It happens in the cereal aisle, usually. A song comes over the supermarket speakers — one that hasn’t been anywhere near a chart in fifty years — and a man who cannot reliably find his reading glasses is suddenly seventeen, in a borrowed car, with the windows down. He could not have named the song a minute earlier. He will not be able to stop humming it for the rest of the day.
Most people past middle age know this ambush. The music from one stretch of life keeps a key to the door, and everything recorded since seems to knock politely and leave. For years that was a thing people said about themselves. It is also, it turns out, a thing researchers can measure.
A word about us before the research: we are writers and parents, not psychologists or memory scientists. What follows is our plain-language reading of one published study and the literature around it — perspective, not clinical advice, and not a verdict on anyone’s memory.
A study of 470 adults and sixty-five years of hit songs
In 2020, Kelly Jakubowski and Tuomas Eerola of Durham University, with Barbara Tillmann, Fabien Perrin, and Lizette Heine of the Lyon Neuroscience Research Center, published “A Cross-Sectional Study of Reminiscence Bumps for Music-Related Memories in Adulthood” in the open-access journal Music & Science. The full study is also held by Durham University’s research repository.
They asked 470 adults in France, aged 18 to 82, about 111 popular songs that had featured in the charts between 1950 and 2015. For each song, participants rated — on ten-point scales — how familiar it was, how much they liked it, and how many personal memories it carried. One detail matters: no music was played. People rated from the song’s title and artist name alone.
The result was strikingly consistent. Every age group showed the bump for songs that charted when participants were between five and fourteen — and in three of the four groups, the songs that charted around ages ten to fourteen carried more personal memories than songs from any other period, with the researchers’ best statistical estimate of the peak landing at about age fourteen. The oldest participants, those 56 and over, peaked slightly later, between fifteen and nineteen, with the elevated window stretching to age 24. The pattern showed up in adults born more than six decades apart.
Psychologists have a name for the general phenomenon: the reminiscence bump, the long-documented tendency of older adults to disproportionately recall memories from roughly ages ten to thirty. What this study showed is how sharply music seems to concentrate that bump in early adolescence — and how the songs’ familiarity follows nearly the same curve.
The stranger finding: their parents’ music
Liking was a different story, and a stranger one. Whether people liked the songs from their own bump years was the least consistent finding in the study. The two youngest groups’ liking ratings actually dipped slightly across their own adolescent years.
What the youngest group did like — significantly more than their own average, and more than the music of their own adolescence — were songs that charted eleven to twenty years before they were born. The researchers connect this to what a 2013 American study called a “cascading reminiscence bump”: the music of one generation’s adolescence echoing in the preferences of their children, passed along, presumably, through kitchen radios and car stereos. The memory a song carries, in other words, may not even be your own first.
Familiarity told a similar story. Participants aged 42 to 55 rated songs that charted up to five years before they were born as significantly more familiar than their own average — music that was simply still in the air when they arrived.
A few tastes, meanwhile, seemed to belong to everyone. Liking for songs from the late 1970s and early 1980s ran higher than average across three of the four age groups, regardless of how old anyone was when those songs charted — and all four groups agreed, with similar coolness, about the hits of the early 2000s. Some eras, apparently, just wear better than others.
What this study does not show
This is one study, with honest limits, and the popular version of this finding routinely overstates it. A few corrections worth keeping.
It does not show that teenage music is remembered more vividly. Participants rated how many memories a song called up — the study collected no actual memories and measured nothing about their vividness or accuracy. The vividness claim belongs to older research on the reminiscence bump generally, not to this dataset.
It does not explain why. The study is cross-sectional — it compared people of different ages at one moment, in 2017, rather than following anyone through life. The authors themselves note that a cohort effect is possible: the oldest group’s later peak might reflect something about growing up in the 1950s rather than something about aging. Theories about adolescent brains, first loves, and identity formation exist in the wider literature, but this study tested none of them.
It is also a particular slice of humanity: French adults, 98 percent native French speakers, 79 percent with higher-education degrees, rating French chart pop. The authors acknowledge that chart hits represent only one of many genres that may matter to a person — the study has nothing to say about the hymns, film scores, or folk songs that may anchor someone else’s adolescence.
And it does not show that everyone is sentimental about their teens. The pull the study measured is statistical, visible across hundreds of people — individual lives vary, and the study makes no claim about any one person’s relationship to their seventeenth summer.
What this can and cannot do
What it can do: explain the cereal-aisle ambush, and perhaps soften a familiar intergenerational complaint. The parent whose memories live in the songs of 1974 and the grandchild whose memories live in 2014 are very likely both reporting the same thing — the view from around fourteen — and both are right about their own window.
It also suggests something gentle about sharing: the music playing in a household does not stop at the person who chose it. The cascading bump implies the songs parents play become, measurably, part of what their children later carry.
What it cannot do: serve as a therapy, a memory test, or a diagnosis. Earlier research cited in the study suggests music can help cue memories in people with conditions like Alzheimer’s disease; what the authors raise strictly as a possibility for future work is the idea of targeting the songs of adolescence in particular — and this study itself involved healthy adults only. A playlist is not an intervention, and worries about a parent’s memory — or one’s own — belong with a doctor, not a streaming service.
The window that stays open
The charts kept moving. The man in the cereal aisle moved with them, mostly — new decades, new formats, new songs that never quite stuck. But somewhere around fourteen, the research suggests, the volume of life was turned up, and what played then got written down differently.
Everything since has been music. What played at fourteen became memory.
