Imagine you arrive at your doctor’s office for an appointment to discuss some recent chest pains. You undergo a series of blood tests and scans, and a week later you return to the clinic, where your doctor reviews the results with you. The condition seems serious, and she briskly recommends surgery for a heart bypass. When you ask your doctor why she is confident the procedure is necessary, she walks you through her thought process, including the possibility that she is wrong and what being wrong might entail, before reiterating her advice that you undergo the surgery. What would you do?
Now imagine that, after you undergo a series of blood tests and scans, the data is fed into an artificially intelligent assistant, which confidently states that the condition seems serious and it would be desirable if you had surgery for a heart bypass. When you ask your doctor whether this is really necessary, she can’t tell you; she doesn’t know why the recommendation has been made. All she can say is that, when fed the full range of test data, the AI has been highly accurate in the past, and that it would be wise to trust it and proceed with the surgery. What would you do?
In the first case, the answer probably seems obvious: if the doctor is confident and able to explain her reasons for being confident, you feel you should trust her advice. In the second, however, it may not be so clear. Many of us intuitively feel that if a person or a machine is going to be making high-stakes decisions on our behalf, we should be able to ask them to explain why they have come up with a particular answer. Many of our legal frameworks—those that ascribe liability and blame for errors—are based on the notion of being able to justify and defend what we did and why we did it. Without an explanation, we are left with blind trust—in each other, or in our machines. Ironically, some of the highest performing machine learning algorithms are often the least explainable. In contrast, humans are rapacious explainers of what we are doing and why, a capacity that depends on our ability to reflect on, think about, and know things about ourselves, including how we remember, perceive, decide, think, and feel.
Psychologists have a special name for this kind of self-awareness: metacognition—literally, the ability to think about our own thinking, from the Greek “meta” meaning “after” or “beyond.” Metacognition is a fragile, beautiful, and frankly bizarre feature of the human mind, one that has fascinated scientists and philosophers for centuries. In the biologist Carl Linnaeus’s famous 1735 book Systema Naturae, he carefully noted down the physical features of hundreds of species. But when it came to our genus, Homo, he was so captivated with humans’ ability for metacognition that he simply annotated his entry with the one-line Latin description “Nosce te ipsum”—those that know themselves.1
Self-awareness is a defining feature of human experience. Take a student, Jane, who is studying for an engineering exam. What might be going through her head? She is no doubt juggling a range of facts and formulas that she needs to master and understand. But she is also, perhaps without realizing it, figuring out how, when, and what to study. Which environment is better, a lively coffee shop or a quiet library? Does she learn best by rereading her notes or by practicing problem sets? Would it be better to shut the book on one topic and move onto another? Can she stop studying altogether and head out with friends?
Getting these decisions right is clearly critical for Jane’s chances of success. She would not want to fall into a trap of thinking she knows a topic well when she does not, or to place her trust in a dodgy study strategy. But no one is giving her the answers to these questions. Instead, she is relying on her awareness of how she learns.
Our powers of self-reflection do not lose their importance when we leave the classroom or the exam hall. Consider the experience of James Nestor, an author and free diver. In his book Deep, Nestor recounts how he traveled to coastal locations in Greece and the Bahamas to report on free-diving tournaments. At each tournament, there is only one goal: to dive deeper than all the other competitors, all on a single breath. To prove that they have reached a particular depth, the divers retrieve a tag with a number emblazoned on it. If they pass out after surfacing, the dive is declared null and void. To be successful, professional free divers must be acutely self-aware of their ability to reach a depth while avoiding injury or even death. Slight underconfidence will lead to underperformance, whereas slight overconfidence may be fatal. It’s telling that a large part of free divers’ training takes place on land, in psychological exploration of their underwater capacities and limitations.2
Or how about the case of Judith Keppel, one of the first contestants on the British TV game show Who Wants to Be a Millionaire? For each question, contestants are asked if they are sure they know the right answer and want to risk their existing winnings on the chance of a higher prize, or if they’d prefer to walk away with whatever they have already won. The stakes are high: being wrong means losing everything you have earned. In Keppel’s case, she faced this decision with £500,000 on the line. The million-pound question was: “Which king was married to Eleanor of Aquitaine?” After a brief discussion with the show’s host, Chris Tarrant, she settled on the answer of Henry II. Then Tarrant asked his killer question, the moment when contestants typically agonize the most: “Is that your final answer?” Success again rests on self-awareness. You want to know if you’re likely to be right before accepting the gamble. Keppel stuck to her guns and became the show’s first winner of the top prize.
What unites the stories of Jane, James, and Judith is how keenly their success or failure hinges on having accurate self-awareness. To appreciate the power of metacognition, we can reimagine their stories in a world where self-awareness is inaccurate. Jane might have erroneously thought that because the fluid mechanics problems felt easy, she could close the book on that topic and move on. She would think she was doing fine, even if this was not the case. A metacognitive error such as this could lead to catastrophic failure in the exam, despite Jane’s raw ability and diligent studying. In Judith’s case, we can identify two types of metacognitive failure: She may have known the answer but thought she did not, and therefore would have missed out on the opportunity to become a millionaire. Or she may have been overconfident, choosing to gamble on a wrong answer and losing everything. In James’s case, such overconfidence may even be the difference between life and death. If he had thought that he was able to handle deeper depths than he was capable of, he would, like a submarine Icarus, have overreached and realized his mistake only when it was too late.
We often overlook the power of metacognition in shaping our own lives, both for good and ill. The relevance of good self-awareness can seem less obvious than, say, the ability to solve equations, perform athletic feats, or remember historical facts. For most of us, our metacognition is like the conductor of an orchestra, occasionally intervening to nudge and guide the players in the right (or wrong) direction in ways that are often unnoticed or unappreciated at the time. If the conductor was absent, the orchestra would continue to play—just as Jane, James, and Judith would continue to plow on with studying, diving, and game-show answering even if their self-awareness was temporarily abolished. But a good conductor can make the difference between a routine rehearsal and a world-class performance—just as the subtle influence of metacognition can make the difference between success and failure, or life and death.
Another reason why the role of self-awareness is sometimes ignored is that it has historically proven difficult to measure, define, and study. But this is now changing. A new branch of neuroscience—metacognitive neuroscience—is pulling back the veil on how the human mind self-reflects. By combining innovative laboratory tests with the latest in brain imaging technology, we are now gaining an increasingly detailed picture of how self-awareness works, both as a cognitive process and as a biological one. As we will see, a science of metacognition can take us further than ever before in knowing ourselves.3
Creating a Science of Self-Awareness
I have been fascinated by the puzzle of self-awareness ever since I was a teenager, when I was drawn to books on the brain and mind. I remember glancing up from one of those books while lying by a pool during a summer vacation and daydreaming about my experience: Why should the mere activity of brain cells in my head lead to this unique experience of light shimmering on the surface of the swimming pool? And more to the point: How can the very same brain that is having this experience allow me to think about these mysteries in the first place? It was one thing to be conscious, but to know I was conscious and to think about my own awareness—that’s when my head began to spin. I was hooked.
I now run a neuroscience lab dedicated to the study of self-awareness at University College London. My team is one of several working within the Wellcome Centre for Human Neuroimaging, located in an elegant town house in Queen Square in London.4 The basement of our building houses large machines for brain imaging, and each group in the Centre uses this technology to study how different aspects of the mind and brain work: how we see, hear, remember, speak, make decisions, and so on. The students and postdocs in my lab focus on the brain’s capacity for self-awareness. I find it a remarkable fact that something unique about our biology has allowed the human brain to turn its thoughts on itself.
Until quite recently, however, this all seemed like nonsense. As the nineteenth-century French philosopher Auguste Comte put it: “The thinking individual cannot cut himself in two—one of the parts reasoning, while the other is looking on. Since in this case the organ observed and the observing organ are identical, how could any observation be made?”5 In other words, how can the same brain turn its thoughts upon itself?
Comte’s argument chimed with scientific thinking at the time. After the Enlightenment dawned on Europe, an increasingly popular view was that self-awareness was special and not something that could be studied using the tools of science. Western philosophers were instead using self-reflection as a philosophical tool, much as mathematicians use algebra in the pursuit of new mathematical truths. René Descartes relied on self-reflection in this way to reach his famous conclusion “I think, therefore I am,” noting along the way that “I know clearly that there is nothing that can be perceived by me more easily or more clearly than my own mind.” Descartes proposed that a central soul was the seat of thought and reason, commanding our bodies to act on our behalf. The soul could not be split in two—it just was. Self-awareness was therefore mysterious and indefinable, and off-limits to science.6
We now know that the premise of Comte’s worry is false. The human brain is not a single, indivisible organ. Instead, the brain is made up of billions of small components—neurons—that each crackle with electrical activity and participate in a wiring diagram of mind-boggling complexity. Out of the interactions among these cells, our entire mental life—our thoughts and feelings, hopes and dreams—flickers in and out of existence.
But rather than being a meaningless tangle of connections with no discernible structure, this wiring diagram also has a broader architecture that divides the brain into distinct regions, each engaged in specialized computations. Just as a map of a city need not include individual houses to be useful, we can obtain a rough overview of how different areas of the human brain are working together at the scale of regions rather than individual brain cells. Some areas of the cortex are closer to the inputs (such as the eyes) and others are further up the processing chain. For instance, some regions are primarily involved in seeing (the visual cortex, at the back of the brain), others in processing sounds (the auditory cortex), while others are involved in storing and retrieving memories (such as the hippocampus).
In a reply to Comte in 1865, the British philosopher John Stuart Mill anticipated the idea that self-awareness might also depend on the interaction of processes operating within a single brain and was thus a legitimate target of scientific study. Now, thanks to the advent of powerful brain imaging technologies such as functional magnetic resonance imaging (fMRI), we know that when we self-reflect, particular brain networks indeed crackle into life and that damage or disease to these same networks can lead to devastating impairments of self-awareness.7
Know Thyself Better
I often think that if we were not so thoroughly familiar with our own capacity for self-awareness, we would be gobsmacked that the brain is able to pull off this marvelous conjuring trick. Imagine for a moment that you are a scientist on a mission to study new life-forms found on a distant planet. Biologists back on Earth are clamoring to know what they’re made of and what makes them tick. But no one suggests just asking them! And yet a Martian landing on Earth, after learning a bit of English or Spanish or French, could do just that. The Martians might be stunned to find that we can already tell them something about what it is like to remember, dream, laugh, cry, or feel elated or regretful—all by virtue of being self-aware.8
But self-awareness did not just evolve to allow us to tell each other (and potential Martian visitors) about our thoughts and feelings. Instead, being self-aware is central to how we experience the world. We not only perceive our surroundings; we can also reflect on the beauty of a sunset, wonder whether our vision is blurred, and ask whether our senses are being fooled by illusions or magic tricks. We not only make decisions about whether to take a new job or whom to marry; we can also reflect on whether we made a good or bad choice. We not only recall childhood memories; we can also question whether these memories might be mistaken.
Self-awareness also enables us to understand that other people have minds like ours. Being self-aware allows me to ask, “How does this seem to me?” and, equally importantly, “How will this seem to someone else?” Literary novels would become meaningless if we lost the ability to think about the minds of others and compare their experiences to our own. Without self-awareness, there would be no organized education. We would not know who needs to learn or whether we have the capacity to teach them. The writer Vladimir Nabokov elegantly captured this idea that self-awareness is a catalyst for human flourishing:
Being aware of being aware of being. In other words, if I not only know that I am but also know that I know it, then I belong to the human species. All the rest follows—the glory of thought, poetry, a vision of the universe. In that respect, the gap between ape and man is immeasurably greater than the one between amoeba and ape.9
In light of these myriad benefits, it’s not surprising that cultivating accurate self-awareness has long been considered a wise and noble goal. In Plato’s dialogue Charmides, Socrates has just returned from fighting in the Peloponnesian War. On his way home, he asks a local boy, Charmides, if he has worked out the meaning of sophrosyne—the Greek word for temperance or moderation, and the essence of a life well lived. After a long debate, the boy’s cousin Critias suggests that the key to sophrosyne is simple: self-awareness. Socrates sums up his argument: “Then the wise or temperate man, and he only, will know himself, and be able to examine what he knows or does not know.… No other person will be able to do this.”10
Likewise, the ancient Greeks were urged to “know thyself” by a prominent inscription carved into the stone of the Temple of Delphi. For them, self-awareness was a work in progress and something to be striven toward. This view persisted into medieval religious traditions: for instance, the Italian priest and philosopher Saint Thomas Aquinas suggested that while God knows Himself by default, we need to put in time and effort to know our own minds. Aquinas and his monks spent long hours engaged in silent contemplation. They believed that only by participating in concerted self-reflection could they ascend toward the image of God.11
A similar notion of striving toward self-awareness is seen in Eastern traditions such as Buddhism. The spiritual goal of enlightenment is to dissolve the ego, allowing more transparent and direct knowledge of our minds acting in the here and now. The founder of Chinese Taoism, Lao Tzu, captured this idea that gaining self-awareness is one of the highest pursuits when he wrote, “To know that one does not know is best; Not to know but to believe that one knows is a disease.”12
Today, there is a plethora of websites, blogs, and self-help books that encourage us to “find ourselves” and become more self-aware. The sentiment is well meant. But while we are often urged to have better self-awareness, little attention is paid to how self-awareness actually works. I find this odd. It would be strange to encourage people to fix their cars without knowing how the engine worked, or to go to the gym without knowing which muscles to exercise. This book aims to fill this gap. I don’t pretend to give pithy advice or quotes to put on a poster. Instead, I aim to provide a guide to the building blocks of self-awareness, drawing on the latest research from psychology, computer science, and neuroscience. By understanding how self-awareness works, I aim to put us in a position to answer the Athenian call to use it better.
I also aim to help us use our machines better—both those that exist today and those that are likely to arrive in the near future. As with your imagined visit to the doctor’s artificially intelligent clinic and its inexplicable advice to have surgery, we are already being forced to deal with complex systems making decisions we do not understand. We are surrounded by smart but unconscious algorithms—from climate forecasting models to automatic financial traders—and similar tools are poised to encroach on all areas of our lives. In many cases, these algorithms make our lives easier and more productive, and they may be required to help us tackle unprecedented challenges such as climate change. But there is also a danger that deferring to supersmart black boxes will limit human autonomy: by removing metacognition from the equation, we will not know why or how certain decisions were made and instead be forced into blindly following the algorithms’ advice. As the philosopher Daniel Dennett points out: “The real danger, I think, is not that machines more intelligent than we are will usurp our role as captains of our destinies, but that we will overestimate the comprehension of our latest thinking tools, prematurely ceding authority to them far beyond their competence.”13 As we will see, the science of self-awareness provides us with alternative visions of this future, ones that ensure that awareness of competence remains at the top of the priority list, both for ourselves and our machines.
Let’s take a look at the road ahead. The big idea of this book is that the human brain plays host to specific algorithms for self-awareness. How these algorithms work will occupy us in Part I. We will see that the neural circuits supporting metacognition did not just pop up out of nowhere. Instead, they are grounded in the functions of the evolved human brain. This means that many of the building blocks of metacognition are also shared with other species and are in place early in human development. We’ll cover both the unconscious processes that form the building blocks of self-monitoring and the conscious processes that enable you to be self-aware of the experiences you are having. As will become clear, when we talk about self-awareness, what we really mean is a collection of capacities—such as being able to recognize our mistakes and comment on our experience—that when bundled together result in a self-aware human being.14
By the end of Part I, we will have seen how a number of critical components come together to create a fully-fledged capacity for self-awareness. We will also be in a position to understand how and why these processes sometimes go wrong, leading to failures of self-awareness in diseases such as schizophrenia and dementia. In Part II, we will then turn to how we use self-awareness in many areas of our lives to learn, make decisions, and collaborate with others. By understanding how and why self-awareness may become distorted—and by recognizing both its power and fragility—we will be in a position to ensure that we do not end up in situations in which it tends to misfire. We’ll dig into several important arenas of human affairs—including the crucial role that metacognition plays in witnesses testimony, in politics, and in science—to see why knowing ourselves, and knowing how others know themselves, is crucial to building a fairer and better society. We’ll explore how self-awareness helps us separate reality from imagination and how, by learning to harness it, it can even help us shape our dreams. We will see that because self-awareness is sometimes absent there are, in fact, plenty of cases in which we humans are also no better than black boxes, unable to explain what we have done or why.
We will also see that, despite these limitations, the human capacity for self-awareness and self-explanation is what underpins our notions of autonomy and responsibility. We’ll explore the role of self-awareness in classroom learning and teaching. We’ll see why in sports it might be better to be less self-aware to perform well but more self-aware when coaching others. We’ll see how digital technology changes our awareness of ourselves and others in a range of crucial ways. Indeed, I’ll make the case that in a world of increasing political polarization and misinformation, cultivating the ability to self-reflect and question our beliefs and opinions has never been more essential. We’ll explore why computers—even the most powerful—currently lack metacognition, and how the increasing prevalence of machine learning in AI means that algorithms for intelligence are rapidly diverging from algorithms for self-awareness. We’ll examine what this might mean for society and how we might fix it, either by attempting to build self-awareness into our computers or by ensuring we can understand and use the machines that we build. However that effort concludes, it may hold the key to solving some of the most pressing problems in society.
By the end of all this, I hope it will be clear why, from ancient Athens to the boardroom of Amazon.com, cultivating self-awareness has always been essential to flourishing and success. But we are getting ahead of ourselves. To unravel the mysteries of how self-awareness works, we need to start with the simplest of building blocks. Let’s begin with two features of how our minds work: how we track uncertainty and how we monitor our actions. These two features may appear simple, but they are fundamental components of a self-aware brain.