Non-musicians can identify minor-key tunes, but only when labeled "sad"

ResearchBlogging.orgIf you've had a lot of musical training, you can probably tell the difference between a major and minor key. If you haven't had much training, even after having the difference explained to you, you're still not likely to be able to make that determination. Listen the following clip. It plays the same melody in a major and a minor key. Can you tell which is which?

But if the question is phrased differently, even non-musicians can reliably tell the difference: When listeners are told that some music (which happens to be in a major key) sounds "happy" and other music (in a minor key) sounds "sad," non-musicians can pick out the difference. With that information in mind, do you want to change your answer about the two samples above? If you do, you're probably a non-musician. If you don't, you either got lucky in your answer, or you are a musician. Either way, it's clear that musicians process "major" and "minor" differently from non-musicians. So what's different about the mental processing of musicians and non-musicians?

A team led by Andrea Halpern created 35 short tunes like the clips above. Each tune was then modified to have a minor-key and major-key variant -- this involved changing just a few notes in each tune. Then three expert musicians rated each clip for musicality and how "major" or "minor" each clip sounded. The 24 best examples of tunes with readily-identifiable major and minor keys were selected for study.

The researchers then played the clips for 18 musicians, with over 8 years of musical training, and 18 non-musicians, with less than 6 months of training. The listeners were first briefly trained either to recognize major/minor keys or "happy" and "sad" music. Several examples of each type were played, and then for seven clips, the listeners were asked to identify whether it was major or minor (or happy or sad), and given feedback on whether their answers were correct.

Finally, they were tested on new clips, while wearing a cap which had 30 electrodes designed to measure electrical activity in the brain. The data was collected on an EEG device.

Regardless of whether they were trained to recognize happy and sad music or major and minor keys, the musicians were extremely accurate, averaging about 90 percent correct. But non-musicians had a different result:


When they were trained in the major-minor method, non-musicians averaged just 63 percent correct for tunes in a major key. What's more, after the experiment, participants were asked what strategy they used to respond, and many non-musicians said they just listened for whether the music sounded happy or sad. Those who used this strategy averaged 75 percent accuracy, while the remainder who used no strategy were just 53 percent accurate -- statistically no better than chance.

Meanhwile the musicians nearly all reported using a major/minor strategy, even when trained in the happy/sad approach. Interestingly, this result can be clearly seen in the data from the EEG:


This figure shows a topographic map of electrical activity from 500 - 600 milliseconds after the first diverging note played in a tune (e.g. the note that distinguishes a minor-key from a major-key). As you can see, regardless of the training they received, non-musicians' brain activity doesn't look much different for minor tunes and major tunes. But for musicians, again regardless of their training, there was a dramatic difference in activity in the temporal and parietal lobes of the brain (the lower back area of your head).

The researchers suggest that the lack of activity in musicians during major tunes may be due to the fact that most Western music is played in a major key: 97 percent of popular American songs, and 73 percent of classical music is in a major key. So the activity occurs when an relatively unexpected key is used -- but only for trained musicians.

Halpern, A.R., Martin, J.S., Reed, T.D. (2008). An ERP Study of Major-Minor Classification in Melodies. Music Perception, 25(3), 181-191. DOI: 10.1525/mp.2008.25.3.181

More like this

Take a listen to this brief audio clip of "Unforgettable." Aside from the fact that it's a computer-generated MIDI performance, do you hear anything unusual? If you're a non-musician like me, you might not have noticed anything. It sounds basically like the familiar song, even though the…
Listen to these two short music clips. Music Clip 1Music Clip 2 Now, can you identify the musical style of each clip? If you said "Classical," you're technically only correct for the first clip. The second clip is actually in the Romantic style (bonus points for identifying the works and composers…
Listen to these two musical excerpts and note any differences you discern: Ave Maria, version 1Ave Maria, version 2 (Source: courtesy of Mayumi Hamamoto and Kyota Ko) If you're a typical nonmusician, you will probably notice some sort of difference between the two excerpts. Maybe one seems to be…
[originally posted April 6, 2005] Listen to these two musical excerpts and note any differences you discern: Ave Maria, version 1Ave Maria, version 2 (Source: courtesy of Mayumi Hamamoto and Kyota Ko) If you're a typical nonmusician, you will probably notice some sort of difference between the two…

Minor, major, in that order. And I have zero musical skills. (Although I like the blues.)

A Korean friend told me Korean music is inherently sad because it can be played on the black keys on a piano.

I said "So can boogie-woogie."

By Nelson Muntz (not verified) on 05 May 2008 #permalink

I wonder if anyone has done a study to see whether traditional folk music from different parts of the world is predominantly major or minor. I imagine a colored major-minor world map... Would be interesting to see!

The major message I'm hearing is that 97% of popular American music, and 73% of classical music is boring. And it's about time somebody else noticed this obvious fact... what? Oh, yes, that's right: I'm the one whose brainscan is in the minor key. So, never mind.

By Matt Platte (not verified) on 05 May 2008 #permalink

In general (assuming "mom and dad" didn't force you), people gravitate towards activities they are successful in. Put another way, most musicians had a natural aptitude which sustained their interest in musical training. Given this, do the differences in brain activity reflect differences in training vs. no training or do they reflect an inherent neurological difference between the two groups, the same difference that results in musical aptitude for one person and lack of aptitude in another?

Hope that makes sense.

"If you haven't had much training, even after having the difference explained to you, you're still not likely to be able to make that determination."

Oh. Oops. Guess I shouldn't have been so cranky with my significant other on that point then. (I always thought it was obvious...I had only a handful of piano lessons and we never got to the discussion of major vs. minor or anything.)

Maybe I should take lessons again. ;)

By wintersweet (not verified) on 05 May 2008 #permalink

"If you haven't had much training, even after having the difference explained to you, you're still not likely to be able to make that determination."

Who says? I have had essentially no musical training & I have little difficulty telling minor from major. Maybe I'm a savant! :)

Middle Eastern music doesn't have "major" and "minor". Instead, they have at least 10-20 "maqams" which are in common usage and Turkish claims well over 100. It's been shown that even people who grew up with the music, however, will still claim that any maqam which isn't "happy" sounding must be "saba" (one of their more "sad" sounding maqams). They can't tell the differences any better than we can. :)

(Major disclaimer: I know very little about music). That said, as I understand it the minor/major distinction is very much cultural. For example, in traditional Nusach (tunes used in Jewish services) minor is much more common and major and has not much connection to whether the music is happy or sad.

Wait a second...playing all black keys on the piano is a pentatonic scale: a major third, not a minor third. It should sound happy. (most children's songs are in a pentatonic "Mary Had A Little Lamb", etc.) I wonder why Koreans think that sounds sad? Must be a cultural thing within the context of song form. Were the subjects in this experiment all American/European?

Anyway, I have to say that the brain scans here were fascinating! I took a break from doing music for a few years after a couple of decades of formal training. When I picked it up again, I could actually feel dormant areas of my brain starting up again in roughly the area shown in the "minor" areas on the scans above. How cool! (I know you're not supposed to be able to feel your brain, but I can somehow in certain situations.)

Grrrrrr.... This crap is so infuriating.

"Major" and "minor" are historically and culturally bounded concepts. They cannot possibly have anything to do with basic human cognition. You might as well draw conclusions about language cognition by comparing American and British spelling. The vast majority of the music ever made cannot be classified as either major or minor. It's a distinction that applies only to a tiny subset of music, made in a limited time and place, which has been made unnaturally prominent by accidents of history.

And you don't need to go so far afield as gamelan or Australian clapping-stick music to find counterexamples. Consider the Christmas carol "God Rest Ye Merry Gentlemen." Joyful lyrics, yet the melody could be considered "minor." Yet it's not minor, it's Aeolian mode, which shares the same notes as a minor key, but carries none of the purely cultural baggage associated with minor keys.

Are scientists studying music cognition willfully ignorant? This isn't difficult stuff to understand.

Interesting - but I think I'm missing something.

Certainly the brain-scans seem to support the idea that there's something different going on in musicians' and non-musicians' brains. But I don't get the initial example.

If a non-musician doesn't even have enough of a concept of minor chords to know they're the 'sad' ones, how could they be guessing which chords are minor or major? From how it's described, it seems like testing how well someone with a small vocabulary can identify obstreperous people ... then telling them what it means, and seeing how much better they do! Right?

I wonder if anyone has done a study to see whether traditional folk music from different parts of the world is predominantly major or minor.

As HP has already pointed out, the answer to that question would be "No". Most traditional folk music doesn't even use our "normal" tonal system (the 12-tone equal temperament).

"So the activity occurs when an relatively unexpected key is used"

I have to disagree with "unexpected". Minor keys are just modes of major keys so there is no unexpected about it.

Segovia sort of nailed this when he explained how he tuned up. He was very critical of the common forms of tuning (5/4 frets, or harmonics) and used to tune using minor chords as they were more "pure" (Guitar Magazine, circa 1973)

It sounds different. Minor chords sound much more harmonious, majors are slightly discordant. And I'm speaking as someone who plays in minor keys all the time, so there is nothing "unexpected" about them at all.

"Major" and "minor" are a division that you can date to around 1700. Before then there were a series of 7 modes (seven-tone scales) that western music was based on. The musicians and composers of the baroque period chose two of those, Ionian and Aeolian, to be the two primary modes which became what we call "major" and "minor".

Where there is no big historical basis for a "major" and "minor" phenomenon, it's a good measure of how nonmusicians reason musically. I would be interested to see a similar study on the old church modes, as they sit on something of a happy-sad spectrum.

Those criticizing the study for allegedly ignoring the cultural boundedness of our emotional associations with major/minor keys are missing the main point of the study. Yes, other cultures don't have the same emotional associations with the tonalities of Western art music. But it's still notable that those accustomed to such tonalities have definite patterns of brain activity in response to the two most often used modes in Western art music. And it's still notable that non-musicians have significantly different responses, at least on the brain activity level, to those modes. I think Halpern et al are fully aware that their findings apply to humans who have been acclimatized to certain tonalities. But they are interested in finding out more about the mechanisms behind such acclimatization, which is why the pattern of responses they find in musicians is interesting. Even if people from other cultures would not have the same brain responses, their findings still say something about how people respond to harmonies they have been acclimatized to.

"There's no love song finer,
But how Strange
the Change
From major to minor..."

Apologies. Someone had to do it.

Interesting study - I like musical studies - but I think the use of the "happy" and "sad" terms might be a little subjective.

Given the locations of the brain activity showed for the two groups, might this difference also (or instead?) reflect the difference between "expert" and "novice" processing? As I understand it, as one becomes an expert, most related processing occurs in the specialized areas of the brain devoted to that skill. On the other hand, people who have little experience with a skill have to activate the "attention" centers, ie, frontal lobes, to "consciously" analyze the experience. I wonder if one might not find a similar distribution of brain activity when comparing, say, an adult and a six-year-old both reading a passage from a fourth-grade reader. The adult would simply "recognize" the text, while the six-year-old would have to puzzle out many of the words individually, which requires frontal lobe involvement.

although there is certainly a lot of diversity in world music, it is notable that the notes of the major triad are not just arbitrary, but consitute the first several notes of the overtone series. So there is some basis for the major scale as a building block of music in physics (less so for the minor scale). The interesting question is whether our brains somehow naturally perceive this or whether our response to "major" is completely culturally determined.

also, slightly OT, but I think it's an oversimplification to say that major and minor came directly from the ionian and aeolian modes. In particular, aeolian is not all that similar to minor as it is used in functional diatonic harmony, where major 6ths and 7ths (which are not found in aeolian) are very common. In fact, I would guess that the dorian mode (which contains a major 6th) is at least as much a precursor to minor as aeolian.

That's very interesting. I'm 16, and when I started learning the piano at 14, I distinguished (once we got onto scales and keys and stuff) between them automatically as "happy" and "sad". I've since learnt that that's a bit arbitrary, but I know a lot of musicians that would say the same.

Not to say that all music in a minor key sounds sad. Just that a typical minor chord does :)

PS: I'm on to play Mozart's Fantasia in D minor. I'm not sure that you'd call it a "sad" piece, either! Perhaps "not cheerily happy" might be a better term ;)

The minor melody is much tastier than that goofy major melody (and I don't call them "the same melody" because they are two similar yet different sets of notes, after all).

See Sean O'Nuallain "Search for mind" 1995 and third edition (2003, P 234). To establish keys, musicians in the real world would insist on listening to much more of the piece

Minor, major. I had enough musical training to tend to dissociate from the "happy, sad" notion, since so much with music depends on context rather than the actual "sound" per se.

By medrecgal (not verified) on 28 May 2008 #permalink

Ninety seven percent? From what orifice did you pull that "fact"?

You were doing okay until you laid that whopper on us.

Keglined: I don't have the article with me, but I'm pretty sure Halpern et al. cite research to support that claim. I'll look it up and get back to you.

And rather than just expressing your enthusiastic cynicism, your comment would probably be taken more constructively if you provided some evidence to back your own assertion.

Okay, I looked up the article in my files. The 97 percent figure comes from a search of the HumDrum toolkit. You may be right to be skeptical: the figure is based on just 107 "American popular songs" in the database. Compared to the 9,806 "instrumental classical themes" identified, that's not much. But still, it's quite clear that much of Western music is in a major key (for example, 7,183 of those 9,806 classical pieces). Do you have evidence to dispute this, Keglined?

The statistics above does not imply that non-musician can identify minor-key tunes. Many sad songs are written in minor key, and that's why the statistics look convincing. I believe if all musicians until now compose sad songs in major key and happy songs in minor key, the topic will become "...only when labeled 'happy'".