Following on from our successful series 'The Psychology Of Music', we invited Ian Cross of the Centre for Arts and Related Studies at the City University, London, to submit an article introducing the science and its significance to the way in which we compose, play, perceive and understand the various different musics that currently exist.

Music is a complex, difficult-to-define phenomenon. Perhaps the best available definition of it is that music is whatever activity involving sound that a culture/community/subculture/elite/A&R man judges to be music. Such a definition is not as flippant as it may seem; globally, a wider range of musics than had ever been previously acknowledged to exist is simultaneously available. The questions to be asked in an attempt to comprehend all these musics have to be: Are there any factors common to all of them? And if there are, how can we best investigate and describe them?

Two instantly obvious factors are people and sound. Sound is measurable, and human physiology doesn't vary that much from culture to culture. So some investigation of how we hear and how we produce sound might give us a third common factor: the physiological constraints on what we can hear and what we can do to produce sound. Going further, the processes of human thought seem to share many common features across communities and cultures. So, joining together these human factors, our understanding of human perception, cognition (thought processes) and performance are of central importance in understanding our multiplicity of musics.

If we can listen to a sound purely as a sound, without necessarily associating it with an object producing it, we find few qualities of the sound as easy to describe as shape, size and colour are in seeing an object; when we see an object, we feel that we can see certain properties in it which are physically measurable. This doesn't seem to be the case with sound.

But what about pitch, or loudness? Well, pitch can be related roughly to the frequency of a sound, that is, the number of times the changes in direction of air pressure gradient which constitute the sound occur per second. What if the sound only lasts a millisecond? What if there's more than one frequency present in the sound? And, though loudness is related roughly to the amplitude of the changes in pressure constituting the sound, a low-frequency sound will seem softer than a mid-frequency sound of the same amplitude. Loudness also depends on context; a brick dropped in the hush after a solo violin concert just before the applause will obviously seem louder than a brick dropped in a hush (?) following Einsturzende Neubauten's farewell chord. So the relationship between what we hear and the physical correlates of what we hear is not particularly simple.

"Most musics can be said to consist of a balance between change and stasis."

Many of these relationships have, however, been more or less worked out. We can now measure the frequencies and amplitudes present in a given sound, perform some calculations, and tell which pitch (if any) the sound is likely to be heard as having. Such knowledge is important and useful; unfortunately, it doesn't enable us to answer many questions about music. The reason for this is that most research aimed at finding out how we relate the subjective attributes of pitch and loudness to the physical attributes of frequency and amplitude uses a very small range of sounds in its experiments. The method used usually involves presenting a listener with two or three sounds at a time, varying the physical relationships between the sounds and measuring the listener's responses to the changes in the physical, acoustic relationships (frequency etc.) between the sounds. This psychophysical or psychoacoustic approach to sound tells us how we hear, but doesn't tell us much about how we hear music, as almost all musics comprise more than just groups of two or three different sounds.

So another possible factor common to different musics must be considered: structure. Most musics can be said to consist of a balance between change and stasis; while one element (eg. rhythm) stays the same, a second (eg. pitch range) may change, and vice versa. Alternatively, all elements of a piece of music may change simultaneously, or all may stay constant. Leading on from this is the idea of a piece of music having a clearly defined beginning, middle and end. However, not all pieces can be said to have clearly defined shapes in time; although a classical symphony or a three-minute single may have self-evident beginnings and ends, other musical forms such as a traditional Javanese gamelan piece or a twelve-bar blues can always occur 'one more time'. This leads to the idea of repetition, which, unlike closure, or definite structural ending, does seem common to almost all musics, though occurring in a variety of ways. Repetition is easily heard as the essence of gamelan music and of the music of recent Western composers such as Philip Glass or Steve Reich, whereas a piece of serial, atonal music may be heard as always changing and never repeating (although consisting almost entirely of 'repetition'). The 'repetitions' in serial music can be thought of as transformations rather than simple repetitions: a transformation of musical material being at its simplest a varied repetition. Transformations may vary the original material (eg. a tune) so much that it becomes unrecognisable, by, for instance, slowing it down or playing it backwards.

"Someone familiar only with Verdi operas may not be able to make sense of a North Indian raga... they may therefore dismiss it as not being music."

The ideas of change, closure, repetition and transformation may seem too abstract or too simplistic to help us understand our musical behaviour. They do, however, help us to describe and define aspects of music which can't be defined in purely physical terms. Perhaps the most explicit example of this is in relation to melody (common to many, but not all, musics). A physical description of a melody could only exist in terms of a list of frequencies, frequency ratios, amplitudes and durations in time. The physical descriptions of performances of 'My Way' by Frank Sinatra and Sid Vicious would differ in almost all respects. Yet the two performances would probably be perceivable as being of the same song, even if the words are not considered. What is unaccountable for in physical or even psychophysical terms is not particularly that two different series of events (frequencies, etc.) should be heard as similar, but that a series of events should have (or be heard as having) some specific identity as a 'melody' in the first place. This is only accountable for in terms of broad human capacities to integrate (connect) events in time and to comprehend structure - deriving or imposing it - in such series of events. One framework within which these capacities can be investigated is contemporary cognitive psychology.

The cognitive psychology of music is probably best defined not in terms of 'what do we know about how we experience music?' but in terms of 'how do we know about how we experience music?' Although the psychology of music is to some extent a body of knowledge about how people experience music, that body of knowledge is constantly changing under the pressure of new ideas and experimental results. The psychology of music consists primarily of models of how we interact with music and of methods for testing the validity of these models. Current psychological models of how we comprehend and act on our surroundings tend to view the process of comprehension as interactive; that is, what we perceive is not determined solely by what exists in our surroundings but also by our ability to impose or extract structure and information selectively from the surrounding world. These models have particular implications for our experiences in listening to and performing music.

"Cognitive psychology is not the only way in which to investigate our experiences of music; it doesn't explain why some music is judged 'better' than other music."

One such simple implication is that music will make sense according to how we derive structure from it, which could be considered dependent on our previous experience of music based on similar principles. So someone familiar only with Verdi operas may not be able to make sense of a North Indian raga (or vice versa); they may therefore dismiss it as not being music. In this case the components which a culturally-skilled listener would identify as occurring in music and as constituting musical structure would be unavailable to a listener unfamiliar with that culture, so that musical structure would seem incoherent or even absent. Similarly, a classically-trained violinst attempting to play traditional fiddle music might find that, although all the notes were there in the right order, their performance didn't sound quite right. What is important in classical violin performance (quality of tone, etc.) might not be important or appropriate to fiddle music. So although the violinist might be able to make some sense of the fiddle music, it may be distorted to fit the violinist's own musical awareness.

So musical structure can be thought of as dependent on at least two factors; the physical phenomena (sounds) constituting the music, and the listener's or performer's ability to comprehend the structure of the sound. As both musics and people vary widely, the terms in which music and people's ability to comprehend it are defined psychologically have to be very general. Hence the use of such terms as 'change', 'closure', 'transformation' etc., which are applicable to a wide a range of musics, and terms such as 'memory' and 'attention', which can be seen to underlie a wide range of human cognitive abilities. Because of this variation between people, experiments on, say, perception have to test numbers of people and have to express their results as tendencies, eg. that most people, most of the time, would perceive such-and-such a piece in such-and-such a way.

In spite of, or perhaps because of, this generalisation, the study of musical cognition has revealed a range of specific principles which could apply in many musical perceptions and performances. For example, in describing the perception of a succession of notes as a melody we can refer to the relation of the notes to an overlearned schema (such as a scale), to the likelihood of successive notes being perceived as connected or 'streamed' (decided by the relative separation in the time and pitch domains), and to the resultant melodic contour. Each of these principles (reliance on overlearned schema, stream and contour formation) seem to be fairly general and can themselves be referred to even broader cognitive principles such as category formation and selective attention. So ideally, purely musical concepts such as scale, key, metre and harmony have to be broken down into very general terms before we can either explore them or begin to understand them in psychological terms.

Of course, cognitive psychology is not the only (and frequently not the best) way in which to investigate our experiences of music. Cognitive psychology doesn't tell us why we make or listen to music; it hints at but ultimately doesn't explain why different cultures produce different musics. But then, it's not intended to. To a large extent, psychology has its own appropriate domain of enquiry. For a psychological understanding of a particular aspect of music to come about, that aspect of music must be expressed in terms which psychological models and methods can be used to investigated anthropologically or sociologically: others, such as musical value judgements, might be better approached via aesthetics or musicology.

However, psychology does have a bearing on all these questions, since at the centre of psychology is an attempt to understand the consciousness which asks the questions. It is not so much the question itself as the form in which it is asked which makes it amenable to psychological investigation. Currently, the study of psychology of music is still in its infancy. Although the human experience of music has been the subject of analysis for over two thousand years, it is only relatively recently that the aim of that analysis has become the understanding of human experience of music which currently co-exist, and are constantly coming into existence. For one of the constant features of music and of our experience of it is change: to misquote an all-too-well-known politician, 'there is always an alternative'.