Acoustic Learning, Inc.
Absolute Pitch research, ear training and more

Phase 1:  Discovering the Ear

August 1 - What's genes got to do with it?

I suppose it's inevitable that I should weigh in on the nature vs nurture argument-- is perfect pitch genetic, or is it learned? Obviously I think it can be learned, or this entire website would be a colossal waste of time; but what about the genetic question?  How does genetics fit in?

Interestingly, genetic researchers-- and by "genetic researchers" I'm referring to the UCSF people and the sources that they've linked to-- are now acknowledging that even if absolute pitch is genetic, there is an "environmental component" to it. Specifically, they believe that there is some critical window of age in which the child will either trigger the ability-- or fail to trigger it-- and that after that window closes, they either have it or they don't. Typically, they say, the trigger is early musical training, which "teaches" the child to have AP.

That view seems to contradict every report that I've read from people with spontaneous AP. Every one of their stories reads the same way: "I always perceived the pitches, and once I got some musical training I learned names for the notes." "Musical training" taught them nothing-- except how to put into words what they already heard. Furthermore, the musical training that they do receive is contrary to their ability, because music study and music theory is based completely on relative pitch-- and teaches that the word "pitch" means "if we say a note is high or low" rather than describing a quality inherent to the note itself. If AP is truly a genetic "gift", I wonder how, then, the genetic researchers believe that the ability could possibly be lost after this "critical age" is passed?  Does a person's gene disappear?

The explanation that I seem to see, from interviews that the researchers have given, is that people with spontaneous AP actually have a "genetic predisposition" rather than a specific on/off genetic gift. That is, they're telling us that AP is not like having green eyes-- which some people have naturally and other people don't-- but it's something more like a preternatural talent for drawing, or for baseball. It's certainly true enough that some people have a "talent" for drawing or sports and can do well almost out of the cradle, while others labor through years of training to reach a competitive level, and still others can't seem to do well no matter how hard they try.

In the movie Pre, about the life of Olympic runner Steve Prefontaine, there's a scene in which Steve's coach explains that Steve has unusually high lung capacity. Steve Prefontaine thereby had a natural advantage over other runners. Clearly "Pre" had a genetic predisposition for running; it meant that he could do effortlessly what they must labor to achieve. But other runners were, nonetheless, still able to beat him.

There definitely seems to be some kind of genetic advantage which makes the incidence of AP more likely. Research has shown that people with AP have certain brain areas which are commonly enlarged, and a recent article describes that there is now known to be a specific segment of the brain which can be attributed to AP. It seems probable that people who are born with these "advantages" will be more likely to have or develop AP.

But what about those people who have to learn it or it goes away?  There's the trick. The geneticists concede that there are people who develop AP. By acknowledging an "environmental factor" they introduce the idea that AP must be learned... yet people with spontaneous AP say unequivocally and emphatically that they never had to learn it (I received a letter from one such woman who said she's terribly weary of having to explain that fact). Why should we think that the presence of some super-normal brain structure, which allows some people to know without being taught, means that no one else is capable of learning? People who do not have super-normal lung capacity can still train to be Olympic runners.  If AP is not an unchanging genetic mandate like green eyes or wavy hair-- if AP is something that people with talent can learn-- you could describe any skill that way!

So what does genetics have to do with it?  Even the researchers don't seem to know what they will do with a genetic connection if they happen to find one.  I'm not sure how the information could be used for anything other than a curiosity, any more than the knowledge "good runners have better lungs" could be applied to anyone's benefit.

August 2 - There are none so blind

For some reason, in the past two or three days I've seen repeated references to the fact that the incidence of absolute pitch is much higher in Japan than it is over here in America.  In each case this has been attributed to the fact that music there is most typically taught by the Suzuki method.

As I read these references, remembering my own Suzuki experience (learning piano from age 4-15, as I mention elsewhere on this site), I kept wondering how on earth Suzuki method could possibly induce AP?  In order for that to happen, the children would have to be encouraged to listen to notes individually, and nothing in my experience supported that.  I learned note names so that I could read sheet music.  I learned that "C" was this key here in the middle of the keyboard.  "C" to me was a dot on a page or a strip of ivory in its set place; I was never taken to the keyboard and informed that this-- plunk-- was the sound of a C.  Although the heart of Suzuki method is indeed listening, Suzuki method is about listening to music, in order to reproduce what you hear-- they have CDs and tapes of the compositions in each new book of sheet music, which the students must listen to carefully.  And, as I touched on yesterday, music comprehension reinforces relative pitch comprehension, not absolute pitch.  You hear melodies.  You hear harmonies.  You learn fingering.  You play chords.  You hear crescendo, decrescendo, and dynamics.  You hear, play, and learn change-- notes changing to other notes, or notes being dampened or emphasized. You don't ever just stop and pay attention to any single pitch.  How could that experience ever create AP, if the entire process is designed to turn your attention away from absolute pitch, to teach you to use relative pitch exclusively?

Tonight, however, I was reminded that my experience is with Suzuki piano.  The larger part of Suzuki method is Suzuki violin and other stringed instruments.  Stringed, fretless instruments.

Although the Suzuki method is still substantially the same, the nature of the instrument itself would make AP more likely to appear.  Instrumentalists who play fretless strings are the most likely to understand what it means to listen carefully to a single pitch-- because their instrument requires them to do so.  The bow only draws across one string at a time, is capable of playing only one note, and because there are no fret guides the only way a musician knows that they're playing the correct note is to listen to the pitch just as a pitch-- just as itself, and not as part of a piece of music.  The piano, with its multiple simultaneous notes and keys that don't require any skill to produce an accurate correct pitch, forces the student completely into the realm of relative pitch-- but the string student is required to listen carefully to single pitches right from the beginning.  This kind of attentive listening, combined with the Suzuki encouragement to listen carefully to the qualities of the prerecorded music, does seem likely to make AP more prevalent.

This brings me back to yesterday's comment about how "good runners have better lungs", genetically.  Of course I wanted to make a rhetorical point-- that a recognizable physiological difference can't be learned or taught-- but I knew as I was writing it that there actually is an answer there.  Once you know that better lung capacity is beneficial to running, you can examine what other runners have been doing to train themselves and figure out how what they're doing might be altered to make their air usage more efficient (or at least more effective).  That is-- once you analyze the genetic "gift", you use your knowledge of why it's beneficial to teach other people how it's done.

But genetic researchers seem to deny that absolute pitch can be taught.  They seem to think you either have it or you don't.  That's why I felt some sympathetic chagrin when I read an interview with one particular researcher.  This woman wanted very much for her child to have perfect pitch, saying "I tried to give my daughter piano lessons by the time she was six years old.  It became obvious to me that she does not have perfect pitch, so I didn't push it."

This woman presented her child with piano lessons-- perhaps the one instrument least likely to induce AP, and, indeed, the instrument most likely to ruin it.  Then she stood back, crossed her fingers and hoped, and when the magic didn't happen she gave up.  She never attempted to teach her child perfect pitch, because she doesn't think it's possible.  She doesn't know how to teach perfect pitch because she doesn't know what perfect pitch is.  But she does know that languages can be learned; she does know how that can be taught.  And wouldn't you know?  The daughter "is doing a beautiful job learning Chinese."

August 3 - When all dead horses are in bed

I don't like the definition of perfect pitch.

I've discovered that there are plenty of people who complain that "perfect pitch" is itself a poor term, because the ability is not actually "perfect"-- and although that point is not incidental, that particular semantic flaw doesn't bother me.  What frustrates me is not the term "perfect pitch", but the fact that every time I encounter the definition of that term, it seems to be this:  "Perfect pitch is the ability to accurately name or produce a tone without an external reference."  Dan Levitin takes it two steps further by adding "...presumably through reliance on a highly developed internal template, or self-referencing mechanism. People with this ability are able to retain absolute information about sounds along the unidimensional continuum of auditory frequency, and they are able to attach labels to these sounds."

The problem that I have with this is that the definition is being attached to the result of the phenomenon, rather than the phenomenon itself.  It's like saying "Turning on a light switch is defined as being able to see every object in the room."  Turning on the light is not unrelated to seeing what's in the room, but this supposed definition tells us nothing at all about the actual causative process-- that of flipping the switch.

By now you've probably encountered-- even if only on this website (although if you've made it here, you've probably read at least a few other articles)-- some stories by or about people who have perfect pitch.  And each of them say the same thing-- their experience of perfect pitch is the perception of the notes.  The identification and naming of the notes is simply something that naturally follows.  Levitin's extended definition is much clearer, but still allows the interpretation that the ability is that of naming notes, or the ability to retain information about pitches-- not the ability to perceive a note's unique characteristics.

This mislabeling makes it more difficult to learn perfect pitch.  If we think that perfect pitch is naming a note, then we play a guessing game that we win or lose at the moment we speak a letter name.  If we think that perfect pitch is remembering a note, then we keep trying to memorize some single note on the scale.  And, indeed, if an accurate definition of perfect pitch were "being able to name a note" or "remembering information about notes", either of these strategies might have a chance of success.  But how can someone ever learn to flip a switch if they spend their time and effort trying to see objects in the dark?

Perfect pitch is neither note naming nor memorization.  Perfect pitch is the sensation you experience when the note is received by your ear.  When learning perfect pitch, we speak the letter name because we have no other way to confirm our perception.  When learning perfect pitch, we attempt to recall information about pitch because we want to make sure we have an accurate mental construct of what we are going to hear.  But our construct is a representation of what we hear-- and if we do not know what we are hearing, our mental construct is flawed.  The goal of perfect pitch training is to become familiar with the experiences of perceiving each individual pitch, and as long as a person continues to think that perfect pitch is "the ability to name notes without an external reference", that person can only learn the wrong things.

August 4 - I'm my own grandpaw

Relative pitch trumps perfect pitch.  Musicians who have trained their ears say this.  They can't learn perfect pitch, they say, because the relationships between the notes are so strong in their mind.  They can learn one note and then use it as a reference, perhaps, to get all the others, but they can't learn perfect pitch-- or, at least, they're severely handicapped when they try-- because they can't "turn off" their relative pitch.  Fortunately, there are software packages which make it impossible for you to use your relative pitch, and that is the only way you can hear the notes "cleanly" without any sense of their relationship to each other, because relative pitch is a terrible obstacle to the process.

To that I say balderdash.

Relative pitch is a great boon to learning perfect pitch, and I think my task of learning would be a heck of a lot easier if I had it.  People do think that relative pitch "gets in the way" of learning perfect pitch, but from all accounts I've heard or read, the reason is because they think that the ability of perfect pitch is the ability to name the notes.  If you happen to remember the note that was just played a moment ago, then now you know for a fact what this new note is.  Dadblast it!  How can I learn perfect pitch and guess what the note is if I already know?  Well, if perfect pitch were indeed "the ability to name a note with no external reference", then you would be sunk-- you have an external reference, and your relative pitch finished the task for you; you are stuck with nowhere to go.  But perfect pitch is not the ability to name notes.  Perfect pitch is the ability to perceive notes, and that perception exists independently of linguistic labels or reference notes.  With an appropriate understanding of perfect pitch, relative pitch is at the very worst merely irrelevant, and at best it's a wonderful help.

There is nothing about hearing the relationships between notes that actually prevents us from hearing an individual note within that relationship.  If we listen to music, play an instrument, or understand relative pitch, we will comprehend that a note is "higher" or "lower" than another note, but we will still recognize that all the A-sharps sound similar to each other; all the Ds sound similar to each other; all the B-flats sound similar to each other-- this is something we all can tell even when we do not actually know the letter names of the notes we are hearing.  If relative pitch made it impossible to hear the identity of an individual pitch, how could you hear, sing, and identify all three of the notes in a random triad?  If it were impossible, you would be able to name the chord, you would be able to construct the chord intellectually, but you would not be able to hear any of the notes within it.  But you can hear the notes.  Relative pitch does not interfere with your ability to hear a note.

The task of learning perfect pitch is to listen to a note; the goal is not to name it, but to hear it.  The hardest part of learning perfect pitch is knowing what pitch you're listening to and applying the label to the experience of listening.  If you have strong relative pitch you've already got that covered; you know what note you're listening to, and that gives you the freedom to relax and listen and think to yourself, "So that's what a B sounds like."  Someone with relative pitch who thinks that their task is to name a note will play the note, speak the note's name, and move on to the next note.  But the name of the note is a linguistic association-- the letter "A" is not a pitch any more than the word "elephant" is a giant mammal.  Relative pitch doesn't stop you from hearing a note.  It helps you name it.

August 5 - If you have to ask...

A couple weeks ago I was stabbing at a few embers in a couple of different newsgroups, and the flash of one particular response interested me.

When I came across the idea of learning perfect pitch last year, the timing was just right-- I had already been thinking about the fact that sensory experience is a "pure idea".  An ordinary object can be described intellectually-- "a plastic tube with a cylinder of ink inside" could reasonably help a person to understand the concept of a "pen", but there is no way to intellectually communicate any physical sensation.  Words and linguistic descriptions of what you receive with your senses merely allow you to represent your own experience of sensation to someone who already understands that experience. When you speak the name of the sensation you want to share-- "heavy"; "prickly"; "blinding"-- the hearer can only understand you by associating your words with his own visceral experience. If the person has never experienced what you're trying to communicate, you can only help them understand by comparison with what they do know-- "tastes like chicken, but stringier"-- or by causing them to experience that sensation directly ("Here, eat this.")

In this newsgroup post, its author indirectly makes one of the most critical points of learning perfect pitch: in discovering how to listen to a pitch, it is very important to hear what a pitch is, but it's equally important to hear what it is not. He describes how his own young children are learning their colors. A child looks at a bicycle and a fruit on a page of red objects, he says, but does not conclude that "red is a name for all things which are fruits or bikes." The child compares the bicycle and the fruit to each other, and compares them to whatever else is on that page-- he understands that the thing they share is "red".  What is different about them is not red-- therefore what is the same is red.  Noting that the objects have different shapes, sizes, features, uses, and everything else that's different among them, the child by process of elimination strips away everything that doesn't match until all that's left is the color.

Here is a series of typed characters. What about it should you remember? What about it is important?


Maybe you picked out the semicolon because it's the only bit of punctuation; more likely you said "How the heck should I know?" All we have here is a random assortment, and I've given you no clue about what to look for. But if I ask you to compare that series to this one--


Okay.  Have you studied the two series?  Are you ready to compare the two? You have successfully memorized the first one flawlessly, so you can now look at the second series and, from memory, go through it letter by letter and identify what's different about it. The better your memory for the first series, the better and more quickly you'll be able to see what's different about the two of them... it didn't even occur to you to try to memorize the first series, did it. If anything, you just looked at them both together, and after a little bit of scrutiny and comparison you saw that the "Y" has been switched for a "Q".

If I told you you had to memorize the first series, you might do it, but it's a bit harder, isn't it? And the semicolon is just a distracting fake-out.

A lot of people think that you can learn perfect pitch by carrying around a tuning fork and striking it every so often to reinforce the pitch in your head. Or you should return to your instrument every so often to check yourself on a single note and see if you "got it right" each time. Or you should just memorize one pitch and relate that pitch to all the other ones that you hear. Memorize one pitch. Reinforce one tone.

That's not gonna help.

If you successfully memorize one pitch-- and I'm not saying it can't be done-- your experience is indiscriminate; you're attempting to recall not only the pitch but all the sound information about that pitch, including (at the very least) the overtones and timbre of the instrument it's played on. You're memorizing "SE48SL8NE8EGLSEGLHY8GUEA32;P38LO4". And if you try to memorize a second pitch as well, then you are trying to memorize the entirety of "SE48SL8NE8EGLSEGLHY8GUEA32;P38LO4" and "SE48SL8NE8EGLSEGLHQ8GUEA32;P38LO4". Again, it certainly can be done-- but isn't it easier just to learn the difference between "Y" and "Q"?  Plus, if you don't know that it's the Y and the Q that you're looking for, you can't and won't pick either of them out as the defining character when you see SDRLI78R97GTSLE74LW4JU5HLQW instead.

I'll be interested to know how many people might choose the semicolon as the "important part" before looking at the second series; the only difference between it and the other characters is the intellectual understanding that it's "punctuation" where the others are "letters and numbers"-- but the brain zeroes right in on it, makes it pop right out, makes you think it's important.  It's possible that this is parallel to trying to recognize notes as "fluffy" or "mellow" or whatever, but I'll let you be your own judge of that.

If you've tried the Ear Training Companion, you've probably noticed that one of the exercises is to play either a C or a D and identify which one you're hearing.  Subsequent exercises add additional notes-- E, F, G, et cetera, until all twelve tones have been added.  But we start with C and D, not just C.  That's because we are not memorizing these pitches.  We are not putting C and D into our intellectual memory.  We need to experience what's different about each new note, and the fewer differences we are required to detect, the easier it is to feel them.

You can only understand a sense experience by comparing it to a similar sensation.  But there is nothing in your experience that you can compare to a pitch-- except another pitch.  So every time you play the D, listen to it and hear two things: D and not C. Every time you play the C, listen to it and hear two things: C and not D. This is why it's distracting to think of the definition of perfect pitch as "identification without a reference note"-- you need reference notes.  You must play the notes, have them both in your short-term memory, and compare them to each other.  And, by comparing the notes to each other on the same instrument, you're doing more than just minimizing the amount of work your brain has to do in order to become familiar with the experiences of "C" and "D". Not only do you hear that C is "not D" and D is "not C" by identifying what's different about them, but by hearing what's similar between them, you also hear that the pitches of C and D are not the piano.

You will never be able to describe this experience. You can't make it work just by thinking about it.  You can't figure it out. It does not happen intellectually, any more than you can "figure out" the taste of salt until you've actually tried it. This only works through repeated exposure-- the brain figures it out for itself, and you will not know why.  The fellow who wrote this post says, somewhat despairingly, "I've been listening on and off to pitches for a long time and have still no idea what I'm listening for." But, paradoxically, it's impossible to know what you're listening for until you know what you're listening for.  For the past two weeks I have been identifying C, D, and E. And each day, I've been listening to each tone, assisting myself in my identification by associating adjectives:  "Okay, C is rounder... D is whiny... E is brighter..." and that has been successful in that it seemed to help me recognize the pitches.  But I recently had the extraordinary experience of playing a D and, well, it just sounded like a D.  No other words but "it's a D" were adequate, or even necessary, to describe what I heard. It "sounds completely different in some way, but not anything I can latch on to and use to identify it," just like the fellow in the newsgroup says.  I listened, baffled, because I knew exactly what I was hearing yet I had no idea what "it" was.

And that's the way it happens.

August 6 - Strummin' with the rhythm that the drivers made

"I wouldn't want perfect pitch. Just imagine-- hearing a concert that was slightly off key! That would drive me crazy. I'm better off not knowing."

I had heard of this argument against learning perfect pitch, but honestly didn't believe that it was for real until I heard two different people (both of them professional musicians) say it directly to me. It seems to be nothing but sour grapes; what it makes me wonder is, how do you know that? If you don't even know how it feels to hear when something's out of tune, what makes you sure that it's going to drive you crazy-- so sure that you won't even consider finding out?

I want perfect pitch because little by little I'm beginning to understand that there is music in the world all around me, music that I never would have even known was there. When I'm in the shower, each drop of the thousands that fall makes a note with its tiny impact. When I'm driving, there's the hum of the engine and the roar of the road and the whoosh of the cars that zip past me. At the office there's the clacking of my shoes on the pavement of the parking garage, the shudder of the elevator, and the harmony of the various fans and lights and machines humming all day inside. Life is a constant symphony, and I never knew it. With the expanded aural awareness that is perfect pitch, I will look forward to waking up each morning and opening my ears to the new day.

But the only rebuttal I've seen echoed with equal fervor is this: "Why should it bother anyone to hear something out of tune? You don't get crazy if your TV set has colors that are slightly off, do you? You don't say 'Oh, that shirt should be bluer than that.'"

This statement doesn't make sense either. Some people actually do go crazy when the colors aren't right. I have a VHS copy of the movie "Yellow Submarine", taped from a 1983 broadcast on a local Chicago channel. I adore this film, and over the years I've watched that tape no fewer than 47 times. But this year I was astonished to see the colors of the DVD, so incredibly bright and distinct, and now that I've seen the DVD I can barely stand to watch the VHS tape for that sole reason (those meanies should be bluer than that).  I only keep it around because the DVD version cut out some footage and wiped a big chunk of background music.

Will perfect pitch drive me crazy? Have I any way of knowing? I think about one piece of music that was practically ruined for me: one day, my voice teacher pointed out what bad technique it is to breathe audibly when singing. The next time I played "Goodbye" by Mary Hopkin (one of my favorite songs, written by John Lennon) I couldn't help but hear her sing: "Please don't wake me up too late.. GAAASP.. tomorrow comes... SUUUUCK.. and I will not be late..." and I turned the song off halfway through. What I used to enjoy about the song is constantly being interrupted.

But the song hasn't changed, and the VHS tape is the same-- what's happened is that my standards have been raised. I don't enjoy them as much because now I know there's something wrong with them. I find that I don't want to hear a singer with such a noisy lack of technique. Yellow Submarine on VHS "drives me crazy" because I can see the murkiness of its screen where I couldn't before.

Is this worth it? I can't say I'm not irritated. I would prefer to be able to watch those extra bits on the VHS tape and not be turned off by the lousy color. I would like to be able to ignore the singer's wheezing. But now I will be able to discriminate among singers whose superior craft I can appreciate. Now I can watch the Yellow Submarine DVD and be unusually impressed by how good it is-- because I know how bad it could have been. By improving my perception I'm driven to seek out quality, I will recognize quality when I see it, and I will appreciate quality that much more when I find it.

And, despite my specific disappointments, I think it's utter nonsense to say "I wouldn't want to have refined tastes because then I wouldn't like garbage any more."

August 7 - Behind door number C

I think the game-show mentality of win or lose is detrimental to the process of learning perfect pitch. The mentality is evident in a "success ratio" featured in some perfect-pitch software, which tells you how many notes you've guessed versus how many you've gotten right. The feature seems to reinforce the implicit mindset that naming notes correctly is the way to win.  But being wrong-- "guessing" wrong-- seems beneficial, often absolutely necessary, to learning perfect pitch. If you think that your goal is to guess correct notes, you may not understand how well you are doing if your ratio shows only 30% "correct". If you are convinced that your goal is to crank that ratio as high as possible, to be right at all costs, you may try harder and end up listening to the wrong things.

At those times you can't succeed, you need to be able to say "I don't know." Last week, a fellow who knows I'm learning perfect pitch approached me and played a note on his guitar. I did not recognize the note, but I wanted to "get it right" and not look ignorant. At the time, I had only familiarized myself with C, D, and E; if there was a chance that this note was one of them, I didn't want to miss it. I listened carefully, I listened hard, asked him to repeat it, and finally heard what I thought was C. I guessed C. It was a G. He graciously pointed out that C is a strong harmonic of G, and suggested that that overtone must have been what I heard. Of course he was trying to be kind, but he made an important point for learning.

Your mind can trick you into producing the wrong result.  If you don't let your ears hear the "wrong" pitch, your mind's desire to hear the right notes will cause you to fail. Because every note is actually a complex mess of many pitches, with the strongest pitch being the "fundamental" that names the note, it's essential to be able to acknowledge when you don't know. When he played me that note on the guitar, although I didn't yet know a G, I was familiar with C, D, and E. I could have listened to the note, heard the pitch of that note, and heard that pitch as "not C", "not D", and "not E". If I had done that, I could have told him with confidence "I don't know." There's nothing wrong with that answer, because I would have heard the pitch correctly. But instead of hearing what was there, instead of listening to the note that was given to me, I tried hard instead to find something that I could recognize. And what I heard was not just the "wrong answer", but it was the wrong thing to listen to.

It's hard to let yourself be wrong.  Very, very hard.  Whether or not you're literally keeping track of your guesses, "I don't know" is a bad guess. Whether or not you get a literal demerit in your success ratio, you know that you're not labeling the note correctly, and that makes it hard to say you can't. But you don't have to label the note in order to learn-- you have to hear it. When you don't know, you need to let yourself hear it as unfamiliar, and not try hard to hear something familiar about it.  That's difficult to do; it's terribly hard on our ego if it seems like we're admitting ignorance.  But the goal of learning perfect pitch is not to win a guessing game; the goal is to listen well.

August 8 - Speak and Spell

Today I finally got some legitimate input about what it means to have "genetic and environmental components".

My rhetorical question of "Does the gene disappear?" is seemingly on target, because genes don't disappear-- but my assumption is wrong. I overlooked that a person's genes are not always expressed.  This isn't merely the "recessive" gene that we all read about in our middle school textbooks, which gets passed on from generation to generation until it finally manifests.  Rather, all genes must be "activated"-- all genes require proteins to carry out the DNA's instructions.  If the proteins ignore the DNA, the DNA is never used.  Environmental factors, like musical instruction, may require the services of a certain strand of DNA, in which case that strand readily fulfills its purpose; but if there's no environmental factor to prompt it then a gene might not work, because there's just no use for it.  Eventually it deteriorates into uselessness.

The question then becomes, if there is a gene for AP, what does it do that makes AP possible?  There seem to be two neurological facts known about AP: that there is a certain area of the brain which is unusually large in those people with absolute pitch and that there is also an area of the brain which is directly associated with naming notes.  But if an AP gene were responsible for the hyper-development of these areas, surely this does not prove that AP can't be learned, but that people with these enhanced areas can learn it.  So which is true-- is perfect pitch something that "people with talent can learn", or is some unusual brain structure absolutely necessary?  I've asked this before, but this time there is a mystery factor in the works.  It may be possible that people who do not have an enlarged or hyperactive "AP center" could be trained to learn, as I've asserted, but it's also possible that without the activation of a specific AP gene the brain may develop in a very different way, leaving the necessary structures for AP incapable of executing their task.

Well, what is that task? No one seems to think that that task is that of hearing the note. Color-blindness is caused by a recognizable physical difference in the eyes' components, but if anyone has theorized even for a moment that the AP gene somehow affects the physical structure of the ear itself, I haven't seen any hint of it.  All research points to the brain.  If there are no physical differences in the hearing mechanisms of the AP and non-AP listener, then the sensation each receives is identical.  No one I've read is arguing that anyone's ears operate differently.  But are we hearing the same thing?  Basically, what we think we hear is influenced by what we expect to hear.  There is, further, explicit versus implicit content:  people can be given precisely the same input, and they agree that they are experiencing the same thing as each other; but they interpret it differently, and come away believing that they have had vastly different experiences.

Philosophically, the task seems to be knowing the pure experience of the pitch.  The author very quickly and firmly asserts that labeling a note is a "conceptual content"; linguistic labeling is an additional layer that need not be present in order to have a "perfect pitch experience".  Perfect pitch becomes, exclusively, hearing a sound and interpreting the "pitch" within that sound.  I have experienced this myself, already, when I heard a note that "sounded like a D". Philosophically, I did have a perfect pitch experience.  Psychologically, I may have perceived the same thing as a person with perfect pitch.

As I believe my brain to be musically ordinary, I conclude that perfect pitch is something we can all hear and interpret.  I can listen to the noises in my environment-- the whir of my air conditioner, the creak of my exercise machine, the beeping of the snooze alarm-- and I can hear and sing their pitches.  But I still can't identify those pitches flawlessly.  This is why learning perfect pitch is, in part, like learning a language. If I see someone with green hair, I am having the same sensory experience as a native Spanish speaker; yet they are able to add the "conceptual content" of Spanish words equivalent to "green hair", and these are words I have not learned.  If naming notes is, just like language acquisition, adding conceptual content to a sensory experience, then the "ability" of perfect pitch must be nearer to the definition that I don't like-- being able to associate the "perfect pitch experience" with the note labels.

Perhaps learning perfect pitch is like learning a language, after all.  The part that is enlarged in an AP brain is the same area as that which processes language and math. These are two skills for which I do have an above-average facility, and that does run in the family. My brother has a PhD in astrophysics and speaks two languages fluently, my mother has her PhD in international business and speaks five or six, and my father is a Great Books instructor (and a onetime professional editor) with a master's degree in literature. Both my mother and brother are "tone deaf", and are unable to sing on pitch.  My father, on the other hand, was once a clarinetist whom the conductor would rely on to tune up the band.  He didn't have perfect pitch, but he had a very keen sense of pitch.

But  I have to ask the question again-- what is the task that is "perfect pitch"? It's not merely perceiving the pitch, because that's something anyone can do.  It's not memorizing a note, because there are many complex waveforms in a note which have nothing to do with a pitch. It's not even memorizing a template of 12 notes, because scales vary from culture to culture (and era to era).  The task of "perfect pitch" is knowing the experience of pitch.  If we know it well enough then we can consistently describe it regardless of the words we use.  In our culture, the purest acceptable way to describe the notes we hear is by one of twelve letter names (okay, each accidental can be spelled two different ways, but you know what I mean). In another culture, there may be more than twelve or fewer than twelve. The AP person who has a "perfect pitch experience" in another culture may still accurately describe their experience. If someone has not learned letter names for their pitches, they will still be able to accurately describe their pitch as "chirpy" or "bright" or "round" or whatever they choose. If someone has learned letter names, they may say "an A that's a quarter flat"-- but that same pitch may have a more specific name in another culture (or have even been called "A" in a past era). But whether or not they have letter names to use, whether or not they use the same letter names, people with AP are able to understand their experience well enough to consistently apply an accurate linguistic description.

Is this comprehension genetic in nature? The "perfect pitch experience" is indeed a universal perception; is it really a genetic quirk that some people can tell you what they're hearing and others are at a loss for words?  I suspect it's just a matter of vocabulary.

August 9 - Bewitched, bothered, and bewildered

Suddenly I am exhausted.  To begin with, I'm overwhelmed by the amount of information that is available to read about perfect pitch.  None of it directly disconfirms anything I've been blathering about, which is both good (on the right track) and bad (nothing to challenge my premises). And also, now, I am reminded of the most difficult part of the process-- recognizing what part of the sound is a "pitch".

I was doing tremendously well identifying three notes.  Having familiarized myself with C, D, and E, it was relatively easy to play a random third, recognize one of the notes, and name the other one by relation. Or I'd recognize neither of the notes and reach one of them by thinking "up one" from E or "down one" from C. But then.. then, just a few days ago, I jumped ahead, thinking that I could just add F to the three notes I already knew. And now my life is hell.

I had read that the sound of a third is the most familiar relationship between any two notes in Western music, but I completely underestimated how potent that sound is. By introducing myself to the F, I am now dealing with both a major third (C-E) and a minor third (D-F) and I have become terribly confused.  Could it be that in identifying C, D, and E, I was instinctively hearing the bottom of a third, the top of a third, and the "other note"?  Now that I have a fourth tone, it's not so easy to make the distinction.  I keep mistaking C for D, E for F, and even F for C!  Worse still, now that I have included an F in my "sounds familiar" category, I can no longer easily tell if notes in any third are unfamiliar. I'm shocked and amazed at how similar the sound is of any given third to any other third.

Of course, now I see much more clearly why people can be frustrated by relative pitch, and want to "turn it off".  On the one hand, relative pitch is not "bad" if your short-term memory helps you identify the next note.  That's something for which we want and need relative pitch.  But because of relative pitch, our brain is completely prejudiced to think that every major and minor third are identical to every other.  Even when I hear the individual pitches inside the third, they mutate to fit my expectations: every "bottom note" sounds the same as every other "bottom note", and every "top note" sounds the same as every other "top note". I find myself hearing G and B and, even though the pitches are totally unfamiliar, their sound is so familiar that I identify them as D and F!  In a lifetime of listening to music, nothing could be harder to filter out than the sound of a pitch's relationship to other notes.  I heard from another fellow who was making good progress in identifying notes, but as soon as those notes were part of an interval he was totally lost.

Now I know how exactly how he feels-- but now I'm more convinced than ever that trying to remove relative pitch from the learning process is a terrible mistake. It'd be like taking all the derivatives and integrals out of calculus because "they're too hard". What this is forcing me to do is to listen so much more carefully, and to figure out how on earth to develop the discrimination which will recognize, separate, and distinguish between the sound of a note and the sound of a relationship.  It's just as true now as it was at the start: the most difficult part of learning perfect pitch is recognizing what it is about the sound we hear that can be selectively identified as "pitch".

Really, really difficult.

August 10 - I say thee neigh

I haven't really made much of an effort to answer "why is it so important to learn perfect pitch?"  Most obviously, I do have something to prove. I also want to fully appreciate the natural music of the world around me.  But the first reason really doesn't have anything to do with the aesthetic, and the second, admittedly, doesn't require the ability of perfect pitch so much as it does an ordinary attenuation.  There's a deeper answer-- an answer which relates to my love of music and my desire to make music.  I hadn't quite figured it out myself until today, but I was sure it had something to do with a neon horse.

I was in London for the summer of 1993, working a summer job at the British office of the Stanford Research Institute.  I usually typed the consultants' reports using any of a half-dozen Macintosh computers, but when I wasn't creating reports or presentations, I had time to twiddle.

One day, I had a lot of time on my hands, and I decided to use it to practice my MacDraw skills. Looking for a subject, I scanned the bulletin board and, underneath a thumbtack, discovered a photo of one consultant's beloved pet horse. I took the photo over to my workstation and began drawing. Macintoshes in that year had just made the transition into color-- even though all of the Macs and their software were color-enabled, all but one of the monitors were standard grayscale displays.  MacDraw is (or, was) a line-and-vector drawing application, not a pixel editor, so I defined areas of light and shadow and, on my grayscale monitor, outlined areas of appropriate shape and gave them appropriate shades. By the end of the day I rather proudly printed (on the grayscale laser printer) a faithful full-page reproduction of the horse's head. The consultant was delighted. The office manager was.. tolerant.

The next morning I decided to look at the file again to see if it needed any further tweaks. I brought myself to the office, plugged in the teapot, sat down at a computer, found the file, double-clicked it... and stared in dismay.  The office's one color monitor was now showing me the ridiculous patchwork jumble I'd made-- a neon rainbow horse.  Instantly I realized that I'd been working with color on a grayscale monitor.  What I'd seen as light grey was actually yellow. What I'd seen as medium grey was actually blue. What I'd seen as dark grey was actually red. Even something I thought was near-white was in fact a light pink. What a mess! I half-heartedly attempted to change a few of the drawn shapes to different kinds of brown, but after only a few minutes I decided that, since I had only the black and white printer, I might as well leave it as it was.

A couple weeks ago I remembered this neon horse; I was sure that it had something to do with perfect pitch.  Maybe the weird clash of colors would be like a weird clash of pitches?  Perhaps someone with perfect pitch would be able to hear more clearly which notes actually belong together? Perhaps someone without perfect pitch would, in all innocence, construct a piece which was relatively, structurally perfect, but which had no coherent theme? I discussed it with a composer friend of mine, trying to think of some kind of parallel between this misbegotten horse and her own profession, but despite my fervency she didn't find any of my suggestions particularly compelling. She wouldn't believe that someone with strong relative pitch, or at least a firm grasp of standard music theory, wouldn't create music just as good or better than someone with perfect pitch.

Last week I found out why she was right, and at the same time I found out why anyone would want perfect pitch.  What I found is a marvelously succinct answer about the beauty of absolute pitch, spoken by someone who is a musician with that ability.  My mistake in trying to understand the horse was making the assumption that a patchwork horse is somehow wrong (Andy Warhol was, of course, famous for creating images with the "wrong" colors), and trying to conclude that absolute pitch means better composition.  "It's utter nonsense to say that people without AP are good or no good," this lecturer says; "For us musicians, it's important to strive our utmost to polish the sense of the sound." When I constructed and printed the horse in greyscale, I was creating something that had complete relative integrity, just as a piece of music should. But even though I created a lovely picture, I was simply unaware of the power and feeling, the bursts of vivacious color, that actually existed underneath the greys. My creation existed with an extraordinary additional dimension that I couldn't even see. My not knowing about the color made it no less splendid a picture; my using what turned out to be the "wrong" colors made it no less interesting to look at.  But if I'd had the entire spectrum at my disposal when I drew it, I could have enhanced it in subtle and meaningful ways beyond simple black and white.

August 13 - A rose by any other name

Here's an interesting analogy that I came across in a Yahoo discussion group.  One of the group members talked about notes' relationships as analogous to, well, relationships. He likened a piece of music to a family reunion, in which you know the relationships of all the people to each other.  For example:  you're aware that Jane is Bob's wife, Millicent is Steven's grandmother, Sharon is Jim's estranged wife, that sort of thing. You can instantly perceive how Jane and Bob belong together (everyone calls them "Janeandbob"), how Sharon and Jim can't come near each other without sparring, how Steven brightens when Millicent boasts about his latest report card. Consonance, dissonance, augmentation-- take your pick.  It's all role-playing.

Role-playing is not a Dungeons & Dragons game, but a term for how the same person behaves differently depending on whom they're with.  Imagine the two roles of the office tyrant who, in the presence of his boss, becomes a spineless toady.  But relationships are not always as drastic and transforming as that.  Sometimes relationships just emphasize existing aspects of our personalities. I personally love the holidays because I see all my family members in different roles. My mother becomes a sister, playful and conspiratorial; my stepfather becomes a son-in-law, respectfully deferent and eager to please; my brother is no longer a titled professor of academic distinction but the little boy who proudly shows off all the things he did in school that day. (I'll let one of them tell you the roles I adopt.) My brother is also son, grandson, and nephew. My mother is also wife, daughter, and sister-- and so is my aunt, although in a very different way!

This family analogy may help me understand what's going on between the notes.  The notes in all the thirds want to sound identical to each other. A note that sounded so familiar all by itself is substantially altered by the sound of its relationship to another note.  But in real life, although each of my family members change when they relate to each other, those changes are not so drastic or so remarkable that I can't recognize them.  I see that this is my mother relating to her parents. This is my brother relating to his uncle.  Their behavior "sounds" different to me, because certain parts of their personalities are enhanced or suppressed through their relationship to the person they're with.  But I know those personalities.  The more thoroughly I know them, the better I understand their relationships, and the more likely I am to understand how they will change in each other's presence.

A violin teacher advised me that I may be thinking wrongly about relative versus absolute pitch. "It is like tasting and smelling. These are sensibilities that are very hard to separate until a neurologist starts placing electrodes on your skull."  I've said before that in order to understand "C", you must be able to hear it both as "C" and "not D".  Once I encountered the sound of the third, I thought there might be a similar approach with relative pitch.  I was trying-- unsuccessfully-- to hear both "C" and "not a major third".  But the problem is that C is a critical part of a major third! The sound of a D exists without C, but the sound of the C-E major third does not exist without the C. You can't have Janeandbob without Bob, and if you try to separate Bob from Janeandbob you force yourself to ignore critical components of his personality.

So how do you recognize Bob? This googly-eyed baby-talking mushhead seems very different from the foul-mouthed maniac you faced across the tennis net this afternoon.  An E may sound more cheery than a D, but it sounds dull and even grumpy next to an F (to my ear, anyway-- but that's a topic for another day).  There are even famous illusions in which a single color appears to be different depending on the color it's next to.  How can you make the distinction? How can you tell that Bob is still just Bob when he's part of Janeandbob?

Perhaps you can't. In the colored image, it's really quite impossible to see the same color-- the illusion relies on that fact.  The color is presented inextricably within its relationships, but if you pay attention to those relationships you will see the same color in two different ways.  In just the same way, when we listen to the thirds A-C and C-E, we hear the same C, but we can't help but hear it in two different ways-- first as the "top" note and then as the "bottom" note.

But I'm intrigued by the fact that the color illusion can be solved by zooming in on the picture. Once you are able to look at the color all by itself, and compare it to another instance of itself, you recognize that it is the same color.  This must be part of the solution.  It's probably a fool's game to attempt to ignore one note's relationship to other notes; we want to be able to hear that relationship in music, and we're too conditioned to not hear it.  It's also probably wrong to attempt to hear the pitch as separate from the sound of the relationship, because the sound of the relationship is largely composed of the sound of the note. What we can do-- one strategy that could be effective-- is to continue to familiarize ourselves with the feeling of an individual pitch, so that we can recognize each pitch inside of a relationship.

But that in itself is not enough, because a pitch will feel different depending on its relationship!  We need to be able to expand our understanding of pitch, as well. We may be surprised when our strong-willed cousin crumbles under the withering judgment of her father, but we understand that that's still part of who she is. We may be surprised when an E suddenly sounds sullen, but that's still part of the identity of an E. Relationships should enhance our understanding of a note, not obscure it.  We must allow ourselves to hear a note as more than its isolated fundamental-- we must eventually learn how it contributes to the sound of its chord.  But whatever consonant or dissonant personalities a note may take on, however much it may seem to change, by comparing the pitch to itself in isolation we shall understand that underneath it all... it's still a C.

August 14 - Slow down; construction zone

I think I'm making progress.  I am getting better at identifying C through F.  In fact, the only time I really get anything wrong is when I do it hastily-- when I hear a note, carelessly decide "Oh, that could be an F" and quickly check my guess.  I then see that I'm wrong, and when I play the two notes I cringe because they sound so different from each other that I feel I've made a very obvious mistake. Careless haste is definitely the worst detractor from success.  It's incredibly difficult to let myself go slowly, though.  I know that people who have perfect pitch know their notes instantly, and I have had the experience of knowing instantly, so why shouldn't I just be able to make a snap judgment? I want to test my ability, not just plod along!

Every song I hear gives me pitches that whiz by.  This is a great motivation; now that I know what it's like to hear a note and feel that it "sounds like a D", it's maddening to hear the pitches in a song and not be able to know them. I'll hear a bright sting, or a key change effected by a specific chord, or a burst of bumbly bass, and I can't tag it.  I feel like I've been given an electric bulb with nowhere to plug it in.  I hear the pitch, and I know that what I'm listening to is a pitch, not a relationship or an instrument's timbre, but I don't recognize the feeling. I just can't connect it. But the more I work on these exercises, the more convinced I become that it should be possible-- the more convinced I become that it's mainly a lack of familiarity that makes it impossible.

Developing that familiarity takes time, though, and I have to let myself understand that it's too early to be able to make snap judgments. The only way that it is possible to make a snap judgment is when you know-- really know-- what it is you are hearing. Anything else is just a partly-informed guess.  If I let go of my desire to see if I'm "there" yet; if I accept that my goal right now is familiarity, not speed; if I don't let myself "guess" but wait until I know before I speak the note name, it's possible to do an identification exercise flawlessly every time.  And as long as I remain patient, and don't let my eagerness to guess override the learning process, I can do it consistently, by using five stages of recognition.

- Instant awareness

Sometimes I'll play the note and just instantly know it for what it is.  Whether this happens by relative pitch, nascent absolute pitch, or short-term memory, I don't care; I know the note.  There is no question in my mind.  I play it a couple times for reinforcement and then I move on.

- Recognizing the pitch

If I'm not immediately sure, the next question I ask myself is "Does it sound like a [letter name]?"  I am NOT comparing the note to any conscious memory.  I just ask the question and let the answer come.

I don't create something in my head and match it to the note.  I let the note enter my ears and represent itself to my mind.  Most of the time it will then just feel right to call that sensation a D, or an F, or whichever.  Sometimes, too, I'll ask "Does it sound like an F" and my mind will tell me no, and that makes me understand that it must be something else; often that means I suddenly do recognize it, and other times I just ask myself again with another letter.

- Relative recognition

If I'm still not sure, then I imagine the notes directly above and below.  Often I will have instant awareness of the new note I imagine, and that identifies the note I'm listening to.

- Cognitive analysis

If I'm not confident about the adjacent notes, I use a list of trigger words which describe a unique characteristic of each note.  (I'll have to explain that later, because it took some method to make this list meaningful.)  I ask myself if I hear that specific trigger.  By the time I get here, I've essentially given up on feeling and I'm just attempting to name the note-- although I pay close attention to the characteristic.

- Memory

And then, finally, if that's not enough, I mentally create each note of the scale, one at a time, using relative pitch, and compare it to what I'm hearing. Of course, the trick there is to successfully recall a C from which to relate the others, but that can become fairly easy to do in the middle of an exercise once you're naming four or five notes. So I hear a C in my head, and play the test note to compare. Then I imagine a D, and compare, then E and F, stopping when the note I play matches the note in my head.

If I take the time to do this process, then it's very hard to make mistakes.  It is absolutely essential-- but oh so difficult-- to wait until I am certain enough that I am naming, not guessing

I've said that it is important to make mistakes because each time you refine your category of "what it is" by excluding "what it isn't".  It may seem odd that I make such an effort, then, not to make mistakes; but notice that by the time I reach "relative recognition", I'm comparing notes to each other again.  I may not name the note incorrectly, but I'm still making comparisons between notes.

But, ultimately, "correct" and "incorrect" don't matter.  I'm not trying to attach labels, I'm trying to develop my perception of the note.  Perceiving pitch is something beyond categorization, and beyond simple memory.  It is something beyond the note's name or its relationship to other notes. It's the indescribable feeling you get from audible frequency.  I've only just begun to comprehend it.  The simple expedient of attempting to label notes-- "F-sharp is 'brighter'"-- is not enough.  It seems very similar to the way one fellow with perfect pitch attempts to describe visual objects.  With descriptive adjectives, you can infer any object you please, and construct that object from what you intellectually knows about it.  But do you actually perceive the object?  Do you actually have a pure construct in your head that represents the object as itself, not as a collection of attributes?  This five-step process that I use starts with the construct and shifts to its attributes.

August 15 - Do you hear what I see?

Despite what you'd think, the definition of perfect pitch turns out to be something that's still hotly debated. When I wrote about the "official" definition not too long ago-- perfect pitch as "the ability to name notes without a reference"-- I was not yet aware that this is merely the most frequently used definition. Anywhere you look, whether it's in a Yahoo forum, a newsgroup, or in scientific study, there's almost always criticism of the actual definition. I attempt to define perfect pitch not because I want anyone else to be wrong, but because I want to learn perfect pitch.

Everything I discuss and conclude on this website, in e-mail, or in conversation online and offline, is dedicated to the goal of learning perfect pitch. I'm not interested in a universal definition of perfect pitch-- I'm interested in the definition which makes it easier to learn. I'm not doing scientific research; I'm not trying to be all things to all people.  If an idea makes it easier to learn perfect pitch, then I conclude that the idea is "correct" until proven wrong.  I'm trying as hard as I can to have my conclusions challenged by anyone with an opinion; please leave your own comments on what I write.  If I have the right ideas, then I'll be able to defend them; if not, you'll be able to tear it down and we'll build something better.

I'd like to use the term "pitch" instead of "pitch color".  It seems to me the term "pitch color" is useful principally to make people realize that the word "pitch" has been corrupted; linguistically the term is analogous to "color smell".  Each of our senses has a specific word to describe the experience of what it perceives: our eyes see color, our skin detects texture, our noses inhale scent, our tongues taste flavor-- and our ears hear pitch. Pitch is the unique nature of a single sound. If you had synesthesia, and saw sounds or tasted colors, "pitch color" would be a literally meaningful term.  But I think it's important to know that we are not trying to add anything to our perception of a note.  We are trying to reinterpret what our ears already hear.  Our ears already give us all the information we need.  A new phrase like "pitch color" can be confusing.  The only word we need-- the word I think we need to redefine-- is "pitch".

August 16 - Leading the blind

A lot of people with perfect pitch think that the talent can be learned by listening to, and memorizing, individual notes. I've gone to some length to explain why I think that won't work.  But until last night, when I had the opportunity to speak with someone who has AP, I didn't really make the connection about why everyone tends to think that this will work.  The reason is: people with AP tend to believe that everyone already knows what a pitch sounds like, and that they just can't remember it.  

People with AP don't know what it's like to not hear a pitch.  When I say that the greatest obstacle to learning is understanding what it means to hear a pitch, I find that people with AP typically don't agree.  They can't imagine how we don't hear it any more than we can imagine how they do. Consequently, people with AP tend to think that a person should be able to just listen to a note over and over again until they remember it. But the problem isn't remembering correctly, the problem is listening correctly.  Until a person actually understands what part of the sound is "pitch", until they actually know what it's like to hear pitch, instead of timbre or relationship or any other kind of complex waveform, they can never gain "perfect pitch".

Although the nature of current perfect pitch instruction is to progress note by note, the ultimate goal is to totally understand the actual sensation of "pitch", independent of any note. We need to grok pitch.  Pitch is the essential comprehension of our hearing.  It must become as meaningful to us as are the concepts of texture, scent, color, or flavor-- something we can understand purely as itself. I can conceive "rough" or "smooth" without thinking of any specific objects. I can imagine the odor of "something rotten" without specifically saying "eggs"; I can picture a "dazzling rainbow" as a formless nothingness of shifting color. I can even, just from hearing a friend's story about jogging and chewing gum at the same time, imagine the hideous taste of a plump fly, without consciously comparing it to any other taste that I do know.

But when you ask me to imagine a pitch, I'll remember an instrument playing a note.  I don't imagine it any other way.


August 18 - Physician, hear thyself

When I've used the term "relative pitch" before now, I've meant the ability to identify intervals.  I carelessly assumed it was the ability to recognize intervals, and didn't think much about it.  A bit ironic, isn't it-- I have been repeating over and over that the term "absolute pitch" is not the skill of naming pitches, and yet I've been describing "relative pitch" as the skill of naming intervals.  I've been analyzing on multiple levels what it means to hear in perfect pitch, but I haven't taken any time to turn that around and learn what it means to hear in relative pitch.

Since I've had the opportunity, in the last three days, to talk with two people who have perfect pitch, I've discovered that they don't listen to notes' relationships in the same way.  It isn't surprising to me that it's harder for them to learn relative pitch than it is for an average listener; because they are first inclined to perceive notes as "different", rather than "higher" or "lower", they first have to learn how to detect relationships before they can hear them well.  But even once they have learned their intervals, and the sounds of the intervals, they still don't hear the notes' relationships-- because to them, the notes don't change.  To someone with perfect pitch, a C is a C no matter what notes you play before or after it.  They can compare a couple of notes to tell you what the interval is, but to them a C-E major third has a different sound quality than every other major third.  Major thirds (or any interval you choose to name) do not all sound identical to each other.  They don't hear the relationship-- they hear an interval.

I found myself talking about this with a person who has perfect pitch, and the conversation was illuminating to both of us.  I said that "if we play a D and then an E, the E sounds brighter and cheerier than the D because it is 'higher'; but if we play an F followed by an E, the E sounds duller and more introspective than the F because it is 'lower'."  But to her they all sound the same every time.  She was astonished because this is not how she hears; I was intrigued because this is how I do.  It leaves me wondering-- what am I listening to that causes me to hear the same note differently?  My mind has trained itself to listen to the wrong thing, and I want to know what that thing is.  Relative pitch doesn't prevent us from hearing pitches.

So what am I listening to?  I'm sure it's a trick of the mind.  A week ago I was trying to identify C-F, and having a terrible time of it.  After I failed repeatedly to name even four in a row correctly (at each failure muttering curses at major and minor thirds for confusing me so) I suddenly realized that none of the notes I'd been given were F.  That's odd, I thought-- I know it's random, but surely by now I would've gotten at least one F!  And then I noticed that I was only working with C, D, and E; and, on my next attempt, I named all twenty in a row.

So I know it's a trick.  I know that my mind is deliberately listening to the wrong thing, just as surely as I know that it's capable of listening to the right thing.  And now I've got to figure out what that wrong thing is.

August 19 - One plus one is puce

What is the difference between sight and hearing?  If you read through this website you'll see many comparisons between the two senses.  Everything I've learned about pitch has a direct parallel with some aspect of color--  so far, everything I know about absolute pitch or relative pitch can be directly compared to "absolute color" and "relative color".  This has been extremely useful for understanding how a person with absolute pitch processes their sense of hearing-- but it is only an analogy.  Color is analogous to sound, but it is not identical.  For psychology and philosophy, for how we comprehend and interpret our senses, that's fine; but on a biological level the senses are definitely not the same.  You can't hear a light wave and you can't see a sound wave.  Even someone with synesthesia, who may "see" pitches, can't truly hear with their eyes (nor see with their ears).

Vision is an objective sense.  Color is determined by a mechanical, mathematical process.  We have three different types of cones in our eyes which are stimulated by either red, green, or blue.  When a lightwave enters our eyes, it applies a measurable quantity of energy to the cones, and that energy is translated into a color, directly.  X red + Y green + Z blue = color.  Although the number of possible colors is still theoretically infinite, because of the infinitesimal variations which could be applied to each component, there's no guesswork and nothing unspecific about it.  Put the right levels in, out comes the color you want.  The computer monitor you're looking at right now is using that principle to fool you into thinking you're seeing millions of colors, when in fact there's only red, green, and blue.

The biology of hearing seems more subjective.  I found a quick summary of the physical ear and was intrigued by its explanation.  A sound wave enters the ear and is transmitted to a fluid, and that fluid is dashed against the wall of the cochlea.  The coiled cochlea houses a membrane-- the basilar membrane-- which senses frequencies along its length, from highest to lowest.  The higher frequencies agitate the liquid so that it most aggressively hits the first part of the membrane, while the lower frequencies create a greater disturbance at the end of the coil.  What we "hear", then, is where the cells of the membrane are most strongly agitated-- but the wave is not absorbed by the membrane in the same way the eyes absorb light energy.  The frequencies in the ear have resonance  that are felt along the length of the cochlea.  There is a critical point of greatest resonance, a spot where the membrane is sympathetic to a specific frequency, but the resonance of the sound frequency is neither channeled nor limited to that precise spot on the membrane.

When a light wave enters our eyes, zap, it resolves into a color.  The rods and cones are charged in a very specific and measurable way, and barring any deficiency (like color-blindness) a color is very precisely calculated.  When sound enters our ear, it generates an extremely complicated pattern of liquid waveforms which resonate simultaneously and varyingly across the length of the basilar membrane.  We can hear the exact same sound repeatedly and, because of its complex interaction with our sense organ, with each repetition interpret the parts of the interaction very differently.

August 20 - There's a place for us

As the interval between two tones decreases, their respective disturbances on the basilar membrane (critical bands) overlap to an increasing extent.
(From Campbell & Greated, 1987: "The Musician's Guide to Acoustics". New York: Schirmer Books, p.58.)

Place theory is the name for how the cochlea recognizes specific frequencies at specific places along its basilar membrane.  If you do a Google search for "basilar membrane" you'll find that this theory is widely and generally accepted, and I certainly don't see any reason to contest it myself.  What it means is that sound as a physical sensation can be quantitatively identified in specific spots along the length of the cochlea.

But pitches don't just limit themselves to their designated spot.  I borrowed the above graphic from UCLA class material because it shows how individual pitches resonate across the inner ear.  An energetic interaction with any surface causes sympathetic vibrations within that surface (just hit a table and watch everything on it shake); sound frequencies interact with the basilar membrane and provoke "critical bands" of response in nearby receptor cells, presumably with the point of greatest excitation being the "hot spot" where the fundamental frequency is registered.  As pitches come closer to each other in frequency, their critical bands also draw nearer and begin to overlap. The nearer the pitches are to each other, mathematically, the greater the overlap.

[The graphic assumes that the cochlea, which is spirally coiled in real life, can be straightened out into a tube for demonstration purposes.  Although I found a research paper filled with unbelievably complex equations demonstrating how the "rectangular model" is mathematically an inaccurate representation, the paper's conclusion was that the coiled model makes the interactions even MORE complex, so that generates even more additional confirmation that the mind must apply cognitive interpretation to the chaotic result of sound wave perception. What it adds up to is that we hear what we think we hear-- literally.]

You must notice that the diagram shows the effect of an individual frequency within the ear. It does not illustrate the effect of overtones or complex waveforms, but of a single identifiable pitch.  Part of our process of learning perfect pitch has been to recognize the pitch within its collection of "overtones", or "timbre"; namely, the extra sounds that the instrument adds to the fundamental pitch. But this diagram begins to show us that once we can screen out all the extra noise, once we can recognize a sound frequency (say, an "F") in isolation, there's still room for error.  Although the "F" vibrates on a specific range of membrane, that same membrane is affected at the same time by every other frequency we hear.  There are no barriers between the vibrations of an "F" and "F-sharp".  They vibrate simultaneously on the same membrane.

Since pitches that are nearer in frequency have more resonant overlap, then the vibration of one pitch is literally shared on the membrane by another pitch.

This overlap suggests a reason why it is easy to be a semitone off when identifying pitches.  The nearer the frequencies, the larger the shared area of resonance.  When the frequencies of two pitches are adjacent, as shown here, the shared area actually becomes larger than the area of the individual pitches!  You can hear an A-sharp and believe that you are hearing an A-- probably because most of what you sense (the black area above) is literally and truly part of the sound of an A.

When you compare three pitches you can see a reason why it's often so difficult to hear the middle tone of a chord.  Look how much overlap there is!  Everything in the shared (black) area of resonance does in fact belong to either the top or the bottom note-- the effect of the middle note is minimal.  There's only a tiny part of the sound (the small white diamond) which is uniquely the middle pitch.  It's that tiny piece of information that is truly the "pitch". No other frequency can possibly produce that infinitesimal sensation.

That's probably why perfect pitch is such a refined comprehension-- according to place theory, a person with perfect pitch recognizes that infinitesimal sensation; they know when the "hot spot" on the membrane has been struck.  An ordinary person can play three different notes, and by careful mental comparison they can eventually locate the hot spot by eliminating everything that belongs to the other two notes.  They can gradually come to hear the "pitch" which is the black diamond between the two notes in figure 1, below.

But a person with perfect pitch can hear a single pitch and, like figure 2 shows, they can recognize the unique character of the pitch (the black diamond) without requiring anything other than their own cognitive understanding of what to listen for.

August 21 - As the nose on your face

Today I have been thinking about how multiple frequencies appear to the inner ear.  The individual frequencies on the basilar membrane were very easy to conceive of and diagram, yesterday, accurately simplified into triangles; but obviously sound enters our ears as a jumbling of frequencies, not as a pure tone.

It's a bit beyond my mathematical capability to guess at what two or more sine waves look like mixed together, so I hunted down a formula on how to create a sine wave.  I made an Excel spreadsheet with one sine wave in it, and played around with what it looked like at different frequencies and sample rates (they all looked pretty much the same).  Then I added a second set of data, and I found another formula to show me the result of the two waves' interaction.  That was more interesting to look at.  Then I realized that I know the actual frequencies of each pitch, so why don't I just put in a couple of those frequencies and see what it looks like?  So I created a minor second, A and A#:

Cool!  So it was immediately obvious why a minor second seems to "wobble" when you hear it.  And as I looked at this graph, I suddenly became curious to know what another minor second would look like-- so I changed the frequencies to D# and E.  And sure enough, this is what I saw:

Elated, I next decided to try a major third-- first A-C#,

and then C-E:

As I suspected-- exactly the same, just compressed.

I shared some of these images with a composer friend of mine; she asked if I wasn't overlooking the fact that different instruments produce visibly different waveforms.  I'm not; the timbres of the individual instruments don't matter in this model.  No matter what the instrument, we're still dealing with the same sine waves.  The timbre of any instrument is composed of multiple overtones, which are themselves sine waves; the "sawtooth" or whatever waveform that results is merely an additive representation of the overtone series.  I could probably model an instrument more precisely, varying both frequency and amplitude of the different overtones in a more complex spreadsheet-- but since the overtone series is the same from note to note, it all cancels out and we're left with the interaction of the fundamental tones.  Two sine waves.

I played around with all the other intervals-- it was exciting to see how the most consonant intervals resulted in the most regular waveforms.

So now I know for sure what I'm listening to that makes all the intervals sound the same.  What I am "hearing"-- what my mind is interpreting as the sound of the chord-- is not the individual frequencies as they stimulate specific places in my basilar membrane.  I am sensing the patterns of how those sensations pulse and shift among those specific places.  Now I know, too, why people with perfect pitch have difficulty with relative pitch, and vice versa.  Relative pitch is apprehension of a changing pattern; absolute pitch is a sensation of specific stimulation.  Absolute pitch makes note of where the membrane is stimulated; relative pitch attends to how the membrane is stimulated.

I'll have to think of a really good analogy for this later; it's not coming to me right now.  But this may be clear enough as it stands.

August 22 - What has gone before

A few days ago, I wrote about a non-musician who has absolute pitch who told me she'd heard that notes "influenced" each other, but she wasn't sure what that meant.  I said that if you play a C and then a D, the D will sound cheerier and brighter because it is "higher"; and if you play an E followed by a D the D will sound duller, perhaps more sullen, because it is "lower".  She wrote back in great astonishment-- "is that what they meant?  I'm blown away!  I don't hear notes like that at all."

Of course, I was amazed at my own casual assumption that things always sound "higher" or "lower" as a matter of course.  I began testing how this worked.  You can do it yourself-- just play or sing C-D, then E-D, and the D will sound recognizably different each time.  If you listen without unusual effort you will hear the difference.  You can't help it.  I spent about ten minutes just singing and playing C-D-E-D-C-D-E-D-C, listening to the phenomenon.  I simply couldn't not hear it.

I speculated it was because the combination of frequencies distorted each other when heard simultaneously.  Or perhaps, since the musical scale is logarithmic in frequency, adjacent notes have exponentially different overlaps on the cochlea.  But these were easy to disprove by playing the notes separately.  I let the top and bottom notes fade out completely before I played the middle note, and the middle note still sounded different.

It wasn't interference, and it wasn't overlap-- perhaps the receptor cells were spaced differently along the cochlea?  Maybe higher notes are literally a shorter distance apart.  When I looked for specific proof of that, though, I found instead a very helpful diagram which explains that all the notes are the same distance apart.

In terms of length... the critical bandwidth is nearly constant at 1.2 mm, within which are located about 1300 receptor cells, and is generally independent of intensity... Twenty-four critical bands of about one-third octave each comprise the audible spectrum.

They're all the same distance.  This was phenomenal!  The physical area for each frequency is the same size as every other.  The critical bandwidths are identical regardless of the frequency.  This means the cochlea is capable of recognizing that two frequencies are "five notes away" from each other because they literally are.  You can measure it.  And this meant that every note overlaps every other note in exactly the same way.  C and D have the same level of overlap as E and D.  So why would D sound different depending on what you played before it?  According to Music, the Brain, and Ecstasy,'s clear that auditory cortex does not act as a sort of tape recorder, pigeonholing each sound as it arrives.  If this were the case, reversing a sound sequence would reverse the pattern of neurological activity it produces.  But it doesn't:  research shows that the reversed sounds generate a unique response.  This implies that auditory cortex does not consider individual sounds in isolation.  Instead, it always interprets sounds in the context of what has preceded.  (p 57)

This lets me explain why a note will sound higher or lower depending on which comes first.  Our senses perceive change-- this is a well-documented fact.  It's called "habituation".  In the winter, I always used to enjoy the sensation of not moving my hand until I could no longer feel my mitten; at this moment, I only notice the whir of my computer fan when I actively listen for it.  I'm sure you've got examples of your own.

When someone without absolute pitch hears a note, they comprehend the entire sensation of the frequency-- frequency and critical bandwidth-- as "the note".

When they hear a second note, what they listen to is what changes.  The person has clear short-term memory of the first note, and since they know that what they have already heard is "C", then everything new, everything that's changed, must be "D".

We hear the "D" as higher because we are only paying attention to a half of the note!

Likewise, when we come down from E to D, we group together the sense of what we've already heard as "E", forcing us to interpret only the lower half of the new note as "D".  So, naturally, it sounds lower.

In order to counter this effect, I tried playing all three notes in sequence-- C D E D C, repeatedly, for about fifteen minutes-- to mentally pry away the C-ness and E-ness to hear just the D.

Didn't work.

I tried playing all three notes in sequence while singing the middle note, to prove to myself that it was the same tone.  And I was amazed to discover that my mind wouldn't cooperate.  I even tried singing along with the piano at different intensities, to make the D pop out dynamically-- but each time I sang C-D, it sounded like I was singing the D too flat, because I heard the C-ness of the D.  When I sang E-D, it sounded like I was singing it too sharply, because the D also sounded like the E.

Then I tried playing the C and the E and singing the D in between them.  That put me on to something.  After maybe ten minutes I found that the D had completely disassociated itself from the C and the E, just like what happens when you repeat a word so frequently that it loses its meaning.  After relishing this sensation for a while, I decided to try identifying C through F, and I blazed through swiftly without a single error.

I'm still not sure how to evaluate that effect, but it seems encouraging.

August 23 - Oranges and peaches

If you've heard anything at all about absolute pitch, you've heard that it's a rare ability, an unusual skill, a strange genetic gift that some people have been mysteriously blessed with.  But did you ever think about why it's so uncommon?  Did you ever wonder why there are so few people that actually have it?

Natural selection, perhaps?

Our thinking about absolute pitch may be backwards.  Most of us think it's some kind of incredible skill to be able to hear pitches absolutely.  Most of us think that relative pitch (just hearing music) is normal.  But absolute pitch is a reflexive response to a single stimulus.  Relative pitch is an analytical synthesis of complex patterns.  Which sounds to you like the more advanced task?

Animals perceive the world tonally.  Music, the Brain, and Ecstasy tells us (emphasis mine):

Much recent research has been devoted to puzzling out the precise roles of columns in various kinds of cortex.  This involves the painstaking labor of inserting minute electrodes into individual neurons in a living brain, then casting.. a tone on the cochlea, to see how a column's neurons respond.  Such work must of course be confined to non-human subjects (mostly cats are used).  Since animals are not musical, this important research technique can be employed only to study basic perceptual mechanisms.  Scientists can locate a column attuned to middle C easily enough, but finding a group of columns maximally responsive to "Singin' in the Rain" is out of the question.

Animals "have perfect pitch".  When they hear a middle C, they hear a middle C, and when they hear music they hear.. a bunch of pitches.

People who have perfect pitch don't always have relative pitch perception.  I have spoken (on-line) to a woman with perfect pitch who says that notes sound the same no matter how you arrange them, no matter in what order you play them.  I have met a woman with perfect pitch who honestly wishes she didn't have it because she can't easily perform musical tasks (like transposition).  I've also read a discussion post from a fellow with perfect pitch who says that, to him, comparing a C-E-G chord to the notes C, E, and G is like calling four quarters "one dollar".  To all of these people, an interval or a chord is just a different name for the same thing-- multiple pitches played simultaneously.  They don't necessarily hear the patterns or the relationships.  They hear the pitches.

Perfect pitch doesn't necessarily have anything to do with music.  It will help a singer start on the right note, and it will help you tune your instrument; and, as one musician has said quite profoundly, it will help you refine the quality of the pitches you present in your musical performances.  But it is not inherently a musical ability.  It is a primal response to sound.  Perhaps, then, it's a type of perception that we have evolved away from...?

Continue reading with Phase 2