The Mind That Watches Itself

Metacognition, rumination, and why “just be more self-aware” is so often the wrong advice.

Some advice sounds wise but lands badly. Be more self-aware. Notice your thoughts. Pay attention to what’s going on inside. For some people, at some moments, that is exactly right. But it tends to be handed to the people it helps least. For them, it can be a little like telling someone who is struggling in deep water to think harder about the water.

Self-awareness isn’t a single dial you can turn up. The power behind it is a psychological term: metacognition. Then when you get a glimpse of how it really works, “be more self-aware” begins to sound more like a misstep. For many individuals who seek therapy, it is not a matter of failing to watch their own minds. It is because they can't stop.

What metacognition actually is

The word was coined by the psychologist John Flavell in 1979. It means, roughly, thinking about thinking. It is the part of the mind that turns back on itself: noticing you don’t understand something, sensing a name is on the tip of your tongue, judging whether you have revised enough, catching yourself drifting off. None of that is the thinking itself. It sits a layer above, keeping an eye on things.

Sounds good, and in many ways it is. The trouble begins when we treat accumulation as a virtue and assume that more is always better. It isn’t a virtue. It is a system. As in any system, it can be well-tuned or badly tuned and can turn on you.

Two levels, two flows

A helpful way to picture it comes from two researchers, Thomas Nelson and Louis Narens. In 1990, they described the mind as working on two levels that stay in conversation with each other.

One is the object level. It does the actual work of thinking, remembering and feeling. The other is the meta-level. It sits above, holding a rough model of what the object level is doing. Two flows of information run between them.

Monitoring is the flow upward: the meta-level reading what the working mind is up to (“this isn’t going in”, “I’m getting anxious”, “I keep coming back to this”). Control is the flow downward: the meta-level adjusting what happens next (“slow down”, “try another way”, “let this go”).

Two flows of information run between the levels. When monitoring keeps running while control falls quiet, you can know you are stuck and be unable to stop.

Keeping monitoring and control separate matters because, in distress, they often come apart on their own. Imagine the monitoring flow working perfectly while the control flow goes slack. You would know, with painful clarity, that you were turning the same thought over and over. And you would be unable to stop. That is close to what rumination actually feels like. The trouble isn’t too little self-awareness; there is plenty of it. What has failed is the steering, not the watching.

For a lot of people, the problem isn’t that they aren’t watching their own minds. It is that the watching has nowhere useful to go.

This is why “notice your thoughts more” can quietly make things worse. If your monitoring is already relentless, that advice asks for more of what hurts. The useful direction is usually the opposite. You don’t need to watch more. You need a different relationship with what the watching turns up.

It’s rarely the thought — it’s what you believe about it.

Modern therapy offers a useful turn here. We tend to assume that distressing thoughts are the problem and that the fix is to argue with them, disprove them, or replace them with better ones. The psychologist Adrian Wells suggests something else. What keeps worry and low mood going is often not the content of the thoughts at all. It is what we believe about thinking itself.

Wells discusses a pattern he calls the 'cognitive attentional syndrome'. It's a way of thinking characterised by worry and rumination, by scanning for threats, and by coping mechanisms that ultimately work against you, such as suppressing thoughts or seeking constant reassurance. Two beliefs sustain it.

On one side are positive beliefs about thinking: the sense that worrying is useful, that if you turn a problem over enough, you will be safe or ready (“if I think it all through, nothing can blindside me”). On the other hand, there are negative beliefs: that the thinking is out of control or dangerous (“I have no control over this”, “dwelling like this means something is wrong with me”). Together they make a trap. The first belief draws you in. The second tells you that you cannot get out and that being stuck there is harmful. That is frightening, and the fear produces more of exactly the thinking you were afraid of.

This reframing alters the work. You don't wrestle each worry to the ground (there's always more to worry about!); you work on the belief that sits beneath all your worries. Typically, the most significant one to loosen is the idea that the rumination is uncontrollable. There’s no reason to win the argument with the thought. It's just a matter of breaking free from the belief that the argument is necessary.

Minds that are wired a little differently

It would be tidy to say that some people are more self-aware than others, and leave it there. The evidence doesn’t allow anything that tidy. For neurodivergent people, especially, the popular story tends to be confident and wrong.

Take autism. It is often thought that people with autism have no idea what is going on in their minds. It's actually far more nuanced than that. When researchers measure it, the picture is genuinely mixed rather than one-directional. One 2023 study that aggregated 17 studies reported a moderate, average decrease in the accuracy of autistic participants' perception of their thoughts. But the same review was cautious: It was far from universal, the effect was far from uniform, and the results were extremely variable depending on which measure was taken and how. Some forms of self-evaluation, such as self-rating of confidence in one's answers, appear to be relatively intact in autistic adults. Others appear to be more impacted. The truth is that it isn’t “impaired”. It is nearer to differently and unevenly calibrated and never goes away.

Something similar holds for ADHD. The familiar finding is a tendency to overrate your own performance, the so-called positive illusory bias, documented most consistently in children. But this resists becoming a rule for an entire group, too. Look closely, and overly positive self-judgement turns out not to be a feature of everyone with ADHD. It clusters in some people while others judge themselves accurately, and many adults with ADHD actually hold a noticeably harsher view of themselves than the facts warrant.

Here, the missing capacity is not usually an issue. It's one that is irregular, vigorous in some places and weak in others. This is important both on a clinical and personal level. Much of the discourse around neurodivergent minds – ‘poor self-awareness' or ‘lacks insight' – takes a narrow, conditional, measurable difference and inflates it into a blanket statement about the capacity for self-awareness. The facts do not back up that conclusion, which is also harmful to live under. Thinking about metacognition as a continuum of "calibrated" rather than as "you have it or you don't" is more accurate and a little kinder.

A gentler way through

What, then, is the work? If the watching mind is not the problem, then what is wrong with more watching?

It's not so much about the content; it's more about a change in the relationship. You're not trying to think less, or just right thoughts or to watch yourself to calm. You are learning to stop the meta-level from repeating your distress to you over and over again. Sometimes that involves simply allowing a thought to come and go. Not to reject it but to refuse the call to follow, for in this way it merely shows itself to be a danger. Sometimes it involves doggedly challenging the belief that the thinking is uncontrollable, typically by discovering, in some small experiments, that it is more controllable than it appeared. And often it means noticing your own monitoring habit, whether it runs hot, or lags, or scans constantly for threat, and meeting that habit with curiosity instead of another layer of judgement.

None of this is quick, and none of it happens just by reading an article, including this one. However, basic reframing is calming and is something to cling to. There was never an intention to monitor yourself more closely. The observing mind is a marvellous thing. It just works best when it is allowed, now and then, to look up from the page.

Curious about your own patterns of mind?

If the relentless-watching, can’t-switch-off picture feels familiar, working through it with someone can make a real difference. I offer integrative, neurodiversity-informed therapy online, across the UK and worldwide.

Book a free consultation →

A NOTE ON THE EVIDENCE

This piece is inspired by the work of John Flavell (1979) on metacognition, Thomas Nelson and Louis Narens (1990) on two levels of monitoring and control, Adrian Wells (1994, 1996; Wells, 2009) on metacognitive therapy and the Self-Regulatory Executive Function model with Gerald Matthews, and Carpenter and Williams (2023) on a meta-analysis of metacognition in autism, alongside the literature on positive illusory bias in ADHD. The two-level model is used here as an illustrative way of describing felt experience; the link between “monitoring without control” and rumination is offered as a clinical illustration rather than as a claim made directly by those authors. Research on metacognition in autism and ADHD is genuinely mixed and continues to develop, and the descriptions here are general rather than diagnostic. Nothing in this article is a substitute for individual assessment or therapy.