On the origin of music by means of natural selection

Jun 18, 2012

Do away with the DJ and scrap the composer. A computer program powered by Darwinian natural selection and the musical tastes of 7,000 website users may be on the way to creating a perfect pop tune, according to new research published today in the journal Proceedings of the National Academy of Sciences (PNAS).

Scientists from Imperial College London have devised a way of producing from noises without a composer. They programmed a computer to produce loops of random sounds and analyse the opinions of musical consumers, who decided which ones they liked. The result is music filled with many of the sophisticated chords and rhythms familiar from modern songs.

The results could also help explain why popular musical trends continuously evolve and why traditional musical forms can persist for thousands of years.

This video is not supported by your browser at this time.
Listen to Dr Bob MacCallum explaining the evolution of evolution of musical loops created by DarwinTunes, accompanied by a selection of loops from the website

The scientists set out to test a theory that cultural changes in language, art and music evolve through Darwinian , in a similar way to how living things evolve. They simulated this cultural evolution by harnessing the power of a 7,000 strong in an experiment that was designed to answer several questions. Can music exist without being the product of a conscious, creative act? If so, what would that music sound like? Does everyone's ideal tune sound the same?

Armand Leroi, co-author of the research and Professor of Evolutionary Developmental Biology from the Department of Life Sciences at Imperial College London, said: "Everyone 'knows' that music is made by traditions of musical geniuses. Bach handed the torch to Beethoven who gave it to Brahms; Lennon and McCartney gave it to the Gallaghers who gave it to Chris Martin. But is that really what drives musical evolution? We wondered whether consumer choice is the real force behind the relentless march of pop. Every time someone downloads one track rather than another they are exercising a choice, and a million choices is a million creative acts. After all, that's how natural selection created all of life on earth, and if blind variation and selection can do that, then we reckoned it should be able to make a pop tune. So we set up an experiment to explain it."

The computer algorithm behind the study, called DarwinTunes, maintains a population of 100 loops of music, each eight seconds long. Listeners scored loops in batches of 20 on a five-point scale from 'I can't stand it!' to 'I love it!'. DarwinTunes then 'mates' the top ten loops, pairing them up as 'parents' and mingling musical elements of each pair, to create twenty new loops. These replace the original parents and the less pleasing non-parents. This process represents one 'generation' of musical evolution. At the time of publication, DarwinTunes had evolved through 2,513 generations.

The scientists then tested the like-ability of loops from different generations by asking listeners to rate them in a separate experiment. Without knowing the generational age of the loops, the volunteers consistently ranked the more evolved music as more appealing, thus independently validating the assertion that the music was improving over time.

Dr Bob MacCallum, another co-author and a mosquito genomics bioinformatician in the Department of Life Sciences at Imperial College London, said: "We knew our evolutionary music engine could make pretty good music in the hands of one user, but what we really wanted to know was if it could do so in a more Darwinian setting, with hundreds of listeners providing their feedback. Thanks to our students' and the general public's valuable input, we can confidently say it does."

Members of the public can continue to help the music evolve, by taking part in the DarwinTunes experiment at darwintunes.org. Individual loops can also be downloaded and used as ringtones or for offline music making.

Explore further: Researchers developing algorithms to detect fake reviews

More information: MacCallum R, Mauch M, Burt A, Leroi AM, "Evolution of music by public choice" Proceedings of the National Academy of Sciences www.pnas.org/cgi/doi/10.1073/pnas.0709640104

Related Stories

Composing Music For the Next Generation

Nov 01, 2006

Georgia Tech professor and composer Jason Freeman is bringing his musical passion to the Internet and letting his audiences shape the music they’ll hear in performance. The Graph Theory Project is an online ...

Scientists study music hallucinations

Jul 12, 2005

Psychiatrists at St. Cadoc's Hospital in Wales have issued the largest case-series study ever published concerning musical hallucinations.

Music therapy may improve schizophrenia symptoms

Nov 01, 2006

Music therapy for psychiatric in-patients with schizophrenia can improve some of the symptoms of the disorder, according to a new study by researchers at Imperial College London and therapists at the Central and North West ...

Recommended for you

HP supercomputer at NREL garners top honor

50 minutes ago

A supercomputer created by Hewlett-Packard (HP) and the Energy Department's National Renewable Energy Laboratory (NREL) that uses warm water to cool its servers, and then re-uses that water to heat its building, has been ...

Turner channels removed from Dish amid pact spat

1 hour ago

Channels such as Cartoon Network and CNN are no longer part of Dish's programming lineup as a deadline has passed for the satellite TV provider and Turner Broadcasting to renew their distribution agreement.

Google's streaming music service adds mood to mix

1 hour ago

Google's music-subscription service will try to anticipate its listeners' mood swings as it amplifies its competition with Pandora, Spotify and other popular services that play tunes over the Internet.

User comments : 33

Adjust slider to filter visible comments by rank

Display comments: newest first

foolspoo
1.5 / 5 (11) Jun 18, 2012
Lennon and McCartney gave it to the Gallaghers who gave it to Chris Martin.

HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA..........HAHAHAHAHAHAHAHAHAHA

what a miserable waste of resources
LariAnn
2 / 5 (15) Jun 18, 2012
So one user (one intelligent individual) can make good music using this intelligence-derived computer and algorithm, but hundreds of listeners (hundreds of intelligences) using it is Darwinian natural selection? Perhaps these researchers have drunk too much of the evolution kool-aid because they think this is "blind variation and selection" when it is clearly and unequivocally intelligence-based and driven.
Parsec
4.1 / 5 (10) Jun 18, 2012
So one user (one intelligent individual) can make good music using this intelligence-derived computer and algorithm, but hundreds of listeners (hundreds of intelligences) using it is Darwinian natural selection? Perhaps these researchers have drunk too much of the evolution kool-aid because they think this is "blind variation and selection" when it is clearly and unequivocally intelligence-based and driven.

That isn't the question the researchers were asking. Obviously the evolution of music is driven by intelligent choices. The question asked (and answered), is... "Is it possible to simulate the cultural evolution of music in a computer program?". Clearly the answer is yes and the author's have done a nice bit of research.
MrVibrating
3 / 5 (6) Jun 18, 2012
The question asked was "is human creativity really what drives musical evolution, or is it consumer choice (ie. polularity)?"

I Agree with LariAnn; it tells us nothing about the causes of the selection pressures - the agents could be replaced with an automated searchable library indexing tunes by popularity, for even more successful 'emergent creativity', such that it is. Unfortunately it's just not the same kind as ours.

The beginnings of an AI that could appreciate music in the same way as us would need to have the same experience of octave equivalence - it would need to have the impression that all factors of two of a fundamental contain no information relative to it. This first step would be the easiest, compared to progressing that to an appreciation of harmony and dissonance, cadence & melody etc.

I've yet to see any research even making a start in this direction....
LariAnn
1.7 / 5 (6) Jun 18, 2012
@Parsec,

This is what Armand Leroi, the researcher and co-author, stated - "if blind variation and selection can do that . . . it should be able to make a pop tune". Well, only if intelligence is involved (in which case it is not blind variation and selection), which they seem to be ignoring in this case.
chromosome2
not rated yet Jun 18, 2012
chromosome2
3.7 / 5 (3) Jun 18, 2012
The variation is blind, the selection is not. Same as IRL.
Deathclock
2.3 / 5 (6) Jun 18, 2012
@Parsec,

This is what Armand Leroi, the researcher and co-author, stated - "if blind variation and selection can do that . . . it should be able to make a pop tune". Well, only if intelligence is involved (in which case it is not blind variation and selection), which they seem to be ignoring in this case.


What chromosome2 said, the variation is random and the selection is guided, this is exactly how biological evolution works.
TheGhostofOtto1923
3.9 / 5 (18) Jun 18, 2012
That isn't the question the researchers were asking. Obviously the evolution of music is driven by intelligent choices.
Hahaha right. The Secret of Fashion: ANYTHING will sound or look good if it is worn or performed by healthy, attractive young people.

Meanwhile old rockers cry out in fury and indignation:
http://www.youtub...pp_video
MandoZink
4.8 / 5 (4) Jun 18, 2012
This trial and error method for selecting the most commonly pleasant ambiance certainly seems to have produced exactly that. However, I see little connection to the process that flickers in the brain of a talented musician/composer. That spark of inspiration that results in a uniquely satisfying aural dance through an emotional maze of sonic delight is missing in this process.

Excuse me. I should have just said the results are pleasant but got no soul.
dtxx
5 / 5 (2) Jun 18, 2012
After listening to the audio quote, I actually don't think the first loops sound bad at all. If you took apart modern music and analyzed it on a track-by-track basis you would hear things that sound very similar. Of course, all this means is that producers will rely on generated/random sequences that they select. Most modern music engines provide for such generation, and it's at the heart of some electronic genres. Even random arrpegiation can be so beautiful, though admittedly this is a much more highly ordered structure than random tones.

It's also interesting that major chord harmonics are the first attribute selected for. As someone who plays the blue harmonica, I can tell you that these are some of the most pleasing and interesting tonal constructions for the human ear. The change between generation 0 and 500 is much more significant than the change betweengeneration 500 and 3000, regardless of rythmic elements.
no1nose
1.4 / 5 (10) Jun 18, 2012
What simplistic crap. Why doesn't the idea of evolution itself evolve into something interesting?
PussyCat_Eyes
1.4 / 5 (11) Jun 19, 2012
That isn't the question the researchers were asking. Obviously the evolution of music is driven by intelligent choices.
Hahaha right. The Secret of Fashion: ANYTHING will sound or look good if it is worn or performed by healthy, attractive young people.

Meanwhile old rockers cry out in fury and indignation:
http://www.youtub...pp_video
- TheGhostofOtto1923

So says those wiki jockeys like the NAZI-LOVER TheGhostofOtto1923 aka TheGhostofOtto1932, idolater of the WW2 NAZI SS Officer, Otto Skorzeny, who became secretary and hatchet man of Adolph Schnikelgruber Hitler. HIS con game is to pretend to be more knowledgeable than anyone else in this website, while rushing to Wikipedia to find out that information, then voting everyone down with his sock puppet names.
Suddenly he's a music expert....an aficionado blah blah blah
TheGhostofOtto1923
4.1 / 5 (18) Jun 19, 2012
Sorry p/r/r/p crossposting is a-ginst physorg guidelines
http://phys.org/help/comments/

-I had to send the CEO of physorg an email. Again. Did he/she/it give you my number yet?

Schnickelgruber. Find that on buzzle?
CardacianNeverid
2.3 / 5 (6) Jun 19, 2012
Yet another thread ruined by PussyTard Pirouette/Ritchie/Russky. Why isn't he banned yet?
Msafwan
5 / 5 (1) Jun 19, 2012
I think after 10,000 generation the music will still sound the same.
Husky
3.5 / 5 (2) Jun 19, 2012
Sometimes music evolves too much, to cater to the widest possible audience and it loses the sharp distinct edges that made it interesting, take for instance mid 90 trance and house wich was really creative and far far out, then it all became popular due Tiesto mainstream in 2000 with lots of layered synths with easy to swallow polished 24 bit melody, boooring! I want my dj Scott Project, my Prodigy and Dritte Raum back, but fortunately somebody will come along and push it into yet another directions like progressive psytrance wich has some sharp edges and a ton of reverb...
Husky
not rated yet Jun 19, 2012
oh and i absolutely can't stand that new skrillex dubcore stuff (maybe i am getting old?) but I respect what they are doing because elements of it will mate with other styles and pieces fall into place.
CardacianNeverid
1 / 5 (4) Jun 19, 2012
take for instance mid 90 trance and house wich was really creative and far far out -Husky

LOL, I've never heard noise described as 'really creative' before!
dtxx
not rated yet Jun 19, 2012
take for instance mid 90 trance and house wich was really creative and far far out -Husky

LOL, I've never heard noise described as 'really creative' before!


Seriously? There are genres that are far more easily assaulted on the attibute of musicality...

http://en.wikiped..._(music)
TkClick
not rated yet Jun 19, 2012
The guys who can sing and play some musical instrument always got most of girls - it's classical Darwinian selection. An ideal tune: it's always good to combine presentation skills with demonstration of sexual interest about subject.
frajo
2.3 / 5 (3) Jun 19, 2012
This article presents a very restricted aspect of music.
No mentioning of the many scandals music brought about when it was new because it was grossly unpopular.
Consumers who can see only a very small subspace of the musical cosmos only select what they are used to by consuming what the industry offers.
In 1913, none of them would have chosen the "rite of spring".

Music is interesting and precious only where it's deviating from the preference of the masses.
That's why Henryk Gorecki, Zbigniew Preisner, Arvo Pärt and their like are more relevant in the musical cosmos than Abba, Gaga and their like.
CardacianNeverid
3.1 / 5 (7) Jun 19, 2012
That's why Henryk Gorecki, Zbigniew Preisner, Arvo Pärt and their like are more relevant in the musical cosmos -frajo

Huh? Who? ABBA were great!
Pattern_chaser
5 / 5 (1) Jun 19, 2012
What a fascinating and worthwhile use of resources! Music may not save us when fossil fuels run out, but it is a significant part of most human people's lives. In my book, that makes it important.
SeanRhys
not rated yet Jun 19, 2012
I agree that this study is a useless waste. Sean Rhys Stirling - life musician (with a very good pedigree).
LariAnn
1 / 5 (3) Jun 19, 2012
The variation is blind, the selection is not. Same as IRL.

Exactly - but the selection is driven by intelligence - in this case, the listeners. So if intelligence is required for this relatively simple program to work, why is intelligence totally denied in the creation and development of something far more complex - life?
Deathclock
2.9 / 5 (8) Jun 19, 2012
What simplistic crap. Why doesn't the idea of evolution itself evolve into something interesting?


You're an idiot.
Deathclock
2.3 / 5 (3) Jun 19, 2012
but the selection is driven by intelligence - in this case, the listeners. So if intelligence is required for this relatively simple program to work, why is intelligence totally denied in the creation and development of something far more complex


Intelligence wouldn't be required if bad music killed people...

Bad genes KILL organisms, good genes help them live in order to reproduce to pass on those genes to their offspring. It really isn't that difficult of a concept to understand and I can't help but think that those that deny evolution do understand it and they are not all that stupid but they have some kind of need to deny it anyway.

Intelligence is not required to determine the path of a river either... only physical laws and processes. Evolution is much like a river flowing in time, and it is guided by the environmental factors that influence the survival of life, much like the path of a river is guided by the geography of the land it flows over.
PussyCat_Eyes
2.1 / 5 (7) Jun 21, 2012
Deathclock....Intelligence is required for the APPRECIATION of a piece of music. The nuances of a difficult passage for those who have studied music will understand how each note is supposed to be performed and the length of time it takes to hold the note. Evolution gave us this gift of cognition to appreciate the soul-moving, emotional aspects of the combined instrumentation of the band or orchestra. Music IS a part of life....from the heartbeat to Beethoven.
Deathclock
2.3 / 5 (6) Jun 21, 2012
I agree with that... it doesn't have anything to do with the objections of LariAnn or my response to them, however.
antonima
5 / 5 (1) Jun 24, 2012
Since the user is trying to create something that sounds 'good' to them, they are most likely going to select for something that they have heard elsewhere before. It will reflect on their taste in music. Its interesting, and perhaps sad to some, to think that in the future music may be made without any instrumental skill. As an intellectual I see nothing wrong with it, if it is your job to pick out the best tunes and then transform them into songs then that takes inspiration and taste in music as well.

Don't know how much of that is actually possible with a darwin tunes however ;)
jibbles
1 / 5 (1) Jun 25, 2012
if this is musical evolution, then we just heard the dodo :)
Tachyon8491
4 / 5 (4) Jun 25, 2012
This "study" is scientistic bullshit (I don't often use invective but just can't avoid it this time...) It has nothing to do with the CREATION of music which involves evolution in a completely different manner from just "democratic selection." Stravinsky's "Sacre du Printemps" was regarded as "incompetent" when first publically performed. J.S. Bach's sons regarded him as "an old dodderer" but are themselves mostly dissolved in history. Bach's genius, his understanding of harmony and counterpoint, his inventive grasp of music as a language reflecting an underlying reality, is unequalled. As a classical musician my opinion of the "music" in this study is that it reflects regressive paradigm, it's primitive, repetitive, non-evolving of a thematic idea, boring as hell - incompetent - what one would expect from the lower segment of the Gaussian distribution... If we had to rely on this process for the creation of worthwhile music, forget it - it's fine for accompanying you in elevators.