Machaut's Mass: an introduction 0193163330, 0198163061, 0198163011, 9780198163060

Citation preview

Machaut’s Mass PI

TO ANNV

MACHAUT’S MASS An Introduction DANIEL

LEECH- WILKINSON

CLARENDON

PRESS

: OXFORD

Oxford University Press, Walton Street, Oxford 0x2 6DP Oxford New York Toronto

Delhi Bombay Calcutta Madras Karachi Petaling Jaya Singapore Hong Kong Tokyo Nairobi Dar es Salaam Cape Town Melbourne Auckland and associated companies in Berlin Ibadan Oxford is a trade mark of Oxford University Press Published in the United States by Oxford University Press, New York © Daniel Leech-Wilkinson 1990

First published 1990 First published.in paperback 1992 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means,

electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford University Press This book is sold subject to the condition that it shall not, by way of trade or otherwise, be lent, re-sold, hired out or otherwise circulated

without the publisher’s prior consent in any form of binding or cover other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser British Library Cataloguing in Publication Data Leech- Wilkinson, Daniel Machaut’s Mass : an introduction. 1. Masses in Latin. Machaut, Guillaume de, ca. 1300-1377 Temiathe.

783.2'1'0924 ISBN 0-19-316333-0 (hardback) ISBN 0-19—816306—1 (paperback) Library of Congress Cataloging in Publication Data Machaut’s Mass : an introduction | Daniel Leech-Wilkinson.

Includes a new ed. of the Mass Includes bibliographical references. 1. Guillaume, de Machaut, ca. 1300-1377. Messe de Nostre Dame. 1990. II. Title.

ML410.G966L3

1990

782.32'—d20

89-23005

ISBN 0-19-316333—0 (hardback) ISBN 0-19-816301-1 (paperback)

Typeset by Latimer Trend & Company Ltd.

Printed in Great Britain by Bookcraft (Bath) Ltd, Midsomer Norton, Somerset

PREFACE GUILLAUME DE MAcHAuT’s Messe de Nostre Dame has some of the qualities of the great neolithic circle at Stonehenge. It functions as an imposing vehicle for worship; for its creator it represents a triumph of construction; while to us it is a work whose magnificence dwarfs all that survives of its age. And yet, while Stonehenge has been excavated, analysed, dated and interpreted—many times over—+Machaut’s Mass remains largely untouched, Its date is uficértain, its purpose unclear; its construc, tion is understood at only the most visible levels; it awaits analysis chalIts interpretation and—above all—convincing restoration. lenge is irresistible. This book can offer no more than a preliminary excavation, as if cutting trenches across the site through features which look ng but a they might prove interesting. What turns up is thus nothi heses sample of the evidence which may survive, the hypot r investiconstructed around it no more than proposals for furthe composer gation. Chapter 1 draws a biographical sketch of the guess at r earlie and attempts to fit the Mass into it, supporting an different the work’s original function. Chapters 2 and 3 look from from the angles at its composition, first in terms of form and then asking if it viewpoint of technique. Chapter 4 examines the result, of dates. The is one piece or several, and suggests a range inappropriaAppendix supports a new edition which (perhaps

tely) is printed at the end. ne who can The text should, I hope, be comprehensible to anyo when they first read music. I have tried to explain technical terms those experts for appear, and I therefore beg the indulgence of , especially in whom the preliminary sections are too simple. Later analysis setting Chapter 3, there are unavoidable stretches of close have been printed out evidence for broader conclusions, and these to skip to the next in smaller type to enable the general reader main point. bar numbers of Reference is made throughout the text to the is the usual c—b, ¢’ the accompanying edition. The pitch system

-classes (C=any (= middle C) —b’, c”, with capital letters for pitch

C, and so on).

Be

PREFACE

To friends and colleagues I owe more debts than I can decently mention; but I hope that at least those thanked below will forgive me for associating them with what follows: they are certainly not to blame for its faults. My longest-running debt is to Richard Andrewes and the staff of the Pendlebury Library, Cambridge, who over the years have provided assistance far beyond the call of duty. I wish to thank also the staff at the Archives Départementales and the Bibliothéque Municipale in Reims, and M. Francois Avril and the staff of the Salle des Manuscrits at the Bibliothéque Nationale for allowing me exceptional access to the Machaut manuscripts. Musicologists are the most generous of scholars, and I have received materials and ideas from a great many colleagues, not least Margaret Bent, Christian Berger, John Caldwell, Tim Carter, Liane Curtis, Lawrence Earp, David Fallows, Sarah Fuller, Maria Carmen Gomez, David Hiley, Jehoash Hirshberg, Peter Lefferts, Christopher Page, Richard Rastall, Brian Trowell, Reinhard

Strohm, Andrew Wathey, Nigel Wilkins and Laurence Wright. I owe an especial debt of gratitude to Stanley Boorman and Iain Fenlon. My colleagues and students at Southampton have been more understanding of my divided loyalties than the inequality of the division deserved. And the University of Southampton, The Queen’s University of Belfast, and Churchill College, Cambridge have generously provided financial support. D.L.-W.

NOTE TO THE PAPERBACK

EDITION

I have taken the opportunity to make a few small changes and corrections to the text and examples, and in this connection I am particularly grateful to Roger Bowers for stimulating correspondence on the Machaut brothers’ memorial plaque. D.L.-W., August 1991.

CONTENTS Abbreviations CHAPTER

I:

INTRODUCTION:

EE

Machaut

1.2

Machaut’s Mass

CHAPTER 2.1

Function

AND

, 7

co SI coon]

9

2: CONSTRUCTION

Formal choices 2.1.1 Introduction

2.1.2 The fourteenth-century tradition Kyrie 2.2.1 2.2.2, 2.2.3 2.2.4

THE MaAss =

1.2.1 Introduction, 1.2.2 Date ’ 1.2.3

MACHAUT

Introduction Chant

KyrieI Christe

2.2.5 Kyrie II and III 2.2.6 Conclusion

2:3 Gloria and Credo 2.3.1 Introduction 2.3.2 Gloria 2.3.2.1 Gloria Amen 2.3.3 Credo 2.3.3.1 Credo Amen

14 14 14 14 17 17 19 24 27 29 29 29 30 37 38 43

DA

Sanctus

45

2.5

Agnus Dei

49

2.6

Ite Missa Est

51

ah Conclusion

§2

Vill

CONTENTS

CHA PTER

aK

3: ELABORATION

Introduction

a2

Four-part conception and composition 3.2.1 The problem 3.2.2 Gloria and Credo 3.2.3. The isorhythmic movements 3.3 Prolongations

CHAPTER

4: COHERENCE

.

80

4.1 Introduction

80

4.2 Mode change

81

4.3 Lower-voice counterpoint 4-4 Four-part harmony

82

4.4.1

Dissonance control

4.4.2 Changing dissonance treatment 4.4.3

Consistent dissonance treatment

4.5 Rhythmic coherence 4.6 Motives?

89 9O

Developments

g2

4.9 Satellite works

Q2

4.10 Conclusion

APPENDIX:

a:

INTRODUCTION

84 85

87

4.7 Quotations? 4.8

83 83

95

TO THE

Sources

EDITION

96

96

Aca Establishing Machaut’s text

96

Sharps A.4 Plicas

98

103

A.§ Transcription

105

A.6 Texting

106

A.7 Pronunciation

110

A.8 Voice types and performance style

II4

A.9 Liturgical reconstructions

118

A3

CONTENTS

CRITICAL NOTES Kyrie I Christe _ Kyrie II Kyrie III Gloria Gloria Amen

’

1x

120 121 123 126 128 131 137

Credo

139

Credo Amen Sanctus

150 153

Agnus I

j

'

:

Agnus II Agnus III

Ite

ae

160 162 165

167

Bibliography

I71I

Index

177

La Messe de Nostre Dame

181

ABBREVIATIONS

abbr. add af

ambig aug All Alll

bb. bf br cf. ch Cr CrA

CT dim F-Pn

fo. fos.

G Gl GIA I I-Pu KI KII KIT

Lb] Add. lig m max Mo MS

F-Pn 1584 abbreviation addition, added after

ambiguous augmented Agnus I Agnus II Agnus III

.

F-Pn 1585 bar bars before breve F-Pn 1586 compare

changed Credo Credo Amen Contratenor diminished F-Pn 9221 France, Paris, Bibliothéque Nationale, fonds frangais, MS no. folio folios F-Pn 22546 Gloria Gloria Amen Ite

Italy, Padua, Biblioteca Universitaria, MS no. Kyrie I Kyrie II Kyrie III London, British Library, Additional MS

ligature minim maxima Motetus

manuscript

ABBREVIATIONS

x1

MSS om pd pl

manuscripts f omitted punctus divisionis plicated, plica rest (or recto if attached to a folio number)

Rad

Reims, Archives départementales de la Marne, Annexe de Reims, MS

Rbm

Reims, Bibliothéque Municipale, MS

Sanctus semibreve sb ‘Tenor y? Triplum ae alr New York, Wildenstein America, of US-NYw United States Galleries verso

Vg

US-NYw, callmark unknown (formerly owned by the Marquis de Vogié) Christe

I

INTRODUCTION

MACHAUT 1.1

AND

THE

MASS

Machaut

Guillaume de Machaut) éxists mainly in our imagination, afate which would probably have delighted him.’ His remains— documentary, literary and musical—for all their extraordinary variety and the networks of international relationships which they suggest, offer us few hard facts on which to base a picture of the man and his life.2 We have no date of birth. That he died in 1377

and that his earliest known employment can be traced back to about 1323 have led to a convenient assumption that he must have

been born around 1300, though he could have been several years e younger or even older. The first surviving record of his existenc dates only from 1330. This is a Bull (or edict) of Pope John xxu, t sealed at Avignon on 30 July 1330, which declares that Machau n positio shall become a canon of Verdun Cathedral when a of John, becomes vacant. The Pope agrees to this at the request isclerk King of Bohemia, for whom (the Bull records) Machaut

4 January and almoner.? Similar documents of 17 April 1332 and and Reims ca1333 promise him further canonries at Arras

to include thedrals, varying their descriptions of Machaut slightly a’s Bohemi of notary and secretary amongst his roles in John

household.’

graciously These are typical documents of their time. The Pope in the future favours a secular ruler by agreeing that at some time ular, see Douglas Kelly, ! On 14-century imagination, and Machaut’s in partic (Madison, 1978). Love y Courtl Medieval Imagination: Rhetoric and the Poetry of bey, Guillaume de Macha nd 2 The most detailed published biography is Arma follows,

l (Paris, 1955). In what Machault, 130?-1377: La Vie et L’Oeuvre Musica

oted by their current shelfthose documents which I have consulted are footn shed treatment. mark. Otherwise, references are to the fullest publi its des 3 The

Bull is published and discussed

in Antoine

Thomas,

‘Extra

littéraire’, Romania, 10 (1881), 328, archives du Vatican pour servir a histoire

330-1. 4 Thomas, op. cit., pp. 328-9, 331-2.

2

MACHAUT

AND

THE

MASS

one of his employees may enjoy additional income from some other (ecclesiastical) institution. Everyone, except possibly the institution providing the goods, benefited from the system. The Pope, at no expense to himself, placed the King in his debt; the King secured the possibility of reward for his secretary, and the secretary could look forward to increased income and to the possibility that, should circumstances change, he could always retire to the community of which he would be nominally a canon but in which—and here was the beauty of the arrangement—he was by no means obliged to reside in order to draw his salary.° By 1333, then, Machaut was looking forward to enjoying no fewer than three cathedral canonries, in addition to a chaplaincy

at Houdain which, the 1332 Bull reveals, he already held. None of

these expected positions had, however, become vacant by 17 April 1335 when the new Pope, Benedict xu, as part of an attempt to curb the absurdities of the benefice system, sealed a Bull rationalising Machaut’s holdings. This allowed him to keep the promised canonry at Reims in return for surrendering those at Arras and Verdun immediately, and for giving up the chaplaincy at Houdain when aposition at Reims became vacant. However, Machaut was permitted to keep an income from the church at Saint-Quentin, which he had acquired (for once) without papal intervention.° Machaut’s turn eventually came at Reims in 1337 when, on 29

January, he was received as canon ‘by procuration’, which indicates that he was not there in person.’ He cannot, however, have been entirely absent throughout the remaining years of his royal employment, since his presence is recorded twice in Reims documents over the next few years:® on 13 April 1340 he was

° For more information on the benefice system as used to support musician s see Craig Wright, Music at the Court of Burgundy 1364-1419 (Henryville, 1979)

pp. 66-4, 85.

° Thomas,

op. cit., pp. 329,

332-3.

An

18th+century

MS

in Reims

(the

“Weyen’ MS), compiled in part from medieval documents now lost, implies that the cathedral Chapter recorded Machaut’s nomination on 1 August 1331 and 16 June 1335. The latter date would fit with that of Benedict’s Bull; the former remains mysterious but may refer to an earlier stage in the appoint ment process. Rbm 1773, fo. 400°, also fo. 284°. ” Rad 2.G.1650, fo. $4’. This is the ‘Livre Rouge’ to which Machabey refers,

apparently following Varin. * Perhaps relevant to these years is Rad 2.G.321, piéce 13, fo. 1, a fragmentary summary of canon law with respect to a person who, receive d as a canon of Reims, continues to serve another patron. It dates from around the 1330s and may refer specifically to the case of Machaut.

MACHAUT AND THE MASS

3

present to witness the installation oath of Jean de Vienne as Archbishop of Reims,’ and on 30 May 1344 he witnessed a land

transfer between the abbot of St Rémi de Reims and John of Bohemia.!° On the other hand, the inclusion of his name

in a

salary list (he was paid the standard canon’s stipend of 60 livres) dated 25 May

and

1 November

1346 tells us nothing

of his

whereabouts." It is probable, however, that by about this date Machaut had settled permanently in Reims; for it was on 26 August 1346 that

John of Bohemia died ’at the battle of Crécy. His was one of the more glorious deaths of’ the fourteenth century. Blind, but determined to fight, the King said a very brave thing to his knights: ‘My lords, you are my men, my friends and my companions-in-arms. Today I have a special request to make of you. Take me far enough forward for me to strike a blow with my sword.’ Because they cherished his honour and their own prowess, his knights consented. . . . In order to acquit themselves well and not lose the king in the press, they tied all their horses together by the bridles, set their king in front so that he might fulfil his wish, and rode towards the enemy... . The good king . . . came so close to the enemy that he was able to use his sword several times and fought most bravely, as did the knights with field, him. They advanced so far forward that they all remained on the

not one escaping alive. They were found the next day lying round their leader, with their horses still fastened together."

If these are the bare facts of the first available slice of Machaut’s life, what can we deduce from them about his background and about the kind of life he led up to that point? John The Bull of April 1335 mentions that Machaut had served

of Bohemia for ‘twelve years or thereabouts’, which takes us back been to 1322 or 23. Clearly, by that date he must already have write exceptionally educated for his time, able to read and ial financ elegantly in Latin and French, drafting the letters and ss of the and legal documents that were the day to day busine only secretary to a great landowner. Such an education could to as clue no have begun in an ecclesiastical institution. We have he was which one, beyond a fifteenth-century suggestion that 9 Rad 2.G.323, piéce 13. 1 Rad 56.H.74, piéce A, fo. 30°. Rad 2.G.1650, fo. 269bis”. rev. edn. (Harmondsworth, 12 Eroissart, Chronicles, tr. Geoffrey Brereton, 1978), pp. 89-90.

4

MACHAUT

AND

THE

MASS

born in Champagne." It is tempting to jump to the conclusion that the village of Machault, 39 km. north-east of Reims, must

have been his birthplace, and that therefore Reims cathedral is the obvious site for his early education; but (despite the confidence of the local tourist board) it is a larger jump than the evidence will allow. An early connection with Reims is nevertheless hinted at by the motet, Bone pastor, whose text celebrates a Reims-based pastor Guillaume, the only candidate for whom seems to be Guillaume de Trie, Archbishop from 1324 to 1334.

It is hard to believe, though, that any provincial cathedral could have offered Machaut the exposure to new developments in literature and music which his earliest surviving works show him already to have enjoyed. He must surely have spent some time in Paris. This would seem the most likely explanation for his early adoption of Ars nova compositional techniques,"* and in particular for direct modelling of at least one early work on an existing motet by the originator of the new style, the Paris-based Philippe de Vitry.’ Whatever his early circumstances, they must have changed dramatically once he entered the service of John of Bohemia. The king was an obsessive traveller, moving back and forth between

his possessions in France, Luxembourg and Eastern Europe with such frequency that most years included at least one visit to each. In addition, during the years in which Machaut was his secretary, he campaigned in Germany, Austria, Silesia and as far as Lithua-

nia, as well as undertaking diplomatic missions in France and Italy.'° And however unpleasant Machaut found the hardships of '° Machabey, op. cit., p. 15. His pupil Eustache Deschamps also lists Machaut amongst the poets of Champagne: A. Coville, ‘Philippe de Vitri: Notes biographiques’, Romania, 59 (1933), 521.

‘* Bone pastor is one of the clearest examples of the influence of Philippe de Vitry’s motet style. : ® The relationship of Machaut’s motet Aucune gent/Qui plus aimme/Fiat voluntas tua/CT to Vitry’s Douce playsence/Garison selon nature/Neuma quinti toni

is summarised in Daniel Leech-Wilkinson, ‘Related Motets from FourteenthCentury France’, Proceedings of the Royal Musical Association, 109 (1982-3), p. 5, and discussed in detail in Leech- Wilkinson, ‘Compositional Procedure in the

Four-Part Isorhythmic Works of Philippe de Vitry and his Contemporaries’ (Ph.D. thesis, University of Cambridge, 1983), published (without chapter 5) as Compositional Techniques in the Four-Part Isorhythmic Motets of Philippe de Vitry

and his Contemporaries (New York, 1989), ch. 2.

'® Machabey, op. cit., pp. 19-34.

MACHAUT AND THE MASS

5

é

these continual expeditions, they must at least have made him into one of the most widely travelled clerks of his time. It is probably not coincidence that as John was increasingly incapacitated by blindness during the 1340s, so we begin to find references to Machaut’s presence in Reims as a resident canon. It is worth remembering that he must himself have been nearing, if not already in, his 40s by now—in medieval terms, approaching old age—so that the sedentary life of a cathedral canon must have seemed increasingly attractive. Thus on 1 January 1352 we find him witnessing the installation oath of Archbishop Hugués d’Arcy, and likewise that of Jean de Craon

on 4 November

1355; but that he was

not confined

entirely to Reims is suggested by his absence from similar ceremonies on 2 May 1353 and 29 December 1374.” In December 1361 Machaut’s house was used by the visiting Charles, Duke of Normandy,

then regent of France, while he

attempted to sort out a quarrel between the Archbishop and City concerning the city defences.'* Reims had already been under siege by the English in 1359-60, a calamity which may perhaps

underlie the texts of Machaut’s last three motets.’” In 1364 he was

assessed for taxes payable on a property in the parish of St Timothée in Reims, which, if nothing else, suggests that he was not penniless.” The impression is reinforced by a payment of 300 gold francs from the Count of Savoy in 1371 in return for a manuscript romance,” and by a record of a business transaction on 15 October of the same year.” Then in 1372 we find Machaut

living in a sizeable canonical residence just to the south of the west end of the cathedral (in the present Rue d’Anjou),” after which

nothing survives until the appointment of his successor, Johannes Rbm 17 Respectively, Rad 2.G.323, piece 15, piéce 17, piéce 16 and

1780,

de Reims, ui (Paris, pp. 75-7. Pierre Varin, Archives Administratives de la Ville 1352. August 18 dated e, referenc 1848), p. 31 n. I, transcribes a further

18 Rbm

1628, fo. 236, part of the Mémoires of Jean Rogier, is the earliest

source I have traced. 19 Leech-Wilkinson, Compositional Techniques, pp. 105~7nt has not been 2 Rad 2.G.191, piéce 1, fo. 141°. As far as I know, this docume previously noted.

21 Machabey, op. cit., p. 66. ; 2 Tbid., p. 67. Brejon de Lavergnée. ‘Note Edith Mariein ed discuss 5, piéce 8, 2.G.31 3 Rad

Guillaume de Machaut, Poete sur la maison de Guillaume de Machaut a Reims’, in 1982), pp. 149-52. et Compositeur (no ed.; Actes et Colloques, 23; Paris,

6

MACHAUT

AND

THE

MASS

Gibourti, who is received in person as occupant of Machaut’s prebend, on 9 November 1377.74 None of this material (except in so far as it is influenced by John of Bohemia) is especially surprising or so exceptional as to set Machaut apart from men of similar rank in other noble households. The most obvious comparison, with Philippe de Vitry, whose career began in a similar fashion at the French court, makes Machaut appear a rather unremarkable figure. Vitry probably started as a clerk in the royal household, but is first heard of in surviving documents as a notary under Charles le Bel (1322-8). By the age of 50, however, when Machaut had already retired to Reims, Vitry had become one of the most powerful figures in the French establishment, a valued aide to the future King Jean ny, under whom he was duly rewarded with the bishopric of Meaux.” Like Machaut, Vitry composed—indeed his musical innovations were largely responsible for the character of French music for most of the rest of the century—and like Machaut he wrote poetry, both short formes fixes and larger narrative works. But if in their origins and in their literary and musical interests the two composers seem to be comparable, in terms of political and intellectual achievement Vitry is by far the more significant figure. Why is it, then, that to the student of medieval literature and

music Machaut seems so much more engaging? Certainly the ageing canon of Reims who exchanges love-letters and lyrics with his young admirer Peronne seems altogether more sympathetic a character than the political schemer who supports crooked

administrators and lives under the protection of German bodyguards. Perhaps also the richness and idiosyncracy of Machaut’s musical and poetic languages seem more intriguing than does the classical perfection of Vitry’s surviving motets. But the main reason for Machaut’s comparative fame today is more prosaic. Exceptionally amongst composers (though not ** Rad 2.G.1650, fo. 54". The Weyen MS records that Machaut died in April 1377: Rbm 1773, fo. 284".

* A. Coville, ‘Philippe de Vitry: Notes biographiques’, Romania, 59 (1933),

520-47, remains the fullest study of Vitry’s life. See also Leo Schrade, ‘Philippe de Vitry: Some New Discoveries’, Musical Quarterly, 42 (1956), 330-54.

* On

Machaut’s

correspondence

with Peronne,

see initially Sarah Jane

Williams, “The Lady, the Lyrics and the Letters’, Early Music, 5 (1977), 462-8. For this view of Vitry see Alexander Blachly, “The Motets of Philippe de Vitry’

(unpub. Master’s diss., Columbia University, 1971), pp. 13-15.

MACHAUT AND THE MASS ,

7

poets) of the fourteenth-century Ars nova, Machaut saw to it that his complete works, both literary and musical, were collected together and copied into beautifully written and illustrated manuscripts for the benefit of his wealthy patrons. Because these manuscripts were so costly many were carefully preserved long after their contents had passed from fashion. And being unfashionable, and therefore largely unopened, several survived in near original condition up to the beginning of the present century. Thus while only about a dozen of Vitry’s compositions are known to survive today—still the most by any other Ars nova composer—of Machaut’s we have 143, very possibly his entire musical output. This extraordinary fact has constantly to be born in mind in assessing Machaut’s relationship with the rest of fourteenth-century music, not least in dealing with the apparently unique Messe de Nostre Dame.

1.2. Machaut’s Mass 1.2.1

Introduction

Machaut’s Mass has often been praised as unique. Richard Hoppin, in his standard history of medieval music, writes that ‘It is his largest single musical work and the only one with a strictly liturgical function.... It is the first complete setting of the Ordinary that is known to have been written as a unit by one of composer. In length it far exceeds any of the compilations fourthe in individual movements that make up other Masses and teenth century. Machaut’s Mass was the only one of its kind, not until some fifty years after his death did complete masses begin to appear in the works of early Renaissance composers.” survives But claims like this are only valid if we assume that what writof fourteenth-century music is representative of what was one of ten. Machaut’s Mass is unique not because it was the only only one its kind—we cannot know that—but because it is the it relates amongst those pieces which we still possess. Exactly how will be to those other ‘compilations of individual movements’ must considered during Chapter 2. But before we can proceed we its of m attempt to face the equally difficult historical proble 1978), pp. 414-5. 2” Richard H. Hoppin, Medieval Music (New York,

8

MACHAUT

AND

THE

MASS

position within Machaut’s own output. When and why did he write it? 1.2.2

Date

It will be one of the conclusions of this study that the Mass must be a late work of Machaut. The evidence is largely technical—to do with style and construction—and so can only fully emerge during the detailed discussion of the work in Chapters 2 to 4.¥ Briefly, the Mass relates closely in harmodnic usage to Machaut’s Motets 21-3, themselves possibly written around 1359-61, but at

any rate plausibly later works than those which precede them in Machaut’s own ordering (nos. 1—20). It also has marked similarities to songs written in the period around 1362-3. And this very approximate placing is supported by the manuscript evidence. Of the five main manuscripts

considerably

of Machaut’s

earlier than the others.

works

Dating

one, MS

from

C, is

c. 1350, it

contains substantially fewer works than the later sources; and the

traditional assumption that this is because it represents an earlier state of Machaut’s complete works (i.e. it includes everything written up to that date) has been broadly confirmed by recent scholarship.” MS C includes neither the Mass nor the last three motets.*’ These first appear in the next major source, MS Vg (and its copy, MS B), dating from the early 1370s. On manuscript evidence, then, the Mass is likely to have been

written between c. 1350 and c. 1372 (the latest date for the paper of MS B),*' while stylistic evidence can probably narrow this down to the earlier 1360s.

1.2.3 Function But for what occasion? Much speculation about the function of Machaut’s Mass has grown around the suggestion—apparently first made as early as the eighteenth century??—that it was composed for the coronation of Charles v in Reims cathedral on * See esp. Sect. 4.9.

” Lawrence M. Earp, ‘Scribal Practice, Manuscript Production and the Transmission of Music in Late Medieval France: The manuscripts of Guillaume de Machaut’ (Ph.D. diss., Princeton University, 1983) (University Microfilms

International 8318466), Ch. 2.

* The omission of Motet 4 from C was accidental: Earp, op. cit., pp. 141-2. *' Elizabeth Ann Keitel, unpublished work reported in Earp, op. cit., p. 108. * Machabey, op. cit., ii. 114.

MACHAUT

AND

THE

MASS

9

19 May 1364. No significant evidence has been offered in support: John m died on 8 April, leaving less than six weeks for the commission to be issued and the work composed, and there is considerable doubt as to whether the coronation liturgy could have accommodated such a mass;** but in the absence of a better

proposal the possibility has inevitably remained open, if only to the extent that writers on the Mass feel obliged to repeat it. It has in its favour the splendour of the occasion, a worthy statecounterpart to Machatt’s musical achievement, and the documented friendship (if such a word is appropriate) between Charles and Machaut. But these, of course, are emotional not historical

arguments: they offer nothing to support the Coronation Mass as a serious hypothesis. At the opposite extreme, Elizabeth Keitel has questioned the relevance of any Great Work—Great Occasion assumptions, emphasizing inconsistencies in the Mass in order to propose that it was assembled—perhaps even for use at Avignon—from movements composed at various periods. Chapter 4, in particular, will suggest that the inconsistencies are less divisive than has been supposed; and the liturgical argument for Southern French

compilation is extremely tenuous. But can one offer any more plausible hypothesis—since there remains no certain evidence— which will place the work between these two psychologically attractive extremes? Given the timespan c. 1350-72,

it might seem reasonable to

that suppose that the piece was written for use in Reims. Certainly e would be one way of reading the title given it in MS Vg—th te source which (perhaps incidentally) provides the most accura de messe la version of the music. Here it is headed ‘Ci commence refers n[ost]re dame’ (Here begins the Mass of Our Lady). If this le to the cathedral of Notre Dame at Reims then the possib answers to this question are considerably narrowed. aAlternatively, the title might be liturgical rather than geogr feast n phical in meaning: a Mary Mass, suitable for use on Maria ciation, Asdays (in Machaut’s lifetime, the Purification, Annun

that case it sumption and Nativity of the Blessed Virgin Mary). In (relatively might have been commissioned for use in any of the Machabey, op. cit., 1. 62-4. 3 The coronation liturgy is briefly described in Guillaume de Keitel, ‘The So-Called Cyclic Mass of 4% Blizabeth

Machaut: 307-23-

New

Ann

48 (1982), Evidence for an Old Debate’, Musical Quarterly,

ie)

MACHAUT

AND

THE

MASS

few) churches and chapels capable of supporting complex polyphony at this time. But it is difficult to see why Machaut, of all people, should have been chosen, other than by his own community, as a suitable person to write a mass. His fourteenthcentury reputation was first as a poet, and only to a much lesser extent (if we are to judge by his representation in surviving manuscripts) as a composer of music setting his own love-lyrics. Even his isorhythmic motets—exceptionally for his time—set mainly amorous texts in the vernacular. Only numbers 19 (in honour of St Quentin) and, more significantly, numbers

21-3,

have strong sacred associations.

Moreover, documents first brought’ together by Armand Machabey point to the tantalising possibility that the piece was in some way connected with Machaut’s will.” According to eighteenth-century copies, Machaut and his brother Jean (also a canon of Reims from 1355 until his death in 1374)*° were commemorated in the cathedral by the following inscription: Guillelmus de machaudio . suusque Johannes frater Sunt in loco concordio . iuncti sicut ad os crater Horum anniversarium . est iuxta petitorium Oratio pro defunctis . diebus sabbathi cunctis Pro animabus eorum . amicorumque suorum Dicetur a sacerdote . celebraturo devote Ad Roellam in altari . missam que debet cantari Pro quorum oratione . cum pia devotione Ad eorum memoriam . percepimus pecuniam Trecentorum florennorum . nuncupatorum francorum Suis exequtoribus . pro emendis redditibus Ad dicte misse crementum . reddituum et fomentum In eadem presentium . solerter venientium Hos fratres salvet Dominus . qui tollit omne facinus®” Guillaume de Machaut and his brother Jean Are joined in this harmonious place as cup to mouth. Their memorial is, according to their wish,

That the prayer for the dead, on every Saturday, For their souls and for those of their friends,

*° Machabey, op. cit., i. 69-70, ii. 114-15. The suggestion is repeated, watered down, in Hoppin, op. cit., p. 420. *° Rad 2.G.1650, fo. 13" and Rbm 1773, fo. 201". *” Rbm 1773, fo. 488” and 1941, p. 94. Transcribed with emendations in Jean

Goy, ‘Note sur la tombe de Guillaume de Machaut en la cathédrale de Reims’, in

Guillaume de Machaut, Poéte et Compositeur (Paris, 1982), pp. 153-4.

MACHAUT

AND

THE

MASS

Te

May be said bya priest about to celebrate faithfully, At the altar by the Roella, a mass which is to be sung. On account of their prayer, with pious devotion To their memory, wg have set aside money— Three hundred florins certified as French— For their executors for improving the income For payment for the said mass and in order to nourish, By that means, those present and diligently attending. These brothers may the Lord save, who takes all sin away.

The eighteenth-century Canon Weyen copied, in his astonishing collection of facts about previous Reims cathedral dignitaries, a large number of epitaphs, often under the title of the mass celebrated in their memory. In the case of the Machaut inscription his title is ‘Messe de Beata’.*® That this is indeed a mass of the Blessed Virgin Mary is confirmed by the second eighteenthcentury copy, which introduces the epitaph with an explanatory paragraph in French: Guillaume et Jean de machaux, tous deux freres et chanoines de l’eglise de notre dame de reims, ce sont eux qui ont fondé la messe de la viérge qu’on chante, les samedis dans la susditte eglise, c’est ainsi qui l’en explique leur epitaphe que l’on voit sur du cuivre proche I’autel de la roélle a la nef.” Guillaume and Jean de Machaux, both brothers and canons of the church of Our Lady of Reims; it is they who founded the Mass of the Virgin which is sung on Saturdays in the above named church, as is explained in their epitaph which may be seen on the brass near the altar of the roella in the nave.

The final surviving piece in this jigsaw of odd-shaped references is a document of 1411 in which the cathedral Chapter agrees to the institution—according to the wishes of a canon, Jean de Verrier —of two masses, the second of which was to include ‘that

prayer for the dead (after the death of our brother [i-e. Jean de Verrier]) namely Inclina domine aurem tuam, as is customarily said in the on Saturdays in the Mass of the Blessed Mary at the Roella said church for the late Guillaume de Machaut, previously canon ¢ of Reims... .”*° Clearly, then, the Machaut brothers founded a sung mass, in

honour of the Blessed Virgin Mary, which was celebrated on 3 Rbm % Rbm

1773, fo. 488”. 1941, p. 94. Machabey, op. cit., p. 70.

Rad 2.G.357, piece 20.

12

MACHAUT

AND

THE

MASS

Saturdays at a side-altar in the cathedral,*! and which apparently continued to be celebrated right up until the eighteenth century.” It is at least a possibility that Machaut intended his own ‘Mass of Our Lady’ for performance at the Saturday Mary Masses that were to be sung for the souls of himself, his brother and their friends after their deaths.* But there were several similar occasions in the cathedral which could equally well have fathered it. For example, a document dated 9 December 1363 makes arrangements for the annual celebration (on 1 March) of a mass of the Blessed Virgin ‘at the altar of the Holy Virgin in the nave of the cathedral near the roella’.** This was to be funded from land held by the deceased Canon Hugues de Juilly died before 11 February 1362), the income to be divided up among those present at the annual celebration. There is no need to propose this as an especially likely occasion for Machaut’s composition—although an endowment of this sort could, perhaps, be as attractive an inducement for the composition of a polyphonic Marian mass as the prospect of the composer’s own posthumous celebration. It simply provides a further example of one of the most common forms of extra-canonical mass. Although in most circumstances plainchant might have sufficed, any such celebration could— given sufficient endowment to pay the singers*°—have offered a suitable occasion on which to perform a polyphonic mass. *' Goy, op. cit., p. 153 and plate I facing p. 156, places it in Machaut’s time beside the sixth pillar on the right of the nave, counting from the West end, and from c. 1417 on the right of the entrance to the choir through the new roodscreen. * It may be no coincidence that the Tournai Mass, whose Credo may have provided a model for Machaut’s (see Sect. 2.3.3 below), now seems likely to have been used for a Mary Mass celebrated at an altar in the right transept of Tournai cathedral from 1349: Jean Dumoulin and Jacques Pycke, ‘La ‘“‘Messe de Tournai”’ et le culte de Notre-Dame 4la cathédrale de Tournai au XIV‘ siécle’, in La Messe de Tournai (Tornacum, 4; Louvain-la-Neuve, 1988), pp. 56-7. *® We know far too little about the format of such ceremonies to take, with

Keitel, Machaut’s use of a Credo and a high-ranking Kyrie as evidence against such a hypothesis (Keitel, ‘The So-Called .. .’, p. 321). * Rad 2.G.444, piéce 1. , © Rad 2.G.1650, fo. 18° * Recent research by Anne Robertson (generously communicated to me while this book was in press) adds substance to the theory outlined here. During Machaut’s time at Reims the Rouelle—a stone commemorating the martyrdom of St Nicasius in 406/7—became one of the most important places in the cathedral, acquiring an: altar and a statue of the Virgin. Documents from this time on record regular Saturday Mary masses there and numerous foundations

MACHAUT

AND

THE

MASS

13

Whether or not it was written for his own commemoration,

the date of Machaut’s mass, the apparently localised nature of his ecclesiastical reputation, and the variety of extra-canonical masses celebrated in the cathedral all suggest Reims as the most likely destination for the ‘Messe de Nostre Dame’.

—

in the wills of their of various sorts by canons and archbishops. As set out med annually on a perfor be to mass founders these generally establish a Mary al. Although memori a as ter specified day until their death, continuing thereaf suggests that above quoted Machaut’s will has not survived the memorial plaque , that his though ting, interes is It his foundation may have been along these lines. canons. than shops archbi of donation should have been so large, closer to those ated celebr were masses al Professor Robertson’s discovery that such memori t’s Machau if that, ce already before the donor’s death offers welcome eviden his during rly regula it heard Mass were one such, the composer would have nt appare the for ons lifetime. Such hearings could, of course, provide occasi in the Commentaries to revisions to the source for MSS A and G hypothesised the Edition.

5 CONSTRUCTION 2.1 2.1.1

Formal choices Introduction

La Messe de Nostre Dame does not, of course, set the whole text of the Latin Mass to music. Rather, Machaut chose six items from the ‘Ordinary of the Mass’, that is, six of those items whose texts

remained the same throughout the church year: Kyrie, Gloria, Credo, Sanctus, Agnus Dei and Ite Missa Est.! And he set them, in

four-voice polyphony, using two contrasting musical styles: one, based on the principles of the isorhythmic motet, for the Kyrie, Sanctus, Agnus and Ite; the other, using simultaneous text declamation by all voices, for the Gloria and Credo. First, then,

we need briefly to investigate the precedents for these basic formal choices. Why these movements and why these styles?

2.1.2 The fourteenth-century tradition The obvious place to look for an answer is in earlier fourteenthcentury polyphonic mass settings.” This is not as easy as it sounds, first because hardly any of it is even approximately datable, and secondly because little of it is at all like Machaut’s setting in style. A good deal of the known repertory survives in manuscripts likely to have been copied towards the end of Machaut’s lifetime—perhaps in the 1360s or 70s;? and much of this must have ' On the history and constituents of the Latin Mass see, in the first instance,

Richard H. Hoppin, Medieval Music (New York, 1978), ch. s. * These are published in Hanna Stablein-Harder, Fourteenth-Century Mass Music in France (Corpus Mensurabilis Musicae, 29, and Musicological Studies

and Documents,

7; American

Institute of Musicology,

1962); Leo Schrade,

French Cycles of the Ordinarium Missae (Polyphonic Music of the Fourteenth

Century, 1, pp. 110-64; Monaco, 1956); and Charles Van den Borren, Missa Tornacensis (Corpus Mensurabilis Musicae, 1 3; American Institute of Musico-

logy, 1957).

* Principally the manuscripts Ivrea, Biblioteca capitolare 115 and Cambrai, Bibliothéque municipale 1328. Their mass contents are listed in Stablein-Harder,

Op. cit.

15

CONSTRUCTION

been composed some time before: it is most unlikely that it was all composed in an unprecedented burst of activity c. 1360. We

can

say

with

reasonable

confidence,

therefore,

that

Machaut was following a tradition rather than creating one. To judge by thé surviving material, it consisted almost entirely of settings of these same six items from the Mass Ordinary. It would seem that by this time composers set in polyphony (or rather were required to set) only those texts, normally sung to plainchant, which recurred without variation throughout the liturgical year. The texts of the ‘Proper’, which changed from day to day, seem to have been almost entirely ignored. It may follow that these pieces were composed for general use and not for a specific or annual occasion; that they; were not tied to coronations or enthronements or individual saints’ days, but could be used at any service of sufficient solemnity’to merit polyphony or whenever a donor had paid for it. It is important to understand, also, that this

general repertory is copied in the manuscripts as individual items, or ordered by text—all the Glorias together, all the Credos, and so on—and not grouped together into five- or six-movement sets. The rare exceptions seem mostly to consist of musically unrelated settings of a Kyrie, a Gloria, and so on, brought together for convenience to provide music for a particular institution. They represent, in other words, selections from the general unordered repertory for use in aservice. For these reasons scholars were slow to notice the musical connections which do exist within this general repertory;* not only between settings of the same text (for instance the related

group of Credos, IV 46-48 and 55-57), but also creating pairs of

movements (e.g. the Sanctus IV 79 and the Sorbonne Agnus) and occasionally larger groups (the Kyrie-Credo-Sanctus group, IV 27-IV 52-Apt 27, and the Kyrie-Sanctus-Agnus movements of the Tournai Mass).> Machaut may also have been responding

his to tradition, therefore, in pairing his Gloria and Credo and also

Sanctus and Agnus settings. Nevertheless, the concept of a single style within which all movements of a Mass could be composed seems to have lain some way in the future. Within the fourteenthlations between * The principal study is Roland Jackson, ‘Musical Interre

29 (1957), PP- 54-64Fourteenth-Century Mass Movements’, Acta Musicologica, The letter-number $7. IV 5 Cf. also the Tournai Gloria with the Credo cit. abbreviations are those of Stablein-Harder, op.

16

CONSTRUCTION

century repertory, three broad styles have been distinguished:® a ‘motet’ style, related to the isorhythmic motet; a ‘discant’ style, where a decorated upper voice is supported by accompanying lower voices; and ‘simultaneous’ style, in which all voices move in near (or strict) homophony. We have already seen that Machaut uses the first and last of these. Tradition is of less help in explaining how he arrived at this selection of contemporary treatments, or why he should have chosen simultaneous style for his Gloria and Credo but motet style for the other movements. Amongst surviving compositions there are no clear trends which associate particular styles with particular texts—there are really too few pieces for any pattern to be established—but several writers have drawn attention to similarities between Machaut’s Mass and the set of movements known as the Mass of Tournai; so that it is possible that Machaut’s

model was not so much general tradition as the example of this specific collection. In fact their similarities are striking only in their Credo settings, which are so closely related that one must have been based upon the other,’ presumably the Machaut upon Tournai since the latter is almost certainly earlier.* Several of their shared features (especially the use of ‘link passages’ to separate formal units) are common in the general repertory, however; and it is quite likely that Machaut was aware of this larger tradition when borrowing them. On the other hand his Mass seems to have had very little influence in return.” The piece is found nowhere else in surviving manuscripts, excepting only a single copy of the Ite Missa Est; and those movements for which there are no close precedents (Kyrie, Sanctus, Agnus and Ite) have no clear successors either, at least

° Stablein-Harder, MSD 7, pp. I5—19. ” The details of this relationship are examined in Sect. Des ak * SeeJean Dumoulin and Jacques Pycke, ‘La ‘Messe de Tournai” et le culte de

Notre-Dame 4 la cathédrale de Tournai au XIV* siécle’, in La Messe de Tournai

(Tornacum, 4; Louvain-la-Neuve, 1988), pp. 21 and 41-57. On purely musical grounds this judgement would remain uncertain, however, for it depends upon rhythmic style and notation, which are unreliable guides in a repertor y as conservative as this. Even works which must be late fourteenth or early fifteenth-century, such as Susay’s Gloria or Murrin’s Credo (Stablei n-Harder,

nos. 35 and $3) are hardly more ‘advanced’.

” It may be that Machaut chose not to release copies other than as part of his

collected works, particularly if the Mass were intended for Reims use.

CONSTRUCTION

17

until the English composer Leonel Power had access to Machaut MS E in the early fifteenth century."® Of course there is a limit to how much we can understand about this piece simply by comparing it to others, by siting it in its musical and liturgical context, and by finding for it an historical reason for being. A great work of music survives the obsolescence of all these, continuing to fascinate and reward the listener long after information about them is lost. If we wish to understand—in an intellectual way —how this is so, then we must turn from the ‘external’

evidence

of documents,

contexts

and

traditions,

to

examine the music itself. If even that cannot answer all the ‘Whys’ which the last few pagés have,raised, it may at least help us towards some understanding; of the equally intriguing ‘How’.

2.2 Kyrie 2.2.1

Introduction Machaut’s Kyrie is isorhythmic. That is, it takes as its startingpoint a plainchant Kyrie melody, allotted to the Tenor voice and set to a repeating rhythmic pattern. Because the chant already

exists—and the composer simply borrows it—and because of the strictly repeating pattern to which it is set, the construction of this voice has to precede the composition of the polyphony which will surround it. It would be impossible to write the other voices without knowing exactly what the Tenor contained. On the other hand, the kind of polyphony which the composer wished to write could very well determine the rhythmic treatment of the plainchant melody and perhaps also the choice of the melody itself, provided that that was not already laid down by liturgical 118 and 141 in © Compare with Machaut’s his paired Sanctus and Agnus, nos. Power had studied the Old Hall MS—a similarity which seems to suggest that also lie behind the Credo and Gloria the that e Machaut’s settings. It is just possibl by Ciconia and pairs —Credo Gloria (?) ury related group of early fifteenth-cent Hughes and Andrew (See . Teramo de (more especially) Antonius Zacharias Musicae, bilis Mensura s (Corpu ipt Margaret Bent (eds.), The Old Hall Manuscr 46; American Institute of Musicology,

1969); Margaret Bent and Anne Hall-

of the Fourteenth mark (eds.), The Works ofJohannes Ciconia (Polyphonic Music

Fischer and F. Alberto Gallo Century, 24; Monaco, 198 5), nos. 3 and 4; Kurt von Monaco, 1987), nos. 3 and 13; (eds.), Italian Sacred and Ceremonial Music (PMFC

4.)

18

CONSTRUCTION

requirements.'' In tracing Machaut’s compositional procedure, then, we must begin with the chant, with its particular melodic characteristics and with the way in which the composer arranges it into an isorhythmic backbone around which to construct the rest of the piece. 2.2.2

Chant

Masses were grouped according to the importance of the celebration which they marked, and different ranks of feast required different plainchant settings. Machaut’s Kyrie is based upon a mode 1 chant for principal or double feasts.'* Like many Kyrie chants it consists of four sections of music: Kyrie, Christe, Kyrie, Kyrie. (The designations I, II, III, for the three sections which set

the text ‘Kyrie eleison’, though not found in the manuscripts, will be used here for ease of reference.) The first is sung three times, then the second three times and the third twice, with the fourth

section acting as a close to the resulting ninefold structure (Ex. 1). Machaut adopts this arrangement in his polyphonic setting, providing four sections of music, marked clearly in all the manuscripts with the appropriate repeat marks: . | | | . after Kyrie I, .|||. after the Christe, and .| |. after Kyrie II. Ex. 1 Kyrie: Tenor pitches Kyrie I

4

Christe

Kiet

z

> a @ oe a

ame

————

SSS

3

Koc

(cont.)

SSS

eS

ao

a2

0

ao

ee

o oe

& o

=;

eee

ee

a=

ie =

a=?

=

' The reciprocal relationship between upper and lower voices is discussed

further in Ch. 3. See GB-Lbl Add. 23935, fo. 433”, headed ‘in toto duplici et duplici’; Add. 39678, fo. 167’, ‘in precipuis’; and for sources closer to Machaut, F-Rbm 224,

fo. 246", and 264, fo. 72", in the latter specified as ‘in precipuis festis’.

'° This repetition scheme

was noted by Ludwig

(Guillaume de Machaut:

Musikalische Werke, iv (Leipzig, 1954), p. 115, but has been consistently ignored

CONSTRUCTION

19

Kyrie chants tend to show considerable variation in form, but always within a basically three-section plan following the textform Kyrie—Christe-—Kyrie. Thus where the final Kyrie group consists of two music sections, as here, they will usually show some common material; and often a similar relationship exists between the first Kyrie and the Christe, or between both of these and the final group. These generalizations apply with some precision to the chant employed by Machaut. As Example 1 shows, Kyrie I and Christe share more than half their material— the first four and the last eleven notes—so that only their middle sections contrast. However,

the differences in these middle sec-

tions do substantially alter the construction of each chant. Kyrie I, after its opening four pitchés, covers a range of a seventh in its next nine notes, cadencing at the end of its second phrase onto e and then returning to a before the final 11-note resolution onto d. The Christe, in contrast, remains within a range of a third below the a from which, like Kyrie I, it began, moving down to the d of

the final 11-note passage only at the end of its middle phrase. The whole Christe chant thus remains within the range of a fifth. The Kyrie II and III chants contrast with the first two sections, though closely linked to each other, in reversing the a-d progression of Kyrie I and Christe. Kyrie II divides into two phrases, Kyrie III differing only in that it repeats the first. Thus in very general terms the arrangement of chants might be represented as follows: Kyrie

Christe

Kyrie

aaa

ay aa

bbb’

and this form is clearly paralleled in Machaut’s construction.

isorhythmic

2.2.3 Kyrie I Having chosen his chant, the composer’s next step comprised a simple arithmetic calculation. Given that the chant, which was to or in recent performances, in favour of alternating polyphony with chant ninefold a which instrumental arrangements. In view of the huge demands a polyphonic Kyrie followed immediately by the Gloria places upon the singers, to difficult preference for alternatim performance is easy to understand. But it is y. ignore the clear manuscript evidence in favour of continuous polyphon tion Further guidance on performance will be found in the Appendix: Introduc to the Edition.

20

CONSTRUCTION

become the melody of the Tenor voice (or color), had ideally to be set to a whole number of statements of the repeating rhythmic pattern (the talea), the number of pitches in the color had to be factorized to give possible quantities of notes for the talea and of numbers of taleae in the complete color statement. For example, a 28-note chant, as here, could be set to two statements of a 14-note

rhythmic pattern, or four statements of seven notes each, or—a further refinement which Machaut sometimes adopted—three taleae of nirle notes each, plus a final note (supporting the last chord of the movement) outside the isorhythmic scheme. Other theoretical possibilities, such as 14 taleae of two notes each or even seven four-note taleae, were unlikely to be used in practice, by this time, because of the monotony and compositional inflexibility of a very short and frequently repeating Tenor talea. A glance at the edition will show that Machaut’s Kyrie I is in this sense very unusual, for the composer has chosen to arrange his 28 chant notes

into seven taleae of four notes each. The result is an exceptionally short talea whose rhythmic pattern is among the simplest to be found in the isorhythmic works of the period, harking back to the early Ars nova works of Le Roman de Fauvel.'4 (As we shall see, however, this apparent archaism does not extend to the other voices and so is no indication of an early date for the piece.) The musical function of the Tenor and Contratenor voices is to provide a clear formal structure and a secure harmonic foundation on which to build the two upper voices (Triplum and Motetus). The Contratenor has to work together with the Tenor to this end,

and must therefore be constructed in such a way as to complement the Tenor rhythms and melody. The Tenor melody being a given, this means in practice that the Tenor and Contratenor rhythms, and the Contratenor melody, must be constructed together as part of a single compositional process if the result is to be properly coherent. The composer, then, must have set out the Tenor melody (either in writing or in his imagination) and composed a Contratenor color against it, in correct two-part counterpoint, at the same time taking decisions about the rhythms of both voices, since the relationship between their rhythms See Leo Schrade (ed.), The Roman de Fauvel (Polyphonic Music of the

Fourteenth

Century,

1, pp. 2-71;

Monaco,

1956, repr. 1984). For more

on

isorhythmic construction see Leech-Wilkinson, Compositional Techniqu es in the Four-Part Isorhythmic Motets of Philippe de Vitry and his Contemporaries (New York, 1989), esp. ch. r.

CONSTRUCTION

21

would determine where they were able to sound together and where separately (one resting). For instance, Example 2 shows the pitches of Tenor (void notes) and Contratenor (black notes). Clearly these details—the number of Contratenor notes, their placing together with or between Tenor notes, and the contrapuntal relationship between the parts—depend upon decisions already taken about the rhythms of the two voices. Equally, the rhythms of each voice are pointless (there is no reason for the composer to choose these rather than any others) in the absence of decisions already made about the Contratenor pitches and about the contrapuntal relationship of the two parts. In other words, the composition of Contratenor pitches and Tehor and Contratenor rhythms has to be a single process, a juggling act during which the composer has to keep control of the given Tenor pitches, the counterpoint which the Contratenor is to make against them (hence ‘contratenor’), and the details of their two talea patterns, for it is the inexorable repetition of the taleae which determines exactly where chords and individual pitches will be possible and where not. Ex. 2 Kyrie I: Tenor and Contratenor pitches @ = Tenor @ =2 Contratenor

és =

=

03 $sys

oS 23s

ze $2-

a a

a 9

aos er

“ae e

wou-o0

@

=

-o—#9

‘wen-sou

_

-

z—— = 1

ap snu-3e

sn- 9p 9u-Tu

- oq

ap

1-apsnu-de

ea

¥

= 5

9 #0

eo

1 -

7

:

>

eh

ae)

op

y

ef

_» 2?Bg

-ed

-

‘sin

eo,

Po

f-o2p

of

m ‘smid-ues

eB

-V

oe

‘snu - mw -op snj - os

eee SS

Tul -Op snq-os

oe

wt? Wo -

og

03 ut

m

te ur

m

OG

#6

=

5:

- wu

Ii

Eg

=

ge

tds

- ids

oy8 ur n3-14-1ds

we - m-ond C2 62

m

ues

=e:

*fegn,

eet

019-ues

Fe

a1 - 2s

o- o1-as-1w

24

of-ex;

‘siq-ou

Le

Ssiq-ou

(19)

‘ais

wrei-2)

-

wei-21-xep

==

wm)

ump,

ed

sin -ed

=== eS u

oof

snp- os

posse

-

m 2

‘In

=

snj-os a,

edt -

©

‘snu

—-)

om

ap -

0

snp -os

og Vv

ou

=o ——

uo -

snw - Is-sh-]e oof

———

snur - Is

LL)

dda gas D

-

#6, _0f#*% t_# Og 2p Cog, Se2 eee

ee

CONSTRUCTION

34

TW:

“NSO

+

aut

ans

=

fas

___9. 72 oon = e =z As

pe sop eze #6 = —,—-76—_2_s_z— Zz ge

(ow)

i

CONSTRUCTION

aS

the time, many important features—in particular its melismas and cadences—are as often as not avoided. Whereas all the chant’s cadences, excepting only the final ‘patris’, are onto e, Machaut makes regular use of clos cadences onto d. Clearly the harmonic structure devised for his polyphony takes precedence over the melodic content of the chant. Secondly, and related to this, there is no clear relationship between the recurring upper-voice melodic outlines and the form (or use) of the chant. Although content to incorporate the chant’s pitches wherever possible, Machaut is giving precedence always to the abstract formal structure devised to shape his polyphony. If this formal structure, has an underlying rationale, then, it

must lie in its relationship with the text, and it must have been with the text that Machaut began. It is, of course, prose: there are

no recurring quantities or rhyme-schemes to suggest a musical form; but one recurring phrase which might have struck Machaut as significant for his purposes is ‘ihesu christe’, the dominant idea

of the Gloria text, and coincidentally occurring 18 syllables from the end and again 18 syllables from the nearest grammatical break to the mid-point (counting syllables). The special musical treatment given to these words may suggest that they were amongst Machaut’s first considerations. If so, it helps to explain the binary structure which he gave to his setting. Looking for ways to split the text up further, Machaut might

have noticed that the first half of the text, thus divided, consists of 14 syntactical units, the second half of 12—near to equal numbers: Et in terra pax hominibus bone voluntatis laudamus te benedicimus te adoramus te

glorificamus te gracias agimus tibi propter magnam gloriam tuam

domine deus rex: celestis deus pater omnipotens domine fili unigenite ihesu christe domine deus agnus dei filius partis

36

CONSTRUCTION

qui tollis peccata mundi miserere nobis qui tollis peccata mundi suscipe * deprecationem nostram qui sedes * ad dexteram patris miserere nobis quoniam tu solus sanctus tu solus dominus tu solus altissimus ihesu christe

cum sancto spiritu in gloria dei patris

s

This is important because, starting from the assumptions that this will be a largely syllabic setting (which must have been one of Machaut’s earliest decisions) and that the text phrases will correspond closely to musical phrases, it follows that the piece will be characterized by many short and clearly defined musical units; and if the text can be divided logically into halves, linked by their placing of ‘ihesu christe’, then the possibility arises that these musical phrases may also correspond between the resulting halves of the piece. A further coincidence of the text’s structure is that each of these halves breaks down semantically into two approximately equal parts (as shown in the spacings above). Thus:

Btcreo terra wick vines anus Doniine deus 32.5. CJOIOURSE a dete e Ouomigne ie tas ikea

gloriam tuam filius patris miserere nobis dei patris

8 units 6 6 6

where one syntactical unit equals one line in the text laid out as above and one musical phrase in Machaut’s setting. The irregularity of the first and third sections (8 against 6 units) is removed by Machaut splitting up two units of section 3 (at the asterisks in the text above) to make 8 phrases in all. It seems probable, nevertheless, that it was the prevalence of 6-

unit sections in the text that led Machaut to fashion six melodic

—harmonic outlines (‘a’ to ‘f? above) as his essential musical material. The correspondence survives most clearly in sections II and IV (not surprisingly, since these were the sections which Machaut left unaltered at six text units each). Thus a simple scheme of I text unit =1 musical phrase (defined by text, rhythm

CONSTRUCTION

37

and cadence)=1 harmonic—melodic unit is preserved almost unaltered in Machaut’s setting of section II: a b c d_ e

f

domine deus rex celestis deus pater omnipotens do-mine fili unigenite ihesu christe | domine deus agnus dei

bb. 30-4 35-9 39-41 42-6 + 47 48-51

filius patris

52-6

Sections I and II, on the other hand, are far from regularly set;

and it seems probable that Machaut was here concerned to vary the rate of musical changé (by varying the relation between these various aspects of text andymusic),

content that the basic 4 X 6

musical plan would operate satisfactorily, without being worked out with ‘isorhythmic’ regularity. The ‘isorhythm’ analogy is not entirely metaphorical. Clearly the four-section form, each section filled with basically identical material (subject to variation at the level of surface details), is comparable to a four-color isorhythmic construction. And it is interesting to note, also, that towards the end of each section (in ‘e’ and ‘f’ material) melodic figures and harmonic progressions recur more exactly than elsewhere, e.g. in ‘f’ material: 2711-29

I

2611-271

Il

$2

531

Ill IV

75-761 1031

761i

$31 77-80

811

54-56 8111-8 3 1031-104

Whereas ‘a’—‘d’ material tends to be expanded by surface decoration, ‘e’ and ‘f’—as the music approaches the end of each section—tend to be expanded by insertion, the ‘original’ material preserving its harmonic and melodic details (compare in the score the phrases aligned vertically in the table above). It is therefore

more memorable and, like the hocket so often used near the end

of a motet talea, emphasizes for the listener the essentially strophic nature of the form. The ‘ihesu christe’ settings, in their instantly recognizable stretched note values, have the same effect on a larger scale.

2.3.2.1 Gloria Amen Although the Amen makes even freer use of its referential chant than does the main body of the movement,

the basic melodic

38

CONSTRUCTION

outline of the plainchant Amen appears to be echoed in the d—g—f-e—d figures of Machaut’s tenor (cf. Ex. 6). And although in rhythmic style it is indistinguishable frdm the isorhythmic movements, the Gloria Amen is not, in fact, strictly isorhythmic in any part. The piece is constructed not upon atalea but rather in two contrasting sections, the first made up almost entirely of breves and longs (arranged, especially in bb. 6-11, in sequence in each part), while the second (bb. 15—26) contains many shorter values, often arranged into syncopated and hocketing figures, so that together the two halves of the piece seem) to refer to the integer valor and diminished sections of an isorhythmic motet. Also unusual in a work isorhythmic in style is the Contratenor which, against all convention, has the most active rhythms and therefore the largest number of notes of the four parts (CT 74 notes, Tr 62, Mo 53, T 34); and it is the Contratenor which creates

almost all of the many dissonances. In these respects it is much closer to the contratenors of Machaut’s songs—and particularly to some

of those

composed

in 1363,

as part of the

Voir Dit

correspondence—than to those of his motets.”4 As Example 5 suggested, it is just possible to read the melodic structure of the Amen in terms of that of the main part of the movement,

and so to see the Amen

as a fifth stanza.

immediately evident is the link between

More

Gloria, bb. 1-4 and

Gloria Amen, bb. 1-5, and, at the other end of each section, between Gloria, 100-4 and Amen, 21-6, which seem to point toa

strong architectural sense of each section as a whole.

2.3.3 Credo Before comparing the construction of the Credo with that of the Gloria it is important to look at its relationship with the anonymous Credo of the Tournai Mass. Although this is a much less impressive setting than Machaut’s it was obviously popular, for it survives in four sources,

three of them from the South

where it may have originated earlier in the century.” As well as “ See especially Ballade 34, Quant Theseus/Ne quier veoir, and Rondeau 17, Dix et sept. For further connections between the Mass and Voir Dit songs, see

; Sect. 4.9. * The sources are B-Tc 476, E-Mn 8, E-BUlh and F-APT. Southern origin is

assumed by Anglés (‘Una nueva versién del Credo de Tournai’, Revue Belge de Musicologie, 8 (1954), 88), Schrade French Cycles, Commentary, p. 131) and Van

den Borren (op. cit., p. III), although there is no decisive evidence . Dumoulin,

CONSTRUCTION

39

the homorhythmic style, the most immediately striking similarity is both composers’ use of reduced-voice ‘link passages’ to define units of form—features which of course Machaut also adopted for the Gloria. In fact, Tournai need not have been Machaut’s source for these: similar passages are common in the general repertory of fourteenth-century mass music;”° so that in using them Machaut was probably referring to an existing tradition. The same is probably true of the opening Triplum figure e’—f"—e’—d’—c’ which is found in several other Credo settings.*’ There has been much dispute as to whether the Machaut and Tournai Credos paraphrase chant.” There are undoubtedly occasional references in both

settings

to the chant

now

known

as Credo

1, most

obviously at the e’—f’—e’—d’-“’ openings,” but also in the Machaut

at T: ‘Christum’ CT: ‘Filium dei’ T: ‘-i unigen-’ (Cr, bb. 21-5) and more approximately elsewhere (e.g. Tr: ‘homo factus est’ bb. 77-9). No consistent paraphrase need be hypothesised, however. Any composer would have been so familiar with the plainchant Credo I that in setting the same text he would inevitably fall into some of the same melodic figures, indeed, it might be surprising if he did not. It seems probable that comparable familiarity lies behind Machaut’s references to the Tournai setting. There is no consistent use of the Tournai material, but its melodic figures and harmonic progressions surface unpredictably throughout (compare, for example, their settings of ‘Patrem om-’, ‘visibilium omnium’, ‘sum Christum’, ‘lumine’, ‘de deo vero’, ‘et incarnatus est de

Huglo and Pycke have proposed that the Tournai Mass was used in Tournai a date cathedral at celebrations of aMary Mass founded in 1349; which supports ’s Machaut for date d suggeste the for its constituent movements well before setting (see Ch. I, n. 44). Ct. 6 See, for example, Stablein-Harder, op. Cit. NOS. 23,28, 33, 30 n-Harder, (Stablei 1328 27 Compare the Tournai Credo, that in Cambrai (Tenor). 44 no. 39) and—at a different pitch level—nos. 42 and (London, 28 See Manfred Bukofzer, Studies in Medieval and Renaissance Music

1951), pp. 52-3, for the strongest objections. in 2 This chant was so well known as to be recorded relatively rarely Lbl paedic manuscripts of the period, although it is included in the encyclo Vatican Add. 23935, fo. 434°, where it is notated a fifth above the modern referred. version. It seems to have been to this ‘higher’ version that Machaut from clear is It 5544, 42, 40, 39, nos. cit., op. r, n-Harde 3 Cf. also Stablei tary to the these that the chant was known at both pitches. (See also the Commen Credo, below.)

40

CONSTRUCTION

spiritu sancto’, ‘nobis sub pontio pylato’, ‘patris’, ‘gloria’, ‘-ficatur’ and ‘per prophetas’).Sometimes they are a tone lower in the Machaut (e.g. ‘“-potentem’, ‘-torem ce-’ and ‘qui propter nos homines’), and sometimes over different text (compare Machaut’s ‘In spiritum sanctum dominum’ with Tournai’s ‘et iterum venturus est cum gloria’). All this suggests that Machaut was falling into similar phrases through the habit of associating text with familiar music, rather than that he was systematically reusing the Tournai model. But despite the substantial number of similarities most of Machaut’s setting is entirely new, including all the most striking melodic and harmonic gestures. The strophic form, of course, is all his own; and it must have been with that that his conscious

compositional planning began. The three sections of Machaut’s Credo are clearly marked off in the manuscripts by ‘double-bars’ followed by a page-turn;” which probably implies that a pause, however brief, was intended after each, making the larger structure of the movement very

clear. Within each of these three sections, three subsections are

marked off by ‘link passages’ placed about one-third and twothirds of the way through. All these formal divisions are explicable in relation to the structure of the text. Patrem—> omnipotentem factorem—> celi et terre visibilium omnium—> et invisibilium

[I. 1]

[link] et in unum—> dominum ihesum christum

[I. 2]

filium dei unigenitum> et ex patre natum ante omnia—> secula

[link] deum de deo lumen de lumine> deum verum de deo vero> genitum non factum—> consubstantialem patri—>

[I. 3]

per quem omnia facta sunt *' MS Gis exceptional in having page-turns in the middle of sections.

CONSTRUCTION

[page-turn]

4I

@

Qui propter nos homines et propter nostram salutem “descendit de celis— et incarnatus est de spiritu sancto ex maria virgine et homo factus est

[link] crucifixus etiam pro nobis sub pontio pylato> passus et sepultus est et resurrexit>

tertia die

,

{II. 1]

(Il: 2]

§

hi

secundum scripturas>

et ascendit in clum—> sedet ad dexteram patris [link] et iterum venturus est cum gloria—

_[II. 3]

iudicare vivos et mortuos cuius regni non erit finis [page-turn] Et in spiritum sanctum dominum- _ [IIl.1] et vivificantem qui ex patre filioque procedit— qui cum patre et filio simul adoratur et conglorificatur> qui locutus est per prophetas

[link] et unam sanctam catholicam et apostolicam ecclesiam confiteor unum baptisma

[IlI.2]

in remissionem peccatorum

[link] et expecto resurrectionem

mortuorum et vitam venturi seculi> [page-turn] Amen

[III.3]

42

CONSTRUCTION

The three main sections, as defined by Machaut’s setting, contain similar quantities of text (107 : 132 : 118 syllables), and are logically self-contained in meaning. This latter seems to have been the overriding requirement, since Machaut could have arrived at sections of more nearly equal length (128 : 111 : 118 syllables) had he started section II at ‘et incarnatus est’, though to do so he would have had to pass over the most logical text division. His choice of subsection divisions must have been directed by a similar search for logical breaks in the text approximately onethird and two-thirds through each section. The only clear departure from this criterion is the placing of the link-bar between II. 2 and II. 3 after ‘dexteram patris’ rather than ‘secundum scripturas’; but again the determining factor seems to have been the very clear change of subject in the text at ‘et iterum venturus est’. Evidently the message of the Credo was fundamental to Machaut’s formal decisions.” Similarly, the harmonic structure of his setting seems to be wholly dependent upon the sentence-structure of the text. The arrows in the text as printed above show where Machaut sets an imperfect cadence, necessarily leading on into the next phrase in search of resolution. It will be seen that in nearly every case these coincide with a grammatical continuity in the text. This suggests that cadential patterns recurring from section to section are likely to be the result of coincidences in the grammatical structure of the text and not of any abstract formal structure imposed upon it. This perhaps explains why Gombosi’s labellings of phrases are so unconvincing: his I.1.a (bb. 1-5), II. 1.0 (bb. 55-61) and III. 1.a@ (bb. 116-21) really have nothing in common beyond their cadence resolutions. In contrast to the Gloria, then, the Credo does not make great

use of recurring musical material through each section. There are

a few recurrences ofa similar sort (cf. bb. 25—33, 85-93, 146-50 or bb. 30-2, 45-8, 90-3, 104-5, 110-13, 129-31, 148-50, 159-62, and also the concluding passages of subsections 1 and 3: bb. 13-1 5; $14, 77-9, 112-15, 132-4), but these account for less than half the material of the piece, and tend to cluster towards the end of each section (another isorhythmic analogy?). On the whole, Machaut * That he chose.to ignore half the link-passage positions in the Tournai Credo may be due to this concern for continuity of text, although it might also relate to a wish for the larger spans of uninterrupted music which make his strophic form possible.

CONSTRUCTION

43

seems to have been more coycerned with harmonic direction— with how he is approaching and resolving each cadence—than with regular repetitions of melodic and harmonic shapes. In this sense, then, the music of the Credo is less formalised than that of the Gloria, but also more closely allied to the words; and if

the movements were written in their performed order (and probability must favour that assumption)*’ this may reflect an evolving view of the purpose and potential of this homophonic and metrically flexible style of text-setting. Machaut’s use of metrical flexibility is worth a moment’s study. It is a feature of both these movements that amongst the duple values which predominate he from time to time inserts passages of one or more three-beat units.** These can appear for several reasons: at Credo, b. 25 he takes an extra beat in order to allow the Triplum descending figure to give birth (unigenitum—‘begotten’—is the word set) to an identical figure in the Motetus; in bb. 86-95 he gives emphasis to the crucial phrases ‘et resurrexit . .. tertiam die ... secundum scripturas’ by lengthening the final chords, whereas the two preceding phrases, although occupying three beats for their first six syllables, conclude with two-beat finals (producing five-beat phrases) to keep the music moving towards the more important text to follow. According to contemporary notation rules this is indefensible, and as such is an endearing indication of the composer’s determination to create as flexible a means of text-projection as his ear could imagine.

2.3.3.1 Credo Amen Despite some similarities between its opening and that of the Gloria Amen, the tenor melody of this section seems not to be derived from chant; for although it begins and ends on d, suggesting the mode 1 of Kyrie, Gloria and Credo, it is otherwise focused firmly on c, anticipating the mode 5 of the Sanctus and Agnus at the same time as looking back to the alternation of d and c chords which underlay the harmonic structure of Kyries II and III; and in this sense the Credo

Amen

serves as a central link

between the two modal halves of the Mass. That these references forwards and back were deliberate is further suggested by the 33 See also Sects. 2.5 and 4.8. 4 Gloria, bb. 12-17, 38, 87-9, 92, 101-2; Credo, bb. 25, 86, 88, 90-5, 104, 116, 120, 124, 133, 146, 149, 154, 159-61. There is, of course, room for

disagreement as to how these should be barred in a modern edition.

44

CONSTRUCTION

apparent quotation of the outline of the Sanctus chant at the words ‘dominus deus sabaoth’ in bb. 23-31 of the Credo Amen

(transposed to c). The rhythmic form of this piece is-no less interesting, though much more conventional, than that of the Gloria Amen. Essen-

tially the movement is panisorhythmic; that is, its upper voices repeat in strict isorhythm along with the Tenor and Contratenor; and it is constructed upon lower-voice taleae which—like those of Machaut’s last motet (no. 23, Felix/Inviolata)*>—exchange parts at their centres. As may be seen in Example 7 (where one-bar rhythmic units are represented by letters), the upper voices also contain repetitions within each talea; and the statements of these repeating phrases overlap the ends of the larger taleae—isorhythm within isorhythm.*° Ex. 7 Credo Amen: rhythmic patterns Key:

a oo ioe

Tr Mo

BieRear Br AlNWis Meal eM ieneae ue, ae apa Sse Barna rae Dg |eens ee aes a 4b Derceh det ccserait. D by hes a ne f

Deg aa Cr:

ae at pat ther Dune eigen iin atTha iy

Ree

Gh as

x,

BS ED na aoa ge iewe

eeu

ee eet

re

ies: J

J

sides pe nee ae roan hte b dd dd odd) db

Only one irregularity disturbs this tightly knit structure. The inevitable result of the exchange of Tenor and Contratenor rhythms in each talea—as in all the motets where this occurs—is that in one or other half of the talea two Tenor notes will have to be harmonised by one in the Contratenor. In the case of the Credo Amen, where the Tenor melody seems to have been composed by Machaut, this is a problem which could normally be avoided simply by ensuring that at these points the Tenor notes * For precursors see Leech-Wilkinson, Compositional Techniques, pp. 132-3. Motet 23 proves to be closely linked with the Mass—see Sect. 4.9 below. * Possibly the earliest example of this in the motet repertory is the anonymous Rachel/Ha fratres (ed. Frank Harrison, Motets of French Provenance (Polyphonic Music of the Fourteenth Century, 5; Monaco, 1968), no. 11, p. 66) Without precise dates it is impossible to say where it originated.

CONSTRUCTION

A5

in question were always able to be harmonized bya single Contratenor pitch. Where this cannot be arranged, however, is in the quotation from the Sanctus chant, and without further modification the pitches a—g (bb. 27-8), which would strictly have fallen in b. 28, would have had to be accompanied by one note in the Contratenor. This is perhaps the reason why Machaut exchanged the Tenor and Contratenor taleae in the final taleasection, causing these two Tenor pitches to fall conveniently against different Contratenor notes. In addition Machaut ,has taken the opportunity to make a further change, in b. 35 of the Contratenor, replacing the imperfect long with two breves, despite the resulting dissonance, in order to improve the voice-leading into the final cadence.” Although based on what appears to be a newly composed color—perhaps because of this—the Credo Amen is in many ways the strictest musical construction in the Mass. The relative freedom from pre-compositional restraints which in the Gloria Amen produced a quasi-isorhythmic reworking of the parent movement’s basic material, very freely used, is here taken rather as a challenge to Machaut’s powers of strict musical organisation for which the (perhaps relatively new) technique of panisorhythm provides an ideal medium.

2.4 Sanctus The Sanctus chant, a mode 5 setting for solemn or semi-double feasts,® has a remarkably convenient structure for isorhythmic setting, which may be why Machaut chose it. It consists of seven phrases of music (I, 2, 24, 2b, 2c, 3, 4) ordered as in Example 8.

Phrases 2, 2a, 2b and 2c provide, by means of their common

material, a refrain between phrases 1, 3 and 4. Phrase 4 could be

seen as an expansion of phrase 1, emphasizing the economical construction of the chant. The Osanna and Benedictus consist of material already stated in the Dominus Deus, leaving the Sanctus statements outside the scheme of the remainder of the movement. 7 Though this is a change which Machaut might not have felt free to introduce into the Tenor had the taleae not been exchanged, the Tenor being traditionally the more inviolable of the two lower voices. 3 Solemn in F-Rbm 224, fo. 247° and 264, fos. 75‘-76', semi-double in GBLbl Add. 23935, fo. 434" and Add. 39678, fo. 173°”.

ee

snjours

SS

l

euuesc,

ee

:

?

Il|

ee (ee

ae ee

I]

NE ee on

0%

snurwog snaq Se =i; 0g

OO 4

Al|

os

SS

J

ee 1

o

snmjsueg

4k

22

ee

J

IA|

J

See

€

a z=

I

ee

=

97 —o=z-

UT

a

25

suTwou

o-2 Zz

l

IIA

—

v

Al

smorpouog]| |,

oe

I

snjoues

9% wae 4

oo eee

SsNyIURSJURYD

l o-2 Zz

Zoe

a Xt

@

XT8

46 CONSTRUCTION

oe aos

x|

euUesO XI

CONSTRUCTION

47

@

And this formal arrangement is matched exactly in Machaut’s polyphonic elaboration. The quantities with which Machaut worked are as follows: Sanctus Dominus Osanna Benedictus Osanna

7+6+7=20 chant notes 45 22 22 22

Since the Dominus Deus 4s but one note more than twice the length of each of the remaining three sections, Machaut’s iso-

rhythmic scheme of ten i1-note taleae plus a final note would seem to be almost inevitable, allowing each of the last four sections of the movement to consist of a whole number of taleae,

thus respecting the formal structure of the chant. Only the opening Sanctus invocations cannot be accommodated within this scheme, and Machaut treats these as a non-isorhythmic introduction, reflecting their function as a melodic introduction to the chant. The only variation which Machaut introduces into this remarkably convenient scheme is the transference of the extraisorhythmic final of the Dominus Deus to the end of the first Osanna,

so that the music can arrive at a full structural close

before the liturgically prescribed ringing of the Sanctus bell, after which the Benedictus follows.” Thus: Talea:

Lp

CnancCaotes

alt?

fi

ele:

EVV

11°

STF

110

(= 45) Text:

Dominus

Wie

+a IL ete:

tu.

or ie

ete:

(= 22)"h Osanna

|Benedictus

The division of the color into 11-note groups, while respecting (or exploiting) the quantities of the chant, does not correspond to any natural melodic grouping within it; and given the large number of talea statements there can be no question of matching rhythmic and melodic cadences in the Tenor with any consistency. Nevertheless, in constructing the lower-voice taleae Machaut seems to have been working towards rhythmic cadential figures in very much the same way as in the Kyrie. The Tenor and ® In order to make a satisfactory cadence here, Machaut omits the rest at the

end of talea VI which would otherwise separate this final from its preceding pitch.

48

CONSTRUCTION

Contratenor define two rhythmic phrases, whose patterns are very similar to those of Kyrie II and II, the first leading to a long in the fourth bar of the talea, the second leading to the seventh

(the penultimate bar) after which a Contratenor link leads on to the start of the next statement, exactly as in Kyrie I. Also as in Kyrie II and III, the second phrase is matched in the upper voices by isorhythmic hocket/syncopation figures, while the first is rhythmically varied from talea to talea. Significant differences between the Sanctus and Kyrie only begin to appear when we turn to look at Machaut’s treatment of the Contratenor color in relation to that of the Tenor and at the resulting harmony. The lower-voice pair offers a remarkable example of skilful voice-leading, due to the way in which Machaut has been able to exploit apparent contradictions between rhythmic and harmonic cadences, often by means of highly unconventional progressions, in order to produce harmony directed over longer phrases than are to be found in the work of any of his contemporaries. Bars 20-6 provide perhaps the clearest example. The Tenor progression from e’ to f’ (bb. 22-4), which might normally be expected to provide a cadence, is interrupted by the rest at the end of the talea through which Machaut _ prolongs the e’ chord into the resolution of the beginning of talea II, thereby turning to good effect the apparently disruptive position of the rest. This extended progression is itself prepared by a sequence of chords which, by breaking contemporary rules of harmony, serve to link the previous c’ unison to the cadence onto f- By means of these dissonant chords Machaut has created a directed progression from c’ to f’ spanning no less than eleven beats, and this itself runs into a progression back to c’ (bb. 24-6). The whole sequence is repeated with the repetition of this segment of the chant at the first Osanna (bb. 48-52) despite the fact that here the progression falls over different notes of the talea.*” There is no question here of the rhythmic situation forcing Machaut to use lower-voice fourths and seconds: the Tenor and Contratenor taleae are so simple that such structural dissonances could easily have been avoided. Machaut is using them solely for the role which they are able to play in joining more consonant chords into a single extended progression. As Chapter 3 will * Note also the similar use of the Tenor rest at bb. 46-8, 54-6. A similarly

elegant progression is to be found in the Benedictus (bb. 76-8), in this case from (Contratenor) f’ to c.

CONSTRUCTION

49

e

suggest, the consequences of this liberation of the lower voices from the limited vision of contemporary counterpoint prove to be far-reaching.

2.5 Agnus Dei The mode 5 Agnus chants chosen by Machaut belong to the same cycle as those of the Sanctus, being constructed from related melodic material. Indeed,/the Agnus I chant might be expressed in terms of the Sanctus melodic phrases (as numbered in Ex. 8) as 1*+2+4, although the first is considerably altered. Like the Kyrie chants those of the Agnus present a simple formal plan:

Agnus

a. boa’

emibea’

aba!

I

Il

I

(=)

the third Agnus being a repeat of the first. In addition, although the first eight notes of Agnus I= III and Agnus II are different, the remaining 26 are identical; and in Agnus I=III the last 11 notes are a varied repeat of the first nine. In both settings Machaut has arranged these first nine notes, carrying the essential text ‘Agnus Dei’, as a non-isorhythmic introduction comparable to that of the Sanctus, further emphasising the links between the two movements. The isoperiodic plans chosen by Machaut for these two movements are closely related: Intro

Taleae

Final

Agnus I= III

Notes Taleae

9

12 I

12 II

Agnus II

Notes Taleae

9

Ad, (he oc bee We

AeA, tN

I ds NT

AE

as are their taleae; though the short taleae of Agnus II contrast sufficiently with the longer sections of Agnus I=TIII for the function of the second Agnus as the contrasting middle section of an ABA form, matching the ABA of the chants, to be clearly audible in performance. Although the lower-voice talea-patterns of the Agnus use similar rhythmic figures to those of the earlier isorhythmic movements, they differ significantly in their greater use of short rests. This produces a faster rate of harmonic change which in

$0

CONSTRUCTION

turn allows a smoother melody line (since the upper voices are not required to maintain a single chord over so long a period). This continues a trend towards increasing lyricism seen already in the Sanctus and before that in the Credo

Amen,

so that if these

movements were not composed at intervals over a period of time it would be difficult to avoid the impression that Machaut was introducing a sequence of techniques of composition which would produce a graded change in style throughout the work from the earliest movements, dominated by an ordered harmonic structure, up to the latest where melodic lyricism is the primary aim. Certainly the contrast in style between the simple rhythms and relatively angular lines of Kyrie I and the double syncopations (at different levels in Triplum and Motetus) and free lyricism of Agnus I is striking, and symptomatic of a significant change in priorities and in compositional technique. Agnus II relates closely to the first in its use of a six-bar invocation followed by a talea using similarly active rhythms to those of Agnus I. Here, however, the talea is both panisorhythmic and extremely short, its six statements producing a litany-like effect which recalls the repetitive nature of the text, and provides an effective contrast to the less audible form of Agnus I=III. The alternation of Tenor and Contratenor pitches seen in Agnus | (caused by the many short rests) is here carried to an extreme: the two parts meet only at the single cadential progression, so allowing a maximum of harmonic and melodic freedom despite the rigidly repeating upper-voice taleae. Machaut makes good use of this, but, as if to compensate for the resulting contrasts, draws

the upper-voice phrases together at each cadence by using the same descending melodic figure (c”—b’—a’—g’) in the first four statements

(bb. 9, 12, 15, 18). The

comparison

with Kyrie I,

which used similarly short taleae, is again revealing. Despite the large amount of common chant material between Agnus I= III and Agnus II their harmonisations are very different, largely because of the contrasting rhythmic context provided by the very short Agnus II talea, requiring frequent cadences. Thus, for example, the e’~f' Tenor progression in b. 11 of Agnus I= III is placed in a larger sequence of lower-voice chords moving towards the 6 of b: 14, whereas in Agnus II the f’ has to function as the resolution of a principal cadence (bb. 12-13). Conversely the Tenor pitches f-g representing a cadence in Agnus I=III (b. 17) are less important elements in a larger progression in

CONSTRUCTION

st

@

Agnus II (b. 20). On the other hand, where the rhythmic contexts

are identical so are the harmonisations, as for example at the central cadence onto 6 (Agnus I=III, b. 13-14, Agnus II, bb. 15-16) and of course the final cadences. The isorhythmic sections of the two settings are thus begun, centred and ended identically, though elsewhere they contrast, and the identical placing of the therefore prominent central cadences may perhaps have determined Machaut’s choice of these related isoperiodic arrangements rather than one or more of the several alternatives available (e.g. 9 + (8 X 3) +1 orgt (5 X §)). In view of the longrange harmonic planning’ démonstrated by Machaut in earlier movements this is perhaps not unlikely. On the smaller scale, the use of lower-voice dissonance as a

directing element within a larger progression is seen here as in the Sanctus (Agnus I=II]I, b. 12), though now supplemented by a no less extraordinary development, in the form of implied harmony—the implied prolongation of lower-voice pitches through their following breve rests. Thus in b.9 of Agnus I the Contratenor e is the implied bass of the second chord of the bar, providing the ‘third’ of the chord, and leading on to the d of the last chord. Similarly—and perhaps more radically—in b. 12 a Contratenor g is implied in the bass as part of the larger progression from f to c’. Likewise in Agnus II, b. 12 the Contratenor a is implied through the second beat, moving down to the g at the end of the bar and so onto the cadential f of b. 13. In each of these cases Machaut is choosing not to exploit the possibility offered by the Contratenor rest for a change of chord but rather to imply in retrospect (by means of the chord of the following beat) a continuation of the previous chord through the silence. The result may, of course, be fortuitous, but together with the

evidence of an unusually ordered approach to harmony seen in these and earlier movements it is perhaps not unreasonable to point to an exceptional interest in and appreciation of the way in which chord relationships are perceived by the listener, an interest which seems not to have been shared to quite the same degree by any of Machaut’s contemporaries.

2.6 Ite Missa Est The

chant

which

Machaut

used

for his Ite is recorded

in

52

CONSTRUCTION

contemporary manuscripts as being as appropriate for principal

feasts,’ and differs somewhat from the chants of the previous movements in that it belongs to mode 6 rather than to mode 5. Machaut’s treatment of it, however, links it firmly to the Agnus

settings. While the choice of two 10-note taleae plus a final seems to have been determined by the natural division of the chant after the tenth note, the rhythms of all voices are close to those of the Agnus; and the suggestion that the two rhythmic halves of the upper-voice talea, defined by the strictly isorhythmic hocket/ syncopation of the last four bars, refer back to the free introduction and isorhythmic sections of the Agnus movements is perhaps validated by the similarity of the cadence which closes the first section (bb. 3-4) to those which close the introductions of both Agnus I=III (bb. 5—6) and Agnus II (bb. 4—6). This same division of four bars of standard rhythms plus four bars of strictly isorhythmic hocket/syncopation is reflected in the harmonic plan of the movement—bb. 1-4 repeating almost exactly as bb. 9-12 while bb. s—8 and 13-16 contrast—which

is itself in part the

product of the similarity between the openings of the first and second halves of the chant. The piece is thus a binary structure not only in overall form, but also within each talea and in its harmonisation, so that the form of the chant is mirrored in all

major aspects of the piece, exactly as in the case of the ABA form of the Agnus. As in the Sanctus, Machaut uses a sequence of lower-voice dissonances to create a larger progression (bb. 6-8) and, in a second unconventional passage (bb. 11-12), delays the resolution of a similar progression onto f by introducing an extra c chord between preparatory dissonance and resolution in order to avoid a full bar based on the final of the mode—a situation acceptable in the corresponding b. 4 because of the introductory function of bb. 1-4 but not here where the music must lead on into the closing phrase of the piece.

2.7 Conclusion This chapter necessarily contains a forbidding concentration of detail; yet it is only from the evidence of bar-by-bar decisions that significant conclusions about Machaut’s musical priorities can be *' See GB-Lbl Add. 39678, fo. 179).

CONSTRUCTION

$3

drawn. We have seen that all the movements of the Mass—even the non-isorhythmic Gloria and Credo—are constructed around carefully arranged forms. Indeed, it is precisely because these forms are so rigidly applied that reconstruction of Machaut’s compositional procedure becomes possible. The isorhythmic movements maintain such a tight relationship between their component parts (chant, lower-voice rhythmic patterns, lowervoice counterpoint, upper-voice rhythms and contrapuntal elaboration) that the order in which they were shaped and assembled, and the ways in which details at any level depend upon decisions made earlier, can be determined with some certainty. In the case of the Gloria and Credo, ‘observing the way in which form is defined by repeating melodic and cadential patterns leads us back to a reconstruction of the procedure by which Machaut generated forms from the semantic structure of the texts. In both cases the technique is perfectly chosen: the isorhythmic movements, having little text, are structured according to musical—formal criteria based ultimately upon their chants; those movements having a lot of words to be set are shaped not around chant but according to the forms of their texts. Within these different approaches a number of trends have been revealed. We have seen that in several respects (most importantly, rhythm and chant treatment) Kyrie I is unlike the rest of the first movement, suggesting that Machaut’s plans changed somewhat before he proceeded. And we have seen that there is a significant difference between the lower-voice counterpoint of the Kyrie and that of the later isorhythmic movements. On the other hand, their rhythmic procedures seem to be very similar; and this must restrain any theory of chronology. Similarly, clear differences have been seen between the procedures of Gloria and Credo, the latter being shaped less by musical recurrences and more by the form of its text. More evidence about all these differences will erherge in Chapters 3 and 4. But even at this stage of the investigation it must be clear that we are dealing with a composer of great skill who knew exactly what he was doing, and who was doing it in a more sophisticated way than has generally been expected of a medieval musician.

sues ELABORATION 3.1 Introduction Chapter 2 was mainly concerned to reconstruct Machaut’s early decisions about form. Most of its evidence came from the lower two voices and in particular from their rhythm. Only in the case of the Gloria and Credo did evidence from the upper voices seem crucial to the planning of a movement. Consequently very little has yet been said about later stages in the composition process, about the filling-in of upper-voice pitches and about the way in which they elaborate the lower-voice framework, producing the complex four-part whole which we hear in performance. As far as isorhythmic construction is concerned, Chapter 2 has provided a model for the examination of formal decisions and also for matters of detail in so far as they depend upon rhythm. The same technique can easily be applied to the upper voices. Rather than go through the same process again, this chapter aims to deal with the more interesting question of how the result is ordered, within a movement, in terms of melodic and harmonic structure. Some

of the answers may in turn shed new light on Machaut’s compositional procedure. It is impossible, though, to get very far without first confronting a question which has occupied musicologists for almost as long as they have worried about the construction of medieval music, namely the way in which composers imagined counterpoint.

3.2 Four-part conception and composition 3.2.1 The problem We have seen that in the isorhythmic movements Macha ut’s construction of the four-part whole must have proceeded in several stages. Because the Tenor pitches were borrowed from existing chant, and because the rhythms of both Tenor and Contratenor repeated regularly, the two lower voices must have been complete before details of the upper voices were compo sed. Machaut cannot have first conceived the full texture and then

ELABORATION

55

worked isorhythm into it: the result could never have been so regular. At the same time, it is possible to exaggerate the extent to which this process had to be made up of successive stages. While it might theoretically be possible to construct the lower voices by first taking the chant, then writing rhythms for it, then rhythms for a Contratenor and only then thinking about the Contratenor’s pitches, the result would necessarily be crude. Because these elements are interrelated, the most effective compositional pro-

cedure would be one in which their various needs were balanced against one another until,an efficient compromise was reached. Thus the composition of Tenor and Contratenor rhythms depended to an important dégree on the kind of counterpoint the composer wished to make between them. Pitches were only possible where rhythms allowed, but rhythms were only useful where pitches worked well. The juggling of these factors was, as we have seen, best performed as a single operation, one in which

the composer switched back and forth between decisions about Contratenor pitches and decisions about Tenor and Contratenor rhythms. Once the lower-voice duet was complete, one would expect a similar process to be applied to the composition of Triplum and Motetus, balancing rhythmic possibilities against requirements of good voice-leading. The difference, of course, is that the com-

poser is now having to test each possibility against three other parts, and not simply against the other of a pair, as was the case in the duet with which he began. But to what extent was this possible for a medieval composer? Should we deduce, from the paucity of precisely aligned score notation surviving from the period, that composers did not habitually think in vertical simultaneities; or—on the contrary—does this imply that their mental control was so complete that score notation was unnecessary? Or—and this may seem the most reasonable assumption— was score used for composition, but on erasable or disposable material, wiped or thrown away once the music had been notated (in parts) onto parchment? A number of untenable assumptions (for instance, that theorists

describe composers as working a part at a time,’ or that com-

! The evidence is set out in Leech-Wilkinson, Compositional Techniques in the Four-Part Isorhythmic Motets of Philippe de Vitry and his Contemporaries (New York, 1989), pp. 18-24, where it is argued that, according to his own testimony, the chief witness for this period, Egidius de Murino, was writing for beginners and not describing the practice of skilled composers.

56

ELABORATION

posers were unable to manipulate in their heads more than two parts at once” have in the past tended to cloud the answering of these questions. Nevertheless, the central question of to what extent composers worked horizontally, on one part at a time, and to what extent vertically, on all four together, requires an answer. It may be a different answer for each composition; but it must in each case influence our understanding of the extent to which a med‘eval musician heard music as an interplay of individually coherent lines and the extent to which it struck him as an ordered sequence of chords. ; The isorhythmic movements cannot help us here, for they are necessarily constructed in stages. But the Gloria and Credo may be more informative. If we can show that, despite their homophonic, chordal texture, they were constructed a part at a time then we shall be able to say that the composer’s priorities are linear—lines matter more than sonorities—and we can hear the more obviously contrapuntal isorhythmic movements in that light. But if the Gloria and Credo prove to have been conceived as a four-part whole, then our view not only of them but also of the isorhythmic sections will be changed. The melodic lines at the top of the texture may seem more important than any other; and the possibility may arise that the isorhythmic scheme and the lower-voice duet were arranged with melodic design and particular harmonic progressions already in mind. 3.2.2 Gloria and Credo If the four voices of the Gloria and Credo have consistently

independent functions; if, in other words, the Tenor behaves as a referential line for Tnplum, Motetus and Contratenor, each

making correct counterpoint with it (if not necessarily with one another), if the Contratenor behaves as a conventional contratenor, making counterpoint against the Tenor, and so on; then it may be said that the parts could have been written successively. * For counter-arguments see Richard Crocker, ‘Discant, Counterpoint, and Harmony’, Journal of the American Musicological Society, 15 (1962), 1-21; Daniel Leech-Wilkinson, ‘Machaut’s Rose, lis and the Problem of Early Music Analysis’, Music Analysis, 3 (1984), 9-28; and BonnieJ. Blackburn, ‘On Compositional

Process in the Fifteenth Century’, Journal of the American Musicological Society, 40 (1987), 210-84.

ELABORATION

$7.

At least, there is no absolute necessity for them to have been composed together.’ But in the Gloria and Credo the four voices do not consistently

have these functions. Dissonances (most significantly, fourths) occur in every combination of upper and lower voices. There is a tendency for the Triplum to provide the top of the texture— particularly in the Gloria—and there is a more marked tendency for the Tenor to be most consonant with the other voices. But there are structural fourths even between this pair (e.g. Gloria, bb. 32, 76, 86; Credo, bb, 11, 25, 44, 46, 118, 130). Even when

consonant with the Tenor thet counterpoint which the upper voices make is often unconyéntional, usually because the essential progression is placed elsewhere in the texture. This is particularly noticeable at cadences, where the essential sixth-to-octave progression is not consistently allotted to any two of the four voices. Structural sixth-to-octaves occur with near equal frequency between Triplum and Tenor and Motetus and Tenor but are occasionally found between Triplum and Contratenor (e.g. Gloria, bb. 14, 76) and sometimes even between three voices, for

example when the e of a sixth e-c’ sharp is resolved by a d ina different voice (e.g. Gloria, 15-16 (Mo—CT sixth resolved by Mo-T

41-2

octave),

(Mo—-T

to Mo-CT),

69-70

(Tr-T

to

Tr-CT)). Even where the Tenor behaves conventionally at cadences, this is sometimes achieved only by awkward voiceexchanges earlier in the phrase (e.g. Gloria, 11-12, 27-9, Credo,

55-7). And ‘voice-exchange’—that is, places where Tenor and Contratenor or Triplum and Motetus leap in order to exchange lines—is a constant feature of both these movements. As well as the cases just mentioned, where it brings the Tenor part back to a conventional tenor function, it is used frequently as an aid to the singers of the other parts, helping them to make the right decisions about where to sing sharps or flats; (all instances of this are discussed in the Commentary.to the edition). A third use of voice-exchange is for the sake of sonority: repeated chords often 3 Such appears to be the case in the most nearly comparable works from the ‘general’ repertory of fourteenth-century mass music, the simultaneous-style Credo

‘Bonbarde’

and

the Sanctus

sanans fragilia.

See H.

Stablein-Harder,

Fourteenth-Century Mass Music in France (Corpus Mensurabilis Musicae, 29, and Musicological Studies and Documents, 7; American Institute of Musicology, 1962), nos. 55 and 66.

58

ELABORATION

exchange voices simply in order to keep the texture ‘alive’, the second chord sounding different even though the pitches have not changed. Even more surprising is that relatively extended progressions, involving dissonances prolonged over several beats, may resolve in voices other than those in which they began. In Credo, bb. 22-5, for example, the e’ reached by Triplum and Motetus at the resolution of the previous sub-phrase (b. 21) is transferred to become an imperfect consonance in the Contratenor (b. 22); is taken up again by the Contratenor and Motetus (in unison) in bb. 23-4, where it is still waiting to resolve into a d-chord; and is then transferred to the Triplum (b. 24, last quarter) where it finally resolves in b. 25. It is difficult to see how this progression could have been so arranged if the parts were composed successively. These unusual features of part-writing are only explicable if we assume that the Gloria and Credo were conceived not one voice at a time, against a separately composed Tenor, but as a four-part whole. And the fact that melodic lines at all levels of the texture (not just the top) migrate freely between voices suggests that it was only at a later stage in the composition process that this whole was allotted to four individual voice-parts. The large numbers of unisons (reducing the final texture to three or two parts) and of unconventional leaps in the bass suggest that Machaut’s initial conception did not attempt to account for the precise workings of

each voice all of the time. (Credo, b. 86, last chord, for example,

was probably reduced from three notes first imagined (the present two plus an a in the Contratenor) when the part-writing became sufficiently clear for Machaut to see that the resulting Contratenor melodic shape a—g—a might encourage a singer to produce an unwanted g sharp.) It is unlikely, then, that all the details of such passages as Gloria, b. 63 or Credo, b. 43, which arise only out of the needs of part-writing, were already clear from the beginning. These sorts of decorative figures are much more likely to have been worked out as the parts were extracted from the sonorities initially conceived. But Machaut must have had a fairly precise idea of the essential chord-progressions from a very early stage. The process of separating out four individual voice-parts from this initial conception was evidently not a simple one. Machaut seems to have proceeded pragmatically, according to convention to the extent that from each chord whichever of the lower voices

ELABORATION

59

was most consonant with all’the others was allotted to the same part, as in a conventional tenor; but this principle had frequently to be set aside whenever better part-writing would result from

some other arrangement. In Gloria, b. 15, for example, Machaut has imagined a progression in which one voice (the second from the bottom of the texture) passes down (a-g—f) to meet the bass e (Contratenor, third to sixth notes, in the final version). When notated in parts,

therefore, the Tenor has no choice but to leap up a fifth to the position of ‘harmonic filler’ just vacated by the Contratenor. Returning the voices immediately to their original positions, at

b. 16, despite the continued bass descent to d, avoids pitch repetition in the Tenor. This seems to have been a factor which concerned Machaut. Gloria, bb. 7-9, 72-3, 101, and Credo, 104—5

are other occasions on which he exchanges the bass line and the second line up between the two lower parts in order to avoid either part having the same pitch in two successive chords. As already suggested, there seems to be no consistent reason for this other than to keep the texture mobile via the change in sonority which inevitably results. Another clear example of part-crossing is Credo, b. 35, where the Contratenor passes above the Tenor

because the result makes a better melodic continuation of the

previous bar. In this case, then, Machaut had no choice but to

place either the Triplum or the Motetus as the third voice down, until passing notes in the middle of the texture allow it (in b. 36) to rise back towards the top. Voice-exchange on repeated notes in b. 37, Triplum and Motetus, finally brings the Triplum back to the highest line in time for the cadence. Another factor tending to modify a simple separating-out of chords according to their vertical order (top note to Triplum, second down to Motetus, etc.) was—as has been suggested—the need to avoid melodic shapes that might encourage solmisations (decisions made by the singers about what we would call sharps and flats) other than those which the composer intended. This was a particular problem with the far from conventional harmonic language which Machaut developed in the Gloria and Credo; and there are a great many examples of part-writing arranged with this in mind.

In Gloria,

bb. 81-2,

for instance,

Tenor

and

Contratenor exchange, breaking up an excellent Tenor line, in order to ensure the c natural in the bass (and thus the Triplum) at the end of b. 82. In Gloria, b. 93 both pairs of voices exchange

60

ELABORATION

their lines in order to facilitate the C sharp to C natural chord progression. At Gloria, b.101 the Tenor and Contratenor exchange for their second notes in order to discourage a Tenor f sharp as its first. Similarly in Credo, b.27 the exchange of Contratenor and Motetus facilitates c’ natural after the c’ sharp of b. 26. And in Credo, b. 8 the exchange of Triplum and Motetus encourages the melodic augmented second f’ natural—g’ sharp; while that of Triplum and Contratenor in the next bar encourages

a Contratenor c’ sharp. Related to this factor is Machaut’s consistent attempt to keep the Tenor line free from any pitches needing to be read as ficta. A glance at the edition will reveal that the Tenor in the Gloria and Credo—in stark contrast to the other voices—contains no sharps or flats at all, save only two cautionary B naturals in the Credo (b. 146 and Credo Amen, b. 26). There may have been some theoretical principle behind this—a feeling that the Tenor should not be subject to any modification—but one valuable consequence is that the singers can safely take the Tenor as a prescriptive line, deciding how to solmise the other lines in relation to its requirements. That this was Machaut’s intention is implied by exchanges such as Gloria, b. 11 (between first and second beats), 27-8, 95-6, which serve no purpose unless to keep c’ sharps out of the Tenor.* Two larger extracts illustrate how such factors can mine the optimal layout of parts over more extended 9 (system 6) shows Gloria, bb. 30-46 as it would Machaut if, as a first step, he simply laid out the parts

combine to deterpassages. Example have appeared to in vertical order of

pitch: Triplum, Motetus, Contratenor, Tenor—the order in which they

begin at b. 30 in the final version (shown in system c). For the Triplum and Motetus, at least, this arrangement seems to have been largely satisfactory: the few occasions on which the Motetus goes above the Triplum (bb. 34, 40 2nd beat to 41 2nd beat, 43) all offer better voiceleading than could be obtained from a simple vertical transcription from the

original

(system

a).2 The

Tenor

and

Contratenor,

however,

exchange frequently. In b. 34 the bass c [sharp] was taken out of the

Tenor, in line with Machaut’s invariable habit; and in order to lead

* The problem of deciding which unsigned pitches should be sung sharp is

considered

in the Introduction

further discussion ° The notation writing. Separate direction depends

to the Edition, Appendix,

Sect. A.3, where

of the ‘sharpless’ Tenor part will be found. of system a intends to imply as little as possible about parttails are used only in order to specify durations; and their only on graphic clarity.

oe

8

wosn

-

op

—

os

¢

a

ee.

oe

z=

eae

o

OW/IL

Tee

ap

ou-m

- og

>

ES?

= SSS

SSS

eS

e

4

7

=

ieee

Sa

- 29 ©xel sn $3]

:ob—o€

SuniM-jied

(#)

ase

=

= : t =

==

ase

Vie

ed

‘eu0[y ‘qq

queUOssIp pue Juasqe

ee ee etal eee

eteteeee ant tape

SSS ==

-

se

od-m-wo1a

clea asa -

*xg6

ELABORATION

62

ELABORATION

easily into it the Tenor and Contratenor are exchanged for the preceding notes as well, back as far as the first three-voice chord in b. 32, where

Machaut is able to engineer a unison between Tenor and Contratenor. (Unisons, of course, are the most convenient point for parts to cross. In this case, because.it makes better voice-leading, the Contratenor probably had a 6 originally, in unison with the Motetus, which Machaut would have changed for the sake of the lower voices.) Having made this exchange the lower parts would have looked as in Ex. roa. At this stage the Tenor in b. 36 might sugge f sharps st to the singer, clashing with Triplum and Motetus; and Machaut therefore exchanges the voices at the unison f in b. 36 so that the return from f to g is now in the Contratenor. Bars 35-8 now look as in Ex. rob; but because the last exchange has brought a cadential c [sharp] into the Tenor line a further exchange is necessary in order to return it to the Contratenor, giving the arrangement shown in Ex. 10c. Machaut made one further change, and it is one of the most interesting. As the last Example shows, not only is the Tenor line tiresomely repetitive—something which the final versions of Gloria and Credo suggest Machaut tried hard to avoid—but also the word ‘deus’ begins two successive phrases with the same pitches (f-g). This offers a classic opportunity for a singer (or a scribe) to skip to the wrong one, either singing (or copying) ‘deus rex celestis’ twice or leaving it out altogether. Machaut therefore exchanges Tenor and Contratenor pitches for the first beat of b. 35, which gives the Tenor line a quite different profile although at the expense of an awkward

Ex. 10 Gloria, bb. 32-8: part-writing

ih

a

32

35

==

[Contratenor]

=F

nae

32 de’

=

“us rex,

ce

==

=

=

=

35

= =

SS

es:

38

SSS

=

“tis

de

- us

SSS

=: c=

=

=:Sine =

38

SS pa - ter om - ni -

po

=

se

tens

ELABORATION

63

é

Contratenor leap, c [sharp] to f natural. Clearly the music was arranged with the habits of singers very much in mind. A rather more complex example is shown in Ex. 11, representing Credo, bb. 19-26. As in Ex. 9, system ais necessarily a reduction of the

final version rather than a certain representation of Machaut’s initial conception: it is impossible to be sure exactly what that contained, although it probably lacked some details of the end result. A possible original state will be suggested below. System b again lays out the parts in vertical order of pitch (giving the bass line to the Tenor in order to keep lower-voice sharps in, the Contratenor), but this time already includes some rearrangement of pagts into the cadence (bb. 25—6) of the sort that Machaut must have’mhade automatically as he began to think about their most effective layout. System bis still too literal-minded, however. The d’—c’ passing-notes in Contratenor, b. 21 need an e’ to lead from: that can be taken from the Motetus, which thus takes the a previously in the Contratenor. In the

interests of better voice-leading into this progression, Machaut also exchanged these voices for the preceding four notes (b. 20). Then in b. 22 the Motetus and Contratenor were exchanged again in order to avoid pitch-repetition between their e’s and bs. In the three-part chord in b. 24 Machaut took what was now the Contratenor up to the Triplum e

in place of Motetus c’ sharp to avoid the melodic clash c’ sharp—a—b—c’ natural. System c records the position so far. Next the rather dull Motetus line which results from these changes is enlivened by the daring

exchange of Triplum and Motetus from b. 22, so that it now covers an

octave in four notes. In performance we hear the texture unmistakably revitalized as the Motetus singer leaps up to the high a’—g’ sharp—a’. Finally the upper voices are returned to their previous positions at b. 25, and beat, as Machaut once again avoids repeated pitches, this time two d's in what is now the Triplum (system d).

The result of this process is a very complex ensemble of partwriting, applied to an already complex span of music whose underlying harmonic structure is now concealed beneath several layers of compositional elaboration. Some of these layers it has been possible to isolate; but the preceding discussion made no attempt to decide which details were integral to Machaut’s conception and which evolved only as he separated out the music into parts. It assumed that the music had its final form before the separating-out into parts began. Yet this seems unlikely. Exactly ructions with 6 Of course it is a relatively simple matter to make such reconst trate what the benefit of hindsight. The example does not claim to demons have influenced Machaut did but aims to show the kinds of factors which must his decisions.

64

ELABORATION

Ex. 11 Credo, bb. 19-26: part-writing

(AS St

ya

z

Sere SSS

SS

SSS 19

SSeS =

SS SS Ee

| a Po

21

a} \ 2

ee

i pyr

i

at

i

Ts

23

eS tee

25

ett tire

Saas

SSS

SS

SS

SS

: =. SE teeta

—forge

how much was finalised before Machaut started to work in parts can never be known; but it may be possible to suggest a model

which, even if its details are hypothetical, will show the sorts of

processes which he is likely to have followed. Example re suggests a possible ‘original state’ for this passage,

ELABORATION

65

é¢

a form it might have taken as Machaut first imagined it. It already contains those passing-notes which are essential to smooth chordprogression—and so were likely to be present from the start—but it excludes details such as the Triplum figures at the beginning of b. 20 and the Contratenor d in its second chord, both of which

help the otherwise weak progression between the two middle chords of that bar. There are two differences which are more significant. Because of the ‘unigenitum’ symbolism in Triplum and Motetus (mentioned above, Sect. 2.3.3) and its, need for an extra beat before the cadence, it seems probable,that ‘b. 25’ was extended as a later

thought, when the repetition of the melodic descent occurred to Machaut as a possibility.’ And that in turn would have involved considerable elaboration of the lower voices into the cadence. Seen in this reduced form there also seems to be a compositional problem in the leading from b. 21 to b. 23. When the final form of b. 21 is considered as a whole, it appears to be doing no more than disguising a rather weak progression from the e-b-e’ chord

—basically the same sonority on the adjacent of b. 21 via e-b-e’—g’

strong beat—into the remainder of the b. 23 progression return-

ing down to e (b. 24). The disguise, however, is brilliant; and,

coupled with the voice-exchange already mentioned, gives new impetus to the otherwise rather e-bound harmony. The details of b. 23’s scoring, in other words, are bound up with its function.

And the possibility that it was conceived only at the part-writing stage—and that therefore the part-writing stage involved important developments in the composition—seems strong. This appears to be borne out by the dissonances introduced as part of the process of elaborating Ex. 11e. The prolongation of the pitch e’ has already been described (p. 58) and the discussion of the part-writing process has explained its migration through the voices (p. 63). The prolongation of a through this passage is just as carefully arranged. From its prominent bass position in b. 20, 2nd chord, it is reinforced as a dissonant note in b. 21 (Motetus) and

again, at the top of the texture, in b. 22. It resolves, but only in the 7 This scenario for b. 25 is supported by problems which the all three main sources had in fitting all the notes over the word spaced by the text scribes. Apparently the final Tr c.o.p. ligature that G read it as ligated to the following long; which suggests

music scribes of ‘unigenitum’ as was so squashed that it was these

two semibreves which were added. See Appendix, Critical Notes, under Credo,

‘Texting’.

66

ELABORATION

most unstable of imperfect consonances, at the beginning of b. 23 and is held as a dissonance in the following chord (Triplum, sth note of b. 23). It is (imperfectly) consonant with, but does not sound in, the first chord of b. 24; and is only taken up again in the second half of that bar, in a Contratenor figure which leads it to eventual resolution in the first chord of b. 25. But unlike the e’,

which is fundamental and present even in the hypothesised original (Ex. r1e), the a, although a powerful cohesive force in the passage, becomes such only in the final elaboration. The dissonant as of b. 21, 1st chord, and b. 23, 2nd chord, and the a’s of b. 22 are

most unlikely to have been imagined right from the beginning: they depend upon details worked out only later, probably in connection with the production of four individual parts. This is not to say that Machaut’s first idea for the piece was necessarily more conventional than his final elaboration: but it does imply that the voice-leading became markedly more subtle as the parts were specified and as the texture came contrapuntally to life. This view of the detailed working-out of Machaut’s initial conception helps to make sense of a question which arose earlier: where in the composition process is the Gloria chant-paraphrase introduced and how are its details determined? The most extended references to the chant (bb. 8-15, 17-22, 24-8, 35-40, $7-63, 68-74, 88-91, 99-101—as shown back in Ex. 6) almost

always migrate between two voices. This might at first seem to indicate an original cantus firmus line broken up during the working-out of parts. But as we saw in Chapter 2 the chant is by no means always present: there are many significant diversions, especially around cadences. And this suggests that from the point at which Machaut began to work out the Gloria’s repeating melodic—harmonic structure he can no longer have been concerned with the chant melody as a continually present thread. In view of the evidence presented above for an initially chordal conception, the most we might provisionally suppose is that the movement was conceived as a chordal elaboration of the chant, its

presence or absence at any moment in the finished piece being subject to the needs of the melodic-harmonic plan. This scenario might be supported by noting that at several points when the chant is absent it would nevertheless be (perfectly or imperfectly) consonant with the pitches present: these cases are marked with ‘x’ note-heads in Ex. 6. And this may suggest that the process of working out the vocal lines involved rescoring chords (and

ELABORATION

67

é

thereby losing the chant notes) where better part-writing would result; which seems to tie in with the process suggested in Ex. 9. Example 9d brings together information from Examples 6 and 9a in order to give some idea of how far the chant survives in the final version, and of the extent to which it might have been present (in the ‘x’ positions) in Machaut’s first conception.® On the other hand, it may be significant that the closest match of polyphony to chant involves shifting syllables (‘deus pater omnipotens’ and ‘ihesu’), for this suggests that the composer was concerned not so much/to harmonize the chant as to ‘reminisce

around’ it. In that case wecannot be certain that Machaut at any stage in the composition process attempted a simple harmonisation of the chant: he simply kept it very firmly in mind as he worked through from text to melodic-harmonic structure to chord sequence to part-writing. The advantage of this view is that it allows the melodic structure to have been developed in immediate relation to the text (as Chapter 2 suggested) and only then to have given rise to the initial conception of the whole polyphonic texture. If the latter came first, as a harmonisation of the chant, it is hard to see how the close formal relationship between text and music could be accounted for. But even if this is too concrete a proposal to be substantiated we can say as a minimum that the Gloria refers to its chant only when melodic structure, cadence-formation and part-writing permit: although Machaut may have had it in mind from the beginning, as the details of the composition were finalised these three dominant factors in the composition process all take precedence over it.

The process suggested here is complex but realistic. It does not require that all details were fully formed from the beginning, but nor does it suppose that so complicated and carefully organized a texture could be produced if parts were added one at a time. The question that now arises is how much of this Machaut could have worked out in his head, before all the parts were complete, and how much required the assistance of notation. The memory of a medieval composer was better exercised than that of his twentieth-century counterpart. Not only was he used 8 Because we cannot reconstruct completely the readings of the chant as Machaut knew it, it is impossible to be certain exactly how far these pitches apply, however.

68

ELABORATION

to remembering more music, but, it is possible that as a consequence his powers of musical imagination were more acute. Even if the latter does not follow, there seems good reason to suppose that once his conception became concrete enough to be specific about chords, Machaut could have held a large part of each of these movements in his head without having to resort to notation. But if this overestimates his powers, any notation used at this stage, before he began to work out details of parts, must have been score and on a single staff (seven lines, with a c-clef on the middle one, would have been sufficient). In estimating whether this would have been necessary a great deal depends upon how much of each movement Machaut needed to work on at once. Did the whole need to be conceived before details of parts were worked out? The lack of any single overall formal or harmonic plan for either movement suggests not. Even in the more obviously organised Gloria the largest unit of form is the ‘stanza’, and this is therefore the most that Machaut

needed to conceive clearly as a whole. If, however, there were no initial score to which to refer, it

seems probable that the process of working out the parts would have been somewhat more difficult, Machaut having to go consciously through rather more of the stages isolated above than might otherwise have been necessary. The more acute his imagination the less the difficulty. In either case, though, the working into parts—being a multistage process and not simple transcription—would have been much more efficiently managed in score than directly into separately notated parts. This need have had no permanent form: Machaut could easily have worked onto erasable tablets allowing alterations to be made at each stage of the process. Some support for the use of score at the part-writing stage comes from the inconsistent placing of mi (sharp) signs amongst the parts. These are by no means always placed in every part that needs them (e.g. Gloria, b. 28, Credo, b. 76); and at present the best explanation for this seems to be that sharps were originally positioned above, beneath or in the middle of a system of score, next to the chord to which they applied but not necessarily next to the relevant parts.” A copyist might then have made overliteral choices in copying into choirbook format. Similarly, * See also the discussion Appendix, Sect. A.3.

of sharps in the Introduction

to the Edition,

ELABORATION

69

é

Elizabeth Keitel has noted that ‘the custodes [of the Gloria and Credo, absent from their Amens] and mixed lines [between the Credo and its Amen in MS E] of the texted portions [present] strong evidence for their once having been in score’.’” The use of score seems probable, then; but it is not essential to the procedure hypothesized. To sum up the discussion so far, it has been suggested that Machaut conceived the Gloria and Credo not part by part but as a rough chordal whole, which was then separated out into four voice-parts according to criteria which can to some degree be reconstructed. Less essentially, a model for the state of Machaut’s original conception has been proposed, and there has been some general discussion of the kinds of notation which might have been used. More importantly, the evidence which has accumulated seems to demonstrate a very sophisticated handling of four-part textures and total control both of vertical sonorities and of the voice-leading which generates them. It invites us to look afresh at the isorhythmic movements to see how this new evidence may

affect our view of their creation.

3.2.3 The isorhythmic movements We have seen that in the isorhythmic movements the Tenor and Contratenor must have been composed to a large extent together in order to make a satisfactory isorhythmic and contrapuntal pair. That the upper voices were also composed as a pair seems to be confirmed by their very close relationship in the hocket passages—sometimes one providing the melodic resolution of a preceding passage, sometimes the other (e.g. Kyrie i, *b.35 continuation of Tr and Mo in Mo; Kyrie III, b. 5 in Tr; Sanctus,

b. 60 Mo, and b. 36 Tr)—and also by passages such as Kyrie II, b. 12 or Agnus I, b. 13 where they are mutually dependent. In addition there are aspects of the four-part writing in Kyrie, Sanctus, Agnus and Ite which suggest that Machaut was finalizing the upper voices with the four-part whole firmly in mind; and after what we have seen of his four-part control in Gloria and Credo this should come as no surprise. Kyrie II and III both open with a remarkable Motetus line, rising from the initial ContraElizabeth

Ann

Keitel, ‘The So-Called

Cyclic Mass of Guillaume

de

ly, 48 (1982), 314. Machaut: New Evidence for an Old Debate’, Musical Quarter See also Sect. 4.1 below.

7O

ELABORATION

tenor a through b-c’—c’ sharp (b. 2) d’—e’-f" sharp (b. 3)-g’ (b. 4), momentarily touching a’ an octave above the starting point (Tr and Mo, b. s) before falling back to g’ (Kyrie II, b. 6, Ill, b. 5). It is

hard to see how such a progression could have been created separately from the composition of the Triplum, or (more controversially) disconnected from the leap of a seventh in the Contratenor which makes it possible. The relationship of the four voices must have been considered from an early stage, perhaps— though not necessarily—early enough for the Contratenor counterpoint to be influenced by whatsMachaut planned to do with the upper voices. The argument may be taken a little further by considering the relationship between Triplum and Motetus isorhythm and four-part harmony. In the isorhythmic section of the Sanctus (b. 16 onwards) the pairs of hocket bars (bb. 20-1 et seq.) are placed between two “pause bars’, so that however they elaborate their lower voices they must lead logically into the three-beat chord which follows them. (Naturally this

is true also of the Tenor and Contratenor, whose faster movement

in

these bars was evidently fixed knowing that the upper-voice hockets would be placed above them.) The Triplum—Motetus rhythmic figure which ends each hocket-bar serves this leading function admirably. Into this carefully designed mould Machaut inserts some cleverly contrive d voice-leading. In bb. 83-6, for example, the Triplum and Contratenor outer voices’ b’ over g (b. 83) resolve only at the far end of the hocket (last note of b. 85 into b. 86), while the essential cadential e’, which gives the progression its strength, migrates from the Tenor (b. 83) to the Motetus (b. 85). Machaut could have arranged the Tenor—Contrate nor pair in these bars without thought for their elaboration, but it seems unlikely, especially in view of the contrapuntal function built into the rhythms of Triplum and Motetus. These same bars, and their continuation, feature another of Machaut ’s

long-range melodic ascents, this time in two layers at once. The unison

Tenor—Contratenor a (b. 85), continued by the Tenor b-c’, is then taken

up by the Motetus d’ (bb. 87-8)-e’-f’ sharp-g’ (90-1)-a’ (again, an octave above its starting-point) which then becomes part of the final

descent in the Triplum (b. 93). The Tenor e’ (b. 83), meanwhile, has

resolved into the Motetus f” (b. 86) which transfers to the Triplum f’—f" sharp (b. 87)-g" (88)—a’ (90)—b' (91) aiming at the top c” in b. 92 from which the final melodic descent begins, c”—b’ flat—a’—g’—-f" (b. 94). Once

again, this proves nothing about the design of the lower voices, but certainly demonstrates that Machaut was composing Triplu m and

Motetus

with continual

reference to the four-part whole,

masterly control of long-range harmonic progression.

and with

ELABORATION

71

@

The possibility that the lower voices were formed with upper-voice progressions at least partially in mind (though not necessarily fully formed as té detail) is opened up by the use of lower-voice rests. An early example is to be found in the Christe. The third and seventh bars of each Tenor

talea have an unusual

rhythmic

figure, a one-beat

note

followed by a two-beat rest during which the Contratenor is free to change the harmony. Before offering an explanation for this it is necessary to examine another special characteristic of this movement. While the surrounding Kyrie sections have final cadences in which the d—a-d'—a' resolution is approached via pungently sharpened Cs and Gs, the Christe seems intended to fungtion as a contrasting central section,

studiously avoiding C sharps and G sharps before Ds and As. In this it finds an echo in the Gloria and Credo, both of which tend to use ‘sharp’ cadences in their opening and closing sections, but tend to avoid them in their middles. To this extent all the D-based movements are shaped in a similar way. In the Christe the ‘natural’ (or recta) final cadence is assured—unmistakably to the singers—by the Tenor rest in the penultimate bar; for the Contratenor g which fills it must be natural on account

of the Triplum and Motetus d’s a fifth above. A similar pattern is made at bb. 10-11 and 17-18. It is at least possible here that the lower-voice taleae were shaped with this function in mind. More interesting examples are to be found in the Agnus Dei. In Agnus I Tenor and Contratenor rests are placed in such a way as to

produce wide leaps between their successively sounding notes: a fourth

in b. 16, a fifth in bb. 4, 11 and 12, a sixth in b. 9, anda seventh in bb. 9,

11 and 12. It seems inconceivable that a composer would engineer such a lower-voice duet unless he knew how he was going to use the upper voices to make it work. Bars 9 and 12 seem to confirm that Machaut knew exactly what he was doing: in each case the leap from low Contratenor to high Tenor and back again is harmonized above by pitches consonant with the note which would be in the Contratenor line if it had a note insteadofits rest—in b. 9 and e and in b. 12 (and arguably b. 11) ag. In other words, Machaut is using the Triplum and Motetus to imply harmony in the lower voices. (This can be confirmed by comparing bb. 11-13 with Sanctus, bb. 20-5 (and see also 49-50) which has the same Tenor—Contratenor progression in reverse with all the notes filled in.) The Tenor and Contratenor must have been composed with the upper voices in mind. In Agnus II Machaut goes still further, alternating lower-voice notes and rests so continuously as to create a single line, the compound melody of whose pitches implies a two-part contrapuntal structure. Again, he must have been absolutely confident that the upper voices would do the rest.

The voice-parts of the isorhythmic movements, then, need be

72

ELABORATION

composed successively only in so far as the isorhythmic process demands. The Tenor and Contratenor have to be treated as a selfsufficient pair only in order to determine the Contratenor pitches in relation to the rhythmic scheme which permits them. But these decisions may already be influenced by what the composer intends to do with the upper voices. The lower-voice rhythmic scheme may be shaped so as to alternate areas of harmonic activity with points of repose (the three-beat pause-chords which occur two or three times in each talea), knowing that the upper voices can be directed from one pause to the next above the Tenor chant notes which will fall on each. The skilful composer will be able to foresee the consequences—what can and cannot be satisfactorily accommodated by the upper voices—of any decision he makes in the design of the lower. It seems likely, therefore, that as he finalises the details of Tenor and Contratenor he may already have a rough idea of what the Triplum and Motetus will do; and when he comes to finalise the details of Triplum and Motetus he

is, of course, considering all four voices as a whole. This would

seem to be the implication of the relationships which we have observed; and in view of Machaut’s procedure in the Gloria and Credo it should come as no surprise.

3.3 Prolongations Because Machaut was working for much of the time with all four parts together, he was able to direct the harmony—and the four lines which create it—more single-mindedly and towards longerterm goals than would have been possible with aless sophisticated procedure. How he manages this deserves further investigation.!! We have already seen several examples of the way in which a single melodic line—usually an inner voice—is led through the four-part texture over several bars of music. Some of these were simple progressions (Kyrie II and III, bb. 1-6, Sanctus, 85—94);

others involved prolongations,

continuations of the effect of a

single pitch (or sonority) through a changing texture, often with the result of intensifying its effect towards eventual progression

into another

(Credo,

bb. 21~s,

Sanctus,

83-5).

Some

further

" An excellent introduction to the workings of fourteenth-century counterpoint is Sarah Fuller, ‘On Sonority in Fourteenth-Century Polyphony: Some Preliminary Reflections’, Journal of Music Theory, 30 (1986), 35—70.

ELABORATION

73

examples may allow us to isolate the principles which underlie this technique. A simple case first. In Sanctus, bb. 22-3 the outer vcices—g in the Contratenor and b’ in the Triplum—are prolonged by means of the intervening chord on a. The outer As, also in Triplum and Contratenor, are one step inside—neighbour-notes to—the g—b’ sonority, and because they return to it they seem to the listener to do no more than decorate its continuing effect through those two bars. In addition, the Tenor e’ in the first chord is taken up in the A-chord by the Motetus, which leads that line downwards by step to fill in the third above the Contratenor absent from the first chord. The resolution in, b. 24 leads the intervals of the third and the tenth, of which the second b-2' chord is composed, outwards

to a perfectly consonant fifth and twelfth on f. But the Tenor also reappears with the high f” as the resolution of the previous e's. It appears, because of this resolution, that the high e’ was ‘present’, even though not sounded, right through b. 23; and the Motetus seems not so much to have led the middle voice to a new position but rather to have created a fourth voice beneath the continuingly implied e’. Thus the reduction to three parts forced by the isorhythm is hardly apparent as the cadential preparation intensifies its need to resolve through these two bars. Example 12 shows how this analysis may be notated so as to save this sort of wordy description. Beams show prolongation, slurs show how one note leads into another; and the decorative pitches are labelled “N’ for neighbour- (or auxiliary) notes, ‘P’ for passing-notes.’ Ex. 12 Sanctus, bb. 22-4

Similar prolongations may be seen at work on a slightly larger scale at the openings of Kyrie, Gloria, Credo, Sanctus and Agnus (Kyrie I, bb. 1-3, Christe,

1-3, ‘Kyrie: IL..3-=3,,6 yrie

wlll. 1-3,

2 Daniel Leech-Wilkinson, ‘Contrapunctus Diminutus and Prolongation’ (paper delivered to the Thirteenth Annual Conference on Medieval and

Renaissance Music, University of Nottingham, 1985) argues that these functions

are illustrated in the counterpoint treatise of Petrus Frater dictus Palma Ociosa (1336) and so were well understood by anyone trained in ‘diminished’ counterpoint. Machaut’s use goes beyond that of the theorist only in its application to four parts instead of the theorist’s two.

74

ELABORATION

Gloria, 1-4, s—8, Credo, 3—5, 6—9, Sanctus, I—5, 48—51 (Osanna),

Agnus I, 1-4, Agnus II, 1-3), firmly.establishing the focal sonority of the following sections. This might suggest the deliberate creation of links between the openings of each movement, comparable to the motto openings of later, ‘cyclic’ masses, were it not for the occurrence of similar progressions at the start of several of Machaut’s songs—for example, Ballade 42 or Rondeau 21. Example 13 analyses the opening of the Sanctus; and Example 14 shows, in reduced form, how the whole of the non-isorhythmic introduction, setting the threefold invocation ‘Sanctus, Sanc-

tus, Sanctus’, is working within a framework of f-c” sonorities. Ex. 13 Sanctus, bb. 1-5

A similarly extended example is the prolongation of a d-a’ sonority through Sanctus, bb. 30-45. This offers a much weaker

continuity than the previous example; for fewer of the sonorities

intervening between the d—a’ ‘pillars’ (at bb. 30, 35, 42, 45) stress

pitches consonant with them. It might be more appropriate in this case, therefore, to speak of the d-a’ sonority being ‘kept in play’ as the music moves away from and back towards it, rather than to suggest that it is fundamentally present throughout. It is between these extremes—of firm prolongation in the first example and looser ‘playing round about’ in the second—that Machaut organizes large spans of music around focal sonorities and simple progressions. The Christe provides a typical example of the way in which prolongations and progressions link up to make a complete musical section. Example 15 shows Machaut taking advantage of the repeating as in the chant (bb. 1-3) and then the recurring gs (bb. 5-8), counterpointed by Contratenor fifths beneath, to bind

ELABORATION

75

Ex. 15 Christe

together each passage within d-a-d'—a’ and c—g-c’—g’ sonorities. These are linked by b. 4 which Machaut makes unstable—and thus directed towards b. 5—by means of its Triplum and Motetus elaboration. The next passage is only a little more complicated. Bars 9-11 lead towards the imperfect consonance on a in b. 12; but the intervention in b. 11 of a more stable D-chord, also

prolonging the melodic a’, ties in with the D-chord reached in b. 13; so that these bars are bound together by both A- and Dsonorities. The remainder of the Christe avoids extended prolongations (perhaps deliberately since the Contratenor in b. 17 could have used ac to link with b. 15, as bb. 1-3 (the same rhythms

focused on d) indicate). Machaut instead directs his progressions towards the Tenor d of b. 18, again supporting a high a’. It is only

then that the high c” is at last introduced (b. 20) in order to allow

an effective melodic descent to the final chord. It is melodic progressions such as this, placed at important points of structure, which are the most interesting features for study; for while the harmonic foci are already determined to a large extent by the Tenor—Contratenor duet, and were laid down by Machaut at the time of its formation, the essential melodic progressions outlined in the upper voices are chosen from a wider field of possibilities. The main restrictions on Machaut’s freedom of choice are the details of the lower-voice chords and the need to fit all four voices within a limited range (usually a twelfth) above the bass. In a D-based section, therefore, major cadences are going to be arranged around either d’ or a’, and in an F-based section around f’ or c” as the only available perfect consonances above the

76

ELABORATION

final, (though c” is not in practice used as a final). These melodic finals are then approached from either the third or the fifth above,

in the case of d’ and f’, or additionally (but rarely) from below in the case of a’. Here in the Christe we see Machaut focusing upon a’ and f-based sonorities) and g’ (above c, e and g) almost out, the main areas of extended melodic motion being

the pitch (above dthroughbetween

focal sonorities (e.g. bb. 3-4, 6-7, 13-15) and at the end where the

Triplum jumps up to top c” for the first time in order to lead more effectively downwards into the final a’. And this contrast between melodic ‘hovering’ around a focal pitch and descending towards the next is a consistent feature of Machaut’s treatment of Triplum and Motetus. Equally consistent is that the Motetus functions for most of the time as a support for the Triplum, and not as a melodic line on an equal footing with it.'* The Motetus fills in the space between the Triplum and the lower voices, only rarely crossing over to take up the melodic thread. Where it does cross it is almost always for the sake of good voice-leading (e.g. Christe, b. 3 et seq., the fast descents from b’ being placed in the Motetus whenever they lead

down to an inner voice) or to avoid mistaken solmisations from

the singers (e.g. Sanctus, bb.2 and 12 where the exchange discourages a Triplum b’ flat). It is hard to avoid concluding from all this that Machaut is organizing his harmonic and melodic prolongations within a treble—bass framework. Agnus I provides a particularly clear example of this from the F-based half of the Mass. The fourths between the Tnplum and

Tenor in bb. 8 and 10 and between Motetus and Tenor in bb. 1,8

and Io indicate that neither upper voice was written so as to make a contrapuntal structure with the Tenor; nor is any voice but the Tenor consistently consonant with the Contratenor. As all the evidence collected in this chapter has implied, the fundamental reference line is the bass, in whichever part it may be. It was against the bass that the upper voices were composed. Similarly, between the upper voices it is the top line—rather than consistently Triplum or Motetus—on which Machaut lavishes most care to create melodic coherence, while the next line down is used

as a supporting filler even at the expense of a more jagged contour. Thus in bb. 8 and 9-10 the Motetus rises to the Triplum ‘’ Compare the discussion of Kyrie I, Sect. 2.2. 3:

ELABORATION

77

level to fill in the top line pitches where the isorhythmic scheme forces a Triplum leap downwards, and likewise in b. 15 provides the resolution of the preceding B-chord where its rhythms offer a longer climactic c” than those of the Triplum. In b. 19 the

Motetus leaps a fifth and back solely in order to provide a sonorous third beneath the last c”. In each case the resulting Motetus line is less elegant than that of the Triplum, and less still than that which they make together. With this in mind we can turn with more confidence to examine the treble—bass contrapuntal structure of Agnus I. In Example 16 staff a accounts for most of the detail of the outer lines, with inner

voices/ shown, wherever

they have extended

progressions inside the texture. Staff b reduces out the detail to show more clearly how the focal sonorities are connected. Ex. 16 AgnusI Frangois Avril, ‘Les Manuscrits enluminés de Guilla es, 23; Paris, 1982), Guillaume de Machaut: Poéte et compositeur (Actes et Colloqu pp. 126-7.

98

APPENDIX

ambiguous points in his working manuscript (S Tr Mo 9~12; see also Sect. A.3 below). None of these sources, therefore, can be definitively called ‘the best’. Vg probably represents the best sense that a particular editor could make of whatever source material Machaut provided. A represents a more literal reproduction of this or something similar, but also contains changes which may have some authority; while G descends from the same exemplar, presumably—if copied in the 1390s—as left to Machaut’s literary descendants. Each source deserves consideration at each point of disagreement; and where genuinely alternative readings

exist a chronology of alterations has to be attempted. Where a plausible chronology can be achieved I have preferred to print the earliest recoverable version, on the grounds that it,is not possible to be certain that later revisions were made by Machaut himself. I have attempted to indicate with bold type, in the listings of variants, where alternative readings exist which performers may legitimately consider.

A.3 Sharps Like most medieval music, Machaut’s Mass is notated without indicat-

ing the placing of all the sharps, naturals and flats which performers would have sung. And although supplying the necessary naturals and flats is no more difficult than in other fourteenth-century works, where to recommend sharps is perhaps the most serious problem facing the modern editor. It is particularly difficult in the D-based sections where cadential E- or C-based sonorities are sometimes signed by Machaut with C sharps and G sharps in all possible voices, sometimes in only some of them (when it is by no means always clear that the unsigned sharps are intended), and sometimes in none. The pattern is so varied

that at present there appears to be no set of principles for the interpretation of written sharps that works. The lists of variants therefore conclude with a commentary on every sharp (or flat) editorially suggested or suppressed, in the hope that performers will wish to investigate alternative possibilities. Those sharps (or flats) which are indicated in A, G or Vg and which seem to represent Machaut’s intended reading (as implied by manuscript evidence and context) are printed in the staff as usual; those which I suggest were intended but not signed are indicated above the staff; those about which I am less certain

but which are still plausible are printed above the staff in parenthesis; and a few more improbable alternatives are mentioned in the commentaries. Some further explanation of my approach to this problem may be useful.’ In establishing criteria for the choices which have to be made it * An excellent introduction to this whole area, and particularly to the mental procedures employed by medieval musicians in deciding how to apply sharps and flats, is provided by Karol Berger, Musica Ficta: Theories of accidental inflections in vocal polyphony from Marchetto da Padova to Gioseffo Zarlino (Cambridge, 1987).

INTRODUCTION

TO

THE

EDITION

99

helps to start by considering the remarkable variety of intervals which Machaut expected his singers to be able to manage: major sixths (e.g. KI CT 9-10), major sevenths (KI*CT

17), lines such as KI CT 21-4, the

diminished fourth in Gl Mo 24-5 (though there may be room for dispute about the sharps here), the Contratenor fireworks in the Gloria Amen matched by hocketing fourths in the upper voices, frequent augmented seconds (Cr CT 37-8), diminished fifths (Cr CT 26-7), augmented fourths (Cr Mo 8), and successive semitones (CrA&: Tei 36-7). And the fact that these are all vocal lines suggests that modern assumptions about what is naturally vocal and what instrumental may be the exact opposite of those of medieval musicians. In the Middle Ages it was the singer who would be sufficiently trained in reading and sufficiently sensitive to contrapuntal context to be able to do these sorts of things; and conversely it would be the instrumentalist whose ‘oral’ tradition and whose heteroplionic ideas about accompaniment would lead him to expect conjunct diatonic lines in anything he performed. The first serious difficulty that singers performing the Mass from the Machaut manuscripts would have met occurs at the end of the first Kyrie, bb. 24-7, where the sharps in the Motetus part raise questions about (at least) the second g’ in Tr 26 and the c and g in CT 25-6. The end

of the Christe,

although

the chant

is the same,

is elaborated

differently (bb. 20-2). Here there are no sharps signed, and the g’s and d's of Triplum and Motetus in bar 21 prevent the Contratenor g from being sharpened. Thus there can be no question of adding ficta sharps. The cadence proceeds exactly as notated. The end of Kyrie II is also fairly straightforward (bb. 14-17). The exchange of Tenor and Contratenor pitches e and g allows the Tenor to be sharpened without either singer Ill being upset by seeing the sequence g-g sharp—a. But the end of Kyrie has (bb. 26-9) seems to have caught out all the scribes. What Machaut into n) transcriptio in done here is to add an extra three breve beats (a bar Tenorthe upper-voice isorhythmic scheme, but without extending the the —Contratenor pair. At some later stage he (or a copyist) turned but n) transcriptio in original CT breve e into a maxima (two bars Only Vg without making clear that the same applied to the Tenor. copied a maxima, and someone soon afterwards erased its right-hand (one bar). half to leave a breve. All the other sources copied a long precisely to Exactly what happened in the exemplar is impossible the reconstruct, but it is likely to have been caused by a need to sharpen nothing If maxima.” g a second part of what could otherwise have been vocabulary, liberal else, this hiatus seems to demonstrate that Machaut’s

notated on one staff in 5 A far-fetched possibility is that the T and CT were Ars Nova (ch. 19) that red different coloured inks (black and red—the remark in g voice may be relevant notes are sometimes used to distinguish a chant-bearin g A, G and E to assume here) and that Machaut corrected only in black, leavin copying of Tr in red, see that only the CT was affected. (For possible mistaken pitch/rhythm variants for Gloria Amen Tr 5.)

100

APPENDIX

as it was, did not include cadential augmented octaves (though see KIII 16 and AII 20 for passing ‘false relations’); it seems to confirm the necessity of sharpening the comparable plainchant g at the end of Kyrie II; and it seems to show that notated ‘sharps really do mean ‘raise by a semitone’. What, then, should happen at the end of Kyrie I? Taking the ends of Kyrie II and III as models, there seems good reason to sharpen the Triplum and Contratenor Gs in the penultimate bar (b. 26). Bar 25 in fact makes perfectly good counterpoint without any sharps; and it is

possible that Machaut decided that the Cs should be sharpened here in order to make the Contratenor line easier to sing (¢ sharp—g sharp instead of c natural—g sharp). But what of the Triplum g’ at the start of that bar: are vertical diminished fifths on the menu? An interesting passage in the Christe (bb. 12-13) rather suggests that they are. The Triplum and Motetus in b. 13, Ist beat have been exchanged in order to avoid the

Triplum figure a'—g'—a’ in which the g’ (above the Contratenor c sharp) would probably have been sharpened. Machaut seems to be doing his best to encourage the singers to produce a diminished fifth as an intended feature of the harmony. And as we have seen (Ch. 3) this technique of exchanging voices in order to force an unconventional result recurs quite often in the Mass, indicating that Machaut has the habits and needs of singers very much in mind as he writes. The problems of the Kyrie are child’s play compared to those of the Gloria; mainly because Machaut takes the cadential vocabulary of the

Kyrie one stage further, so that what was previously a cadence constructed over an e-d progression in the lowest voice is now constructed over a motion up from ¢sharp to d, with parallel octaves or twelfths between the outer voices. The problem of C sharps and G sharps is thus present in almost every cadence. We have already seen (Ch. 3) that the Gloria and Credo must have been conceived as a four-part whole, that it was only at a later stage in the composition process that this chordal whole was allotted to four individual voices, and that this seems to have been done in such a way as

to leave a Tenor voice with no notated sharps (Sect. 3.2.2). The parts are so arranged that wherever the other voices have a sharpened C or G (or both) the Tenor always has a neutral E. The only Cs or Gs in the Tenor occur in chords without sharps in the other parts, and in a contrapuntal context where the other parts might plausibly be sung natural. It seems unlikely that the Tenorista was supposed to be incapable of singing sharps correctly: even if plainchant were his staple diet the sharps necessary at the Kyrie final cadences (and the exceptional cautionary naturals in Credo, b. 146 and Credo Amen, b. 26) were presumably within his competence. It is possible, then, that Machaut arranged the

part in this way so as to help the singers and choirmaster in rehearsal to decide which unsigned notes were to be sharpened and which not.

INTRODUCTION

TO

THE

EDITION

101

Machaut must have recognized that the language of this work was extraordinary and liable to be misunderstood. We have already seen him offering help to the singers through voice-exchange. It may be that the Tenor is another such aid, offering a more consistent thread of certainty running through this most hazardous work. It certainly is a help. Gloria, bb. 23~7 offer, as an example, one of the most difficult passages. The Triplum c’ sharp resolves, if at all, only at the

d’ two notes later. The Motetus c’ sharp notated in b. 26 seems to imply that the preceding c’, in b. 25, was natural. This is confirmed by the unsigned Tenor c, from which it follows that the Contratenor g in b. 25 must also be natural. The status of the second Contratenor g, in b. 26, is

less clear. We have already seen that harmonic diminished fifths are possible and sometimes intehded by Machaut. Equally, melodic augmented seconds also happen, usually in the sequence FG sharp—A, as here. On the other hand, that pitch sequence is extremely unlikely for the Triplum in b. 26, not least because there seems to be a correlation

throughout the Mass between the placing of sharps and the duration of notes. Unusual sharps tend to be placed only before long notes—perhaps again a concession to singers. Thus the Triplum g' here is likely to be natural, and therefore the Contratenor also. The fs in b. 27 (T and CT) require the Motetus c’ to be natural; and the lack of a sign before it seems

to confirm the conclusion suggested by the placement of sharps throughout the piece, that sharps apply only to the notes immediately following; they are always ficta, in other words, and never involve hexachordal transposition sharpwards.° Hence the lack of F sharps, either notated or required, in these movements. Finally in this example we are left with the Contratenor c in b. 24. From the examples already seen I think we must conclude that a melodic diminished fifth in the Contratenor, c sharp-g, is a lesser evil than a harmonic augmented octave with the Triplum cc’ sharp, and that therefore this Contratenor ¢ must be sharp. At the same time I think we are entitled to ask why it is not signed with a sharp. One would have thought that precisely because it is such an unusual melodic progression the composer would have felt obliged to indicate to the singer what was required. Yet it is a disturbing pattern throughout these movements that sharps tend to be notated only where they are most likely anyway, and tend to be omitted where they would look most strange. This of course suggests scribal intervention. Scribes might insert sharps into an unsigned part where they looked most likely; or, scribes might omit sharps from a generously signed part where they looked least plausible. (This latter hypothesis would provide a better years 6 Exceptional in this respect is scribe of MS G, working perhaps twenty they that g assumin flats and sharps after A and Vg, who seems to have copied ’s Machaut not was this that lasted until the end of the line. A and Vg suggest view.

102

APPENDIX

explanation of Gloria 23—7.) But in Machaut manuscripts A, G and Vg we are not dealing with careless or ignorant scribes, These are people who, as we know, copied with considerable care what was put in front of them, And what was put in front of them was very close to Machaut, That the copies A, G and Vg used as their examplar(s) were at least confusing is suggested by numerous variants, The end of Kyrie III has already been mentioned. Considerable confusion also arose over the meaning of vertical strokes through the staves: sometimes copyists assumed that they were rests (as they often were), sometimes section divisions (e.g. Gl 56, S 48-9—see variant listings under ‘Pitches and Rhythms’); sometimes they copied them singly, sometimes as a pair (Cr 115). Even the order of sections seems to have been open to dispute: A copied the whole of Kyrie III after theSChriste before realizing his mistake and erasing the entire page to take Kyrie II. G also had doubts as to which music belonged to which Kyrie section (see KILI 5-6 variants); and a possible explanation for variants and erasures in the text-music fit at the join of the second and third ‘Sanctus’ invocations (S 9=11) is that the source for at least A and Vg indicated bb. 11-15 only by an instruction to repeat bb. 15. It seems quite possible, then, that all these three sources were compiled from materials which, whatever their format, were not a clean final copy and which retained inconsistencies, ambiguities and indecisions. This seems to provide a reasonable explanation for a variety of problems with sharps in the Gloria and Credo. If the source for (e.g.) Gloria 39-43 was provisional, that at least explains why there is a ¢’ sharp in all three manuscripts in the Triplum of b. 40 but not elsewhere, even if it does not tell us what Machaut expected the singers to do with the rest of the passage. As in the example of the Kyrie endings, it would seem that for an answer to this problem one can only rely upon empirical solutions of the sort that performers might have arrived at in rehearsal. Thus it seems probable that by b. 43, whatever has happened in bb. 41 and 42, the Cs and Gs are now

natural once more. It seems

likely also that the Contratenor g in b. 41 must be natural, partly by analogy with the Tenor line in b. 42 (f-g-d), partly because the lower voices’ progression from the end of b. 41 into b.42 seems to be essentially minor third g-over-e down to minor third frover-d, and partly because of the Contratenor fourth g—d. Their melodic contexts (D-C—D) may argue for the sharpening of Cs in Mo 41 and 42 and CT 42; but what are we (and what was the fourteenth-century singer) to make of the first beat of b. 41? Sanctioned by each main source and almost certainly, therefore, in Machaut’s copy, the Triplum sharp in b. 40 cannot readily be ignored. The editor cannot simply put a natural over the staff and wish it away. So the Contratenor cin b. 40 will have to be sharp. But then what? If Machaut wanted C naturals at the beginning of b. 41 he could have exchanged the voices in order to encourage them.

INTRODUCTION

TO

THE

EDITION

103

The only other possibility is that he did not mind: because these are very short passing-notes it did not matter. Uncomfortable as it may be, I believe that we must accept not only that surviving versions of the Mass are provisional but also that in at least some cases Machaut was not concerned to provide a single solution, but rather was content to leave performers to choose from amongst various possibilities. The best that an editor can do in such cases is to outline the range of possible choices. In determining the limits of this range a number of points need to be borne in mind. The parts as notated are not self-sufficient; that is, a singer could not always have inferred from his part alone which unsigned notes needed to be raised; this could only have been established in rehearsal.

The modern editor must likewise consider contrapuntal context in order to establish a plausible reading. He must be prepared to add sharps to produce perfect octaves, and to produce perfect fifths except where the parts are arranged (or the counterpoint is directed) in such a way as to encourage the production of a diminished fifth. I am less certain, however, that the editor should introduce sharps where none is signed in any voice, either to smooth out unconventional melodic lines or to make entirely unsigned C—D progressions sharp. It seems probable that both unconventional augmented and diminished melodic intervals and natural C-D progressions are part of the intended language of the piece, the latter (as Ch. 3 suggested) tending to occur in the middle of phrases and movements, and sharp C—D progressions at the end. It is with these principles in mind that I have—often with great reluctance—arrived at a final choice of sharps to recommend in the edition itself. I hope, with the aid of the foregoing examples and the

detailed commentaries appended to the Critical Notes, to have made the

process of choosing as transparent as possible: every decision should be open to change.

A.4 Plicas Machaut’s Mass is one of the last French works to use the plica; and by this late stage its meaning is not always clear. Its traditional function, filling a skip of a third with a passing note, can only apply in 40 per cent of the signs notated in the manuscripts. The remaining 60 per cent fifths, include unisons, ascending and descending seconds, fourths and

and notes followed byarest. Of these the first and last (unisons and notes before rests), if they indicate anything like the traditional plica, can only be ornaments returning to their starting pitch. Vg, alone among the scribes, makes a graphic distinction between these two classes of plica, marking those at the top of a falling third with

as a downward tail or curve to the right ( [), while the rest are notated 27 Tr r ions—C simple down-tailed breves ( f). (There are three except

104

APPENDIX

i,’ S Tr 65 v, and AIII Tr 9 iv—all down-tailed breves at the top of

falling thirds, but no cases of curved plicas in other contexts.) It is also interesting that of the 13 plicas copied in all three main sources (B copies Vg, and E transmits none) all but one are at the top of falling thirds, as are half the six transmitted by two manuscripts. In other words, the scribes were likely to agree about falling-third plicas but tended to make independent decisions about the others. Tables 1 and 2 will make the distribution clearer.* Vg appears the most independent (in line with other aspects of his work) and the most enthusiastic for plicas— he is the only scribe to transmit plicas in the Sanctus—and for this reason it would be unwise to place too much faith in the graphic distinction he makes between falling-third plicas and all others: there can be no certainty that the distinction is Machaut’s. For the same reason one must be wary of all plicas which are unique to him.

Table 1: Plicas A

G

Vg

Cr Al All

Alll A

G

A G

KI Cr

Mo 31, T 961i, CT 146 ii, Tr 156 Tino) eolet Ort et Siti 2 tt Tet at Ont Tr it it Grass Glamis

Vg

Cr

Mo 112 iv

Vg

Cr

Mo 181, CT 771i

A G

Vg

104 i, Mo 108 i, Mo 111 iti, Mo

Gl Gr

Prs2i Cr 435-1 107% (CE 116 11

Gl Cr

Te $2 i Oy TEs, Creer

Tia

Gl

Gio

Cr S

727 1, Cl 65 16 F777 u, dr tos i, Mo rog 1 Paro4 1, Mo 116 vi, Mo 1591 Mo 37 u1, Mo 45 ii, T 53 i, Tr $7 iv, Tr 65 v, Mo 65

Al

Tr 13 1, Mo 19 ii

iv, T 84 iii, Tr 90 iv All

Alll

SriO Mie lieleaicwlae2 tele

Tr9 iv

’ Lower case roman numerals refer to the symbol number within each bar. * In Table 1 falling thirds are italicised. Note that Vg was not accessible for bb. 56-115,

and that Vg plicas listed for those bars are taken from B. B,

however, does not follow Vg’s precise note-forms. In Table 2 falling thirds are bracketed, and are subsets of the unbracketed numbers.

INTRODUCTION

TO

THE

EDITION

10§

Table 2: Plicas quantified KI

Gl

Cr

BONE”

is.

6 (5)

A A

(1)

2 (1) I 2 (1)

G G

Vg Vg

:

A

G

Vg

3 3 (1) 8 (1)

2 I

S

8 (3)

AI

AII

Alll

(2)

(2)

(2)

artes

(1)

J /

With these points in mind the edition notates plicas in three different forms. (J. j)) and( J. J) show falling-third plicas except those transmitted only by Vg, with the parentheses implying that their performance is optional (their meaning being uncertain, although the traditional passing-note is supported by variants at Cr 3 and 156—see Critical Notes). < J.) > indicates falling-third plicas unique to Vg. t above a note indicates all other plicas. As to how this last group should be performed we have no evidence: rising thirds may be filled with a passing note (e.g. Cr CT 6s i), larger skips may be filled with an intermediate pitch (Mo 45 ii, f’) or even a melodic link (Cr T 77 ii, e.g. crotchet d, two quavers e,f), and at unisons and notes followed by a rest performers may wish to experiment with mordents or other discreet ornamentation. I have omitted entirely any special indication for up-tailed longs ( @ ). As Ludwig observed,’ they are unlikely to be plicas: they occur only at the bottom of the staff, above text, and thus are almost certainly simple longs whose tails were directed upwards only in order to avoid colliding with the text. There are, however, two exceptions: AII Mo 1 v has an

up-tail in A and G but is on the middle line; and B copies the AII T 1 uptailed long in Vg as an ascending plica ( ). These raise enough doubt to persuade me to list all up-tailed longs in the Notes (in parentheses after ‘Plicas’) for the use of those who—perhaps rightly—remain unconvinced.

A.5 Transcription The parts have been laid out in the order: Triplum Motetus Tenor Contratenor 9 Friedrich Ludwig,

1928), p. 47*.

Guillaume de Machaut: Musikalische

Werke, ii (Leipzig,

106

APPENDIX

The original notation is full black without any coloration, and is written on five-line staves. In the edition note-values have been reduced to a quarter of their original length. Original clefs and clef-changes are listed in the Notes. Ranges are indicated (for analytical purposes) at the beginning of each section, immediately after the clef. For the total ranges see p. IIS. Barring is in regular three-beat units in the isorhythmic movements (corresponding to the original modus). In the Gloria and Credo Machaut works irregularly with duple, triple and quadruple rhythmic units; and barring in the edition has followed these, aiming as far as possible to avoid ties and their implications of intermediate downbeats. Assuming that the music will not normally be conducted I have avoided intrusive modern

time-signatures, but have included simple numerals

(2, 3, 4)

where the bar-length changes in order to aid the singers’ counting in rehearsal. Beaming of quavers is editorial (the original values were minims) and aims to be as neutral as possible. Although the manuscripts often group minims (for example || 1..1, transcribed here as JJJJ J J), this is for purposes of texting or as a reminder of the position of the underlying beat, not an indication of phrasing. Ligatures have been indicated with the usual horizontal brackets above the notes which in the manuscripts formed a single symbol. Ligature variants are listed in the Notes. Bars are numbered in fives above the system. Isorhythmic taleae are numbered with Roman numerals beneath the system.

A.6 Texting For the same reasons that pitches and rhythms can vary from source to source, so too can the layout and quantity of text. In each source some sections are fully texted in all four voices; some have text in Triplum,

Motetus and Tenor but only the first word or two (an incipit) in the Contratenor; while in others the text (entered first by a different scribe) was ignored by the music scribe when he came to set out the notes belonging to it. Once again it is necessary to try to establish the composer’s intentions from an examination of the surviving variants. The pattern of texting amongst the three main sources is shown in Table 3. This rather confused situation is explicable in the light of Machaut’s compositional procedure. Because of the way in which the rhythms depend upon the quantities of their texts, the Gloria and Credo must have been texted, throughout the process of composition, in all voices

(or, as Chapter 3 and the texting of the initial T and CT rests at ‘Patrem’ suggest, beneath a system of score). Consequently, the Gloria and Credo were copied with four texted voices thereafter. The other movements,

INTRODUCTION

TO

THE

EDITION

Table 3: Texting Kyrie I

A

G

Vg

maaan (-------- yo Rents once ceee

eeee)

aleNe Ley rd

Eee

(--------) pe ENEae

Christe

Kyrie Il

(INCIp IE Re Sts peo

(

Kyrie III

ae

OES

(

(

poSea lee

ESE

Bee a

Tapas Yee

) 0 ae )0

J ae ee sen

yo anne ) 0 ae

Gloria, Credo

Sanctus

Dominus

Se Maat

kserate,

ano.

Osanna

fncipi)-_

Gndipit)

Gncipin

(incipit)

(incipit)

(incip. + etc.)

Benedictus

108

APPENDIX

SSS

(incipit)

No TUS TD rete

(incipit)

(

ee

)

et

ar

(incip. + etc.)

ee

ee fee a Se

Fe

aaa

Se

te

oS

(incipit)

(incipit Agnus II

eee

ieee

Agiitisl: sy jo Aouonboay ay YM

‘sdo ovy = vp uvaioseyiAg YUM

UE IY} VION > “s1u20 06 ‘£¢z 2 9SZ auOIUIIS INEUTOIYD & ‘sJUaD HII JO YIPIM & puL gror : LEIZ ONE 9Y} SeY SUOMas DTUOUIIEYUS 4

aaoge uaaid stj3ey gq :dieys y pur rey y IVY D ‘dieys

¢_ 9+ + OI

SCI +

quaureroduia) yenbo WIOIJ $JUID

zoL719g6rgov

UWIOIJ SUID)

?

ip

ne

Lip

Aguonbo1j

Sse] SaNeyoeyy OJ ofeos uvsioseyiAg :¥ eTqey

aH

o'60£ £°£67 o:orr LeLEy €-DZe ooLe o' Stor S:1z$ S6¥ L'097 ¥'g3lz 1'16E LLVe

Aouonbe14

£°6L7 6 €°V6z S-1br 1°61¥ 1'SOQV £7 L964 gre 8cS 9197 TOme S°ZLE zee ¥-Z6E

4H

: gvoz Lgiz

‘

Vil oO

‘yuourerioduiay ‘eagt-yosd Aue ye sojeos uvr03ey34qg sonpoid oj paururesidoid 2q 03 szastsayjuds-iojnduros a[qeue pynoys sone pue sopusnbary oy], sy], :sa0Ny Jenba wroyy syus0 UT a0URISIP 2Y}3 SILOIPUT YTYM sI9UNy STUOIDe[9 YIM pasn aq p[noys (uumnjoo yyy oya Aprepnonsed) syuso ur somBy

s1u99 Ul [easoquy

960 71S

$9:18 LEC 8:6 Gs : :67cL vias 19$9 re

Tz 6:91 Ol EC QZ: £vz

Oey 0} 3

966906918 oOOozI OIII 2.

+ OI

v6zvoz f

Oo

VII 06

vil

06 vil 06

06 06 VII 06 VII

06

107, EDITION THE TO INTRODUCTION

118

APPENDIX

noticeably sharper than in modern tuning. It hardly needs to be said, though, that these figures represent nothing more than a model. In practice, singers will make innumerable adjustments; but they should be adjustments to a Pythagorean and not an equal-tempered norm. At first it may feel a little strange, but with practice performers will find the music and the tuning encouraging each other. Experience suggests that dynamics can be as loud as is comfortable for the singers. This is not music designed to fill Reims cathedral, but it should be clearly audible in the vicinity of a side chapel or the choir. The many short rests and syncopated figures suggest that articulation in the upper voices should avoid long=breathed legato; while the extended vocalisations of Tenor and Contratenor can make discreet use of dynamic shading and accented attacks, always listening closely to the direction of the upper-voice pair. As with dynamics and articulation, we have no certain evidence on speeds, but a rate of minim (in the edition) = 95 allows texted quavers to

be articulated without breves becoming interminable.

A.g Liturgical reconstructions Any performance of the Mass requires plainchant intonations at the start of the Gloria and Credo and probably (but see Sect. A.6 above) Ite; and in the edition these have been supplied, the Gloria and Credo

from

contemporary chant manuscripts, the Ite from the Tenor color. In modern times the Kyrie has usually been sung alternatim (KI-chant-KIchant—X—chant-KII-chant-KIlI),

taking

the chant

from

the Tenor

pitches of KI, X, X, KII respectively. Repeat markings in the manuscripts, however, imply that Machaut intended the polyphony to be repeated; and this has been recommended in the edition.

Performers wishing to place Machaut’s music within the context of a full liturgical reconstruction face a huge task, for few of the materials have yet been published. A model for the musical elements in the service will be found in the Parrott recording mentioned above, which includes all the music for High Mass on the feast of the Nativity of Our Lady as sung in Reims cathedral. A model for an entire service, including spoken material

and ceremonial

instructions,

may

be found

in W.

Thomas

Marrocco and Nicholas Sandon, The Oxford Anthology of Music: Medieval Music (Oxford University Press, 1977), item 11, which describes a pre-Reformation Easter Mass from Salisbury cathedral. More detailed information is published in Nick Sandon, The Use of Salisbury, i (Newton Abbot, 1984); but the reader should bear in mind that habits at

Reims were significantly different.

For an authentic reconstruction it is still necessary to go back to Reims service-books. The following, all in the Bibliothéque municipale at Reims, may be helpful:

TO

INTRODUCTION

THE

EDITION

119

é

MS 217, 224, 264, 265,

noted noted noted noted

Missal, Reims cathedral, C13 Missal, Reims cathedral, C15 Gradual, St Thierry, C12 Gradual, St Denis, C12

285, Sequentiary, St Thierry, C16 295, 330, 333, 344,

Lectionary, Reims cathedral, C11—15 Ordinal, Reims cathedral, C14—15 Processional, Reims cathedral, C18 noted Pontifical, Reims cathedral, C14

An invaluable guide to the arrangement of medieval service-books is Andrew Hughes, Medieval Manuscripts for Mass and Office: A guide to their organization and terminology (Toronto, 1982). Daunting as the task of assembling a service must seem, the result will more than justify the effort. It is only in a complete liturgical setting that Machaut’s polyphony can make its true, miraculous effect.

CRITICAL

NOTES

For each section of the Mass the following information is given, as appropriate: Section title Folio numbers: alphabetically by source. Intonation: lists chant manuscripts used and variants. Part names: lists variants, by source and then by part. Texting: lists all instances in which the alignment of text and music is other than that given in the edition, arratged by voice, bar number, note number (lower-case Roman numerals), text (between inverted commas), source. Variants from MSS B and Eare not included (see Sect. A.2 above) unless they offer additional information. Ligatures: arranged in the same way as ‘Texting’, but including B and E variants.

Clefs: initial clefs listed by voice (variants in parentheses), subsequent clef-changes listed by source, voice, bar, next note following new clef,

new clef.'

New lines: listed by source, voice, bar, note, (new page). The ‘new page’ folio references (indicating page-turns only) specify the page (MSS B, E, Vg) or column (A and G—columns a and b on each recto, c and d verso) as appropriate.

Pitches and Rhythms:

differences from the edition ordered by bar

number but listed as voice, bar, note, variant, source.” Symbols which

are faint in the manuscripts and erasures are included in the listing as an aid to users of microfilm. Plicas: lists all plicas, ordered by bar but listed as voice, bar, note, source.

(Followed bya list of up-tailed longs, arranged in the same way.) Sharps and Flats: variants are given first (including variants in horizontal and vertical position), ordered by bar number but listed as voice, bar, ' The order of elements—e.g. under ‘Texting’ voice, bar, note, source, but under ‘Clefs’ source, voice, bar, note—is determined by the route which most

readers are likely to follow if they need the information. Thus text overlay or the placing of sharps and flats are most likely to be questioned directly from the edition, while clef or line changes are more likely to be investigated in relation to questions about a particular source. * It is hoped that by placing voice-name first and source last in each entry the reader will be enabled to search by voice or by source as well as by bar reference.

CRITICAL

NOTES

note, variant, source. A commentary

I2I

follows on each case of doubt,

setting out arguments for all solutions which seem possible.

Commentaries are appended to other categories as appropriate. é

Variants printed in bold type are those which performers may wish to consider as plausible alternatives to the readings of the edition. Arabic nurherals refer to bar numbers, lower-case roman to symbol numbers (counting notes and rests) within each bar. The pitch system is C-c—c’ (= middle C) —c”.

A list of abbreviations will be found on pp. xii—xiii.

Kyrie I A B EG

Vg

fo. 11°.xxxvily” ; fo. 281° r i fos C.banj-Clikv ws fo. 125” (facsimile in Heinrich Besseler and Peter Giilke, Musik des Mittelalters und der Renaissance (Musikgeschichte in Bildern, III/s; Leipzig, 1981), p. 61) fo. CClxxxiij’ ‘Ci comme[n]ce la messe de nostre dame’

Texting Tr

Mo

T

CT

1-2 ‘Kyrie’ AEG 1—3 ‘Kyrie’ Vg 15 ‘e Vg 26 —27 ‘Leyson’ E af 26 —af 27 ‘eleyson’ A 1-2 ‘Kyrie’ AEG Vg 17 ili, bf 20, af 24 iii, 27 ‘e’ Vg 25 ‘ley’ (of ‘eley’) AG af 27 ‘ley son’ Vg 1-3 ‘Kyrie’ A 1—2 ‘Kyrie’ G Vg 2ear ashe, Vg bf 21 — 22 ‘eley’, 25 ‘son’ G 24 ‘eley’, af 27 (new line) ‘son’ A 1-2 ‘Kyrie’ A G Vg af 19 ‘e’ Vg 22 ii—24i‘eley A an-4 ley, 24,1. som. VE 26 1 ‘eley’ ii ‘son’ G

Commentary

Although Vg spaces ‘Ky-ri-e’ as three syllables in KII Tr and Mo, it may have been sung more usually as two (‘Ky-tie’); this would help to

CRITICAL

122

NOTES

explain the failure of the text scribes to space it or the music scribes to overlay it with any care here. Although I have followed Vg’s hint, performers may wish to experiment with two-syllable pronunciation. Similarly, ‘ley’ appears throughout to be regarded as one syllable, confirming Dr Wright’s opinion that by this time it was no longer spoken as a diphthong (see Sect. A.7 above). For the ‘breaking’ of the final CT ligature see Sect. A.6 above. The repeated ‘e’s in Vg relate to the ‘or’s traditionally provided by text scribes as an indication to the drawer of staves of the extent of the untexted T and CT parts on each page (see Earp, pp. 186-7). Here the final ‘leyson’s accomplish that; and therefore the ‘e’s can be read both as reminders and as text.

Ligatures Tr Mo alt

267EG 26-7AEG 25-7 AEG

Clefs TeLoer

Mo

C3 (C2 B Vg)

T Cie

C4 Gs.(CAE)

BN

(GAR

okey C5

New lines A

a Mo 8, 19, 25 ay 18 CY “10,20 air 2s Mo 24 iv Abate) ir

E G

Mo 12, 23 ill ct 21 CTT edt eo att

Vien Bo

Tes, Mo

nbs 20

Pitches and Rhythms CT

2no rest, dotted long E

Ie

Berioie)s}

Mo

4 tail very faint G

Mo

5 af ii extra br b B

CRITICAL

GE Mo Gr eLsr: i Mo

NOTES

123

10 af ii erasure G

I1iom ii long G 12 bf i void br f’ later add {to correct CT?) G 14—151 br long B

15 HEE

16 26 Cr 27 itr: 27 Mo All af 27 Mo

af iii three-br r E ii— 27 ed (from T) E max BG max BG repeat symbols :|||: AB

EG Vg

Sharps and Flats: variants slor Lr Mo

4 # very faint G, no $ E

‘Sik CX Mo Cr Mo GE GE Mo Mo

6 # in e-space B, noZE 10 # bfi A G, no ¢ B Vg" 1nofE 17 # in b-space G inofAEG 20 above i # erased from c’-line A 22 # in b-space G, # bfi A B G Vg 25 # in d'-space G, no fE 26n0of AEG

spABGVg,nopbE ‘ 6 # in d’-space A G, no #'E

Commentary

Tr iv g’ # possible, but natural perhaps preferable on account of melodic continuation through f’. 18 CT ii natural because of melodic continuation to g. 22 Tr ii must be natural with T f, vi natural preferred after 11. 25 Mo ii forces CT #, though Mo natural a possible alternative (# may originally have referred only to 26) 26 Tr iv and CT ii naturals may just be possible, particularly if 16 Tr iv and 25 Tr and CT were sung natural. 26 Mo ii # is not negotiable, however. 16

Christe

A B E G

fo. 1i°ij.xxxix" fo. 282° (as KI) (as KI)

Vg fo. CCl xxxiij’

124

CRITICAL

NOTES

Texting Tree

1 Xa

2 ave'ste’ A

1 “Xpriste’ G I—2 v ‘Xpiste’ Vg 7 iv, 8, 91, af II ix, 15, 17 iv, af 21 iii “e’ Vg

Mo

20 v—21 iii ‘eleyson’ G 21 vi—22 ‘Leyson’ E af 21 —22 ‘eleyson’ A 22 ‘leyson’ Vg 1-3 i ‘Xpiste’ A Vg I—2 ‘Xpiste’ G 10 fi; af 17 Hi, It iv, al 17-4) at 20 € Ve

nt

CT

21 iii ‘eley’ G 21 iv ‘eley’ A 1 ‘Xpiste’ A G I1—2 ‘Xpiste’ Vg 15 ‘e Vg 18—21 i ‘eleyson’ A 20-21 ‘eleyson’ G bf 1 ‘Xpiste’ G b= 2) Apisic

81-92 passim, 97, 100, 109, 160-7 Credo 14-16, 17n., 29-30, 31, 43-5,

$0, 53, 56-69, 71, 72, 73-4, 78-9, 80-95 passim, 97, 100, 102, 106, 110,

115, 118, 131, 139-52 Gloria 14-16, 17n., 19n., 29-38, 39,

motet

composition

22-3,

25, 27, 37, 38,

44, 87 see also isorhythm Murrin, Iacobus 16n. musica ficta/musica recta 60, 71, 99, IOI

see also sharps and flats

42-3, 45, 53, 56-62, 66-9, 71, 73-4,

York, Waildenstein collection, see Machaut, Guillaume de, manuscripts,

78-9, 80-95 passim, 99, 100, 102, 106, I10, 115, 118, 131-6, 137-9 Ite missa est 14, 16, 29, 51-2, 69, 83-02 passim, 109, 118, 167-70

notation 18, 43, 55, 67-9, 80-1, 98-110 passim, 120-70 passim see also Machaut, Guillaume de; manu-

Kyrie 14-29, 31, 43, 47, 48, 50, 53,

New

Vg

scripts

69-7007 ie 72,0813) 745,07 O18,

81-95 passim, 97, 99, 100, 102, 109,

Sanctus

Ordinary

of the

mass,

see

mass

music,

chants for the Ordinary

IIO, I1§, 118, 121-31 14-15, 16, 17n., 24n., 29, 44,

45-9, $0, $2, 69, 70, 71, 72; 73-4

Paris, Bibliothéque nationale, fonds fran-

76, 78, 81-92 passim, 98, 102, 104,

cais, MSS 1584-6, 9221 & 25546, see Machaut, Guillaume de, manuscripts Parrott, Andrew 115-16, 118 part-writing 55-72, 75-7, 99-101, 103,

109, 153-9 see also cadence; chant; chronology; harmony; hocket; instruments; iso-

rhythm;

link passages;

mass music; melodic

part-writing; rehearsal;

manuscripts;

writing; mode;

performance

rhythm;

sharps

practice; and

flats;

130, 136, 148-9, 150, 152

performance practice

18—19n., $7, 59-60,

63, 98-119 passim see also Critical Notes, Commentaries

Peronne (d’Armentiéres?) 6, 92

solmization; text-setting

Petrus Frater dictus Palma Ociosa

motets, see motet composition Nes que on porroit (ballade 33) 94 Ploures, dames (ballade 32) 92-5

73

Phillipe de Vitry, see Vitry, Phillipe de plainchant, see chant; mass music, chants

Quant je ne voy (rondeau 21) 74 Quant Theseus/Ne quier veoir (ballade

for the Ordinary Power, Leonel

17

34) 38n., 94 Tu/Plange/ Apprehende/C T-(motet 22) 8, 10 Voir Dit, Le 38, 94n.

Machaut, Jean de 10-12 manuscripts,

see Machaut,

manuscripts

Guillaume

de,

Rachel/Ha fratres 44n. rehearsal 100, 102-3 Reims I—5, 9-13, 97, 114-15, 118 rhythm 36, 38, 43, 53, 87, 90, 106, 110

see also isorhythm Roman de Fauvel, Le 20

179

INDEX

Sanders, Ernest H. 25 n.

é

Sandon, Nick 118 | Savoy, Amadeus VI, Count of, see Amadeus VI, Count of Savoy sharps and flats 57, 59-60, 62, 68, 71, 76,

120, 121-2, 124, 126, 128, 132, 140-1, 149, 154-5, 160, 163, 167-8 Trie, Guillaume de, see Guillaume de Trie Tournai mass 12n., I§—I6, 24n., 38-40

130-I, 135-6, 138-9, 145, 147-50,

Verrier, Jean de, see Jean de Verrier Vienne, Jean de, see Jean de Vienne

152, 159, 162, 165, 167, 169

Vitry, Phillipe de 4, 6-7

97, 98-103, 120-1, 123, 125-6, 127-8,

solmization 59, 60, 76, 159, 165 Sorbonne mass 15 Stewart, Rebecca 116

Douce/Garison/Neuma 4n.

Susay (Iehan de?) 16n.

Weyen, Canon II Wright, Laurence 110-14, 122

text-setting

Zacharias de Teramo, Antonius

15, 22, 27, 29, 31, 35-7, 39-43%,

49, 50, 53, 65, 67, 79, 80-1, 106-10,

17n.

Guillaume de Machaut

La Messe de Nostre Dame

LA MESSE DE NOSTRE DAME Edited by Daniel Leech-Wilkinson

GUILLAUME DE MACHAU (c. 1300 —1377) Es

Kyrie

ranges

, XYRIEI

f

st a

ee =

aioe

St

ace

ge

———

fae

ease

:

=

a

BSS

ae

)

Se

SSS re

SS

SSSS

f=

——— it

Printed in Great Britain

© Oxford University Press 1990 OXFORD

UNIVERSITY PRESS, MUSIC DEPARTMENT, WALTON

STREET, OXFORD

Photocopying this copyright material is ILLEGAL.

OX2 6DP

LA

MESSE

DE

sa

=

oO

fs

aSsSoEp

rE

a

a

ae

Sing three times

==

Ss

ese}

DAME

Ss

=a

&

NOSTRE

te

-—

=—_-

i

fe

:

ee

—,

oo

[

=

- ley -

==

=

Sc

=

=

E

6 o-

3

e-

- son

(#) ———

SSS

ley -

- son

i

Ili CHRISTE

(—S—

ste

Chri

es

=e

Seay

Chri

=

ste

=

= SSS ———————— 4

oo

oa

=

:

SF

=

3

pa

: cua

é

SS

=e

az

——

a

SSS

ze

ieee

=

ae eee

ae

3

7

SSS

=

SS

=

|

i

=

7

LA

MESSE

DE

NOSTRE

185

DAME

SS

See

=

5

SS SS SS Ss

SSS SS

) 5

3

2S

=

je SSeS

ae See

=

ee SS

=

i

186

LA

MESSE

ee

f= ee

DE

NOSTRE

DAME

=

Ss

————

—

SS ee fee + Se

ees —

SS ———

Zz

#

Cet

Se

SS

5

Se a

a

j

- ley -

- son

ee

———

- son

Se Fe

$l]i9

Iii! 9

se

=

=e

SS

————

iL eGhL GL >

ae

dll

i

as

ate

_—S

4

- son

'

SS

CS

Lo

=

=

=:

Sing twice

- son

RUC Yates

~ ley -

soe es

Ss

ae

Haas Pe

=,

SS

Z

e

LA MESSE DE NOSTRE DAME

SSS

= os

187

SS

Sa

SS ———

c= =

=

—

Fae

je

(aS ee i =

==

= SS a

:

ZS

St

3

188

LA MESSE

DE

NOSTRE

DAME

aoaeee

Se

=son.

d

= =

==

|

ley

- son.

€

SSS

I

Se

e

suet

(8)

=

HO) =

SS

Seta eer Car

ota

f f

e

e

ley

- son.

- ley -

- son.

o-

|=

j

et

ae

e-

Gloria t

TRIPLUM

¥ Gh

SSS

esi ”:

MOTETUS

Et

in

4 - Ap =

:=

5

TENOR

=

Et

CONTRATENOR

3

Et

in

4 : Z- =

f=

Et

in

ter -

- va

16 =

-

ho

-

fz

pax

at =H = +

lo: =

F

=

iP

pax

5 =aE =

ter -

- ra

bus

ies

pax

a

bo

-

ne

vo

SS

-

bus

bo

-

ne

vo

—

-

ni

-

bus

bo

-

ne

vo

mi

-

ni

-

bus

ne

vo

—— ho

=

- 7a

.

a

ei

4 aa

ter -

fe

pax

rest,

_——SS_= mi

= fz =

- 1a

6 =

Sa ho

ho

= zt

ter -

f =

i

at

6: =

in

4a

ge

ex-cel-sis de,

-

SS

a bo

SSS -lun-ta- tis; Lau - da

-

mus

te;

- lun- ta - tis; Lau - da

-

mus

te;

- lun-ta - tis; Lau-

da

-

mus

-lun-ta- tis; Lau - da

-

mus

SSS

-

-

Ss Be

-

ne - di - ci-mus

-

ne-di-ci- mus _ te;

te;

-

ne - di - ci-mus

te;

-

ne- di - ci- mus

SS

SSS ei SSS SSS SS= te;

-

LA

MESSE

DE

NOSTRE

DAME

189

190

LA

és

Qi

MI “