On Attention, Part Two: What the Science Actually Says

In the first essay in this series, I described the experience of sitting down with a book and finding, three pages in, that I had forgotten what I was reading. I called this the substrate problem: attention is not one skill among many but the foundation under most other cognitive skills, and when it erodes, the skills above it erode too.

In this second essay, I want to look at what the research actually says. Because the popular science version of this story — phones are destroying our brains, multitasking is impossible, the younger generation is doomed — is partly true and mostly simplified. The truth, as always, is messier and more interesting.

What the research does say

There are a few findings that have held up well across many studies, and I want to be honest about each of them.

Task-switching has a cost. When you switch from one task to another, there is a measurable cost — your performance on the second task is worse than if you had not just been doing the first. The cost is real, sometimes substantial, and it is the closest thing to a settled finding in this field. If you are writing a report and you check your email, the next sentence you write will be slightly worse than the one you would have written otherwise.

Heavy media multitaskers perform worse on attention-control tasks. This is the famous Ophir / Nass / Wagner study from 2009, replicated several times since. People who report doing a lot of media multitasking — having many browser tabs open, watching TV while scrolling, texting while reading — perform worse on tasks that require filtering out distraction. The correlation is robust. The direction of causation is not.

Phone presence reduces cognitive capacity. A 2017 study by Ward, Duke, Gneezy, and Bos found that having your phone on the desk — even face-down, even powered off — measurably reduces your performance on cognitive tasks, in a dose-response way (the more you depend on your phone, the more it costs you). This is a finding that has stuck with me.

What the research does not say

Here is where I think the popular version of the story goes wrong.

The research does not say that attention spans have collapsed to eight seconds (this is a marketing statistic from a Microsoft report, based on a survey, not a measurement of attention). It does not say that the younger generation is cognitively broken (cross-generational comparisons are hard to do well, and the ones that exist show modest effects, not catastrophe). It does not say that multitasking is impossible (we all multitask all the time — walking and talking, cooking and listening — what is costly is rapid task-switching between demanding tasks).

And it does not say that the effects are permanent or that the brain is permanently changed. The brain is changing all the time. It changed when we learned to read. It changed when we moved from scrolls to codices. It is changing now. The question is not whether it is changing — it is — but whether the change is in a direction we want.

The honest version

If I had to summarize what we actually know, in a single sentence, it would be this: the constant presence of attention-demanding devices makes sustained focus harder, in a way that compounds with use, but is reversible with practice.

That last clause is important. The substrate is not gone. It is atrophied. Atrophied muscles can be rebuilt. The same is, tentatively, true of attention.

In the third essay, I want to talk about what that practice might look like — drawn from old sources (monastic traditions of lectio divina, Buddhist practices of single-pointed concentration) and newer ones (the cognitive science of deliberate practice, the modest literature on attention training).

For now, I want to leave you with the most useful finding I have encountered in this research, which is also the simplest: when people are asked how distracted they were during a task, their answers predict future performance better than any objective measure of distraction. In other words, noticing that you are distracted is itself the first step toward not being distracted. Attention can be trained. But only if you can see when it has wandered.