Perfect pitch may not be perfect after all

Child On White BackgroundIf you can name a pitch, either real or imagined, without any kind of comparison to a reference pitch, you have perfect pitch, or in the terminology of scientists – absolute pitch.  Non-musicians tend to think that perfect pitch is a kind of ultimate credential for musicians; if you have perfect pitch, you must be an exceptional musician.  In fact, only a small number of musicians have absolute pitch and most musicians know that having it can be a mixed bag.

If you’re a singer learning a piece with unusual pitch intervals, absolute pitch can be very helpful.  However, if you are that same singer and working with musicians who are singing or playing out of tune, having absolute pitch can be excruciating.  And for any musician with perfect pitch who has to transpose to another key, perfect pitch is often a hindrance.  But up until now, musicians and non-musicians alike have shared the belief that perfect pitch is some sort of absolute.

Now, a new study from the University of Chicago shows that may not be case.  You may have read about this study when it came out earlier this summer because it sparked a lot of interest, and news reports have been posted on many websites and blogs.  Steven Hedger, psychology grad student, and Shannon Heald, a post-doc, were talking about absolute pitch in the lab one day while Hedger began playing a piece of music on a MIDI keyboard.    Unbeknowst to Hedger, who has perfect pitch, Heald used a pitch wheel to very slowly flatten the pitch about a third of a semitone (a semitone is the equivalent of two adjacent keys on the piano).  Hedger never noticed.

That made them think that maybe perfect pitch is changeable.  So working with Howard Nusbaum, University of Chicago professor who researches brain plasticity (or how the brain changes in response to learning or experience), they designed an experiment to see if the same thing might happen with other people who had perfect pitch.

They divided 27 people who had been identified as having absolute pitch into two groups.  Both groups did a pre-listening session in which they listened to a series of individual notes played on a violin.  They were asked to identify the name of the note and indicate whether it was sharp, in tune, or flat.  Then group one listened to the first movement of the Brahms First Symphony.  They were asked to notate melodic themes to ensure that they were listening closely.  Over the course of the movement, the pitch was slowly detuned at the rate of about 2 cents/minute.  There are 100 cents in a half step, and by the time they had listened to the 15-minute movement, the pitch had been lowered by 33 cents, or a third of a semitone.  They listened to the remaining three movements at the detuned pitch of 33 cents low.  No one noticed the change in pitch.

In a post-listening session, group one again listened to individual violin pitches as they had in the pre-listening and identified them as sharp, in tune or flat.  An additional section of notes with a piano timbre was added at the end.  The researchers compared the pitch identifications in the pre-and post-listening sessions and discovered that, after listening to the detuned symphony, participants rated flat notes in the post-listening session as in-tune and notes at pitch as out-of-tune.  During the 30 minutes of hearing the Brahms 2nd, 3rd, and 4th movements at the lowered pitch, the participant’s brains had become adjusted to a flattened version of “absolute pitch.”

After doing the same pre-listening as group one, a second group listened to four pieces of what the researchers call phase music. (I heard an excerpt from their study and it’s not phase music as in Steve Reich, but that really doesn’t make any difference for the purposes of the study.)  The phase music used only five pitches (G, A, B♭, C, D).  The researchers wanted to determine if pitches that weren’t heard would also be affected.  The listeners were asked to notate the main melodic figure in the phase piece.   During the first piece of phase music, about 15 minutes, the pitch was flattened by 33 cents.  They then listened to the remaining three phase pieces at the detuned pitch of 33 cents low for another 30 minutes.  Again, no one noticed that the pitch had been flattened.

This group also did the post-listening, although in this case, there was a section of French horn notes at the end instead of piano notes, as there had been for the first group.  As with the first group, notes that they judged as flat in the pre-listening, they now heard as being at pitch, and pitches they heard as in tune in the pre-listening were now heard as out-of-tune.  Whether they were hearing the five pitches used in the phase music or pitches not used, the results were the same. Nor did it make any difference whether the subjects were hearing the individual notes in the timbre of a violin, piano, or French horn, or hearing the Brahms or electronic phase music listening examples, timbre made no difference.

The researchers cite this study as evidence of brain plasticity – the ability of the brain to change, and note that a person with absolute pitch does not have some sort of template of frequency values in her brain that is unchanging.  The ability known as absolute pitch depends on the standard of tuning and depends on frequent reinforcement of that pitch standard.  So Mozart’s absolute pitch would be different from ours.  Tuning forks used during Mozart’s time were at A = 421.3 or 421.6, rather than A 440 as is used today, so Mozart’s perfect pitch was based on A 421.

Did the test subjects’ pitch sense remain changed after the experiment?  No.  Listening to a detuned piece for thirty minutes doesn’t compare to years of brain reinforcement of the correct pitch. But it did temporarily change the brain’s pitch perception.  This is just one of hundreds of studies about absolute or perfect pitch that have been done over the past 20 or 30 years.  There is a lot of interest in what absolute pitch is and how one develops it.  So stay tuned (pun intended) for next week’s post and we’ll talk more about the puzzle of perfect pitch.

If you want to track down the actual study, it was published in the June issue of Psychological Science.

6 responses to “Perfect pitch may not be perfect after all”

  1. Jill Kelman Avatar
    Jill Kelman

    I have often wondered if absolute pitch was based on the tempered scale or the original pitches ie: c# lower then db.

    1. LS Avatar

      Interesting question. AP appears to be based on whatever tuning you are studying at a young age (other factors also which I’ll talk about next time). A child who grows up using a piano that is tuned a half step low may develop AP based on that lowered half step. So it would seem that C sharp or D flat would be detectable by a string player with AP. Anyone have any experience with that?

  2. Michael Taylor Avatar
    Michael Taylor

    This is really interesting. Thinking from a broader perspective, is a potential takeaway here that brain plasticity can have an unseen negative consequence in performance as any skills/perspectives you have learned become neutered in any particular moment/action/project/etc where if you don’t fully recognize slow change is occurring in your environment, your own actions/perspectives will be wrong/distorted.

    If that’s the case, I wonder what the tool would be to help avoid this?

    1. LS Avatar

      Interesting question, Michael. Reply will follow soon.

  3. Avatar

    Fascinating. I do not have what is commonly called PP, but when I was doing a lot of solo work, I never had trouble with modulations, even when singing without accompaniment. My brother does have PP, at least of a type. I recall, when he was practicing for his Grade 8 Ontario Standard music exams, his teacher sending him to the next room, striking a chord, and asking him what it was. After a moment’s thought, he would tell her, even to the point of saying: diminished Seventh of C minor, for example. The Chinese are said to have a higher rate of PP than other ethnicities, due to their pitch-based language. It would be interesting to know how many Chinese as first language were in this study, or to repeat it in that ethnicity.

    1. LS Avatar

      Hi Loren,
      Replies to both yours and Michaels comments will come in the next post.