The Capacity to Be Alone Starts With Someone Else in the Room

In 2014 Timothy Wilson and his colleagues at the University of Virginia published eleven studies in Science about a simple instruction: sit alone in a room and think. Most people did not enjoy it. In one study, participants who had already said they would pay money to avoid a mild electric shock were left alone for fifteen minutes with a button that delivered it. Two thirds of the men and a quarter of the women pressed it at least once.

The paediatrician and psychoanalyst Donald Winnicott had a name for the thing those button presses seemed to be missing. He called it the capacity to be alone, and in a short paper in 1958 he argued it was one of the clearest signs of emotional maturity. His account of where it comes from is a paradox. People learn to be alone by being alone in the presence of someone else.

What Winnicott meant by the capacity to be alone

The capacity to be alone is the ability to rest in your own company without feeling deserted and without needing to fill the time. Spending a lot of time on your own is a different thing. Someone can live largely alone and never have it, and someone surrounded by people can.

He separated it from two things it is often confused with. One is withdrawal, which is a defence: getting away from contact because contact feels dangerous. That is the move R. D. Laing would describe two years later in people for whom being understood felt like being swallowed. The other is loneliness, which is the felt absence of someone. The capacity to be alone is neither. It is closer to being at ease in a room that happens to have only you in it.

For Winnicott this mattered well beyond spare afternoons. A person who has it can let a relationship go silent for a while without reading the silence as abandonment. Discontinuity in a bond does not feel like the end of it.

Being alone with someone else there

The paradox comes from infancy. Winnicott’s picture is a very young child absorbed in itself, with an adult nearby who is reliably there and is not demanding anything. The child is alone, in the sense that nothing is being asked of it and it is free to drift. It is also not alone, because the drifting is only safe while someone dependable is in the background.

He called the relationship between those two people ego-relatedness, a relationship in which one of the two is alone and the other’s presence is what makes that possible. Over time, if the background presence is steady enough, the child takes it in. The reliable other no longer needs to be in the room, because something of them is now carried inside. Winnicott traced the steps as a movement from the infant being able to say I am, to being able to say I am alone, without the second sentence meaning I have been left.

This is the same caregiver who appears in Winnicott’s good enough mother: present, responsive and not intrusive, someone who does not fill every pause. The capacity to be alone is what that kind of presence leaves behind. Where it was missing, or arrived too unpredictably, the empty room never becomes neutral ground. It stays the place where someone should be and is not.

What the shock study did and did not show

The Wilson result became one of the most quoted findings in popular psychology, usually as a line about people preferring pain to their own thoughts. The research since has made it stranger and less tidy.

Later in 2014, a group led by Kieran Fox published a commentary in Frontiers in Psychology that went back through Wilson’s data and found little support for two of its headline claims: that people typically disliked the thinking period, and that many preferred shocks to being left alone with their thoughts. In 2022 Andreas Eder and colleagues ran four experiments and replicated the basic pattern. Plenty of people shocked themselves. But they chose strong shocks even when a harmless option was available, and they did it even when they had been helped to think pleasant thoughts. The more likely motive was curiosity about the shock, not escape from the mind.

The same year, Aya Hatano, Kou Murayama and colleagues asked people to predict how much they would enjoy sitting and waiting, then had them do it. Across six experiments, people expected to enjoy it less than they actually did, and that low expectation was enough to make them choose checking the news online instead, even though their experience of the two was not reliably different.

So the firmer finding is not that solitude is unbearable. It is that people predict it badly. They walk into the empty room expecting less of it than it gives them, and reach for something else before finding out.

Why the empty room can feel like being left

None of these studies tested Winnicott. They measured adults on an afternoon, not the history of how those adults came to sit in a room. What his account offers is a reading of the prediction itself. For someone with the capacity to be alone, an hour with nothing scheduled is an hour. For someone without it, the same hour carries an older meaning, the absence of a person who was supposed to be in the background, and so it gets filled before it can be felt.

The Hatano finding is suggestive here, though it cannot carry the weight of proof. When people in that study avoided the waiting, what they chose instead was the news: other people’s voices, somewhere to be addressed. Whether that is a hunger for stimulation or for company, the data cannot say.

What is settled, and what is not

The settled part is narrow and solid. People consistently underestimate how bearable, and how engaging, time alone with their own thoughts turns out to be, and the famous shock result is weaker and odder than its headline. Winnicott’s account is a different kind of claim: clinical, built from observation and from the consulting room, and never tested the way the laboratory studies were. It explains the shape of the difficulty well. It does not prove where the difficulty comes from in any given person.

What it leaves open is the question the shock study accidentally asked. Fifteen minutes in a bare room, and many people want something, anything, to happen. The research can measure how long they last before reaching for the button. It cannot say whether the room felt empty because nothing was happening in it, or because nobody was.

FAQ

What is the capacity to be alone?

It is D. W. Winnicott’s term, from a 1958 paper in the International Journal of Psycho-Analysis, for the ability to rest in your own company without feeling deserted and without needing to fill the time. He treated it as a sign of emotional maturity, and distinguished it from withdrawal, which is a defence against contact, and from loneliness, which is the felt lack of someone.

What does “alone in the presence of someone” mean?

Winnicott’s paradox is that people learn to be alone by being alone while someone else is there. An infant absorbed in itself, with a reliable adult nearby who asks nothing of it, is having that experience. Over time the reliable presence is taken inside, and the person can be alone without the other person in the room.

Is the capacity to be alone the same as being an introvert?

No. Introversion describes a preference about stimulation and company. Winnicott’s idea is about whether solitude feels safe, and someone can prefer company and still have it, or prefer solitude and lack it. They are different questions about the same empty room.

Did people really prefer electric shocks to being alone with their thoughts?

A 2014 study by Timothy Wilson and colleagues found that many people gave themselves a mild shock during fifteen minutes alone, including two thirds of the men. Later work complicates the headline: a 2022 replication found people chose strong shocks even when a no-shock option was available, which looks more like curiosity than escape, and another 2022 study found people consistently underestimate how much they will enjoy simply sitting and thinking.