Monday, October 17, 2011

Chord Discrimination in Pigeons - Searching for the Evolutionary Origins of Human Perception of Music

Source: Brooks, Daniel I. and Cook, Robert G. Chord Discrimination in Pigeons. Music Perception: An Interdisciplinary Journal, Vol. 27, No. 3 (February 2010), pp. 183-196

Retrieved: October 17, 2010, from JSTOR


Though we may consider ourselves to be evolved and advanced today, there is substantial evidence to suggest that humans as a species originated from the most simple and primitive of life forms. From this it logically follows that the components that make us what we are have their roots in more primitive life forms, so to discover how we arrived at our current state we must look to the past and examine the biological and cognitive processes of even our most distant animal relatives.

It is for this reason that the study done by Daniel I. Brooks and Robert G. Cook entitled Chord Discrimination in Pigeons is useful to those who wish to understand the origins and evolution of human perception of music. The cognitive processes that help us identify the melodic, harmonic, and rhythmic components of music must have had some precursors in non-human primates, according to Brooks and Cook.

The focus of this study is interval perception in pigeons. Brooks and Cook note that interval perception is an interesting perceptual skill to study because it is so important to the way humans perceive differences between individual pitches and melodies. Finding out how it functions and develops in other species is the first step to understanding how it developed in humans.

Studies on musical perception and discrimination in animals have been done on a number of species, including songbirds and primates, but they have never been done on non-songbirds such as pigeons. Two new experiments were reported in this article. Experiment 1 involved training the pigeons with chords developed from the C major scale. Pigeons were trained in a go/no-go task to distinguish a C major triad from 4 other triads, each of which differed from the C major triad by one semitone (the no-go triads were C minor, C suspended 4, C flat 5, and C augmented). Pigeons were given food reinforcement when they pecked after hearing a C major triad, and no reinforcement when they pecked after hearing one of the other four triads. The training took place over fifty sessions. As the training progressed, 3 out of 5 pigeons successfully learned to discriminate among the five triads. The augmented triad was shown to be the easiest for the pigeons to identify as no-go (no pecking), while the other triads proved to be of variable levels of difficulty. This experiment revealed, for the first time, that non-song birds are able to discriminate between triadic chords differing by only one semitone.

In Experiment 2, a second set of triads was added to the discrimination test. These triads were the same type, but were based on a D root. The assumption was that if the pigeons has learned the general harmonic configuration of the chords (the relations between the notes), then they should be able to transfer this recognition to chords with a new root. This test proved more difficult for the birds, and while they were still inclined to recognize the augmented trias as no-go, they were not as successful in general in identifying go versus no-go triads.

According to Brooks and Cook, these results suggest that pigeons are able to identify different frequencies but are not as adept at identifying the relationships between different frequencies played simultaneously. Previous research suggests that in the auditory domain, birds and mammals differ in their ability to use absolute versus relational stimuli. This is especially so with regard to their capacity to process the absolute value of pitch. It is speculated that in general, birds tend to recognize more the absolute value of pitches, while mammals tend to recognize relationships between pitches.

When comparing the results of this study with other studies done on similar topics, such as the discrimination of chords in song-birds, or the discrimination of consonance and dissonance in humans, two interesting points present themselves and were mentioned in the general discussion section of the article. Firstly, as it has been shown through studies that song-birds can distinguish between chords, looking only at research done on song-birds might encourage one to assume that this ability could be the result of biological mechanisms responsible for learning songs, and therefore necessary for mating and survival as a species. However, since pigeons are non-song birds and this study suggest that they possess a similar ability to distinguish between chords, this is not necessarily the case. It seems that this ability is widely shared across birds as a class rather than just belonging to particular species for which it is a necessary survival skill. Secondly, further observations were made when the Brooks and Cook did studies on humans, asking them to define the relative consonance and dissonance of the same set of chords. Interestingly, humans and pigeons seemed to agree on many things, for example that the augmented 5 triad was the most different, or easiest to distinguish from the major triad. This may suggest while cultural conditioning plays a role in our perception of harmonies, there is a unified account that can be made of harmonic perception among all species.


Neurological descriptions and explanations of absolute pitch in humans are still unclear and unproven. In many ways this phenomenon is a mystery for neuroscientists and musicians alike. It is interesting, and also somewhat refreshing, to learn from this article that some animals may find using absolute pitch easier and more natural than using relative pitch, while in my experience the reverse is most often the case for humans. When I was in my first year in university, one of my professors said to the class ‘ok, if you have perfect pitch put up your hand.’ When those with the skill identified themselves, a sort of smile crept over his face that seemed to say to me ‘be envious class, these are the special people.’ Indeed my experience with opinions about absolute pitch from a cultural standpoint is that it is highly valued and admired, more so than relative pitch.

Do other species have a reversed hierarchy of importance for the skills of relative and absolute pitch? Perhaps this is the case for pigeons. So it causes me to wonder, why does this happen? Why do species prefer one skill over the other? What does this say about the nature of the two skills? Is absolute pitch a learned skill, or is there a gene that bestows some species, or some people, with the power to instantly recognize frequencies? If it is learned, then why do some people seem to display the skill so quickly and accurately that it appears to be automatic, while others can only use the skill in certain situations and with much less accuracy? If it is genetic, then do humans with true absolute pitch actually have a gene that was passed down from their aviary ancestors, while others discarded this gene in favour of some sort of ‘relative pitch gene?’ Is the absolute pitch gene recessive? Could it possibly go extinct like the gene for red hair?


Chairat said...

First of all, having absolute pitch myself, it amuses me immensely to learn that I may, to some degree, share this trait with the very creatures that I used to frighten into flight when I was much younger...

Amusing reflection aside, I think that absolute pitch is overrated. I agree with you that, from a cultural standpoint, absolute pitch seems to be more highly valued and admired than relative pitch. But personally, I find absolute pitch a hindrance at times, not only when listening to period performances in which A is disturbingly lower than 440Hz. For example, I still remember struggling to identify intervals during the ear training section of my RCM practical exams. I could tell exactly what the two pitches of a certain interval were, but then I actually had to count the steps between them to identify that interval; I had not learned to recognize intervals by their distinct overall qualities. This was a serious mistake because it implied that I did not properly conceptualize the “space” between the notes. And, as my teacher pointed out to me several times over the years: “Music happens between the notes”.

I find it very interesting that the augmented triad should prove the easiest to distinguish from the major triad in both pigeons and humans. Had the researchers established the augmented triad as the “go” chord, instead of the major triad, I suppose that they would have observed some impressive results. Among the five triads that the researchers used, I think that the augmented triad is inherently the most distinctive-sounding because it is the only “symmetrical” triad in the set, being made up of two major thirds. This means that only four distinct augmented triads are possible (in a system of twelve pitches, all equal “distance” apart), making them much rarer (i.e. easier to identify) than the other triads.

Federico said...

Some people with perfect pitch struggle in contextualizing syntactically the pitches that they hear; others, while able to associate note names to frequencies, cannot sing the pitches that they hear. My understanding is that perfect pitch is mainly a mnemonic trait (of a special kind). I wonder how much social expectations and each person's sensibility and personal history play in the development of perfect pitch. For example, the fact that, among Chinese musicians, there is a higher rate of perfect pitch than among European musicians, makes me wonder whether perfect pitch is acquired in China more easily than in Europe by virtue of a more systematic and efficient ear training. Also, the rate of perfect pitch is higher among people who are blind from birth: does that mean that pitch perception is somehow also related to sight? Or to other senses?

Also, why did certain places develop tonal languages and others not? any underlying genetic influences such as increased prevalence of perfect pitch before that?

The experiment shows pigeons as acquiring, or possessing, perfect pitch. I think the distinction between acquiring and possessing constitutes an interesting point, which makes me wonder whether the brain processes perfect pitch information in the same area/s where relative pitch is processed, or not.