The Origin of Music: new light on an old hypothesis

Graham Pont

The University of New South Wales, School of Science and Technology Studies, Sydney 2052 Australia. Fax 612 313 7984

The birds taught us to sing: thus the ancient poets explained the art of music. As a system of animal communication, regularly pitched and inflected sound was possibly also the basis of spoken language. Statistical comparison of human and avian pitch profiles may throw new light on the origin of music and human communication in general.

The theory that human melody was acquired from the birds was well known to the ancient Greeks and Romans but the idea may go back to much older poetry and myth. There can be no doubt that birds and human beings have copied each other's behavior for untold millennia: our music abounds in imitations of birdsong, from the exquisite trill of the nightingale in Handel's L'allegro and Il Penseroso to the evocative call of the cuckoo in Bethoven's Pastoral Symphony. It has recently been hypothesized that Mozart´s G Major Piano Concerto (K453) incorporates a theme from the singing of his pet starling (The Independent, February 1991). In his Natural History, Pliny claimed that the nightingale was the only bird the notes of which are modulated in accordance with the strict rules of musical science - hardly surprising when many of the best human musicians have modeled their song on that or the nightingale.

In taking their inspiration from the birds, composers might well be continuing an age-old collaboration between these preeminently musical species. Birds are well known too for their abilities in copying not only the songs of other birds but also human song, speech and even dance. The intimate connection between the song of the Beautiful Fruitdove (Ptilinopus pulchellus) an the thought, language and sensibility of the Kaluli tribe in New Guinea has been brilliantly documented by Steven Feld (Sound and Sentiment, 1982/90). Could this association be a living fossil - a survival- of a prehistoric stage in human development, when we acquired a communication system based on that of the birds? The pervasive role of birds in classical mythology, and angels in Christian belief, may point to remote analogous developments in the northern hemisphere. Another clue might be the widespread, if not worldwide association between birds and the oldest known musical specialists, the shamans, who could understand the language of the birds and imitate their flight. At the beginning of the age of science, these magical powers were still being attributed to Pythagoras.

Yet melody, properly so-called, is not a significant part of our direct evolutionary inheritance. In The Descent of Man, Darwin drew attention to the singing of the gibbons Hylobates agilis and H. leuciscus but little else to that the progenitors of man... endeavored to charm each other with musical notes and rhythms... Generally speaking, the primates are not very musical. Our nearest relatives, the chimpanzees, cannot speak or sing and the vocalizations of the others cannot be compared with the artistry of human and avian melodists. From a broad evolutionary point of view, human song - like our linguistic ability - appears to be a recent and largely unprecedented intrusion in the natural history of the primates. How then can explain humanity's sudden acquisition of its marvelous vocal arts?

It is at least possible that we acquired the basic method o pitch control and inflection through imitating birds, whose melodic art has been universally admired. The growth of human music may, indeed, have been the result of direct imitation rather than slow evolutionary development - a relatively quick technological transfer (possibly within the lifespan of Homo sapiens sapiens). This supposition could account for the curious fact that they produce sounds by means of completely different physiological systems -the syrinx and the larynx.

Our hypothesis is enhanced by Konrad Lorenz's observation (Behind the Mirror, 1973/77) that birds are 'our only colleagues in the art of music and imitation' - which certainly limits the range of possibilities! Since there is no other terrestrial musicality that can be compared with avian and human song; and, since feathered bipeds and much older that featherless bipeds, the traditional belief that 'the birds taught us to sing' seems all the more reasonable. This conclusion in turn might persuade us to reconsider the old theory that spoken language was derived from music, which has long been considered the 'natural' language of humanity.

The existence of whale symphonies does not invalidate our argument. Despite their extraordinary musical powers (unequaled in range and second only to humanity's in duration and variety of performance), the influence of whales did not extend much beyond the mariner's poetic impressions of the mermaid's song. There can only have been very limited contact between human beings and whales but our earth-bound predecessors lived in direct and intimate contact with avian songsters throughout the evolution of the species. In tribal societies living close to nature, the regular concerts of the birds would have become familiar to the growing human, even before birth.

In The Evolution of Bird-Song (1895), Charles A. Witchell reports may examples of birds imitating human music. Black birds nesting in his garden during 1888 heard the piano being played for an hour or two each day and soon developed a song based on the common chord of C mayor. Variations of this song were still being sung (possibly by the same bird) four years later. Given their powers of mimicry, it is not surprising that the birds share our music - and perhaps even our aesthetic enjoyment of it (Hartshorne, 1973). As Witchell points out, English birds have been about a thousand years! In the New England National Park, Australia, territorial calls of the resident lyrebirds include still-recognizable snatches of human melody acquired in the '30s by a pet lyrebird from its flute-playing owner.

The science of melody has recently been advanced by the statistical study of pitch profiles. The method developed by Denys Parsons (in The Directory of Tunes and Musical Themes, Cambridge, 1975) compares melodies in terms of their three basic movements in pitch from note to note: 'Up' (a rise in pitch), 'Down' (a fall in pitch) and 'Repeat' (no change in pitch). By comparing the first three notes o more than 10,000 classical and popular tunes, Parsons discovered a new law of western melody (New Scientist, 24 March, 1977). Parsons' method was designed for, and has been extensively tested on human song but there is no reason why it cannot also be applied to birdsong - at least to birdsongs which have been reliably transcribed into musical notation. Producing such transcriptions by ear and hand is a very laborious business - and is not done much these days. But, within the limits of conventional notation, a technically competent observer can produce approximate transcriptions which are musically and scientifically acceptable. Perhaps the century's most famous birdsong-fancier in this line was the composer Olivier Messiaen. (Catalogue d'oiseaux, 1959).

The availability of four substantial collections of birdsong in musical notation, two English and two American, made it possible to perform some preliminary experiments on the melodic relationships between human and avian song. Research has suggested that the 'preferred pitch profile' of a given culture or region can be remarkably stable: the 'Up-Up' profile that Parsons found to be the consistently preferred for the first two intervals of modern European melody is also the most frequent in traditional German folksong, Gregorian chant and (as far as the very small sample indicates) ancient Greek music (Pont, New Scientist, 24 January, 1990). Parsons found that the overall order of preference among western composers in their use of the nine possible initial profiles is also amazingly consistent. Much more research is needed, however, to determine whether this kind of pitch profile preference is really part of the 'deep structure' of human melody.

If human and avian song have been interacting since time immemorial, then one would expect to find significant similarities of pitch profile preferences between the melodies of the two species. Given Parsons' Law for the order of initial pitch profiles in recent western music, one would have to predict that the order for British birdsongs would be similar. A preliminary summary of my work shows that, according to Witchell's and Gardiner's scoring of British birdsongs, the preferred avian profile is not 'Up-Up', but 'Up-Down', the second most preferred by western musicians. Overall, however, the ordering of the nine possibilities is very similar. Comparing the rank orders according to Spearman's test, p = 0.85 which is high (the probability of this agreement being 0.0023%). As Parsons himself has pointed out; the statistics suggest that birds and human beings are 'from the same population'. These samples, of course, are far from perfectly matched: it would be interesting to compare the Witchell and Gardiner figures with those derived from a sample of more specifically English composers or, better still, with the traditional or most enduringly popular human melodies from the area inhabited by the avian sample.

But, given the limitations of the experiment, these first results are at least encouraging: they point to the possibility of much more precise and extensive observations of musical relationships between human and avian cultures.

The results of the same test repeated with two the collections of American birds songs (Cheney, 1892 and Saunders, 1935) show the 'Up-Down' profile again to be very prominent but this time rivaled or surpassed by two other profiles, 'Down-Up' and 'Repeat-Repeat'. When these figures are compared with those obtained from 724 North American Indian melodies, the correlation of pitch profiles is surprisingly low (p = 0.167). But the prominence of the 'Repeat-Repeat' profile in both American Indian melody (about 40%) and in the birdsongs of that country (nearly 25%) possibly indicates what we are looking for: regional co-variation of human and avian pitch profiles. The higher levels of frequency for 'Repeat-Repeat' in both human and avian song of America, as compared with those for western music and British birds, is very suggestive. This difference could point to a general distinction between the musics of the two regions: that is, human and avian melody in America might be more monotonous ('Repeat-Repeat') than melodious ('Up-Down', 'Down-Up').

These figures yield another interesting result: they indicate that human and avian songs of both the old and new worlds fall into two distinct groups, monotonic and melodic (which correspond roughly to chanting and singing). Here the statistical comparison of pitch profiles not only confirms the traditional division of human melodies but also extends it to geographically related birdsongs. The proportions are identical: for our sample, 63 % of human songs and 62 % of avian songs are melodic (Fig. 1, Pearson's Chi-squared 0.04, probability that they represent the same population = 0.84).

If this kind of correlation were supported by further studies, it would strengthen the supposition that human patterns of pitch profiles are avian in origin. This in turn might also help to explain the origin of human music in general, and perhaps have some light to shed on the origin of language.


For their generous advice and collaboration, the author thanks Mr. Denys Parsons (Highgate, London); Dr. Nigel Nettheim and Dr. Jim Franklin (University of New South Wales) and Dr. Julian Monge-Nágera (University of Costa Rica). This research was supported by the Australian Research Council (1991-3).


S. P. Cheney, Wood Notes Wild. (Boston, 1892)

Charles Darwin, The Descent of Man. (2nd ed. London, 1875)

Steven Feld, Sound and Sentiment; Birds, Weeping, Poetics and Song in Kaluli Expression. (Philadelphia, 1982/1990).

William Gardiner, The Music of Nature (London, 1832)

Charles Hartshorne, Born to Sing; an Interpretation and World Survey of Bird Song (Bloomington and London, 1973)

Denys Parsons, The Directory of Tunes and Musical Themes (Cambridge, 1975)

Aretas A. Saunders, A Guide to Bird Songs (New York, 1935/51)

Charles Witchell, The Evolution of Bird Song (London, 1896)