science tries to explain music

Socializing and general posts on wide-ranging topics. Remember, it's Poststructural!
Post Reply
User avatar
chas
Posts: 7697
Joined: Wed Oct 10, 2001 6:00 pm
antispam: No
Please enter the next number in sequence: 10
Location: East Coast US

science tries to explain music

Post by chas »

I got a link to this in my inbox this morning:

https://www.newsweek.com/what-makes-pop ... er-1470073

I'll put my thoughts, which includes the spoiler, a little further down.















I find the conclusion bizarre. They're saying that surprises and uncertainty make a good pop song. But almost by definition, a pop song is the most predictable music there is. More specifically, they say that expecting one thing to happen, and hearing another, is the most pleasurable. So that means they're not talking Schoenberg, but to take a couple of examples, Yes and Genesis didn't have pop success till they dialed down the variations a lot.

I'm interested in other folks' thoughts.
Charlie
Whorfin Woods
"Our work puts heavy metal where it belongs -- as a music genre and not a pollutant in drinking water." -- Prof Ali Miserez.
User avatar
Peter Duggan
Posts: 3223
Joined: Tue Aug 30, 2011 5:39 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: I'm not registering, I'm trying to edit my profile! The field “Tell us something.” is too short, a minimum of 100 characters is required.
Location: Kinlochleven
Contact:

Re: science tries to explain music

Post by Peter Duggan »

'Next, the team stripped the songs of melody and rhythm, leaving only the chords.'

Not clear what that means when I'd regard harmonic rhythm as integral to the effect of a chord progression...

Since there's potentially a huge difference between leaving the original textures, voicings and timings and just abstracting the chords in some other form, I think we'd need to know which. Not to mention that our enjoyment of music is surely governed by the piece (or song) as a whole rather than some isolated component, even when clearly influenced by the latter. And, since the best music typically gets better and better with repeated listenings, by which time you're going to be eagerly anticipating your favourite bits including 'surprises', I'm not convinced by what I read there. Of course that could be the fault of the report rather than the study, which I've glanced at but am not currently motivated to read, but I wonder if they're really measuring what they think they're measuring?
And we in dreams behold the Hebrides.

Master of nine?
User avatar
Nanohedron
Moderatorer
Posts: 38202
Joined: Wed Dec 18, 2002 6:00 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Been a fluter, citternist, and uilleann piper; committed now to the way of the harp.

Oh, yeah: also a mod here, not a spammer. A matter of opinion, perhaps.
Location: Lefse country

Re: science tries to explain music

Post by Nanohedron »

Peter Duggan wrote:... I wonder if they're really measuring what they think they're measuring?
That's my question, too.

I get the impression that rather than findings, they're making assumptions based on some pretty hazy preconceptions: Why, for example, does the study equate surprise with pleasure? There's such a thing as unpleasant surprise, too. And sometimes it even works: "Let It Be", for example, has a somewhat jarring chordal variation toward the end, but it succeeds in its way, and having matured and gotten used to it, I now look forward to it as a well-placed oddity, but it still jars, so the label of "pleasure" is very open to interpretation. I don't believe the variation was at all essential to the song's success.

Of course the study's mathematical formulae are way over my head, so I would have been grateful for accompanying audio examples to illustrate their points, but instead, only at the very end of the report just before the references (as an afterthought, it would seem) there is but one - ONE - MIDI sound sample provided (here) that bears no discernible relationship to ABBA's "Knowing Me, Knowing You" as I recognize it. They tell me that's what it is, but listen for yourself and come to your own conclusions. If they've pared it down to essentials, I fail to detect what those essentials, or the study's criteria for them, would be. Supposedly the sample relates to Figs. 1 and 2, but why then is the sample tucked away at the dusty, lonely end of the report, far from the meat of the matter? Seems a bit conveniently placed if you want it to be ignored. Given its glaring musical demerits, perhaps that explains its placement in the overall composition of the report.

"Knowing Me, Knowing You" would be a great example of surprise elements, because the song's chock full of them. But it would be far better to analyze the song as it is, rather than deflate it down to what appear to be arbitrary, unrecognizable, and unconvincing points of reference. Personally, I think the researchers wouldn't know what music was if it hit them in the face.

I think it's a lost cause to look for hard scientific answers as to why some music is more popular than others. Indeed, how are we even going to define "popular"? Numbers alone? What about cultural context? And what about the phenomenon of herd mentality? Some songs are popular for no good reason, to my thinking.
"If you take music out of this world, you will have nothing but a ball of fire." - Tribal musician
User avatar
Peter Duggan
Posts: 3223
Joined: Tue Aug 30, 2011 5:39 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: I'm not registering, I'm trying to edit my profile! The field “Tell us something.” is too short, a minimum of 100 characters is required.
Location: Kinlochleven
Contact:

Re: science tries to explain music

Post by Peter Duggan »

Nanohedron wrote:They tell me that's what it is, but listen for yourself and come to your own conclusions.
Oh dear, oh dear, oh dear... what did I say about harmonic rhythm and and just abstracting the chords in some other form?

Afraid your slightly closer look at what they've done simply backs up all my concerns and confirms my suspicions about the value of the whole exercise!
And we in dreams behold the Hebrides.

Master of nine?
User avatar
Nanohedron
Moderatorer
Posts: 38202
Joined: Wed Dec 18, 2002 6:00 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Been a fluter, citternist, and uilleann piper; committed now to the way of the harp.

Oh, yeah: also a mod here, not a spammer. A matter of opinion, perhaps.
Location: Lefse country

Re: science tries to explain music

Post by Nanohedron »

Peter Duggan wrote:Afraid your slightly closer look at what they've done simply backs up all my concerns and confirms my suspicions about the value of the whole exercise!
Well, I too had my doubts about the whole premise, so my mission was simply to find relevant audio examples - not too much to ask for, one should think - but the more I had to comb through all the fine-feathered gobbledygook without finding any, the less I thought that the study should ever be taken seriously, because it was looking more and more like all hat and no cowboy. And then, at the very, very end, what should my crawl through the desert of arcana be rewarded with? A big, steaming pile of B.S. And I almost missed it, too. Thanks loads, boffins.

Mind you, I'm always one to look to science first, but I also recognize that anything can go off the rails, and for my money, these folks sure did it.
"If you take music out of this world, you will have nothing but a ball of fire." - Tribal musician
Tunborough
Posts: 1419
Joined: Sun Dec 05, 2010 2:59 pm
antispam: No
Please enter the next number in sequence: 10
Location: Southwestern Ontario

Re: science tries to explain music

Post by Tunborough »

What's the point of using popular songs if all you're doing is extracting isorhythmic chord progressions from them. As you noticed, the chord progression says very little about the song it comes from. They might just as well have used computer-generated chord progressions. Not to mention that all of the songs were written before over half the study participants were born.

Here's another news report on the subject.
The problem is that there are often external factors that are difficult to control that could affect the results of these studies — including the background, age and musical experience of the participants themselves.

In the case of the Current Biology study, the songs that were selected range in date from 1958 to 1991, but the median age of the study's participants was around 25.
This article notes a different study from McGill University. They at least played real music, and asked people whether they liked it. Preferences were correlated with music of intermediate predictability and intermediate uncertainty, neither too low nor too high. They also noted a relationship between uncertainty and predictability that, "suggested preferences for more predictability during more uncertain contexts." I take this to mean we can cope with uncertainty, but we like it better when we get what we expected in the end, analogous to "happily ever after" in stories.
User avatar
Nanohedron
Moderatorer
Posts: 38202
Joined: Wed Dec 18, 2002 6:00 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Been a fluter, citternist, and uilleann piper; committed now to the way of the harp.

Oh, yeah: also a mod here, not a spammer. A matter of opinion, perhaps.
Location: Lefse country

Re: science tries to explain music

Post by Nanohedron »

Tunborough wrote:They also noted a relationship between uncertainty and predictability that, "suggested preferences for more predictability during more uncertain contexts." I take this to mean we can cope with uncertainty, but we like it better when we get what we expected in the end, analogous to "happily ever after" in stories.
How many times have I been asked by patrons to play songs that were more familiar to them?

Sometimes it seems a study can't see the forest for the trees. This might have been one of those cases.
"If you take music out of this world, you will have nothing but a ball of fire." - Tribal musician
User avatar
chas
Posts: 7697
Joined: Wed Oct 10, 2001 6:00 pm
antispam: No
Please enter the next number in sequence: 10
Location: East Coast US

Re: science tries to explain music

Post by chas »

Nanohedron wrote: I think it's a lost cause to look for hard scientific answers as to why some music is more popular than others. Indeed, how are we even going to define "popular"? Numbers alone? What about cultural context? And what about the phenomenon of herd mentality? Some songs are popular for no good reason, to my thinking.
Agree with all you and Peter have said, but this paragraph captures my thinking, especially the bolded part. I think they're trying to know the unknowable, possibly verging on that which shouldn't be known. Trying to quantify and define beauty is kind of un-beautiful.

In the words of Albert Einstein: “It would be possible to describe everything scientifically, but it would make no sense; it would be without meaning, as if you described a Beethoven symphony as a variation of wave pressure.”
Charlie
Whorfin Woods
"Our work puts heavy metal where it belongs -- as a music genre and not a pollutant in drinking water." -- Prof Ali Miserez.
PB+J
Posts: 1306
Joined: Sun Mar 25, 2018 5:40 am
antispam: No
Please enter the next number in sequence: 8
Tell us something.: I'm a historian and the author of "The Beat Cop:Chicago's Chief O'Neill and the Creation of Irish Music," published by the University of Chicago in 2022. I live in Arlington VA and play the flute sincerely but not well

Re: science tries to explain music

Post by PB+J »

from an idealist point of view, things are good as they approach an ideal

From a realist point of view, things are interesting when they are simultaneously like and not like

So a sunset is a appealing because it's simultaneously like and not like other sunsets. One version of "Chief O'Neill's Favorite" is good and interesting because it's just like but also not like, all other versions, in surprising ways. "Surprise" is something most of us aren't to pleased with, most of the time, because surprises are often unpleasant, but Claude Shannon, the Father of "Information Theory" and digital computing, argued that there is no information without surprise. In English, the "U" following the "Q" contains no information, because there's never any surprise or uncertainty in its appearance.

So a certain amount of surprise is the key. Many people who know nothing about Irish traditional music say "it all sounds alike," but the more you know about it the more you are delighted by small surprises in the melodic line.
User avatar
Nanohedron
Moderatorer
Posts: 38202
Joined: Wed Dec 18, 2002 6:00 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Been a fluter, citternist, and uilleann piper; committed now to the way of the harp.

Oh, yeah: also a mod here, not a spammer. A matter of opinion, perhaps.
Location: Lefse country

Re: science tries to explain music

Post by Nanohedron »

chas wrote:Trying to quantify and define beauty is kind of un-beautiful.
It's like performing an autopsy to explain being. I suppose it's always worth a try, though, if for no other reason than to underscore the reality that in the end, aesthetics and calculation serve different perspectives that have little bearing upon each other, except incidentally. Now bear in mind that my thoughts on this are just my own, and they are merely intuitive; while my stance is by no means anti-intellectual, I am also no mathematician, so do feel free to point out my errors and faults of logic: Let us say that there are two formulae that rightly arrive at the same conclusion; if one is more elegant, it doesn't mean the other's conclusion is less true. Yet we are attracted to elegance, so elegant formulae are preferred. Why? I suggest the reasons are subjective - not that there's anything wrong with that. Conversely, composition in an artistic work may be readily found and plotted, and it may explain what makes the work strong or weak, but even unattractive works can have strong composition; composition alone cannot explain why a piece works, or is even praised, for that matter. The theme of famous artists being unappreciated in their own lifetime is not based on fantasy. And we know very well that what was valued at one time can fall out of favor in another. How, then, are we to measure in any absolute sense why something succeeds or fails? Can it even be predicted? Apart from making educated guesses, I personally doubt it.

On the opposite end of the spectrum, I've always thought that Keats grossly overstepped when he wrote, "Beauty is truth, truth beauty." Now, I fully realize that it was the product of a time when the idealized was often considered superior to the real, but even that does not absolve Keats: thanks to its frankly startling premise, the popular mind found in it a bright, shiny object. Other poets have debated whether it is a blemish, and continue to do so; its being "in character" with the poem seems to be its main defense. As a reader, I deemed it a reckless abuse of poetic license, because the concept being so memorable and from a lauded poet, people reverently quote or paraphrase it as if it were profound and to be taken as an esoteric article of faith; in reality, though, the Emperor has no clothes. It doesn't take much to see that it's actually glib nonsense, and we shouldn't be afraid to say so. In fact, I'll bet Keats himself was well aware that it was utter rubbish; how could a writer of his parts not be? We enshrine it only because of who penned it, but without Keats' name to cloud our judgment, anyone should be able to see it for the didactic gibberish it really is, unworthy of the currency it enjoys. If I wrote it, I would have been laughed out of literature. Lack of a muse aside, I thankfully also have my self respect to prevent me from such rococo frothings, except maybe as a joke. And maybe the formula was indeed intended as a joke, only we didn't get it. We'll probably never know. Instead, we just squirm in its shadow and mutter, "Erm, yes, well ... it's mysterious, but it's Keats, and surely he was wise." I don't have that kind of patience. Sorry if I've raised any hackles by giving poor old defenseless Keats a pounding (well-deserved, IMO), but consider: Truth may indeed be beautiful, but not every time; how else do we speak of "the ugly truth" without any sense of contradiction? Rather than opening a door, Keats' trite but catchy bromide invites us to put on blinders in the midst of a mirage that illuminates nothing. That's what makes me suspect it may have been a joke all along. As for "beauty is truth": not even close. Beauty may be collectively agreed upon, but it is nonetheless entirely subjective - I'm reminded of a fellow who swore that the Hollies were superior to the Beatles in every way - so how can what is not universal ever be truth? One cannot explain attraction so easily; we are also fascinated by horrors. I've seen it for myself. Some people, due to their value system, are repelled by fine art because they find it vapid. I've seen that for myself, too. One language's conventions for good prose composition can be contradictory to another's. And you guessed it: I've seen that for myself, as well. Am I not to believe my own eyes, but live instead in hope of the day I might come to understand Keats' opaque yet curiously esteemed mantra? I say that day will never come for anyone. It's opaque because there's nothing more to be understood than what's right in front of you: just decorative words that sound nice because they're rosy. There is no truth in them; try to find it.

Clearly, aesthetics are personal and sociological. This very rant is a good example of the personal: not everyone will agree with me, even though they'd be wrong. :wink: Cosmetic scarification is a good example of the sociological: where one culture valorizes it, another finds it an abomination. One cannot accurately calculate for any of it, because neither can one say what beauty really is.
chas wrote:In the words of Albert Einstein: “It would be possible to describe everything scientifically, but it would make no sense; it would be without meaning, as if you described a Beethoven symphony as a variation of wave pressure.”
I like that. :)
"If you take music out of this world, you will have nothing but a ball of fire." - Tribal musician
User avatar
daveboling
Posts: 4868
Joined: Sun Feb 17, 2002 6:00 pm
antispam: No
Please enter the next number in sequence: 8
Location: Huntsville, AL

Re: science tries to explain music

Post by daveboling »

Years ago, I read an interview (could have been in Downbeat) with Paul Desmond, who played alto sax on many of the Dave Brubeck Quartet's best recordings, and wrote a few of them. When asked about improvisation, he said he had found that the improvised melody was best received when about 50% of the melody was predictable to the ear of the listening audience, and the other half a surprise. I have no idea how he determined this, but he said his belief in this colored how he structured the melody line.

dave boling
I teleported home one night
With Ron and Sid and Meg.
Ron stole Meggie's heart away
And I got Sidney's leg.
-- Douglas Adams

'Bundinn er bátlaus maðu'.
PB+J
Posts: 1306
Joined: Sun Mar 25, 2018 5:40 am
antispam: No
Please enter the next number in sequence: 8
Tell us something.: I'm a historian and the author of "The Beat Cop:Chicago's Chief O'Neill and the Creation of Irish Music," published by the University of Chicago in 2022. I live in Arlington VA and play the flute sincerely but not well

Re: science tries to explain music

Post by PB+J »

daveboling wrote:Years ago, I read an interview (could have been in Downbeat) with Paul Desmond, who played alto sax on many of the Dave Brubeck Quartet's best recordings, and wrote a few of them. When asked about improvisation, he said he had found that the improvised melody was best received when about 50% of the melody was predictable to the ear of the listening audience, and the other half a surprise. I have no idea how he determined this, but he said his belief in this colored how he structured the melody line.

dave boling

I think that's exactly right: improvisation often works by exactly this kind of tension between the expected and the unexpected. I've been teaching myself "Chief O'Neill's favorite" and as usual there are twenty different versions online, all somewhat different. I enjoy hearing the variations once I have a version in my head that I think is authoritative
User avatar
Nanohedron
Moderatorer
Posts: 38202
Joined: Wed Dec 18, 2002 6:00 pm
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Been a fluter, citternist, and uilleann piper; committed now to the way of the harp.

Oh, yeah: also a mod here, not a spammer. A matter of opinion, perhaps.
Location: Lefse country

Re: science tries to explain music

Post by Nanohedron »

PB+J wrote:
daveboling wrote:Years ago, I read an interview (could have been in Downbeat) with Paul Desmond, who played alto sax on many of the Dave Brubeck Quartet's best recordings, and wrote a few of them. When asked about improvisation, he said he had found that the improvised melody was best received when about 50% of the melody was predictable to the ear of the listening audience, and the other half a surprise. I have no idea how he determined this, but he said his belief in this colored how he structured the melody line.

dave boling
I think that's exactly right: improvisation often works by exactly this kind of tension between the expected and the unexpected. I've been teaching myself "Chief O'Neill's favorite" and as usual there are twenty different versions online, all somewhat different. I enjoy hearing the variations once I have a version in my head that I think is authoritative
I agree with this general principle too. A certain amount of predictability gives a familiar, even comforting frame of reference; something to hang your hat on, as I've put it before. Pure, wandering improvisation can be impressive and even admired, but I don't believe it has as much audience appeal as when a thematic root can be counted on to add dimension to the mix. You get a sense of give-and-take, of departure and return.
"If you take music out of this world, you will have nothing but a ball of fire." - Tribal musician
Post Reply