The Uses And Costs Of A Wandering Mind

AOn a wet Tuesday in 2014, a driver on the Marlow branch line ran two caution signals in succession and told the inquiry afterwards that he had been thinking about his daughter's school application. Nothing in the sequence was difficult and nothing had been forgotten; his attention had simply gone elsewhere and come back, and he could not say when. Psychologists call this mind-wandering: an episode in which thought uncouples itself from the task in hand and from the room the body is sitting in. Experience-sampling studies, in which a phone beeps at random moments and asks what the respondent was thinking about, put the phenomenon at between a quarter and a half of waking life; a 2016 survey of 3,400 adults run by the Kessel Institute recorded off-task thought in 38 per cent of the moments it sampled. The dispute among researchers is not over how often this happens, which almost nobody contests, but over the verdict: whether this is the mind failing, or the mind doing something else that matters.

BIngrid Sollberg, who directs the attention laboratory at Kessel, has built her case on the losses. Between 2011 and 2019 she ran thirty-four experiments in which readers were interrupted at random and asked whether their thoughts had been on the page; in every study, those who admitted to wandering answered fewer comprehension questions afterwards, and nothing in the procedure allowed the lost lines to be recovered. The eyes told the same story. Measurements of pupil size showed that the eye stops responding to what it is crossing while the reader is away, and Sollberg's tracking data record the eyes travelling on across the text for eleven seconds on average after understanding has stopped. Perceptual decoupling is her term for that state, and its most troubling feature is that the sense of having followed the argument survives it intact. Wandering, on her account, is not a mode of thought but a failure of the machinery that keeps thought on task, and the students who wander most in her studies also score lowest on tests of working memory.

CTomas Reyner, who works with people whose problems arrive without deadlines, thinks the laboratory has chosen its own battleground. Between 2015 and 2020 he collected 214 accounts of creative breakthroughs from engineers, composers and mathematicians; in 61 per cent of them the solution surfaced not at the desk but during some undemanding activity — walking, showering, riding a bus — and always after a stretch of effort that had ended in failure. He also disputes the picture of wandering as idle fantasy. When his team sampled the content of episodes rather than merely counting them, roughly two-thirds concerned the respondent's own future, arrangements to be made and conversations to be had, which Reyner calls autobiographical planning. Tasks that punish any break in attention, he points out, are exactly the tasks psychologists like to carry into a laboratory, because they yield clean numbers; a discipline that studies wandering only in such settings has decided its verdict before collecting the evidence.

DThe two shared a platform at a workshop in Bergen in 2018 and left with a joint paper rather than a winner. Both conceded that they had been arguing about different situations. The paper sets out two conditions under which a wandering episode repays its cost: deliberate work must already have carried the problem to an impasse, and the activity that fills the interval must make few demands of its own. Where either fails — where nothing has been prepared, or where the gap is filled by a second exacting task — the episode returns nothing. The cost, by contrast, is fixed by the task that has been left: it rises in proportion to how severely that task punishes a lapse, which is why a signal missed on a railway and a daydream on a garden bench are not the same event. Neither quantity, the authors stressed, is a property of the wanderer.

EThe framework leaves people poorly placed to police themselves. In 2021 Sollberg probed 240 undergraduates at random during a forty-minute lecture; those who had described themselves beforehand as rarely wandering were off task at 41 per cent of the probes and everyone else at 39, a gap smaller than the error of the measurement. Worse, in 44 per cent of the episodes the probe caught, the student reported having had no idea until that moment that they had left — zoning out, as Sollberg labels it, to distinguish it from the wandering people notice as it happens. She takes the finding to dispose of self-report altogether. Reyner accepts the numbers and refuses the conclusion: that a group's self-descriptions predict nothing, he replies, does not establish that no practised individual ever catches the moment of departure, and the study that would settle it has not been run.

FThe remedies that follow are unglamorous and, awkwardly, compatible. Sollberg's are structural: shorten the block of continuous monitoring, require a physical response every few minutes so that departure declares itself, and stop designing jobs in which nothing is asked of an operator for an hour. Reyner's is a matter of timing — put the undemanding interval, the incubation walk, immediately after an impasse rather than at the end of the day, when the problem has gone cold. Several engineering firms now do both, on different parts of the same working day. What neither can settle is the classroom, where a lapse costs a paragraph of the lesson and may at the same time be the only planning a fourteen-year-old does all week; Sollberg counts it as loss by default, Reyner calls the question open. Until it is closed, the safest thing to say is that whether a wandering mind is wasting the hour or rescuing it is decided less by the wanderer than by the hour.