The biological foundations of musical improvisation
Consideration of the biological origins of music is enjoying a revival after many years of suspicion of biological and universalist ideas in musicology and the anthropology of music. For example, Ellen Dissanayake has recently derived humanity’s overall tendency towards music-making from mother-and-infant vocal behaviour. She writes that ‘the enjoyment and capacity of producing musical notes are faculties of indispensable use in the daily habits of … mothers, and their infants, and … it is in the affiliative interactions between mothers and infants … that we can discover the origins of the competencies and sensitivities that gave rise to human music.’1 ‘Together, mother and baby practice and perfect their attunement by engaging in mutually improvised (jointly constructed) dyadic interactions’.2 These are ‘literally, not just metaphorically, musical.’3 Dissanayake proceeds to parallel specific central features of these mother-infant interactions with certain specific central features of humanity’s music which she believes derive from them. Infant and adult musical activities, she argues, both exhibit tendencies towards what she calls ‘regulation’, ‘disruption and repair’ and ‘heightened affective moments’.4

One point of apparent similarity between infant-and-mother musical dialogue and wider human musicality is, however, touched on only briefly by Dissanayake.5 I refer to the similarity whereby both spheres of activity embody a capacity for a broadly improvisatory, experimental or exploratory manner of musical production. No other species of mammal produces even a tiny fraction of the unstereotypic vocal play found in the human baby.6 Babbling infants up to the age of 18 months are not trying to imitate songs they hear. Instead they seem to be experimenting with the melodic use of intervals.7 Since much adult behaviour develops from infant behaviour this factor in infant development may provide a clue to the nature of human musical activity more widely. If, at the infantile stage, music begins in exploration, perhaps music will never fully shed this exploratory trait as the human person grows older. Perhaps, in other words, exploration will remain a normal attribute of humanity’s musical activity. Perhaps the music-making which arises in later life might reflect consistently not merely babbling’s musical attributes in general, but its exploratory musical attributes in particular. In the light of this, it is remarkable that a speculative connection between babbling and improvisation in adult music-making is hardly ever made, and even less explored.8 Improvisation, the creative realisation of a musical product as far as possible within the time of its sonic production, is, after all, pre-eminently the exploratory and experimental mode of adult humanity’s music.

Moreover, just as infant babbling seems to be a biological universal, ethnomusicologists have sometimes identified improvisation as a universal of world musical culture. Bruno Nettl observed some years ago that ‘[i]f the concept of improvisation can be said to be at all viable, it should be considered one of the few universals of music in which all cultures share in one way or another.’9 Others have noted more recently that ‘[a]lthough it may be surprising to a Westerner, improvisation is at the centre of rather more musical traditions in the world than composition … It is actually only in Western music that there has ever been any significant period in which improvisation was not central.’10 Perhaps the reason why improvisation in musical cultures is universal or central, is that musical cultures are constructed by human beings who all encountered music-making first of all in their exploratory infant babbling.

The similarities of production, between the exploratory nature of infant babbling and the musical improvisation widespread in human cultures, suggest that there may be a single process underlying both. There is, in other words, a natural continuity from the vocal exploration of babies to the musical improvisations found in adults’ (and older children’s) musical cultures. These improvisations seem to be where infant babbling as early exploratory vocalising are somehow preserved, in a transmuted form, into the later life of the person. Later musical improvisations thus recall and extend infantile engagement with the mother, the music-making which first integrated us with the interpersonal factor of our human existence. Not surprisingly, given this interpersonal dimension, much adult musical improvisation is group-based rather than solo in nature. While not all such improvisation is group-based (the solitary improviser is a well-known phenomenon in Western music) it would seem that, worldwide, most is.11

If later improvisation comes from early babble, then it will also unconsciously recall it, in a manner analogous to the psychotherapeutic recall of the early stages of the life of the analysand. This may help to explain the success of improvisational musical therapies. In words of Mechtilde Langenberg: ‘What is involved in the product of musical improvisation? The living process of encounter and relating provokes feelings, fantasies and images that are signs for the staging of subjectively experienced, internalized, interhuman relational experiences and conflicts from earlier developmental phases.’12 Likewise the use of improvisation in the context of psychological healing has been appropriated by Don Cherry and Nana Vasconcelos.

Improvisation and the scored work in the Western musical tradition
The history of human music is mainly a history of musical improvising. Obviously, in claiming (following Nettl, above) that improvisation is a universal feature of human music, it is not implied that this musical universality of improvisation is equally evident or prominent in all musical cultures. Within the confines of specific cultures, other factors may intervene so as to veil the natural human preponderance of improvisation in music-making. Western Europe is an example of a culture in which the improvisatory essence of human musical behaviour has become veiled. It has become veiled in the West by the preferred emphasis on performance from the notated score. Although explicitly improvised music does still occur in the West (e.g. the jazz improviser, the folk singer) it is not immediately apparent how improvisatory practices occupy a central position in Western musical culture. Therefore in considering music as it arises in the West, there is little or no hint that improvisation may be a general feature of human music.

In saying that, in the West, improvisation is not culturally prominent, it is not implied that the overwhelming urge which human music has to assume an improvisatory form has been obliterated in this culture. In Western music, the major outlet for music’s improvisatory impulse is found in the ‘expressiveness’ which musicians bring to the performance of a notated score. This expressiveness entails exploratory performance decisions, which are a form of improvisation. We may consider an example of this. When a crescendo marked on a musical score is performed, the exact process whereby the loudness of the music grows, and indeed the overall degree by which the loudness grows both involve exploratory decision-making by the performer in performance. There will always be some element of ‘making up the crescendo as we go along’, of exploring our way into the crescendo, and it has generally escaped notice that this phenomenon is implicitly improvisatory. Clearly, the crescendo is not improvisation in the sense of being ‘an improvisation’. Yet, each crescendo will inevitably be unique, because the performance of the crescendo always requires more detail in relation to the crescendo than can be held in the performer’s memory or be represented by the score. This situation is not to be viewed as some kind of problem. On the contrary, the inspired performer does not aim to make each performance the same, and will actually want to bring the crescendo to new life in every performance. This will give rise each time to a unique crescendo. Without realising it, he or she wants the performance to be improvisatory, even though not ‘an’ improvisation. In short, the written crescendo is not just implemented. It both has to be, and seeks to be, ‘improvised out’ into real life. It goes against the grain of Western performance practices to wish to render each performance of a given work identical (something impossible in any case). We sense that to attempt this would be to deliver the music up to the listener as a lifeless parody of the lived involvement we expect from a performer in performance, which is essentially an improvisatory involvement.

This illustration from the crescendo may be enlarged to shed light on the whole process of musical performance from the notated score. As Bruce Ellis Benson has recently noted: ‘Performers – even when performing music that is strictly notated – do not merely “perform” but also “improvise” upon that which they perform.’13 When we perform a notated composition, therefore, we always rethink the music musically as we perform. Just as our crescendo was constructed as a form of improvisatoriness, so, holding together a musical piece as a homogeneous unity, or easing contrapuntal voices at the keyboard into a balanced relationship, or reacting to unforeseen situations arising within a performance, are all improvisatory activities.14 We have to explore our way into any successful performance, and this process of rethinking the music musically is present even when, in terms of conventional cultural judgements, an exact repetition of the same piece is considered to be taking place. Thus we do not feel it necessary to say, each time we hear a performance of Beethoven’s Waldstein Sonata, that it will involve newly modified expressive decisions. We simply describe it as another performance of the Waldstein Sonata. Nevertheless, the interpretation of this score will normally display improvisatory activity compressed into the microscopic domain of expressive adjustments. We listen ‘through’ such adjustments.15 Because in the case of the Waldstein Sonata we listen ‘through’ the expressive adjustments, we perceive the music as ‘the same’ piece as on the previous occasion it was performed. Yet despite this cultural judgement of identity which we thereby pass upon each performance, each performance is not in fact identical at all with any other, but, on account of these expressive adjustments, a guided improvisation within parameters defined by the score.

Once it is seen that improvisatoriness embraces musical expressiveness in general, improvisation is seen to be foundational to musical production even in cultures which, like that of the West, are dominated by the score. Even where, as in the West, the scored performance has taken over from improvisations as the major methodology of musical production, the improvisatory element remains pervasive, taking the form of expressive decisions brought to the performance of the score by the performer. While a mechanical device like a pianola might reproduce the Waldstein Sonata the same way each time,16 human pianists will normally choose to introduce expressive freedoms when performing it. This insight is perhaps preserved unconsciously in the words used in some European languages, including English, to describe the process of the work’s performance: the words ‘play’, ‘jouer’, and ‘spielen’. We do not simply ‘do’ a performance: we ‘play’ it.17 These ‘playful’ words implicitly define the scored work in terms of something which, in the course of performance, absorbs improvisation and retains improvisation within itself.18
I have proposed so far, in essentials, that music-making may have roots in the exploratory activity of babbling, a fact echoed in the dominance of improvisation as a method of music-making, when music is considered in world terms. I have also suggested that this human musical impulse to explore can be sublimated in the form of exploration applied to pre-existent musical structures, as when a Western musician performs expressively from a score. What is being claimed, in short, is that improvisation tends to inhabit all music. In other words, it is a musical universal.

Improvisation as a musical universal
What exactly is a musical universal? Music is commonly thought to exist always in the form of real or imagined sound – this sonic property is one of its universals. However, it is possible that music has certain general properties additional to this and, if it does, these properties are also correctly referred to as musical universals. David McAllister has proposed some speculative examples of musical universals. First, music tends to be organised into units which possess a beginning, a middle and an end. Although music with some kind of unexpected beginning or ending is possible, conventions of melody or rhythm usually indicate to the informed listener that a piece is beginning or finishing. Second, most music establishes a key, or tonic, of some kind, from which it starts and to which it will often return at the end. Third, music tends to give some impression of goal and direction, in such a way that the informed listener’s expectations are aroused by the music, and then satisfied.19 McAllister speaks of these as properties which music often has or tends to have. This is an important point in relation to musical universals.

Because they can be suppressed or contravened, universals only tend to be present in music, yet they are no less truly universals on this account. As Leonard Meyer says, although ‘[u]niversals are immutable, constraining relationships in the physical, biological and cultural worlds … their action can be qualified or contravened.’20 For Mantle Hood, universality is ‘high probability of occurrence.’21 Universals are therefore to be seen as major tendencies in human music, and not characteristics invariably present in every musical utterance.22 Undoubtedly, although it would not make for an idiomatic performance, the pianist in the Waldstein Sonata is physically capable of repressing his or her exploratory (or improvisatory) response to the music in performance, and playing the notes woodenly.

An example of the paradoxical idea of how a human universal can be contravened can be taken from speech. In every known language, humans inflect the voice in what is called speech intonation.23 The fact that voice pitch goes up and down in this way is a speech universal. However, anybody who feels like it may try to speak at a level pitch without intonation (for example, in the manner of a science-fiction robot) and in this way he or she may suppress or modify a universal. To consider improvisation as a musical universal is therefore not to claim its presence in all musical manifestations, but to assert that every sounding musical event seeks in the first place to possess an improvisatory character, unless the latter is suppressed or overridden. Elements of improvisation will normally be present in any musical utterance but need not be present. Worldwide, therefore, musical behaviour seeks to have an improvisatory dimension unless that dimension is qualified or contravened. Music, paradoxically, is improvisatory except when it is not – in much the same sense that human beings have two legs except when they do not, or that the sun always sets at night, except (in a few areas of the world) where it does not.

Sometimes the deliberate suppression of exploration/improvisation in performance is even an idiomatic feature of a performance style. In some Western pop-music styles, the live performance is widely expected to ‘sound like the record’. This entails that the live performers are precluded from attempting any on-the-spot variety in their performance decisions. Fans want to hear the music performed just like the recording they already know. They want to hear a live copy of a recorded performance. In other words the improvisatory impulse is expected to be overridden here, in deference to a specific performance expectation. Such restriction of exploratory musical elements nevertheless takes place against the cultural horizon of Western European society in which exploratory performance is the norm for musical performance. The restriction, in the pop music referred to, is a local, subcultural departure from this wider cultural norm, rather than the absence of a norm, and is a good example of how a musical universal can be suppressed. Where, as in these pop-music styles, exploratory behaviour is placed under constraint in a musical performance, this is not because such expressiveness is absent from the wider musical culture in which that performance is set, but because it has been intentionally repressed within that subcultural context. One would not say, for example, because a certain photograph uses only black and white images, that the photographer or the intended viewers of the photograph do not understand what colour perception is. On this analogy, the pop musicians described here, in overriding the form of expressive exploration which tends to inhabit all human music-making, are like photographers who have confined themselves to black and white.

A similar point can be made regarding music by Steve Reich and Philip Glass, where it is precisely the absence of expressive/exploratory flexibility which the listener chiefly notices. This reinforces the idea that the cultural context, within which the music of Reich and Glass is received, involves wider assumptions by the listener regarding the normal presence of expressiveness in music-making. In this music of Reich and Glass, improvisation/exploration is consciously missing for the listener, or present in absence. It is not wholly ‘other’ in relation to what is being heard, in the sense that would be implied by its being outside the experience or sound-conceptions of the listener. If it were wholly ‘other’ it could not be missed. Stated another way, the suppression of musical exploratoriness is here a kind of reverse-expressive decision making, covert reference to the exploratoriness/improvisatoriness which universally characterises music. The performer knows he or she is suppressing something usually present, rather as a person who deliberately speaks in a monotone knows that he or she is suppressing the normal intonations of speech. The listener to Reich and Glass also knows that something is being suppressed. Just as no amount of immersion in black-and-white photography ever causes us to forget that the real world exists in colour, so no world in which all music had been deprived of its exploratory dimension – equivalently an improvisatory dimension – could be a human world.

Conclusion
I have used the traditional image of ‘turtles all the way down’ to characterise my idea of the presence of a single principle active within all levels of music making: improvisation. The phrase has, I think – speaking under correction – Hindu origins and refers to the idea that the world rests on the cosmic support of turtles standing, one upon another, all the way down. In other words, the phrase conveys the idea of a single founding principle in the cosmos: turtles. I have sought to propose a perspective from infantile development out of which improvisation’s musical universality might be substantiated. I have suggested that all music-making possesses in the first instance an exploratory/improvisatory character. This exploratory attribute of music-making goes back to the human infant’s first explorations of its interactive social world, from where human music-making first emerges. This inbuilt musicality of infancy is never ultimately supplanted in later life by some entirely different musicality devoid of exploration in the moment, even though it is also true that this primary exploratory character of music may be suppressed. In life we tend to want to go on doing with music what we did with it when we were babies, namely exploring; this is indeed our first instinct, whenever we make music. Music is an anamnesis (a calling up of the past into the present) of our infant selves. Likewise it is infantile exploratory vocalising which furnishes the deep exploratory impulses which seek to find their way out in human music-making. The musicality of the infant shapes musical cultures into their exploratory forms as improvisatory sound cultures. It is probably no accident that, just as the mother-and-baby exploratory musical interaction is emotional, so the sound exploration found in later musical expressiveness is widely perceived as contributing an emotional dimension to the music. ‘Mother-infant dialogue seems to be the prototype for a kind of fundamental emotional narrative that adult music … can grow out of, build upon, exemplify, and sustain.’24

Notes
1. Ellen Dissanayake, ‘Antecedents of the Temporal Arts in Early Mother-Infant Interaction’. In The Origins of Music, Nils L. Wallin, Björn Merker and Steven Brown (eds), Cambridge MA: MIT, 2001, p. 389.
2. Ibid, p. 391.
3. Ibid, p. 394.
4. Ibid, pp. 395-396.
5. See ibid, p. 403.
6. See Terrence Deacon, The Symbolic Species: The Co-Evolution of Language and the Human Brain, London: Penguin, 1997, p. 251.
7. See John A. Sloboda, The Musical Mind: The Cognitive Psychology of Music (1985), Oxford: Clarendon, 1993, p. 202.
8. I attempt this in my Musical Improvisation. Heidegger and the Liturgy: A Journey to the Heart of Hope, Lampeter: Edwin Mellen Press, 2003.
9. Bruno Nettl, ‘Thoughts on Improvisation: A Comparative Approach.’ The Musical Quarterly 60, no. 1 (1974): 1-19, p. 4.
10. Hazel Smith and Roger Dean, Improvisation, Hypermedia and the Arts since 1945, Amsterdam: Harwood, 1997, pp. 55-56.
11. See R. Keith Sawyer, ‘Improvised Conversations: Music, Collaboration, and Development’. Psychology of Music 27, no. 2 (1999): 192-204, p. 193.
12. Mechtilde Langenberg, ‘On Understanding Music Therapy: Free Musical Improvisation as a Method of Treatment’. The World of Music 39, no. 1 (1997): 97-110; p. 98.
13. Bruce Ellis Benson, The Improvisation of Musical Dialogue: A Phenomenology of Music, Cambridge: Cambridge Univ. Press, 2003, p. 26.
14. See David Elliott, Music Matters, Oxford: Oxford Univ. Press, 1995, p. 59.
15. Something similar happens when, in Schenkerian listening, we attend ‘through’ foreground prolongations and embellishments to the middleground or background levels.
16. One may note in passing that even a pianola roll is created by a human person who introduced into it, initially, expressive decisions not specified by the score.
17. I am told by Irish speakers that this transferred sense of ‘play’ does not arise in Irish: one does not ‘play’ (‘imir’) a musical piece.
18. Of course, the authority to introduce improvisatory/expressive decisions into the performance of a notated score may be delegated, in the case of group music-making, to a conductor or director who is not a first-hand contributor to the sound being made. If there is no external conductor or director of a musical ensemble (e.g. in a string quartet), an elusive interpersonal dynamic among the players informs the achievement of their collective expressive freedoms.
19. See David P. McAllister, ‘Music’. In Encyclopedia of Cultural Anthropology, vol 3, David Levinson and Melvin Ember (eds), New York: Henry Holt, 1996, pp. 825-826.
20. Leonard B. Meyer, ‘A Universe of Universals’. The Journal of Musicology 16, no. 1 (1998): 3-25, p. 5, footnote 6.
21. Mantle Hood, ‘Universal Attributes of Music’. The World of Music 19, nos. 1-2 (1977): 63-69; p. 66.
22. See Bruno Nettl, The Study of Ethnomusicology: Twenty-Nine Issues and Concepts, Urbana: Illinois Univ. Press, 1983, p. 40.
23. See Daniel Hirst and Albert Di Cristo, ‘A Survey of Intonation Systems’. In Intonation Systems: A Survey of Twenty Languages, Daniel Hirst and Albert Di Cristo (eds), Cambridge: Cambridge Univ. Press, 1998, p. 1.
24. Dissanayake, ‘Antecedents of the Temporal Arts’, p. 404.