Le Magnus liber organi de Notre-Dame de Paris, vol. 1, Les quadrupla et tripla de Paris.
The edition offers transcriptions of the three four-voiced compositions (two organa and one clausulae), of thirty-eight cantus-firmus settings for three voices (organa and clausulae) from manuscripts Florence, Biblioteca Medicea-Laurenziana, Plut. 29.1 [F] and Wolfenbuttel, Herzog-August Bibliothek, Helmstedt 1099 [W2] (Wolfenbuttel, Herzog-August Bibliothek, Helmstedt Con [W1], which is, of course, accounted for in the transcriptions and the commentary, is not identified in the table of contents, though it contains forty percent of these compositions), of four clausulae in F (one of which is also in W1) on segments of cantus firmi not known to have been "organized" for three voices, and of fourteen compositions for three voices from other manuscripts (mostly Montpellier, Faculte de medecine, H 196 [Mo] and Bamberg, Staatliche Bibliothek, Lit. 115 [Ba]). The transcriptions are presented calendrically, "combining office and Mass chants into a single series" (p. lxxi), and with O and M numbers (according to Friedrich Ludwig's numbering) given in the critical apparatus. This arrangement is essentially that in F, but expanded by the insertion of each of the Notre-Dame clausulae after the appropriate organum. (It is debatable to what extent this purported liturgical use of the clausulae reflects original intention.) in addition, all instructive table in the introduction presents the compositions in calendrical order, but separated into two groups, for Temporale and Sanctorale, respectively, with Mass settings following Office settings for each occasion, as applicable. O and M numbers, which would have been helpful, are not given.
Several compositions from other sources are excluded for various reasons, chief of which K non-Parisian origin. Regrettably, this applies inter alia to the three-voice hocket clausula Portare in Paris, Bibliotheque nationale, lat. 11411 and to the four-voice In seculum clausula in a Salzbburg manuscript (Universitatsbibliothek, M II 345). Whatever the origin of these pieces, they are valuable compositions, no less so than the In seculum longum in Ba that is included and identified, in accordance with Anonymus IV's report, as made by a Spaniard (pp. x, 262, 352).
The transcriptions are preceded by material in both French and English versions - a general preface; acknowledgements; and an introduction containing commentary on the repertory, the sources (very full and informative and with individual bibliographies for each), and on the edition, and notes on performance. They are followed by the appropriate plainchants, impeccably edited by Michel Huglo, and the sort of exhaustive Apparatus Criticus kin English) characteristic of Oiseau-Lyre editions. The bibliography of the Apparatus unfortunately fails to include a good many titles cited in the preface and the introduction, and with respect to the source bibliographies one troublesome inconvenience is that any work cited for two or more sources is fully identified only the first time, where it is given a siglum by which it is cited subsequently; decoding such sigla can involve laborious searches.
The two quadrupla and two of the tripla are properly identified as having been composed by Magister Perotinus, in accordance with the few remarks by Anonymus IV in De mensuris et discantu. (It is gratifying to read Roesner's defense of the existence of Leoninus and Perotinus against Hendrik van der Werf's speculative attack on their historicity; see "Anonymous IV as Chronicler," Musicology Australia 15 : 3-25.) It would have been good if the editor had expanded the sparse indications in the treatise that these were not the only organa for more than two voices by Perotinus, by including in the introductory material a brief account of earlier scholars' (particularly Ludwig's) suggested additional attributions to him.
On the other hand, Roesner's use of the title Le Magnus liber organi to stand for all Notre-Dame polyphony is an unjustifiable expansion of the brief but specifically focused remark in our medieval reporter's treatise that Leoninus "made [confected] a large book of organal polyphony concerning the Gradual and the Antiphonary for the enhancement of the Divine Service." (That the term fecit in effect means "composed" is apparent from its subsequent uses, as for instance with Perotinus's production of clausulae.) This was no book entitled Magnus liber organi - an opus 1, as it were - but it apparently was the first (a large one) of a number of volumes of polyphony accumulating in the library of Notre-Dame; in fact, in his descriptive list of these volumes Anonymus IV did not refer to it as "the Magnus liber" but identified it by naming its first composition. The title for Leoninus's production, though incorrect, has become acceptable through ingrained musicological custom. (There is no reason, however, to doubt Anonymus IV's testimony regarding the erstwhile existence of the book and "to steer younger scholars away" [Van der Werf, in Musicology Australia, 22] from an assumptive mirage.)
Roesner's remark that "the magnus liber is generally regarded by present-day scholars as consisting of organum duplum and clausulae" (p. lix) is ambiguous, leaving the reader with the surprising impression that all the hundreds of extant clausulae are part of the magnus liber. It is more likely, however, that he here takes the term to apply to the largely premodal discant passages occurring as parts of organa dupla in one or more of the three extant versions of the magnus liber. (It strikes me as desirable, indeed essential, to maintain the customary musicological distinction between the discant portions and the generally modal and on the whole separately collected clausulae.)
Extending his somewhat problematic assessment he presently states that "the magnus liber organi is a great book of polyphony,' organum here meaning ... musica mensurabilis itself" (p. lix) and concludes that the contents of the Florence and Wolfenbuttel manuscripts taken as a whole represent the magnus liber organi of Notre-Dame in different stages of its development" (p. lx), subsequent, presumably, to an irrecoverable original collection, whose version of Leoninus's organa might already have contained revisions. These books consist or consisted of "the different volumina ... that held the repertory" (p. lix). The extant Gesamtausgaben, did not come into being before ca. 1240, that is, in or near the year of death of the person with the "most likely identity for Perotinus" (Craig Wright, Music and Ceremony at Notre Dame of Paris, 500-1500, [Cambridge: Cambridge University Press, 219). It seems appropriate to venture the assumption that this event would have stimulated the inscription of comprehensive up-to-date collections (especially F) of the astounding repertory that began with that big book by Leoninus. (None of those books, however, seems to have had an author's title.) One wonders whether the anachronistic magnification of Anonymus IV's reference to a magnus liber (produced ca. 1180) was due to a nationalistic desire for a titre magnifique for the publication of the series (under Roesner's direction) of which the book under review is the first volume. The series title reveals a very large undertaking: MUSICA GALLICA: Edition des oeuvres du patrimoine musical de France....
It is possible that early in his career "Perotinus composed in organum purum as well as discantus" (p. lix), but the absence of evidence for Roesner's claim that it is "obvious" is not surprising, especially in view of Anonymus IV's identification of Perotinus as optimus discantor etc., better at writing discant than his predecessor, whose organa contained long sections of organal polyphony not subject, according to the thirteenth-century commentator Johannes de Garlandia in his De mensurabili musica, to the rules of discant. More unexpected is Roesner's subsequent assertion that there is no reason to suppose that [Leoninus] wrote only for two [voices]" (pp. lix-lx). Anonymus TV singled out Perotinus as, inter alia, composer of polyphony for more than two voices (up to four) and described Leoninus as optimus organista, who, therefore, cultivated a style necessarily restricted to tenor and duplum. There seems, then, to be good reason for the deduction that it was Perotinus (presumably in the 1990s, some years before Leoninus's death) who first recognized that the nature of discant, as experienced in the composition of caudae of conducti a 2, made possible the composition of organa for more than two voices and, in fact, required its application to the upper parts of organal sections in organa cum alio [cantu], as Garlandia put it. (In his "Johannes de Garlandia on organum in speciali," Early Music History 2 : 140 Roesner has argued that alius [another part] and organum here should be taken to refer to the tenor and upper voice, respectively, of organum purum. The following circumstances argue against this reading: (1) in contrast to the double meaning of discantus, which is both one of the three species of polyphony and a voice discanting in recto modo, i.e., mensurally, against the tenor, another discanting voice, or both, organum does not mean an organal voice; (2) the tenor could not have been referred to as alius, since Garlandia designates it as primus cantus, a term embedded in a long and continuing tradition - e.g., Anonymus VII; dignior pars; (3) in his discussion of organal polyphony Garlandia uses the expression organum quantum ad discantum; this can only be synonymous with organum cum alio, i.e., for three voices.)
The artfully interlocking counterpoint of organum for three (or four) voices forbids the sort of modernizing manipulations to which Leoninus's organa were subjected, first and foremost by Perotinus. It stands to reason, therefore, that these compositions suffered "relatively few changes from source to source ... so that the musical texts are more-or-less stable by medieval standards, permitting something approaching a 'critical' edition" (p. lx). Yet, asx) a good deal of the subsequent discussion will make apparent, this beautiful book reminds us that, alas, we continue to, and to some extent will probably always, confront uncertainties of knowledge and empathy, causing us to make some dubious choices in our attempts to present the repertory in transcription.
Roesner (p. lxxxiii) offers three options for the performance of the da capo of the Alleluya respond after the verse: (1) chant; (2) repetition of the polyphony, followed by the chanted rendition of the iubilus (or of the entire respond); (3) abridged performance of the polyphony, followed by chant. In support of the second (and occasionally the third) alternative I cite the admittedly later and more complex English Alleluya setting edited as No. 70 in English Music of the Thirteenth and Early Fourteenth Centuries, ed. Ernest H. Sanders (Polyphonic Music of the Fourteenth Century, vol. 14 [Monaco: Oiseau-Lyre, 1979]), which provides a new polyphonic version for part of the da capo and then adds a shorter alternative da capo, marked Vel si brevitas hore deposcit, dicatur sic. (Also see Frank Ll. Harrison, Music in Medieval Britain [London: Routledge and Paul, 1958, 2/1963], 61-3, and English Music for Mass and Offices (I), ed. Frank Ll. Harrison, Ernest H. Sanders, and Peter M. Lefferts (Polyphonic Music of the Fourteenth Century, vol. 16 [Monaco: Oiseau-Lyre, 1983], xv).
"F provides the copy text for most of the editions in this volume" (p. lxxi), that is, for all settings of which it is a source. Roesner recognizes the editorial issues associated with the customary preferential choice of one source, when it furnishes one of two or more alternatives whose impartial consideration might well lead to a more desirable editorial decision (p. xciv). There are, however, cases in which he chose not necessarily preferable versions. For instance, the five-note figure with which the duplum and triplum of No. I (Viderunt) begin appears in the three upper voices altogether seven times, in all but one of the first seven phrases. Of the three manuscripts preserving all of this famous quadruplum, Madrid, Biblioteca nacional, 20486 [Ma], a Spanish source, is the only one that presents this figure with two ligatures (3 + 2) throughout; London, British Library, Egerton 2615 [LoA] does so four times. The remaining three occurrences in LoA as well as all seven in F are spelled as quinarie. Without accounting for the diverse notations in the Commentary the edition offers all these phrases according to Ma (3 + 2), which seems odd, not only because F is claimed as copy text, but because the visual gesture of consistently applied five-note ligatures can be seen as having implications for the performance of these phrases.
More significant is the case of No. 9 (Alleluya: Dies sanctificatus), whose discant setting of no(bis) exists in two versions (W1 and N Cambridge, University Library, Ff. 2.29 [Ca], an English source). Notwithstanding the declared function of F as copy text, the Continental version, regarded as a variant (p. 325), is excised from the organum and presented as a separate clausula (No. 10), though it has no such existence in the sources (and was therefore excluded from the count of organa and clausulae in this review). Hence, the Ca version is given as part of No. 9, "because it seems more consistent with its context" (p. 323) a statement that, because of the English variant's slower and more stately rhythms over double longs in the tenor, is contradicted by the evidence on the printed page and by comparison of No. 10, 13-15, with No. 9, V. 759-61. Not only, therefore, are these two versions representative of the cases that show the slower counterpoint over double longs in the tenor to be a modernization, but the four old-fashioned unsupported fourths in the clearly older No. 10 are reduced to one in Ca. (The fourth in 307 on p. 61 is due to an error in the transcription.) Finally, the English source is later than W1 and F (Max Lutolf, Die mehrstimmingen Ordinarium Missae-Satze [Bern: Komm. Paul Haupt, 1970], I, 213; his book is cited in the bibliography).
Roesner follows convention by basing his interpretation of rhythm (pp. lxxxv-xc) on the system of the rhythmic modes as transmitted by Johannes de Garlandia and his followers. This reliance is not without risks, since we have no earlier description of the modal system, and the appearance of the second mode, which must have caused its establishment, has not been documented before ca. 1210. As Roesner points out (p. lxiii), most of the repertory was surely composed earlier. The rhythm of much of the music in the repertory under discussion must, therefore, be seen as still protomensural and, more specifically, premodal. For the often problematic reading of its notation the only didactic guide is the highly elliptic old part of the Discantus positio vulgaris. Roesner's frequent choice of "upbeat first mode" (as against second mode) is evidence of his full awareness of premodal notational configurations, which, however, should not be seen as disruptions or alterations of modal patterns (p. lxxxv); the latter often turn out to be an anachronistic and misleading conceptual framework. To give another example of this sort of contextual Gestaltnotation, there are chains of binary ligatures, set off from one another by strokes; each unit signifies two longs and a breve rest (as in Roesner's transcription of No. 1, 205-8 and 243-50, but not of 71-74, triplum, where nonetheless the same meaning seems indicated). Another more frequent notational device, that is, the addition of several ternary ligatures to a single long, should not necessarily result in an automatic third-mode transcription (whether the regular or the original "alternate" form), but should often be seen as predecessor of the alternative sixth-mode notation described by Garlandia (see this writer's "The Earliest Phases of Measured Polyphony," in Music Theory and the Explotation of the Past, ed. Christopher Hatch and David Bernstein [Chicago: University of Chicago Press, 1993], 43-50). "Roesner's perception of this situation is evident in his sensible and sensitive discussion of the rendition of rhythm in performance "remarkable for its lack of rigidity (p. xcvii); in the recognition of alternative possibilities throughout the critical commentary; and in some of his transcriptions. Evidently, however, there are other applicable passages. In general, it often seems that the ambigous some of his transcriptions with faster rhythmic motion and short rest, where context and notation suggest that option, as for instance in No. 4, 21-50 (cf. Gloria, 54-55), 65ff., and a few passages in the verse and in the Gloria Patri; No. 5, 32-84, and others.
Roesner omits from his transcriptions the long strokes coincident in all voices of the two quadrupla and accounts for the (in most cases) only in the critical commentary as fines punctorum, as Anonymus IV called them, a term that appears to be in conflict with their uncommon use in these works Their retention as barlines going through the system seems essential because of their potential significance for performance and analysis. Discussion of a few other subtopics, such as plication, must be reserved for another forum.
The one problematic aspect of the edition concerns the many cases in which a change (or repetition) of a pitch and a change of syllable in the tenor occur in conjunction with a phrase ending in the upper voices, mostly written as a binary. ligature. The latter and the note in the tenor are preceded in the sources (and the edition) by a small stroke (tractus), indicating the syllable change) it is indistinguishable from tile stroke marking the end of a phrase, separated as a rule from the next phrase by a rest. In relation to the new note in the tenor the first and last notes of such concluding ligatures in the counterpoint are respectively dissonant and consonant. Like many of his predecessors Roesner takes for granted that a syllable stroke is followed by each singer's intonation of the subsequent note simultaneously with the others'; he views the resulting "strong cadential dissonances," a feature that is unreported by medieval commentators, as "basic to the Parisian style,, (p. lxxxv).
Forty years ago Manfred Bukofzer in a trenchant and generally disregarded review had called such dissonant cadences in conjunction with a syllable change an "optical illusion" (Notes 12 : 236). It had caused William Waite in his transcription of the W1 version of the Leoninian organa (The Rhythm of Twelfth-Century Polyphony [New Haven: Yale University Pres, 1954]) to render such ligatures as second-mode dissonant cadential interruptions of the premodal or first-mode flow of the duplum. Roesner, heightening the dissonant effect, makes both ligated notes ternary, longs, perhaps induced to do so by the thickening of the first note of some of these ligatures. This device seems to have originated in organum per se (organum duplum), particulary as an occasional graphic indication of the lengthening of the penultimate note of a phrase ending with a perfect consonance - one of the three categories of lengthening Johannes de Garlandia reported for such organa. In the post-Leoninian repertory it must often be understood similarly as a mensurally insignificant fermata or ritardando before the consonant cadential coincidence, where, as Anonymus IV informs us, the stroke concerns exclusively the change of syllables, but has no effect otherwise (nullum tempus significant). (Roesner himself asserts that "vertical strokes that are used for a purpose other than to indicate a rest or a breath mark do not have significance for sounding articulation" [p. xcix].)
Several notational and transcriptional factors reinforce the view of the syllable stroke as a mensural phantom: (1) the absence in the sources and in this edition of dissonant cadences not involving syllable changes; (2) the beginning of the verse of No. 9 (and its partial concordance No. 14), in which Roesner correctly coordinates the tenor note with the second (consonant) note of the unligated beginning in the upper voices - a singular remnant, not particularly suitable to organa for more than two voices, of the type of ligated beginning common in organa dupla; (3) the corrosion of phrase parallelism by the attribution of mensural effect to the syllable stroke. In addition to such cases in Waite's transcription of W1 of which I cited one as example 6 in "Sine littera and cum littera in Medieval Polyphony," in Music and Civilization: Essays in Honor of Paul Henry Lang, ed. Edmond Strainchamps, Maria R. Maniates, Christopher Hatch knew York: Norton 1984), 220, there are several such instances in this edition, for example, the triplum of the Gloria setting of No. 27 the three phrases after 40), the respond of No. 27 (the two phrases after 22), the Gloria setting of No. 36 (last two phrases); (4) the many noncadential (midphrase) syllable changes involving ligatures, each preceded by a tractus. As the continuity of the counterpoint precludes the shifted lengthening applied by the editor to cadential situations, some of these syllabications are left unsynchronized in the edition (e.g., No. 2, V. 550; No. 18, 32-35; No. 20, 2-3; No. 34, 4-5), while in the majority of such transcriptions Roesner moves syllabic articulation to the last note of the ligature (without comment); (5) the fair number of consonant syllabic cadences involving not ligatures, but either single notes or plications, so that it is the last note, standing by itself, that is preceded by the tractus; (6) several concordant cases in which one scribe preceded a ligature of more than two notes with a syllable stroke, while another separated the last note from that ligature and transferred the tractus accordingly.
All these items, taken together with the testimony of Anonymus IV, argue against the apparent fallacy of the dissonant cadence. Roesner's apt observation that "modern notions of consistency had little relevance in the writing down of accidentals in the sources of this period" (p. xc) is applicable to the issue of syllabication. Like b mi and b fa, often placed rather hap-hazardly like beacons, the syllable strokes seem to have been one of the "accidental" elements of the basically melismatic notation, inserted as closely as convenient and to be worked out by the signaled performer. (These conceptual and notational habits were traditional since the twelfth century, ultimately going back to the more demanding needs for coordination of the separately notated parts in the Winchester Troper and the two earliest Aquitanian sources.)
Of the two alternatives - preceding the cadential syllable change in the tenor with that in the upper voices or making them coincide by disregarding, that is, breaking, the ligatures in performance for the sake of simultaneous pronunciation - the latter is clearly the one whose adoption is indicated; it is also consistent with Roesner's own predominant way of transcribing noncadential syllable changes (see the fourth item in the preceding list) and with his exceptional transcription of at least one cadence (No. 13, 211). The latter must be recognized by performers and scholars as the chief model for cadential syllables, changes.
This approach is confirmed, lastly, by the meticulous treatment of syllabication in the reformed notation of the In substitutes in F (ff. 178-183'), which, because of the often quite rapid succession of syllables, conscientiously avoids ligatures inconsistent with proper declamation. Far from indicating a sudden radical change in declamatory practice, it thus removes remnant conventions of the traditional melismatic notation of polyphony from the sort of syllabic context at issue and so is related to motet notation (Sanders, op. cit., These snippets are evidence of the process of abbreviation of Leoninus's organa reported by Anonymus IV, which tightens for shrinks, to avoid connotations of modern psychology-see Musicology Australia 15 , 14) the Leoninian originals by substituting extremely concise discant passages; they have no other purpose. (To the best of my knowledge abbreviate does not mean "to make a redaction" in medieval Latin; apart from "to abbreviate, to abridge" it can only mean "to make an abstract, to give a brief written account.")
Finally, the inevitable unsystematic list of some misprints. P. lxviii, n. 17, 1. 3: 34, not 35; p. lxxxiv a: 13, not 1; No. 1, 8-10, duplum: incorrect ligation; e, not d. for the third note; Verse 4 triplum: beam missing; No. 4, Gloria Patri 102, triplum:, not c; No. 5, Verse 44, triplum: d, not e; No. 27, 77, triplum: second note d, not c; No. 32, Verse 197: tenor note one beat too late; hence, the duplum and triplum phrases end with a quarter-note and an eighth-note rest, thus becoming four-beat phrases, like the others surrounding them; No.35, 94: labiis, not libiis; No. 36, Verse 265: no plica; No. 38, 127, duplum, d, not c; p. 314 a, Notes, 1. 6: e, not c.
No edition of music in pre-Franconian notation will be without its problems. Both the praiseful and the critical comments in this review are forms of respect for the editor and of admiration for the repertory in the service of which he expended prodigious time, labor, and expertise. I am delighted to call attention to Roesner's acknowledgment of his particular gratitude to Margarita Hanson (p. lxii), whose name appears on none of the title pages, but without whose will, skill, and dedication to an inherited cause the fine series of Oiseau-Lyre editions published during the past twenty-odd years might well not exist.