As I sing, I am reading these adiastemmatic “cursive” notations in the Graduale triplex, as copied from L and C by hand by the apparently indefatigable Marie-Claire Billecocq and Rupert Fischer, who—as far as I can see—make no mistake relative to the facsimiles in Paléographie musicale of L and C (E has a facsimile in its own publication). In the Graduale triplex, the notations of L and C (or E) are placed respectively above and below the four-line staff; on the staff is placed the version in square notation derived directly from the Vatican Graduale, with, however, the rhythmic marks of Dom Mocquereau. All of this makes it easy to compare several text-states of the Gregorian Mass Propers.
It has been the experience of scholars and singers using the Graduale triplex that there are numerous slight differences of pitch between the early manuscripts L, C, E transcribed in Graduale triplex and the version printed in square notation on the staff (the Vatican version); and also between the earliest versions themselves.
The traditional summary description of the staffless adiastemmatic notation of L, C, and E as well as other early chant books is that this notation does not indicate pitch. More precise descriptions, however, have always pointed out that what the adiastemmatic notation does not have is exact sizes of intervals; it does in fact show the direction of changes in pitch, for instance in scandicus (three pitches ascemding) and climacus.(three pitches descending). In those cases the meaning of the signs (which is confirmed by collation) is visually evident; in other cases the direction is indicated by convention. There are cases where a visual clue is apparently not confirmed by collation, and these may be problematic.
The changes of pitch that are indicated (and in which the meaning of the sign in its general use is confirmed) occasionally disagrees with the version in square notation, and this provides a real variant. A simple example of such variants occurs when L or C (often both at once) provides a group of two pitches in a pes, such as FG, for a syllable that Vatican gives as a single pitch, F.
It is this type of variant, numerous in the repertory as a whole, that is usually acceptable in each version. A new edition of the Graduale would presumably take account of such differences, by way of bringing the edition into accord with tenth-century manuscripts in preference to the eleventh- and twelfth-century diastematic manuscripts used for the Vatican edition. (A start has been made in this direction in the Beiträge zur Gregorianik 21, 1996). There are plenty of indications that such variants, when observed over a broad base of manuscripts, may show melodic shifts over time, comparable to the sound-shifts that mark the early medieval development of some European languages. The MI/FA shift has long been observed. Perhaps related to that shift is a difference between so-called Roman and Germanic “dialect” also involving a MI/FA option, although I believe that a chronological study of this difference, observing more closely the behavior of the early staffless notations, will show it to be not synchronic and regional but rather diachronic in nature.
Another common context involves two consecutive descending clivis, for instance db ca. When Vatican Graduale reads these as dcca, presumably in a source from after 1100, it may reflect an elevation of the b to c, and that would be clear in staff notation; but it is not clear in staffless notation, and in this case the notation of L or C often shows no such equivalence in pitch between the end of the first clivis and the beginning of the second, either by relative placement (which in staffless notation would not by itself be convincing), or by using an oriscus or the figure called pressus, which is readily available and in fact used as an alternative in some cases. I draw the conclusion that if L or C does not notate a repeated pitch, such was not desired, and that the reading should be db ca.
A similar case involves a porrectus, for instance dbc. Sometimes when Vatican Graduale prints dcc (as a “clivis strophica”), L or C has a porrectus which can be read as dbc, and the difference can be ascribed to an elevation, after 1100, of b to c. In general such elevation of MI to FA can be understood as a melodic anticipation—sounding the c expressively before actually arriving at it. I find the anticipation to be slightly rhetorical in effect, and I believe eliminating it in favor of the MI (b) gives the line a much more plastic, lyrical effect.
In yet another context for the MI/FA, FA/FA variant, C often writes a trigon b[?]ca for L cca. On the face of it C is singing MI FA RE. But the sign called trigon involves an extremely intricate argument that the trigon is and always was a graphic convention for FA FA RE, in spite of appearance; so read, L and C are the same. The argument about the trigon, however, seems not conclusive; hence, graphic appearances and also context suggest the reading MI FA RE, bca, for C, hence different from L.
I conclude that the singer has an option, subject in large part to his or her musical judgment at the moment of performance. I have exercised this option, while continuing to pay close attention to context and scribal preference.
With this kind of conclusion in view, I believe it to be tedious as well as fruitless to discuss in print the numerous instances of pitch differences that turn up in comparing L, C, and Vatican. Anyone can tell by observation what is being sung, and few will be persuaded—or should be persuaded—by whatever reasons I might provide. How it sounds is the only test. Still, I myself find the options of notation endlessly fascinating, and I recommend the study of them as a profitable subject.
In the present performance I modify the pitch configurations of the square notation of the Graduale triplex whenever it seems to be called for by the staffless notation of L and C. I do not thereby claim—or am even concerned with—a systematic reading intended to establish a more correct version; I want merely to follow as closely as I can whatever L or C had in mind. Since the variants are slight, they do not make much difference in the overall conduct of the melody, and for the listener not following the notation they may pass completely unnoticed.