An Australian premiére of the »Gothic«?

In late 2003 rumours were in circulation that the 2004 Brisbane Festival would be putting on what is arguably the most ambitious symphonic work ever composed, the Symphony No. 1 in D minor (the Gothic) by the English composer Havergal Brian (1876–1972). This ultimately did not go ahead, possibly because the work has a notorious jinx associated with it: the very size of the work invariably brings on cases of cold feet amongst sponsors and promoters. While the Festival had given the green light to the project, it appears advice was solicited from various people on its feasibility, and who advised in the negative. The Berlioz Requiem was proposed instead; further attempts to remount the Gothic again in Brisbane, for the 2007 festival, also came to nought.

It takes only a brief look at the score of the piece to realise its difficulty. It takes a much more detailed analysis to show that is in fact manageable, and even reducible to a smaller body of instrumentalists than the outrageous demands of the composer’s specification (and the die-hard school of Brian purists, some of whom would prefer no performance at all rather than an abridged or reduced one). Hence, in April 2004 I wrote a letter to the Havergal Brian Society advising that the Brisbane Festival might be more willing to support a reduced performance. This was backed up by an analysis of the logistics of performing the symphony, which for the sake of its extreme length was published separately.

Executive summary:

It is feasible to perform some (if not all) of the Gothic symphony by Havergal Brian, without needing to engage a gigantic orchestra (i.e. 200+) and choir (i.e. 600+) in pedantic adherence to the specifications laid down in the full score, and even in a worst-case scenario, should be possible without huge financial loss. (Any work of this magnitude – such as a Gurre-lieder or a Mahler 8 – is fundamentally a loss-making concert, and should be mounted by a festival with ticket sales hedged to recover most of the loss.) This would involve scaling back the fifth movement (Judex, lasting 15′) and possibly omitting the sixth and final movement (Te ergo quæsumus, lasting 35′).

The raw numbers are as follows:

Terse executive summary:

You don’t have to be mad to consider doing this work, but it may help*.

The case for a reduced »Gothic« revisited

This brief letter advanced two possibilities for a small orchestra for the Gothic, which essentially corresponded to options D2 (a) and C2 (d) above:

(1) Use only two off-stage bands in the Te ergo.

(2) Omit the Te ergo, and use no off-stage bands in the Judex.

The full analysis is found in the article below on logistics, amplifying the particular details and giving substance to the arguments given very much in passing.

Comparing Havergal Brian’s Symphonies

Another essay written for the Havergal Brian Society taking another view on the performability of the symphonies generally.

The logistics of the »Gothic« symphony (as published in Newsletters 188 and 189 of the HBS)

Lest any reader believe I am not prepared to back up the claims in this article, let me state that I have typeset some (short!) sections of the Gothic in full score, and have assisted the Scores Sub-Committee of the HBS by scoring one of the smaller symphonies, complete. New vocal scores for the entire choral Te Deum, which would be of enormous assistance in the ways I have alluded to below, are in preparation.

Although I take it that the person proposing to “put on” the Gothic had considerable funding promised (and then withdrawn) by the Brisbane Festival, that is not to say the situation is impossible. The story seems to be that various people in the choral/orchestral scene were asked for opinions, who said the work would not be possible (I would very much like to know who was asked!). Since writing this article I once again performed Gurre-lieder, this time with the Melbourne Symphony Orchestra using about a dozen players fewer than the composer’s specification. The experience only confirmed my opinion that any of the large Australian orchestras really ought to be able to tackle the Gothic in some fashion. The essay looks at most of the possibilities in great detail as well as looking at the option of a “performing version”.

As mentioned in the article, orchestral extravagances are viewed with especial concern by the Arts Administration bean counters, so why would any of the ABC orchestras consider performing this? These paragraphs below, removed from the essay with its more general consideration of logistics, specifically address the Australian orchestra and choral situation.

Firstly, though Brian himself still remains a marginal figure, the Gothic itself is fairly well-known, its notoriety in being the largest symphony ever written no doubt assists. Like works such as Mahler’s 8th or Schönberg’s Gurre-lieder it shares the reputation of an unforgettable, once-in-a-lifetime experience, and would have to be promoted in this way. Sell-outs of the Mahler and Schönberg occur because the unusualness of the works themselves generate intense interest and word-of-mouth amongst the concert-going public, and the Brian would have to rely on the same principle. The work is yet to be heard outside the UK/Europe except on recordings, which should if anything drive interest higher in hearing the Australian premiére, and would be of national interest to concert-goers (and from a broadcast perspective, the ABC would want to record any performance for possible commercial release). As many commentators of Brian’s work have observed, the Gothic is the most crucial work in his œuvre, and until ways are found to perform the work more often, the remainder of the composer’s work, some of which certainly merits attention, will likewise remain unknown.

Secondly, while concerts are more expensive than ever to put on, some of the attendant costs can be partially reduced. The largest cost in monetary terms is the provision of an adequate number of orchestral rehearsals. One of the chief difficulties in exploring this repertoire would be general lack of familiarity with Brian’s idiom; this could be remedied by a bulk purchase of the existing CD recording (just about to be re-released), and issued in advance to the main orchestral players at a cost well below that of say, losing a few hours of rehearsal time owing to unfamiliarity with the size and scale of the music in the earliest stages of rehearsal. This also applies to a lesser extent to the choir, who may also wish to take advantage of obtaining recordings in order to drive down the cost of rehearsal as much as possible. lt stands to reason that practice parts should be made available relatively early on.

Technology should also be enlisted to make the learning process easier for choristers, who have to sing music of formidable difficulty which is no less important to the success of the symphony than the orchestral music. In practice then, where it is possible choristers should be assigned to a specific part within a choir well in advance and issued with practice score and rehearsal CD, again in the interest of minimising wasted rehearsal time. Once the full choir is assembled, repetetive notebashing of unlearnt parts of the Te Deum is sadly impractical without causing rehearsal schedules to balloon out expensively; there is simply too much music in too many parts. The choir is essentially divided into four eight-part units, so the challenge is to adequately span the resources of multiple choirs across 32 or so individual sections.

The second part of the work uses a larger orchestra and sets the ancient Latin text of the Te Deum in three movements: Te Deum laudamus, Judex, and Te ergo quaesumus. The fifth movement sets one single line of the text, Judex crederis esse venturus, which occurs just over halfway through the hymn, so dispensing with either the last movement or both of the last two movements leaves the musical setting of the text half incomplete. There is some debate whether performing an abridged Te Deum is musically adequate, against the more pragmatic assertion that the music more often than not remains unperformed because the Te Deum as a whole is too difficult. The author however is of the pragmatic school, and would very much prefer the possibility of an abridged Te Deum to the alternative of “all or nothing”. At the very least, the author believes a fallback position should always be to simply perform the orchestral movements of Part One alone. Although the orchestral-only section of the work would be much less of a draw card in terms of publicity, it would still constitute an Australian premiére. To date none of the 32 symphonies in Brian’s œuvre have received a performance in Australia.

One specific Australian issue is the presence of the extremely rare pedal (contrabass) clarinet, and whether the sole instrument of its kind in Australia may be obtained (it belongs to the military). It may be expedient to rewrite one each of the bass clarinet and contrabassoon parts to cue in the notes that would be played by it. Equally well, if the pedal clarinet is obtainable, its part could accomodate an eliminated second bass clarinet.

A scaled-down orchestra for a full performance adds up to either 156 or 168 players in total, depending on whether 2 or 4 off-stage bands are used. This is only about twenty/thirty more players than the AYO used to play Gurre-lieder in January 2003, and about twenty-five/thirty-five more than the MSO used to play the same work in 2004, or Mahler’s 8th Symphony for the centenary of Federation in 2001. The difference is largely due to the minimum of 16 off-stage brass required for the climax of the Te ergo.

(a) with 2 bands, 156 players
31 WW (8.8.10.5) – 41 brass (8.7.5.5, 25 on-stage + 4.4.4.4, 16 off-stage) – 4 (14) timps, 12 perc (inc. celesta), 2 harps, organ, and 65 strings (17.16.12.12.8)

(b) with 4 bands, 168 players
31 WW (8.8.10.5) – 53 brass (8.7.5.4, 24 on-stage + 8.8.8.5, 29 off-stage) – 6 (20) timps, 10 perc (inc. celesta), 2 harps, organ, and 65 strings (17.16.12.12.8)

Total numbers including choirs

The recommendation for a minimum of 500 adult singers (supposed to be mainly professional) on the Havergal Brian Society website appears to me to be a slight over-reaction, especially if the orchestra is reduced in size. I would expect 200 professional singers to be heard over an orchestra of 150, and 300 such singers to be heard over an orchestra of 200. The real difficulty is in obtaining them!

This is not to say that 500 singers is a bad thing, only that the so-called “Symphony of a Thousand” (a phrase which could apply to the Gothic just as well as Mahler’s 8th) doesn’t literally require a thousand performers. If the singers are largely non-professional singers, then my estimates of a minimum of 200 to 300 singers (matched to a 150 to 200 piece orchestra) should probably be increased up to 300 to 400, or two adult singers per orchestral player.

There is also a sizeable part for children’s choir, for which a minimum of 100 voices is deemed suitable, though the commercial recording of the Gothic would suggest the contrary; in the very first entry of the Te Deum the sopranos and altos from one of the adult choirs completely dominate the children’s choir, who are initially at the very bottom of their range. In the worst-case the parts given to the children’s choir could be sung by the upper voices of the adult choirs, however the presence of children in works such as this (e.g. Mahler 3 or 8, the Berlioz Te Deum) implies a different aesthetic purpose than can be achieved by substituting adult choirs. 150 or 200 voices seems a more sensible number to me, though I will concede this is quite debatable. Almost all of the work for the children’s choir is in the fourth movement; they do not sing at all in Judex, and in only one brief episode of the Te ergo.

The total minimum number of performers appears to me to be about 500 or 600, with an orchestra of just over 150 players matched by between 200 and 300 adult singers, and a children’s choir consisting of 150. More extravagant performances could use the full 200 strong orchestra, with three or more singers for each orchestral player.

Considerations for finding choirs and capable singers

Finding four large SATB choirs capable of mastering the demands of the choral music will be a major challenge, and also in ensuring that enough truly competent singers can be spread across each polyphonic part, as the music approaches 40 vocal lines at times in the a cappella section of Judex. (The description of “two large double choirs” proves inadequate when the composer proceeds to divide each double choir – traditionally notated SATB–SATB – into 2×SSAATTBBBB!) However, as I said above, 500 adult singers seems to be a ridiculous “minimum” number – I have not heard of performances of Tallis’ 40-part motet “Spem in alium” with literally dozens of singers to a part. Beyond a certain number, adding more singers does not really increase the volume or accuracy of a choir, but merely diffuses the sound as the strong singers are spread more thinly through the group and increases the choir’s inertia.

Experience proves that in massed symphonic choirs, there are very few true leaders in all-round terms of confident sight-singing of pitch, subdivision of rhythm, and general musicality to responsively sing their part in concert with a full symphony orchestra conducted by someone who is probably standing fifty metres away on a podium (and it is this that separates the professionals from the non-professionals). For the non-professionals, ample rehearsal time is necessary to bring their performance up to the required standard. For a work of this difficulty, the preparation of rehearsal CDs seems evident, with the logistical issue that 16 (or possibly 32) separate CDs are required (if they are of the kind with each singer’s individual track isolated in one stereo channel with the other parts and accompaniment separate).

However, the two principal symphonic choirs in Australia – which it might be argued are the Sydney Philharmonia and the Melbourne Chorale – each have a core body of strictly auditioned, experienced choral singers. I refer to the Philharmonia’s Motet Choir and the Chorale’s Ensemble, each of which consists of about 40 voices. It would be desirable to have about 100 singers from choirs such as this to form the bedrock of the two mixed-voice double choirs, so that there would usually be a minimum of two competent singers dedicated to each polyphonic part in the choir music.

This is not to say that experienced and competent singers cannot be found in other symphonic choirs such as the Royal Melbourne Philharmonic, or the choirs of the Australian Intervarsity Choral Societies Association, or in specialist small choirs, but the process of finding these singers is necessarily more difficult as these choirs do not have identifiable sub-groups comprising solely of the more talented singers.)

The choir music is not uniformly demanding or evenly difficult throughout the Te Deum. The fourth movement beginning Te Deum laudamus is probably the easiest, though it has extensive a cappella sections requiring solid intonation. The adult choir splits into only 16 parts at maximum, and the children’s choir in 2 parts. On occasion there are four-part divisi in sopranos and altos. In this movement, one of the subsections of the second double choir (denoted II B), is never independent of the rest of the choir.

The fifth movement’s setting of Judex crederis esse venturus is more difficult again, and begins with an extended a cappella section which immediately splits each of the soprano, alto, tenor, and bass parts in eight; another passage requires sixteen bass parts (mostly moving in parallel). The music is dissonant, chromatic, and difficult. The children’s choir does not feature in this movement.

The sixth movement beginning with a tenor solo Te ergo quæsumus is the most diverse movement musically, and contains both the simplest and probably the most difficult music for the choir to sing. The children’s choir reappears for a brief instant, and their music contains a short solo for a girl singer. The adult choir usually splits into 8 or 16 parts, and briefly into 30 parts at the most difficult passage in the whole piece, which encompasses the entire adult choir and full orchestra:
Choir II, pp. 202–213 of the full score, fig. 280–300
Choir I, pp. 210–213, fig. 293–300
I would suggest ensuring the core singers who are able to sing this horrendous passage are explicitly voice-placed, regardless of the abilities of the rest of the choir; or as an alternative (in spite of Brian’s horror of such doubling, which mainly concerned the a cappella music) consider using the un-utilised off-stage brass to help double the non-professional choirs at this point, where possible, since the 8 instruments in each group correspond neatly to the 8-part division of each section of the choir.

The following a cappella section for men’s voices, ”Salvum fac populum tuum”, also divides the tenors and basses into about 20 parts and presents a level of difficulty similar to the opening of Judex. If there is a shortage of quality tenor voices, some (not all) of the altos should be hand-picked to augment them.

With specific regard to the choirs of the Australian Intervarsity Choral Societies Association (AICSA), there are some considerations for the 2007 festival planned to be held in Brisbane. These are that the Gothic would require almost all of the entire 2 week rehearsal schedule, and would thus preclude programming a first week concert of any great difficulty requiring non-trivial amounts of rehearsal time. This would also suggest that if a first week concert is desired, then the length of music requiring the choir must be short, and the repertoire must consist of relatively easy music. An orchestral concert with a small amount of choral participation might be ideal; e.g. “lollipops” such as Borodin’s Polovtsian Dances on the back of other orchestral music. In the words of the chorus master for the recent Melbourne intervarsity: “it has to be fun!”

* The terse summary is something of an in-joke. Brian prefaced the Te Deum with a quote from Goethe’s Faust: “Wer immer strebend sich bemüht, den können wir erlösen”, which might be translated as, “Whomever strives with all his might, that man we can redeem”; but was more loosely paraphrased by a friend of Harold Truscott’s with respect to this symphony as, “If you can stick this one out you deserve a medal.” It occurs to me that that summary might well correspond to anyone who manages to wade throughs this morass of information.

With the final words on this page, I would like to acknowledge the work of Professor Paul Rapoport and the late Harold Truscott (principally for their Two Studies, published by the HBS), and others whose writings on the Gothic have likewise illuminated the work for me: Malcolm Macdonald, David J Brown, and in particular the late Deryck Cooke, whose chapter on choral symphonies (in the Pelican “Choral Music” book edited by Arthur Jacobs) first introduced me to the Gothic, and to the music of Havergal Brian.

© 2004–08 Philip Legge