Music and memory: How the songs we heard growing up shape the story of our lives

ResearchBlogging.orgi-87c02757b578c9b709e8404461aea03c-janata0.jpgOne of the first things I did after my 90-mile hike with Nora in the North Cascades was play some music on the car stereo. We'd been in the wilderness for seven days, and other than birdsong, we hadn't heard so much as a note for the entire time.

Matching our intuitions about music, researchers have found that music is an important influence on our memories. We associate songs with emotions, people, and places we've experienced in the past. This isn't to say that music is the only influence on memory: the photos I took, the sights I saw, and the words I wrote about my hike will also help to preserve it in my mind for many years to come.

But it's not easy to parse out exactly how music evokes memories. If I listened to "Rock Lobster" on the drive down from Hart's Pass where we finished our hike, will "Rock Lobster" be associated with that memory, or with my birthday party in college where I danced wildly to the same song? Does music have a more powerful effect on memory than other influences, like images, words, or smells? We don't know, but a group led by Petr Janata has taken an important first step in understanding how music can affect memory.

The researchers collected over 1,500 "preview" clips from the iTunes Music Store's listing of the top 100 pop and R&B songs from each year over the past couple decades. The idea was to have a sampling of the songs college students were most likely to have heard while they were growing up. They recruited 329 students to listen to the clips. Each student heard 30 songs randomly selected from the most popular songs that came out when they were between the ages of 7 and 19 (so an 18-year-old would hear clips chosen from a different set of songs than a 29-year-old).

Each song was rated for familiarity and like/dislike, and then the students were asked if the song evoked any memories for them. They indicated what emotions they associated with the song, whether the memories were about person, place, or event, and what words they associated with the memory. Finally, for each memory, they were allowed to type in a description of what they recalled.

On average, the students recognized about half of the songs. So did the songs they were familiar with evoke stronger memories? This graph tells the story:


As the songs increased in familiarity, so did the strength of the autobiographical memories associated with the songs. Very familiar songs were more likely than not to be associated with a memory. Overall, about 30 percent of the songs elicited memories. Did these memories come from the songs of a particular period in the students' childhood? It's often seemed to me that people prefer the songs that came out when they were teenagers, even when they are 50 or 60 years old. Is this observation born out in Janata's team's data? Not exactly. Take a look at this graph:


The dashed line shows the total number of songs played per year in the students' lives. The solid line shows how many of those songs elicited memories. The proportion of songs presented to memories elicited remains roughly constant throughout the study period -- a song that came out when you were seven was just as likely to elicit a memory as a song that came out when you were seventeen.

While songs that were rated as "pleasing" were more likely to evoke memories than "not pleasing" songs, there was no consistent pattern as to which songs were pleasing -- it seems to come down to personal preference: songs you like are more likely to be associated with memories. But do the preferred songs evoke the memories, or do we like the songs because we associate them with memories? Since this is only a correlation, this study can't tell us.

This study also can't tell us how accurate the memories are -- it only gives us an idea of how music is associated with memories in the present. That's not at all irrelevant, since our memories, accurate or not, are how we experience the past. Still, as the researchers point out, it would be very interesting to compare the types of memories evoked by music to memories from other sources, like pictures or words. A large number of the memories in this study were about dances or cars -- places where people are likely to be listening to music. Maybe memories associated with pictures are completely different!

Petr Janata, Stefan Tomic, Sonja Rakowski (2007). Characterisation of music-evoked autobiographical memories Memory, 15 (8), 845-860 DOI: 10.1080/09658210701734593

More like this

If my twentieth high school reunion last year was any indication, we seem to hang on to the music we listened to as adolescents longer than any other time period. Everyone was dancing to "Purple Rain" and "Rock Lobster" like the music written in 1984 was the best ever written. A 1996 study…
Last week's Casual Fridays study elicited quite a few confused reactions. Listeners heard short music clips and then were asked to indicate which color they associated with each piece from a list of 13 colors. So were we able to come up with consistent results? Yes. For some of the clips, our 377…
WHEN talking about our feelings, we often use expressions that link emotions with movements or positions in space. If, for example, one receives good news, they might say that their "spirit soared", or that they are feeling "on top of the world". Conversely, negative emotions are associated with…
In the past, I've often wondered how journalists pick which studies to write about. The obvious answer is that they pick studies that will get readers or viewers, but given how little their stories correspond with the research they're writing about, it seems to me like they could pick any study and…

I don't understand the first graph. What is "memory association," and why is it highest for the "no association" variable?

Sorry if that was unclear. "Memory association" indicates the percentage of listeners indicating that they had a particular type of memory associated with a song. "No association" means they had no memory to go with the song. So nearly all listeners had no memories associated with unfamiliar songs, but when the songs were very familiar, most listeners had either strong autobiographical or somewhat autobiographical memories associated with those songs.

So, all the more reason to use songs in the classroom, yes? I know several science teachers who advocate using songs to teach science because the stuff seems to stick better in kids' heads. Here's a post with some sciencey songs:…

and a podcast about science songs (the density song, as performed by my coworkers at the Exploratorium, is priceless)…

It seems that Whistler's second questions is not well responded, and that's also what I want to know.

I can see why listening to music whilst we learn could sometimes help implant the memory in the brain better.... however I also know that for some types of clerical tasks, I cant pay attention if there is music playing, let alone remember what I was doing.....however, I would probably remember what music was distracting me...

ppl do associate or tag songs to their memory ....but can they exactly place what actually took place during the event of the song....may be not !....its a blurred figment of memory that they associate that particular song with.....may be songs actually act as illusionists by taking them back to their teenage and make them rejoice with this small figments of memeory!

ppl do associate or tag songs to their memory ....but can they exactly place what actually took place during the event of the song....may be not !....its a blurred figment of memory that they associate that particular song with.....may be songs actually act as illusionists by taking them back to their teenage and make them rejoice with this small figments of memeory!

Interesting article. I was thinking about this earlier this month and how certain songs do bring up specific memories for me based on 'strong' emotions I felt when I heard the song

I agree 110%. And I'm sure of this because I experience it on a daily basis. Truth is, doesn't matter whether the music is good or bad. It acts like a recorder on all possible levels, absorbs my perceptions and freezes them, just so put them on reply the next time I hear the same song. I can recall what I was seeing, smelling, thinking while I was listening to some chorus. Scary, actually. Like being in a time machine for a second and then back again.

Past couple of decades? If they went back further, they would probably find an even stronger correlation. In the 50s, 60s, and 70s, everyone listened to the same songs at the same time, before radio format fragmentation and the proliferation of alternate media. I've always said the songs that were Top 40 when a person goes through puberty and gets their first kiss and first heartbreak are the songs that will stay with you throughout your life. Baby Boomers will continue to support "classic rock" and "good time oldies" formats because there are enough of us.

I'm no scientist, but I was a radio disc jockey for decades.

Music totally Rocks. Amazing when a song comes on and it reminds me instantly of something from my past. Pretty cool indeed.


By Ron Dishman (not verified) on 19 Aug 2008 #permalink

I can't believe you needed to do a study to realize stronger memories are evoked by familiar songs than unfamiliar ones...DUH

Great article and graphs. I agree 100%. I like to listen to classic rock stations as they bring back a time when life was more relaxed. A good tune can mentally calm the nerves.

thanks for the article.

It's interesting to see someone else write this out. Songs have always brought up strong feelings of emotions and memories for me. I really do hate it so much because it's not only songs i heard years and years ago but recent songs as well (within the last 3 years).

For example,
If i hear Life Styles of the Rich and Famous, Bloody Valentine, or The Anthem, by Good Charlotte, I immediately think of my first summer job in high school. I would listen to the songs on the white school bus to work. I can remember everythign about it, the color of the seats, the sound the bus made, the sound of the air conditioner in the bus, and of course, feeling nervous about going into a situation I wasn't familiar with (first job) it's really crazy.

On the other hand, If I listen to Chasing Cars by Snow Patrol I think about driving around with my ex. It almosts shatters my heart every time. I feel like I can't listen to that song anymore without feeling the pain of that breakup. My memories have ruined such a great song.

My best friend made a video about 4 years ago and edited together a bunch of clips of me and my friends to the song, The Restless by The Matches and of course every time I listen to this song, my eyes tear up because I miss my friends so much and I can't help but worry about my best friend being in Iraq.

I guess this is a normal event for a lot of people.

I am delighted to see that someone is blogging about this study! I was an RA in the Janata lab when the study was being conducted, and the research always stuck in my head throughout the years as I compiled playlists to capture a certain memory, or a particular time in my life. To this day, that is the single best way I know of to recall a memory and to be brought back to one moment in life.

By Julia Udell (not verified) on 19 Aug 2008 #permalink

Does anyone know if there was any sort of adjustment for the length time a person might be exposed to songs relative to how likely they were to have memories associated? It would seem that while someone might not have formed a bond with a song that came out when they were seven, there's a good chance that song would have popped up from time to time through the years. Completely unscientific but I think of all the "80's, 90's and Now!" stations I run into on the dial and how many moms keep them alive.

Would love to read the full study. Great article.

I'm with Justin above. Maybe I'm missing something, but isn't it common sense that familiar songs are also the songs we have memories associated with? As Justin said, "DUH"!

Pretty weak premise for a study. What's next, a study to determine whether the people we're most familiar with are also people we have more memories associated with? Sheesh!

By Tinderbox (not verified) on 19 Aug 2008 #permalink

Catholic theologian Hans Kung listens to Beethoven when he's thinking and Mozart when he's writing. Beethoven is more Dionysian, filled with more searching and conflict. Mozart is more Apollinarian, filled with harmony and completion.

By Richard L.A. S… (not verified) on 20 Aug 2008 #permalink

I am the author of Training With A Beat, a well respected book for trainers, speakers and educators on the application of music in learning situations. This article definitely "sings." It is true the pull of music is ancient and deep on us. It is also the case that the audio signal is often MORE important than visual input. Thank you for reporting on this interesting study. Lenn MIllbower, The Learnertainment(r) Trainer

There is association and integration of memory inputs with the reward (and punishment) center. Moreover, memories aren't stored as such, but rather, it is first broken down and compared to 'similar experiences that is already present ' and then 'linked' to these traces, through association fibers.

Songs may usher memories in a way similar to that of a search engine 'indexing' and retrieving pages.

As such, scientists have used Mozart to improve memory. I would go for Metallica to ERASE my unpleasant memories! Great article!

#1, #2, #4,

I'm not sure, but I think the percentage at the memories associations axis doesn't state the percentage of listeners indicating having a certain type of memory for a song. If this would be the case, it should either be 100% for each familiarity segment (100% of songs should either evoke no memory or some memory), or there is a hidden type of memories not included in the graph (because, for unfamiliar songs for instance, 45% has no memories, about 5% has autobiographical memories, but what about the other 50% of listeners?).

Instead, I believe the percentage refers to the percentage of songs which fall into a certain familiarity segment (unfamiliar, familiar, very familiar) and evoke certain type of memories (no associations, somewhat autobiographical, strongly autobiographical)

On average, across all listeners, most songs evoke no memory associations, some evoke somewhat autobiographical associations, and only a few evoke strong autobiographical associations. More interestingly, those songs which do evoke autobiographical associations are the familiar songs, and especially the very familiar songs.

(And somewhat less fascinating :p, on average, more songs are rated as unfamiliar, only some as familiar, and even fewer as very familiar, which makes sense ^^)

This is what I think at least ^^ , but I'm not completely certain atm.

As far as the second question is concerned, I think Miss Cellania has got a good point about the fragmentation of music fans. Since this study was done with popular songs, the drop off in familiarity with these songs after 15 may just be due to the fact that after this age, people tend to be more particular about what kinds of music they listen to.

I remember when I was 14 or 15, I stopped listening to popular music altogether. So I have very strong memories related to the music I listened to when I was 16, but they were not the top R&B hits of that era.

I don't know that this holds true for everyone. First of all I never listened the music of my teen generation I grew up listening to my parents music which I still prefer. Secondly I can't stand listening to music when I am studying or focusing on a task, or trying to listen to someone, I find it annoying in the extreme!! I am also terrible at remembering music, I just don't have that great an interest in it. My memory is more linear, connected to dates and events that have nothing to do with music. Nor am I the only person those memory works like this, so I am not sure that all of this isn't just pie in the sky!!