I have just stumbled across a review in Trends in Cognitive Sciences, vol 9, pp. 26-31 (2005) entitled “Absolute pitch: perception, coding and controversies” by Daniel Levitin and Susan Rogers of McGill University, Montreal. I found it very interesting and so I have resurrected this thread, quoting below many of the authors’ provocative assertions. I have no special expertise in this area and won’t attempt to defend or further interpret any of the statements. My occasional comments appear in italics.
Absolute pitch (hereafter referred to as AP vs. relative pitch, RP) is defined as the ability either to identify the pitch of a tone presented in isolation or to produce a specified pitch without external reference.
AP occurs in 1 in 10,000 people.
Sometimes regarded as a mark of musicianship, AP is in fact largely irrelevant to most musical tasks. Being unable to turn it off, many possessors of AP perform dramatically poorer at judging whether a melody and its transposed counterpart are the same, a task that many non-AP musicians accomplish with ease. Some find songs transposed to unfamiliar keys disturbing, as if one day you went to the market and found bananas to be purple and tomatoes blue.
Growing evidence suggests that many people in addition to AP’ers may have stable, long term auditory memories for pitch. This is 1 of the components of AP, called “pitch memory”, while AP possessors have the additional component of “pitch labeling”. Some of the evidence: When people without AP are asked to identify a pitch or key, the modal (single most common) response is the correct one (see figure). [I personally am not quite convinced by this. The range of response varies over 4 octaves. Does the tendency of the population to center over the true pitch mean that people possess rudimentary AP (some memory, without labeling), or that the population tested is as likely to guess high as low, and so centers around the true pitch. I did not check the original article to see if they had considered this.
] If asked to sing popular songs that currently exist in only 1 key and thus have an objectively correct pitch, most people sing at or near the correct pitch. Quite often, this is tied to the lyrics. For example, people tend to pitch the word “hotel” correctly at G when asked to sing “Hotel California”. [no accompaniment or external references are present, remember.
AP is automatic [not further defined in this article
], categorization occurs without deliberation, is accompanied by large differences in speed of performance compared to non-AP attempts, and can be done easily while engaging in other tasks.
AP’ers as a group do not have exceptional pitch acuity. Also, AP is neither “absolute” nor “perfect” in the ordinary uses of those words. AP’ers often make octave errors and semitone errors (confusing tones 6% apart). It appears AP is not an all-or-none ability, but exists along a continuum. Musicians not claiming AP score up to 40% on pitch labeling tests, where 1/12, or 8.3% is chance. Even true AP’ers, who score better than 90% in naming pitches, are no better than other musicians at noticing when one tone is out of tune with respect to another. Hence, there is nothing “perfect” about AP – it is just the ability to place tones within nominal categories with a high rate of success.
Some AP’ers can only label tones produced by one particular instrument (usually piano, of course, and sometimes called ‘absolute piano’). This suggests that their internal template is bound up with the timbre of that particular instrument, so presumably a life-long concertina player might have absolute pitch specific to concertina [insert joke here
Some people have AP for only a single tone – often their tuning note. They can get good scores on AP tests by using RP for all tones but their one internal referent, but are betrayed by the increased reaction time for the other notes. Some people develop AP on mistuned instruments and will make constant errors thereafter. [This is truly sad.
Many animals, wolves for example, use AP to identify members of their own pack. Starlings and rhesus monkeys try first to solve pitch tasks with AP, but can move to RP if necessary. Monkeys, but not songbirds, are able to use “octave equivalence” in performing tasks.
AP is acquired before the age of 9; no known cases exists of an adult acquiring it. Because piano instruction starts with the white keys, many AP’ers are best at these notes. Speakers of tonal languages (e.g., Mandarin Chinese) are more likely to have AP.
Controversy exists as to whether AP acquisition requires explicit training or can result from incidental exposure to music. Most AP’ers report having acquired the ability without remembering how or when it occurred.
Edited by Stephen Mills, 31 August 2005 - 01:47 PM.