Bernardus Claraevallensis (Bernard of Clairvaux)
(Treatise on Singing the Gradual)
Set before the Cistercian Gradual in the manner of a prologue. [HOMMEY Supplementum Patrum, I, p. 27.]
Just as we have forewarned the notators of the Antiphonaries, so also we forewarn those who are about to notate the Graduals, and we beseech and adjure both the one group and the other not to separate notes that are joined, nor to join notes that are separated; so that just as in the singing, so also in the manner of performance, with respect to pauses and divisions, uniformity may be preserved. Let it be understood that those notes alone are called and are finales (finals) which are supported by the full range both of the depositio (descent) of the plagales (plagal modes) and of the elevatio (ascent) of the authenti (authentic modes) — that is, they have four notes below and seven above themselves. The first of these is low D, and the last is high C. Apart from these, and those alone which naturally lie between them, you will find none above or below to which there is not lacking either the fullness of descent or the perfection of ascent.
So that you may be made more certain about the individual cases, know that the first mode, namely the protus, which contains the first and second tones, has two finals: low D and high A. For those chants of this maneria (mode-group) which have a whole tone above the fifth, the necessary final is low D, because above its fifth it naturally has a whole tone. But for those which have above the fifth not a whole tone but a semitone, the proper final is high A, because at the same interval it has not a whole tone but a semitone. For those chants also which descend two whole tones below the final, this is likewise necessary. Having therefore diligently considered this property, you will not wonder that those chants which used to end on D, when notated through B, are naturally placed on high A, nor will this transposition seem to you superfluous or tedious.
The second mode, namely the deuterus, which contains the third and fourth tones, likewise has two finals: low E and B quadratum (B-natural). The first of these is the proper and necessary final of those chants of this mode-group which have a semitone below the sixth. The other properly serves those chants which have below the sixth not a semitone but a whole tone, and also those which descend three whole tones below the final. Note that many authoritative churches — among which I believe are the churches of Amiens and Therouanne — just as they have psalms in their introits, so also have them in their postcommunions. Although you may find few introits among them, you can find many postcommunions. Those which can only end on B-natural can in no way receive the psalm-chant of the third or fourth tone, since the psalms of those tones have a whole tone above the fourth degree, and B-natural has above its fourth degree not a whole tone but a semitone. We also, just like them, have many introits which naturally must end on B-natural, having everywhere a semitone above the fourth degree; but because there they cannot receive the psalm which has, as we said, a whole tone above the fourth, they are therefore transferred to the lower final through B rotundum (B-flat). For this reason you will find the psalm-chants of the third and fourth tones so modified that they may serve both finals.
The third mode, namely the tritus, which contains the fifth and sixth tones, likewise has two finals: low F and high C, for those chants which at that same interval have not a semitone but a whole tone. But because we have no introit which can end on C, you will therefore find the psalm-chant of the fifth tone modified, although it cannot be adapted to both finals.
The fourth mode, that is the tetradus, which contains the seventh and eighth tones, has only one final: G. You will find the psalm-chant of the seventh tone modified only in its melodic pattern, because it was making its pause on a note where it cannot begin, which is not permitted for any chant. For a chant can pause only on those notes on which it is able to begin; and especially on those which most frequently and most fittingly serve as the beginnings of chants.
Furthermore, with the superfluous multiplicity of differentiae (psalm-tone endings) eliminated, each tone retains only one differentia, and that one alone which, ending on the finals, confirms its mode-group either in itself or in its melodic pattern. For this reason, in the eighth tone that differentia has been retained which, bending back down to low D, returns through F to the final G: for by this means it cannot be confused with ending on the final of another mode-group.
Now, while the differentiae of six tones end on their proper finals, the differentiae of only two — that is, the third and fifth tones — did not end on their own finals, and therefore could just as conveniently be assigned to other tones as to their own. These have been brought back to the norm and now end on their proper finals. Besides these matters, there are certain responsories, namely Christus factus est, Ecce sacerdos, Exiit sermo, and similar ones, of which there is an abundant number.
In these, the responsory is of the sixth tone, but the verse is of the fifth; whereas you would far more fittingly attach the verse to its responsory if it were of the same tone, as is the case in the other responsories of the Gradual itself. Moreover, in the responsory which is of the sixth tone, there is the fault of contradiction: because the melodic progression is plagal, while the composition is authentic — which is not permitted. For this reason you will find these responsories quite fittingly reduced to the fifth tone. Although, therefore, emendations of this kind may seem contrary to custom, nevertheless, since nature prevails over custom, let not the change of usage displease you as much as the integrity of nature when properly observed pleases you — so that you who were unwilling to admit dispensation in the rule of living may not grant dispensation for falsehoods or faults in the rules of singing, which ought rather to be cut away than dispensed.