This article presumes to give some guidance on the choral component of the Gothic symphony to the chorus masters and concert conductor planning a performance.

Several years into my career as a second tenor with Crouch End Festival Chorus, I decided I should take singing lessons, the better to contribute to that fine body of musicians. After some weeks of light Italian operetta (my teacher was a light Italian), I proudly marched up to the music director and informed him of my commitment to the choir etc etc. “Oh yes, he said. I thought I could hear you!” The moral of the story being that, even in good symphony choruses, a headcount is not necessarily an indication of strength or quality. I must hasten to add, for those readers unacquainted with this choir, it was pretty good when I was there – successes which it achieved despite my contribution rather than because of it. Since I left, it has become excellent. Crouch End Festival Chorus, in my day, could summon 120
singers of a standard sufficient to deliver a good performance of A child of our time (I might add we acquitted ourselves excellently in repertoire as recherché as The Plague, Jeanne d’Arc au bûcher and Harmonium). When it came to Belshazzar’s Feast, the tenors in particular, struggled. The problem, apart from the high tessitura, is the amount of the work that is written for two choruses, halving the number of singers on each line. I daresay that today’s, much improved, CEFC would have no difficulty in this area but it is a truism that, in the UK generally, men are less attracted to choral singing than women and, as a result, our amateur choruses can be a bit light on the chaps. And, if you’re any good as a choral male, you have the pick of choirs to sing with, thus perpetuating the excellence of the best and
putting a permanent brake on the less good. A child of our time has very little divisi in its four part choral writing and, although Tippett’s melodic lines can appear ungrateful when one is learning them, curiously, with practice, they become as easy as the spirituals which also form part of this work.

I am envisaging a performance of A child of our time. Apart from the four soloists, in front of me is an orchestra of 21 players with strings to match, and a choir – could be as many as 100. I replace that image with one of Belshazzar’s Feast. One soloist, an orchestra of thirty (plus organ) and strings to match, and a choir – definitely well over 100, not only to support the eight part choral writing,but also to provide sufficient weight to balance some massive orchestra tuttis – not to say the two additional seven piece brass groups.

Now, shift up another couple of gears to the Gothic. At the front of the stage are the ‘standard’ four soloists. On stage are eighty players (including organ) and a body of strings to match that vast array of wind, brass and percussion. This is the equivalent of two complete symphony orchestras – and then some (check the photograph of the recording session printed in the CD booklet of the original Marco Polo release). Before I visualise the choral forces, let’s look at the score. Part two is described in the full score as being for ‘Double Chorus, Orchestra, and four extra Brass Orchestras’. (So much for the ‘Soloists’ then.) Passing over precisely what the word  extra’ means, we flip the pages to see what Brian’s writing for a ‘Double Chorus’ is like. (I need to ask some readers for their forbearance at this point. I appreciate that I making my point with Brianic emphasis.) It’s clear that much of the ‘double chorus’ music consists not just of two SATB choirs, but that each of these is further
divided. In reality, there is far more divisi into four SATB choruses than there is double chorus writing in Belshazzar’s Feast.

It is often said that Mahler deployed his large orchestras like a painter, requiring many instruments in order to tease out subtleties of colour, of light and shade. Brian, though, is more of a Walton man – not averse to throwing the whole orchestra at the listener. The fact is that the choral forces needed to balance the Gothic’s instrumentation are going to be massive if they are even to be heard. But there’s worse. Three other factors make marshalling adequate choral forces for a performance of the Gothic harder than one would like. They are duration, difficulty and divisi.

Firstly, the duration: it is a long sing. Part two lasts well over an hour. While it’s true that there are stretches without choral involvement and other sections in which not all the singers are performing (and these seem to be approximately evenly divided between singers), your average singer is probably singing for forty minutes. (For what it’s worth, The dream of Gerontius has 31 minutes of choral singing.) It requires stamina and staying power.

Secondly, it’s difficult. As the Naxos recording painfully shows, not only are the lines ungrateful, they move an awful lot of the time in tones and semitones, usually downwards. It must be difficult to avoid the impression, in the immortal words of David Temple, chorus master of the Crouch Enders, of a herd of cows on the way to the abattoir. With difficulty comes uncertainty and, with that, loss of volume and clarity. (Removing difficulties and uncertainties associated with reading, or even understanding, the copies is the prime reason for the society’s production of the vocal score.) I have yet to discover whether sufficient familiarity with Brian’s writing enables it to click into place in the way Tippett’s does.

Thirdly, the division into four SATB choruses is a given. There is then divisi within these 16 parts. The most extreme example of this occurs in the fifth movement, from fig 160, in which each of the four bass parts is divided a4, making 16 bass lines alone. (The tutti choral forces are in 37 parts at this point.) Brian deploys the basses in four groups of four, and, while there is some doubling, there are spatial reasons (which I will come on to) why this cannot be used to solve the problem. There are sufficient other passages in which one or other of the ‘basic’ sixteen choral lines is divided, at least a2, for the need for a strong choir to be clear.

Given the four factors of orchestral weight, duration, difficulty and divisi, it seems an inescapable conclusion that the minimum choral forces needed for the Gothic are four symphony choruses, each of the standard of my friends in Crouch End. Candidate choirs in the UK might be CEFC, the London Symphony Chorus, the City of Birmingham Symphony Chorus, Brighton Festival Chorus, and so on.

Before going on to make some helpful suggestions, let me add two further considerations. Firstly, Brian’s ‘handedness’. I don’t know if Brian was left
handed, but the music sounds as if he were. Speaking as a left handed pianist, I know such a person will be more interested in what the left hand is doing than a right handed player is. Brian’s music often seems to need to be read from the bottom up to be understood. In the Gothic, where such a heavy musical load rests on the choruses’ shoulders, this means they have to be able to enunciate and deliver that bottom up writing. In other words, the men have to be particularly strong yet, as I suggested at the beginning of this article, this is the area where amateur choirs are weak. I feel there is no alternative but to supplement the four symphony choruses with extra men, perhaps in the form of a couple of male voice choirs. Secondly, the children. Brian also asks for a children’s chorus, singing either in unison or divided a2, boys and girls. His use of this choir is idiosyncratic: not at all in V and only briefly in VI. However, in IV they have a deal to do. Part two opens a capella with half the women singing a3, forte, with the children, ff, singing a counter melody. However, by fig 41, the children are up against half the full adult choral forces and, by fig 44, the adults are ff and children are a2, against big, forte ‘stately and majestic’
music throughout the orchestra. If the children are to be heard, there will have to a lot of them, too.

Going back to visualising the performance. An area the size of a small playing field stretches into the distance accommodating the orchestra. Surrounding, and behind, them in a vast semicircle are the four choirs, each some 150 strong, such that, of you were relatively close to the stage they would occupy close to a 180 degree field of view. Behind each of the choirs is its brass and timpani group. (And somewhere are a couple of hundred kids.)

This spatial disposition is crucial in clarifying Brian’s music. Indeed, the music is written so that it can be presented in this way. (The choirs are labelled IA, IB, IIA, IIB , from left to right.) The ultimate spatial effect is the section in V (fig 203) where each of the four choirs, accompanied by its offstage band comes in, in turn. There is simply no point to this if the audience doesn’t get the thrill, as in the Berlioz Requiem, of one massed force after another crashing on the shore like tidal waves. [The full score copyist blotted his copybook by bringing in the choirs not in the order IA, IB, IIA, IIB in one magnificent sweep from left to right, but in the order IA, IIA, IIA (again) and IIB. This is an example, albeit the most egregious, of where the society has silently corrected what
are manifest errors in either the full score, the extant vocal score or both.]

Brian is inventive in the way he deploys his choral forces in differing groupings. Sometimes the women en masse (eg fig 270), sometimes in eight parts (1 before fig 323) where the spatial layout of the singers is essential for the audience to hear the polyphony. Then there’s the section (fig 113) where the tutti chorus alternates with either IA or IB (Tu rex gloriae Christe, etc) culminating in a tutti, pp, declamation of Tu de victo mortis (fig 124), followed by the same, ff, but from only IA (a quarter of the forces, and divided a8) only [*1].

There’s a lengthy section (2 after fig 280), for choirs IIA + IIB, balanced elsewhere by a dominance of IA + IB. An interesting episode is the one concluding with the massive male outcry, fff, Salvum fac populum tuum Domine (ends fig 320). Brian has the men in 32 parts (1 before fig 320), but is actually replicating the same eight parts across the four choirs. He could have assigned one tenor part and one bass part to each of the four choirs, but the sound would
have been completely different. As written, it is a massive wall of sound which launches the women singers’ and orchestra’s response (they can’t even wait for the men to stop singing before they start!).

Compare that with the fortissimo Non confundar in aeternam right at the end of the work (fig 427). Here the eight part writing (for tutti choirs) is written on four staves, two parts each. For reasons I want to expand on in a future article, I am confident that the top part on each stave is for IA + IB, and the bottom part for IIA + IIB. The audience thus hears a homogenous sound, for sure, but it is made up of two distinct chords, one to their left and the other to their
right.

Returning to that section of V (fig 160) which calls for four groups of four bass parts each, one solution to providing 16 strong bass lines when only eight are sounding at any one time might be to require individual singers to perform more than one part. However, the music is more problematic. At 1 after fig 160, IIB basses enter a4. In the next bar, IA basses enter a4 with different material though, at the following bar, they are singing the same four parts as IIB. The trouble is they are half a mile away, spatially. Brian wants the audience to hear two distinct groups which are then answered by the other two groups, also singing their same material. We need to remember Brian’s interest in renaissance music for these effects to seem entirely appropriate and natural.

My view is that the choral forces pose the most difficulty in mounting a performance of the Gothic. Of course, booking and paying an orchestra, a hall, a publicist and so on are huge hurdles but, whereas a good enough orchestra will play the music and good enough soloists will sing their parts, if the choir can’t deliver the goods, the performance is compromised. Unlike Gerontius, it isn’t a work for soloists and orchestra where the chorus is thrown in for a bit of light relief. Part two, at least, is equally divided between chorus and orchestra (I’m tempted say the soloists are thrown in for a bit of light relief).

So the first thing is to encompass the size of the enterprise. I suggest not thinking of this as Mahler 8 plus [*2]. Mahler obviously demands a cataclysmic finale but the effect of Veni creator spiritus is not solely, or even largely, determined by the size of the choir. Brian’s is music on a completely new scale: it is truly a ‘symphony of a thousand’ where the previous work to trade under this name was merely exercising marketing hype. The second is definitely to believe you can do it. It’s been performed in a church by amateurs for heaven’s sake [Hanley, Staffs in 1978] with fewer people in the audience than were performing. If they can do it, you can. Be aware of the tipping point effect. Confidence is catching and enough good singers will fortify the rest. Then, there are workarounds.

Going back to the 4 × 4 part bass problem (fig 160), a workaround would be not to double up IA and IIB on the grounds that they are singing together, but to
double up IA and IB, and IIA and IIB, on the grounds that they never sing together in this passage. Some degree of spatial interest is maintained, and choruses singing these parts should be expected to divide their basses a2 on occasion. (One still needs plenty of good basses however.) Where the number of parts is reduced (and there are passages of simple SATB or 2 × SATB), rest singers by not deploying all of them all the time in these passages if the orchestral volume permits.

[1] There’s something immensely powerful in a large chorus singing very quietly. Equally, a (relatively) small number of singers, in comparison, singing loudly has an emotional kick: voices singing in the wilderness.

[2] Mahler requires an orchestra of 50, plus 7 offstage instruments.

NL199 © 2008 Jeremy Marchant

BACK TO TITLE PAGE

070312 Havergal Brian - the official website HOME