Is Auditory perception a learned process or innate?

[jkeny]

I'm not talking about the organs, I'm talking about the ear/brain interface; perception. I believe somewhere along the line, before I can even remember it, my Dad probably told me "that sound" was a car horn, put a name on it, but then I recognized car horns, in all their variety, for what they were going forward. Didn't have to learn it, the tools were already there. I just had to be told what "that' was.

Of course you're welcome to disagree.

Tim

Tim, you are dealing with this at too simplistic a level.
What I'm talking about is much more fundamental than your example. Before you were even able to recognise the sound of a car horn (although you didn't know the name of it yet) your auditory perception had to first analyse all the other sounds that were happening around & were causing pressure waves to hit your ear all at the same time & jumbled together. It's like a big jigsaw puzzle being thrown at you & you have to work out what pieces fit with what & then be able to isolate the individual elements in the jigsaw picture. But all of this is being done on a set of jigsaw pieces that are changing all the time & you have to find the solution in realtime & make sense of the elements in this moving picture.

This is the job that auditory processing has to do but we are only born with the tools not the technique for doing this. So I'm talking about the subconscious learning of the techniques that allow us to achieve this fundamental aspect of auditory processing - what is called auditory scene analysis. This is the internal auditory reference map that we are talking about - where the rules of what goes with what & what can be expected to be where are found. It's these rules that we reference subconsciously to decide if something makes auditory sense or not - how close a playback system comes to a realistic illusion.

It's somewhat like grammar - we aren't taught it - we learn the rules through listening & experimenting & eventually form subconscious rules which allow us to correctly form sentences that will mostly be understood by others (because they have the same set of internal rules).

 

[jkeny]

If a dozen people sit down and listen to a twango instrument for an hour, and then a week later they listen to a recording of the twango instrument, although they each have different hearing and ear brain interfaces, I suspect they will all recognize the sound of that twango when it is played the most realistically possible by whatever speaker manages to hit most of those sensory buttons.

How can they have different ear brain interfaces & yet independently agree on the same ranking? They must have the same set of internal rules by which they judge the sound & therefore they must have built these rules using very similar ear/brain interfaces?

 

[esldude]

Hold on, I've always seen it said that bias influences our perception of what we hear, particularly sightedness - how can "a natural world bias doesn't disagree with reality" - a bias is a bias - it means that it is influencing what we perceive so therefore is changing our perception from what it would be without the bias.

A bias at seeing a larger hollow instrument and expecting it to put out a lower note is not contradicted in reality as it will put out that lower note. The bias is still there, but it isn't wrong in such a case. Any cognitive maps developed from such experiences aren't wrong in practice either. It is the source of them.

Yes but this is not due to anything other than the balance mechanism in the ears are picking up signals which the eyes are not agreeing with - it's the McGurk effect but with different senses at play.. I'm not sure the point you are making here?

It is a bit more than the McGurk effect. But the point is senses that work well, even with reasonable bias, are fooled by artificial conditions. Flying an aircraft that creates G force in one direction and tells the eyes something different is not a condition we find ourselves in enough to adapt.

There is only one solution that I see which I already used in a reply to the same point you made in your previous post What I mean is that over the long term we seem to be able to tease out the correct perception from the various skews that the biases might be introducing.

I think it the reverse, the bias is reasonably correct. Which means in artificial conditions they can be wrong yet we trust them. We aren't teasing them out the correct perception, the perception is simply usually correct without teasing.

 

[Mosin]

I think most audio perception is rooted in survival, so a product of evolution, not training. But that's just what I'd guess. I don't know.

Tim

Primitives who live in the Brazilian rain forest can distinguish between over forty different species of animals by the smell of their urine. That ability is critical for their survival. Those from industrialized societies who are put into the same environment cannot distinguish between those scents. What does it mean? Did the natives learn to smell, or do we filter scents in order to avoid an overload? Or, did we not bother to learn the nuances because they aren't necessary for our survival?

Apply the same to hearing. Do we learn how to hear? Or, do we unlearn how to hear, only to learn part of it again for our own amusement?

I vote both.

 

[esldude]

How can they have different ear brain interfaces & yet independently agree on the same ranking? They must have the same set of internal rules by which they judge the sound & therefore they must have built these rules using very similar ear/brain interfaces?

Not hard to imagine, and again it is the real experience in the physical world that ties it together.

For an over simplified example, 4 of the people are young with excellent hearing. Two are old with bad high frequency hearing loss. One fought in the war and has tinnitus wiping out most of his upper midrange perception. 1 is a DJ and loves him some heavy, heavy bass with a predilection for that. 2 are in between with some hearing loss at high frequencies, but not much. 3 are musicians who hear okay, but mostly pay attention to the musicianship.

All are exposed to the same physical reality of the twango. All might very well give the same preference vs what they heard in physical reality though all heard it differently. True the guys with no highs might not notice a speaker with missing highs. But in those ranges they still hear they likely will select the same basic overall preference.

 

[NorthStar]

I did not take the time to grasp the first original post in depth, so I omitted my vote for now.

One thing is for sure though; I don't feel the same intensity of the music playing when I'm shaved and not. ...I prefer when I'm shaved. ...And I just don't know why exactly...I don't try to understand...it's just like that.

Women don't even think about that.

 

[microstrip]

(...) I was going to start a poll to ask two easy questions & maybe discuss some issues leading from the answers:
1. - Are we born with our auditory perception ?
2. - Is our auditory perception learned/formed due to our exposure to the real world?

(...)

John,

I have recently read that some aspects of our auditory perception are learned/formed. Height of the source emitting sound is one of them. We do not have an innate capability of localization of height, we perceive it mostly by changes of timbre due to reflections on the boundaries and by spectral cues generated by reflections within the pinna. However, compared to inter-aural differences responsible for horizontal localization these elevation spectral cues are weak and vary considerably between listeners. It is known that people who suffer changes in their ears geometry due to accidents or surgery have to re-learn how to perceive source of sound elevation.

Most "height cues" are said to be personal and are impossible to predict. Perhaps this can explain our eternal disagreement about the perception of height in stereo.

BTW, I am addressing natural sounds, not manipulated recordings to artificially create elevation.

 

[Phelonious Ponk]

Tim, you are dealing with this at too simplistic a level.
What I'm talking about is much more fundamental than your example. Before you were even able to recognise the sound of a car horn (although you didn't know the name of it yet) your auditory perception had to first analyse all the other sounds that were happening around & were causing pressure waves to hit your ear all at the same time & jumbled together. It's like a big jigsaw puzzle being thrown at you & you have to work out what pieces fit with what & then be able to isolate the individual elements in the jigsaw picture. But all of this is being done on a set of jigsaw pieces that are changing all the time & you have to find the solution in realtime & make sense of the elements in this moving picture.

This is the job that auditory processing has to do but we are only born with the tools not the technique for doing this. So I'm talking about the subconscious learning of the techniques that allow us to achieve this fundamental aspect of auditory processing - what is called auditory scene analysis. This is the internal auditory reference map that we are talking about - where the rules of what goes with what & what can be expected to be where are found. It's these rules that we reference subconsciously to decide if something makes auditory sense or not - how close a playback system comes to a realistic illusion.

It's somewhat like grammar - we aren't taught it - we learn the rules through listening & experimenting & eventually form subconscious rules which allow us to correctly form sentences that will mostly be understood by others (because they have the same set of internal rules).

I like simple examples, but that doesn't change my point. I think we are born with most of this fundamental auditory processing ability, and the rest of it develops naturally, in infancy. It is not, as far as I can tell, a "learned process," but I'm not sure it matters much. What are you driving at with this line of reasoning?

Tim

 

[esldude]

Jens Blauert says in his book Spatial Hearing:

During the development of the individual the auditory world differentiates itself. Auditory events at first diffuse in their locatedness, become more precisely defined spatially; the correspondence to the visual world and to the other senses also become more precise.

This just after he is discussing that the idea auditory hearing has to be learned through experience to ascribe location to sound is incorrect. So it sounds like his opinion is that ability to hear and perceive spatially is innate to the hearing mechanism which develops itself early on.

 

[Phelonious Ponk]

Jens Blauert says in his book Spatial Hearing:

During the development of the individual the auditory world differentiates itself. Auditory events at first diffuse in their locatedness, become more precisely defined spatially; the correspondence to the visual world and to the other senses also become more precise.

This just after he is discussing that the idea auditory hearing has to be learned through experience to ascribe location to sound is incorrect. So it sounds like his opinion is that ability to hear and perceive spatially is innate to the hearing mechanism which develops itself early on.

At the very least, it develops so early that i'm not sure it makes any difference. Does anyone remember learning how to differentiate sounds coming from above? How to recognize their mother's voice? Whether or not we're born with it is almost a moot point.

Tim

 

[microstrip]

Perfect pitch is one very interesting subject of auditory perception. Although we have several theories, it is generally accepted that it is a innate process that can not be learned.
But there is disagreement - some people say that it is a learned process that only develops in the very early years and that we mistakenly accept it is innate for a small percentage of people.

 

[jkeny]

A bias at seeing a larger hollow instrument and expecting it to put out a lower note is not contradicted in reality as it will put out that lower note. The bias is still there, but it isn't wrong in such a case. Any cognitive maps developed from such experiences aren't wrong in practice either. It is the source of them.

I think you re playing fast & loose with the word bias - a bias is tendency towards some skew AWAY from the correct position - it's not a tendency towards the correct position - a bias is always wrong, no? What I am saying is that a low note played on a large instrument will sound deeper than the same note played on a smaller instrument due to our cognitive bias. Take away the sight of these instruments & the two note will sound the same. But in laying down our internal mapping we don't have to close our eyes to do so - we do so with all biasing intact & functioning in the real world. How does this not influence the internal map we all have created?

It is a bit more than the McGurk effect. But the point is senses that work well, even with reasonable bias, are fooled by artificial conditions. Flying an aircraft that creates G force in one direction and tells the eyes something different is not a condition we find ourselves in enough to adapt.

I think it the reverse, the bias is reasonably correct. Which means in artificial conditions they can be wrong yet we trust them. We aren't teasing them out the correct perception, the perception is simply usually correct without teasing.

Again, you are using bias incorrectly - there is no such thing as a "reasonably correct bias" - a bias can only draw you away from the correct view. A bias is like a filter - which is more correct looking through a filter or looking without the filter? Which is more correct auditory perception without bias or with biasing factors?

 

[jkeny]

Primitives who live in the Brazilian rain forest can distinguish between over forty different species of animals by the smell of their urine. That ability is critical for their survival. Those from industrialized societies who are put into the same environment cannot distinguish between those scents. What does it mean? Did the natives learn to smell, or do we filter scents in order to avoid an overload? Or, did we not bother to learn the nuances because they aren't necessary for our survival?

Apply the same to hearing. Do we learn how to hear? Or, do we unlearn how to hear, only to learn part of it again for our own amusement?

I vote both.

Good point moisin. Yes we are born with the underlying mechanisms necessary to sense the world - how we educate/train these senses does depend on what we are exposed to on a daily basis & what is important for our survival. I'm sure there have been animal experiments which have investigated this - which wouldn't be allowed now - confining an newborn to a restricted acoustic environment.

 

[jkeny]

Not hard to imagine, and again it is the real experience in the physical world that ties it together.

For an over simplified example, 4 of the people are young with excellent hearing. Two are old with bad high frequency hearing loss. One fought in the war and has tinnitus wiping out most of his upper midrange perception. 1 is a DJ and loves him some heavy, heavy bass with a predilection for that. 2 are in between with some hearing loss at high frequencies, but not much. 3 are musicians who hear okay, but mostly pay attention to the musicianship.

All are exposed to the same physical reality of the twango. All might very well give the same preference vs what they heard in physical reality though all heard it differently. True the guys with no highs might not notice a speaker with missing highs. But in those ranges they still hear they likely will select the same basic overall preference.

Identifying an instrument is different from ranking speakers in order of preference. You can be colour blind & still identify grass, but not be able to differentiate the colours between red & green.

Coming up with the same preference ranking for speakers means that we are all mapping in pretty much the same way. Therefore we have all worked out any biasing in this mapping to come up with the same internal map.

 

[jkeny]

I like simple examples, but that doesn't change my point. I think we are born with most of this fundamental auditory processing ability, and the rest of it develops naturally, in infancy. It is not, as far as I can tell, a "learned process," but I'm not sure it matters much. What are you driving at with this line of reasoning?

Tim

Training the nascent senses by exposure to the real world is what I mean by "learned process", Tim. As the example by moisin showed, we pay attention & internalise what we are exposed to that is of importance to us.

I'm trying to investigate how we are not affected by biasing in the "training" of these senses.

 

[jkeny]

John,

I have recently read that some aspects of our auditory perception are learned/formed. Height of the source emitting sound is one of them. We do not have an innate capability of localization of height, we perceive it mostly by changes of timbre due to reflections on the boundaries and by spectral cues generated by reflections within the pinna. However, compared to inter-aural differences responsible for horizontal localization these elevation spectral cues are weak and vary considerably between listeners. It is known that people who suffer changes in their ears geometry due to accidents or surgery have to re-learn how to perceive source of sound elevation.

Most "height cues" are said to be personal and are impossible to predict. Perhaps this can explain our eternal disagreement about the perception of height in stereo.

BTW, I am addressing natural sounds, not manipulated recordings to artificially create elevation.

Yes, micro, I think our perceptual senses are developed in stages - the fundamental, basic building blocks, like frequency/amplitude/location we may actually be born with - aren't newborn babies tested in these ways? Development of these basic signals into meaningful representations of the auditory world is what comes later based on our experiencing of the natural world & it's operation.

One of the fundamental functions that we learn is how to analyse & differentiate this jumble of signals into an auditory scene - the so-called auditory scene analysis. This is critical to our understanding of the world - there is some evidence that this is one of the (perhaps the main) issues in autism.

It's the "learning" of this function which fascinates me & hence my questions.

I've said it before but I believe this is akin to how we learn the rules of grammar through experiencing the natural world of speech. Nobody teaches us these rules we somehow all develop them through teasing out the examples of speech that we encounter as children. I believe Noam Chomsky's theory of universal grammar is that the brain is hard-wired to "learn" this "mental grammar" or the ability to construct intelligible sentences.

 

[Phelonious Ponk]

I think you re playing fast & loose with the word bias - a bias is tendency towards some skew AWAY from the correct position - it's not a tendency towards the correct position - a bias is always wrong, no? What I am saying is that a low note played on a large instrument will sound deeper than the same note played on a smaller instrument due to our cognitive bias. Take away the sight of these instruments & the two note will sound the same. But in laying down our internal mapping we don't have to close our eyes to do so - we do so with all biasing intact & functioning in the real world. How does this not influence the internal map we all have created?

Again, you are using bias incorrectly - there is no such thing as a "reasonably correct bias" - a bias can only draw you away from the correct view. A bias is like a filter - which is more correct looking through a filter or looking without the filter? Which is more correct auditory perception without bias or with biasing factors?

You've both got it wrong. Bias is prejudice against or in favor of a thing. It doesn't change the definition if the bias lines up with reality. That's the definition of bias. What it means to audiophiles depends on the audiophile, it seems, and therefore doesn't mean much at all.

Tim

 

[jkeny]

You've both got it wrong. Bias is prejudice against or in favor of a thing. It doesn't change the definition if the bias lines up with reality. That's the definition of bias. What it means to audiophiles depends on the audiophile, it seems, and therefore doesn't mean much at all.

Tim

I'm using it in the sense that it is typically used in audio forums - a cognitive bias interferes with the true/correct experience of a stimulus, therefore it can't be a bias towards truth or correctness it is always AWAY from this truthful or correct experience of the stimulus.

Using this as the definition, it applies equally to natural sounds as it does to reproduced sounds - whenever a bias is at play there is a skewedness away from the correct experience of the sound.

 

[Phelonious Ponk]

I'm using it in the sense that it is typically used in audio forums - a cognitive bias interferes with the true/correct experience of a stimulus, therefore it can't be a bias towards truth or correctness it is always AWAY from this truthful or correct experience of the stimulus.

Using this as the definition, it applies equally to natural sounds as it does to reproduced sounds - whenever a bias is at play there is a skewedness away from the correct experience of the sound.

You are, in my opinion, an by definition, using it in an inaccurate sense that could lead you to the conclusion that removing that bias can only result in negative conclusions, and that's not true. You can expect to hear accurate reproduction and hear it, even when it is not there. MLs are a perfect example in these discussion.

Tim

 

[microstrip]

(...) It's the "learning" of this function which fascinates me & hence my questions. (...)

The F. Toole book is filled with references of this "learning". Sometimes we "learn" how to ignore things - needed for the "suspension of the disbelief" .

In order to analyze this subject we must use an hierarchy to put some order in the debate - Toole refers to the Blesser and Salter (2007) perspective in his book, considering sensation, perception and meaning. From the Sound Reproduction

Sensation At the lowest level is , an indication that the organism reacts
to a sound—a detection threshold. This is probably quite well related to
physical measurements of the sound.

Perception The next level is , which incorporates cognitive processes
embracing cultural and personal experiences. Here we recognize what it
is that we heard, and perhaps initiate a process of adaptation. This
means that some features in measurements may be neutralized by
adaptation, and no longer be relevant.

Meaning At the highest level of response to sound, we attribute to the
recognition, and this can range from irrelevant to highly relevant, from
undesirable to good. Depending on the informational content of the
sound, we may choose to pay attention or to ignore it.