Bruno Nettl (6)

Chapter 5

DESCRIPTION OF MUSICAL COMPOSITIONS

Analysis of music in the American music curriculum denotes mainly the description of the over – all structure of a piece of music, and of the Interrelationship of its various sections. In most cases, indeed, it is the fitting of this structure Into a preconceived mold.

The typical textbook of analysis does not set out to teach the student the principles and procedures used In describing music; rather, it attempts to teach him the characteristics of certain specific forms and then to show how well – or how badly – Individual compositions fit these forms. The fact that relatively few pieces fit the formal outlines which they are supposed to has not deterred the writers of texts and some teachers of "form and analysis" from continuing this approach. The result is that there is little methodology available for analysis of the over – all forms of musical compositions, and even less for descriptions of smaller segments of music such as scale, melody, and rhythm, and hardly any for the description of timbre, dynamics, and tempo. The absence of methodology In this case would seem. to be a great handicap In the study of Western musical culture, In which composers have written about their own methods and Intentions, and in which a body of music theory in writing has long existed to guide the interpreter of music. How much greater a handicap must be the absence of such methodology to the student of music in other cultures, in which preconceived ideas as to the intention of composer or performer are likely to do infinitely more damage than they do in the description of Western music. This chapter does not attempt to provide the methodological foundations which we need. It will attempt to survey the approaches which are evident in ethnomusicological descriptions of music, to evaluate the contrasting work of a few scholars, and to provide some guidelines for the student who wishes to undertake the description of individual pieces of music and of entire musical styles.

We are faced with description of two kinds of musical units: the individual composition, and the corpus of several compositions. Of course the description of the single composition comes first, both logically and methodologically. Before undertaking the discussion of descriptions of bodies of music, we must study the problems of describing individual compositions. And before a scholar can describe, statistically or impressionistically, a body of music, he must describe the individual compositions in that body. To do otherwise would be to invite gross errors and false conclusions. The problem is, of course, that an infinite number of things could be said about a piece of music, and that we would like to restrict our statements to those things which are somehow relevant to discovering the essentials of a style, and to distinguishing it from that of other composers, nations, historical periods, or cultures. We are faced with two processes: first, the inspection and analysis of the material, in which objectivity and reliability are the primary requisites; second, the description of the music, which results from the analysis, and which must, above all, be communicative, and which must make some concessions to the reader's frame of reference.

In recent years few have been so concerned with the problems involved in the description of music as Charles Seeger. His publications (1951, 1953, 1962) have probed the fundamental problems and the underlying assumptions of musical analysis, and while they do not present blueprints for the practical side of our work, they are invaluable as critiques of the thought processes of musicology. Seeger is concerned with the distinction between music as an event and music as part of tradition (1951), and between the knowledge which one may gain from music directly ("music – knowledge") and the knowledge which comes from talking or writing about music ("speech – knowledge"). He shows a healthy degree of pessimism about the possibility of using language to make musical sense: There have, 1 must say, always seemed to be excellent grounds for a hearty distrust of all talking about music. On the other hand there is the inescapable mandate of modem scholarship that there is nothing that cannot be talked about, provided only it be done in the right way. The musicologist is the one who obeys this mandate and tries to find the right way. (Seeger 1953:370).

Seeger summarizes a set of 21 working hypotheses for describing music, hypotheses which seem lofty and remote from the mundane work of counting notes and intervals, but which help to present the musicologist's work In broad perspective. Among these hypotheses are such safeguards of logic as his direction to exploit. the various possible approaches In terms of opposites,, i.e., to make a "structural" as well as a "functional" analysis; to use quantitative as well as qualitative criteria for data.

Seeger (1962) is also concerned about the problem of vocabulary in musical description. While his publication of Concepts and conceptual operating techniques" necessary for describing musical tradition does not Include terms which can be used In musical description itself, it does emphasize the need for clear thinking and for defining the basis and the limits of the work under consideration. It is hard to disagree with Seeger, for it cannot be denied that the terminology of musical description is Inadequate. Based on terms used by composers, and approaching its object from a prescriptive rather than a descriptive view, this terminology has become hopelessly muddled so that communication among musicologists has suffered greatly. And In ethnomusicology, a field In which no prescriptive terminology from cultures under Investigation is to be expected, the use of terms and distinctions based on the thinking of Western practical musicians seems to make even less sense than in the study of Western music history. It is a temptation to throw up one's hands, to say, “never mind describing the music, just let it speak for itself”. And this may be a solution for those who wish to understand individual musical events without reference to their cultural environment, and without reference to the other musical events in the same and in other cultures. But where comparison, and even the possibility of comparative work, are intended, it is necessary to reduce the musical information to a form of communication which is readily understood and in whose terms comparison can be made. Thus Seeger's distinction between "music – knowledge» and "speech – knowledge» may perhaps be applied as follows: Music – knowledge may be sufficient for the understanding of a piece or a body of music in itself; speech-knowledge is necessary for comparison.

Instead of using speech, we might find the use of mathematical symbols a useful possibility, and, to be sure, the use of nonlinguistic symbols, such as letters and graphs, for musical description is widespread and should be encouraged. But mathematics as a way of communicating is not nearly as widely understood as language and, in the same way that music-knowledge must be translated into speech, mathematical symbols might have to be transl!1ted again, making it necessary to resort, as before, to the use of words. Inadequate as language may be for describing a nonlinguistic form of communication such as music, we find that it is still the most promising tool.

The typical textbook in analysis does not intend to give the student a tool whereby he can describe music however he may find it. Rather, it describes a body of music in terms which are applicable especially to it, assuming that the best way to introduce a student to musical structure is to allow him to see description of one kind of music in terms which apply to the style of that music. It seems to be taken for granted, in the best of these books, that different styles will cause the student to evolve different terms and methods. This seems unexceptionable; but students of other musics must be blamed for having used the methods derived from several European styles on the non-Western musics for which they should have derived (presumably from their "music-knowledge" of these musics) terms and methods appropriate to them. It would seem to be the task of the ethnomusicologist to derive a method which is equally applicable to all music.

Three Approaches to Description of Style In the next group of paragraphs we will describe three approaches to the description of music. These three, labeled here "systematic," "intuitive," and "selective," appear in many publications, yet it is often difficult to identify them. The reason for this difficulty is that we cannot always tell, from a published description of style, what approach the author actually took in analyzing, and that combinations of approaches are perhaps the overwhelming majority. The approaches as outlined here are, to an extent, abstractions, but it is useful to distinguish among them as an introductory procedure.

The Systematic Approach. One approach to describing music is to identify all possible, or many, or, for practical purposes, a selected group of aspects of music, and to describe each of these aspects in an individual composition, or in a body of musical composition which, for one reason or another, are assumed to have something in common justifying their description as a unit.

The usual procedure in this method is to divide music into a number of so – called elements. In the teaching of music theory these are, most frequently, melody, rhythm, meter, form, and harmony or polyphony. In musicological studies a less practical but more objectively scholarly arrangement can be made; none is generally agreed on, but the following outline reflects the tendencies found in the majority of descriptions of musical style: Pitch Rhythm Interrelation of pitch and rhythm scale (enumeration of scale of tones). 

Scale of note values (enumeration) relationship of parts intervals (melodic and scaler)  meter thematic material melodic contour sequences of values polyphony formulas tendencies texture.

Timbre tempo

Many studies of bodies of music are devoted to only a few of the elements of music enumerated here; but the exhaustive studies touch on all of these (except where a given musical style does not contain material relevant to such description, such as a style without polyphony, or a study based on written notations in which no indication of timbre is given).

Many scholars are of the opinion that one would be able to produce an adequate description of a whole composition by religiously describing the elements of music of the composition in the form of this outline. Since the possibility of the analyst's being absolutely exhaustive is remote, there have been some attempts to analyze music with electronic computers. No doubt such an analysis would be exhaustive if properly programmed, but it would, in turn, have to be translated into ordinary musicological language. There is, moreover, no way of knowing whether a computer could distinguish between characteristic and non-characteristic aspects of the music unless the programmer were able, in advance, to make the distinction himself. IBM punch cards have been used to identify tunes in the British folk-song repertory which are genetically related (Bronson 1949), and this approach indicates that computers and other information – storing devices could be of limited use in the description of music. It is necessary to remember that description of the individual elements of music, without a consideration of the interrelationships and the points of correlation, among them, could give a misleading total impression. This applies, of course, whether mechanical aids are used or not. It is conceivable, for example, that two musical compositions produce exactly the same description without being identical except in certain superficial ways. Similarly, it is possible to find, in a statistical analysis, that 60 per cent of the compositions in a given corpus have a certain type of scale, and that 60 per cent have a certain type of rhythm. But this statement would not indicate that only 20 per cent of the compositions have both the rhythm and the scale indicated, and that the presence of both in the same song is an exception rather than a rule.

The Intuitive Approach. An alternative to the systematic, elements – of – music approach is one which attempts to identify the most striking, the most important aspect of a piece of music, or of a musical style. This procedure seems to be most rewarding in Western music, in which one can sometimes identify the composer's wishes and intentions. The question "What was the composer trying to do?" can be answered either by recourse to the composer's own statements or by the informed listener who, as a member of the composer's own culture, may be in a position to make valld statements about his music. To adopt the procedure of describing the most striking feature of the music would seem less advisable for non – Western music (or any music outside the student's frame of reference), since one would inevitably be struck by features in the music which either conflict or coincide with his own frame of reference. For example, he may be struck by the peculiarity of the scale, which may sound out – of – tune to him; or he may be struck by the fact that the kind of polyphony found is similar to a kind which he knows, etc. It is doubtful that a reliable description of non-Western music can be produced with this method; moreover, it would be difficult to be sure that such a description is correct.

On the other hand, the intuitive approach may function as a check on the more reliable, but perhaps impersonal, systematic approach.

Thus the student who has described the elements of music in a composition, one by one, could then proceed intuitively to find important features in the music which may have been omitted.

The Selective Approach. Many ethnomusicological studies do not attempt to describe a piece or a body of music in its entirety but, instead, analyze only one or a group of related aspects. There are studies of scale and melody in the music of a given tribe (e.g., Brandel 1962), or of rhythm (e.g., Hendren 1936), or of melodic formulae (e.g., Hood 1954). Some of these studies are purposely selective of the aspect of music which they treat; others are selective because their authors have assumed certain aspects of music to be more fundamental than others. While no scholar can be blamed for studying one aspect of nature or culture which interests him, and nothing else, it is perhaps appropriate to criticize the selective approach as regards its contribution to ethnomusicology at large. The student who is beginning to work in the field of ethnomusicology, or the scholar who is trying to get a broad view of the musical culture of a people, fares better with a systematic, holistic approach to description of music. His great interest in rhythm should not allow him to neglect description of the melodic aspects of music, and the fact that a people exhibit in their music an unexpected kind of polyphony should not blind the student to the intricacies of the melodic formulae. The interrelationship among the various elements of music should be constantly kept in mind. After practice in systematic description of all aspects of a musical style, the student may wish to go into yet more detail in an individual aspect of the style. But this should, in my view, be done only as a part of what must, at least ideally, constitute a broader approach. If scholar A describes the scales of a music and leaves all else, he should at least approach these scales from the point of view that either he or scholar B will some day also study the rhythm of that music. To assume that successful and relevant study of one element of music can be accomplished is to neglect the overweening fact of the close interrelationships among the various aspects of music. Thus a description of only the scales of a musical culture tells us very little about that music. Scales do not live by themselves, and selective approaches to musical description should be considered selective for practical reasons only, not on the basis of scholarly principle. The manner in which selective description was done during the first decades of ethnomusicology indicates the great interest of the early scholars in melodic aspects of the music. And while considerable light was thrown on the music of non-Western cultures by this approach, the fact remains that a somewhat false impression was frequently given. Notions such as the inevitable simplicity of rhythm in non-literate cultures, the lack of classification of polyphonic styles, and the almost complete absence of terminology for description of timbre are partly the result of this selective approach to description of musical style.

Some Published Examples of Descriptions of Style One way of learning about the description of music is to examine some of the analyses of and commentaries on non-Western and folk styles. Thus, before discussing the terminology of musical description and proposing some specific procedures, we should hav~ a look at some of the publications which describe musical style. We are mainly concerned, here, with descriptions of style in which attention is paid to the individual composition; descriptions of the styles of groups of compositions is the subject of Chapter 6. Our task in this section is to see how descriptions of style in individual songs have been presented and how some scholars arrived at their techniques of description. The type of description with which we are concerned has been made almost exceptionally for vocal music; for that reason we are omitting the discussion of descriptions of instrumental music, but these could be expected to follow the same principles.

It is possible that an examination of published descriptions of musical style may not be indicative of the kind of work which is generally done in musical analysis. After all, few scholars are able, for reasons of space, to publish the complete results of their musical analysis. Few scholars would even wish to publish their complete analyses: a great deal of musical analysis (done by the student with the purpose of simply understanding the music) involves features which can easily be seen or heard, and the purpose of most published descriptions of music is the enlightenment of the reader or listener who will find some of the simpler aspects of the description unnecessary. This is especially the purpose of most analyses of individual compositions. As we move from analysis of single songs to descriptions of bodies of music, less can be found out through simple inspection of transcriptions, and the results of analyses, digested, are more essential to the reader. The following paragraphs are a discussion of certain selected published descriptions of music, beginning with description of individual compositions.

The tendency to describe each composition within a large collection is most evident in studies of folk music, particularly that of the British and British – American traditions. Cecil Sharp (1932) already made use of this practice, although on a selective basis. Each of the tunes in his classic collection of English folk songs found in the mountains of Virginia and neighboring states has the description of its mode and scale, in abbreviated form, according to Sharp's special system. Other aspects of music are not described, however, and it is obvious that Sharp, like many of his contemporaries, was most impressed by the relationship which the melodic structure of these tunes seems to bear to that of medieval music.

A more nearly complete description of individual tunes by Schinhan appears in a collection (Schinhan 1957) of songs similar in style to those published by Sharp (1932).

Schinhan, after each tune, gives the scale (according to his system of classification), names the tonal center, and gives an indication of the interrelationship of the sections of each song with the use of letter – schemes widespread in musical literature. At the end of his collection, he goes into greater detail in describing the melodic structure, reproducing the scale of each tune, and indicating the number of times each tone is used in the song.

This very thorough description of scale (without accompanying description of melodic movement, formulae, etc.), forms a curious contrast to the neglect of the rhythmic aspects of the music. It is indicative of trends in ethnomusicology that Sharp's collection does not go beyond description of the mode of each tune, while Schinhan summarizes the individual descriptions in tables and statistical charts.

Flanders (1960) publishes analyses modeled largely after Schinhan’s procedure, but with some added features. The structure of each tune is given according to the usual letter – scheme, and the relative lengths of the individual sections or phrases are stated by formula.

Rhythmic structure is classified according to five main types which appear in the British – American ballads treated here. A combination of note – value relationships and meter is used as the basis for this classification. The melodic contour is described by a key word, such as "arc," "pendulum,' "undulating.'. The scale, transposed so that the tonic will be G is reproduced for each tune.

In each of these examples, British – American folk song tradition is involved, and the form of the description of the tunes conforms to the general style of the music. Thus, it is possible to classify rhythm and meter as falling in one of five main categories (corresponding essentially to poetic meters) ; it is possible to show the interrelationship of sections, since the songs can easily be divided into sections of approximately equal length, corresponding to poetic lines, and since there is considerable repetition or recurrence of individual sections so that letter–schemes are meaningful.

It is more difficult to describe the style of individual compositions in non-Western music, in which the concepts developed in Western music – theory, which form the basis of the descriptions given above, are not as easily applied. Perhaps it is for this reason that printed collections of non – Western music are not as frequently accompanied by the kind of tune – by – tune musical description as are those of folk songs. Roberts (1955) has published a collection of Nootka Indian songs containing what is essentially song – by – song description, even though the arrangement is such that the analyses do not appear with the transcriptions. Ninety – nine songs are involved. Roberts has concentrated on scale and over – all structure, paying less heed to rhythm. In contrast to the students of American folk song, she seems to be most interested in the form of the songs, giving the letter – scheme for each tune at least twice, first in a tabular arrangement in which the number of sections, phrases, or parts is the main criterion, and again near the end of the book, where the form of each song is laid out in even more detail. Finally (p. 209) a table giving the characteristics of each of the ninety-nine songs is presented.

Another example of non – Western music described song-by-song was published by Christensen (1957) in a study of music in New Guinea. Rather than relying on the formulaic presentation of Schinhan (1957), Flanders (1960), or Roberts (1955), Christensen gives a commentary on each of his fifty – two songs. He includes mention of over – all form, melodic contour, scale, rhythm, and – this is somewhat unusual – manner of singing, timbre, and tempo. This information is given again, in terms lending themselves more easily to comparison, in an appendix consisting of tables.

The examples of descriptions of individual songs discussed here are all more or less in the systematic category, i.e., they examine various aspects of the music one by one, giving their interrelationship but avoiding an impressionistic statement of what is most striking or what the composer was trying to do. Perhaps a reason is that the intuitive approach, applied to individual songs, may result in statements which have more the ring of "criticism" than of description. When this approach is applied to a body of music, it may be more valld, since it contains at least the statistical validity of numbers. An example of this critical sort of approach to description of individual songs is provided by Peacock (1954).Here, a "musical analysis" of nine songs is presented. Song no.1 is described as follows: Mode of G (Mixolydian) ; tonic G. This is a short song of a fisherman....The melody consists of a single phrase, very pretty, but of a more restricted scope and with fewer ornaments than the other songs. Its bucolic color comes from the use of the Mixolydian mode (G)" (Peacock 1954:135). Rhythm is not mentioned, and the form is given only by implication. Emphasis is on the mode, which the author obviously considers the essential element of this song. But his song no.3 is described in di:fferent terms: "Mode of D (Dorian); tonic D. The form of this love song is similar to that of the song which follows A – B – A. Here however the second phrase varies as it is repeated, with pretty ornaments, and, at

the end, the theme A comes back slightly altered." Aside from the value judgments ("pretty"), this description is objectionable because of the discrepancy between the elements of music described in the two songs. If song no. lhas bucolic color caused by the Mixolydian mode, what color is caused by the Dorian mode in song 3? Can the author be sure that he is describing elements of music which are actually the important ones, or is he simply reflecting his own background and interests? Since he probably could not tell what the composer was trying to do, he might have produced a more successful description by using a systematic outline. The intuitive analysis in Peacock (1954) is, we should say, probably indicative of earlier practices than his date of publication indicates. While the intuitive approach to analysis may occasionally show important features of music which do not turn up in more systematic descriptions, it is evident that, at least in music outside the Western cultivated tradition, it must be used in conjunction with the systematic.

It is obvious, from an examination of the systematic, song-by-song descriptions of musical styles cited here, that each description is made in a frame of reference based on the describer's knowledge of what the style is at least likely to contain. Making a successful and communicative description of a song without some presuppositions, without some things which are taken for granted by describer and reader alike, seems impossible. Attempts would probably founder on the elaborate terminology needed for establishing concepts and units, beginning with questions such as "What is a tone?" "What is meter?" etc.

That such questions should be asked, and, indeed, that they have not been asked or answered sufficiently, cannot be denied. But if the task at hand is the description of music, then such fundamental questions may get in the way. The ethnomusicologist must take advantage of one of the universaIs of culture: he must take for granted that his reader will know what music is, and his description must fall within that frame of reference. But he should not take for granted is that he and the reader can correctly identify the relevant, the significant units and distinctions in any musical style.

Here again we face, as we did in considerations of transcription (Chapter 4), the distinction between music as a conglomeration of sounds, and music as communication with meaningful and non-meaningful units and distinctions, and operationally, between the phonetic and the phonemic approaches.

There is little that can be advised in a general way to solve this problem and to give direction to the student. It may be possible to differentiate between the phonetic and the phonemic distinctions in music, but such differentiation must come before actual description, rather than resulting from it. Perhaps one example of s:uch differentiation may show its importance in description of style. Let us consider the environment of the individual unit. In the case of scales, it is not enough, for example, to state the number of different pitches which occur, or, in the case of rhythm, the number of note values. Attention should be given to the environment of each, to the notes which come before and after each one, to see whether the occurrence of one cannot be predicted from that environment. Thus, in examining certain eighteenth – century pieces, we find that the raised seventh degree in a minor scale occurs only when followed by the higher tone, while the lowered seventh is found when followed by a lower degree. The two sevenths, used altemately, could thus be considered different manifestations of one tone, not two separate tones, and while each has its distinct function, each is a complement of the other. Obviously, for this reason the two sevenths are considered simply the seventh of the melodic minor scale. Similar but also much more complex situations may be found in non-Western musical styles.

General Terminology and Procedure

While there is no generally accepted set of terms, or definition of these terms, in ethnomusicology, a large proportion of the literature does conform to an approximate standard in terminology. It seems appropriate to comment briefly on some of these terms and concepts.

Tonal Material. The procedures for describing the tonal material are fairly well established. First comes description of the scale, i.e., the tones which appear, without consideration of their role in the melody. (Ethnomusicologists have used the word "scale" to mean something rather different from what it means in traditional music theory. But since ethnomusicologists have also used the word to mean several different things, let us for our purposes define it as those tones – and the intervals among them – which are used in one or a group of compositions.) The first classification of scales is simply an enumeration, so that the terms ditonic, tritonic, tetratonic, pentatonic, hexatonic, and heptatonic simply indicate the number of tones in the scale. Octave duplications are normally omitted, although this may be a questionable procedure, since all cultures may not consider tones an octave apart to be so close in identity as do Western musicians.

Simple enumeration of tones tells us something about the music, but not really very much. Two scales containing five tones may be as different as 9 a – flat b – flat c d – flat and 9 a b de. An indication of the intervals among the tones of the scale is important as well.

Terms such as pentachordal, tetrachordal, and hexachordal indicate the number of tones as well as the fact that these scales are built out of successive seconds. We then proceed to a description of mode, another concept which has been used and defined in a number of different ways. For our purposes, mode is the way in which the tones of a scale are used in a composition. Thus, when we say that a song uses certain tones, we have given its scale; when we say that certain tones are important, certain ones appear only before or after particular others, and a specific one functions as the tonal center, we have given at least a partial description of mode. Scale and mode are usually presented on a staff, with the frequencies and the functions of the tones indicated by note – values – a practice; started by von Hornbostel. The tonic is usually given by a whole note, other important tones are indicated by half notes, tones of average importance by quarters, rarely used or ornamental tones by eighths and sixteenths. Brackets and arrows can be used to indicate other distinctions, such as the tones used by each voice (see p. 162) in a polyphonic piece or the tones used in the individual phrases of a song.

Here we begin to approach the problem of tonality, which in ethnomusicology is a difficult one. (See, for example, Apel 1960:304.) That tonality also presents a difficult problem in Western music, with the exception perhaps of the music composed between 1700 and 1800, is evidenced by the large amount of literature composed of arguments about definitions. Nevertheless, in Western cultivated music one can talk about this elusive concept because the native composer and theorist is available as a source. For music in other styles, acoustic criteria based on the idea that certain intervallic relationships will bring about in man a feeling of tonal center, simply because of the inevitable effect of acoustical laws, have been postulated (see Hindemith 1945).While these may be valid, we cannot easily test the members of non – Western cultures for their reactions to descriptions of tonality based on acoustics. I do not wish to give the impression that I doubt the validity of the acoustic criteria of tonality; but I cannot find that they make a great contribution to the description of non Western musical

styles. If we are to talk about tonality at all, in ethnomusicological description, then we should use tonality as a concept directly descriptive of the music. And while counting the frequency of the tones in a song, and calculating the interrelationships of the tones in terms of their positions in the song, may produce results which violate the acoustical criteria of tonality, the information thus gained does tell something concrete about the song itself. To study tonality in any other way seems both risky and potentially meaningless; but at the time of writing, ethnomusicological descriptions of tonality which do more than simply enumerate the tones in the scale and indicate their relative frequency are rare.

Following are some methods which ethnomusicologists have used to identify tonal centers and to distinguish a hierarchy of tones in a piece: 1) Frequency of appearance is perhaps the most widely used criterion. 2) Duration of notes is sometimes used, that is, those tones which are long – whether they appear frequently or not – are considered tonal centers. 3) Appearance at the end of a composition or of its subdivisions is thought to give tonic weight to a tone. Initial position is also a criterion. 4) Appearance at the low end of the scale, or, again, at the center of the scale, may be a criterion. 5) Intervallic relationship to other tones – for example, appearance at two octave positions (while other tones appear only once), or appearance a fifth below a frequently used tone – is another criterion sometimes used. 6) Rhythmically stressed position is a further one. 7) We must never neglect the possibility that a musical style will contain a system of tonality which can only be identified by means other than those already known and used. An intimate acquaintance with the music of such a style would seem to be the best insurance against ignorance of such a system.

Although the above criteria often conflict, most scholars, in their identification of tonal centers, seem to rely on a combination rather than a single one.

Other aspects of melody have been less formalized in description than scale and tonality. Melodic contour, for example, is usually described by very general terms, such as "arc," "pendulum," "gradual descent," etc.

While a more comprehensive system of classification would be helpful here, the use of generally understood terms has advantages over specialized and rigid systems which – as in the case of scales – sometimes obscure rather than amplify the music to be described.

Rhythm. Since ethnomusicologists have paid less attention to rhythm than to melody, the methods for describing rhythm are much less well developed than those for movement in pitch. Sachs (1953) has attempted to provide some of the techniques which are required. But the concepts of intensity and stress in music have proved themselves so elusive that little real progress has been made. The study of length as rhythmic function is better understood, and length is easier to describe than stress patterns. Thus the best way to begin a description of rhythm is to count the various note values and describe their functions and environments, much as was done for the individual notes in the description of scales. Formulas and repeated patterns should be identified and noted.

Stress patterns, and ultimately meter, are more difficult to describe, and published transcriptions are not always reliable in this respect because of the desire of many transcribers to identify meter, and to show that their music has some regularity of meter. It is useful to distinguish among pieces which 1) have a single metric unit repeated throughout, 2) are dominated by a single metric unit but diverge from it occasionally, and 3) are not dominated by any single pattern. The terms "isometric" and neterometric" have been used to distinguish the first from the other two kinds.

The definition of meter is bound to plague the ethnomusicologist, since it is a concept essentially confined to Western music and derived from special types of Western poetry.

Rather than approach meter intuitively, the student is advised to identify it by using such objective criteria as stress and repeated patterns in note – length. It is true that he will sometimes turn out to describe something which is not exactly the same as the meter of Western classical music, but at least he will be describing an aspect of the music at hand rather than something which he only assumes to be present.

Tempo, a further aspect of rhythm, has usually not been described at all. It is usually indicated in transcriptions by an "M.M." marking, but this is only part of the notation, not of a description. Kolinski (1959) and Christensen (1957) have attempted to provide techniques for describing the speed of music. Their systems, essentially, express tempo in terms of the number of notes per minute (on the average), and this approach seems the best worked out so far.

It is possible to determine the average number of notes per minute from a transcription with a metronome (M.M.) marking. If, for example, the marking is q=84, one should count the number of quarter – note equivalents in the piece and divide by 84. This gives the number of minutes and fractions there of which the piece took to perform. Then one should count the number of notes and divide by the number of minutes. The answer is the average number of notes per minute. This method of indicating tempo does not take into account changes of tempo; if these occur, and if each change of tempo is easily identified, each section with its own tempo should be treated separately in the way described above. Also, the aspect of tempo involving the length of beats, or the pulse, is neglected. But beats cannot easily be identified (or distinguished from half or double beats) unless the composer can identify them for you, or unless a percussion instrument performs them, or unless the notes of the melody are regularly the equivalent of beats.

Form. Form, with the specialized meaning as the interrelationship of sections, and the total structure of the piece including the interrelationship of melodic and rhythmic elements, has been classified in several ways. Unfortunately, labels such as "miniature sonata," "reprisenbar," etc., taken from Western music and not descriptive but simply comparative in their function, have frequently been used and have obscured the form as it really appears.

Two main problems face the describer of musical form: 1) the identification of thematic material, on which the rest of a piece is based, and 2) the identification of divisions in the music, that is, of sections, motifs, and phrases. The first of these, thematic material, tends to occur in longer pieces only, and for the ethnomusicologist it is primarily of interest in the study of Oriental cultivated music. The idea that there is a theme which is stated at the beginning, and on which the ensuing material – perhaps a whole piece – is based, applies mainly to European music composed after 1700. In the sonata form the distinction is very specific. But the notion that there is, somehow, a hierarchy of musical material in a longer piece, that there is primary material which is the composer's basic idea and on which other, secondary material is based, may be valid for other musics as well. No specific way of describing this phenomenon has been published, but the possibility of its presence should be kept in mind.

Dividing a piece into sections is necessary for describing its form. Criteria for division are repetitions (i.e., a portion of music which reappears can be considered a unit) ; phrasing and rests (i.e., rests and dynamic movement such as a decrescendo may indicate endings of units) ; modified repetition such as a repeated rhythmic pattern or a transposition; units of the text in vocal music, such as words or lines.

The relationship among sections of a composition in which thematic and non thematic materials are not distinguished is usually indicated by letters. Thus, a piece which has four different sections would be designated as A B C D. The length of each section, in terms of measures or note values, could be shown. Related but not completely identical sections are indicated by superscript numerals: Al and A2 are variants of the same material; Ba is a new section which contains some material from section A. It is often convenient to indicate some other relationship of a specific nature. It is possible to indicate the interval of transposition by figures in parentheses; thus, A(5) is A transposed down a fifth. Devices of this sort are applicable to specific styles of music, and each style will suggest special ways of presenting its form.

Description of forms as a group is made by general statements of tendencies. Statements of this type might be made as to the number of different sections in a piece, the relationship among the lengths of the sections, the degree to which material presented at the beginning recurs, the extent to which special techniques such as transposition or variation are found, etc. All of these are properly part of a description of musical form, but no specific procedures or formulae for these statements have been evolved or generally accepted.

Other element 8. Hardly any framework is available for the description of timbre and dynamics; but perhaps this is no great disadvantage, since preconceived systems of classification such as those found for scales and meters sometimes tend to obscure rather than facilitate understanding of the nature of the musical phenomena involved.

Harmony and polyphony do have some existing classifications into which these elements of music, as found in non- Western music, can be fitted for descriptive purposes.

While these classifications certainly have their usefulness, they present the temptation of being made into procrustean beds. Moreover, we must not assume that a satisfactory description of harmony or polyphony has been arrived at simply because we have classified the music according to the criteria given below. It is quite likely that important aspects of the texture might thus pass unnoticed. But polyphony is such a complex aspect of music that some initial classification is probably even more useful here than in other aspects of music.

Western music makes rather sharp distinctions between polyphony, in which the interrelationship of two or more voices as melodies is paramount, and harmony, in which the succession of simultaneous intervals or chords is more important (but even in Western music all polyphony has harmonic aspects while in all harmony there is some interest in the melodic relationship among the voices).Distinguishing between harmony and polyphony, however, depends partly on the existence of music theory. In non-Western and folk music, we usually cannot make this distinction, since normally we can only view the material as outsiders without recourse to the composer's own point of view. We can't tell whether the non-Western musician conceives of the music as harmonic or contrapuntal, and to superimpose our own view is irrelevant. Perhaps we should assume that all music in which more than one pitch is heard is essentially contrapuntal, unless we know that the concept of chords is actually present, since such music is usually the result of several instruments or voices performing individual melodies, rather than of one musician performing chords.

Whatever the case is, it seems best initially to class all non- Western music in which more than one pitch is heard at a time as one type of music, which we may for convenience call polyphony.

The two simplest approaches to description of polyphony are examination of the over – all relationship among the voices, and study of their note-by-note interrelationship. The over – all interrelationship can be described in terms of the relative importance of the voices, and of similarity or difference of their content.

Two voices may be of equal importance or one may accompany the other. One may have a progressive form, without repetition, while the other has frequent or constant repetition. One may move through a large range and a scale of many tones while the other may be restricted. Again, the two or three voices may perform the same musical material at different times (imitation), at different pitches (parallelism), or in different variants or speeds (heterophony). The number of voices and the interrelationship of their tone colors – are all of the voices sung or are they performed on one kind of instrument? or is the music performed by a combination of instruments? – are relevant to a description of polyphony.

The harmonic aspects of polyphony are best studied through an exact accounting of the intervals found among the voices. If the progression is note by note, such calculations are relatively easy. It remains, after simply counting intervals, to indicate the kind of position which each occupies within its metric unit, phrase, or section, and which intervals are its neighbors. We may, by this kind of consideration, arrive at structural definitions of consonance and dissonance, based not on the acoustic properties of intervals but on the position which they occupy in the music. (See Kolinski 1962 for a lucid discussion of consonance and dissonance in world music.) If the progression is not note-by-note, more complex ways of stating the kinds of harmonic intervals must be devised. In the field of polyphony, the tendency to impose the standards of Western music theory on descriptions of non-Western music are perhaps the most tempting.

Manner of performance has been recognized since the early descriptions of non-Western music as an essential aspect of musical style, but it has not always been adequately defined.

The assumption has been that a musical performance, as an event, consists of the music itself, which possesses a certain degree of permanence, and the way it is performed, which can be superimposed on the music itself. This idea, of course, stems from the Western cultivated music tradition, in which it is possible to separate what the composer has indicated in the notation and what is added by the performer. In traditional music this distinction exists only by conjecture, and while some of the phenomena usually covered in a description of manner of performance are important, the difference between them and the other elements of music should not be pushed too far. Ordinarily, timbre, vocal quality, and ornamentation are included here. The idea that "manner of performance" is somehow not an essential part of music presents problems in musical analysis which can be illustrated in a consideration of ornamentation. In some periods in Western cultivated music history (the Baroque period, for instance), the insertion of trills, turns, and other ornaments was taken for granted and not specified by the composer. These ornaments could, in a sense, be considered non-essential, since their exact placement and nature were not specified, although their existence in the music at large was considered essential and is a hallmark of the style. But in non – Western music, it is not usually possible to distinguish between ornaments (trills, turns, etc.) which are essential to the music and others which are superimposed. Comparing different performances of one song may illuminate this matter, and informants may be able to make comments. But in a description of one piece of music, the only obvious difference between ornaments and non-ornaments is length: ornaments are made up of shorter notes than non ornaments. In view of this fact, while the term "ornament" may be admitted to indicate certain melodic features such as trills, because it may be descriptive to Western musicians, the concept of ornamentation as a special, nonessential or optional aspect of the music cannot be accepted without evidence from the music's cultural background.

Description of vocal technique lacks an adequate vocabulary. Designations such as "natural" or "unnatural" are meaningless, since all music is a cultural rather than a natural phenomenon. General terms such as "tense," indications of the use of falsetto and of the part of the vocal range employed, statements of comparison with other cultures are useful.

Imitations of animal or instrument sounds can be noted. Obviously, ethnomusicologists have so far been delinquent in providing ways of measuring and describing various aspects of music. The "manner–of–performance" aspect can expect some help from the mechanical and electronic transcribing machines, when these are improved and more readily available. In several papers delivered orally, Alan Lomax has distinguished among about ten different kinds of vocal technique. These play a major role in his classification of the world's music into about ten areas (Lomax 1959). But there still looms the difficulty of communicating to the reader the character of a vocal style. Commonly used terms such as "tense," "relaxed," "pulsating,» "pinched – voiced,» etc., are very general and, moreover, seem to have meaning only to those already acquainted with different singing styles. One is tempted to follow the example of a student of South American Indians who stated that one singer, according to his compatriots, sounded like a cow, and added, "He did."

Other aspects of music remain, but the foregoing are the most obvious. Having commented on the customary ways of describing them, we may proceed to some sample minimum descriptions of individual musical compositions.

Examples of Analysis

It is not easy to teach musical description to the student of ethnomusicology. Much of what he may have learned in the theory of Western music may be useful, some of it will have to be modified, and nothing should be taken for granted. Description of individual pieces of folk and non-Western music, in ethnomusicology, must be more detailed than conventional analysis of Western music. A whole book could be written about each song; the samples here present only a decent minimum. In general, impressionistic statements should be avoided or labeled. And whether a whole piece, an excerpt, a variant of another piece, or an improvisation are being described depends, of course, on what has been done in the field and in transcription. Thus we see again how closely the various techniques of ethnomusicology are interrelated. While both listening and reading notation are necessary for description, the following paragraphs give samples of description of short pieces from notation only.

(Figures 6 and 7 were transcribed by the writer, and thus the descriptions include material on vocal technique and timbre.)

No.1. (Fig. 6) Arapaho song, not otherwise designated. Sung several times, but only one rendition in the transcription. Original pitch not given.

Scale, Mode, Tonality: Tones – g e c G E. This scale is considered tritonic, with two of the tones repeated at octave transpositions. The relationship of the tones is triadic, and the intervals between them are two minor thirds, a major third, and a perfect fourth. The frequency of the tones (in terms of quarter-notes – this is one way of stating frequency) is as follows: g – ll, e – 2, c – 4, G – 12, E – 2. The fact that the two "G's" and the two "E's" exhibit identical frequency is interesting. The note 9 appears slightly raised in the third measure, but there seems to be good reason for considering this pitch simply a variant of the tone 9 rather than an independent tone. The tonal center is G because this tone appears most frequently and constitutes the beginning and the end.

Range: minor tenth; Melodic contour: Generally descending, with gradual lowering, of the center of pitch. If the three – part form indicated below is accepted as the basic form of this piece, the first part centers about the tone g, the second begins on 9 and ends on G, while the third centers about the lower G.

Rhythm: One note value (quarter – note) dominates the song, appearing 29 times (counting eighths followed by eighth rests). Four eighth-notes and a dotted eighth-sixteenth figure are the only other note values. The grace-note in the last measure and the pulsations on the notes in the first three measures add to the rhythmic variety. This variety decreases between the first and second halves of the song.

Meter: Uneven, with units of four, five, and three quarters appearing. Stresses appear after bar lines.

Tempo: Quarter – note equals M.M. 120. Thus, according to the formula of Kolinski (1959) which designates tempo as the average number of notes per minute, the tempo of this song is 140. The duration of the song is slightly over 15 seconds.

Texture: Monophonic.

Form: Sections are not easy to separate. According to the points at which rests occur, there are three sections, which could be labeled ABa Cb, indicating that the second section has material from the first, and the third, material from the second. According to the presence of repeated material, which indicates to some extent the length of independent units, there are five sections, marked by full bar – lines: A B A C D, which have the lengths, in terms of quarter – notes, of 5 – 4 – 5 – 7 – ll. Some of these sections could be subdivided, a new subdivision's beginning with each heavily stressed note, so that a third way of describing the form of the song (according to full and half bar lines), is a A B1 A C B2 D C B2. Characteristic of this song is the presence of a closing formula for each of the three main sections, labeled B1 or B2, and the recurrence of certain formula, such as c – G – E.

Timbre: Considerable tension on the vocal chords; falsetto on the high notes; pulsation on the longer notes in the first half.

Intensity: Quite loud, but diminishing in the lower and later portion. No.2 (Fig. 7) One stanza of a song in British – American folk tradition, "The Elfin Knight," collected in southem Indiana.

Scale and Mode: Hexatonic, with two octave duplications:

The center of the range is used more than the extremes. The mode is related to the heptatonic Dorian and Aeolian modes.

Range: Major ninth.

Intervals: In scale, major seconds, one minor second, one minor third. In the melody, 23 unisons, 1 minor second, 8 major seconds, 3 minor thirds, 3 major thirds, 2 perfect fourths, 1 minor sixth, 1 major sixth.

Melodic contour: Generally undulating. Sections 1 and 2 have arcs, section 3 is ascending, section 4 an inverse arc.

Meter: Triple; the notation in 6/8 is probably no more justified than would be a notation in 3/8. One measure has an extra eighth; otherwise isometric.

Rhythm: Note values, counting each rest as part of the preceding note (this procedure may be justified here since the point at which a note stops and a rest begins varies from stanza to stanza): eighth 35

fourth 4

dotted fourth 1

half 1

sixteenth 1

dotted eight 1

The song is dominated by eighths, with longer notes appearing near the endings of sections.

Tempo: Average of 152 notes per minute.

Texture: Monophonic.

Timbre: No indication; no ornamentation.

Form: Four sections, coinciding with lines in the poetry. The sections are almost equal in length, and their interrelationship is expressed by A Ba C D.

Scale, Mode, Tonality: Pentatonic, major seconds and minor thirds among the tones. The tonality is difficult to stabilish, and may be described as changing with each line, because of the melodic material is transposed in each line. On the basis of note length and the importance of the cadence, the last tone of each line could be considered its tonal center.

Possibly the repetition of rhythmic patterns and the basic identity of the four sections made the use of a strong tonal center (by the composer) unnecessary.

(Number of occurrences, section by section.)

Thus the tone of greatest prominence could be the final or the pre–final tone of each section.

Range: Minor fourteenth.

Melodic intervals: Unisons (8), major seconds (19), minor thirds (13), major thirds (I), perfect fourths (12), perfect fifths (6).

Melodic contour: Each section is an ascending and then descending arc; the overall

relationship of the four sections also describes an arc of sorts because of the considerably higher average range of the third section.

Meter: Duple, evidently with major stress every fourth quarter.Bruno Nettl - Theory and Method in Ethnomusicology 96

Rhythm: Note values – 40 quarter – notes, 8 eighths, 8 dotted quarters, 4 halves.

Halves appear only at section endings, dotted rhythms only in middle of section. The rhythmic arrangement is a repetition of a rhythmic pattern for each section, i.e., isorhythmic.

Tempo: Average of 153 notes per minute.

Texture: Monophonic.

Form: Four sections, of equal length and identical rhythmic pattern.

Each section is a transposition, with some variation, of the first. Using superscript and subscript numerals to indicate approximate intervals of transposition upward or downward, the form can be expressed as A A2 – 3 A 5 A " – 3. If the interval of transposition is calculated according to the tones in the song's scale rather than the diatonic scale, the pentatonic intervals of transposition would be A A2 A 3 A 2, with the third section a variant.

Polyphonic music requires the same kind of description of the individual voices as is done for monophonic music. This accomplished, a special description of the cumulative effects of the combined voices, and of the interrelationship of the voices, may be attempted.

The samples of description below involve only the specifically polyphonic aspects of the musical examples examined. It is assumed that an analysis of the individual voices has already been made.

No.4. (Fig. 9).There are three voices, the bottom one a repeated drone, the upper two moving in parallel thirds (major and minor, in a diatonic scale), each with a range of a perfect fourth. The middle voice occasionally reaches the pitch of the drone, but the upper one stays at least a major third above the drone. The rhythmic aspect of the polyphony is note-against note. Since the drone provides the note of greatest frequency, it (the note C) can perhaps be considered the tonic. Since the interval C – E, sometimes with G added, appears most frequently at section endings (according to the transcriber's bar lines), it may be possible to assume even a tonic chord, C – E – G. There is evidently a hierarchy among the voices, the upper two a melody, the lowest an anchor.

No.5. (Fig. 10). There are two voices, evidently of equal importance, with ranges which overlap on only one note. The rhythmic arrangement is note – against – note, with only occasional exceptions. The piece is presumably not complete; only an excerpt seems to be given in the transcription. The harmonic intervals have the following frequency: octave (6), major sixth (3), perfect fifth (9), major third (3), unison (1). At the beginnings of measures, which seem to be rhythmic as well as formal divisions, octaves or fifths appear.

Figura 9 e 10

The two voices do not perform the same thematic material, and tonality seems to be difficult to define. Distribution of the tones between the voices is given here:

It is obvious, from the complex problems posed even by these very simple polyphonic pieces, that the description of polyphony is a complicated and detailed task. We need not wonder, perhaps, that most published descriptions of polyphony are relatively cursory and impressionistic, that statistical expressions of what transpires are rarely available, and that these descriptions are most frequently based on a single aspect of polyphony, such as the over – all relationship among the voices, or the intervals, or tonality.

Conclusion

The descriptions of musical styles of individual compositions presented here also indicate to how great a degree the student wishing to describe musical style from a written notation is dependent on the quallty of the transcription and the information which is given with it. Notes on vocal technique are rarely given, tempo markings are often omitted, indications as to the portion of a piece transcribed are often not there. In polyphony, the number of performers, the kinds of voices used, reinforcement with instruments, the presence and specific rhythms of accompanying percussion instruments are often lacking.

Thus, while the ethnomusicologist is frequently obliged to describe music which he knows only from paper, he is better off if he can have on hand recordings of the identical pieces, or at least of pieces in the same musical style.

Basic as the description of individual compositions is, it is of much less general interest than the description of bodies of music. The informed reader of ethnomusicological literature is usually able to provide, at a glance or within a few minutes, the kind of analysis which we have just given. Describing an individual piece of music is, then, only a moderate service to the informed reader. It is primarily done as a step toward describing a body of music, and for the benefit of the person who is doing the describing. Perhaps for this reason, descriptions of individual compositions are not common in the literature; and presumably such descriptions are much more readily available in an author's notes leading to publication than in manuscripts submitted for publication. Of course, the longer a composition, the more useful is a description of it; but curiously, the shorter and simpler a piece is, the more likely is a description of it to appear in print.

Descriptions of the long musical forms of oriental cultivated music are sadly lacking, but they are not as rare for European folk tunes. In any case, the ethnomusicologist's greatest service in musical description is in the statistical or intuitive blending of individual descriptions into a description of bodies of musical creation – of musical styles as determined by composer, function, community, culture, or historical era.

Bibliography

Apel, Willi and Ralph T. Daniel (1960). The Harvard Brief Dictionary of Music. Cambridge: Harvard University Press.

Branrlel, Rose (1962).The Music of Central Africa; an Ethnomusicological Study. The Hague: M. Nijhoff.

Bronson, Bertrand H. (1949)."Mechanical help in the study of folk song," Journal of American Folklore 62:81 – 90.

Christensen, Dieter (1957). Die Musik der Kate und Sialum. Berlin: author.

Flanders, Helen H. (1960).Ancient Ballads Traditionally Sung in New England, vol. 1. Philadelphia: University of Pennsylvania Press.

Hendren, Joseph W. (1936). A Study of Ballad Rhythm. Princeton:Princeton University Press.

Hindemith, Paul (1945).The Craft of Musical Composition, book 1. London: Schott.Bruno Nettl - Theory and Method in Ethnomusicology 98

Hood, Mantle (1954). The Nuclear Theme as a Determinant of Patet in Javane8e Music. Groningen: J. B. Wo]ters.

Kolinski, Mieczyslaw (1959)."The evaluation of tempo," Ethnomusicology 3:45 – 56. (1962)."Consonance and dissonance," Ethnomusicology 6:66 – 75.

Lomax, Alan (1959)."Folksong style," American Anthropologist 61: 927 – 954.

Nettl, Bruno (1956).Music in Primitive Culture. Cambridge: Harvard University Press. Recommended reading, pp. 45 – 89.

(1960).Cheremis Musical Styles. Bloomington: Indiana University Press.

Peacock, Kenneth (1954)."Nine songs from N ewfoundland"m Journal of Amerjcan Folklore 67: 123 – 136..

Roberts, Helen H. and Morris Swadesh (1955).Songs of the Nootka Indian's of Western

Vancouver lsland. Philadelphia: Transactions of the American Philosophical Society, new series, vol. 45, part 3.

Sachs, Curt (1953). Rhythm and Tempo. New York: Norton.

Schinhan, Jan Philip (1957).The Music of the Ballads. Durham, N. C.: Duke University Press. (The Frank C. Brown Collection of North Carolina Folklore, vol. 4).

.Seeger, Charles (1951)."Systematic musicology; viewpoints, orientations and methods," Journal of the Amercan Musicological Society 4:240 – 248.

(1953)."Preface to the description of a music," Proceedings of the Fifth Congress of the lnternational Musicological Society. The Hague: Trio.

(1962). "Music as a tradition of communication," Ethnomusicology 6:156 – 163. Sharp, Cecil J. (1932). English Folk Songs from the Southern Appalachians. London: Oxford

University Press. Reprinted, Oxford University Press, 1952.Bruno Nettl - Theory and Method in Ethnomusicology 99

~0~