Robot Jazz
February 13, 2007 6:45 PM Subscribe
I love YouTube posts not for the video, but for the comments - here's proof that the robot is smarter than the average YouTube member:
thegetawayplan9 (3 hours ago)
Fuck you asshole. Music majors have lives if anything computer science majors are the ones with no lifes because they never have friends. And are you even in college or are you some 15 year old kid?
(Reply)
posted by Muddler at 7:00 PM on February 13, 2007 [1 favorite]
thegetawayplan9 (3 hours ago)
Fuck you asshole. Music majors have lives if anything computer science majors are the ones with no lifes because they never have friends. And are you even in college or are you some 15 year old kid?
(Reply)
posted by Muddler at 7:00 PM on February 13, 2007 [1 favorite]
I know a few saxophone players who are definitely gonna get some email with a link to this.
posted by flapjax at midnite at 7:05 PM on February 13, 2007
posted by flapjax at midnite at 7:05 PM on February 13, 2007
I think this video would be useful in classrooms as a demonstration of how music is burglarized when it is stripped of expression.
Coltrane: 1
Robot: 0
posted by vacapinta at 7:09 PM on February 13, 2007
Coltrane: 1
Robot: 0
posted by vacapinta at 7:09 PM on February 13, 2007
It's a horrible choice of music for a robot... the notes aren't nearly as important as the dynamics and expression when you're playing Coltrane. Maybe the robot would fare better with Kenny G?
posted by tehloki at 7:13 PM on February 13, 2007
posted by tehloki at 7:13 PM on February 13, 2007
Sounds robotic to me.
Here is Captured by Robots that drezdn mentioned above. The guy is a mad scientist! The website is crap. The band is a Disney ride gone crazy.
posted by lee at 7:19 PM on February 13, 2007
Here is Captured by Robots that drezdn mentioned above. The guy is a mad scientist! The website is crap. The band is a Disney ride gone crazy.
posted by lee at 7:19 PM on February 13, 2007
If you have to mess with Giant Steps, I prefer this version.
on preview: tehloki is right, it is a horrible song for a robot. But I don't think Kenny G would be much better (too sappy). A robot needs something it can dance to, like Herbie Hancock's Rockit. [YouTube]
posted by LeLiLo at 7:19 PM on February 13, 2007
on preview: tehloki is right, it is a horrible song for a robot. But I don't think Kenny G would be much better (too sappy). A robot needs something it can dance to, like Herbie Hancock's Rockit. [YouTube]
posted by LeLiLo at 7:19 PM on February 13, 2007
Robots have already mastered the music of Sha Na Na. Those robots were called Sha Na Na.
posted by Astro Zombie at 7:30 PM on February 13, 2007 [2 favorites]
posted by Astro Zombie at 7:30 PM on February 13, 2007 [2 favorites]
If you have to mess with Giant Steps, I prefer this version.
That's not what Giant Steps looks like at all.
posted by rxrfrx at 7:32 PM on February 13, 2007
That's not what Giant Steps looks like at all.
posted by rxrfrx at 7:32 PM on February 13, 2007
First robot DJs and now this.
Humanity is doomed.
O, cruel fate, to be thusly boned... Ask not for whom the bone bones, it bones for thee.
posted by sparkletone at 7:34 PM on February 13, 2007
Humanity is doomed.
O, cruel fate, to be thusly boned... Ask not for whom the bone bones, it bones for thee.
posted by sparkletone at 7:34 PM on February 13, 2007
I think this video would be useful in classrooms as a demonstration of how music is burglarized when it is stripped of expression.
Yup. Coltrane - soulful expression - emotion = this.
It's a robot playing notes, no more like Coltrane than a cold plucking robot "playing" Watermelon in Easter Hay or Maggot Brain, etc.
Robots can do many things with equal or greater success compared with humans, but adding the human element to what technically or scientifically qualifies as music is not yet one of them. I've saved this link as a perfect illustration of that.
Mark my words, music will be the last Mechanical Turk.
posted by rollbiz at 7:46 PM on February 13, 2007
Yup. Coltrane - soulful expression - emotion = this.
It's a robot playing notes, no more like Coltrane than a cold plucking robot "playing" Watermelon in Easter Hay or Maggot Brain, etc.
Robots can do many things with equal or greater success compared with humans, but adding the human element to what technically or scientifically qualifies as music is not yet one of them. I've saved this link as a perfect illustration of that.
Mark my words, music will be the last Mechanical Turk.
posted by rollbiz at 7:46 PM on February 13, 2007
Nobody's said ColTron yet?
posted by emelenjr at 7:48 PM on February 13, 2007 [1 favorite]
posted by emelenjr at 7:48 PM on February 13, 2007 [1 favorite]
Bebop Cola - Comes in a variety of flavors: Root Bird, Gilberto Grape, Artie Shawberry, Kiwi Holliday, Peachmo, John Cola-trane, Don Wild Cherry, Vince Guavaldi, Dave Bruberry, Cab Colaway, Dexterade, Nina Lemone, Mango Reinhardt, Getzberry, Fizzy Gillespie, Marian McPineapple, Or'ngette Coleman, Mingus Dew, Plain, and Diet Plain
posted by furtive at 8:22 PM on February 13, 2007
posted by furtive at 8:22 PM on February 13, 2007
Well said vacapinta. The geek in me appreciates the engineering and programming that went into the robot, but the terrible butchering of the music was painful to listen to. I couldn't watch it all the way through.
But if they got together with someone who knew a few things about music and something of the physical technique that goes into playing sax, I suspect it'd be possible to produce a more realistic performance. It'd involve an artificial diaphragm and some freaky "mouth" apparatus, but it'd be interesting.
An aside: Was anyone else reminded of Data's music recitals from Star Trek TNG?
posted by aladfar at 8:28 PM on February 13, 2007
But if they got together with someone who knew a few things about music and something of the physical technique that goes into playing sax, I suspect it'd be possible to produce a more realistic performance. It'd involve an artificial diaphragm and some freaky "mouth" apparatus, but it'd be interesting.
An aside: Was anyone else reminded of Data's music recitals from Star Trek TNG?
posted by aladfar at 8:28 PM on February 13, 2007
But if they got together with someone who knew a few things about music and something of the physical technique that goes into playing sax, I suspect it'd be possible to produce a more realistic performance. It'd involve an artificial diaphragm and some freaky "mouth" apparatus, but it'd be interesting.
Well said mostly, but it's not even that. A more realistic robot mouth and the like isn't going to help, in my opinion. I honestly don't think that the human element of Giant Steps could be recreated when approached technically and mechanically. Nor could the masterful tracks I referenced above or others I didn't bother to cite. How someone feels determines how they solo, how they approach every moment of every note. While the blues are technically rudimentary, there's a big difference when I riff on the progressions and when SRV/B.B. King/Muddy Waters/etc/etc. did/do.
It's the reason why those who still look at music as a true art form wouldn't be fooled by a song churning propaganda music machine (a la 1984). Not yet, at least...
posted by rollbiz at 8:43 PM on February 13, 2007
Well said mostly, but it's not even that. A more realistic robot mouth and the like isn't going to help, in my opinion. I honestly don't think that the human element of Giant Steps could be recreated when approached technically and mechanically. Nor could the masterful tracks I referenced above or others I didn't bother to cite. How someone feels determines how they solo, how they approach every moment of every note. While the blues are technically rudimentary, there's a big difference when I riff on the progressions and when SRV/B.B. King/Muddy Waters/etc/etc. did/do.
It's the reason why those who still look at music as a true art form wouldn't be fooled by a song churning propaganda music machine (a la 1984). Not yet, at least...
posted by rollbiz at 8:43 PM on February 13, 2007
Was anyone else reminded of Data's music recitals from Star Trek TNG?
Yep.
Anybody else here a saxophone player? (aladfar?) From an engineering standpoint, I was particularly impressed by robotic control of embouchure, which is necessary when playing Coltrane, even in a form such as this. It ain't just blowing air into a tube.
But I agree-- having some more "mouth" flexibility (bends, vibrato), and maybe adding some "noise" to the timing (playing some notes late, some early) would help a lot.
Plus, it's not a very good transcription-- it sounds like it's translated twice: from record to sheet music, then from sheet music to computerized sheet music. If a computer could analyze the record and get Trane's timing right (i.e., not going through restrictive music notation), it would be a lot more impressive.
posted by supercres at 8:49 PM on February 13, 2007
Yep.
Anybody else here a saxophone player? (aladfar?) From an engineering standpoint, I was particularly impressed by robotic control of embouchure, which is necessary when playing Coltrane, even in a form such as this. It ain't just blowing air into a tube.
But I agree-- having some more "mouth" flexibility (bends, vibrato), and maybe adding some "noise" to the timing (playing some notes late, some early) would help a lot.
Plus, it's not a very good transcription-- it sounds like it's translated twice: from record to sheet music, then from sheet music to computerized sheet music. If a computer could analyze the record and get Trane's timing right (i.e., not going through restrictive music notation), it would be a lot more impressive.
posted by supercres at 8:49 PM on February 13, 2007
How odd. Figure there's plenty of electronically-generated music, that can be broken down into simple patterns, that still retains a feeling of soul. Coltrane wrote Giant Steps for a person and the robot can't play it how it's meant. Why can a synthesizer play what someone else wrote and retain the soul written into it?
Is it a hurdle of soul or ingenuity that makes it sounds so hollow right now?
posted by griphus at 9:11 PM on February 13, 2007
Is it a hurdle of soul or ingenuity that makes it sounds so hollow right now?
posted by griphus at 9:11 PM on February 13, 2007
Why can a synthesizer play what someone else wrote and retain the soul written into it?
I assume because the music you're talking about was written for synthesizer.
As an example, Switch on Bach was a great technical achievement at the time, and is remembered as such, but I'd argue it's about as much fun to listen to has ColTron here. Maybe better because of all those crazy Moog sounds, but quite frankly it's expressionless and, well, robotic, compared to an orchestra playing Bach.
On the other hand, modern electronic music - music that was intended to be performed on electronic instruments - can be soulful because it was written for the purpose.
posted by Jimbob at 9:29 PM on February 13, 2007
I assume because the music you're talking about was written for synthesizer.
As an example, Switch on Bach was a great technical achievement at the time, and is remembered as such, but I'd argue it's about as much fun to listen to has ColTron here. Maybe better because of all those crazy Moog sounds, but quite frankly it's expressionless and, well, robotic, compared to an orchestra playing Bach.
On the other hand, modern electronic music - music that was intended to be performed on electronic instruments - can be soulful because it was written for the purpose.
posted by Jimbob at 9:29 PM on February 13, 2007
Jimbob- the question remains, though. will our ability to program robots to do things ever come close to mimicking the way a person would do them on an aesthetic level?
Figure we could hook a (good) sax player up to a thousand machines, monitor limb movements and breath use, break the session down to figures and feed it into a robot that can simulate it.
posted by griphus at 10:55 PM on February 13, 2007
Figure we could hook a (good) sax player up to a thousand machines, monitor limb movements and breath use, break the session down to figures and feed it into a robot that can simulate it.
posted by griphus at 10:55 PM on February 13, 2007
Then you just have a video of a sax player playing being conveyed by a robot.
posted by tehloki at 11:13 PM on February 13, 2007
posted by tehloki at 11:13 PM on February 13, 2007
I don't understand all this talk about "expression" and "emotion". There's no emotion in playing an instrument. What you're really saying is that the robot is poorly constructed and programmed for the task set before it. It has nothing to do with it not being human.
posted by Target Practice at 12:35 AM on February 14, 2007
posted by Target Practice at 12:35 AM on February 14, 2007
I'd be very interested to see a robot pass a turing test of musical expression. At one level, it's something that any cd player could do. But more interesting would be to have a jam with a robot and see if it gave you that sense of communication. I don't see in principle why not, and I think it would be a step towards giving robots emotions.
And once they have emotions they'll be just as fucked up as the rest of us.
posted by leibniz at 12:50 AM on February 14, 2007
And once they have emotions they'll be just as fucked up as the rest of us.
posted by leibniz at 12:50 AM on February 14, 2007
Figure we could hook a (good) sax player up to a thousand machines, monitor limb movements and breath use, break the session down to figures and feed it into a robot that can simulate it.
Surely we could hook a golfer up to similar machines and make a mechanical Tiger Woods and we would be achieving the same thing; it doesn't really tell us much about music. Philosophically, to me, a robot able to recreate a performance on a physical instrument would be technically triumphant, as "Switched on Bach" is, but a robot able to create an interesting musical performance based on nothing but sheet music would be a real development.
posted by Jimbob at 3:54 AM on February 14, 2007
Surely we could hook a golfer up to similar machines and make a mechanical Tiger Woods and we would be achieving the same thing; it doesn't really tell us much about music. Philosophically, to me, a robot able to recreate a performance on a physical instrument would be technically triumphant, as "Switched on Bach" is, but a robot able to create an interesting musical performance based on nothing but sheet music would be a real development.
posted by Jimbob at 3:54 AM on February 14, 2007
Target Practice, unless my sarcasm detector is failing me, it sounds like You Just Don't Get Music if you don't recognize the expression and emotion a human player can bring to a piece of music. Much moreso by a musician like John Coltrane, who I'd wager wasn't looking at sheet music when he originally performed on that recording. Hell, much moreso by any musician who's not in the Look How Fast I Can Follow Directions camp.
As for ColTron, it's Look How Fast I Can Follow MIDI Instructions.
... and as a musical performance, it sucks.
posted by emelenjr at 4:45 AM on February 14, 2007
As for ColTron, it's Look How Fast I Can Follow MIDI Instructions.
... and as a musical performance, it sucks.
posted by emelenjr at 4:45 AM on February 14, 2007
Emotion in music? First, a machine will be able to imitate it. The real test will be improv, but even that will just be a matter of the right programming.
posted by dreamsign at 5:42 AM on February 14, 2007
posted by dreamsign at 5:42 AM on February 14, 2007
I don't understand all this talk about "expression" and "emotion".
I didn't know we had Vulcans on MeFi!
nola: Thanks, that was great.
posted by languagehat at 6:11 AM on February 14, 2007
I didn't know we had Vulcans on MeFi!
nola: Thanks, that was great.
posted by languagehat at 6:11 AM on February 14, 2007
There's no emotion in playing an instrument.
Hmmm....
posted by jimmythefish at 7:15 AM on February 14, 2007
Hmmm....
posted by jimmythefish at 7:15 AM on February 14, 2007
emelenjr
I honestly have no idea what the word "expression" is supposed to mean in this context. As for emotion, you cannot tell me that it takes a specific emotional makeup to play certain notes a certain way. Any sounds that can be made by any person playing an instrument should be replicable by an automaton, given that the latter is properly programmed and that it has the necessary motor ability.
posted by Target Practice at 10:03 AM on February 14, 2007
I honestly have no idea what the word "expression" is supposed to mean in this context. As for emotion, you cannot tell me that it takes a specific emotional makeup to play certain notes a certain way. Any sounds that can be made by any person playing an instrument should be replicable by an automaton, given that the latter is properly programmed and that it has the necessary motor ability.
posted by Target Practice at 10:03 AM on February 14, 2007
No soul.
posted by gnutron at 10:04 AM on February 14, 2007 [2 favorites]
posted by gnutron at 10:04 AM on February 14, 2007 [2 favorites]
natch, Target Practice.
The bending of a note, a strained bit of phrasing -- these might be the results of emotion but it isn't necessary that they proceed from it for a truly talented mimic.
That's one of the reasons why improv would be a much greater feat. It would have to mimic various styles and "emotions" in the music and weave them together in a way that made aesthetic sense (to us).
posted by dreamsign at 10:51 AM on February 14, 2007
The bending of a note, a strained bit of phrasing -- these might be the results of emotion but it isn't necessary that they proceed from it for a truly talented mimic.
That's one of the reasons why improv would be a much greater feat. It would have to mimic various styles and "emotions" in the music and weave them together in a way that made aesthetic sense (to us).
posted by dreamsign at 10:51 AM on February 14, 2007
You would first have to teach the robot when bending, sustaining, syncopation, and all those other emotional/distinctly "human" musical techiques actually makes sense and sounds effective. You could just throw a random pitch-bending and hip-swaying algorithm in there, but you'd just end up getting something firmly resting at the bottom of the uncanny valley.
posted by tehloki at 12:10 PM on February 14, 2007
posted by tehloki at 12:10 PM on February 14, 2007
Nobody said anything about random anymore than hitting random notes is playing a melody. But emotion is an internal state. A machine can be trained to mimic the externals and patterning is obviously part of that. Improv would just be necessarily more complex and involve something approaching judgment.
posted by dreamsign at 12:20 PM on February 14, 2007
posted by dreamsign at 12:20 PM on February 14, 2007
Sorry, make that -- importantly -- resembling judgment.
posted by dreamsign at 1:12 PM on February 14, 2007
posted by dreamsign at 1:12 PM on February 14, 2007
Well I thought it was cool. There is one thing musical robots can't do unfortunately... heroin.
"I didn't know we had Vulcans on MeFi!"
then you haven't been paying attention.
posted by vronsky at 1:12 PM on February 14, 2007
"I didn't know we had Vulcans on MeFi!"
then you haven't been paying attention.
posted by vronsky at 1:12 PM on February 14, 2007
Hey guys, me again. Sorry about this post.
It should go without saying, but I wasn't posting this thinking that it was musically wondrous. It made me laugh a lot. Since I've done a few good posts, I thought it was time for a silly one-linker.
In general, I agree with Target Practice that words like "expression" and "emotion" are bandied about a little more than they deserve to be. Not that music isn't emotional (of course it is), but the split-second decisions and reactions that go into the stylistic flourishes of jazz/blues music are the product of having a solid understanding of craft, plus years of trial and error. At the end of the day, Elvis Costello doesn't feel emotions more deeply than I do; he knows how to make music better than I do.
On the other hand, modern electronic music - music that was intended to be performed on electronic instruments - can be soulful because it was written for the purpose.
Yes, I agree. And a lot of 20th century music is so much more about the creative mind that devised a system for creating music than about the music itself. I think a robot could theoretically play a jazz solo sufficiently well, but I don't think that a robo-Xenakis could invent a computer program for creating music.
posted by roll truck roll at 1:37 PM on February 14, 2007
It should go without saying, but I wasn't posting this thinking that it was musically wondrous. It made me laugh a lot. Since I've done a few good posts, I thought it was time for a silly one-linker.
In general, I agree with Target Practice that words like "expression" and "emotion" are bandied about a little more than they deserve to be. Not that music isn't emotional (of course it is), but the split-second decisions and reactions that go into the stylistic flourishes of jazz/blues music are the product of having a solid understanding of craft, plus years of trial and error. At the end of the day, Elvis Costello doesn't feel emotions more deeply than I do; he knows how to make music better than I do.
On the other hand, modern electronic music - music that was intended to be performed on electronic instruments - can be soulful because it was written for the purpose.
Yes, I agree. And a lot of 20th century music is so much more about the creative mind that devised a system for creating music than about the music itself. I think a robot could theoretically play a jazz solo sufficiently well, but I don't think that a robo-Xenakis could invent a computer program for creating music.
posted by roll truck roll at 1:37 PM on February 14, 2007
I think robots could quite easily compose and play smooth jazz muzak.
posted by tehloki at 2:11 PM on February 14, 2007
posted by tehloki at 2:11 PM on February 14, 2007
This discussion actually ties in to something I was thinking about recently. I was wondering whether it would be possible right now to program a robot that performs a piece of sheet music exactly as written, and pull it off in a way that sounds human.
posted by drezdn at 2:17 PM on February 14, 2007
posted by drezdn at 2:17 PM on February 14, 2007
I'd like to hear the "Giant Steps" recording with Coltrane's track muted, replaced by ColTron. Anyone feel like whipping that up?
posted by emelenjr at 3:30 PM on February 14, 2007
posted by emelenjr at 3:30 PM on February 14, 2007
I've done this before, where I write an insightful post just when the thread is about the stop forever and so basically no one reads it. But having a phd focusing on the subject of how music expresses emotions, (and particularly jazz) I actually find it painful to read some of these comments that are tentatively trying to figure it out. I resisted making a big comment above. But if anyone actually wants to hear some of the explantions for these issues, or to read my work, please email me.
posted by leibniz at 1:14 AM on February 15, 2007
posted by leibniz at 1:14 AM on February 15, 2007
For pete's sake, just make the comment. Those of us who care are still reading. What's the point of sharing your wisdom with one e-mail correspondent when you could be sharing it with the community, and leaving it as part of the record so that those who find this thread in future will get the benefit of your learning?
posted by languagehat at 6:09 AM on February 15, 2007
posted by languagehat at 6:09 AM on February 15, 2007
Yes, leibniz, please share your thoughts. I think we could all use some schoolin'.
Anyway, doesn't everyone use the my comments page? I always follow discussions to the bitter end, even if I've stopped posting in them.
posted by roll truck roll at 11:49 AM on February 15, 2007
Anyway, doesn't everyone use the my comments page? I always follow discussions to the bitter end, even if I've stopped posting in them.
posted by roll truck roll at 11:49 AM on February 15, 2007
Speak, Leibniz, Speak! Okay, now roll over.
Actually, roll truck roll is taking care of the rolling. He rocks, too. leibniz, however, is too busy feeling the acute pain of having to read comments by non-PHD holders. And that, my friend, is a special kind of pain that only PHD holders can truly suffer.
You can email him with specific questions about the pain, though.
posted by flapjax at midnite at 4:42 PM on February 15, 2007
Actually, roll truck roll is taking care of the rolling. He rocks, too. leibniz, however, is too busy feeling the acute pain of having to read comments by non-PHD holders. And that, my friend, is a special kind of pain that only PHD holders can truly suffer.
You can email him with specific questions about the pain, though.
posted by flapjax at midnite at 4:42 PM on February 15, 2007
He doesn't have a PhD, if that makes any difference. He's doing one.
posted by Wolof at 1:35 AM on February 16, 2007
posted by Wolof at 1:35 AM on February 16, 2007
ok, that's funny. Alright, the actual proof is long and complicated, but it seems that jazz improvisation is a case of extended cognition (where mental processes are partly carried out outside the head cf. moving scrabble tiles around to help figure out what word to play). Or read this paper:
http://consc.net/papers/extended.html
Moreover, it's often the extended cognition of emotions - which is much more unusual (yes emotions can be cognised -ever deliberately alterered your emotional behaviour in order to change your emotion?). But this is not just the case of using the music to influence your emotion in a feedback loop. No, the patterns of the music produced stand in the same relation to your emotional experience as your ordinary bodily changes do. In most cases, they act as elaborations of your bodily state. However, in cases known as 'flow states' it appears that the music takes the dominant role. It becomes the locus of control for the progress of the emotion and the representational content of the emotion (i.e. how things are going for you). Musicians experience this as self-dislocation, appropriately so, because their self has literally become extended into the music.
The upshot of all this? sometimes the music isn't just the expression of an emotion. It IS an emotion (partly).
So far as the basic emotional capacity of music is concerned, this is grounded in the deep resemblance that musical patterns bear to feeling patterns and the brain's automatic tendency to process cross-modally (sights, sounds, touch all become patterns of movement). However this is not experienced necessarily as a resemblance effect. Rather the simulation mechanism of the listener is triggered, which is what we use to recognise emotions in people. This mechanism is a bunch of systems using mirror neurons, which fire both when movement (of conspecifics) is observed and when actually performing the action oneself. So overall, you process the movement as if you were doing it yourself. The generates a neural level of emotional arousal (which can sometimes contagiously overwhelm the perceiver) but which more often leads to the perceiving the feeling IN whatever movement the mechanism is tracking i.e. the person or the music.
That robot just doesn't trigger my simulation mechanism very well because all those tiny inflections in coltrane's real solo are what give a sense of the subtle and complex movements of a body like mine. This is not to say it's impossible though. And like I said above I'd be interested to see just what it takes to get a robot to convince in this way. I've never heard of anyone being being able to produce emotive music without using their own sense of feeling. Then again it is possible to hear a tap dripping in an emotive way if you imagine very hard that a person is flicking the water from a brush (thus triggering your simulation mechanism).
Ok, I have to get back to work now. I am trying to show that groups can share single emotional states by extending their emotions in the way I described above. Hope this lived up to your expectations of phd research.
posted by leibniz at 2:01 AM on February 16, 2007
http://consc.net/papers/extended.html
Moreover, it's often the extended cognition of emotions - which is much more unusual (yes emotions can be cognised -ever deliberately alterered your emotional behaviour in order to change your emotion?). But this is not just the case of using the music to influence your emotion in a feedback loop. No, the patterns of the music produced stand in the same relation to your emotional experience as your ordinary bodily changes do. In most cases, they act as elaborations of your bodily state. However, in cases known as 'flow states' it appears that the music takes the dominant role. It becomes the locus of control for the progress of the emotion and the representational content of the emotion (i.e. how things are going for you). Musicians experience this as self-dislocation, appropriately so, because their self has literally become extended into the music.
The upshot of all this? sometimes the music isn't just the expression of an emotion. It IS an emotion (partly).
So far as the basic emotional capacity of music is concerned, this is grounded in the deep resemblance that musical patterns bear to feeling patterns and the brain's automatic tendency to process cross-modally (sights, sounds, touch all become patterns of movement). However this is not experienced necessarily as a resemblance effect. Rather the simulation mechanism of the listener is triggered, which is what we use to recognise emotions in people. This mechanism is a bunch of systems using mirror neurons, which fire both when movement (of conspecifics) is observed and when actually performing the action oneself. So overall, you process the movement as if you were doing it yourself. The generates a neural level of emotional arousal (which can sometimes contagiously overwhelm the perceiver) but which more often leads to the perceiving the feeling IN whatever movement the mechanism is tracking i.e. the person or the music.
That robot just doesn't trigger my simulation mechanism very well because all those tiny inflections in coltrane's real solo are what give a sense of the subtle and complex movements of a body like mine. This is not to say it's impossible though. And like I said above I'd be interested to see just what it takes to get a robot to convince in this way. I've never heard of anyone being being able to produce emotive music without using their own sense of feeling. Then again it is possible to hear a tap dripping in an emotive way if you imagine very hard that a person is flicking the water from a brush (thus triggering your simulation mechanism).
Ok, I have to get back to work now. I am trying to show that groups can share single emotional states by extending their emotions in the way I described above. Hope this lived up to your expectations of phd research.
posted by leibniz at 2:01 AM on February 16, 2007
ok, that's funny. Alright, the actual proof is long and complicated, but it seems that jazz improvisation is a case of extended cognition (where mental processes are partly carried out outside the head cf. moving scrabble tiles around to help figure out what word to play). Or read this paper:
http://consc.net/papers/extended.html
Moreover, it's often the extended cognition of emotions - which is much more unusual (yes emotions can be cognised -ever deliberately alterered your emotional behaviour in order to change your emotion?). But this is not just the case of using the music to influence your emotion in a feedback loop. No, the patterns of the music produced stand in the same relation to your emotional experience as your ordinary bodily changes do. In most cases, they act as elaborations of your bodily state. However, in cases known as 'flow states' it appears that the music takes the dominant role. It becomes the locus of control for the progress of the emotion and the representational content of the emotion (i.e. how things are going for you). Musicians experience this as self-dislocation, appropriately so, because their self has literally become extended into the music.
The upshot of all this? sometimes the music isn't just the expression of an emotion. It IS an emotion (partly).
So far as the basic emotional capacity of music is concerned, this is grounded in the deep resemblance that musical patterns bear to feeling patterns and the brain's automatic tendency to process cross-modally (sights, sounds, touch all become patterns of movement). However this is not experienced necessarily as a resemblance effect. Rather the simulation mechanism of the listener is triggered, which is what we use to recognise emotions in people. This mechanism is a bunch of systems using mirror neurons, which fire both when movement (of conspecifics) is observed and when actually performing the action oneself. So overall, you process the movement as if you were doing it yourself. The generates a neural level of emotional arousal (which can sometimes contagiously overwhelm the perceiver) but which more often leads to the perceiving the feeling IN whatever movement the mechanism is tracking i.e. the person or the music.
That robot just doesn't trigger my simulation mechanism very well because all those tiny inflections in coltrane's real solo are what give a sense of the subtle and complex movements of a body like mine. This is not to say it's impossible though. And like I said above I'd be interested to see just what it takes to get a robot to convince in this way. I've never heard of anyone being being able to produce emotive music without using their own sense of feeling. Then again it is possible to hear a tap dripping in an emotive way if you imagine very hard that a person is flicking the water from a brush (thus triggering your simulation mechanism).
Ok, I have to get back to work now. I am trying to show that groups can share single emotional states by extending their emotions in the way I described above. Hope this lived up to your expectations of phd research.
posted by leibniz at 2:14 AM on February 16, 2007
http://consc.net/papers/extended.html
Moreover, it's often the extended cognition of emotions - which is much more unusual (yes emotions can be cognised -ever deliberately alterered your emotional behaviour in order to change your emotion?). But this is not just the case of using the music to influence your emotion in a feedback loop. No, the patterns of the music produced stand in the same relation to your emotional experience as your ordinary bodily changes do. In most cases, they act as elaborations of your bodily state. However, in cases known as 'flow states' it appears that the music takes the dominant role. It becomes the locus of control for the progress of the emotion and the representational content of the emotion (i.e. how things are going for you). Musicians experience this as self-dislocation, appropriately so, because their self has literally become extended into the music.
The upshot of all this? sometimes the music isn't just the expression of an emotion. It IS an emotion (partly).
So far as the basic emotional capacity of music is concerned, this is grounded in the deep resemblance that musical patterns bear to feeling patterns and the brain's automatic tendency to process cross-modally (sights, sounds, touch all become patterns of movement). However this is not experienced necessarily as a resemblance effect. Rather the simulation mechanism of the listener is triggered, which is what we use to recognise emotions in people. This mechanism is a bunch of systems using mirror neurons, which fire both when movement (of conspecifics) is observed and when actually performing the action oneself. So overall, you process the movement as if you were doing it yourself. The generates a neural level of emotional arousal (which can sometimes contagiously overwhelm the perceiver) but which more often leads to the perceiving the feeling IN whatever movement the mechanism is tracking i.e. the person or the music.
That robot just doesn't trigger my simulation mechanism very well because all those tiny inflections in coltrane's real solo are what give a sense of the subtle and complex movements of a body like mine. This is not to say it's impossible though. And like I said above I'd be interested to see just what it takes to get a robot to convince in this way. I've never heard of anyone being being able to produce emotive music without using their own sense of feeling. Then again it is possible to hear a tap dripping in an emotive way if you imagine very hard that a person is flicking the water from a brush (thus triggering your simulation mechanism).
Ok, I have to get back to work now. I am trying to show that groups can share single emotional states by extending their emotions in the way I described above. Hope this lived up to your expectations of phd research.
posted by leibniz at 2:14 AM on February 16, 2007
You see!? This thread is over! Argghgh
I'm never going to do anything anyone asks for ever again.
posted by leibniz at 1:02 AM on February 17, 2007
I'm never going to do anything anyone asks for ever again.
posted by leibniz at 1:02 AM on February 17, 2007
No, no, it's extremely interesting! But it's a lot to digest at one go; I suspect we're all mulling it over. Personally, I find it pretty convincing, given what little I know about how the brain works. Don't go away mad!
posted by languagehat at 4:54 AM on February 17, 2007
posted by languagehat at 4:54 AM on February 17, 2007
Hey, leibniz, remember, the thread ain't over 'til the fat lady sings. But...
she ain't singing anything now cause she's too busy trying to figure out the complex emotional mechanics behind human musical vocal expression. To put it simply, she's gotten, er, over-analytical...
She'll probably get back to singing before long, though.
posted by flapjax at midnite at 4:54 AM on February 17, 2007
she ain't singing anything now cause she's too busy trying to figure out the complex emotional mechanics behind human musical vocal expression. To put it simply, she's gotten, er, over-analytical...
She'll probably get back to singing before long, though.
posted by flapjax at midnite at 4:54 AM on February 17, 2007
This whole notion of the musical communication of somatic affect (or at least its representation) between performer and listener* seems fairly solid from where I sit.
Which is why if I want people to dance when I play, I stand up.**
*Most people underestimate the role of the body in listening. It's a lot more than something only the ears and brain do. Resonant bodily cavities, bone conduction, effect on respiration, etc., all play into this — I would go so far as to characterise music as a form of coerced respiration in certain scenarios.
**Note — example is limited but extensible.
posted by Wolof at 3:39 AM on February 19, 2007
Which is why if I want people to dance when I play, I stand up.**
*Most people underestimate the role of the body in listening. It's a lot more than something only the ears and brain do. Resonant bodily cavities, bone conduction, effect on respiration, etc., all play into this — I would go so far as to characterise music as a form of coerced respiration in certain scenarios.
**Note — example is limited but extensible.
posted by Wolof at 3:39 AM on February 19, 2007
drum machine
roll - this immediately reminded me of your post
posted by vronsky at 4:22 PM on February 26, 2007
roll - this immediately reminded me of your post
posted by vronsky at 4:22 PM on February 26, 2007
« Older Please don't let it be cloudy, please, please | One of the men said that it was the most delicious... Newer »
This thread has been archived and is closed to new comments
When the video ends, you can click through to a video of a robot playing the trumpet. This robot does some foot tapping and grooving.
I remain unimpressed. Call me when they get a robot to play the bagpipes.
posted by flarbuse at 6:54 PM on February 13, 2007