Get lessons or coaching

I teach in Oregon City and online videochat. I work with all ages and levels and a variety of styles. I specialize in creative exploration, the psychology of music, and conscious music practices. Visit the lessons page to learn more.

Friday, February 10, 2012

Absolute vs relative pitch — my take

I was practicing guitar — in this case, sight-reading through a book of intermediate/advanced exercises in all different keys — when I found myself, as usual, distracted by reflective ideas about how I was processing the music. In this case, I found myself modulating back-and-forth between processing the notes in an absolute manner versus a relative manner...

Sitting down to write this, I worry about how to take on one of the most controversial issues in all of music in a short article. Entire books and websites and learning programs have been made on this topic. I do not have time or space to review them all or cite all the vast amounts of research, but the Wikipedia articles are a good starting point: Absolute pitch / Relative pitch.

There is more than enough evidence to conclude that all normal people are sensitive to both absolute and relative pitch. Some rare people are highly sensitive to absolute pitch, including the ability to explicitly name any pitch regardless of context — a skill often called "perfect pitch." All my personal acquaintances who have this  ability are children of piano teachers or at least started music training at a very young age. It seems that full development of "perfect pitch" requires explicit connection between consistent sounds and consistent names during a critical period in early childhood. Nevertheless, everyone seems somewhat sensitive to absolute pitch (see The Levitin Effect).

In his wonderful book Sweet Anticipation, David Huron argues (among a great number of other insights) that absolute pitch is unremarkable (it is easy to understand a neuron firing for particular pitches, and absolute pitch is how most animals recognize sounds). In contrast, the uniquely human ability to recognize relative pitch patterns is more cognitively significant. He supposes that when most children pick up on the idea that songs can follow patterns independent of their exact frequency, sensitivity to absolute pitch is actively suppressed in favor of relative focus.

David Huron furthermore asserts that relative pitch may be more about tonal categories than about absolute intervals. This is the ultimate view of relative pitch. Neither the precise frequencies of pitches nor the objectively measurable distances between pitches are of important consequence in music. Instead, the important factor is our cognitive representation of a pitch's context, such as the tonic or dominant, part of a major or minor scale, the root or third of a chord, or simply higher or lower (by a lot or a little). Huron undertook numerous empirical studies that support this view, such as the response time of musicians instructed to place a single pitch in a subjective internal tonal context.

Thus the problem with the oft-taught emphasis on learning even relative intervals out-of-context: Major thirds and minor thirds, for example, may be objectively different, but they both have a somewhat major feeling when conceived in a major-chord context (or minor in a minor-chord context). Similarly, using songs as mnemonics for intervals is problematic. True, My Bonnie Lies Over the Ocean starts with a major sixth, but it only works well in the same context of the fifth to the third of a major key. This mnemonic might be partly functional in other contexts, but it sure sounds strange to me to try to think of My Bonnie when going from the third to the root of a minor chord! I don't know with certainty that everyone hears this way, but David Huron and I do, and so do all the students I've taught. I am basically in agreement with Huron's views, and I recommend his book to anyone interested in learning more.

So, how does this relative pitch emphasis relate to music teaching, learning, and performing in everyday context?

There are a wide range of opinions on whether music instruction should emphasize absolute versus relative pitch. In many ways, the symbolic systems we use are biased in one way or the other.

The alphabet naming system is mixed. Though letters can be fixed to A440, 12EDO temperament, and concert C; they can also be transposed for different instruments, guitar capo positions, and non-standard tunings. Still, the letter system is rarely or never used in a truly relative way where a certain letter consistently represents the tonal center. Once a tuning is chosen and a transposition is set, the letter names do not move to follow relative changes in the music.

Solfege (Do-Re-Mi etc) is often taught in a truly relative manner (at least in the U.S. and Britain). In other schools of thought or countries (e.g. Spain) it is fixed, thus being simply a substitute for the alphabet system.

The Western staff notation system is unspecified. The staff is fully capable of being a completely relative representation, but it is most often taught to instrumentalists in an absolute manner. Singers, on the other hand, are often taught a relative interpretation of the staff. And dealing with this dichotomy in my own subjective experience is what inspired me to write this article...

When students first learn to read staff notation, they are usually taught in an absolute-pitch manner. This usually means an instruction of mechanical action to do with their instrument whenever they see a particular symbol. As they advance, they are introduced to more and more symbols and then to the complex issue of key signatures. In different keys, visual representation is thus necessarily somewhat relative. Two sets of notes may look the same aside from key signature, and the difference is primarily in the relationships and where the tonic is located. Thus, students can either learn all the different keys as a set of absolute patterns (e.g. knowing which sharps are in the key of E), or they could learn to think more relatively.

In a relative-pitch approach, a key signature only specifies something about the relationships in a scale and where the tonal center is marked. This works great for singers who just need to know where the tonic is and then they can sing the same pattern for any scale as long as they treat it relatively. Of course, because of the historic Western bias of the notation, the "key" is typically either major or minor and there are strict requirements for what can be in the key signature at the beginning. But, in principle, it would be possible to have a staff represent a scale of any number of notes and of any variety, as long as the singer learns what the scale sounds like.

I believe that the absolute-pitch tying of a note to an exact mechanical movement and exact sound is analogous to the "perfect pitch" naming skill. This absolute pitch focus is useful when needing a particular sound in an unexpected context, but it is cumbersome and hampering when playing otherwise predictable patterns in unusual keys.

Different instruments are more or less absolute in orientation. Wind and brass instruments are inherently tuned to certain primary keys and use specific mechanical motions for different notes. These instruments do not easily transpose, but, with extensive training, wind players can learn to mentally surpass the tyranny of absolute pitch. In contrast, the Chapman Stick is a string instrument that uses no open strings, no sympathetic drone strings, and is tuned in parallel intervals (the same proportional pitch change from string to string). The Stick is therefore an optimal relative-pitch instrument. Stick players have no need to learn multiple key signatures. The exact same scale shape will work identically in any key. Once a starting point is determined, a Stick player can read staff notation by simply following the relationships. The guitar is a compromise: an instrument with some consistent relationships and patterns but some other idiosyncrasies that favor keys around the open strings.

A more relative or more absolute perspective on pitch will lead to differences in how one composes and improvises, and this will lead to music that favors one approach over the other. And, as much as I wish that songbooks and music teaching method books were printed with numerals instead of letter names for chords; as much as I would like to see transcriptions of international music styles made by simply mapping different scales onto a basic staff; and as much as I dislike the awkward irregularities of the Western pitch naming and notation systems... my wishes not reality in the world of music today...

So as I was playing through this book of exercises in all keys, I found myself wondering whether I ought to be thinking relatively or absolutely. Is the goal to train myself to relate an exact spot on the guitar to that particular note given in the key of D-flat? Or would I be better off to just recognize the key, find my position, and then think about where the note falls in relation to the key?

Ironically, the key-signature system highlights the value of a relative-pitch mindset (because we can see the common patterns across keys instead of just thinking absolutely about a single key), yet it is also unnecessarily cumbersome for relative pitch. For a purely relative system, we might as well keep the tonic always on the same line and just mark what scale goes around it and where to tune the tonic.

In my experience, the absolute vs relative mentalities create fundamentally different sight-reading experiences. Even playing a bit from memory, if I look at my guitar and think about the absolute pitch, it is a different perspective than seeing the relative patterns. And though I can do both and have some awareness of both at all times, switching is often confusing. It is most comfortable to pick one mindset or the other.

My subjective conclusion: relative pitch feels far more musical.

I can easily see the practical values of absolute pitch. By always connecting to precise sounds, even the most unexpected music can be transcribed by someone with "perfect pitch." Likewise, instrumentalists who know what to do when they see a symbol are capable of playing modern atonal compositions. Communicating with musicians in a modern Western context involves knowing letter names and reading precise musical notes. I've even had some awareness of my own sense of absolute pitch: a certain ineffable surprise upon hearing a familiar song in an unfamiliar key, for example, or an intuitive sense for guitar tuning.

But absolute pitch focus feels like recognizing particular paint colors instead of seeing a beautiful compelling image. Identifying the brush stroke angle and exact colors may make it easier to copy a painting, but this could be a purely technical exercise. I would rather notice my profound emotional response to art and then try to make my own art to replicate those expressive qualities, even if this focus meant missing some of the technical details.

These perspectives may be two sides of the same coin, and they certainly coexist for most of us, but they do not feel the same. To learn to sing barbershop harmony, my friend with "perfect pitch" told me that he had to learn to focus on the relative pitch and blend. Furthermore, he claimed that he stopped tracking the absolute pitch, though he can quickly stop and check the absolute pitch in his mind if someone asks. In suppressing his "perfect pitch," he learned to be open-minded and appreciative of music even when it deviates from his absolute expectations. And he became a better musician in the process, though he still has trouble sight-reading if the key is transposed from what is on the paper. At the risk of presumption, I think my experience in thinking absolutely vs relatively in sight-reading guitar music is the same underlying phenomenon as my friend's "perfect pitch," only perhaps a weaker version. And while I respect his skill, I'm not envious. I'm going to continue trying to diminish my own absolute-pitch sense in favor of thinking more relatively.

Having reflected on these issues, I am even more confident in emphasizing relative pitch in my teaching, even from the very beginning. I have many students use capos to change the pitch of songs, for vocal reasons, for small hands on large guitars, or just for novelty. I had been concerned that this might impede them in some way, but now I think that suppressing absolute pitch in favor of relative is actually positive. I really believe a focus on patterns and relationships is a more musical, more meaningful approach. I may be positioning myself in opposition to the loud voices who proclaim the value of learning absolute pitch, but I believe they work to be loud because they are fighting against the tide. The meaning and experience of music has nothing to do with whether a pitch is A or B or 100Hz or 120Hz. The natural state of music is more relative than absolute. Just as our eyes easily accommodate different lighting, our ears accommodate different contexts. Musicians and teachers who over-emphasize absolute pitch (such as symbol=letter=mechanism) can't see the forest for the trees.


  1. Really interesting stuff, Aaron. I'd never thought about it, but it mirrors my experience as well. Over the course of nearly 20 years of formal piano studies, I always sight read passably well (though never very fast) using absolute pitch, or what you call a 'technical' interpretation of the notes on the page. Since then, I have spent much more time singing (both barbershop and chamber choir), and now I tend to sightread relatively - give me the tonic and off I go. Working this way, I can take an unseen and unknown part (usually baritone, which always makes things worse), and read it almost perfectly the first time through. On the rare occasions that I do sit down to play something more classical at the piano, it takes me a certain amount of time to 'switch gears' and return to a manner of sight reading that works for me there. I really need the two different modes of thinking to perform what is basically the same skill at the piano and vocally.

  2. Thanks for the comment, Martin! I guess my thought is to go even further: try playing the piano while maintaining your relative mindset from barbershop. It will be hard at first, and there are times you'll have to go back and rely on absolute, but I think you'll find it more musically satisfying and insightful if you can do it...

  3. Aaron, I understand that "My Bonnie" begins with a major 6th, but when you say " but it only works well in the same context of the fifth to the third of a major key. This mnemonic might be partly functional in other contexts, but it sure sounds strange to me to try to think of My Bonnie when going from the third to the root of a minor chord!", I'm lost. How is it that it only "works well" in the context of a fifth to the third of the major key? I know what a fifth to a third of the key is, I just don't know what it means by saying it "only works well in this context". It seems to me it's a major six no matter where you start. Secondly what do you mean by going from third to the root of the minor chord? The "My" is a fourth under the root. "Bonnie" is a third above the root. I think I've done a fairly good job of confusing myself about this. Can you figure out what it is I'm not figuring out?

  4. Thanks for the comment, Lance. I'll try to be specific. Now should I be relative (like #s) or absolute (letters) for my examples? I'll go with relative.

    Say you have a song in a major key. There's a M6 interval from the 5 of the key up to the next 3 of the key, like in "My Bonnie." But if the song has a spot where you have to go from the 1 to the 6 of the key, it is a bit weirder to use that mnemonic. The issue isn't that the interval is anything other than a M6. The problem is, by thinking of "My Bonnie" you don't just think of an abstract M6, you think of the song and the strong implication that "Lies" as in "My Bonnie LIES" is the tonic (and that "My" is the 5 and "Bon-" is the 3)! So telling someone to think of "My Bonnie" may make them change where they feel the tonic belongs. For some, it may be hard to even connect to "My Bonnie" for a 1 to 6 of the scale because there's cognitive dissonance between (a) thinking of "My Bonnie" which *definitely* starts on the 5 for that song and (b) hearing the same note as the tonic in the context being sung. It's as though you are being asked to imagine singing "My Bonnie" if it started on the tonic (and what the heck could that possibly mean?!?)

    It's not so hard to think "My Bonnie" for notes 1 to 6 of the scale when coming to the subdominant. In that case, it is as though the key moved to the subdominant momentarily. Not *too* weird.

    But in my example of a minor scale, it is a M6 from the scale's minor 3rd up to the tonic above. Try playing a minor scale and holding a minor chord, or having your quartet sing a minor chord; then move from the m3 of the chord up to the root. Does that sound like "My Bonnie?" It doesn't to me!

    The idea is: we don't really hear intervals abstractly. In fact, in terms of pure distance, we mostly notice just small steps versus larger leaps, and it is hard to tell one leap apart from another purely by distance. Instead, we actually identify specific intervals by relating them to a tonic, by conceiving of them as having tonal categories. There have been scientific tests showing significant tendency to miss changes to songs if the changed notes still fit the same harmonic context and intervals are still similarly sized and same direction.

    I don't think we should bother trying to teach singers to recognize intervals out of context. We should teach them to recognize intervals IN CONTEXT. And we can't expect that to transfer to totally different contexts. Singers should need to learn what 1 to 6 sounds like in a key separately from 5 up to 3 because they are musically different — they are cognitively different... well, partly because it's still true that they are both a M6, but they are not cognitively interchangeable, at least not easily.

  5. Good thoughts, although I'm biased because I tend toward the same ideas. I'm accurate to about a minor third in picking pitches out of the air, but my relative pitch is pretty good - and that's what I use when I play and when I listen.

    By the way, 100 Hz is closest to a G.


  6. Interesting discussion!

    In a similar vein, I wonder if anyone else feels the same as I do with an octave.
    Of course a unison is the most consonant interval, and quite often it is said that an octave follows not far behind in terms of consonance.

    But if I sing an octave above a definitely established tonic, it doesn't FEEL like a completely consonant, in fact I feel a lot of musical tension there despite the top note being of the same musical pitch class.

    Does anyone else feel that?

  7. Hi Mike,

    Thanks for the thoughts. You should check out my other post about octaves and 12ths. I don't think octaves are less consonant per se, but singing is never absolutely precise. I suspect your feeling is driven more by the sense the lower frequencies are just more relaxed and thus more effective settled resolutions. I.e. the higher octave adds more energy and so feels more active. But the term "consonance" generally refers more to whether it smoothly blends and less about whether it's maximally relaxing or resolved, although that might be just semantics.