When the Mind Wanders, the Eyes Stay the Course

You read this sentence. And this one. But then, once you see in your periphery the convoluted syntax stringing this sentence together—multiple clauses, an em-dash sandwich—your thoughts begin to veer. Is it too early for lunch? The deli meat is probably still okay. Probably. Isn’t that what all the salt is for? By the end of the paragraph you realize that you don’t have the faintest idea what you just read. Or, worse, you don’t realize it and plow on mindlessly.

Reading is relatively well-studied; mind-wandering during reading—despite its prevalence—less so. But UC–Santa Barbara psychologist Jonathan Schooler and his colleagues have been doing their best to remedy this state of affairs.

In a typical mind-wandering study, undergraduate participants read from a literary heavyweight such as War and Peace, pressing a button each time they catch themselves losing focus. But of course readers are imperfect detectors of their own mental flights—when they’re off task they’re, well, off task. Researchers expect this, so participants are also prompted every few minutes and asked to make a decision about whether they were just zoning out. The idea, then, is to use both self-catches and prompted catches to measure how often a participant veers off task, as well as how capable she is of noticing.

The results themselves don’t tend to surprise. In a 2010 study, for instance, four undergraduates read Sense and Sensibility in its entirety (though not in a single sitting). The readers caught themselves zoning out an average of 22 times over the course of the experiment. They also reported zoning out after 9% of the prompts. These numbers seem quite respectable given that it took participants 15 hours to read Austen’s novel. But other studies, with other reading materials and different procedures, report much higher incidences of mind-wandering. Cigarette cravings, which distract participants, and alcohol, which incapacitates them, predictably lead to more zoning out.

(A short digression: what drew me to Schooler’s work in the first place were his titles. Consider “Lost in the sauce: The effects of alcohol on mind-wandering,” and “Out for a smoke: The impact of cigarette craving on zoning out during reading.” I can only hope that such titles as “Blanking on Benadryl” and “Puzzled by pot brownies” are in the queue.)

But as fascinating as I find the topic at hand—and indeed what is more fascinating than a mind at play?—the studies themselves have always bothered me, relying as they do solely on participants’ judgments of their own mind-wandering. Cognitive psychologists are loath to trust people’s thoughts about their own thoughts because they aren’t always accurate. In addition, I had the concern that in asking participants to catch themselves mind-wandering, researchers were changing important things about the mind-wandering itself. To some extent, this problem is inescapable: any time we bring participants into the lab, they know we’re interested in something. But if we can encourage participants to assume that a study is about something that it is not, the hope is that the variable that we are interested will fly under the radar. It is not easy, needless to say, to convince participants that a researcher is interested in anything other than zoning out while continuously questioning them about whether or not they are zoning out.

I was so bothered by this a few years ago that I spent an ill-conceived month attempting to design a better mind-wandering measurement. I never came close to succeeding (not least because my participants found War and Peace enthralling), but I was delighted to stumble across work from Schooler’s lab published last year in Psychological Bulletin and Review.

When we read, the speed with which we move our eyes is influenced by a number of factors, including the length and frequency of a given word. When our mind drifts, it seems, we continue to read, but now the well-established relationship between, say, word frequency and reading speed disappears. With the mind left out of the loop, our reading becomes insensitive to the features of the text itself. In this newer work, researchers used the lack of a relationship between text features and reading speed (by looking for the speedy reading of difficult words) to successfully predict moments of zoning out.

With enough detailed knowledge about the unique signature of the distracted eye, researchers hope to be able to catch readers in the act, and send their thoughts back to John Willoughby and those Dashwood sisters.