Exploring metacognition

Metacognition literally means cognition about cognition or thinking about our own thinking. In our research, it is a broader topic than that. It refers to reflection on our own mental states, our own perceptions, our memories and our decisions. For instance, we can introspect about whether a memory of last year's birthday party is actually real, or perhaps we have just imagined it. We can reflect on whether we have made a good or bad decision, and these are all aspects of metacognition enabling us to self-assess our own mental processes.

One question we have is what is happening in the brain when we self-reflect. For a long time, we have been interested in the role of the prefrontal cortex in the human brain, the area of the brain right toward the front of the head.

The task that we have developed to study metacognition using brain imaging involves people making simple judgments about what they see on a computer screen and then evaluating how confident they feel in making those judgments. We can quantify how good your metacognition is by the extent to which your confidence is tracking your underlying performance. What we find is that people who have better metacognition have subtle differences in the function and the structure of particular circuits in the prefrontal cortex. Brain activity in these networks is linked to the formation of this metacognitive judgment; it predicts how confident you will feel in individual decisions or actions that you make. We think that this research can start explaining why, if you have damage to these networks, you might have problems with self-awareness, because you are damaging the brain circuits that are creating this model of how you are performing in a particular task.

Brain damage or dysfunction to these regions can occur for a variety of reasons. These include stroke and the consequences of having brain surgery to remove tumors, but it can also occur for more subtle reasons. For instance, dementia or certain psychiatric conditions might lead to subtle changes in how the circuits within the prefrontal cortex and other areas of the brain are working, and we think that is one reason why, in a number of different psychiatric and neurological conditions, you might see these changes in metacognition.

Metacognition, we think, plays an outsized role in how we make decisions and how we navigate our daily life, because it allows us to form estimates of how much capacity we have, how much skill we have in particular areas of endeavor. One nice example comes from the world of sports, extreme sports in particular. In freediving, people have to estimate very precisely what they will be capable of.

The types of computational model we build in our lab are algorithmic models that are trying to explain aspects of people's behavior. We can build models that characterize people's metacognitive judgments. These are models that simulate probabilities of success on various tasks, and we try to derive equations that predict how people will rate their own confidence in those same tasks. We think there is a deep connection between representing a probability of success and how you evaluate your confidence; the fluctuations in your performance inform the model you have of yourself.

Stanislav Petrov is a fascinating character. He was a lieutenant colonel in the Soviet military in the 1980s, and he was on duty one morning in September 1983 when he was monitoring these early-warning satellites that told him there was an incoming missile attack. The doctrine at the time required him to pass this information up the chain to allow for an immediate response. Rather than doing this, he paused and reflected on what he was seeing — these alarms going off — and thought to himself, this is more likely to be a false alarm than a real attack. He was proved right, and he arguably did more than anyone else in human history to avert a worldwide nuclear disaster.

Our lab and other labs around the world have found links between mental health symptoms such as anxiety, depression, compulsivity and changes in how we see ourselves — changes in these metacognitive processes.

One area we are very interested in is the promise of developing new treatments and new interventions to improve or reshape metacognition in beneficial ways. There is a growing line of research looking at the benefits of meditation for self-evaluation.

There are two ways of approaching the debate over free will. One is through the lens of physics and fundamental properties of the universe — the debate about whether there is really any randomness in the universe or whether everything is predetermined. That is not a useful perspective on whether humans have a psychological sense of free will. Regardless of how the universe is structured and how determinate it really is, that does not have much bearing on whether we as humans feel like certain choices we make are free in the sense of acting for our own reasons, or somehow under duress, or an action that we took without thinking, or for reasons other than those we wanted in the first place. That notion of free will is very tractable psychologically because we can start understanding why someone might make a certain choice in a certain situation. I think metacognition has a role to play in that explanation, because often the choices that we think of as being most free and in line with our own wishes and desires are those which we reflectively endorse. If I make a decision about where to go on holiday, and then after making that decision think to myself, that was a good decision — that is exactly how I would want my holiday to unfold, then that can be considered a decision that was entirely in line with my own wishes and desires. That is separate; it is unrelated to whether those wishes and desires were in some sense caused by previous events. The freedom has its locus within me as an agent rather than being a fundamental property of any physical system.

Like many psychologists, I am worried about the consequences of misinformation for how people form beliefs about the world. Our research on metacognition suggests that an important component of how we form beliefs is the capacity to reflect on whether we have an accurate belief or not.