[PROLOGUE]

[lines 1–42]

1–10 The fitting and honest entreaty of our friends counts as a command. When I have been urged by such entreaty towards what needs to be compiled, written and taught about the art of music, I have replied that for the most part people appropriately skilled in this art are to be found in any important ecclesiastical foundation, and if I were to presume to teach some trifling thing about music they would perhaps attribute it to my arrogance or to habitual rashness, and the question would be put to me, both in my presence and in my absence, which is sometimes put to the foolish and to the ambitious, namely whether they would call for a better bread than a wheaten loaf and a better drink than wine.

11–19 However, I have frequently noticed so many of my friends and pupils wandering badly in the path – that is to say in those things which are the principles of music, namely the knowledge of intervals which are produced either from equality of pitch or from raising and lowering – I thought that I would try to be a help and source of counsel for their obvious ignorance, principally so that they should be able to sing properly constituted chant in a well informed way, and should receive as much honour when they sing amongst the uninformed as do the most expert.

20–33 The following are the matters which I shall not discuss because they are beyond the power of boys and need a more thorough investigation: whether music be a liberal art or not; the subiectum of music and the distinctive property of that subiectum; how the intervals are founded and established according to the properties of numbers; the consistency of music.

Like many writings of the Middle Ages, the Summa musice survives in a ‘late’ manuscript copied by a scribe whose interest in the material before him was partly that of a collector and antiquarian. The treatise is known from one manuscript, now number 264/4 of the Archiv des Benediktinerstiftes in St Paul in Lavantthal. A paper manuscript, dating from the early years of the fifteenth century or possibly from the very last years of the fourteenth, it could be French and may be Parisian. Two scribes can be traced in it, the copyist of the Summa musice being the main hand; unfortunately, his work is inelegant and so highly contracted that it is very difficult to read in many places. Gerbert remarks upon these difficulties in the third volume of his Scriptores where he prints the text of the Summa musice and indeed of most of the other treatises in the codex. Little is known for certain about the history of the book prior to his time.

Ulrich Michels has already provided a description and inventory of the manuscript. In addition to the Summa musice, the book contains all or part of several treatises by Johannes de Muris, the remarkable Tractatus de differentiis et gradibus cantorum by Arnulf de St Ghislain, a fragment of the second chapter of Franco's Ars cantus mensurabilis and some minor pieces. One of these, on f.30v, incorporates a few metrical lines of the Summa musice. Most of these materials reflect a fully developed taste for the apparatus of scholasticism with its formal questiones pursuing an ideal of thorough and dispassionate enquiry.

Geoffrey Chaucer, in The Prioress's Tale, relates how a young boy went to school to learn ‘doctrine’, that is ‘to syngen and to rede’. Learning his Latin from the Psalms and prayers in the Primer, the boy hears the other children singing Alma redemptoris mater as they ‘lerned hire antiphoner’. The kind of schooling which Chaucer describes in this Tale was the foundation of many productive clerical lives spent in the cloister, in the cathedral close or in the service of a magnate. To sing and to read: that was the ‘doctrine’ that promising young boys were taught in medieval Europe, and a man of clerical education never became so grand that he outgrew the simple compliment of the Old French poets: il scet bien lire et chanter.

The purpose of this book is to provide a new text and translation, briefly introduced and succinctly annotated, of a manual written c. 1200 for masters teaching boys to sing plainchant and to appreciate Latin eloquence. The work of two authors whose names have hitherto been obscured, it is a practical manual which has been known since Gerbert's pioneering edition of 1784 as the Summa musice. It is remarkable for many things, not least for its highly unusual form comprising prose chapters followed by versified treatments of the same teaching using a wealth of figurative language. Among the many sections of special interest are a chapter on polyphony and several chapters on composition. There are practical hints for performers of chant, and the treatise also refers to musical instruments in an unusually informative manner, discussing the way in which they are tuned and mentioning their use in the training of novices.

The Summa musice is a practical manual for teaching boys to sing from plainchant notation. Like the Disticha Catonis, a famous collection of schoolroom maxims which is quoted almost verbatim in the first lines of the work, the Summa musice is a schoolbook. We can also read it as a literary work designed to kindle a flame of devotion to Christian history and to the writings of the Ancients: Horace, Ovid, Virgil and more besides. The literary ambitions of the Summa musice are plain enough in the authors' decision to construct their treatise from passages of prose alternating with versified and highly figurative statements of the same information (of which more below), but it is also part of the literary nature of the Summa musice that it is such a verbal work. There is not a single table or diagram of the sort to be found in other treatises on music theory; there is only one musical example and the authors refer to their reluctance to mix musical notation and prose (1935–6).

Perseus and Petrus have also pruned away most of the thorny terms and topics of chant theory. They never use the Greek string-terminology which gives many treatises an alarming appearance (‘proslambanomenos’, ‘hypate meson’, and so on) nor the accepted terms for mathematical relations and proportions like ‘sesquitertia’ and ‘superbipartiens’. They never broach the kind of speculative topics that pleased the fancy of some theorists (a comparison between neumes and the metrical feet of Classical prosody, for instance), and they avoid certain questions, such as the division of the semitone, which lead so many other theorists into technicalities.

The supposed reference to unheightened neumes

Since at least 1905, when the second volume of Wagner's Einführung in die Gregorianischen Melodien was published, it has been widely assumed that a passage of the Summa musice refers to the continuing use of unheighted neumes. Following the indication in Gerbert's edition, Wagner accepted the Summa musice as a work of Johannes de Muris and he therefore interpreted the apparent reference to unheighted neumes as a sign that the treatise was written in Germany since the use of staff notation was fully established in Italy and France in the lifetime of Johannes de Muris. In 1926 Besseler alluded to this argument and tacitly approved it. Smits van Waesberghe rejected this interpretation in his edition of the De musica by John ‘of Affligem’, but in recent years it has surfaced again. In his survey of the music treatises by Johannes de Muris, Ulrich Michels states that in the milieu of the Summa musice there were two systems of musical notation in use, one (with staves) that could be relied upon and another imperfect system (without staves) which could not. Michels also repeats Wagner's contention that on these grounds the Summa musice would appear to have been written in an area of German speech.

This is not quite correct. Having discussed the various forms of neumes, the authors of the Summa musice have this to say: ‘But to this point (sed adhuc), a chant can only be imperfectly (minus perfecte) recognised by these signs, nor can anyone learn a chant from it in solitude’ (536–8).

The Summa musice as we have it incorporates several layers of interpolation. At some stage an interpolator lengthened the Metrum of Chapter XX, splitting it apart in three places and inserting verse mnemonics, cast in hexameters, of the kind to be found in many medieval manuscripts (especially German ones) and mostly concerned with modal finals, the modal classification of various antiphons and related matters. Gerbert printed the interpolated lines as if they were part of the Metrum, but passages such as these destroy the rhyme scheme and seem very unlikely to have been part of the original text:

It is possible to demonstrate that these lines, and others like them, entered the text of the Summa musice some time after the second-layer chapter of c. 1225–37. At line 2123 the author of the second-layer chapter gives the total number of verses in the whole treatise as 860. The text edited here with the mnemonic hexameters deleted comprises 864 lines, and given the likelihood that the very last couplet is not to be included in the count (it is a kind of colophon, as noticed by the scribe who sets it well apart) then the total number of hexameters in this edition is 862, only two lines higher than the correct sum. Even with these deletions made the number of lines in the text is too high, so it is plain that interpolations have been made since the author of the second-layer chapter made his count.

Four techniques have been proposed to solve the problems of transcribing complex polyphonic phenomena. All of them require, for one reason or another, that the investigator spend time in the field. The first is propounded by the American school following Mantle Hood, and is called performing. It involves learning how to play traditional music and using the knowledge acquired in this way to transcribe it. The second requires the kind of special technical equipment developed by A. M. Jones for his own use. The third involves the use of films, which may be either silent or synchronised with a sound track. These are subjected to frame-by-frame analysis. The fourth and last consists of a special use of conventional stereophonic sound recording equipment to obtain a measure of separation of the parts in the polyphonic whole.

We will now discuss each of these techniques in turn and point out what we take to be their individual advantages and drawbacks.

PERFORMING

As we have said, this technique involves learning to play an instrument used to perform the music under study, and using the acquired knowledge as a basis for analysis. In the preceding Book, we noted that Gerhard Kubik (1964a) recommends this method. He advises the investigator to get advanced knowledge of how to play an instrument from the traditional musicians in their own cultural setting. Kubik himself followed this procedure in studying the xylophone in Uganda and the harp in the Central African Republic. Mantle Hood, who first propounded this method, has, of course, also used it. In The Ethnomusicologist (1971: 235–7), he describes how he went about learning drum formulae from an Ewe master drummer in Ghana, to gain an understanding of the nature of African polyrhythm.

ON MUSIC BASED ON THE TACTUS

African cross-rhythm is extremely complex, and is characterised essentially by the permanent sense of tension it creates: different interwoven rhythmic figures are repeated cyclically and uninterruptedly. The interlocked structure may involve irregular accentuation as well, making it extremely difficult for even the trained listener to analyse it by ear. This is what Apel (1970: 214) calls ‘conflicting rhythmic patterns’.

No matter how complex it may be, however, African rhythm always has a simple ultimate reference: the pulsation as defined above, whose function is similar to that of the Medieval tactus. African practice thus resembles the practice in use in the West during the Middle Ages and the Renaissance, when the temporal organisation of music was based solely on this neutral, unmarked, intrinsic element.

Let us recall that the tactus as mere unit of duration had the sole function of providing the performers with equidistant reference points. As Schünemann says, that is why ‘in olden times, the person beating time could only specify the tempo, not accentuation … He beat regularly with upward and downward motions of the hand like a metronome’ (1913: 54). Given the strictly metric (in the primary sense of the term) nature of this governing principle, Auda says, ‘In the thirteenth as in the sixteenth century, the laws of metric measure and rhythmic measure were the same’ (1965: 141).

The tactus made it possible to obtain a direct coordination of the durations in all the parts of a polyphonic work without recourse to an intermediate level in the hierarchy: ‘Until about the seventeenth century, there were no two-, three-, or four-beat measures, only a succession of tactus corresponding more or less to what we call a “beat” today’ (Chailley 1967: 102).

This second Book is divided into two parts. In the first I attempt a classification of African polyphony, and in the second I consider the present state of the subject, in the light of previous studies. My approach to Central African polyphony is based upon fieldwork, and upon initially empirical and always pragmatic observations of the various musical procedures current in a specific geographical area.

It might have seemed more logical to begin with a summary of previous discussions before attempting a classification of these musical procedures. The reason which led me to adopt a different approach is the following one: a summary of the problem could not be attempted until one had defined more exactly what this ‘problem’ was, or, in other words, what the term polyphony could be taken to mean here.

The first part of this book is devoted to a classification of modalities of organisation, because it was this that enabled me to distinguish, out of all the types of multi-part musics observed, those which ought to be considered polyphonic. The second part, on the other hand, is a survey of historical records and of all the specialised modern treatments of African polyphony – such as will be defined below – to which I had access. So my approach, while basically synchronic, does nevertheless enable the diachronic aspect to be considered. This other, historical, point of view does indeed lead to an important observation: namely that a number of the musical procedures that are to be found today in an apparently restricted cultural area are not only still quite widespread in Africa today, but were equally widespread as far back as we have reliable testimonies, that is, up to nearly five hundred years ago.

In Book I, we have shown that traditional Central African music consists essentially of the strictly periodic repetition of similar material, with variations. Polyphonic music is naturally no exception to this principle. The overall structure of Central African polyphony of whatever type can be defined as an ostinato with variations. This definition is wide enough to cover all the observed melodic and rhythmic variety of this music, and still remain the shortest and most concise way of describing its cyclical nature. To avoid confusion, a clear distinction must be made between the ‘tools’ involved in producing this polyphony and the musical techniques employed.

The tools are the human voice and the musical instruments, used individually or in combination. These tools may vary not only from one ethnic group to another, but even within a single population, according to the type of repertory, which will itself be linked to a specific social function. The techniques are the structured ways in which these tools are actually used to produce polyphony.

Years of field work have convinced me that the techniques involved in the production of Central African polyphonic and polyrhythmic structures are few in number.

From a purely systematic, but still abstract, point of view, all polyphony and polyrhythmics in this region can be classified into four types:

1. (1) strict polyrhythmics

2. (2) polyphony produced by hocket

3. (3) polyphony produced by melodic instruments

4. (4) vocal polyphony

These are the basic elements, but the musical reality is more complex and requires further discussion. This is because all of these types, with the exception of hocket polyphony, which works differently, can be combined in various ways.