Why do humans like jazz? (evolution of music, entropy, and physics of neurons)

I think it's a safe assessment to say that most people don't like jazz and pop culture makes us pretty clear even Jazz enthusiasts understand that their music is an acquired taste and I think the best example of this are all the Giant Steps memes that were popular a few years ago the reason that people thought these were funny is because they knew that any normal person would think it sounds terrible to take John coltrane's chord changes and put them on some pop song but God's listeners got the enjoyment of claiming that it was good music we spent almost all of human history making music more organized more structured and classical music was the peak of this but in the last 100 years Jazz has thrown all that away is this some kind of manifestation of the second law of Thermodynamics in our culture are we destined for art to become more chaotic and abstract why do people like gas at all or an even better question why do humans like music at all and this is a hard question to answer because they're seemingly unrelated aspects of music right Rhythm Melody Harmony if there was any theory that attempted to explain why we like music it would need to link all three of these together I'll give you a spoiler we don't have that explanation yet I can't tell it to you in this video because it's just something we don't know but although it's not nicely wrapped together in one Theory we do have some well-educated biological reasons for liking different aspects of music in this video we'll look at some of the evolutionary and physiological reasons that we play music and we'll also look at some of the math and physics that our brains experience when we listen to it soon after Charles Darwin introduced his theory of natural selection he himself tried to apply it to humans interest in music his theory was that early humans used song to attract mates like in birds or cicadas or more closely related Gibbons he didn't provide a physical benefit to our survival that music would have given us but that's okay because sexual selection often exaggerates traits that wouldn't have any purpose otherwise in Stephen pinker's book how the mind works he argues that music is evolutionary cheesecake humans of course wouldn't have any reason to enjoy cheesecake in the wild but there were incentives to have a taste for sugar and fats and oils so cheesecake is a byproduct of these other adaptations pinker's hypothesis is that music never served any purpose to humans it's simply a byproduct of our linguistic ability and our ability to process sounds and find dangers in the environment and the natural rhythms in our bodies because our brains adapted to do the things that music has music is just this Indulgence of ours I'm not entirely convinced but Pinker is getting somewhere with the mental adaptations he mentions that might have led to music especially with the influence that motor control has on a sense of rhythm the human leg is essentially a pendulum so it has a specific resonant frequency when used correctly our legs can be very efficient in fact you could make a simple walking device that's powered only by gravity even when walking on flat land or uphill it takes an impressively small amount of energy to move our bodies but only when we're walking in time with our legs resonant frequency so gravity can do most of the work the better our sense of Rhythm the more affectionately we move ostriches are also bipedal but we don't see them playing drums so what's different about us we have two legs but we're also social creatures we hunted and gathered in groups footsteps are pretty loud and when you have a group of say a dozen humans all walking at possibly different rates they're working hard so they're all breathing loudly possibly at different rates then it starts to mask out important sounds like prey stepping on a twig or a snake in the bushes one thing we can do to alleviate this problem is we can synchronize our steps on our breathing with other members of the group if all of these predictable sounds are lined up with each other then that gives us spaces of Silence to listen to our environment it also makes it easier for our brain to filter out unimportant sounds because they're predictable this would have incentivized two adaptations one is for us to synchronize our steps our breathing tapping a finger with the people around us and with sounds that we hear this is still observed in people today the others for our brain to become better at recognizing and locking on to sound sources with different tempos and becoming comfortable with those in order to filter them out now this starts to Look a Lot Like Music proponents of this Theory believe that not only would this have given us the skills we need to play music it would have also required us to play music in groups in order to practice these abilities some Modern cultures even recognize this practical value that music has for example drumming traditions of the away people in West Africa a web music relies on very complex polyrhythms a musician might be playing a rhythm like this with one hand and their other hand or another musician might be playing a 4B pattern underneath simultaneously there might be a six beat pattern that's being played and still there could be a 2B pattern a three beat pattern any of these could be offset by a beat it gets very complicated and there's a lot going on this music acts as a kind of moral education the belief is that by practicing until you're comfortable with these conflicting rhythms you're preparing yourself for the inevitable conflicts of life if you focus on one Rhythm or if you block the others out then that's a sign of mental weakness you're avoiding reality we get valuable mental exercise by playing complex music around 500 BC Pythagoras carried out some experiments with a monochord an instrument with one string and a movable bridge in the middle he concluded that when the ratio between the lengths of the strings could be expressed as small integers foreign the sound was consonant it sounded good but when the ratio could only be simplified to high integers [Music] it sounds dissonant conflicting string length is inversely proportional to frequency so this could be said about any sounds the tuning of scales varies between different cultures but universally music features at least some notes that fall on these low integral ratios especially when we play chords multiple notes at the same time so the observation that Pythagoras made holds up we like chords with simple frequency ratios but why once Sound reaches your ear it vibrates your eardrum which in turn vibrates the ossicles three bones that pass these vibrations along to your cochlea inside the cochlea is the basilar membrane and this is a strip of tishu that runs along the length of the cochlea the basilar membrane is designed so that the stiffness and other properties vary along its length so different parts of it resonate at different frequencies near the base of the cochlea responds best to high frequencies and at the tip it responds best to low frequencies all along the basilar membrane are these sensors called hair cells because they're each in a different position on the membrane they each respond best to a different frequency so effectively the cochlea performs a foyer transform it separates audio signals into different frequencies each Carousel is connected to a neuron which sends a signal to the brain saying that it heard this frequency neurons communicate primarily through electrical signals when a neuron receives chemicals called neurotransmitters from a sensory cell or from another neuron those trigger ion Gates which move positively charged potassium and sodium ions inside and outside of the neuron so there's a flow of current into the neuron at the same time all the charges that are accumulating on the inside and outside of the neuron are only separated by the thin cell membrane so this forms a capacitor on the edge of the neuron and the current source is charging up this capacitor of course the cell membrane isn't perfect at holding back the ions so some of the calculus are going to leak through this means that the membrane acts as a resistor so now we've turned our neuron into an RC circuit and we can analyze it just like we would in a physics class the key value that we're interested in is the voltage across the membrane the reason that we're interested in that is because once this reaches a certain threshold it'll trigger voltage-gated ion channels to discharge the neuron and then it'll send neurotransmitters to the next neuron and repeat the whole process so here's the equation for our neuron the input current which again depends on the other neurons and sensory cells that our neuron is connected to equals the leakage current plus the charging and both of these depend on the voltage which is what we want to solve for I realize this might feel long-winded but I promise this all comes back to music so let's say you're listening to a chord with two notes and they're both peer frequencies that means that two of your hair cells are being triggered and each one of those sends a signal to one sensory neuron we'll say these two sensory neurons hook up to one interneuron which takes a signal to your brain what we're going to do is we'll take our neuron equation and apply it to these three neurons the hope is that once we solve it we'll be able to plug in different frequencies for different chords and hopefully we'll see some difference in the signal that goes to your brain between good chords and bad chords so we'll start with neuron number one since it's connected to a hair cell the input is just a sine wave at Whatever frequency the node is but there's also a lot of noise in our brains there's so many random factors that could change the input current so we'll also add a term here that represents random noise neuron number two is exactly the same but with a different frequency for a different note neuron number three gets its input from the first two neurons and again the way it works is the input neurons will normally send close to zero current until they fire then they'll instantaneously send the pulse of current so we'll use a direct Delta function to model this it's a function that's zero everywhere except at the moment the neurons fire of course we'll have to solve for neurons one and two to figure out those times this system of equations can be and has been solved and here's the solution I don't think it's particularly enlightening so instead of solving it let me walk you through what typically happens and I say typically because that noise that we included makes the solution slightly random the current signal coming from the hair cell is generally not high enough to trigger the sensory neurons on its own so it takes the addition of our noise to actually fire during the first cycle of the sound wave that we're listening to the neuron is charging up so the moment that it's most likely to fire first is at the peak of the sine wave when the current input is highest if it didn't happen to fire at that time then the next most likely cancer is going to be at the next Peak so if we make a probability distribution of the sensory neurons firing times it'll look something like this a high peak after one cycle of the sound wave and then they get smaller after that on round number three the input from a single sensory neuron is also generally not high enough to trigger it and because of the resistor or charges leaking across the cell membrane if there's no constant current input then it'll eventually discharge so in order for neuron number three to fire it needs to receive a signal from one neuron and then really soon after receive a signal from the other neuron this needs to happen before it has time to discharge so the more often the signals from neuron 1 and neuron 2 line up the more often neuron 3 will fire and send a signal to your brain we can use this to make a probability distribution of neuron number three's firing times but of course it depends on the relationship between the two frequencies that you're hearing so here are some probability distributions for small integer chords you can see that they're pretty regular the signal that your brain gets is organized and predictable but here are some probability distributions for large integer chords as you can see they're much more fuzzy it's not predictable when that neuron is going to fire we actually have a way of measuring this fuzziness it's called information entropy or Shannon entropy after its inventor to introduce it let me show you this picture this is the Audi sibo message and in 1974 we sent this picture through radio waves into the cosmos I guess as an attempt to introduce ourselves to whatever aliens might find it but pretend that you're an alien and your job is to watch the data from a radio telescope and notify someone if you see a signal that looks like it's from Aliens most days you'll just see something like this random noise but then one day you see this which one of these signals gives you more information clearly this one even though you don't know the meaning it's so organized that it must be an intelligent message see you already have an intuition for entropy a signal that appears more organized is more likely to contain information a high entropy signal like this is probably just noise but a low entropy signal like this tells us something if you were just shown each of these signals then the low entropy one carries more information now here's the counter-intuitive part let's say that you know that both of these signals are from Aliens they're both intentional now which one gives you more information this one does the one with higher entropy see the low entropy organized signal follows simple rules you could recreate it by only knowing a few things but to recreate the high entropy signal you would need to know each bit so you actually gain more information by understanding the messy signal the messy signal is ambiguous but decoding It ultimately gives you more information the entropy of neural signals reaching your brain is low for consonant low integer chords it's high for dissonant High integer chords and this makes sense in a lot of ways I mean if you hear a C major chord on a piano then of course it was intentional it carries a simple message and it's unlikely to happen by chance somebody is probably reading music and playing it on the other hand if you hear three adjacent chromatic notes then it could just be that something fell on the piano on the surface you might not gain information from it but if somebody was reading music that directed them to do that then it would carry a profound amount of information because there are hundreds of bad chords and only a few good chords when it's less organized you have more to work with nevertheless our brain prefers the unambiguous case and that's why we like certain chords was this guy from foreign that was advice from jazz saxophone player George garzon in case he didn't catch that he said that one of his favorite piano players played a G minor chord as Just A and B flat a traditional G minor chord sounds like this and this is that piano player's way of playing it [Music] sometimes the music that we listen to lets us know the exact answer it tells us if we're supposed to feel happy or sad or unresolved or scared but with just an A and A B flat you don't know what you're supposed to be feeling think of all the chords that fit over those two notes [Music] even with some context and Melody you still need to think to figure out what the musician wants to say and that's the definition of high entropy it's no coincidence that according to our analysis of neural firing times this is a high entropy interval when Claude Shannon introduced the concept of information entropy he called it that because the disorganization of information is clearly analogous to the disorganization of matter which we call entropy and thermodynamics and statistical mechanics but maybe there's another similarity between the two in matter of entropy always increases on global scale and this is just a result of statistics if you drop food dye into water there's only one state where all the dye molecules form the word bird but there are trillions of states where the molecules look random so over time they'll tend to look random this is the second law of Thermodynamics but maybe human culture follows a second law of information I mean modern films music visual art literature all of it depends on ambiguities that are left up to us to understand a single spoken sentence can contain so many layers of information that are completely absent something like a computer programming language even day-to-day functions like determining of somebody's lying or if they understand you are all difficult to process because human speech has such high entropy but listening to music might be our way of training for that so jazz music and Away drumming really aren't that different they both train us to process difficult information that might be the best benefit that music gives us of course you can't listen to Hard music all the time

Channel: Physics for the Birds

Views: 657,407

Keywords:

Id: Gc5eICzHkFU

Length: 17min 48sec (1068 seconds)

Published: Wed Dec 21 2022