Startups
You pick up your phone for a quick look. An hour vanishes. What do you remember? A man cutting marble, someone’s hot take on interest rates, a dog on a skateboard. Nothing sticks. Yet you feel drained, as if you’d done hard labor. That mismatch between effort and learning has a scientific explanation—and willpower has little to do with it. Reading a book is one continuous task. Your brain loads a setting, characters, and an argument, then reuses that setup for hours. A feed is the opposite. Every post brings new subject matter, tone, and rules. Your attention machinery must reconfigure before it can process anything. Worse, each item demands a verdict: Is this true? Funny? Aimed at me? That judgment happens before you can dismiss it—and repeats endlessly. Psychologists Joshua Rubinstein, David Meyer, and Jeffrey Evans broke down this reconfiguration in a 2001 paper. They identified two stages: goal shifting (deciding to switch tasks) and rule activation (swapping mental instructions). Each switch costs a few tenths of a second. That sounds trivial until you count the switches in an hour of scrolling. Their experiments used arithmetic and shapes, not phones, so applying this to feeds is an inference, not a direct measurement. Switching also leaves a mental residue. Sophie Leroy, now at the University of Washington Bothell, coined the term “attention residue” in 2009. In her experiments, people pulled off one task and onto another performed worse on the second because part of their attention lingered on the first—especially when the first task was unfinished. Scrolling fits that pattern perfectly: each post is dropped mid-thought, replaced by the next. But Leroy’s participants did lab tasks, so the phone parallel is an analogy, not proven fact.
Gloria Mark at UC Irvine has tracked how long people stay on one screen since 2004. Back then, it was about two and a half minutes. Recently, she told UC that the average is roughly 47 seconds, with a median of 40. That means half of all screen stretches last less than 40 seconds. This figure comes from separate studies over two decades, not one group followed throughout. It measures screen behavior, not reading ability. People still read long books—just not in a browser tab. Does the medium matter? Pablo Delgado of the University of Valencia and colleagues pooled 54 studies with over 171,000 participants. They found a consistent, modest advantage for paper, especially for informational texts and timed reading. The gap grew over the years studied. Maryanne Wolf of UCLA argues that skimming crowds out slower processes: drawing inferences, making analogies, testing truth, and adopting another’s viewpoint. Move your eyes fast enough, and none of that fits. Fatigue has its own explanation. In a 2008 CHI conference study, Gloria Mark, Daniela Gudith, and Ulrich Klocke interrupted people mid-task. The interrupted group finished faster with no quality drop—but reported more stress, frustration, time pressure, and effort. The brain compensates for fragmentation by spending harder. Those participants were working, not scrolling, and the equivalent test hasn’t been run on phones. Still, the pattern is clear: your brain isn’t built for the feed’s relentless reset.