A
reminder that the Scores Subcommittee is currently (2005) transferring the
existing vocal score of the Gothic into a Sibelius 4 computer file. At this
stage we need to ensure a very accurate copy, with only obvious mistakes
corrected. Once this is completed we will ‘synchronise’ this score with the
current FS. We know that the current FS contains thousands of errors.
However, if we seek to make the VS ‘right’ before we have produced our own
edition of the FS, we will simply introduce many discrepancies between the
FS and VS which will only hold up rehearsals, create confusion and,
incidentally, diminish our standing in the eyes of the music profession. The
opening bars of the Judex offer an interestingly complicated
page of vocal score with which to show off our setters’ skills (in this
case, Alan Marshall). Ex 2 reproduces (c 90%) the whole page of the VS for
choirs C and D, set using Sibelius. I must reiterate that, at this stage,
the setter’s job is to reproduce what is in the existing VS, only making
obvious
corrections. For the purposes of clarity in this article I have undone some
of the corrections Alan made. Please don’t think this is the final version
of the page—it isn’t; but it does serve to illustrate some interesting
points.
HBS vocal score
Firstly may I draw your attention to some of the items covered in part 1 of
this series. Reading from the top of the page, we have:
- Choir A, B cues: these help the singers of choirs C and D to time their
entries
- One vocal line per stave
- Use of lettering to distinguish choir names (A, B, C, D) and voice parts
(S, A, T, B)
- The children’s music (this is on two staves when the singers have music)
- Orchestral cues: again, these help singers to orientate themselves. Choir
and orchestral cues only appear on pages where there is a need for them
- Rehearsal piano: perhaps I didn’t emphasise enough in part 1 that this is
a big issue in creating the VS because the current VS doesn’t have anything
at all for the accompanist to play except a few orchestral cues.
Current vocal score
It is worth studying ex 1 carefully to see what inconsistencies and problems
may trip up the setter.
1 Bar 2. Noting that the time signature is 3/2, choir A sopranos’ and altos’ music doesn’t add up. Sibelius makes it difficult to set this as written: it forces the arithmetic to be right or, alternatively, for the time signature to be changed for this bar. The general appearance of the page makes it likely that choir B’s Ju- comes in after A’s second Ju-, followed by C and D so one might deduce that the rhythm should be crotchet-crotchet-semibreve (rather than final minim) and then look to other parts for confirmation, for example by inspecting how notes are aligned vertically. The problem here is that the only other parts with music here (choir B, SA) are also wrong in this respect! As will be seen later on, one of the problems is the ambiguous value of the ‘minim rest’.
2 Bars 4-5. As usual with Brian manuscripts, supposedly vertically
aligned hairpins are of varying length and start and finish at different
times. Ex 2 shows our decision. Either the hairpins start at definitively
different times, in proportion to the moment the choir enters, say, or they
are simultaneous
(anything else, if executed accurately, will sound indecisive).
3 Bar 6. Choir A, altos: again this bar doesn’t add up. However, the
clear evidence from the layout of the other 14 parts is that the values
should be semibreve-minim. This type of error is silently corrected by the
setter.
4 Bar 6. Choir D, SA: the Ju- doesn’t have the ^ accent that this
syllable has in all other parts on this beat. Again, this would be
automatically corrected by the setter.
In all cases, we have a published score to consult (albeit one containing
many errors!) and from which we might expect good guidance, if not a ruling.
This is a luxury the SSC doesn’t usually have when seeking to resolve this
sort of confusion in other manuscripts. And, regarding point 1 above, bar 2,
choir B is wrong in the full score, too (all parts)!
Irresolvable discrepancies
There are more discrepancies where it isn’t obvious whether the VS or FS is
to be preferred. In the absence of another reference document, an expedient
decision will have to be made.
5 Title. This may be nitpicking; I raise it because the FS title Judex crederis venturus is surely wrong—it should either be Judex crederis (as in the VS) or Judex crederis esse venturus.
6 Tempo. The FS has Adagio molto solenne e religioso, while
the VS has Adagio molto. Perhaps a small point, given the difficult
of creating a very solemn and religious mood in this dissonant polyphony,
but this marking lasts 27 bars—including a soprano solo—for most of which it
surely is achievable.
7 Bar 3. Choir C, SA: the ambiguous minim rest strikes again! The FS
has note values minim (rest)-semibreve, while in the VS there is semibreve
(rest)-minim. Thinking the music through, the VS version has the entries
increasingly frequently (B comes in five minims after A, C three minims
after B, and D one minim after C) which will build up the tension and seems
compositionally plausible—indeed, an accelerando slightly later would seem
to reinforce this dramatic idea, though unfortunately this accel. doesn’t
appear in the VS! (See next point.) The FS entries of C after B, and of D
after C, are equally paced, which raises in my mind whether the entry of B
one minim after the second Ju- in A can be ‘right’ in this scheme.
8 Bar 4-5. The FS has an accel. poco marking (cancelled at the
start of bar 6 by Tempo) which is simply missing from the VS.
9 Bars 5 and 6. A tenuto mark over the -dex syllable in each bar of the
FS is omitted from the VS (all parts). This discrepancy applies to all the
tenutos in these phrases, so either one of the copyists (of the current VS
or the FS) has made a conscious decision—or they are working from
different mss (a truly ghastly thought in all its implications).
Performances with or without the tenutos will sound different, and my
personal view is to follow the FS. Incidentally, the lack of a tenuto over -dex
in bar 1 (choir A) is consistent with the FS.
10 Bar 7. A crescendo which starts on cre- in the FS (all parts) does
not start until the second crotchet of the following bar in the VS (not
shown in exx 1 and 2).
Performance clues
Interestingly, notice that ex 1 includes the pencil markings of some valiant
soprano in choir B. I deduce she was in a UK performance, rather than the
Marco Polo recording, because she has written STAND at the beginning of the
first system. Note that (a) the singer has marked a little pause at the end
of bar 5—an interpretative point the SSC copyist will certainly ignore!— and
(b) she has marked the beats in her parts (1 2 3 1 2 3 …) even though it is
quite simple. She gets it wrong in bar 2, the very bar which is wrong in the
printed score, implying to me the presentation of bar 2 in the published VS
caused her uncertainty. I don’t have any conclusions to draw from this yet,
save to point out the wide variety of mistakes and discrepancies in
evidence. These range from those which can simply be corrected without
comment, via those where, for the sake of producing a viable performing
version in a finite time, decisions must be taken on the basis of
plausibility without convincing evidence, to those where there is genuine
puzzlement over which of two sources might be ‘right’.
© Jeremy Marchant 2005
270709 Havergal Brian - the official website HOME