"The chasm between music theorizing and music making has not narrowed at all since Plato."
Sloboda (1982) Exeter University

In fact, it has probably widened as a result of the so-called 'prescriptive' view of music theory. The purpose of a proper music theory should be to enable people to think coherently about musical problems. An important sphere of research centres around the intersection between the physical world of music and our perception, cognitive processing, and mental organising of ordered sounds. This sort of research leads us to a music theory which can help to explain musical phenomena without dictating what must be done to obtain them.

The last decade, in particular, has witnessed a sudden interest in the empirical study of music on the part of both scientists and musicians. Along with recent advances in computer technology there has been a flourish of research into auditory shape analysis, attention in music, the organisation of memory for musical information, and the skills used in an actual performance, and so on. As a result of these sparks of interest on the part of psychologists and musicians, there has been a spontaneous expansion of study in the broad area of information processing in music.

Music can simply be understood at all its levels of complexity in terms of grouping: grouping of sound frequencies to form specific notes and textures, grouping of these notes to produce chords of clustered sound, grouping of individual tones or chords along the length of music to form phrases and rhythmic clusters. From the micro to the macro-level, music is based upon this fundamental principle. The complexities which arise in music are not from other strategies but from interactions between pitch, rhythm, timbre, texture and so on. A music theory that does not account for these, and the perceptual processes which are required to analyse them, cannot possibly claim validity in real-world music making. And, after all, isn't this what we as practical musicians are interested in?

Over the last decade, many studies have been done in several of the levels from which musical processing can be approached. Recent advances (relatively speaking) in computer technology have enabled scientists to generate and analyse complex sound stimuli with versatility and precision. Thus, it has become possible to explore such issues as auditory shape analysis, attention in music, the organisation of memory for musical information, to mention just a few. This same technological development has led composers to experiment with the computer as a compositional tool. By so doing, they have raised many questions in perceptual and cognitive psychology that are of both practical interest to composers and theoretical interest to psychologists. As a result of this flourish of interest on both the part of scientists and musicians, there is currently a rapid expansion of work in the area of musical processing, and perhaps more importantly, symbiosis between scientists and musicians is a rapidly growing phenomenon.

Although the fundamental question of perception is essential to the full understanding of music theory, what is most important to the skilled instrumentalist is the knowledge of how music is actually performed in the real world, as opposed to the laboratory. It is because of the lack of technology, specialised measurement techniques, and experimental control that the study of music performance has lagged many years behind the study of more basic questions.

Our purpose here is to draw together all the diverse and scattered knowledge that has accumulated, specifically over the last decade, on musical processing. We will be looking not only at music itself, but also how it is processed by the listener and the performer, the effects of different types of music on our moods and behaviour; musical preference and appreciation; the neuroanatomy of musical function, and brain disorders which result in musical abnormalities; mental plans for action and performance; the role of feedback in performance; musical ability; illusions in music; rhythm; and psychoacoustics.

The overall aim is to produce a 'new theory of music' which overcomes the double barrier that currently exists between musical practice and music theory, and between music theory and those other fields, such as computing, which touch music at some point or other. There has always been a problem of communication; the chasm between thinking about it and doing it. It is one thing to know when and how and where to pass the thumb in fingering, say, the C-major scale on a keyboard, and altogether another thing to do it smoothly at moderate and fast tempos.

The 'traditional' theory of music has been going on for so long, 2000 or 3000 years, that it is no surprise that it is a very confusing body of knowledge. The history of music theory can be seen as a special pleading for this or that musical style. That must be changed. What is needed is a music theory that is bound by style, that is valid for world music, not just its European, Chinese or Indonesian varieties.

Music Performance

Let's first look at four central aspects of music performance which relate to the skilled musician. How do we best characterise what it is that someone knows when he or she performs a piece of music? How does that person acquire that knowledge? How does the performer use perceptual feedback to control his or her performance? And, lastly, how is music performance affected by social and situational factors?

Whenever someone performs a piece of music, he or she is induced in a process of translating a mental plan or representation into action. Everyone has those plans to a certain degree! They allow us to hum tunes as we go about our daily routines. But what kinds of plans are required to execute these actions? A simple possibility would be a list of actions in a specific sequence and of specific durations. However, this simplistic approach cannot account for the differences in performance which are found on different occasions using the same equipment. Human beings do not perform music in exactly the same way on each occasion. In fact, the variation which can be heard can be classified as unsystematic and uncontrolled, or systematic and controlled variations, the latter, of course, being effected by the performer, for instance, many people are perfectly able to sing a song from any starting point (within their vocal range) and comply with requests to sing louder or faster. Furthermore, some people who know a tune very well through singing it, can, if they are competent enough, transfer the tone successfully to an instrument, or if literate, write it down in notation.

These observations suggest that the 'musical' plan must be somewhat more abstract than a list of defined actions, and must be capable of varying infinitely in many dimensions. It has been suggested that a performer's plan may be a listing of items in which the duration, pitch, and intensity of each note is specified relative to other notes. Before the plan can be executed as action, absolute values for each pitch, duration and intensity would have to be specified, and the appropriate motor actions would subsequently be determined.

In a literal kind of transcription (for that is what playing from music, or a mental representation of it is all about), a fluent response sequence may be derived from a representation that contains programs with knowledge about how to translate notation into physical response. Motor programs are mechanisms 'within the nervous system' that enable co-ordination of a movement sequence, and these may become more complex and extensive with experience and practice. The programming can bethought of as becoming rather complex as more lines are added in modular fashion to accommodate further skill acquisition.

There is a further distinction between 'planning' and action, and 'programming' it: plans can be carried around in the head of particular pieces or anything else for that matter, whereas programs can be thought of as mechanisms for interpreting plans and executing them in a particular mode. If the expressive mode is the voice, then the plan is executed via the motor programs which organise the movements of the vocal chords. If the mode is the keyboard, then the motor programs direct the placing of the fingers and make the ballistic movements required to 'land' on the correct notes. Because the plans themselves may be extensive, the programs may interpret them in a succession of fragments, moving along the plan as if on a conveyor belt. However, in sight-reading music from a score, it is the score which is the basic plan, and decisions of expression and dynamics must be improvised during performance. Such a performance may lack the refinement of expression that develops with practice, but it can be reasonably fluent so long as the player can look ahead of the notes he is playing. We will discuss this particular point later as it has been well studied using electronic keyboards and computers. Lastly, in a freely improvised performance there is little or no distinction between planning and programming as the plan is spontaneously created and there is a continuous translation between conception and motor response.

The performance variations which we have so far talked about share a common property, that of 'context-free' application. In other words, each variation applies in the same way to each note. For instance, when a tune is played faster, each note is shortened by the same proportion. Or if it is played in a different key, each note is raised or lowered by the same proportion. However, not all variations in music are like this, some are called 'context-sensitive'. They are often called 'expressive' and include such techniques as rubato, variations in attack and decay, fluctuations in pitch and timbre. These are the variations which prevent performances from appearing 'dead-pan', and are applied at the discretion of the performer in the light of the current situation. It is these that we will turn to in the next article.