Perceptions of musical octaves are learned, not wired in the brain

I was skeptical of his prior dissonance study. I feel there is a bit of accommodation by the participants to maximize the surprise of the researcher. Getting goods for singing funny.

The paradigm I would have liked to see is where two tones are held constant and the participant moves a third tone to the location that sounds the best or the worst.

It is remarkable work. But I don't believe it. Octaves are just too basic a phenomena to be viewed as a cultural invention.

This study does not show what the title claims.

Alternate title: Tsimané people do not perceive melodic harmonies.

This is also the case for some non-Tsimané people (google Amusia). There could just be a genetic basis here.

For comparison, you'd need a Tsimané child raised in Western culture, or reverse.

>The researchers acknowledged in the study that the results might reflect differences in how people sing, and not in how people perceive pitch. But they argued that the totality of collected data, including some more direct measures of octave perception, put the weight of evidence on the side of a perceptual explanation. [..] Moreover, if the Tsimané’s performance in the tests has more to do with singing ability than their perceptual experience, then it would mean that all people have absolute pitch

Since I only have access to the abstract: How did they argue this in detail? Because that is something that immediately sprung at me. In the process of learning to sing, reproduction of relative intervals is far easier (and much earlier) learned than correct absolute positioning °) which actually takes some effort to train.

And the reasoning in the second paragraph eludes me completely. If I haven't completely misunderstood something, this has nothing to do with absolute pitch, and everything with tonal memory, which - again - won't work if it isn't trained: https://en.wikipedia.org/wiki/Tonal_memory

°) EDIT: Apology, the choice of words is actually incorrect. It's still relative, but with reference to the notes heard before, in contrast to the reproduction of intervals where you have your own previous sung note as (control) reference. Simply put: it's much harder to learn to hit the first note (and thus the general pitch) of the sequence correctly than the intervals in the sequence.

Quite a bold title. My understanding (while limited) is that the human ear senses sound by means of tiny hairs in the inner ear that vibrate in resonance with incoming sound waves. I've always believed that the reason certain intervals are more harmonious to us than others was because there would be some overlap in hairs that resonated for the different harmonious pitches. The same way guitar strings will resonate with other strings that are tuned an octave, or perfect fifth, or fourth higher, these ear-hairs ought to behave by the same physical laws and principles.

As such, I'm a bit skeptical of the method and conclusion of this research. It seems like physics to me, not psychology. Ability to reproduce pitches and intervals does not strike me as entirely related to what they were trying to find out.

How about instead playing different soundbites, of consonant and dissonant pitches, and asking for their opinion in some way? Maybe even allowing the participant to find a relationship by some sliding instrument themself? That would seem more aimed at studying the Tsimané's ability at "perceiving octaves."

Maybe my interpretation of the header and conclusion is wrong. Perhaps the idea is that _some training_ is needed in order to perceive octaves, the same way _some training_ is needed in order to learn how to walk. That I might be able to wrap my head around.

Music is definitely an acquired taste. My family was not very good at nor into music. I didn't understand the point of instrumental music until my mid teens. Before that I only listened for the lyrics.

On a different note, people often think of the octave as the most fundamental interval, but the most fundamental is the very same tone twice. Even here people like a little dissonance. Two tones at nearly but not exactly the same pitch will produce a pleasantly shimmering chorus effect.

Anyway music is subjective and there's is no sound that is better than everything else in all respects and contexts.

I wonder if this is more a problem with the article than the study. I haven't read the study. It feels oddly described. Like, are they saying that when a Tsimané sings a song (solo, without instruments) on Tuesday, and then later sings the song on Friday, it's guaranteed to always start on the same pitch, without hearing some sort of reference note from an instrument? If so, that's basically perfect pitch.

We all have comfortable vocal ranges, and I know generally where to start the Star Spangled Banner to hit both the high notes and the low notes, but since I have relative pitch and not perfect pitch, my starting point is still always going to vary within a minor third or so.

Incidentally, I noticed in my college ear training courses that I (I have excellent relative pitch) would routinely score higher than the folks with perfect pitch. There's something about perfect pitch that can be really distracting to musicians for certain exercises, as it's not quite as flexible as relative pitch. So maybe that's related to the Tsimané.

Also, there are aspects of different notes that very much are noticeable in physical reality and don't need to be culturally learned. In an old singing group of mine, we would have fun with a game where we would stand really close to each other, face to face, and sing unison, and then one of us would start to go slightly sharp or flat while the other would try to stay steady. There's an ugly "beating" sensation when this happens that goes away when you come back to unison. It's not just theory, you hear/feel it. That's present to a lesser extent with octaves, and then fifths, through the harmonic sequence.

Those subtleties are of course covered up when you're using separate instruments or timbres, but using different timbres doesn't disprove the underlying presence.

I find this hard to believe from a mathematical point of view - octaves are as far as you can get from arbitrary which is I guess what gives the headline it’s shock value.

> The acoustic structure of octaves is always the same: The frequency of a note in one octave is half the frequency of the same note in the octave above

This is not a great definition; our understanding of frequencies came well after we were using the concept. The idea of a "string of half the length with the same tension" is much more natural, and is the easiest mechanical analogy. The discovery of this is very natural because you get sympathetic vibrations -- you can feel the other string moving when you get consonance across octaves. It shows up (along with fifths) in most other instruments -- blowing twice as hard in a flute or halving the tube size, or compressing the embouchure in brass instruments.

While I don't think it's intrinsic in the human brain, it is intrinsic in most musical instruments, and it's kind of intrinsic in the way sound works, so it would be very surprising if it didn't feature in the music made by a culture, and thus work its way into the appreciation of music in general. The tests described here are interesting but might just point to the fact that there is less instrumental music (produced or listened to) in this culture than in others.

I made a website that describes natural harmonic (just) tuning systems:

http://just-tunings.info/learn

I do need to add a section on cultural consonance/dissonance but..

Octaves are not cultural. Resonance is not just an aspect of music or even animals, but literally everything in nature that makes sound. Dropping a rock in water will create sound, and frequencies that resonate with each-other will stick around longer than those that fight, teaching us that these sounds belong together.

An octave is the simplest possible relationship between two frequencies, and it can be heard in almost every natural sound imaginable. We learn they belong together from almost every sound we hear.

As an aside, western tonality uses the twelfth root of 2 because if you just start hammering randomly on those tones, they'll get along extremely well regardless of whatever the biggest bassist vibration is. The big bassy frequency being the one that every other frequency needs to get along with (harder than it sounds). It's sort of a fudgy form of relativity that allows things to vibe together mostly, but not perfectly.

I feel looking for musical scores in peoples brains is a great place to start decoding. We clearly load data into a conscious signal somehow like CAN-bus but it must also be stored maybe on actin filaments in the dendritic spines as loops we load in? Also mentally we can change the music and add any number of instruments over the top change the tune and deviate the course of the melody, add vocal harmonies from crowds of people and violins etc. so where does that part happen?, in the signal?. can it be 'heard'? . and is the music 'composer' happening in the same part of the brain as the language generating part or is it just similar capabilities in a different part of the brain?. Decoding must be possible. I feel if they had like millions of brains all wired up to computers that were using AI to find match patterns . i.e. audio patterns from songs in brains data. they could brute force it? Is getting people to sing stuff and making assumptions weak science?

I'm a bit skeptical of this

The concept of Octaves (as a lot of concepts in music) comes from physics. Because one of the most basic instruments you can have is a string under tension.

Pluck the string: note

Pluck the same string pinched in the middle: same note one octave higher

Not to mention Note + Note(octave higher) will have the harmonics overlapping, so it will sound better since there won't be any beating.

It reminds me of how tuning can depend on the era. Sometimes even on the composer. When you're used to hearing something in a particular way, the slightest change can be jarring.

https://www.youtube.com/watch?v=kyQaSFgnVI8

Musical octaves are wired into physics. If a note is at frequency X Hz, then the note one octave above it is 2X Hz. Of course due to the harmonics of the string, a string vibrating at X is also vibrating at 2X, so when you play a note at X Hz, then you are also hearing the note one octave above at the same time.

So... it doesn't need to be learned... When you hear a middle A, you're also hearing the A above for sure. And when you hear the A above by itself, you're hearing one part of the A below.

Of course, understanding that what you're hearing is a called an octave is almost certainly learned. But in my experience, any one who is not tone deaf will naturally lower or raise their voice by an octave if you ask them to sing a song 'lower'.

I never understood octaves. Two sounds of different pitch that are supposed to sound "the same" somehow? Obviously if you shift a melody by multiplying every note by 2 will keep it sounding the same melodically, even though it's immediately obvious it's higher pitched. Beyond that, I have never been able to perceive the "sameness" between equivalent notes on different octaves. When I was younger and played adventure games, sometimes there would be a puzzle involving listening to a short melody and playing it back on some instrument. Literally the only way I could solve those was by analysing the original melody with a spectrometer, then analysing the instrument, and figuring what sequence of keys to press that way.

mmmm I'm skeptical of a singular study in which a certain tribe of people are asked to modulate a sound to an octave within their singing range disproving octave equivalence being hardwired.

Wanna know why? Our aural circuits have frequency binning, much like an FFT, and they "ring" at certain frequencies, and an octave is either 1/2 or 2x a given frequency/"note". Closer examination of this system under stimulus is what I would consider a gold standard in this field of research. The form of research undertaken in this study suggests many things, and the researchers have chosen a particular conclusion and it's not necessarily indicated by even their own data.

https://www.britannica.com/science/ear/Analysis-of-sound-by-...

Given the current propensity for thinking that frequency detection is pre-cortex (lower level than speech decoding) and available rather early on in the evolutionary chain, I'm going to be interested in physical studies, not social science.

Perhaps it depends on the environmental noise you are exposed. In nature, most noise is percussion-like, non-rational harmonics, so octaves don't make much sens. OTOH listening to electric motors and combustion engines since the day you are born, makes you 'like' octaves, fifths and thirds, since these rotational machines generate these intervals while doing noise.

I hope that’s true because I still don’t have a clue what an octave is (or a melody or harmony or any other music related word).

It was a surprise to me to learn that octaves have a physical. mathematical basis. I discovered this as a teen reading about how to make music on a computer, where the relationship of notes to frequency was explained. To my ear moving up an octave had always sounded like an arbitrary increase in pitch.

Did they try to sing along to the original notes? Perfect pitch needs to be practiced, but it should be immediately apparent when you sing out of tune if you're even slightly musically inclined.

I'm a little frustrated with the lack of discussion of frequencies in this article, which seems like a key bit of background to understanding this work. An "octave" is two sounds where one of them has exactly double the frequency of the other. I think it makes sense that octaves aren't hardwired, however it is fundamentally a mathematical property of frequencies.

I'm not sure how I feel about the thesis; can children tell if something is twice the size of another or do you need the concept of "two?" I think like other people I initially found this a little unlikely because octaves are innate physical and mathematical properties, but there are tons of physical/mathematical properties that require a lot of study to understand intuitively.

Seems like most of the disagreement in the comments here are due to people having different notions of the word octave. This should be no surprise as there are numerous interpretations of the word according to Merriam-Webster:

1: an 8-day period of observances beginning with a festival day

2a: a stanza of eight lines : OTTAVA RIMA

b: the first eight lines of an Italian sonnet

3a: a musical interval embracing eight diatonic degrees

b: a tone or note at this interval

c: the harmonic combination of two tones an octave apart

d: the whole series of notes, tones, or digitals comprised within this interval and forming the unit of the modern scale

e: an organ stop giving tones an octave above those corresponding to the keys

4: the interval between two frequencies (as in an electromagnetic spectrum) having a ratio of 2 to 1

5: a group of eight

So maybe everyone's right.

The Tsimane man could probably identify the distinction between octaves of bird calls in his habitat. The experimenter would be in the opposite chair hearing only a handful of notes.

As a music teacher and music therapist, I have observed this behavior for three decades. It's especially true with people that have zero musical background.

I'm not too surprised by this. I've had awful relative pitch all my life, and just recently started ear training. I've also for fun read a bit about how absolute pitch develops in young children.[1]

Bit by bit, I'm realising how much of what's taken for granted by practitioners of Western music is in fact learned.

As a concrete example: ever since I was a child, I've heard people speak of major and minor triads as sounding happy or sad. That seems like something that's just obvious to a lot of people. It's never been obvious to me, and it lead me to think I was (at least partially) tone deaf; if I can't even hear something that is (evidently) that obvious to so many people, how can I ever get better at truly listening to music?

Now that I've started ear training, I'm beginning to mentally associate particular sounds with these chord qualities[2] to the point of actually being able to identify them better than a flipped coin would. (Which is a huge milestone for me!)

But it's taken a lot of practise. And I still don't hear them as "happy" and "sad". If I try to listen for "happy" and "sad" chords, I go back to mischaracterising these chord qualities just as much as before. I've simply never learned that association, for whatever reason.

----

It goes on and on: I played a lot of piano (from sheet music) when I was young, and practised the C major scale up and down and up and down, resulting in the interval from the tonic to the major second being etched into my brain very strongly. And I suspect this is what has made it surprisingly hard for me to internally hear a half-step from the tonic. To my brain (before I started ear training), the minimal possible step from the tonic was to the major second. There just weren't any sounds my brain could produce in between.

Learned. With plenty of ear training, I'm now in addition learning to be able to reproduce a minor second too.

Even more speculatively, I've noticed there are some specific half-step intervals I hear as whole steps. These are from E to F, and from B to C. These are half steps, but they are outliers in how often I attempt to classify them as whole steps. I suspect this is because of said C major scale practise: somewhere in the back of my mind, I might associate these tones (in the absolute pitch sense) with steps of the C major scale and therefore think of them as whole steps. I don't have absolute pitch in any useful sense, but it wouldn't surprise me if my brain, somewhere in a back compartment somewhere, has retained the sound of these tones and associated them with adjacent steps in the scale I practised so much.

----

Circling back to octaves: Yes, there are physical bases for considering octaves as a special case of some sort, namely that if you strip out the base resonance frequency of one tone on a string or wind instrument, what you're left with are its overtones, which also makes up its octave's frequencies.

But the fact that this relationship is something that makes these tones sound "the same" might very well be learned. The fact that this level of consonance is even desirable might very well be learned. I can picture cultures in which that level of consonance is considered uninteresting, hollow, and weak. Why would one deliberately care to seek something like that?

----

[1]: By listening a lot to music that extensively and unpredictably uses the 12 tones of Western music in A440 equal temperament tuning, children can learn to recognise these specific 12 tones in that tuning – but it won't help with any other frequencies!

[2]: I now hear major triads as a combination of hollow and triumphant, and minor triads as fuller and more epic. Some parts of it make sense, others do not.