The last two pieces of the latest St Martial manuscript tend to use rather longer melismas (thus ex.10 has on average 12 notes of melisma to one tenor note, as compared with four in ex.9 and ten in ex.11). It is in completing this development that the most significant achievement of the so-called Notre Dame school seems to lie, and it can be readily observed in the Magnus liber, the most important work of the period. This is a collection of two-voice plainchant settings for liturgical use, arranged for the church year in two cycles containing, respectively, the solo sections of the most prominent responsorial chants of the Office and of the Mass. The sheer consistency with which it was carried through, from all points of view, makes possible a much more precise understanding of its nature. The very selection of chant material itself constitutes a conscious limitation when compared with the diversity of material in the St Martial manuscripts.
The Magnus liber was attributed by Anonymous IV (CoussemakerS, i, 342; ed. Reckow, i, 46) to the optimus organista Leoninus, who is now identified with the poet Leoninus. The original version of the Magnus liber was liturgically designed for Notre Dame, Paris (Husmann, MQ, 1963), and was therefore probably compiled in Paris about 1170 (seeMagnus liber, §2). Stylistically the original form of the work is very difficult to determine because it survives only in versions that date from the 13th century. These versions differ from one another in certain ways, revealing a general tendency for existing organum sections to be replaced by discant sections in a more recent style and hence in modal rhythm. This has been seen as evidence of a historical process whereby organum style – already regarded as outmoded towards the end of the 12th century – was superseded by the more fashionable discant. Even if this thesis be rejected as too linear a view of history there are nonetheless several grounds on which it can be argued. In the first place, it is widely accepted that modal rhythm was not an ‘invention’ at a single point in time, but resulted from several lines of development which were then formulated as a system, apparently after the model of the School of Grammarians in Paris (Flotzinger, AMw, 1972). There is no evidence of the complete system of six rhythmic modes before about 1180. Secondly, the procedures of extensio and fractio modi (seeRhythmic modes) were clearly not components of modal theory to begin with, but were connected with later attempts to subordinate other phenomena, including organum, to the system; indeed, they did much to hasten the obsolescence of the modes towards the end of the period of Notre Dame music. Finally, square notation, the basis of modal notation, which apparently developed in the Ile de France from the northern and central French neumes, did not appear in recognizable form before the last third, if not the very end, of the 12th century.
It is clear that this development resulted from a specific need. Stäblein’s suggestion that this development can be observed in the St Martial manuscripts, and that the beginnings of a modal rhythmic interpretation are already detectable in the later versions of these, does not call for an earlier sequence of dates above – if anything, it calls for the contrary. Thus it ought to be clear that the Magnus liber cannot have been conceived with modal rhythm in all sections; equally well, the sustained-note sections as they survive sometimes exhibit series of ligatures characteristic of modal rhythm, but are at other times far more ambiguous. This stage of development seems to correspond with the threefold classification of polyphony made by Johannes de Garlandia about 1240 as organum, copula and discantus. These three categories were abstracted with two considerations in mind: the relative amount of movement in the two voices, and the rhythm of the upper voice. On this basis, the three categories can be characterized thus: organum as a sustained-note style without modal rhythm, copula as a sustained-note style with modal rhythm, and discantus as a style in which both voices move in modal rhythm (see fig.4).
9. The style of Parisian organum.
The decisive changes which occurred in the early Notre Dame period thus become clear in retrospect. The repertory is represented by the Magnus liber (with due reservation as to the composite nature of the form in which it survives) and the examples in the so-called Vatican Organum Treatise (ed. Zaminer, 1959). This style of organum is the natural outcome of certain tendencies found in the St Martial repertory and the Codex Calixtinus. Discantus developed a new characteristic, that of modal rhythm, and to that extent gained a new lease of life (cf Magnus liber, ex.1). An essential feature of the new developments was the clear increase in the range and scope of the melisma above each tenor note. This could go so far as to necessitate not only holding the tenor note for a corresponding length of time, but frequently also repeating it several times. This was very seldom written out in the manuscripts; occasionally it was indicated by placing a rest stroke alone without a preceding note, and in some cases by drawing several vertical strokes through a specially elongated note shape. Otherwise the reiteration of the tenor note was apparently taken for granted. For the time being the general character of the setting still derived from polyphonic extemporization: the melismas of the upper part seem to enjoy complete freedom of movement, and to constitute an ornamentation or paraphrase of an underlying note-for-note setting of the vox principalis.
However, there are distinct and interesting points of contrast with St Martial practice. First, the notes of the underlying harmony were sometimes clearly regarded as either starting- or end-points, linked by melodic phrases. The phrases may be formulaic or more extended; they may develop with reference to a mode or to a single note; they may be associated particularly with the openings or with the cadences of sections. It also seems that more importance was attached to meaningful progressions in the underlying harmony, fulfilling the requirements of the old discant theory. Whether these are real developments or merely differences of quality or interpretation has yet to be determined. Secondly, there was an increase in the melodic autonomy of the melismas. They seem to delight in unfolding around an underlying melodic framework. This in its turn is frequently directly related to the sustained tenor notes and suggests other underlying constructions within the melisma or clausula.
Another feature of this music that is important for the future is the difference in types of melodic repetition. In earlier times, if they operated at all they preferred to do so with melodic particles that were only similar rather than identical. But then there occurred a sudden increase in the use of identical phrases, repeated, moreover, either at the same or at a different pitch. Melodic movement still included wide leaps, acceptable as a legacy from plainsong tradition; they might be upward or downward leaps, perhaps several in succession, compensated by movement in the opposite direction through the intervening notes, by currentes, etc.
It goes without saying that copula, with the upper voice in modal rhythm above a sustained tenor note, had a completely different melodic structure, dependent on the new rhythm. But this in itself elucidates the close interdependence of melody and rhythm.
The motion of the upper voice, or duplum, was largely restricted to the range above the tenor, which emphasized the tenor’s supporting function even more. Its range expanded somewhat, tending particularly to centre on a higher register which often necessitated a fifth or even a sixth staff line. The relationship of the duplum to the tenor part was nearly always that of a flanking movement. One can speak of contrary and parallel motion only with reference to the underlying harmonic framework. One should note here the succession of identical perfect consonances, which were quite permissible at a time when it was neither obligatory nor customary to disguise them with ornamentation. One should also note the appearance of parallel imperfect consonances, particularly 3rds, in exactly the same way.