In the course of its history ethnomusicology has come to include a whole series of different areas of scholarship. In the classical sense of the word it is generally understood to mean the study of the music of illiterate peoples and social groups which pass down their music orally. It encompasses all the musical expressions of the so-called primitive races, together with what we in the Western hemisphere term traditional folk music. But, from its very beginning, ethnomusicology has also included the study of the music of the non-European high cultures. However, its scope is not limited merely to music in its narrower sense. It also takes in dance, as well as other quasi-musical means of communication such as by drum language. Some of its proponents see ethnomusicology as a branch of musicology, while other scholars lay particular stress not only on the question of musical forms, but also on the view of music as a form of human action within a particular socio-cultural context.

The widening spread of communications across the whole globe, progress in the field of electronics and the boom in the media, particularly since the Second World War, have brought a wider crossfertilization of various cultures as well as the growth of entirely new sub-cultural areas. While one group of ethnomusicologists might feel particularly bound during such a period of change to seek so-called 'authentic, traditional music', another group is conducting research into the products of acculturation or into the types of music common to much broader population bands. It is this latter group of researchers which is beginning to leave behind the traditional areas of music research and more and more is making the focal point of its increasingly sociologically rooted studies city-based musicians and their reciprocal relationship with 'hits', pop music, classical music and 'genuine' folk music.

The different routes being taken in research might be seen as a quadripolar field of tension between tradition and progress, musicology and anthropology or sociology. Running alongside, above and even right through them, there is also an area of research which is essentially rooted in psychomusicology and psychoacoustics and which seeks to investigate the psychosomatic effects of music. The reason why this type of research and its practical applications, such as musical therapy, are expanding so rapidly· lies above all in the speed with which the relevant electronic apparatus for acoustic analysis and synthesis is being developed.

All areas of ethnomusicology, however, share as their common base, and regard as by far and away their most important source, the sound recording. It is easy to explain the reasons for its dominant role. In the case of historical musicology, dedicated as it is to the study of the musical products of the West which have been handed down in written form, source material comes in the shape of musical notation of compositions which can be visually studied and, therefore, analysed and described. In the case of music passed down by oral tradition, however, this written, visual record has first to be made before analysis and description become possible. In the past, trained scholars might well have been able to produce extraordinary transcriptions of music directly after hearing it, but it was only with the invention of sound recording techniques that a medium became available with which scholars were offered an opportunity to examine the music in detail by the means of a play-back after the event. Just how essential sound recording was to become, was demonstrated by J.A. Ellis' studies of various musical scales which are generally regarded as the beginning of ethnomusicology. He showed that the Western tonal system was definitely not a 'natural' one, let along the only valid one, but rather that the vast majority of non-European tonal systems differed enormously from it and from each other.

Thus the European system of notation which serves perfectly for performing artists familiar with our culture, was suddenly shown to be inadequate for the detailed work of transcription of musics of foreign origin. Despite certain adaptations, this drawback continues to exist. Consequently, aural transcriptions ceased to be of central importance to many scholars who, instead, gave greater credence to automatic transcriptions (see also section 7) as a basis for objective analysis.

But sound recording provides more than just the basis for the transcription of a melody. It provides a more or less accurate picture of all the physical phenomena which come under the heading of music, and reproduces music in all its complexity of melody, harmony, rhythm and timbre. It has become, therefore, the basis of a musicology which goes well beyond the study of melody and it is interesting to observe how Western historical musicology, which for so long had used notation as its basic source, has only recently discovered the value of sound documentation and has begun to conduct systematic studies of interpretation.

The history of ethnomusicology is, therefore, inseparably bound up with that of sound recording and sound archives.

In 1890 the anthropologist Jesse Walter Fewkes pioneered the use of the phonograph when he recorded songs of the Passamaquoddy Indians. In Europe, the Hungarian Béla Vikár is regarded as its pioneer for his work in recording traditional folk music and dialects. With the opening in 1899 of the Phonogramrnarchiv (Phonographic Archive) of the Academy of Sciences in Vienna, the first scientific sound archive in the world was established, having as its aim the systematic collection of this new type of source material by the production, acquisition and preservation of sound recordings for all areas of scholarship. This was followed in 1900 by the Société d' Anthropologie in Paris; in 1900-05 by the Phonogrammarchiv in Berlin, whose first director was the ethnomusicologist Erich M. von Hornbostel; and in 1902-03 by a sound archive in Leningrad. Alongside these archives, collections of sound recordings were built up in museums and libraries and later by radio stations which, especially in England and Japan, paid particular attention to traditional folk music.

While in the early days phonographs were used for field recordings, from the 1920s onwards the gramophone also came to be used. But it was not until the 1950s and the development of transistorised tape recorders that devices were introduced which made phonographic field research possible on a grand scale.

Throughout Eastern Europe the systematic phonographic documentation of each nation's musical wealth has been conducted on an enormous scale and in most of these countries the total number of ethnomusicological recordings exceeds the 100,000 mark. In the West, in the meantime, many regional research institutions were directing their efforts towards recording the musical folklore of their countries. Again, it was the great centres of ethnomusicology at Berlin, Vienna, Paris, Amsterdam, Washington, Bloomington and Los Angeles, to name but a few, which paid particular attention to recording the musical cultures of non-European countries. Happily, these have been joined in recent times by a growing number of research centres in African and Asian countries. Encouraged not least by UNESCO policy, with programmes such as The Ten Year Plan for the Preservation and Promotion of the Performing Arts and Music in Africa and Asia, African and Asian ethnomusicologists are increasingly receiving the support of their governments which, more than ever, are taking an interest in the study and preservation of their national cultural wealth as an essential component of the heritage of the whole of humanity.
 

Within the space of this short survey it is not possible to deal with the question of archival programmes for every method of approach and every school of research in ethnomusicology. However, a few observations on the principal possibilities may usefully serve for illustrative purposes.

Particularly at a time of rapid cultural changes, more and more research should be given over to traditions which are threatened by these changes and may be in danger of disappearing altogether. In cases where musical styles or musical instruments are in danger of dying out because of the age of performers or because new generations have adopted different styles, even a researcher who takes an unromantic view of life and does not pass qualitative judgements on cultural changes, will be formulating recording programmes which will be devoted to at least the documentary preservation of these aspects of music. Where there is no list of priorities for such urgent programmes, one can be drawn up based on a general survey. If, on the basis of a thorough survey, a list of priorities is established, then it is essential that a balance be retained between personal research interests and inherent necessities as seen from the viewpoint of the culture in question. Ethnocentric preconceptions and western concepts of art should not be allowed to dominate one's views, and care should be taken to ensure that the wishes and self respect of the individual proponents of tradition are given due regard. Minorities should be given particular attention. The smaller their numbers, the greater consideration they should receive, particularly when they have been subject to pressures of acculturation and integration.

As the organizational pre-requisites of recording vary enormously from case to case, it is possible here to give no more than a few general guidelines. Where it is a case of programmes being conducted well away from the site of the research institution itself, it will be necessary to remain in the field for some time, although just how long depends on how well acquainted the researcher is with his informants or with the subject of his research. The researcher who is willing to make do with a recording specially arranged for his purposes, with its consequent lack of immediacy and its lower degree of authenticity, can generally put together appreciably more material in a shorter space of time than one who insists on contextual authenticity. Despite the lower level of productivity, however, it is easy to understand why the latter approach is nonetheless becoming increasingly popular. For, unlike Western concert music, most ethnic music simply cannot continually be restaged without losing certain ritual, psychic or atmospheric conditions which are pre-requisites for the creation of genuine music.

If a recording programme is designed to produce comprehensive documentation, then a twelve month stay in the field, or at least several extended visits at different times of the year, will be necessary if all the festivities which take place in the course of a year are to be taken in.

By contrast, programmes devoted to the study of the music of the present in its social context and acculturated state require a different approach. Intricately bound up with this approach is the demand that recordings of music should be absolutely live, so that one captures not only the music itself but the whole occasion i.e. its atmosphere and every interaction between the performers and the audience. With this form of sound documentation the foundation is laid for a type of musicological research going beyond the mere examination of melodies. If such research studies are to be established, one should first of all gain a thorough familiarity with the culture or region in question by examining previous studies. As a practical exercise in familiarizing himself with this approach fieldwork within the researcher's own home area could also be useful. A network of reliable informants who can provide advance notice of events such as weddings, funerals and so on is indispensable. A rapid form of transport, which generally means a car, and a willingness on the part of the researcher to give freely of his spare time, are just two of the basic pre-requisites for this sort of programme. Although the researcher has to be prepared for any chance eventuality, his programme still has to fit into a framework which will provide a regional, stylistic and contextual balance in much the same way as a well-planned sampling does.

Whatever the subject of research, it should be remembered that sound recording provides an ideal basis for diachronic studies. When several recordings are to be made of one informant, one subject or of similar situations, the (relative) objectivity of sound recording allows one to draw certain valid conclusions about stability and change; that is about the dynamic nature of culture. This facility to record the same subject a number of times over also helps us to distinguish more easily between chance and normative occurrences. One can trace both the development of an individual musician and his style, as well as the development of the whole musical context in which a tradition has grown up. Just how long one should leave between one recording and the next depends on the subject under investigation, but one should never leave too long a gap. Similarly, it is impossible to lay down a general rule on the number of recordings which should be made of one subject; this may be expected to vary between one study and another.

We have already touched upon the main difficulties of choosing a subject for recording when we discussed how research interests can sometimes conflict with the inherent viewpoint of the culture. In the same way, different levels of awareness and understanding can dominate in these cultures, with the result that political or social authorities, for example, might recommend a different viewpoint from that of the musician. The researcher might well find himself in the middle of a potential conflict situation, when faced with the problem of choosing subjects or occasions for recording. Each recording situation, therefore, demands a measure of skill in trying to avoid conflicts of interest or in balancing them out among the various groups within the society which is being examined, if a suitable order of priorities is to be established for research programmes.

The physical-technical definition of sound recording is dealt with briefly in the technical chapter of this publication. Methodologically, a recording of a musical event cannot be seen separately from the role of the researcher.

By his very presence the person recording exerts an influence on his subject, regardless of whether he sets the recording up (i.e. is responsible for the production taking place at all) or whether he is merely attending a musical performance which would have taken place anyway. Apart from the general behaviour of the person making a recording, the deciding factor in the question of the degree to which a recording is influenced can lie in the technical circumstances under which it is produced. In the classical area of ethnomusicology the 'subject' is likely to be a musician bound by the society in which he lives and so, as a rule, it is better to avoid recording in a studio (which might well be free from extraneous noises, but which places the musician in unfamiliar surroundings). This rule may be ignored in cases where the informants are quite used to speaking into a microphone or in making demonstration recordings of instruments. In general, however, recording should be done on location; that is in surroundings and in an atmosphere with which the person being recorded is familiar.

All location recording should begin as unobtrusively and with as little fuss as possible, keeping the number of those involved in the recording to a minimum. Just as the use of a radio car can arouse false expectations, an official government vehicle can cause unnecessary anxiety. Also, no more than two people at once should normally produce the recording. The extent to which measures taken to improve technical quality (for example whether to record in a quiet, seldom used room in the house, or to remove ticking clocks and so on) might disturb the openness of the musician and in turn affect the recording has to be decided on the merits of each case. In the same way one should avoid too much technical gadgetry or drawn out tests for correct microphone positions. However, one should never, at any cost, interrupt a recording because of actual or suspected faults, as quite commonly happens in commercial recordings.

The reduction in technical gadgetry is directly related to the fact that, until recently, in the field of ethnomusicology most recordings were produced in mono, using only one microphone. Increasingly, however, stereo equipment is being used for various recording techniques. Apart from the standard spaced and coincident microphone techniques (or AB-and XY stereophony), binaural techniques are coming to be used more and more. In this technique, similar to dummy head recording, the microphone distance is approximately 17 cm, corresponding to the average distance between the human ears. This has the advantage that, relatively speaking, microphone locations are not critical and quite acceptable recordings can be achieved even from 'impossible' positions which inevitably occur in some live recording session. This technique has even stood the test of concerts in which the music is performed with electronic stage amplification and the recording has to convey the sound as it is heard by the audience. This method of recording is especially useful because it can be carried out by just one person and requires scarcely any more complex equipment than for a mono recording. Of course, recordings of this sort can only really be fully appreciated by listening to them on headphones.

In cases where one is already quite familiar with a particular culture or its representatives, more intricate techniques might be used in order to carry out some kinds of investigations, such as using multitrack tapes to improve the clarity of individual voices for transcription purposes. Such intricacies, however, require careful preparation as they constitute a serious incursion into the actual musical situation and should, therefore, be kept to an absolute minimum. In contrast to practices in commercial recording, all aesthetic alterations to the original sound, such as filtering, should be avoided.

While getting something down on tape is important, keeping written and, if need be, photographic documentation is also an essential aspect of recording. The written record should include the date and place of the recording, details of the subject, descriptions of the content of the recording and of the circumstances which led to its being made, as well as the technical data of the recording itself. Playing a recording through as soon as possible after it has been made is a highly recommended course to follow, because then the details of the event concerned can still be remembered clearly and the opportunity remains to interview the informants again if necessary. It is for this reason the tape recorder is being used more and more often and because it has proved to be extremely valuable if informants are also allowed to give their personal views on music in the form of an interview.

If the coverage of a subject or programme is to be comprehensive, then it is often necessary to exploit sources other than the recording activities of one's own institution in order to acquire appropriate acoustic material. In this respect one should consider co-operating with outside researchers who have similar research interests and who, in return for help with practical aspects and equipment, will supply material to the archive. It would make good sense for those institutions which are responsible for preserving the cultural heritage of particular regions or countries to follow such a policy and to do their utmost to collect historical material as well. For this reason they should make contact with other institutions and private collectors, as well as with radio stations, record producers and foreign sound archives. Newly established sound archives in Third World countries have already begun to enhance the historical content of their national collections by systematically collecting previously recorded material from these other sources.

Sound recordings rarely come in such a form that they can immediately be stored in the recommended fashion, and so the production of archival or security copies lies at the heart of the work of archival processing. While this work is being done recordings can also be re-examined to establish whether the whole of the raw material or just a selection should be kept. The watchword in this process should always be to alter the actual recording as little as possible. Recordings should retain their original length; for instance, if several songs were performed and recorded one after another, then the whole recording should be kept as an integral unit. Comments and reactions at the beginning and end of a recording can often be very revealing and, therefore, should also be copied. One should always avoid any element of aesthetic filtering when transferring a recording although, of course, a written note should be kept of any of the technical shortcomings of the original.

The best method of keeping tapes (an important point for retrieval purposes) has proved to be that of copying the recordings onto reasonably sized spools (minimum 18 cm diameter) in the sequence in which they were recorded or accessioned and to use signals on the tape to distinguish individual recordings from one another (see also the technical chapter of this publication). Accessioning should follow a single numerical sequence, because even small institutions will find it difficult in the long run to continue classifying recordings according to regions, groups, informants or content. This method also mixes up recordings of different technical origins, making technical quality control impossible. What is more, if one classifies material according to cultural theories, these arrangements may often become obsolete.

In contrast to libraries, where catalogues are just used to locate fairly standard and well described published items, academic research archives are obliged to employ more complex methods of registering the contents of their holdings. For this the best method to use is a form resembling a questionnaire, giving details of the production of the recording, its content, what textual and musical transcriptions have been made and what illustrative or photographic documentation accompanies it. In multi-disciplinary archives these forms need to be relatively open-ended but, in institutions bound by limits of subject matter and region, very detailed lists of headings can be worked out.

The question of what form indices should take is something which has to be decided according to the scope and content of the archive. In the field of ethnomusicology, it is the performing musician who normally takes on the role of composer in Western classical music and so he will be the focal point of any finding aid. Performers should be classified according to ethnic groups, but cross-references should always be made according to neutral geographical and political headings. In regional archives, specializing in traditional music, the range of such indices should go down to the level of individual villages and parts of villages. There is also every sense in indexing collections according to musical instruments, musical styles and genres, and perhaps titles in the case of instrumental pieces and songs. Difficulties arise when attempts are made to index collections on the basis of the socio-cultural context of the music, because the categories in each case will be determined by the culture in question and the outlook of the researcher, both of which factors may change with time. The best index, therefore, will be one which does not over-emphasise detail and, in the absence of an anthropological thesaurus, one whose content is not too closely bound to one school of research.

A standardized system of classification along the lines of that used by librarians for indexing books is probably a long way off yet, and for the moment the problems of cataloguing edited source material (such as musical records) alongside books, using the same standard scheme for both, are enormous. Ethnographic acoustic material, which in part calls for completely different classification criteria, is likely to resist any binding form of standardization for a long time to come. Of course, as far as the user is concerned, the most convenient form of access to information is likely to remain the computer. Computers, however, only really serve any purpose if they are to be used for dealing with a massive complex of subject matter and if rapid information retrieval is genuinely necessary, in which circumstances utility and expense are balanced against each other.

Even though the pioneers of ethnomusicology were skilled physicists, chemists and lawyers, nowadays it is essential to employ academically trained staff to carry out research, recording and archival work. Employing enthusiasts and self-taught people as assistants can often pay dividends, but the programme planners and policy-makers should always be people trained and qualified in ethnomusicology. As indicated in the technical chapter it is extremely important to employ technicians, even in relatively small research units involved with the collection of acoustic material and its preservation for future generations. While in a small unit the technician can sometimes be employed on routine copying work, in a larger operation he will probably require a technical assistant for this. But recording, particularly location recording, should be left to academically trained personnel, if the need to keep to a minimum the external influences exerted on a recorded event is to be achieved in practice, and if technology is always to be made to serve research responsibly and conclusively. Even in producing archival copies it is best to use academic staff and not a technician. Not only is this labour-saving, it can also provide a further opportunity to select and document the material. It is especially important to follow these principles if the archive's technical staff have been recruited from radio organizations; their broadcasting experience might be used to persuade the often impressionable archivist that certain commercial studio practices are 'professional'. The job description for an archive technician, therefore, should always specify a service technician rather than a sound technician or sound engineer.

A professional collector/archivist should be capable of recording, accessioning and cataloguing about fifty hours of ethnomusicological material a year. If he is just accessioning other people's material, rather than recording it himself, he should be able to deal with up to a hundred hours of recordings. The work capacity can be increased, although not necessarily at an even ratio, if routine jobs are delegated to a research assistant. Archivists might find it quite useful to share one such assistant between two of them. Experience has also shown that for every three or four professional archivists, there should be one administrative assistant.

Finally, it should be noted that from a practical point of view it can be very helpful if actual members of the group being studied are used to conduct research. It is particularly worthwhile considering this when selecting staff for research programmes relating to minority groups.

Because their range is so large, conventional methods of evaluating ethnomusicological material cannot be covered by this study. However, it is worthwhile devoting some space to the subject of automated methods of evaluation which are becoming more and more feasible with the developments being made in the field of electronics. The most prominent of these methods are melography (melodic transcription) and sonography (acoustic analysis using a sonograph).

The melograph, which comes in many varied forms, gives a continuous transcription of a melody in diagrammatic form, showing the pitch and length of each note. Apart from the basic difficulty of converting this diagrammatic read-out back into a format in which it can be widely understood, it also presents other major obstacles which prevent its wider application. It requires, for instance, recordings having an excellent signal-to-noise ratio, which is not always available. It also registers interference as part of the signal, it records some notes as being an octave higher than they actually are and, finally, it cannot be used for polyphonic music. Before the automatic read-out can be interpreted, therefore, it requires close critical examination.

While machines for the automatic transcription of two or multi-part music have not yet passed the prototype stage, the sonograph is already being widely used. This device provides an analysis of frequency against time, albeit only in relatively short sections of music lasting no more than a few seconds, but it provides a precise graph of individual harmonics. It is eminently suitable for delineating acoustic phenomena and, although restricted by the fact that it is laborious to operate, it can at least be used, for short pieces, for transcribing polyphonic music. It needs to be remembered, of course, that converting its read-out into conventional musical notation is extremely complicated. The applications of the sonograph go well beyond musicology and it has become a standard piece of equipment in phonetic and bio-acoustic analysis.

Finally, mention should be made of the various digital analytical and synthetic processes which are now emerging and which are being developed in various research projects. These techniques will bring decisive advances in the automation of analysis, even within the next ten years. Besides frequently requiring huge financial outlays, these methods will need intensive practical research and continual critical analysis. Only when we are prepared to provide the means of meeting these conditions can a proper appraisal of these techniques be made. This would also be an opportune point at which to establish from the outset a broad, even inter-disciplinary base for co-operation. It is essential for the success of audio-analytical programmes to insist on high technical standards or at least a thorough documentation of whatever standards are in operation, if one is to avoid the risk of interpreting any imperfections which occur in a recording as part of the original signals.

In conclusion we should take a brief look at some of the legal problems involved in collecting, accessioning and distributing ethnomusicological material.

Generally, contracts or at least agreements are essential at the following levels:

between the musician and the person making the recording;

between the person making the recording and the archive;

between the archive and other institutions or individuals to whom material is to be given.

In every case the institution should endeavour to reach an agreement which allows it the maximum possible control over the material it acquires and, as far as practicable, which precludes any possibility of the material being abused. The first problems are likely to arise in reaching agreement with musicians. Artists working in Western cultures will be quite sympathetic in this respect, but in the case of illiterate peoples it will inevitably cause some measure of confusion and arouse false expectations, which in turn will eventually detract from the genuine and honest nature of the informant and, therefore, of the end product of the recording. In all such cases it is best to remunerate the musician in accordance with accepted local norms thus effectively purchasing the rights over the recording's usage. In the field of European folk music it is often customary to pay the musician nothing if the recording is being made for academic use, on the understanding that it will not be used for commercial purposes. In allowing access to an archive, one should take care to remember that composers have a penchant for taking traditional music as the basis for adaptions, which then go on to earn them money, while the originator of the music receives nothing. It is the duty of institutions with holdings of this sort to give full consideration to the 1nterests of all parties concerned. In this they are faced with the problem of copyright which, being based on Western concepts of culture, is often difficult to apply to music in the ethnological field.