The Lyrics of Richard de Semilli: a Critical Edition and Musical Transcription.
The other manuscripts include the rest of the well-known KNPX group: Paris, Bibliotheque de l'Arsenal, 5198 (K), and Bibliotheque Nationale fr. 845 (N) and fr. 847 (P). The first of these contains the first ten songs and N the first eight, in the same order as X as far as the first eight are concerned and with attributions to Richard. P has the first ten songs but in a different order: Nos. 1, 2 and 8 are with the anonyma of the manuscript, while the remainder are given with attribution in the order 7, 6, 5, 4, 9, 10, 3. The only other manuscript, Bibliotheque Nationale fr. 24406 (V), contains the first eight songs, in the same order as in KNX but without attribution and with different melodies.
The editor observes that 'manuscripts KNX are closely related while manuscripts PV are somewhat less closely related to this group'. This is evidently not meant to imply that PV are themselves closely related: their more distant relationship with KNX is based on different criteria in each case. The most striking feature of V is its preservation of different melodies, despite the fact that it contains exactly the same songs as N. Although they are without attribution, moreover, they are preserved as a group within a section devoted to songs arranged by author, unlike the unattributed songs of P, which are dispersed among other anonyma. P, on the other hand, has the standard melodies, that is, those of KNX.
It is undoubtedly a weakness of this edition--the only serious one, in fact--that the melodies of V are not included. This is the more unfortunate in that these melodies are not otherwise obtainable, except of course to scholars with a microfilm. It is somewhat disingenuous to excuse their omission in these terms (p. 10): 'This manuscript regularly presents melodies that have nothing in common with the rest of the manuscript tradition, and it shows certain peculiarities that are difficult to explain (no repetitions where one would expect them, many hypo- and hypermetric lines, and so on). This is characteristic of the whole manuscript, and the notations of Richard's melodies are good examples of its idiosyncrasy.' It is true that it would have been difficult to accommodate them on the pages facing the text, as the editor has done with the KNPX melodies, without compromising the elegance of the layout; but they could have been added separately, in which case the occasional remarks on their peculiarities would have made better sense to the reader.
Before any further criticisms are essayed it should be made plain that the presentation of the edition, as to both verbal and musical texts, is in most respects exemplary. The bibliography of the songs is handled with the utmost professionalism; the orthography of the poems is helpful and follows the standard practice laid down by Foulet and Speer (though not all editors still accept the need for every aspect of the modernization that they recommend). The melodies are given in the form of stemless rounded note-heads without any rhythmic suggestions, a method that, short of a facsimile of each source individually, is still the most generally useful and the least prejudicial to unhindered study. A very minor criticism is that the treble clef used lacks the necessary figure 8 to indicate that the original notated pitch corresponds to the octave below. In the notes on melodic variants I would question only what are here called 'solfege syllables', in practice the use of Italian (not French) note-names but with 'ti' for 'si'.
The introductory comments on versification, musical structure, and language of the texts are thorough and in conformity with accepted wisdom; and the same goes for the commentaries on the individual songs. But a discussion of the first song, 'L'autrier chevauchoie delez Paris', will serve to show that the concerns of a tidy-minded editor do not necessarily reveal all the considerations that a scholar might like to ponder. For a start this (like No. 4, 'Mult ai chante') is a song with multiple refrains: the melody incorporates only the first refrain, and each one has its own verse form, which raises the question of how the remainder were to be sung. Does the melody of the first refrain show any sign of having originated separately from the body of the stanza? These questions may be difficult or impossible to answer, but here they are scarcely addressed, except for an observation that in neither song are the melodies of the remaining refrains found in any other source, though in several cases (song 1, refrain 2; song 4, all refrains) the words alone are. Are the melodies for the first refrain of each song in KNPVX also unique, and are those in V different from those in KNPX? The answers to these questions, predictable though they may be, are also evidence, even if of a negative sort; but they are not given here.
The relationship between the words and the music of the stanza itself in the first song is also more complex than the description here suggests. It is perhaps unusual, at least in the courtly song, but it brings the problems of the caesur and of elision into sharp focus. Each stanza has eight ten-syllable lines. Normally each line in such a poem will be divided by caesura into 4 + 6 or 6 + 4 syllables, as is in fact the case in song 4. 'But many of the lines in this poem are of the taratantara type [a curiously vague expression] which divides the line into two five-syllable halves (5 + 5).' It is also pointed out that a number of lines 'are one syllable short, having the division 5 + 4'. Although some of these lines can be regularized by substituting hiatus for elision of mute '-e', others can not, and the editor rightly resists emendation. What she has not pointed out is that every line of the type 6 + 4 is in reality 5' + 4: that is, every such first hemistich ends in post-tonic '-e', '-es' or '-ent'. Consequently, lines of the type 5 + 4 (seven out of 40) are essentially similar; it is rather that lines of the type 5 + 5 (sixteen out of 40) have an extra syllable, and it is in the second hemistich that it occurs. (Johnson lists ten lines of the type 5 + 4 by assuming elision at the caesura; three of them, however, are really 5' + 4.) What the poet consistently avoids is the scheme 5' + 5.
It is worth looking at the tune to which this poem is set in order to assess the overall significance of these factors (Ex. 1). The form is AA'AA'BB'BB', followed by CC'C for the link-line and refrain, these last three lines being applicable to the first stanza only. A prime indicates a variant ending (clos) in both A and B; in B it also indicates a variant opening. In fact A and B are identical from the fifth syllable onwards, so that an alternation of the same ouvert and clos persists throughout the stanza. One further subtlety is that the variant openings of phrase B simply reorder the notes of the preceding ending, close or ouvert as the case may be. One could hardly attain a greater degree of unity than this.
[MUSICAL SCORE OMITTED]
What are the consequences of imposing this highly schematic melody on a poem that possesses a sophistication of its own? As it happens, they can be studied very easily, since the variables of versification occur within the constant portion of the musical line, and since all three forms of verbal line occur within the first stanza. The first five lines are of the standard type (6 + 4 = 5' + 4); lines 6 and 8 are 5 + 5; while line 7, which is 5 + 4, has been modified by the editor to 6 + 4 by removing the elision in 'conme il'. This is quite unjustified. First, all three manuscripts with this tune (KNP, since X lacks the tune and the first 28 lines of this song) elide 'conme il' and compensate by allocating e d to 'ne' in ligature. (The procedure in V is not discussed.) Second, the emendation creates a line that is unique in being 6 + 4 and not 5' + 4. It is not musically impossible, but it is quite unnecessary from either the musical or (I would suggest) the verbal point of view.
What emerges from a restored reading of this stanza is that a basic melodic shape could easily be adapted to a variable line of verse. In lines 1--5, which represent the norm, the first hemistich moves towards a tonic accent on g, which is plicated and followed by an e, post-tonic, that is evidently weaker. In lines 6 and 8, where there is no post-tonic syllable at that point, the note e has the rather different function of introducing the second hemistich; the plica would perhaps be elongated in performance, and the e, though still relatively weak, initiates rather than concludes a phrase. In line 7 as restored, the first line behaves as in lines 6 and 8, and e is igated to d to introduce the last four syllables. Evidently this was thought more stylish than a g f e ligature on 'soit', perhaps because that would have negated the plica. The link-line and the refrain-lines, which are longer and have a different melody, emphasize e as representing a tonic accent at the caesura: in line 9 there is no post-tonic, in line 10 it is elided, and in line 11 it is retained, the second hemistich being one syllable short. It should be noted that in NP the music of line 11 is almost identical to that of line 9, and there is no particular reason why it should differ at all.
The editor's discussion of the tonal balance of this tune reveals an approach that appears to rest on anachronistic assumptions about musical behaviour. She writes (pp. 31--32): 'The melody is interesting in that it shifts from one chain of thirds, re--fa--la, to another, do--mi--sol. The musical phrases A and A' begin with a variant of a recitation on la and have re and la as clear structural tones. In B and B' this chain is reduced to a third, re--fa, and the secondary chain, do--mi--sol, becomes more apparent, particularly in the second half of the line. These lines serve as a transition, in the first stanza at least, to the refrain and its introductory verse, where the chain do--mi--sol comes to dominate. The piece ends with do as the final tone.' There is of course a recognition here that this scheme can only be demonstrated for the first stanza: it is not in fact the piece that 'ends with do', though it could have been. But apart from that point, the whole triadic focus seems inappropriate. Certainly theme A has a strong opening fifth, d--a, but f is a weak feature, and the ouvert version already emphasizes g e c e, which in the clos becomes g e c d. In phrase B the opening axis is d--g; when in B' this becomes e c e g it is counterbalanced by a return to the dorientated clos. The focal points of the melody of the stanza, namely, the ends of each hemistich, exhibit an alternation between g--e and g--d throughout. This seems quite satisfying internally, and beside it the refrain, even with its connecting line, seems rather tacked on (as one would expect): the equivalent alternation here is e--c, e--B, and e--c again. Whatever the aesthetic value of multiple refrains, they were not conducive to musical integrity.
The remaining songs are far less problematic. In songs 2 and 7, which in KNPX (but not V) have the same melody, there is an ascending plica that creates a melodic augmented fourth (if interpreted as a b[natural]). This occurs only in K, but the odd thing is that it is used in line 4 in song 2, but in line 2 in song 7. For the other line in each case, and in NPX, a descending plica is used (see Ex. 2, where other variants are also noted). The editor may well be right in choosing the 'harder readings' of K; and as there is not much aesthetic reason for preferring one of them over the other the melodies of songs 2 and 7 may as well be left as they stand in K. But it is unlikely, in fact, that both are correct, if the concept of correctness has any validity at all in this repertory; and a brief discussion of the issue would have been welcome.
[MUSICAL SCORE OMITTED]
Song 4, as we have seen, like song 1 has multiple refrains, and poses the same questions as to their musical setting beyond the first. It is not quite correct to say that 'the melody has the widest range of Richard's songs', since that of the preceding song, 'Par amors ferai chancon', has exactly the same range, a minor tenth. Both of these songs illustrate the editor's procedure as regards the notation of b fa, which is simply to record it as it occurs in the base manuscript--K in each case--and to discuss variants in the notes. This is a sound method, but she offers no considered view as to the interpretation of the passages concerned, nor any supplementation of the indications in the sources. In songs 5 and 6 (which share the same melody in KNPX) it is clear enough that the flat signature applies throughout.
The editor has sensibly chosen MS K, despite some imperfections, as base for the first eight songs. For song 9, which is in KP only, she has taken P as base for the text, since it is more complete, and K for the tune, since P has a major error; because the issues are straightforward there is no danger of confusion. In song 10 the words of the two manuscripts (KP) are printed separately, but the music is taken only from P, with the variants of K noted. Song 11, as previously mentioned, is in X alone, and being imperfect at the beginning has lost virtually all its melody.
This volume, then, contains a great part of the evidence that can be brought to bear on the question of how these songs were once sung. There are many imponderables, those of rhythm and nuance among them, to which the answers may never be found. But as if to compensate for this uncertainty, there is the unspoken implication that the precise form taken by a musical setting is less important than the fact that these poems, and all others like them, were at any rate sung. We do not know who devised the KNPX melodies: there may be a presumption that it was the poet who did so, or at least gave them his blessing. The melodies of V carry less authority, not because the manucript is in a minority of one as against four but because it regularly exhibits a divergence from other manuscript traditions where these exist. But even the KNPX evidence leaves enough loose ends open (and not only in the case of Richard de Semilli) to give the impression of a casualness that is at variance both with later canons of artistic word-setting and indeed with those of contemporary ecclesiastical chant. It is this flavour of insouciance about the whole matter (after all, far more manuscripts lack melodies than have them) that makes the accumulation of evidence more significant historically than the attempt through editorial choice to arrive at the best or most authoritative versions; and it is in the light of that consideration that the omission of the melodies of V seems particularly regrettable.
|Printer friendly Cite/link Email Feedback|
|Publication:||Music & Letters|
|Article Type:||Book Review|
|Date:||Aug 1, 1994|
|Previous Article:||The Music of Bela Bartok.|
|Next Article:||The Harp in the Middle Ages: The Symbolism of a Musical Instrument.|