registration without learning

Comments and questions about AP Avenue.
aruffo
Site Admin
Posts: 1694
Joined: Tue Dec 14, 2004 12:09 pm
Location: Evanston, IL

registration without learning

Postby aruffo » Mon Jun 11, 2012 1:39 am

An interesting item that I post here mainly to remind myself of its existence, as documented in

Hintzman, D.L. & Curran, T. (1995). When encoding fails: Instructions, feedback, and registration without learning. Memory & Cognition, 23(2), 213–226.

From the first paragraph of that publication:

Repetitions of a stimulus can "register"—as revealed in increasing judgments of frequency—without the subject becoming better able to discriminate it from another, highly similar but distinct, stimulus (Hintzman, Curran, and Oppy, 1992). In this article we present evidence that the latter failure reflects a powerful cognitive bias against learning more about an already-familiar item's structure, and that it may take overt responding and error-correcting feedback to overcome this bias.

aruffo
Site Admin
Posts: 1694
Joined: Tue Dec 14, 2004 12:09 pm
Location: Evanston, IL

Postby aruffo » Fri Jun 29, 2012 6:49 pm

And from the same paper:

Unless subjects are required to overtly predict the distinguishing feature and are thereby confronted with their errors, they may be oblivious to their failure to master the task. This account applies equally to learning the structure ofan individual stimulus and to associating items with arbitrary category labels like A and B (Experiment 2). If the subject sees TRUCK-A, and if both TRUCK and A are highly familiar in the experimental context, there may be no basis for realizing that the association has not been learned.

There is an important difference between the present idea and the ART model, however. That model assumes that perceptual experience drives two kinds oflearning: (1) ifthe mismatch between bottom-up and top-down sources of information is above a certain threshold, this triggers creation of a new template; and (2) if the mismatch is below the threshold, the old template is tuned. Our proposal, in keeping with Johnston and Hawley (1994), is that when the mismatch is relatively minor-as when a small percentage of bottom-up features are missing from the template-even tuning of the old template may not occur. Indeed, to the extent that the system is unable to differentiate between bottom-up and top-down sources of information, a reliable signal for tuning presumably would not exist.

WildcatShred
Posts: 15
Joined: Tue Jul 01, 2008 4:54 pm
Location: Lexington, KY

Postby WildcatShred » Mon Mar 25, 2013 9:22 pm

This definitely seems key, Chris. APA allows us to pick out C, but not to learn it.

I can't figure out why the concept formation never occurs separate from APA, though. Granted, colors have a more meaningful/useful role in our lives...but sounds are all around us. It would seem that if we actively choose to pay attention to them all the time, similar to the way we do in APA, that we would eventually acquire absolute pitch.

I suppose in order for concept formation to occur, there needs to be meaningful purpose in distinguishing pitch, as well as frequency and variety of presentation.

Just thinking out loud...

Is there anything (albeit dissimilar) other than color, that requires new concept formation as an adult? And can we learn anything from it?

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Mon Mar 25, 2013 11:13 pm

WildcatShred, the issue right now is that we don't have the necessary perceptual experience to build concepts and interpretations upon. People with absolute pitch categorically perceive chroma. The concepts they learn for pitch are associated with pitch categories.

For example, the white piano key left of the two black keys is associated with the C pitch category. The letter "C" itself is associated with the same pitch category. Without pitch categories, any concept formation will be of a single discrete point on the chroma spectrum we learn to hear in APA -- and we lose awareness of such a discrete point after enough time spent away from APA.

I do have questions about the process Chris will attempt to use to create pitch categories (see Amount of Time to Go Through All 12 tones), but nonetheless, they must be formed. That is the current state of affairs. And waiting for Chris to wrap up that PhD :D

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Tue Mar 26, 2013 12:54 am

DISCLAIMER: In the above post I made the assumption that hearing a particular chroma wears off after not playing APA for awhile. In reality, I don't know whether or not this is true (or whether APA really gets you to hear chroma for that matter), but it's a way of looking at the issue I'm currently toying with.

WildcatShred
Posts: 15
Joined: Tue Jul 01, 2008 4:54 pm
Location: Lexington, KY

Postby WildcatShred » Tue Mar 26, 2013 4:26 pm

I agree with Chris, Zac. I think you have to hear Chroma by necessity in APA.

The only other option would be that you memorize what a single note in every octave sounds like, such that you can pick them out as though they are individually different from one another. This doesn't match up with the accounts of the experiences of people playing the game that talk about "getting in the zone" and not knowing where a note is within a chord, but at the same time getting a certain feeling that let them know the note they were looking for was present. I myself have had the same experience. This seems to indicate the perception of chroma.

I agree with you, though...APA requires we perceive it, but not categorically. So the question still is...how do we communicate to our brains that they need not just experience pitch but also categorize it.

Repetition seems like an obvious ingredient, but certainly not the only one or we would all have this ability.

I'm trying to rack my own brain for how this worked with other things for me as a child. The way it happened for me was...

I would learn that a firetruck was red. I didn't totally understand the concept of red, but I completely understood the concept of firetruck...and not only that, but firetrucks are COOL to little kids, so that makes them important. So if I were to look at an apple, and someone said "What does this have in common with a firetruck?" I would probably think that was a silly question at first. You eat apples. You don't eat firetrucks. What would they have in common, and why would one ponder such a thing? But if I thought about it enough, and observed them enough, eventually I would realize something about them LOOKS similar, and then aha! I would begin to form the concept of Red. The concept would begin by association with apples and firetrucks. But eventually I would see more and more things that I needed to understand as Red...then the associations would become secondary, and "Red" would live independently as a category in my mind.

Perhaps this is the approach that should be taken. After learning to perceive chroma, we must repetitively experience that perception in contexts that we subconsciously deem important to recall. Perhaps singing the tones spontaneously might help, but it seems sensical to me to practice picking out the pitches in (first simple, and gradually more complex) melodies within APA. Melodies are more meaningful than random smatterings of notes. It puts the pitches in a meaningful context.

In the meantime, maybe it would be beneficial to use pitch paths at the same time as APA? Not sure...I know some people complained about the trigger melody associations interfering with their perceptions. But as discussed, children start to perceive categories of color based on associations, and then over time the associations fall away and the categories are strong enough to stand on their own.

I enjoy your enthusiasm on this Zac. I was there in about 2008. I read every word Chris wrote and spent the summer trying to figure this out. You are motivating me again.

I will check out the other thread.

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Tue Mar 26, 2013 5:22 pm

Hey WildcatShred, I'm glad that I'm motivating you to take an interest in perfect pitch again :wink: It's something I'm extremely passionate about and would give anything to learn in my lifetime. I won't elaborate too much, as it could end up taking a lot of space and time, but believe you me: this is something I'm dedicated to working out. Only going to stop if it turns out to be impossible (which Chris doesn't think so, but I'm not really sure of).

I should clarify what I said previously. Enough perceptual differentiation should necessarily lead to chroma perception. We all agree on this (you, Chris, and I). The reason I am unsure APA completely isolates chroma is because I am unsure there is enough perceptual differentiation and because I think there is a better way to implement perceptual differentiation. First off, there are some pretty crazy timbres in day-to-day life that perfect pitchers can name the pitch of. For example: sirens, odd bird calls, horns, and people's voices. Secondly, the way APA is set up right now has you differentiate two groupings of sound (once you get to higher levels, 101+) with multiple pitches, and each grouping will contain the target pitch. In the thread I mentioned above I write about the idea of a "chroma key", which I think would be an improvement on the perceptual differentiation learning process.

Here, in a nutshell: APA gets you to compare two sets of sound by separating each grouping into an egg. It's then up to the listener to figure out that the sound at the beginning of one grouping is at the end of the other grouping and perceptual differentiation occurs. I figured a more effective way to perceptually differentiate your target pitch would be to press a key at the beginning of one grouping to trigger the C sound and press the same key again at the end of another grouping to trigger the C sound. Pressing the "chroma key" would hopefully get the listener to compare the two sounds directly. The trick is just getting the brain to compare to sounds -- perceptual differentiation naturally follows.

A final point: You're attacking the categorical perception problem the way Chris initially did. He sought out the function of colors and their associated concepts, hoping that using pitch in the same way as colors are used would lead to categorical perception. Phase 17 of his research documents this. However, after some emailing back-and-forth with Chris several months ago, he let me know he has rethought the categorical perception problem and, likewise, altered his plans for developing categories. His new idea is far more in line with what I believe will lead to categorical perception (using colors as a model for developing categories always seemed off to me, especially my first few times reading Chris's articles).

Our perception of color is extremely complex -- not only in the brain but also at the input level of our eyes. Our eyes have red/green/blue cones and rods (this is an over-simplification as color and black/white vision are actually separate in cones and rods). But basically, I don't like color as a model for chroma categorization because it doesn't only rely on how our brain interprets the stimulus, like perfect pitch, but also how our visual sensory system (eyes) are built.

Instead, there has to be a pattern and novel presentation to the stimuli we are dealing with, chroma, that should lead to categorical perception. And categorical perception should come before concepts and associated objects can be properly learned. Children don't learn the names of sounds and develop perfect pitch consequently. They categorically perceive chroma and learn the names for the different sensations they hear for each pitch category. And the patterned presentation of chroma Chris intends to use as of now is a normal distribution of pitch about an idealized central pitch chroma. Evidence that this method may work is found in phoneme categorical learning research. It's a logical next step, and until we try it out and see what happens, we won't really be able to make much real progress moving forward. So I continue to train on APA every day, as it's still a necessity in the learning of perfect pitch, and wait until Chris wraps up his PhD and hops back on the perfect pitch research train -- starting off with some normal distributions around centralized pitch loci.

P.S.
It's awesome that Chris had spent 6 years researching perfect pitch before coming up with APA. And that you've been following him along since 2008. It's comforting to know there are others as dedicated as I to the pursuit of this elusive ability. I'm only 17 right now, so I remain hopeful (most of the time) that we'll crack this nut in the coming years.

WildcatShred
Posts: 15
Joined: Tue Jul 01, 2008 4:54 pm
Location: Lexington, KY

Postby WildcatShred » Tue Mar 26, 2013 6:43 pm

That makes sense, Zac. Structurally, there IS a difference. To what consequence? who knows.... But the phoneme approach makes a lot of sense.

I feel the same way that you do about the ability. I have some friends with perfect pitch, myself...and my life would be much different if I had this perception, certainly.

There are a couple of things that give me hope that it IS possible. Our brains DO perceive chroma if we are aware of changes in pitch. That is self-evident.
Also...there ARE occasions when I will hear a random environmental sound and know the pitch....especially if I am in a groove of working on my ear training frequently. There is no consistency to it though, and this rarely happens when listening to music.

I'm wondering and anxious about this new approach. Hopefully Chris is done soon!

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Tue Mar 26, 2013 11:02 pm

WildcatShred wrote:Our brains DO perceive chroma if we are aware of changes in pitch. That is self-evident.
Also...there ARE occasions when I will hear a random environmental sound and know the pitch....especially if I am in a groove of working on my ear training frequently. There is no consistency to it though, and this rarely happens when listening to music.


Unfortunately, what we relative pitchers perceive as changes in "pitch" are really just changes in -- and a movement in the same spectrum (tone height) -- not a change along categorically perceived cyclic spectrum (chroma).

As for relative pitcher's ability to recognize environmental tones... it's unclear whether or not we're activating the same areas of our brain as perfect pitchers. Certainly, tone naming does not activate the same areas of the brain as perfect pitch recognition, but we don't know whether APA changes the area we use in our brain or not.

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Tue Mar 26, 2013 11:04 pm

To Chris: I know you keep up to date reading the forums but just can't find the time to write lengthy responses.

Could you just give me a thumbs-up or a thumbs-down on everything I've written so far (especially the long post)? Just want to make sure, in your opinion, I properly understand it all.

Axeman
Posts: 151
Joined: Wed Jul 20, 2005 4:20 pm
Location: New Zealand

Postby Axeman » Wed Mar 27, 2013 4:41 am

Kia ora koutou,
Hello everyone,

Lately while trying to train myself to hear scale degrees I have noticed that my ear gets tricked by the succession of notes because the intervals between two successive notes can throw me off the feeling of the degree of the note as it relates to the key center. I have listened to the audio file a number of times and while I get a fairly good result most of the time I tend to get tricked in specific ways. One is when the 3rd degree is followed by the 7th degree in the same octave. Another is when the 2nd is followed by the 3rd in the octave below. The former fools me because the interval is a fifth which is a perfect interval and almost sets up a feeling of a new key in my ear the 7th sounding like the 5th of the new key. The latter is a large interval and I suppose just causes me to lose the feeling of the key note. These things happen consistently enough for me to think that these are situations where the method of training has become ineffective because of the 'similarity of stimulus not being able to be discriminated'.
All of this seems to suggest the truth of this threads initial quote and the concept it enunciates.
Repetitions of a stimulus can "register"—as revealed in increasing judgments of frequency—without the subject becoming better able to discriminate it from another, highly similar but distinct, stimulus (Hintzman, Curran, and Oppy, 1992). In this article we present evidence that the latter failure reflects a powerful cognitive bias against learning more about an already-familiar item's structure, and that it may take overt responding and error-correcting feedback to overcome this bias.

I liked the idea of the apple versus fire truck analogy. Maybe the failure of APA to teach categorical perception (CP) is that the task being asked to be performed is too one dimensional ( i.e. hear the chroma in the trigger melody/chord and hit the right coloured egg). The thing that strikes me about the firetruck verses the apple is that understanding that they are red in colour has little to do with their function (firetruck for playing with, apple for eating) and so noticing the redness is incidental. But the redness has greater significance if say the adult asks the child what colour the apple is because in responding correctly he/she earns the admiration of the adult. This is high motivation for CP. Again, if there is a significance to noticing the colour of an object then the motivation for the brain to perceive the colouris greater. E.g. 'Red means stop.' or " The red one is the phillips screwdriver" (amongst a bunch of other coloured screwdrivers) - both have different reasons for being chosen but the task or concept gives the redness significance leading to CP perhaps. [/quote]

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Wed Mar 27, 2013 7:15 pm

Axeman wrote:I liked the idea of the apple versus fire truck analogy. Maybe the failure of APA to teach categorical perception (CP) is that the task being asked to be performed is too one dimensional ( i.e. hear the chroma in the trigger melody/chord and hit the right coloured egg). The thing that strikes me about the firetruck verses the apple is that understanding that they are red in colour has little to do with their function (firetruck for playing with, apple for eating) and so noticing the redness is incidental. But the redness has greater significance if say the adult asks the child what colour the apple is because in responding correctly he/she earns the admiration of the adult. This is high motivation for CP. Again, if there is a significance to noticing the colour of an object then the motivation for the brain to perceive the colouris greater. E.g. 'Red means stop.' or " The red one is the phillips screwdriver" (amongst a bunch of other coloured screwdrivers) - both have different reasons for being chosen but the task or concept gives the redness significance leading to CP perhaps.


Just because a stimulus is significant does not mean it will begin to be perceived categorically. If I offered to pay you $1000 every time you passed a categorical perception test of brightness, you would not develop categorical perception just because you want my $1000 (just like the kid who wants the admiration of the adult). My thoughts on categorical perception of color vs. categorical perception of chroma remains the same... The physical sensory systems we have for each and the areas of the brain we use in the perception of each are too different to use color as a model for learning to categorically perceive color.

I emailed with Chris about giving sound meaning in our day to day lives. A previous proposition was that computer programs be opened by singing a particular pitch. Or a certain tone played means it's a certain time of day on an alarm clock. In both of these cases, because the sounds are significant, a listener will pay attention to them. However, the listener will simply learn to sing a tone at a certain height and will learn to recognize a particular sound object, not the pitch. Making sound meaningful WILL lead to learning, but relative listeners will use the faculties they already posses to extract that meaning: they will learn the height and overall feeling of a sound object. Categorical chroma perception must come before concept formation...

And my thoughts on the color analogy... Phonemes are learned via the loci method (Chris, does the learning process have a formal name yet?). But observing normally distributed lengths and brightness probably won't lead to categorical perception of those spectrum. Length, brightness, and color are too far removed from chroma to be used as accurate models. To me, phonemes seem like the way to go.

And just so my other post doesn't get buried...

To Chris: I know you keep up to date reading the forums but just can't find the time to write lengthy responses.

Could you just give me a thumbs-up or a thumbs-down on everything I've written so far (especially the long post)? Just want to make sure, in your opinion, I properly understand it all.

aruffo
Site Admin
Posts: 1694
Joined: Tue Dec 14, 2004 12:09 pm
Location: Evanston, IL

Postby aruffo » Thu Mar 28, 2013 12:47 am

I doubt it.. I think you're better off reading the (scientific) literature to make sure you understand it directly, rather'n my second-hand assessment.

zacxpacx
Posts: 157
Joined: Tue Oct 23, 2012 3:51 pm
Location: United States

Postby zacxpacx » Thu Mar 28, 2013 3:42 pm

Very well. Could you point me in the direction of the text regarding loci of phonemes? I know you have compiled a list of almost all relevant literature to perfect pitch, but research as new as the loci learning isn't in your listing.

WildcatShred
Posts: 15
Joined: Tue Jul 01, 2008 4:54 pm
Location: Lexington, KY

Postby WildcatShred » Fri Mar 29, 2013 9:19 am

Zach,

If he does so, please post it here. I want to read as well.

Thanks.


Return to “Absolute Pitch Avenue”

Who is online

Users browsing this forum: No registered users and 1 guest