There Is No Peak Age

Four Tests, the Same People, the Same Day

Psychologists Joshua Hartshorne (MIT) and Laura Germine (Harvard) collected results from 48,537 people, ranging from adolescence to old age, through the public tests on testmybrain.org and gameswithwords.org, supplemented with norms from standardized intelligence and memory tests gathered over several decades. What sets this apart from older research isn't the tests themselves but the sample: the same people, tested on the same day, with the same method, at any age between fifteen and eighty, which makes the peaks of different abilities directly comparable instead of stitched together from separate studies with different populations and methods.

The result, published in Psychological Science, splits the mind into abilities that crest one after another, not together. Processing speed — how fast you recognize and match symbols — climbs steeply and peaks at nineteen, then declines relatively quickly. Working memory keeps climbing until roughly twenty-five to thirty-five, stays near its peak for a while, and then declines very slowly. Recognizing emotion on someone else's face climbs even more slowly, peaks around forty to fifty, and declines noticeably faster after sixty. Vocabulary and general knowledge kept climbing for the entire span the study tested: at seventy, in this study's data, it still hadn't stopped growing.

From the peak of speed to the edge of vocabulary is more than fifty years, in the same mind.

The gaps between peaks aren't equal — they widen. Six years separate the peak of speed from the start of the working-memory window; fifteen years separate the start of that window from the start of the emotion-recognition window; at least twenty-five years separate that one from the point where vocabulary was still climbing, at the end of testing. The mind doesn't peak once — the gaps between its peaks stretch wider the further you go.

Your Age, Through Four Windows

Type your age, or drag the slider. Each colour shows where that ability stands right now, relative to its own peak.

What Ripens Fast, What Ripens Slow

The distinction isn't new in psychology: back in the 1940s, Raymond Cattell already spoke of "fluid" intelligence — raw processing and problem-solving speed, sensitive to age — and "crystallized" intelligence — vocabulary and accumulated knowledge, which behaves like a savings account, not a sprint. What Hartshorne and Germine's study added is resolution: between fluid and crystallized sit at least two more distinct steps — working memory and emotion recognition — each keeping its own calendar.

The plausible physiological explanation has to do with what matures last in the brain. The prefrontal cortex, involved in processing speed and attentional control, matures relatively early and starts losing synaptic density from young adulthood on. The circuits behind emotion recognition and social judgment depend on accumulated experience — thousands of faces seen, thousands of situations interpreted — which is likely why their peak arrives so much later. Vocabulary has, in principle, no physiological ceiling to its growth: every new word learned adds to the existing stock, for as long as the mind stays able to register new information.