Creating the vocal score of the Gothic – 1

Jeremy Marchant

The Gothic symphony, as in all other respects, presents unique challenges for anyone producing a vocal score. Unsurprisingly, these largely come under the following headings
- complexity of the choral writing
- the piano part

But first, a bit of background. The Scores Subcommittee (SSC) agreed to produce a vocal score of the Gothic when a performance in Brisbane performance looked likely. To speed things up, no fewer than five setters (I prefer the historic term engravers, even though we all use the Sibelius computer program) were allocated to the score, two each on movements IV and VI, and one for V. A sixth person, me, is pulling the five contributions together to make a uniform document and adding the piano part; and at least one more person will be needed to proof read the final document. My credentials include editing and engraving work for Chester Music and Schott, as it happens mostly of choral works. Brian is my first dead client.

This basically means how many parts are there? A large majority of works for symphony chorus and orchestras (as opposed to more expert chamber choruses) simply have a basic four part SATB choral layout, with a few soloists thrown in. What complexity is restricted to occasional passages where one or more of the voices splits into two parts. Then there are works requiring two SATB choirs, such as The dream of Gerontius.

In all cases, the engraver needs to work out the values of the basic parameters which will govern the layout of the score. How many staves per page is the basic measure. The staff size (ie height) has to be the same on all pages, but some pages in Gerontius will require many vocal parts (see below) whilst others only a few. In the latter case perhaps two systems can be accommodated on one page. What is the largest staff size which will fit both the densest pages and the multi-system pages? On the whole, what it says on the title page is a good indication to the engraver of what staff size to choose to accommodate the maximum possible number of vocal parts, without that one section compromising the rest of the score. For example, in the Elgar, the composer divides his choral forces in different ways: sometimes tutti SATB (fig 74, opening of Praise to the holiest…), but usually SATB main choir and SATB semi chorus, though sometimes two equal SATB choruses (eg fig 88 + 4 bars, latter part of Praise to the holiest…). Only at the very end (from fig 133), is it a little more complex: SSAA semi-chorus and SSAATTBB main choir though, as in both choirs the two soprano parts appear on one line, in practice only ten staves are needed. There are few, if any passages, where more than one soloist is singing, so the maximum complexity is reached here with the soprano solo, ten choral staves and two piano staves, resulting in a thirteen stave system. The engraver of my Novello copy of the vocal score has chosen a stave size which is big enough to be visible (though the sung text is a tad small) and which fits the system clearly and neatly with no congestion of symbols.

The chosen staff size therefore allows many pages to contain two systems, eg pages 70 to 81 (ie from fig 32, Low born clods…), even where an extra staff needs to be slipped in—eg, from fig 74+4 which is briefly SSATB. One way of squeezing two systems onto a page occurs when a voice is divided by the composer. For the singer, clarity demands two staves, but, if the two parts are similar (eg Vaughan Williams writing in parallel thirds), the disadvantage of having one stave with double headed notes is outweighed by the value of being able to have two systems per page. This isn’t just to save paper: singers like to see where they are going, so the more that is visible on a double page spread the better. The editorial decision should be that, if the parts are anything other than simple (lots of accidentals, different intervals), and always when they have a different rhythm, they should be on different staves for ease of reading.

Once the score is set on computer, the engraver should experiment with slightly different staves sizes. The name of the game is to have as few pages as possible so that the number of page turns is minimised. But page turns must occur ‘naturally’—with system breaks (ie changes of complexity), often points where a new section starts, or the mood changes. For example, taking Gerontius again, the end of Praise to the holiest… is a nice long ff chord on ways, stretched out over the whole page; immediately after (fig 101, p148) the chorus is silent (and orchestra pp) so that point is the only place for the page turn (though, of course, all choristers will have pencilled in TURN at the beginning of p147 so that they actually turn the page during the loud bit, and look to the conductor for the cut off).

A small reduction in staff size can reduce the length of the vocal score by whole numbers of pages; sometimes increasing it has no effect except to increase legibility. I’ve used Gerontius as example mostly because readers are more likely to have access to copy of that score than the vocal score the Gothic, and also because it is an example of a composer deploying his vocal forces flexibly, requiring the engraver to find solutions to a variety of layout challenges. Nevertheless, an engraver with access to computer software such as Sibelius will have no difficulty is defining parameter values for a vocal score of this complexity.

So, we come to the Gothic. The full score describes the ‘Plan for finale’ as ‘Double chorus, orchestra, and four extra brass orchestras’. On that basis, the score might be no tougher to lay out than Gerontius. At least the vocal score acknowledges the existence of a solo quartet and ‘Girls and boys’! And, of course, we know that the (over) modest ‘Plan for finale’ hides the fact that each half of this double chorus is itself a double chorus and Brian is often writing for SATB SATB SATB SATB. Add the piano, and that’s 18 staves. Using the same process as with Gerontius, the engraver needs to know where the music is at its most complex and multiparted. In the Gothic that section is famously V, Judex crederis… particularly fig 160 to 167 (starting 2:57 into disc 2 track 1 of the Lenard CD) where the double chorus is divided SSAATBBBB SSAATBBBB SSSAATBBBB SSAATBBBB. One issue the engraver should have at the forefront of his or her mind is: what is the vocal score for? When the Scores Subcommittee was founded, its whole raison d’être was (and still is) to produce high quality scores which would present any peruser with no problems whatsoever in investigating the music (and therefore no obstacles in deciding to perform it). This attitude has been applied to the vocal score of the Gothic. The answer to the question, what is the vocal score for?, is: to enable the singers to learn the music (and secondarily to be an aide memoire during performance). Anything which might make the singers’ jobs harder is to be avoided if at all possible. Absolute clarity and unambiguity of image are the orders of the day.

With all that in mind, a slightly radical solution was taken to the problem of how to put 38 choral parts (luckily the soloists and children are silent in this movement) plus piano onto one page and still make it legible in rehearsal. A different solution was adopted by the publishers of the existing vocal score. Here there are two parts per stave which are labelled SATBB SATBB SSATB SATBB on the first page and SATBB SATBB SATBB SATBB thereafter. Given a larger page size, these 20 stave systems are accommodated with a staff size almost that of Gerontius’s. However, the result is that the dense passages are not legible. The existing Gothic VS is neatly handwritten. Computer setting would increase legibility, but squashing two difficult parts on one stave still does not make the singers’ task easy. And, although the complexity of the Judex is exceptional, there are many passages of SSAATTBB-plus-piano where two systems of this don’t fit our given A4 page size if we were to use the stave size of the current VS (slightly less tall than A4). Even if it were computer set, the complexity of the music would render the score unreadable and hence not fit for purpose.

So, editorially, we decided to produce the VS in two volumes, rather than use an illegibly small stave size, or press UMP to sanction a page size larger than A4—it would have to be a lot larger than A4 to be of any use, which would give the singers handling problems as well as substantially increasing production costs and hence hire charges.

One volume contains one half of the double chorus (also known as choir I, or choirs A and B, in the full score) and the soloists; the other volume the other half (choir II, or A and B in the full score, C and D in the VS) and the two part children’s choir. This approach enables complete legibility and the ability to print each singer’s music on their own stave (almost without exception—it’s always a balancing act). And there’s room to provide cues from the other volume. This became an essential requirement once a decision had been taken on nature of the piano part. For example, it allows choir II to follow the first occurrence of the soprano solo in V so that IIB’s basses can get a bit of help with timing and pitching of their entry. (One thing in the back of my mind is the phenomenon of having the score impose itself between the singer and the music: with poorly laid out scores one never loses a certain unease, a slight tension, as a performer which is always apparent when you’re not quite sure of what’s on the page—and I write as a very average choral singer thinking about other average choral singers roped into performing the Gothic.) One immediate problem will probably have occurred to the reader: how can the chorus master rehearse the choir if he or she has to work form two different volumes simultaneously? Here we have to consider the way in which the choirs are recruited. We know that weight of sound is essential for a good balance with the Gothic orchestra and weight means numbers. My own judgement is that the double chorus required by this work is effectively four choirs of the size and quality of say, Crouch End festival Chorus or the London Symphony Chorus (typically numbering 135 each). Such a choir, if it were recruited into a performance of the Gothic, might decide it wanted to contribute across the whole double chorus but, in many passages, it would be allocating very few individuals to each part, simply because of the depth of divisi in the writing. This would tend to result in less confidence in rehearsal and thus a less confident performance, hence less presence as well as a generally unhappy experience for most but the best singers. With a two volume VS, any one choir can still be half of IA+IB, say, or all of IA; and greater cohesion in performance can obtained whilst still giving any one choir a reasonably coherent musical experience in rehearsal.

The piano

It is an interesting fact that vocal scores of choral works with orchestra include some sort of reduction of the orchestral music for the repetiteur to play, whereas the rehearsal piano part for unaccompanied works is a reduction of the choral music, or nothing. So, let us consider the concept of a piano reduction of part two of the Gothic. Who will make it? What exactly will it be? Will the repetiteur be able to play it (ever)? If they could play it, how much help will it be to the choir? Given our premise that the purpose of the vocal score is primarily to help the choirs rehearse and learn the music, it seemed to the SSC that the piano part should consist of the material which would be most useful to the singers. And therefore we decided it should be a reduction of the vocal music on the grounds that this was sufficiently complex that the repetiteur would welcome a nice two handed version, and the singers would benefit form a confident and accurate accompanist.

And, in the same way that we have added cues in each volume from the other volume’s choral parts, each volume also has orchestral cues so that, for example, as the chorus sings its final Ah at the end of V, the descending violin figure underneath the Ah is shown followed by the three bars of timpani to help the tutti entry on Judex.

The piano part is still under construction and I hope that a more detailed exposition of its genesis will form the basis of a subsequent article.

© Jeremy Marchant 2005

© Newsletter, 2005