Few would dissent from the view that music has a central and almost essential place in personal cultural life but the reasons for its universality remain obscure. John Barrow offers some possible explanations. There have been cultures without counting, cultures without painting, cultures without science, cultures bereft of the wheel and the written word, but never a culture without music. Music, scented sound, is there in the jungle, in the city, between our ears and at our fingertips.
Without consciously learning its rules we can respond to its rhythms. Musical ability among the very young, like mathematical genius, can be alarmingly sophisticated - out of all proportion to other skills. Some people find that music is a necessary accompaniment for the successful completion of other activities. But its definition is not so easy because music covers a vast range of sound levels and frequencies - from simple repetitive drumming to symphonic works of enormous complexity in which the mental powers and dexterity of scores of individuals combine to represent the patterns encoded in the score.
When one finds transcultural human activities, like writing, speaking and counting, that display common features, we can ask if they have evolved from simpler activities whose raison d'etre is more obvious. If the simpler predecessor of today's complex activity endowed its exponents with a clear advantage in life, because it made them safer, healthier or just plain happier, then it is likely to become more prevalent because of its cultural transmission or, if it derives from some inheritable genetic trait, by becoming more likely to survive and be inherited.
At first sight it is not easy to see what advantage a penchant for Beethoven or the Beatles confers on its possessor. What could have been the utility of such an abstract form of sound generation and appreciation? This is a difficult riddle. Musical appreciation might be merely a by-product of mental attributes evolved for quite different purposes. There are plenty of possibilities to choose from. The earliest, most spontaneous of human sounds is the cry of a baby at birth, when hungry, or distressed. Sounds that are responded to in circumstances of great intimacy and emotion. But humans of all ages retain the ability to make similar sounds and emotional cries and there is no similarity between those cries and music. Indeed, we recognise the instinctive reaction to crying to be one of irritation, unease, or distress - just the reaction we might have expected experience to have reinforced upon our ancestors, but not our response to most forms of music. A clue to music's antecedents may lie in its emotive power. In civilizations ancient and modern, the world over, we find the sound of music whenever there is a need to increase group bonding or inspire acts of courage. But herein lies a paradox. For we find that music calms the overwrought human mind as effectively as it can rouse it. This dichotomy suggests that we will not find the source of musical performance or appreciation in so peculiar a function as the arousal or subduing of specific human emotions. Nor, despite an ancient tradition going back to Pythagoras, does there seem to be a deep connection between mathematics and music. Mathematics is the study of all possible patterns; but music is more than patterns in sound. It is resonant with something. Music can stir mass emotion, inspire nationalism, and religious fervour. Mathematics can do none of these things. The "music of the spheres " is silent and abstract.
A ubiquitous source of sound is the inanimate natural world: the wind, the rush of running water, the crash of thunder. Alas, these are things that one seeks to be heard in spite of; they hardly qualify as templates for human emulation unless you were attempting to camouflage your presence while hunting or hiding. More promising models are sounds from elsewhere in the living world. Mating calls and complex bird-song play a well-defined role in the evolutionary process: mates are attracted and territory demarcated. Darwin thought that music had its prehuman origins in mating calls.
The most impressive feature of music is its temporal continuity. Whereas art displays pattern in space, music offers patterns in time. Just as the mind has developed acute pattern recognition abilities, so it possesses exquisite sensitivity to nuances of sound over a far greater range than it does for visual images. This is part of a wider facility. We have developed ways to make sense of time in ways that transform chains of events into a history. Legends and traditions first played this role and complemented the human understanding of events. The spatial order exhibited in painting or sculpture is heightened when endowed with a temporal aspect. This is why films are often more appealing than still photographs and why children can find video games so addictive. Unchanging images leave the viewer to look for themselves. They can look again and again, first following one sequence, then another. But music has its own sequential order of perception. It has a beginning and an end. A painting does not. Thus we see that music may be associated with a need to structure time or have derived as a by-product of an advantageous adaption for a structuring of time. What sort of advantage could this offer?
Sequential "timing" is something that lies at the heart of all manner of complex human activities, from juggling to starting a car, which require meticulous coordination of eye, brain, and hand. Take crossing the road: we receive visual and sonic information about vehicles moving in different directions at different unknown speeds, viewed at differing angles under variable light; we need to evaluate whether there is an interval of time sufficient to cross the road and then move appropriately. Viewed in this light, it seems astonishing that we ever manage to make it from one side of the street to the other. The brain has clearly developed an extraordinary facility for sequential and parallel timing of different movements that will combine to produce a rapid, single continuous activity like the serving of a tennis ball. Music could be a by-product of the complicated mental circuits that evolved primarily in response to the adaptive advantage offered by an ability to coordinate body movements in precise, continuous, and rapid response to outside changes. The obvious limitation of this type of explanation is the absence of musical inclination in those apes who are the planet's unchallenged gymnastic champions.
If we have evolved to cope with the changing patterns of a complex environment, there may be naturally occurring forms of complexity which our brains are best adapted to apprehend. Thus, artistic appreciation might emerge as a by-product of those adaptations. An interesting aspect of music sound has recently come to light and suggests how this might occur.
Physicists and engineers prosaically refer to sequences of sound as "noises". A useful way to distinguish them is by their spectrum: this is a measure of the distribution of intensity over wave frequencies (just as the spectrum of light revealed by a prism displays the distribution of light with colour, which is just another word for frequency in this context). An important feature of many noise spectra is that their intensities are proportional to a mathematical power of the sound frequency over a very wide range of frequencies. In this case, there is no special frequency that characterises the process - as would result from repeatedly playing the note with the frequency of middle C. Such processes are called scale-free. In a scale-free process, whatever happens in one frequency range, happens in all frequency ranges.
Scale-free noises have intensity spectra that are proportional to inverse powers of the frequency, 1/fa, where a is a positive number. Their character changes significantly as the value of a changes. If noise is entirely random, so that every sound is completely independent of its predecessors, then a is zero, and the process is called "white noise". Like the spectral mixture that we call white light, white noise is acoustically "colourless" - equally anonymous, featureless, and unpredictable at all frequencies, and hence at whatever speed it is played. When your television picture goes haywire, the "snow" that blitzes the screen is a visual display of white noise produced by the random motion of the electrons in the circuitry. At low intensities, white noise has a soothing effect because of its lack of discernable correlations. Consequently, white noise machines are marketed to produce restful background "noise" that resembles the sound of the gently breaking ocean waves. White noise is invariably "surprising", in the sense that the next sound cannot be anticipated from its predecessor. By contrast, a scale-free noise with a = 2 produces a far more correlated sequence of sounds, called "brown noise". This is also unenticing to the ear; its high degree of correlation gives it a predictable development, like a musical scale. It "remembers" something of its history. Brown noises leave no expectation unfulfilled, while white noises are devoid of any expectations that need to be fulfilled. But midway between white and brown noise, when a = 1, lies the special case of "1/f noise". It is special because such signals are moderately correlated, and hence possess 'interesting' patterns over all time intervals. They combine novelty with expectation in an optimal way.
In 1975, Richard Voss and John Clarke, two physicists at the University of California at Berkeley, serendipitously discovered that many classical and modern musical compositions are closely approximated by 1/f noise over a very wide range of frequencies. Appealing music exhibits an optimal level of novelty at the spectral level: neither too predictable like brown noise, nor randomly unpredictable like white noise. This may be telling us important things about the mind's first adaptations to the world of sound. But if musical appreciation is a by-product of a more general pattern-processing propensity of the brain, why are our senses heightened by 1/f noises? It is significant that the world around us is full of variations with 1/f spectra. One reason is the prevalence of sequential processes in the natural world. Benoit Mandelbrot has claimed that our nervous system acts as a spectral filter, preferentially passing 1/f noise to the brain while filtering out white noise at the periphery to prevent the brain being swamped with uninteresting random background noise about the world. An optimal response to signals with 1/f form might well be the best investment of resources that the system can make for information gathering, or it may be the simplest way for the nervous system to decode vibrations in the inner ear.
Music is the purest art form. Our minds receive a sequence of sounds woven into a pattern - largely undistracted by the other senses. The fact that a wide range of music exhibits a 1/f spectrum for its variations in loudness, pitch and interval, across the whole range from classical to jazz and rock, suggests that this appeal arises from an affinity for the statistical features of natural noises, whose detection and assimilation were adaptively advantageous to humans. This affinity extends to many varieties of non-western music; there is a good approximation to 1/f noise in all of the musical traditions studied: from the music of the Ba-Benzele Pygmies, traditional musical culture could be found where there was a significant, habitual deviation from the 1/f spectral form.
One should not regard this argument as totally reductionist, any more than one should take seriously music-lovers' claims that music is a transcendental form with charm beyond words. Our minds, with their propensity to analyse, distinguish, and respond to sounds of certain sorts, yet ignore others, have histories. Musicality seems most reasonably explained as an elaboration of abilities that were evolved originally for other more mundane, but essential, purposes. Our aptitude for sound processing converged upon a sensitivity for certain sound patterns, because their recognition optimised the reception of vital information. With the emergence of that more elaborate processing ability we call consciousness, has arisen an ability to explore and exploit our innate sensibility to sound. This has led to organised sound forms that span the range of pitches and intensities to which the human ear is sensitive.
Those forms diverge in their stylistic nuances, as do the decorations in peoples homes, from culture to culture. But the universality of musical appreciation, and the common spectral character of so much of the sound that we embrace, behoves us to look at the universal aspects of early experience for an explanation. Had the sounds that fill our world been different in their spectral properties, we would have developed a penchant for sounds with quite different structures - structures that from the spectral perspective would have been more surprising or more predictable.
Alternatively, and more likely, is that the source of 1/f noise in our environment is a consequence of statistical processes of such generality, that this form of noise would be ubiquitous in any environment - terrestrial or extraterrestrial. In that case, we might have been overly pessimistic in believing that music could not be used to communicate with extraterrestrials. If, as we might expect, their environments display a cacophony of vital 1/f-spectral variations, they should have evolved a special sensitivity to them. When transformed into the appropriate medium, they might well appreciate some of our music - which is just as well because the Voyager spacecraft,now heading out of the solar system to the stars, contains an elaborate recording of terrestrial sounds. Ninety minutes of music was included - Bach, Beethoven, rock and jazz, together with folk music from a variety of countries. The senders did not know it, but it all has a 1/f spectrum.
John D Barrow is professor of astronomy at Sussex university. His book The Artful Universe will be published by Oxford University Press in October.