary to form an OBE: 1) “disintegration on the level of the self-model brought about by a failure to bind proprioceptive, tactile, and visual information about one’s body”, and 2) “conflict between external, visual space and the internal frame of reference created by our vestibular organs.” We see the body in a different position than we feel it. The author describes some experiments done with Olaf and a colleague where he – the author – was the subject – utilizing virtual reality hardware to try and create a whole body version of the rubber-hand illusion. The results were not stellar but there was identification with a mannequin by many of the subjects in the manner of the rubber-hand illusion. The bodily self occupies space while the seeing self does not. They can be effectively separated, says the author. The seeing self can be turned off by closing the eyes, so is not the essence of selfhood. The author suggests that we don’t know, but the closest we can get to an essence of selfhood is: “location in space and time plus a transparent body image.” Attentional agency, or self-directed attention allows for global availability, and thus a sense of selfhood.

Conscious experience as the self-model is an interface, a control device, a virtual world that allows us better control of the outside world. The ego is a functional illusion. The human brain can be compared to a modern flight simulator – continually constructing and updating an internal model of external reality by filtering through the senses and past experience. The difference is that our brain as a simulator is updated at such fast rates and degrees of accuracy that we don’t recognize it as a model. The pilot of our simulator is the non-existent ego.

Experiments with “phantom limb” victims (amputees who can still sense their amputated limbs) have succeeded in getting them to feel a sense of movement in those limbs, but only when their eyes were open to the illusion given them. Since, coordinated bodily movements were needed for survival, the bodily self-model developed to increase speed and coordination – first among animals and then perhaps conceptually among humans.

There is another odd phenomenon, known as “Alien Hand syndrome” where one loses control of a limb. The hand seems to have a “will” of its own. The feeling of control, of self-willing actions, and of actions willed by another – varies in different psychiatric maladies. The author asks on an ultimate philosophical level- Are there willed actions or just events that occur? Our sense of ownership (of our body) supports our sense of self which then supports our sense of agency, ie. will. The author notes that: “the Ego is a neurocomputational device for appropriating and controlling the body – first the physical one and then the virtual one.” Attentional agency, or directing (willing) the attention is a result of the consolidation of the ego tunnel. Infants cannot do it – as their ego tunnel is not yet consolidated. In dreams there is often consciousness without attentional control. Drunkenness and senility may be other examples. Cognitive agency is selecting the contents of your mind – experiencing yourself as a thinker of thoughts – a la Descartes (pun intended). Moving the body, focusing attention, and selecting thoughts all involve effort, so agency is associated with effort.

Agency can be hallucinated as ESB experiments confirm. We can think we willed or intended an action when we didn’t. 1990’s trickery experiments by psychologists Wegman and Wheatley caused people to think they were willing actions when not and they concluded that the phenomenal experience of will consists of three principles:

“The principle of exclusivity holds that the subject’s thought should be the only introspectively available cause of action; the principle of consistency hold that the subjective intention should be consistent with the action; and the principle of priority holds that the thought should precede the action “in a timely manner””.

The sense of agency, the conscious experience of will, allows us to see ourselves as pursuing goals and leads to optimizing means to ends. It also allows us to know that others also have will of their own.

 There is a discussion on determinism and free will which the author agrees can be compatible. He says that determinism is favored by most modern professional philosophers yet our own phenomenology suggests free will or so it seems. Metzinger thinks that what science tells us or will tell us eventually about free will, will contradict what we feel subjectively about it. He does admit though that the phenomenology of will is not at all worked out. We assume that free will exists, treat one another as autonomous agents, and even base our legal system and societal rules on those assumptions. Dynamical self-organization according to the laws of physics is the mechanism of the brain and all man-made rules and assumptions of will may be illusory, or so neuroscience suggests.

In examining the dream state he notes an example where a person has a dream that they had an OBE and woke up and told someone about it – but it was a false awakening. This confounds the appearance/reality distinction. He suggests that it demonstrates that consciousness is never more than the appearance of a world. The dream tunnel also creates the appearance of a world but one without sensory-motor activity. Dreams are unstable and tend to lack focused attention. The sense of will is reduced. Emotions, however, can be enhanced in dreams. In lucid dreams we may become aware of the ego tunnel. We know that we are in a simulated world. Memory, a sense of agency, and sensory experience may remain intact in a lucid dream. In fact, the sense of will can be strong – as when I was lucid dreaming I was experimenting with the dream state willing different things – which is also a technique of Indo-Tibetan dream yoga practices. Dr. Stephen LaBerge of Stanford University was able to devise a means where lucid dreamers could signal with their eye movements that they were having a lucid dream. The author calls these rare examples of trans-tunnel communications. Psychiatrist Alan Hobson has worked with discovering a neural correlate of lucid dreaming. There is an interview with Hobson. He thinks that dreaming is “our subjective awareness of our dream activation in any state of sleep.” Activation is highest in REM sleep so that state is most associated with dreaming. Stroke and frontal lobotomy victims often report a cessation of dreaming depending on which parts of the brain were affected so dreaming, or dream recall, is dependent on coordination of brain parts. Asked about the evolutionary function of dreaming, Hobson first suggests that there may not be one. If dream recall was adaptive we would have more of it. Other possibilities of the function of dreams he suggests are “enhancement of motor learning, the regulation of dietary and thermal calories, and the improvement of immune functions.” Since mammals are homeothermic, brain activation during sleep helps to regulate brain temperature in species that face variances in temperature and environment. Hobson also notes that Freud’s psychoanalytic dream analysis and wish fulfillment functions have mostly been refuted by neuroscience.

The Ego Tunnel has not only an inward perspective but also a social aspect. The theory is that new layers to the self-model allowed the transition from biological to cultural evolution. The transition in the self-model from first-person singular (me) to first-person plural (we) allowed for the beginning of moral agency. The new discipline of “social neuroscience” deals with the social aspects of neurology. Neurologists first discovered a set of neurons that respond to visual perception of objects in our environment. These canonical neurons are related to the use of an inner motor vocabulary that deals with possible actions related to an object. These neurons, he says, are precursors to the self-model. Another related group of neurons, called mirror neurons, are activated when we observe behavior in others and match or emulate it. We map the actions of others onto our own model. We also note that others also have goals. Mirror neuron systems can also be activated for emotional states. This also gives a sort of neural correlate for empathy. We can simulate the physical and emotional states of others using our own self-model. The author thinks that the body-model was an adaptation and that social cognition, stemming from the precursor actions of mirror neurons, was an “exaptation”, a new use for the previously developed adaptation. Canonical neurons could be the link between motor activity and primitive inner conceptual language. Mirror neurons could be the precursor to communicating meaning to others. He speculates that the precursor to language was gestural communication rather than animal calls. Gestures communicate meaning better than calls. We understand gestures because we do them too, having learned them from mirroring or mimicking others. Interestingly, he notes, the word for grasping (at least in Latin) is related to the word for concepts – and even now we say that we “grasp a concept.” This is a possible relationship between motor-activity and understanding meaning that may hark back to the formulation or adding of a new layer to our self-model. There are clear neural relationships between movement of the hand/arm and the mouth. It seems sensible to conclude that body language is the foundation of verbal language. Mirror neurons may explain other phenomena as well – the infectiousness of yawning and of laughter, the group behaviors of schools of fish, flocks, of birds, herds, and even human social conformity. These can be seen as forms of synchronization with one another.

There is an interview with Vittorio Gallese, a neurophysiologist and an authority on mirror neurons. He expounds on his shared manifold hypothesis – which is our link with others in understanding them through observation. He calls this shared manifold an expanded idea of empathy. On the phenomenological level there is a sense of similarity and familiarity when we encounter others – that we are part of a shared social sphere. Functionally, there is “embodied simulation” of actions, emotions, and sensations observed in others. Subpersonally, there is the action of the mirroring neural circuits. Generally, when we are with others, we become “we-centric”, at least to some extent. Mirror neuron activity suggests that we empathize unconsciously and automatically by modeling others’ behavior. Such systems may have evolved through the need to accurately read the intentions of others to increase survivability. He mentions “orders of intentionality”, ie. “I know that you know that I know, etc.” We learn best by imitation which may well involve mirror neurons. Even repeatedly hearing the narration of a story may involve mirror neurons so they are activated in language processing. Imitation, and the empathizing the underlies it, is very important to cultures with strong oral traditions. Mirroring neural circuitry is likely being effected by new forms of media as well, from books to cinema to internet, says Gallese. He also notes that cognitive neuroscience and philosophy of mind deal with the same problems but utilize different approaches and so should be utilized in tandem.

Next, there is a discussion of Artificial Ego Machines, which refers to AI systems that simulate ego tunnels in some way. This part of the book I found to be speculative and not all that interesting, maybe because it is in its infancy. There are a lot of “what ifs” here. He examines the ethical implications of developing artificial ego machines as well – noting that we would be better off working on reducing human suffering first.

The final section on Consciousness Technologies is interesting. Ultimately, the Ego and its Tunnel is a model of a model of reality. There can be many ways to model reality. We can see the self as process – we can say we are “selfing”. He does seem adamant in concluding that there is no “ghost in the machine, no Ego in the Tunnel, no definable entity we can call self. There is only “dynamical self-organization.” He suggests that we cannot really understand consciousness without radically transforming our state of consciousness. He notes that:

“Obviously, the evolutionary process that created our bodies, our brains, and our conscious minds was not a goal-directed chain of events.”

He sees our evolutionary development as the result of “bottom-up self-organization.”

He does note that knowledge of the tunnel by groups of scientists is actually invading the tunnel. Consciousness of our consciousness is affecting our consciousness. The fact of mortality is an existential conflict. We are selfless ego machines longing for immortality. Are we the evolutionary products of goal-less chance?

The author suggests that a “philosophically motivate neuroanthropology” may gain in importance as neurological advances weaken traditional religious views. What is a human and what should a human become may be the key questions of this science. Will weakening of religion weaken ethics? I doubt it, as ideas like ethics and compassion seem intuitively important, regardless of religious dogma. The author talks of a “recursive” process that may alter the contents, structure, and function of our self-model:

“The world evolved world-modelers. Parts began to mirror the whole.”

“More important, the world evolved self-modelers who were able to form groups…… Through science, the dynamic processes of self-modeling and world-modeling were extended into the symbolic, the social, and the historical dimensions: We became rational theory-makers.”

The author considers the idea of consensual agreement on what constitutes desirable states of consciousness. Shamans, yogis, mystics, drug users, and others have explored consciousness through the means available: herbs, drugs, fasting, drumming, dancing, fear, pain, ecstasy, etc. Now we have at least crude neurological explanations for many of these states. With newer drugs and brain manipulation techniques (neurotechnology) we have opportunities for “cognitive enhancement.” I remember the beginnings of the “smart drug”, or “nootropics” scene in the 80’s but not much seems to have been done since then from my perspective – but maybe I’m wrong, maybe I just have not kept up. Persinger’s “temporal-lobe” theory of religious/mystical experience and personality shifts was an early attempt at finding neural correlates to these experiences. In a study from the journal Nature it was revealed that scientists that use cognitive enhancers, preferred methylphenidate (Ritalin). in order to stimulate focus, concentration, and memory. Others preferred modafinil. The author examines the ethical implications of cognitive enhancers and the possibilities that there may be tens of thousands of variant drugs on the markets, white and black, in the years to come. Drug policy will have to determine which brain-states will be illegal and which legal.

The internet is merging with our self-model. On-line addiction can be problematic but the trade-offs are acceptable. The internet, for all it is, is also a consciousness technology. Attention management is an issue we face. Attention is finite. We can only “attend” to so many things at once or in the space of a day. The author proposes childhood development in attention management through such forms as training in meditation. Development of a rational and ethical “consciousness culture” is another avenue he explores.

Great book. Difficult at times. Lots of terminology and abstract ideas.