The Folk Process Scientificized

Subject: The Folk Process Scientificized
From: Amos
Date: 10 Dec 09 - 11:02 AM

The power of music to evolve and find (or not) survival in its new forms is now subject to scientific analysis in a new analysis program in which you are allowed to participate. NewScientist has this report:

"...Here's how it works:


MacCullum's computer program creates a randomly generated pair of "Adam and Eve" "songs"--brief loops of sound. They mutate, recombine and reproduce to form a base population of 100 descendants.


Participants act as the force of natural selection by listening to the songs and rating them, from "I love it!" through "It's OK..." to "I can't stand it". For every 20 songs, the 10 worst rated die off, while the 10 best rated go on to reproduce at random, with each "mating" producing two new songs. Each daughter song inherits a mixture of the parents' computer codes, just as a biological organism inherits a mixture of its parents' genetic codes.

headphones04b.jpg"The 'chromosomes' in DarwinTunes are actually tree structures of code," the researchers explain. "There is only one tree structure per song, that is, they are 'haploid'. During recombination a small number of tree nodes are chosen at random in one parent (each node has a 1 in 1000 chance of being chosen). The same number of nodes are then chosen at random in the other parent."


Then random mutation comes into play. Each node of a daughter's code has a one in 1500 chance of mutating. "Eighty per cent of the mutations are 'point mutations' which alter the value of a single atomic piece of information (e.g. note length, note position, wavelength multiple). The remainder are 'macro mutations' which swap, copy, insert, delete or replace part of the tree structure."


When 20 new songs are born, the 10 parents die off, and the process continues in the new generation.


MacCallum and Leroi modelled their experiment on laboratory studies of the evolution of microbes and worms, but they believe the process is similar to how music evolves in real life: "A thousand bands bang away in a thousand garages, mutate the musical idiom of the day, and test the results in the market. Most fail, but a few succeed and popular music continues its relentless evolutionary march."


MacCullum and Leroi are already seeing progress. "The result is more musical than the starting population," MacCullum says. "It was always possible that everyone would pull in different directions and it would go nowhere." The researchers have also been surprised to see how strong the effects of random mutations are on the outcome.


They are hoping to keep the experiment running for at least another week - so now is your chance to be a force of nature, and select.

To participate visit http://darwintunes.org/evolve-music"

Subject: RE: The Folk Process Scientificized
From: theleveller
Date: 10 Dec 09 - 11:31 AM

You'd think people would have better things to do with their time - like maybe writing some songs themselves, or finding a cure for cancer :)

Subject: RE: The Folk Process Scientificized
From: Bill D
Date: 10 Dec 09 - 12:32 PM

mercy! I did this for about 10 minutes, and I am not sure what they are really rating. They sure don't get close to Scottish ballads... *grin*

I suspect they will find that 'most' people don't like cacophony and grating combinations of rhythms....for $19.95,I'd tell them this.

Subject: RE: The Folk Process Scientificized
From: Jerry Rasmussen
Date: 10 Dec 09 - 12:43 PM

The computer process sanctified.

Jerry

Subject: RE: The Folk Process Scientificized
From: Bill D
Date: 10 Dec 09 - 01:10 PM

ah, Jerry... I doubt that sanctifying the computer process is theologically possible.... *grin*

Subject: RE: The Folk Process Scientificized
From: GUEST
Date: 10 Dec 09 - 02:36 PM

"You'd think people would have better things to do with their time..."

When I think of the amount of time and resources that go into the production of music which people either ignore (if they're lucky) or resent (like me), then it seems to me that research which identifies what people dislike is well worth it.

It baffles me why restaurant mangagement, for example, thinks I want to listen to some guy gurgle the infected phlegm in his throat while I eat. Yet they play music where that happens all the time. If it takes a scientist to get through to such managers, then I consider that a good thing.

Subject: RE: The Folk Process Scientificized
From: theleveller
Date: 10 Dec 09 - 03:54 PM

"It baffles me why restaurant mangagement, for example, thinks I want to listen to some guy gurgle the infected phlegm in his throat while I eat. Yet they play music where that happens all the time. If it takes a scientist to get through to such managers, then I consider that a good thing."

I've always found that politely saying, "will you turn that shit off or I'm leaving" works quite well. I came up with that solution without the need for a computer programme.

Subject: RE: The Folk Process Scientificized
From: Artful Codger
Date: 10 Dec 09 - 03:58 PM

I think it's bunk theory. Most significant evolution is not the result of accumulating minor changes (which often cancel each other out), surviving because one change is more appealing than another. Rather, larger forces tend to create quantum shifts and override "quality" more often than not. Consider the influence of intentional revision, publication, localization, community "celebrities", invasions, grafting and sheer forgetfulness (those flagrant gaps in collected versions)...

Furthermore, there is a presupposition in the theory that evolution is always in a positive direction. But very often the "folk process" results in devolution--rendering superior versions (in terms of quality and appeal) blander and duller at they're passed along by the forgetful and less gifted.

If this were not so, most folk songs would surpass the best of Shakespeare in their power and felicity of language.

Subject: RE: The Folk Process Scientificized
From: Wesley S
Date: 10 Dec 09 - 04:02 PM

Scientificized is a word?

Subject: RE: The Folk Process Scientificized
From: Artful Codger
Date: 10 Dec 09 - 05:37 PM

"Scientificized" was the favored variant produced after four generations of mutation, using the evolution process. ;-}

Subject: RE: The Folk Process Scientificized
From: GUEST,Gerry
Date: 10 Dec 09 - 05:52 PM

Hooray for New Folk,
New-ew-ew Folk,
It won't do you a bit of good to review folk.
It's so simple,
So very simple,
That only a computer can do it.

Subject: RE: The Folk Process Scientificized
From: Uncle_DaveO
Date: 11 Dec 09 - 09:23 AM

Furthermore, there is a presupposition in the theory that evolution is always in a positive direction.

"The theory"?   If you mean Darwin's theory, you're wrong. He made no such assumption. He specifically pointed out that the (temporarily and situationally) fittest might very well, in the long run, turn out to be inferior under varying conditions, and get crowded out by a population with another variation which has more flexibility and staying power.

If, on the other hand, you're referring to this study, there's no particular theory yet. There is, of course, a hypothesis that a similar-to-Darwin phenomenon may turn up in the development of musical scraps, but that's merely a hypothesis being tested.   After testing, if the results seem to produce an evolution similar to biological evolution, a theory may be developed. That's what scientific experimentation is all about.

Dave Oesterreich

Subject: RE: The Folk Process Scientificized
From: Artful Codger
Date: 11 Dec 09 - 12:53 PM

Regardless of whether it can produce similar results--monkeys with typewriters come to mind--, we already know that such a hypothesis is false (except in some small neighborhood of time, in which case my response is "Duh") because it doesn't incorporate what we already know to occur: quantum shifts and regression (as in revised Darwinian theory). One can develop chess programs which outplay humans, but it doesn't mean they use the same reasoning processes to achieve that outcome. So, while it may be "scientific" in the sense of technical, it has little bearing on the actual evolution of folk songs.

It also ignores the role of fashion changes over time, which is a crucial factor in determining survival.

And how can they evaluate if the results are "similar", anyway? Similar to what? What are their criteria for comparison? How many Victorians are involved in the voting (so that their tastes are reflected)? Sounds like one of these studies where the people proposing and structuring the study will cry "Eureka!" regardless of what results they end up with (short of a Babelfish translation).