Philip Legge
I should begin by congratulating the Scores Subcommittee on its progress with the vocal score of the Gothic. Many valuable issues are raised and the article points to further ones, most of which I hope this article addresses. I was unaware the old vocal scores describe the four semi-choruses (or are they demi-semi?) as A, B, C, and D. I believe this convention is far superior to the cumbersome I A/I B/II A/II B in the Cranz full score, and should be retained. The chorus master can more quickly shout ‘choirs C and D’ than unambiguously refer to ‘II A and II B’, and there is less danger of confusing multiple different Is, IIs, As and Bs—especially if asking for individual voice parts, Bass 1 and Alto 2 say, which are occasionally described in shorthand as B1 and A2! The orchestral conductor may invariably still refer to the choirs by virtue of reading the full score’s nomenclature, but by that stage it should be obvious that ‘II A’ = C and ‘II B’ = D. And eventually the new VS will be incorporated into a revised full score anyway…
Dividing the VS into two volumes is a brilliant idea, and has many ramifications. In many places the writing will reduce to the traditional four vocal staves plus piano, albeit with differences between the music found at any particular bar in the two volumes. The covers should be given different colours to plainly distinguish volumes I and II by sight. I suspect volume II will be slightly thinner—in movement IV, choir D is never truly independent of the remainder (except by virtue of being silent!); the children’s choir is tacet in V, and in VI appears on just 2 pages of the full score. In fact their part is slight enough to justify a separate, cut-down VS—even for a complement of only one hundred children. This would be a smaller, lower-cost print run, as without the adult voice parts their score will be much shorter, thinner, and importantly, lighter. In practice, adult mixed choirs almost always rehearse separately from children’s choirs anyway, except for the few final general rehearsals when all of the disparate choirs are brought together.
I am not implying for a moment that the parts for children’s choir should be removed from the second volume: if in the dire event that a performance of the Gothic could not obtain a children’s choir, then the burden would naturally revert to the adult choir, being the core singing body. It makes sense then for choir II to be utilised: in movement IV they are slightly less busy. In any case the chorus master may wish to exercise his or her prerogative by reinforcing the children’s choir with some adult voices, because of the ever-present issues of audibility (or confidence), so their music must be printed in the main VS.
Even if the adult choir for the Gothic is as small as three hundred voices—which must be close to a real minimum number, if not the HBS’ recommended one—then nonetheless it is almost certain to have been formed by combining two or more smaller symphonic choirs of between 100 and 200 voices, each of which will usually have its own conductor and repetiteur(s) (and emergencies who deputise from time to time). Thus the overall chorus master who pulls the forces together will not have much of a worry rehearsing any particular component choir in the earlier stages, or finding a repetiteur; his problem is at the final combined rehearsals where he will have to juggle the two vocal scores simultaneously, as well as the full score.
It might be desirable to ensure the page layout of each volume is similar, so that the music doesn’t look radically different when comparing the two at any point. The chorus master will have his or her work cut out advising the conductor of the many errors in the vocal parts of the full score: a vexed question will be how many discrepancies between the choir and orchestra parts are allowed to remain in overhauling the vocal score, before a new full score can be prepared. Fortunately the task of proofreading the new VS may be split between different individuals, just as the typesetting was. This would still need to be limited to a small group with access to the older, handwritten VS: although the composer’s manuscript full score of the Te Deum is missing, the vocal short score is not, and this may also provide some help with the wide range of clerical errors; the Cranz full score is no help, being frequently and obviously wrong (here I know I am flogging a horse that is long dead). There are also some curiosities of Brian’s vocal writing—and the errors in the actual text that is sung!—that I wonder can actually have been intentional on HB’s part. The most obvious textual discrepancy which should be corrected is the phrase et rege eos et extolle illos—according to his Modern Mystic article on how the Gothic came to be written, Brian said he obtained the Latin text and a parallel English translation from a musician at Brighton College surnamed Allen. Less obtrusively wrong, movement IV ends with Patrem in the FS (presumably correct in the VS).
In several places there are both missing or incorrect hyphenations and conjunctions: both de victo and vene randum are one word, but fac-cum is actually two; both qu-os, qu-emadmodum have one fewer syllable while Dei often requires an extra one. Finally there are misspellings of the authentic text, probably because the text which Brian received was itself in error, of which credentilus instead of credentibus is the most obvious. No one expects to see such mistakes reading a Bruckner or Kodály Te Deum; so it should be likewise for Havergal Brian.
The previous article finishes with some rhetorical questions regarding the nature of the piano reduction, which are not unlike the questions I had to ask myself when commissioned to prepare a new VS with piano reduction for the Gurrelieder choruses (a work I’ve now performed twice, in each case with the entire choral forces using my edition, rather than the traditionally unhelpful choir partbooks). There is the extremely valuable point that the reduction should help provide an adequate musical experience, enabling the average chorister to feel some level of confidence when rehearsing the work. With something as forbiddingly difficult as the Gothic, this is probably essential; the piano part is perhaps the most important feature of the new vocal score, as I will endeavour to illustrate.
Something like Mahler’s eighth—at times also a hard slog for the singer—has less intrinsic difficulty since Mahler has virtually been canonised since the turn of his fortunes, and choristers can easily listen to recordings so they have the ‘end of the tunnel’ clearly in sight. Brian’s idiom and his obscurity as a composer are more likely to provoke a mixed reaction in the collective choral mind—’is this really great music, or just rubbish?’—that is not helped by a repetiteur fumbling for notes in having to sight-read multiple vocal staves to create the accompaniment. (Some players have the skill to combine disparate parts like this, others do not.)
‘Will the repetiteur be able to play it (ever)?’
The traditional piano arrangement for Gurrelieder by Alban Berg (which does not appear in the sub-standard choir partbooks) is designed for a virtuoso pianist: to be fair to Berg, it was supposed to replace the orchestra for a chamber performance rather than support the choir. Although only one of the six repetiteurs in the recent Australian performances actually found it unplayable, it should not be assumed that all repetiteurs are thorough professionals capable of rendering at sight a densely polyphonic texture that seemingly requires three (or more) hands. It is better to always attempt to simplify such an inherently complicated work.
‘If they could play it, how much help will it be to the choir?’
Again I believe some piano reductions fail the test of usefulness by unnecessary complication or providing inaudible cues: with an orchestra the size of the Gurrelieder or the Gothic, not all of the polyphonic lines are going to be audible, especially from the bass- and brass- heavy acoustical perspective where the choir normally finds itself, at the back of the orchestra. When the accompaniment is not simply doubling the choir parts, it must articulate the most clearly audible swathes of harmony, melody, and rhythm which will help the choir to co-ordinate their part; simply duplicating Brian’s chromatic (sometimes polytonal) counterpoint would arguably be making the task more difficult. I believe the specific example cited by Jeremy (the final pages of the Judex) actually demonstrates this clearly. The cue of a descending violin line would be almost impossible to hear as a violin line, if the choir is (as usual) placed behind the orchestra. Enough of the upper brass and percussion are also playing and would be positioned closer, so as to obliterate the sense of the line occurring at the pitch of the upper strings: the violins and violas are also positioned to project their sound out into the auditorium rather than back towards the choir. The same line would actually be perceived as a bass line, doubled in the lower octaves by the massed trombones, tubas, and the weightier lower strings, while the most clearly audible texture, both in rhythm and block harmony, would remain the upper brass. The musical example [next page] provides a conjectural short score of this passage: the choir texture is often spread across two octaves, but is reduced to a single handspan for the repetiteur, so occasionally a bass part is heard in a higher octave, or a soprano part in a lower one. While the choir sings Ah the piano can articulate the prominent horn and trumpet rhythm, which also provides a helpful pick-up for the next bar. The left hand is able to fill in the wandering bass line and the descending violin line as accompaniment, but at the more prominent lower octave(s). Five bars from the end, the choir is essentially in organum, so only one hand in each piano part articulates the choir’s rhythm. This allows the brass and timpani rhythm to be thrashed out by one repetiteur, and the ascending bass line by the other. For greater sonority the piano parts are usually separated by an octave, although each is complete in itself.
In the final five bars of Judex there are subsidiary chromatic harmonies in the upper woodwinds and strings which could form part of an accompaniment: however they would be dramatically less prominent in performance, as well as being in direct tonal opposition to the choir, and thus quite justifiably are omitted. Conversely soloists may be expected to stand at the front of the orchestra close to the conductor, so the parts of the vocal score which form ‘solo arias’ can and should rely more on the string parts for the accompanying texture, because these singers will be standing directly in front of them. Finally, singers tend to listen to whatever recordings are available, so the Naxos CD is likely to remain the benchmark for the foreseeable future: the piano reduction should be designed to be readily followed with the recording (which often spotlights certain instrumental lines). A caveat is that a phrase easily heard on a CD may be impossible to hear in concert, or that what is ‘audible’ will vary considerably between individual recordings and performances, as well as owing simply to placement—one choir might be relegated to a gallery (with the off-stage bands, say), while another is sharing the extreme back of the concert platform with the heavy brass and the hordes of the percussion battery.
‘What exactly will it be?’
Being a vocal score, it should cover all 1529 bars of Part Two, despite there being lengthy passages for orchestra alone and a lesser need for piano reduction of the same; fortunately there is no real need to score the 825 bars of Part One, though a ‘Rolls-Royce’ solution might do so. Although the solo quartet has been allocated to volume I, there is probably a need to indicate what the soloists are doing in the actual ‘solo aria’ sections in volume II—for the Soprano, the Judex and Aeterna fac; Tenor, Te ergo quaesumus; and Bass, Dignare Domine—otherwise the choristers of Choir II will simply be following a piano reduction for long stretches of time, when the main focus of interest is the solo line. For this reason I do not believe these should be inserted in volume II as cues; the solo line only adds one stave per system at a time (unlike the instances where the solo quartet is used, which would require four staves). The poor old contralto soloist only sings for 48 bars in the whole thing, and would thus never appear in volume II at all.
There are not many lacunae in Part Two, aside from the two big orchestral interludes in Judex: the orchestral fanfare at the start of the Te Deum is the only such passage in movement IV—which along with the shorter 10 bars of fanfare before figure 101 would be pointless to omit, the rest being firmly choral. The few gaps between the end of a solo in the Te ergo and the next choir entry, eg, between the Tenor finishing at figure 267 and the choir entry at 271, or again the Soprano solo finishing at 278 and choir II entering at 280, are precisely the places where you would want the accompaniment to continue, to ensure the choral entries don’t get lost! The clarinet march is an obvious interlude, but it would be fairly trivial to notate and play: on the other hand, the Varèse-ian outbursts of timpani and brass at the end would require all of the virtuosity of La Main Gauche himself.
Only at full general rehearsals without orchestra (or the case of a particularly large choir split across both choirs I and II) would two repetiteurs be needed at the one time, since different choirs normally rehearse separately. Despite this, the piano accompaniments for choir I and choir II can be designed to be complementary to one another as well as reducing the respective voice parts into a playable texture. For example, when choir I alone is singing with orchestra, piano I should provide the vocal reduction, while piano II reduces any salient orchestral lines; if the choir needs to hear what the orchestra is doing, the answer may be found in the other volume’s piano part.
In a full choir rehearsal without orchestra, (a piano call involving the conductor meeting the vocal tutti is standard practice here), the two vocal scores allow accompaniment by four hands on two keyboards, with each choir providing and being supported by a repetiteur; if only one actual piano happens to be available, an electronic (and realistic sounding!) keyboard is an adequate and easily obtainable alternative.
© Philip Legge 2005
Newsletter, 2005