About the Antiphoner
This manuscript is one of the earliest complete antiphoners, athough the majority of the chants are abbreviated. From the Basilica of the Holy Savior in the Benedictine monastery of Saint-Martial, Limoges, it was acquired on September 5, 1730, by the Bibliothèque du Roy, along with 203 other manuscripts that were preserved at St. Martial. The date of the manuscript is the subject of much debate by scholars. Some features of the manuscript suggest a relatively early date. Spellings (ae) and letter-forms (the loop over the e; the open a) often reflect an early usage, out-dated even in the late tenth century. In the original layer of the manuscript, the Office structure itself had not yet reached the final medieval form. In addition, the first translation of St. Martial is mentioned (fol. 91v), but not the second, which took place in 994.
Other features point to a relatively late date. The gathering formed by fols. 105-110 is a palimpsest, with the antiphoner written over a notated processional. The notation of the processional is diastematic, though still lacking a dry-point line. Thus, the state of Aquitanian notation by the time Paris 1085 was copied was not as primitive as the unheighted neumes (probably due to lack of space) in the antiphoner suggest. Erasures in the office of Martial (fols. 77r-78r) show that the manuscript was still in use in 1029 when Martial was elevated from a confessor to an apostle, but there is no sign of the new office written for him at this time to replace the older, non-apostolic one. Furthermore, there is no mention of the dedication of the new basilica at the monastery the year before, in 1028. Thus the manuscript seems to have dropped out of use about this time. The numerous additions in the text and the margins by many different hands indicate a manuscript that was in use for a long period of time. This suggests a date for the original layer of several decades before 1029, to allow the manuscript to change hands at least once during its working life. An original date in the 980s seems reasonable.
The manuscript shows every sign of being copied by a cantor for his own use. Only the responsory verses are regularly copied out in full, while the majority of the antiphons and responsories are abbreviated. One feature of the manuscript is the presense of multiple responsory verses, usually two, for each great responsory. The responsory verses also receive the majority of the musical notation. At every point, it is clear that the scribe was intimately familiar with the material he was copying. Not only are there very few mistakes, but the text is highly abbreviated, requiring a reader who already knows the repertory well. Some of the musical notation is probably part of the original layer, but much of it has been copied later, along with several additions to the text in various hands. Most notable among these later additions are the Roman numerals indicating the mode of the chants.
The text of the manuscript is small and often very difficult to read. Furthermore, the manuscript has been bound too tightly so that the inner margins are not visible in the microfilm. The index was checked against the manuscript itself through a Summer Stipend from the National Endowment for the Humanities. For their help and permission to work from the manuscript, the librarians at the Bibiothèque nationale de France are also gratefully acknowledged.
About the Index
Nearly all of the chants except for the responsory verses are abbreviated. This leads to problems in identification of the correct CAO number. In many cases, these problems can be resolved for responsories by reference to the repetendum cue, or for antiphons because the scribe often includes later words in the chant. (The scribe clearly knew his material and on several occasions takes the incipit just far enough to distinguish it from another similar chant.) In other cases, the reference to liturgical position and to other similar sources may be used to determine the most likely CAO number.
Rubrics are sparse, but the layout of the text is clear and consistent, so that in the main text it is usually clear what genre a chant is. However, marginal additions seldom have any kind of rubric, and their genre must be guessed from context. Many of these marginal additions are also abbreviated to almost nothing. Identifying these chants was a major problem, and many of them remain a mystery.
Chants written in the margins have been put in their proper liturgical place in the index as much as possible: cues to the left of the location number indicate their actual position on the page:
L left margin
R right margin
T top margin
B bottom margin
Other letters are used to provide further information:
A a later addition
X a chant that has been erased
In the manuscript, the mode for most of the chants is indicated by Roman numerals, including those chants without notation. Mode has therefore been indicated in the index as follows:
* no notation and no Roman numeral indicating mode
3 notated, mode 3
3* no notation, but Roman numeral indicates mode 3
? notated, but mode unknown
r the notation indicates a formula for a short responsory
Because the Roman numerals and the musical notation are rarely part of the same layer, sometimes they seem to clash. For this reason, the mode of notated responsory verses is taken from the responsory verse formula, and not the Roman numeral mode indicated for the respond. Thus, some responsories have verses in a mode different from that given for the respond. A few responsories have verses in two different modes, each indicated by the notated responsory verse formula.
“*” appears in the differentia column for the majority of the antiphons. 215 items have some kind of differentia, either a notated differentia or a notated antiphon verse. These differentiae are written in non-diastematic neumes, and usually consist of the final two syllables of the formula (“amen”) only. This makes the identification of the differentiae extremely problematic. In the index, each differentia formula has been assigned a letter followed by a number, which represent a particular series of neumes. In most cases, this is followed in the extra field by a possible mode and differentia name, which corresponds to the differentia names used in the indices of Toledo 44.1, Toledo 44.2, and Paris Lat. 1090. A complete table of these differentia formulas can be found in Collamore, “Aquitanian Collections of Office Chants: A Comparative Survey.” These differentia assignments should be understood as conjectural.
In a few cases there is some uncertainty about the Roman numeral designation of a chant. In these instances, the mode field is marked "?*" and the conjectured mode is marked in the second two spaces of the extra field as 1?, 2?, 3? etc.
This manuscript is unusual in that it frequently includes short responsories. While it is often difficult to distinguish between a great and short responsory when the text is abbreviated, in many cases it is possible to make an educated guess as to which category the responsory belongs. In cases where it is reasonably clear that a short responsory is intended, a “B” has been entered into the second last space of the extra field.
Some of the incipits in the index do not make sense in Latin. These incipits represent the letters that appear in the manuscript as closely as possible, but, because of the great difficulty of reading the text at that point due to deterioration of the ink or careless handwriting, these letters could not be put together into a sensible phrase.
Chants not found in CAO are assigned numbers beginning with “ssm” (Sancti Salvatoris Mundi).
- Collamore, Lila. “Aquitanian Collections of Office Chants: A Comparative Survey.” Ph.D. dissertation, The Catholic University of America, 2000.
- Grier, James. “Roger de Chabannes (d. 1024), Cantor of Saint Martial de Limoges,” Early Music History 14 (1995): 53-119.