Repertory and Sources
Gregorian chant comes to us in a repertory marked by a degree of closure, consistency, and homogeneity unique in European music. At the same time, it comes completely lacking a cognate theoretical system, or any documentation except the chant books themselves; neither does it have a verifiable historical development; and little specific data concerning its composition or performance is preserved from the time of its appearance. Its only historical context is a liturgical one, and the sources for this context present as many problems in understanding as the chant books themselves. While the liturgical context has taught me much, I have not found in it what I wanted to know about the music. I find that the best source of information about the music is in the music itself—in the way it is written down, in the way it sounds when performed.
The Gregorian repertory can be defined by inclusion in one of the comprehensive chant books whose physical preparation can be dated to the ninth century; the provenance of these is generally given as north of the Alps. Hesbert’s Antiphonale missarum sextuplex gives the six best known and most comprehensive chant books in a six-column parallel edition, and this forms the basis for definition of the repertory, and for its further exploration.
The Sextuplex presents the data of its sources in their inconsistent and often confusing reality. Modern scholars can and do draw differing conclusions from the data, and intense discussion will continue on the many questions concerning what historical reality is represented by the Sextuplex.
One conclusion has been to imagine that there was an archetype in the form of a manuscript book, no longer extant, from which all extant books containing words with or without melodies were ultimately derived.
An alternate conclusion is that the archetype took the form of a consensus among liturgists north of the Alps during the period ad 750–850, as to the liturgical calendar and the assignment of Proper words and melodies to it.
Still another alternative, taking account of the lack of documentation in support of either of the two mentioned, imagines that the agreement among the extant chant books represents the highly integrated practice of only a small group of liturgists and singers.
Clearcut choices among these possibilities are rendered increasingly difficult by a series of second-order options: each of the three possibilities might be imagined to take account of words as well as music, or music only, or words only; or the options might be imagined to apply to various time tables involving decades or centuries before ad 800, and various degrees of preservation of the repertory; finally, the options might be imagined to apply to a written or an unwritten maintenance of the repertory.
There is no agreement, then, among modern scholars as to what happened in the eighth and ninth centuries, and it seems to me that none can ever be reached. So, while the Sextuplex may indeed represent the only basis for an eventual agreement, I do not look to it for one; rather, I use the Sextuplex for a much more limited purpose—to supply a brief past to the earliest chant books with musical notation. These books appear around or soon after ad 900, at the end of the century in which had appeared the books without melodies that form the Sextuplex. Two of the earliest and most important notated books are Laon MS 239 (“L”) and St. Gall Cod. 359 (“C”). Both are included in the Graduale triplex. These have to be supplemented by other similar books in specific cases, since, although closely correlated in liturgical calendar and specific repertory with the Sextuplex in general, they are not exactly correlated with any one of its six books.
The most direct and simple application that can be made of the Sextuplex to L or C is when the words of a chant in L (or C) appear in the Sextuplex; this shows that the item of chant had been in the repertory for some time, say, at least fifty years, before L (or C) was prepared.
That much seems to be a simple statement of fact; no such statement, however, can be made for the melody in L (or C), since no melody accompanies the words in Sextuplex. That the melody in L (or C) was in the repertory can be assumed; but in the absence of documentation it is only one possibility, and furthermore the complex relationships between L, C, and the Sextuplex raise a number of other possibilities. Some of these are listed here; the first case is as just described.
Second case: L or C has a melody whose words are not in the Sextuplex; here the possibilities are that the item (words and melody) was not in the ninth-century repertory, or alternatively that the melody in L or C was in the ninth-century repertory but sung with a different set of words in the Sextuplex (or in some other ninth-century source).
Third case: L or C lacks a set of words that appear in the Sextuplex; these words may turn up in another (later) chant book, with a melody, which then may be a ninth-century melody.
Fourth case: L or C (or other later MSS) may differ among themselves as to the melodies they provide for words from Sextuplex.
Fifth case: sets of words absent from L or C may appear in other ninth-century sources.
There may be other cases, but these are sufficient to show the difficulties of establishing the repertory of melodies, in spite of the relative fixity of the assignment of words to the calendar.
As a point of method in reading melodies with words from L or C, I include in the Archive only melodies whose words are documented for the ninth century in the Sextuplex, and I assume them to be ninth-century melodies. This presents a repertory of melodies known to be from the ninth century (or earlier); other melodies also may be from the ninth century, but that would have to be established in each case by argument; and for that the repertory of the Archive can be used as a basis for stylistic comparison.
Having allowed for the complexities and uncertainties, I need to add that these concern only a small minority of the more than 500 chants in the repertory; the great majority show a degree of concordance among all the sources that has been universally taken to show that almost all the melodies in L and C were used in the ninth century with their words as these appear in Sextuplex. This defines the ninth-century repertory of Gregorian chant.