|
Music historian and Schola Discantus director Kevin Moll is one of the foremost experts in Franco-Flemish Renaissence polyphony in the world today. His five Lyrichord CD's to date have all received high praise from the critics and public alike. In the following essay, moll describes some of the difficulties and rewards of Resurrecting and recording Early Music in the 1990's. From Manuscript to Compact Disc By Kevin N. Moll Reprinted from Fanfare September/October 1997 , Volume: 21 , # 1 The phenomenon of the best-selling "Chant" CD has heightened general awareness of the spiritual and therapeutic effects of early music, and of its potential relevance to modern-day life. Incidentally, I prefer to define early music as European art music cultivated before about 1600, a chronological division running counter to a more popular conception of the term as encompassing music composed through the eighteenth century and even into the nineteenth, which has gained wide currency. The latter definition, however, is inclusive to the point of absurdity, such that works by Hildegard and Haydn can be lumped together under the banner of "early music." Furthermore, there are solid historical and stylistic reasons for regarding the distinction between the prima prattica (vocal polyphony of the Palestrina style) and the seconda prattica (the proto-Baroque practice of monody) as the most definitive "old-new" dichotomy in music history, at least through the early twentieth century. Following this line of thinking, an "original-instrument performance" would not necessarily be equivalent to an "early-music performance." While the medieval and Renaissance eras of music history (ca. 800-1600) are not generally well known to classical-music aficionados, this period has bequeathed to us -- in addition to the plainchant repertoire -- a legacy of hundreds of surviving polyphonic compositions, which historically proceed in a direct line to the universally acknowledged masters of Western music beginning with Bach and Handel. This corpus of polyphony (especially that composed before 1500) exists largely in handwritten manuscripts, which often are wonderfully "illuminated" with decorative miniatures, so that their visual aspect is as stunning as their aural. Performing early music for modern audiences, however, is rarely a matter of reading the music straight off the pages of the original manuscript. On the contrary, to examine the aged parchment folios on which the surviving music is typically transmitted is to see a set of musical hieroglyphs, generally analogous to modern music notation, but strangely unfamiliar to all but a handful of specialists. Due to their foreign appearance, the original entries must be carefully reconstructed into modern format before they can be reliably understood. The resulting "modern edition" is, of course, only a means to an end, and must itself be translated into sound in order to become a meaningful artifact for the vast majority of people. But most early music can be realized only through the medium of highly trained performers, and even then the result always runs the risk of being out of touch with modern scholarship on contemporary performance practices. The foregoing comments illustrate that what the public hears on an early-music CD (or reads in its program notes) can hardly reflect fully the demanding and intricate sequence of tasks that its production probably entailed. Of the total time and effort such a project necessitates, the actual recording phase represents only a relatively small percentage, since significantly larger amounts of time are required in both preproduction (score preparation and rehearsal) and postproduction (digital editing). While I do not propose to discuss here the recondite engineering aspects of production, I would like to acknowledge the recording engineers of our various CDs: Gerald Gold, Alan Mattes, Jay Kadis, and Anthony DiBartolo, and also the digital-editing engineers: Perry R. Cook, and more recently, Jonathan Norton. The sound of the final products is to a great extent the result of their expertise. In the following I shall attempt, based on my experiences as director of the vocal ensemble Schola Discantus, to describe the series of decisions and events that must precede a historically aware recording of period works. I view "historical awareness" as a less grandiose aim in performing early music than "authenticity." The former merely claims to take contemporary performance practices seriously into consideration, whereas the latter can amount to an assertion that one is attempting to approximate period music "as it really sounded." Since our performances hitherto have been confined exclusively to Franco-Flemish sacred music written between about 1300 and 1500, I will focus on the problems of preparing that repertoire for performance. Given that music from the pre-Baroque period lies almost completely outside the realm of most people's listening experience, it is imperative, I think, for the performer to be in a position to choose pieces that most deserve to be presented to the public. But how does one make such judgments? Being well-versed in the discography of early music is only the barest starting point for selecting works knowledgeably. It is much more crucial to know the range of works that haven't been recorded, and to develop a capacity to judge which are most significant. This demands that one be conversant with the boundaries of the early-music repertoire, which in turn presupposes an intimate familiarity with the original sources (manuscripts or early printed editions) compiled during the period or shortly thereafter, and no less with the critical editions of music that have been made in modern times. Once one is sufficiently oriented in a general sense, certain standard reference sources (such as the Répertoire Internationale des Sources Musicales, or RISM, catalogs) can guide further exploration in any given direction by providing an overview of groups of sources and their contents. In choosing which works are to be brought to life, two factors would seem to be paramount: historical importance and aesthetic value. On the one hand, certain groups of pieces are sadly underrepresented in the music-historical literature and in the discography. Obviously, such situations need to be remedied. An example of our attempt to address this historical concern is our French Sacred Music of the 14th Century disc, recorded in 1994; this compilation represents the first comprehensive sampling ever released on CD of the important corpus of Mass music cultivated in royal and papal circles during the period -- a repertoire exemplifying a crucial stylistic watershed in the history of medieval polyphony. On the other hand are those works which, while possibly not remarkable in terms of their historical significance, demand a hearing by virtue of their pure beauty (admittedly a subjective quality, but no less powerful for being imperfectly definable). An example of this second, aesthetic concern is our recording of La Rue's Missa De Sancta Anna -- a cycle I have long regarded as one of the most sublime polyphonic compositions of all time, rivaling the best efforts of Palestrina, Victoria, Lassus, or anyone else. The most compelling candidates for performance, naturally, would be those pieces that fulfill both criteria. As might be expected, such works are not common, but from among them I would cite the Missa Se la face ay pale of Guillaume Dufay (d. 1474) and the Missa Prolationum of Johannes Ockeghem (d. 1497). We have not recorded either of these, however, since multiple modern recordings of both are readily available. Instead, our goal heretofore has been to expand the discography, concentrating especially on works of demonstrable historical significance, another example of which is the Marian Mass of Reginaldus Liebert -- a scintillating cycle that otherwise would surely have languished for many more years awaiting a recorded performance. I mentioned above that a systematic study of sources is indispensable for an understanding of the repertorial problems of early music, but such issues still do not bring us very close to the challenges of actually performing the music. For to perform period pieces adequately -- or even to discuss them competently -- requires in addition that one be conversant with a series of specifically musical issues, including the theoretical and notational principles of both chant and polyphony. As is the case with the musical sources, such knowledge can only be attained through dedicated musicological study. One of the signal accomplishments of music scholarship in the last hundred years has been the transcription and publication of a large body of pre-Baroque music. (Parts of the following discussion are adapted from my program notes to the CDs entitled Johannes Ockeghem: The Two Three-Voice Masses (LEMS 8010) and French Sacred Music of the 14th Century (LEMS 8012). I would like to thank Nick Fritsch, the president of Lyrichord Discs, for granting permission to include the material here.) A few early-music ensembles, it is true, have made a point of reading off the original sources, sometimes even incorporating improvisation, but such attempts are still the exception. Most performers prefer to read from published transcriptions in modern notation, in which the rhythmic note values have been reduced, and the original movable C and F clefs have been replaced by the standard treble and bass clefs. But paradoxical as it might seem, the existence of a critical edition of a given work is far from being a guarantee that all performance questions have been answered. On the contrary, many modern editions neglect to provide authoritative information on crucial issues, and older editions sometimes advocate performance practices that have since been discredited. Informed performance decisions, therefore, presuppose a familiarity with the modern literature on performance practice and with contemporary music theory. In the event, the original manuscripts (whether in published facsimile or on microfilm) must frequently be consulted to verify individual readings or editorial policy. The value of a solid music-historical training is not that it will necessarily lead to a definitive answer to a given problem, but that it provides the tools to find a workable and defensible solution. Once attained, this knowledge should not, however, preclude one's reserving the right to make decisions on the basis of perceived musical necessity or individual taste. The first question that must invariably be asked in a performance of late medieval polyphony is: to what performing media do we assign the various parts? In the relatively recent past it had been standard practice to allot certain lines to instruments, either alone or doubling the voices. During the 1960s the New York Pro Musica recorded a Mass by the fifteenth-century composer Jacob Obrecht utilizing a veritable symphony orchestra of period instruments, as well as soloists and a large choir of over twenty singers. In the last twenty years scholarship has questioned the historical validity of such a conception of performing forces (just as it has revised the tradition for using huge choruses in Bach's B-Minor Mass or in Handel's Messiah). Instead, it seems clear from payroll and other archival records that sacred polyphony during the period was normally performed a cappella. The quantity of singers on a part is also a question, but it seems likely that in the major court chapels of the late medieval era, solo or choral performance involving a small number of singers was common practice, often with more singers assigned to the upper parts than the lower. Our performances of liturgical polyphony have proceeded from that principle. In any performance of music incorporating words, the performer can scarcely afford to ignore their delivery and their relationship to the notes that carry them. This is particularly true in early music, where the words are often fitted to the music in an ambiguous or sometimes even an illogical manner. Therefore, text underlay often becomes one of the most daunting aspects of preparing a given work for performance. In the fourteenth century and well into the fifteenth, only those voice parts having a share of the upper melodic profile were typically provided with words. For a long time scholars had assumed that the untexted lower parts were intended for instrumental perform-ance, but this presumption has been severely challenged in recent years, as is noted above. The later the repertoire, the more likely it is that a given work will have all voices texted in the source. Yet even as early as the fourteenth century, certain pieces are texted in all voices in the manuscript. This circumstance provides concrete evidence that supporting voices had at least the potential to be performed vocally. The problem thus lies in determining how to realize a vocal performance of these untexted lines. Often there are simply not enough notes in a given part to fit the number of syllables in the text. In our recordings we have adopted a variety of solutions to this situation: sometimes we have added as many words of the liturgical text as would fit the number of notes, taking care to follow the text declamation in the texted voice(s) as closely as possible. In movements with short texts, such as Kyries, this procedure often works perfectly. Alternatively, untexted parts sometimes are simply vocalized on a neutral syllable (such as "ah") or on various syllables in succession. In working with the French Mass repertoire of the fourteenth century, I came to the conclusion that the grammatical articulation of the text closely parallels musical structure as defined by melodic and contrapuntal cadences, and I have seen no reason to doubt that the principle applies in later music as well. This provides a key to determining text underlay in many troublesome cases. In recent years an increasing number of early-music ensembles have been experimenting with period pronunciation of texts. Thus, in our French Sacred Music CD, we attempted to render the Latin texts not according to the Italianate pronunciation customarily used in modern choral singing, but according to certain principles of fourteenth-century French pronunciation. As with so many other aspects of historical performance, however, this decision only opened up a new area of seemingly limitless research possibilities, leading to the question of whether it is possible, or even desirable, to strive for an "authentic" reconstruction of vernacular pronunciation in a particular time and place. In any case, the majority of listeners will be more or less unaffected by pronunciation, and when one considers the valuable rehearsal time involved in retooling the singers, it might be argued that this time would be better spent working on ensemble balance, tuning, rhythmic coordination, and other musical problems. Determining the performing pitch of works is a critical factor of performance, for which the actual vocal ranges of the individual performers must be taken into account. Normally we have performed pieces at their written pitch, based on the modern international pitch standard where A above middle C equals 440 Hz. A number of modern studies have demonstrated that there can have been no pitch standard in early music; yet often a given group of singers performing at A=440 is perfectly suited to the tessitura of voice parts in works from the period, such that transposition either way would actually put the singers into less comfortable ranges. We have, however, found it necessary to make several exceptions. For example, in our recording of Ockeghem's Missa Quinti toni, the vocal registers of the available singers rendered it expedient to transpose the work up a minor second; the Mass is written in F, so that the resulting sounding pitch is FJ. Determining the pitch of the companion piece on that CD, the three-voice Missa Sine nomine, was considerably more problematic, since in its notated form the tessituras of every voice part lie much higher than is typical for the period. Given this fact, we did not hesitate to transpose the entire piece down fully a fifth, which puts the Mass into much more characteristic voice ranges. In our French Sacred Music recording, every selection was transposed down a major second (so that a tenor part ending on written G would sound a modern F). The main purpose of this transposition was to stabilize somewhat the sonority of certain works in which all parts have a relatively high ambitus as notated, but it also had the effect of placing less strain on the tenors and countertenors in their upper registers. In order to maintain the relative tonalities of the various movements on that disc, all pieces had to be transposed equally -- not just the problematic ones. No aspect of performance affects the hearer with more immediacy than tempo. This, indeed, is an issue of great complexity and it continues to be the subject of musicological controversy. Generally speaking, tempos revolve around a basic pulse, or tactus, denoted by the time signatures O and C, which probably represent some 60-72 beats per minute in triple and duple mensurations, respectively. These basic time signatures each have an equivalent "diminished" or "proportional" signature (Ø and cents), which most often is interpreted as representing a faster tempo. When occurring in successive sections of music it is questionable what the relationship between the two is intended to be, especially in music from earlier periods. This topic has occasioned a great deal of research and debate in recent years. Formerly, most scholars had asserted that the diminished time signatures should be performed precisely twice as fast as the tactus, based on certain instances where both mensurations are used simultaneously in different voice parts in a clear 2:1 ratio (this practice is advocated in Gilbert Reaney's edition of the Liebert Mass, Early Fifteenth Century Music, in the series Corpus Mensurabilis Musicae, vol. 3 (American Institute of Musicology, 1966), pp. 64-94; see, for example, the successive sections of the Kyrie, pp. 66-69). Many modern editions stipulate this proportion, but we have found that it often results in musical absurdity, making it impossible to find a convincing tempo for either section. Our solution has normally been to interpret the diminished time signatures as being faster than the tactus in a proportion of approximately 4:3 -- a course for which there is also some theoretical evidence. However, musical context often seems to dictate a flexible conception of relative speeds, and sometimes even of the tactus itself from movement to movement. Another component of performance that affects the listener strongly, if subliminally, is dynamics, or relative levels of loudness. There are never any dynamic markings in the manuscript sources, and this is one area in which medieval music theory sheds very little light. It should not, however, be thus inferred that the music is to be performed at a constant volume level. On the contrary, many factors contribute to a dynamic understanding of individual movements, and even of entire Mass cycles: such elements include the meaning of the text, liturgical function, local musical texture, and rhythmic context. Indeed, the very architectonic nature of individual movements often builds in its own interpretation of dynamics, as when a composer brings in all voice parts toward the end of a movement and increases their rhythmic activity so as to effect a "drive to the cadence" -- a phenomenon common in the music of Dufay and Ockeghem. The question of musica facta, or application of accidental inflections not present in the manuscripts, is easy to minimize by simply adopting the solutions of a given editor. Most modern editions do, in fact, offer supplemental accidentals, but often the criteria for applying musica ficta are not clearly stated by the editor, so that one cannot be sure that a coherent rationale underlies the suggested sharps and flats. Even when such principles are stated by an editor, they may be open to serious question. It is therefore advisable for the performer himself or herself to be cognizant of the basic rules and conventions of ficta as discussed by the contemporary theorists. And again, anyone preparing works for performance would be remiss if she or he were not well acquainted with the work of authoritative musicologists. My own researches into counterpoint and musical structure have indicated that the most common uses of musica ficta are the following: 1) To achieve the proper contrapuntal motion at cadences and other major points of articulation (usually set in conjunction with the grammatical syntax of the text or with changes of musical texture, mensuration, etc.). This voice leading is most typically expressed as a major sixth progressing to an octave in some two-voice parts. If the first interval occurs diatonically as a minor sixth (say, a lower B to a higher G), then either the upper pitch must be raised (by applying a sharp), or else the lower pitch must be lowered (by applying a flat). 2) To avoid harmonic clashes or awkward melodic leaps, especially tritones (which occur in the white-note scale as F against B), and cross relations (simultaneous or contiguous notes in different voices that manifest differing inflections of the same note; for example, C-s versus C#). 3) To reflect the solmization of a voice part according to the system of six-note scale segments (hexachords) discussed by contemporary music theorists. This criterion is applicable primarily when the melodic nodes of a given passage lie clearly within one or another hexachord. All three of the above criteria are subject to varying interpretations. In the first case, there is often some question of what constitutes a significant articulation in a given musical context. In the second case, the personal taste of any given editor or conductor will inevitably play a large role in determining what kind of melodic contours, harmonic clashes, or cross relations are acceptable. In the third case, an ironclad hexachordal solmization can sometimes be deduced for a given passage, but more frequently a variety of solutions are more or less equally viable, especially when possible accidental inflections are taken into account. Added to these considerations is the even more subjective criterion mentioned by contemporary theorists -- accidentals that are placed causa pulchritudinis (for the sake of beauty). This term may at times have been used by theorists to refer to some obscure technical principle, but in general it appears simply to have sanctioned one's adding accidentals at will as a coloristic effect. This latter interpretation is supported by many apparently capricious accidentals that actually exist in the musical sources -- particularly those stemming from the late-fourteenth and early-fifteenth centuries. The above discussion has illustrated how the means of solving such crucial issues as text underlay, musica ficta, performing pitch, and relative tempos are often left unspecified or imperfectly specified even when modern critical editions of music are used. It is both laudable and efficient for performers to strive for uniformity in solving problems caused by the indeterminacy of period notation; but whenever a conflict arises between adhering to theoretical stipulations or to musical sense, we have tended to let ourselves be guided primarily by the sounding result, not by the abstract formula. In light of the considerable challenges required to bring the music to a performable state, it may fairly be asked why one bothers to go through such exertions. And what, precisely, does early sacred polyphony offer to the typical devotee of "classical music"? My answer would lie in the philosophical, even spiritual, response that the best of this music elicits in most listeners -- a distant but powerful reflection of a flourishing musical culture that took its art quite as seriously as any succeeding age. Never, perhaps, has music gloried more in the pure play of tones or more perfectly accorded with philosophical ideals of music, which indeed are absolute in every sense of the word. Considering the unparalleled magnificence of other fine arts at this time -- one need look no further than the achievements of Leonardo da Vinci (1452-1519) -- is it really credible that such a civilization could have been satisfied with music of anything but the very highest quality? In this, the five-hundredth anniversary year of the death of Johannes Ockeghem, the Franco-Flemish polyphony of the fifteenth century is steadily attaining the universal recognition I feel it is destined to command. Yet only within the last few years has the discography of even such a major composer as Ockeghem become complete (that is, that recordings are available of all the works he is known to have written). A host of other composers, famed in their own time, remain virtually unknown to everyone but musicologists and a fringe group of enthusiasts. For example, something like seven-eighths of Pierre de la Rue's surviving Mass output -- which numbers over thirty -- has never been recorded. Even more astonishingly, a significant amount of fifteenth-century music still awaits transcription and publication in modern editions. Contemplating the artistic status of the various composers in Western music is as though viewing a mountain chain stretching far into the distance. The relative proximity of figures like Stravinsky, Wagner, Chopin, Beethoven, and Mozart renders their monumentality obvious to us. But much farther off is another series of peaks, receding beyond the horizon, and among which are Josquin, Ockeghem, and Dufay. Because they are so far removed, their true magnitude is hard to gauge, but when one troubles to focus one's vision more securely, they begin to appear rather immense. RECORDINGS OF SCHOLA DISCANTUS ON THE LYRICHORD EARLY MUSIC SERIES: |