How does musical background affect absolute pitch ability?

Article of the day: Vanzella, P & Schellenberg, E.G (2010) Absolute Pitch: Effetcs of timbre of note naming ability. PLoS One, 5 (11), e15449. (FREE TO ACCESS!)

Today I was alerted to a new article published in PloS One on the subject of Absolute Pitch (AP). I have always been interested in this musical skill – mostly because it is something I don’t seem to possess despite having started my music lessons at a very young age. Not that this torments me on a daily basis you understand! – But I would like to know more about why some people have it and some don’t, and how possession of this ability affects other cognitive processes.

So, first question, what it AP? AP refers to the ability to identify or produce a musical tone on demand without the need for an external reference (i.e. the ability to sing a ‘C’ or name a tone as a ‘C’ without hearing any other tones as a guide). Research tells us that the incidence of this ability in the general population is very rare at probably less that 0.01%, but for individuals who started musical training before the age of 4 the incidence is far higher at up to 40%. This fact tells us that musical training is a very important component in AP, as might be speaking a tonal language as AP incidence is higher in Asia compared to Europe. But not everyone who started music lessons early has AP (present author included!) so it is possible there is also a genetic component. As usual with human abilities there is likely a mixture of nature and nurture going on.

A new paper about the importance of our musical training background and later AP ability was released this week by Patricia Vanzella and E. Glenn Schellenberg. These authors were interested in a number of interesting questions regarding AP including:

1)      Why might AP possessors be worse at identifying tone names from sung tones compared to instrumental (usually piano) tones?

2)      Is AP ability affected when an individual learns a fixed pitch instrument (piano) compared to variable pitch instrument (violin)?

3)      Is AP ability affected when an individual learns a fixed do scale (as in do-re-mi) when do is always a fixed pitch (as is the case in Latin countries) compared to when they learn a moveable do scale when do always refers to a tonic but can by any pitch depending on the scale (as is the case in Germanic and English speaking countries)?

The authors developed on online test for AP and their final population sample of 198 contained individuals who scored above chance on note naming on one of four tested timbres. There were a comparable number of men and women (107 men and 91 women) and the mean age was 30 years. Interestingly only half the sample reported that they know they possessed AP, which indicates a wider issue with relying on self-report of AP for population statistics. Maybe it is more common than we currently believe? And of course the majority had begun music lessons when they were very young, although a small proportion had begun lessons after their teenage years. The majority were piano players, but around a quarter were violin/viola player. This division in training would allow testing of the question of musical experience in AP, as piano tones were used in the final test but not violin/viola tones.

For the main test the authors played 96 one second samples of tones on one of four timbres; piano, synthesised voice, natural voice and sine waves. Again, the presentation was by an online test available in both English and Portuguese (the first author conducted the tests from Brazil). After hearing the tone the participant had to press a key on a visual keyboard to indicate the note they thought they heard. Because a fixed do system is used in Brazil, the labels in the Portuguese version of the program were in solfege with Portuguese spellings.

The results showed marked variation in AP ability both across individuals and across timbres. Consistent with previous research, the best result were for the piano tones compared to all other timbres (M = 83%). Second best performance was in pure tones (77%). Performance was then significantly worse for, but did not differ between, natural (73%) and synthesised (75%) voice. This latter result indicates that the vibrato of a natural human voice does not affect AP performance.  The authors combined the results to determine if there was a critical age for starting music lessons and AP ability: They found that those who began lessons before the age of 7 were significantly more likely to possess AP. Also, they found that playing piano but not other instruments (e.g. violin) was associated with unique predictive variance in possession of AP ability (this difference was evident for all 4 timbres). Finally, having a history of a fixed or moveable do system had no effect on the AP results.

This is the largest scale study that has been interested in individual variation in AP performance, and how your musical background might affect your ability. The findings indicate that AP possessors have more trouble identifying tones from voice compared to piano tones, a result which the authors claim might have been caused by interference between music and linguistic/paralinguistic processing systems. They argue that when we hear a voice we automatically process the linguistic and affective components of the sound (possibly related to activation of ‘voice-sensitive cortical regions’ in the brain) and that this might get in the way of simply identifying the pitch, which, following this line of argument, would be a distinct cognitive process. Furthermore the evidence from the present study points to a sensitive period for exposure to music lessons in order to develop accurate AP ability (before age 7), and the authors also hypothesise that this sensitive period may be extended in time if piano is the first instrument studied (since a piano is a fixed tone instrument).

In summary this is a really nice study, well written, showing how you can use the internet to gain a large amount of information about a rare condition. Of course, lab controlled tests are preferable in most cases but studies such as this present us with lots of fascinating hypotheses to test, in this case with regards to AP ability. We are getting closer to understanding what it is about our background in musical training that might lead to this unique skill. Hopefully there is someone out there investigating the genetic side too.