INTRODUCTION: THE STUFF OF THOUGHT

Deep inside the Lascaux cave, past the world-renowned Great Hall of the Bulls, where Paleolithic artists painted a colorful menagerie of horses, deer, and bulls, starts a lesser-known corridor known as the Apse. There, at the bottom of a sixteen-foot pit, next to fine drawings of a wounded bison and a rhinoceros, lies one of the rare depictions of a human being in prehistoric art (figure 1). The man is lying flat on his back, palms up and arms extended. Next to him stands a bird perched on a stick. Nearby lies a broken spear that was probably used to disembowel the bison, whose intestines are hanging out.

FIGURE 1. The mind may fly while the body is inert. In this prehistoric drawing, dated approximately 18,000 years ago, a man lies supine. He is probably asleep and dreaming, as hinted by his strong erection, characteristic of the phase of rapid-eye-movement sleep, during which dreams are most vivid. Next to him, the artist painted a disemboweled bison and a bird. According to the sleep researcher Michel Jouvet, this may be one of the first depictions of a dreamer and his dream. In many cultures, the bird symbolizes the mind’s ability to fly away during dreams—a premonition of dualism, the misguided intuition that thoughts belong to a different realm from the body.

The person is clearly a man, for his penis is fully erect. And this, according to the sleep researcher Michel Jouvet, illuminates the drawing’s meaning: it depicts a dreamer and his dream.1 As Jouvet and his team discovered, dreaming occurs primarily during a specific phase of sleep, which they dubbed “paradoxical” because it does not look like sleep; during this period, the brain is almost as active as it is in wakefulness, and the eyes ceaselessly move around. In males, this phase is invariably accompanied by a strong erection (even when the dream is devoid of sexual content). Although this weird physiological fact became known to science only in the twentieth century, Jouvet wittily remarks that our ancestors would easily have noticed it. And the bird seems the most natural metaphor for the dreamer’s soul: during dreams, the mind flies to distant places and ancient times, free as a sparrow.

This idea might seem fanciful were it not for the remarkable recurrence of imagery of sleep, birds, souls, and erections in the art and symbolism of all sorts of cultures. In ancient Egypt, a human-headed bird, often depicted with an erect phallus, symbolized the Ba, the immaterial soul. Within every human being, it was said, dwelled an immortal Ba that upon death took flight to seek the afterworld. A conventional depiction of the great god Osiris, eerily similar to Lascaux’s Apse painting, shows him lying on his back, penis erect, while Isis the owl hovers over his body, taking his sperm to engender Horus. In the Upanishads, the Hindu sacred texts, the soul is similarly depicted as a dove that flies away at death and may come back as a spirit. Centuries later doves and other white-winged birds came to symbolize the Christian soul, the Holy Spirit, and the visiting angels. From the Egyptian phoenix, symbol of resurrection, to the Finnish Sielulintu, the soul bird that delivers a psyche to newborn babies and takes it away from the dying, flying spirits appear as a universal metaphor for the autonomous mind.

Behind the bird allegory stands an intuition: the stuff of our thoughts differs radically from the lowly matter that shapes our bodies. During dreams, while the body lies still, thoughts wander into the remote realms of imagination and memory. Could there be a better proof that mental activity cannot be reduced to the material world? That the mind is made of a distinct stuff? How could the free-flying mind ever have arisen from a down-to-earth brain?

Descartes’s Challenge

The idea that the mind belongs to a separate realm, distinct from the body, was theorized early on, in major philosophical texts such Plato’s Phaedo (fourth century BC) and Thomas Aquinas’s Summa theologica (1265–74), a foundational text for the Christian view of the soul. But it was the French philosopher René Descartes (1596–1650) who explicitly stated what is now known as dualism: the thesis that the conscious mind is made of a nonmaterial substance that eludes the normal laws of physics.

Ridiculing Descartes has become fashionable in neuroscience. Following the publication of Antonio Damasio’s best-selling book Descartes’ Error in 1994,2 many contemporary textbooks on consciousness have started out by bashing Descartes for allegedly setting neuroscience research years behind. The truth, however, is that Descartes was a pioneering scientist and fundamentally a reductionist whose mechanical analysis of the human mind, well in advance of his time, was the first exercise in synthetic biology and theoretical modeling. Descartes’s dualism was no whim of the moment—it was based on a logical argument that asserted the impossibility of a machine ever mimicking the freedom of the conscious mind.

The founding father of modern psychology, William James, acknowledges our debt: “To Descartes belongs the credit of having first been bold enough to conceive of a completely self-sufficing nervous mechanism which should be able to perform complicated and apparently intelligent acts.”3 Indeed, in visionary volumes called Description of the Human Body, Passions of the Soul, and L’homme (Man), Descartes presented a resolutely mechanical perspective on the inner operation of the body. We are sophisticated automata, wrote this bold philosopher. Our bodies and brains literally act as a collection of “organs”: musical instruments comparable to those found in the churches of his time, with massive bellows forcing a special fluid called “animal spirits” into reservoirs, then a broad variety of pipes, whose combinations generate all the rhythms and music of our actions.

I desire that you consider that all the functions that I have attributed to this machine, such as the digestion of food, the beating of the heart and the arteries, the nourishment and growth of the bodily parts, respiration, waking and sleeping; the reception of light, sounds, odours, smells, heat, and other such qualities by the external sense organs; the impression of the ideas of them in the organ of common sense and the imagination, the retention or imprint of these ideas in the memory; the internal movements of the appetites and the passions; and finally the external movements of all the bodily parts that so aptly follow both the actions of objects presented to the senses. . . . These functions follow in this machine simply from the disposition of the organs as wholly naturally as the movements of a clock or other automaton follow from the disposition of its counterweights and wheels.4

Descartes’s hydraulic brain had no difficulty moving his hand toward an object. The object’s visual features, impinging on the inner surface of the eye, activated a specific set of pipes. An inner decision-making system that was located in the pineal gland then leaned in a certain direction, thus sending the spirits flowing, to cause precisely the appropriate movement of the limbs (figure 2). Memory corresponded to the selective reinforcement of some of these pathways—an insightful anticipation of the contemporary idea that learning relies on changes in the brain’s connections (“neurons that fire together wire together”). Descartes even presented an explicit mechanical model of sleep, which he theorized as a reduced pressure of the spirits. When the source of animal spirits was abundant, it circulated through every nerve, and this pressurized machine, ready to respond to any stimulation, provided an accurate model of the wake state. When the pressure weakened, making the lowly spirits capable of moving only a few threads, the person fell asleep.

FIGURE 2. René Descartes’s theory of the nervous system stopped short of a fully materialistic conception of thought. In L’homme, published posthumously in 1664, Descartes foresaw that vision and action could result from a proper arrangement of the connections between the eye, the pineal gland inside the brain, and the arm muscles. He envisaged memory as the selective reinforcement of these pathways, like the punching of holes in cloth. Even fluctuations in consciousness could be explained by variations in the pressure of the animal spirits that moved the pineal gland: high pressure led to wakefulness, low pressure to sleep. In spite of this mechanistic stance, Descartes believed that the mind and the body were made of different kinds of stuff that interacted through the pineal gland.

Descartes concluded with a lyrical appeal to materialism—which was quite unexpected, coming from the pen of the founder of substance dualism:

To explain these functions, then, it is not necessary to conceive of any vegetative or sensitive soul, or any other principle of movement or life, other than its blood and its spirits which are agitated by the heat of the fire that burns continuously in its heart, and which is of the same nature as those fires that occur in inanimate bodies.

Why, then, did Descartes affirm the existence of an immaterial soul? Because he realized that his mechanical model failed to provide a materialist solution for the higher-level abilities of the human mind.5 Two major mental functions seemed to lie forever beyond the capacity of his bodily machine. The first was the capacity to report its thoughts using speech. Descartes could not see how a machine might ever “use words or other signs by composing them, as we do to declare our thoughts to others.” Reflexive cries posed no problem, as a machine could always be wired to emit specific sounds in response to a specific input; but how could a machine ever respond to a query, “as even the dumbest person can”?

Flexible reasoning was the second problematic mental function. A machine is a fixed contraption that can only act rigidly, “according to the disposition of its organs.” How could it ever generate an infinite variety of thoughts? “It must be morally impossible,” our philosopher concluded, “that there should exist in any machine a diversity of organs sufficient to enable it to act in all the occurrences of life, in the way in which our reason enables us to act.”

Descartes’s challenges to materialism stand to this very day. How could a machine like the brain ever express itself verbally, with all the subtleties of human language, and reflect upon its own mental states? And how might it make rational decisions in a flexible manner? Any science of consciousness must address these key issues.

The Last Problem

As humans, we can identify galaxies light years away, study particles smaller than an atom. But we still haven’t unlocked the mystery of the three pounds of matter that sits between our ears.

—Barack Obama announcing the BRAIN initiative (April 2, 2013)

Thanks to Euclid, Karl Friedrich Gauss, and Albert Einstein, we possess a reasonable understanding of the mathematical principles that govern the physical world. Standing as we do on the shoulders of such giants as Isaac Newton and Edwin Hubble, we understand that our earth is just a speck of dust in one of a billion galaxies that originated from a primeval explosion, the big bang. And Charles Darwin, Louis Pasteur, James Watson, and Francis Crick showed us that life is made of billions of evolved chemical reactions—just plain physics.

Only the story of the emergence of consciousness seems to remain in medieval darkness. How do I think? What is this “I” that seems to be doing the thinking? Would I be different if I had been born at a different time, in another place, or in another body? Where do I go when I fall asleep, and dream, and die? Does it all arise from my brain? Or am I in part a spirit, made of distinct stuff of thought?

These vexing questions have perplexed many a bright mind. Writing in 1580, the French humanist Michel de Montaigne, in one of his famous essays, lamented that he could find no coherence in what past thinkers had written about the nature of the soul—they all disagreed, both on its nature and on its seat within the body: “Hippocrates and Hierophilus lodge it in the ventricle of the brain; Democritus and Aristotle, throughout the body, Epicurus in the stomach, the Stoics within and around the heart, Empedocles, in the blood; Galen thought that each part of the body had its own soul; Strato lodged it between the eyebrows.”6

Throughout the nineteenth and twentieth centuries, the question of consciousness lay outside the boundaries of normal science. It was a fuzzy, ill-defined domain whose subjectivity put it forever beyond the reach of objective experimentation. For many years, no serious researcher would touch the problem: speculating about consciousness was a tolerated hobby for the aging scientist. In his textbook Psychology, the Science of Mental Life (1962), George Miller, the founding father of cognitive psychology, proposed an official ban: “Consciousness is a word worn smooth by a million tongues. . . . Maybe we should ban the word for a decade or two until we can develop more precise terms for the several uses which ‘consciousness’ now obscures.”

And banned it was. When I was a student in the late 1980s, I was surprised to discover that, during lab meetings, we were not allowed to use the C-word. We all studied consciousness in one way or another, of course, by asking human subjects to categorize what they had seen or to form mental images in darkness, but the word itself remained taboo: no serious scientific publication used it. Even when experimenters flashed brief pictures at the threshold of participants’ conscious perception, they did not care to report whether the participants saw the stimuli or not. With a few major exceptions,7 the general feeling was that using the term consciousness added nothing of value to psychological science. In the emerging positive science of cognition, mental operations were to be solely described in terms of the processing of information and its molecular and neuronal implementation. Consciousness was ill defined, unnecessary, and passé.

And then in the late 1980s everything changed. Today the problem of consciousness is at the forefront of neuroscience research. It is an exciting field with its own scientific societies and journals. And it is beginning to address Descartes’s major challenges, including how our brain generates a subjective perspective that we can flexibly use and report to others. This book tells the story of how the tables have turned.

Cracking Consciousness

In the past twenty years, the fields of cognitive science, neurophysiology, and brain imaging have mounted a solid empirical attack on consciousness. As a result, the problem has lost its speculative status and become an issue of experimental ingenuity.

In this book, I will review in great detail the strategy that has turned a philosophical mystery into a laboratory phenomenon. Three fundamental ingredients have made this transformation possible: the articulation of a better definition of consciousness; the discovery that consciousness can be experimentally manipulated; and a new respect for subjective phenomena.

The word consciousness, as we use it in everyday speech, is loaded with fuzzy meanings, covering a broad range of complex phenomena. Our first task, then, will be to bring order to this confused state of affairs. We will have to narrow our subject matter to a definite point that can be subjected to precise experiments. As we will see, the contemporary science of consciousness distinguishes a minimum of three concepts: vigilance—the state of wakefulness, which varies when we fall asleep or wake up; attention—the focusing of our mental resources onto a specific piece of information; and conscious access—the fact that some of the attended information eventually enters our awareness and becomes reportable to others.

What counts as genuine consciousness, I will argue, is conscious access—the simple fact that usually, whenever we are awake, whatever we decide to focus on may become conscious. Neither vigilance nor attention alone is sufficient. When we are fully awake and attentive, sometimes we can see an object and describe our perception to others, but sometimes we cannot—perhaps the object was too faint, or it was flashed too briefly to be visible. In the first case, we are said to enjoy conscious access, and in the second we are not (and yet as we shall see, our brain may be processing the information unconsciously).

In the new science of consciousness, conscious access is a well-defined phenomenon, distinct from vigilance and attention. Furthermore, it can be easily studied in the laboratory. We now know of dozens of ways in which a stimulus can cross the border between unperceived and perceived, between invisible and visible, allowing us to probe what this crossing changes in the brain.

Conscious access is also the gateway to more complex forms of conscious experience. In everyday language, we often conflate our consciousness with our sense of self—how the brain creates a point of view, an “I” that looks at its surroundings from a specific vantage point. Consciousness can also be recursive: our “I” can look down at itself, comment on its own performance, and even know when it does not know something. The good news is that even these higher-order meanings of consciousness are no longer inaccessible to experimentation. In our laboratories, we have learned to quantify what the “I” feels and reports, both about the external environment and about itself. We can even manipulate the sense of self, so that people may have an out-of-body experience while they lie inside a magnetic resonance imager.

Some philosophers still think that none of the above ideas will suffice to solve the problem. The heart of the problem, they believe, lies in another sense of consciousness, which they call “phenomenal awareness”: the intuitive feeling, present in all of us, that our internal experiences possess exclusive qualities, unique qualia such as the exquisite sharpness of tooth pain or the inimitable greenness of a fresh leaf. These inner qualities, they argue, can never be reduced to a scientific neuronal description; by nature, they are personal and subjective, and thus they defy any exhaustive verbal communication to others. But I disagree, and I will argue that the notion of a phenomenal consciousness that is distinct from conscious access is highly misleading and leads down a slippery slope to dualism. We should start simple and first study conscious access. Once we clarify how any piece of sensory information can gain access to our mind and become reportable, then the insurmountable problem of our ineffable experiences will disappear.

To See or Not to See

Conscious access is deceptively trivial: we lay our eyes on an object, and seemingly immediately, we become aware of its shape, color, and identity. Behind our perceptual awareness, however, lies an intricate avalanche of brain activity that involves billions of visual neurons and that may take nearly half a second to complete before consciousness kicks in. How can we analyze this long processing chain? How can we tell which part corresponds to purely unconscious and automatic operations, and which part leads to our conscious sense of seeing?

This is where the second ingredient of the modern science of consciousness kicks in: we now have a strong experimental handle on the mechanisms of conscious perception. In the past twenty years, cognitive scientists have discovered an amazing variety of ways to manipulate consciousness. Even a minuscule change in experimental design can cause us to see or not to see. We can flash a word so briefly that viewers will fail to notice it. We can create a carefully cluttered visual scene, in which one item remains wholly invisible to a participant because the other items always win out in the inner competition for conscious perception. We can also distract your attention: as any magician knows, even an obvious gesture can become utterly invisible if the watcher’s mind is drawn to another train of thought. And we can even let your brain do the magic: when two distinct images are presented to your two eyes, the brain will spontaneously oscillate and let you see one picture, then the other, but never both at the same time.

The perceived image, the one that makes it into awareness, and the losing image, which vanishes into unconscious oblivion, may differ minimally on the input side. But within the brain, this difference must be amplified, because ultimately you can speak about one but not about the other. Figuring out exactly where and when this amplification occurs is the object of the new science of consciousness.

The experimental strategy of creating a minimal contrast between conscious and unconscious perception was the key idea that cracked wide open the doors to the supposedly inaccessible sanctuary of consciousness.8 Over the years, we discovered many well-matched experimental contrasts in which one condition led to conscious perception while the other did not. The daunting problem of consciousness was reduced to the experimental issue of deciphering the brain mechanisms that distinguish two sets of trials—a much more tractable problem.

Turning Subjectivity into a Science

This research strategy was simple enough, yet it relied on a controversial step, one that I personally view as the third key ingredient to the new science of consciousness: taking subjective reports seriously. It was not enough to present people with two types of visual stimuli; as experimenters, we had to carefully record what they thought of them. The participant’s introspection was crucial: it defined the very phenomenon that we aimed to study. If the experimenter could see an image but the subject denied seeing it, then it was the latter response that counted—the image had to be scored as invisible. Thus, psychologists were forced to find new ways of monitoring subjective introspection, as accurately as possible.

This emphasis on the subjective has been a revolution for psychology. At the beginning of the twentieth century, behaviorists such as John Broadus Watson (1878–1958) had forcefully ousted introspection from the science of psychology:

Psychology as the behaviorist views it is a purely objective experimental branch of natural science. Its theoretical goal is the prediction and control of behaviour. Introspection forms no essential part of its methods, nor is the scientific value of its data dependent upon the readiness with which they lend themselves to interpretation in terms of consciousness.9

Although behaviorism itself was also eventually rejected, it left a lasting mark: throughout the twentieth century, any recourse to introspection remained highly suspicious in psychology. However, I will argue that this dogmatic position is dead wrong. It conflates two distinct issues: introspection as a research method, and introspection as raw data. As a research method, introspection cannot be trusted.10 Obviously, we cannot count on naïve human subjects to tell us how their mind works; otherwise our science would be too easy. And we should not take their subjective experiences too literally, as when they claim to have had an out-of-body experience and flown to the ceiling, or to have met their dead grandmother in a dream. But in a sense, even such bizarre introspections must be trusted: unless the subject is lying, they correspond to genuine mental events that beg for an explanation.

The correct perspective is to think of subjective reports as raw data.11 A person who claims to have had an out-of-body experience genuinely feels dragged to the ceiling, and we will have no science of consciousness unless we seriously address why such feelings occur. In fact, the new science of consciousness makes an enormous use of purely subjective phenomena, such as visual illusions, misperceived pictures, psychiatric delusions, and other figments of the imagination. Only these events allow us to distinguish objective physical stimulation from subjective perception, and therefore to search for brain correlates of the latter rather than the former. As consciousness scientists, we are never as pleased as when we discover a new visual display that can be subjectively either seen or missed, or a sound that is sometimes reported as audible and sometimes as inaudible. As long as we carefully record, on every trial, what our participants feel, we are in business, because then we can sort the trials into conscious and unconscious ones and search for brain activity patterns that separate them.

Signatures of Conscious Thoughts

These three ingredients—focusing on conscious access, manipulating conscious perception, and carefully recording introspection—have transformed the study of consciousness into a normal experimental science. We can probe the extent to which a picture that people claim not to have seen is in fact processed by the brain. As we will discover, a staggering amount of unconscious processing occurs beneath the surface of our conscious mind. Research using subliminal images has provided a strong platform to study the brain mechanisms of conscious experience. Modern brain imaging methods have given us a means of investigating how far an unconscious stimulus can travel in the brain, and exactly where it stops, thus defining what patterns of neural activity are exclusively associated with conscious processing.

For fifteen years now, my research team has been using every tool at its disposal, from functional magnetic resonance imaging (fMRI), to electro- and magnetoencephalography, and even electrodes inserted deep in the human brain, to try to identify the cerebral underpinnings of consciousness. Like many other laboratories throughout the world, ours is engaged in a systematic experimental search for patterns of brain activity that appear if and only if the scanned person is having a conscious experience—what I call the “signatures of consciousness.” And our search has been successful. In one experiment after another, the same signatures show up: several markers of brain activity change massively whenever a person becomes aware of a picture, a word, a digit, or a sound. These signatures are remarkably stable and can be observed in a great variety of visual, auditory, tactile, and cognitive stimulations.

The empirical discovery of reproducible signatures of consciousness, which are present in all conscious humans, is only a first step. We need to work on the theoretical end as well: How do these signatures originate? Why do they index a conscious brain? Why does only a certain type of brain state cause an inner conscious experience? Today no scientist can claim to have solved these problems, but we do have some strong and testable hypotheses. My collaborators and I have elaborated a theory that we call the “global neuronal workspace.” We propose that consciousness is global information broadcasting within the cortex: it arises from a neuronal network whose raison d’être is the massive sharing of pertinent information throughout the brain.

The philosopher Daniel Dennett aptly calls this idea “fame in the brain.” Thanks to the global neuronal workspace, we can keep in mind any idea that makes a strong imprint on us, for however long we choose, and make sure that it gets incorporated into our future plans, whatever they might be. Thus consciousness has a precise role to play in the computational economy of the brain—it selects, amplifies, and propagates relevant thoughts.

What circuit is responsible for this broadcasting function of consciousness? We believe that a special set of neurons diffuses conscious messages throughout the brain: giant cells whose long axons crisscross the cortex, interconnecting it into an integrated whole. Computer simulations of this architecture have reproduced our main experimental findings. When enough brain regions agree about the importance of incoming sensory information, they synchronize into a large-scale state of global communication. A broad network ignites into a burst of high-level activation—and the nature of this ignition explains our empirical signatures of consciousness.

Although unconscious processing can be deep, conscious access adds an additional layer of functionality. The broadcasting function of consciousness allows us to perform uniquely powerful operations. The global neuronal workspace opens up an internal space for thought experiments, purely mental operations that can be detached from the external world. Thanks to it, we can keep important data in mind for an arbitrarily long duration. We can pass it on to any other arbitrary mental process, thus granting our brains the kind of flexibility that Descartes was looking for. Once information is conscious, it can enter into a long series of arbitrary operations—it is no longer processed in a reflexive manner but can be pondered and reoriented at will. And thanks to a connection to language areas, we can report it to others.

Equally fundamental to the global neuronal workspace is its autonomy. Recent studies have revealed that the brain is the seat of intense spontaneous activity. It is constantly traversed by global patterns of internal activity that originate not from the external world but from within, from the neurons’ peculiar capacity to self-activate in a partly random fashion. As a result, and quite opposite to Descartes’s organ metaphor, our global neuronal workspace does not operate in an input-output manner, waiting to be stimulated before producing its outputs. On the contrary, even in full darkness, it ceaselessly broadcasts global patterns of neural activity, causing what William James called the “stream of consciousness”—an uninterrupted flow of loosely connected thoughts, primarily shaped by our current goals and only occasionally seeking information in the senses. René Descartes could not have imagined a machine of this sort, where intentions, thoughts, and plans continually pop up to shape our behavior. The outcome, I argue, is a “free-willing” machine that resolves Descartes’s challenge and begins to look like a good model for consciousness.

The Future of Consciousness

Our understanding of consciousness remains rudimentary. What does the future hold in store? At the end of this book, we will return to the deep philosophical questions, but with better scientific answers. There I will argue that our growing understanding of consciousness will help us not only resolve some of our deepest interrogations about ourselves but also face difficult societal decisions and even develop new technologies that mimic the computational power of the human mind.

To be sure, many details remain to be nailed down, but the science of consciousness is already more than a mere hypothesis. Medical applications now lie within our grasp. In countless hospitals throughout the world, thousands of patients in a coma or a vegetative state lie in terrible isolation, motionless, speechless, their brains destroyed by a stroke, a car accident, or a transient deprivation of oxygen. Will they ever regain consciousness? Might some of them already be conscious but fully “locked in” and unable to let us know? Can we help them by turning our brain-imaging studies into a real-time monitor of conscious experience?

My laboratory is now designing powerful new tests that begin to reliably tell whether a person is or is not conscious. The availability of objective signatures of consciousness is already helping coma clinics worldwide and will soon also inform the related issue of whether and when infants are conscious. Although no science will ever turn an is into an ought, I am convinced that, once we manage to objectively determine whether subjective feelings are present in patients or in infants, we will make better ethical decisions.

Another fascinating application of the science of consciousness involves computing technologies. Will we ever be able to imitate brain circuits in silico? Is our current knowledge sufficient to build a conscious computer? If not, what would it take? As consciousness theory improves, it should become possible to create artificial architectures of electronic chips that mimic the operation of consciousness in real neurons and circuits. Will the next step be a machine that is aware of its own knowledge? Can we grant it a sense of self and even the experience of free will?

I now invite you to take a journey into the cutting-edge science of consciousness, a quest that will guarantee deeper meaning to the Greek motto “Know thyself.”