Researchers at the University of Pittsburgh had to throw out Kafka. They were building a reading experiment and needed a book that volunteers would drift away from, but The Trial kept pulling people in. Jane Austen solved the problem. Sense and Sensibility got the job instead. Psychologist Erik Reichle picked it for being undemanding enough to lose focus on without being unbearable, as reported by ScienceDaily.

Four undergraduates then sat through a dozen or more hour-long sessions with the novel on a screen and an eye tracker logging every flick and pause. They were told to press a button marked Z the moment they noticed they had stopped taking anything in.

Everybody knows the feeling. You reach the bottom of a page, realise nothing has landed, and go back up to find the last sentence you actually read.

What the eyes do when nobody is home

Cognitive scientists call the general phenomenon mind-wandering, and the reading version has its own label: mindless reading. Reichle worked with Andrew Reineberg and Jonathan Schooler on the study, published in Psychological Science, and the headline result is that the eyes carry on regardless. They keep moving across the page after the mind has left the building.

They just move differently. In ordinary reading, your eyes pause longer on rare or awkward words, because the pauses are being steered by the language processing going on underneath. During mindless reading that link dissolves. Fixations stretch out and stop responding to how hard the words are, as though the eyes had switched to autopilot.

Small sample, one experiment, four readers. Worth holding lightly. But it does suggest a tidy diagnostic: the eye movements go dumb before the reader notices anything is wrong.

Nearly half of waking life

How often does this happen away from the lab? Matthew Killingsworth and Daniel Gilbert went looking with a smartphone app, pinging around 2,250 adults at random moments and asking what they were doing, how they felt, and whether they were thinking about something other than the task in front of them. Roughly 250,000 responses came back.

People reported being mentally elsewhere 46.9 per cent of the time, in a paper published in Science in 2010.

The mood finding got more attention than the frequency one. Mind-wandering predicted how happy someone felt more strongly than the activity itself did, accounting for about 10.8 per cent of the variation in moment-to-moment happiness against 4.6 per cent for the activity, per the Harvard Gazette. Wandering at one sampling point predicted lower mood at the next, while low mood did not reliably predict wandering later, which is the pattern you would expect if the drift were doing the damage rather than reflecting it.

That is one dataset from one team, and the causal question is still argued over. Treat the 47 per cent as the solid part.

Why the brain unplugs on purpose

None of this is the brain malfunctioning. Jonathan Smallwood and Jonathan Schooler, reviewing two decades of work in the Annual Review of Psychology, describe a process called perceptual decoupling: attention pulls back from the senses so an internal train of thought can run without being interrupted by whatever is in front of you.

Which explains the Austen result rather neatly. Holding a daydream together takes resources, and the cheapest place to take them from is the incoming text.

The bit that argues for the defence

Here is the study people reach for when they want mind-wandering to sound like a feature rather than a bug. Participants tackled a divergent thinking task, took a break filled with either a demanding job, an undemanding one, plain rest, or nothing, then had another go. The undemanding break produced the most drifting and the biggest improvement second time round, Benjamin Baird and colleagues found.

Read the small print, though. The gain showed up only on problems the participants had already seen. Letting your mind roam did not make them more creative in general. It helped them get further with something already sitting in the background.

Whether any of it can be trained

Michael Mrazek and colleagues at UC Santa Barbara put students through a two-week mindfulness course and tested them either side of it. Reading comprehension scores on the GRE went up, working memory capacity went up, and the gains ran statistically through reduced wandering rather than around it. Mrazek’s group put the reading improvement at roughly 16 percentile points, as documented by ScienceDaily.

Two weeks, undergraduates, one lab. Replication has been patchy across the wider mindfulness literature, so the honest reading is promising rather than proven.

Where this leaves the page you keep re-reading

One awkward finding runs through all four studies. Catching yourself is the rare part.

The Pittsburgh volunteers pressed their button sometimes, but the random probes kept turning up episodes nobody had flagged, and the whole design of the smartphone study rests on interrupting people, because they will not report the drift unless someone asks.

Which makes noticing the interesting variable. Roughly half of consciousness is spent elsewhere and the machinery seems to be earning its keep while it is there, so the target worth aiming at is the gap: the stretch between leaving the page and realising you left. That gap is where the re-read comes from, and so far it is the only part of the whole business anyone has shown you can shorten.