Comparing Brian’s symphonies

Philip Legge

Comparing Havergal Brian’s Symphonies

Philip Legge

Despite the best intentions of musicians and conductors knowledgeable of Havergal Brian’s oeuvre, or the persuasive arguments and financial support offered and given by the Havergal Brian Society and its members, there still is considerable reluctance on the part of orchestras to put any of the symphonies before the public with any great frequency. One need not conjecture any elaborate conspiracy theory to account for this current neglect, as the dearth of performance that Brian’s music received in his own lifetime prevented any of the symphonies from becoming firmly established in the repertoire as symphonic "cannon fodder". Also, the scale (and consequent notoriety) of the Gothic has perhaps been construed by programmers of orchestral concerts and subscription programmes as being representative of the remainder.

So I will not attempt to justify programming Brian’s works on qualitative grounds, as many others have argued that cause with great eloquence, but will take what I hope is a fresh tack by concentrating purely on the logistics of performing the symphonies. As I see it, the most important task that the HBS can undertake to establish Brian’s symphonies as repertoire pieces is the electronic publishing of the most important works; hopefully all of them, before too long. Both James Kelleher and Myer Fredman have touched on the inconvenience of conducting from illegible and haphazardly photocopied facsimiles of Brian manuscripts, and this partly explains Mark Henegar’s comment (in NL 170) that Brian lacks a "champion" conductor, as Bernstein (or I would argue, Bruno Walter) was for Mahler; but Mahler’s symphonies were most definitely typeset and published at large! I suspect that until the works are better represented in print, Brian’s visibility will continue to suffer, and the symphonies will remain little known to most conductors.

The existing published works are few and far between. After the initial philhanthropic burst of printing by Cranz in the 1930s, nothing much appeared to happen in the next three decades, while a torrent of works flowed from Brian’s inkpen and biro. In the 1970s Musica Viva published several of Brian’s manuscript full scores in facsimile but these are now out of print. Manuscripts are useful as original source material, but interpreting Brian’s cramped handwriting is naturally more difficult than reading a professionally-typeset full score. Although most of Brian’s works are available for hire from United Music Publishing, only a few of the symphonies may be purchased. Brian isn’t even considered to be a 20th century composer, judging by the list found at the UMP website, despite him writing the Fantastic Variations in 1907 and revising his 8th Symphony in 1971.

Although there have been performances of all 32 symphonies, the vast majority of the parts must almost certainly be handwritten and possibly little better than Brian’s full scores in legibility. Indeed, some of the current parts were originally provided by Brian himself, and while he was an experienced copyist, he did not always provide optimal page turns or cues, probably owing to haste. A further disincentive is that discrepancies between full score and parts are reportedly commonplace, especially with accidentals, and no orchestra wants to hire materials that require expert preparation to fix - or worse, waste expensive rehearsal time.

Therefore, the Society’s Scores Sub-Committee is to be highly commended for beginning to transcribe the music into what is the state-of-the-art in music publishing today, namely the "Sibelius" format. (To date this includes Symphonies 23 and 24; I am a quarter of the way into the typesetting of Symphony 7.) As my knowledge of the symphonies is incomplete I would not attempt to designate those needing priority attention, though I am aware that the author of the three-volume study of Brian’s symphonies has rated 3, 7, 8, 16, 19, 22, 27, 29 and 30 as the most important to Brian’s reputation as a symphonist. Currently only two of these are listed for sale, and - you guessed it - only in facsimile. UMP should be encouraged to allow the SSC’s editions to be made available for sale as print-on-demand full scores and cheaper-priced study scores, as many other music publishers are now able to do with electronically-typeset music.

So programmers of orchestral concerts, take note! All of the symphonies are available for hire from United Music Publishing: but by programming a symphony and notifying the Havergal Brian Society sufficiently far in advance, there is a reasonably high probability that the score and parts will be typeset by the Society’s team of editors before the work is put before your orchestra, with the savings in time and utility resulting from using performing materials which are legible, clear, and internally consistent.

Lastly, at the moment there seems to be a ridiculous chicken-and-egg situation to be overcome: although some new editions of the symphonies are ready to be used, UMP is not advertising their availability until they have been professionally rehearsed, so that any deficiencies are discovered. It seems to have escaped notice that by not advertising that the works are available, it makes them much less likely to get professional rehearsal!

It should be firmly stated that the majority of the works require a normal enlarged symphony orchestra possessing triple woodwind and a dozen brass players: the same 100-strong orchestra that Strauss or Mahler normally wrote for. One symphony (number 18) requires merely double woodwind and ten brass players; seventeen symphonies (6-8, 11, 17, 19, 21-27, 29-32) require no more than triple woodwind and a dozen brass; and ten symphonies (5, 9, 10, 12-16, 20, 28) range from triple woodwind up to but not exceeding quadruple woodwind and sixteen brass. This includes all but the first four, which I will discuss separately. All 32 are set out in tabular fashion below.

Brian almost always specifies predictable woodwind doublings, which should not prove a problem for any symphony orchestra and are not given in the table as to do so would have unnecessarily complicated it; the footnotes amplify this topic. Brian’s typical brass section includes the usual horn quartet, either three or four trumpets, and more often than not a triad of tenor trombones with a bass tuba, rather than two tenors and a bass trombone; if two tubas are listed, then the second is usually a euphonium in practice.

Brian writes for a sizeable but not unmanageably large percussion group, commensurate with the orchestration and comparable with other twentieth-century repertoire. Of the latter twenty-eight symphonies, the common theme would be one or occasionally two timpanists; the usual battery then consists of cymbals, triangle, bass drum, side drums, tambourine, glockenspiel and xylophone, and on average one other miscellaneous piece of percussion as given below in the footnotes. There is usually a single harp part, and sometimes a part for celesta and/or second harp. A few of these symphonies require organ or piano; Symphony 5 requires a baritone soloist; none of them are choral symphonies.

So the bottom line is, all of the twenty-eight last symphonies can be played by a professional symphony orchestra of a hundred players; they are rewarding, albeit challenging, for orchestras to play; none of them are longer than three-quarters of an hour; and most simply, they contain much good if relatively undiscovered music, so what more should be needed to convince a symphony orchestra to play these symphonies, apart from a willingness to entrust conductors with the exploration of unfamiliar territory?

Without any further ado, here follows a table showing the performing resources required for the symphonies - their year of completion, duration, and the total number of orchestral players required. Only the four early symphonies require more inflated resources, so for a discussion of these, and qualification of any other specific points, please consult the footnotes; the only other item to point out is that "Symphony ½" is actually Part I of Symphony 1, without its choral finale.

Symphony table

Symphonies 1 to 4

Symphony 1 (the Gothic) and the other three early symphonies really require more explanation than can be given here. They are obviously works more appropriate to a festival than the average subscription concert; they last about 50 minutes in duration, aside from the First which is twice that length complete, or about 40 minutes for the orchestral-only Part I, sometimes dubbed the "demi-Gothic". The requirements for the orchestral-only and choral versions may be compared directly by reading the first two columns of the table.

Suffice it to say that the Third scales up as being the only "quintuple woodwind" symphony by Brian, if two piccolos, four flutes, and four oboes doubling two cor anglais (i.e., ten players) are considered to average out as five! The Second is exceptional in needing sixteen French horns in the third movement, and six or eight horns elsewhere; technically it would be very difficult to do with fewer than sixteen players.

Then there are the two massed choral symphonies: the Fourth, "Das Siegeslied" (Song of Victory) with its setting of Luther's Psalm 68 in German, and the Gothic with its vast setting of the Latin hymn Te Deum. Each of the two choral symphonies requires a substantial, well-rehearsed choir for full effect.

The four early symphonies not only require larger numbers of woodwind doublings; the choral symphonies also involve the instrumental rarities, such as alto flute, oboe d'amore, bass oboe, basset horn, and pedal (contrabass) clarinet. Lastly, for want of comparison in describing the orchestra for the Gothic, the on-stage forces may be justly compared with Schönberg's Gurre-lieder - which requires about 140 players; the off-stage bands compare with Berlioz's four brass orchestras in his Grande messe des morts - another 36 players; and for the choral writing, the only suitable comparison in terms of polyphony is Thomas Tallis' 40-part motet Spem in alium, including like Brian, no fewer than 16 unique bass parts!

Winds and Brass
In the symphonies with double, triple, or quadruple woodwind, there will almost always be one of each of the following doublings: piccolo, cor anglais, bass clarinet, and contrabassoon. More often than not, the cor anglais and the bass clarinet are dedicated instruments replacing rather than doubling the corresponding oboe or B flat clarinet, whereas the piccolo and contrabassoon players are expected to swap back to the normal flute and bassoon respectively. The limited number of exceptions to the rule above are symphony 13, which has 2 flutes doubling piccolos, and symphonies 20 and 28, each of which also includes a dedicated E flat clarinet. Brian's latter symphonies do not require Wagner horns, cornets, bass trumpets, or any other unusual brass.

A string ensemble of 60 players (as for Symphony 22, for which Brian specified is given as a default number, though for some symphonies this may be too few (especially the first four) or too many (Symphony 18 in particular). As I have access to only 3 full scores (Symphonies 1, 3, and 7) I am unable to provide much detail on the amount of divisi work required of the string section: 4 denotes that the respective sections (violin 1, violin 2, viola, `cello, and double bass) should each be capable of executing 2-, 3-, and 4-part divisis. Symphonies 2, 3, and 4 appear to have much more demanding divisi work, in basic effect making the normal 5-part string ensemble into a 10-part ensemble by default, which is then subdivided further.

Timpani and Percussion
Likewise, a percussion ensemble of 8 players (separate of timpanists) is assumed as a default, with the exception of the symphonies where an explicit number has been specified by UMP (but it is very likely their enumerations are incorrect). A thorough analysis of the score of each symphony would be required to calculate the minimum number of players that are actually needed to play every instrument in combination in a given work, and without a complete collection of full scores it is rather futile to speculate what these minima are likely to be. The author would be delighted to hear from you if you can provide a precise specification of the minimum number of players and their disposition, for any of the symphonies. This is not obtainable by simply adding up the number of percussion instruments. Similarly, it would be useful to know precisely how many drums are needed by timpanists, particularly in works where more than one player is required.

The percussion row entitled "Other" includes all the infrequently-used percussion instruments to be found in the symphonies: Low-tuned bells in Symphonies 1 (bells in C, D, and E); 2 (in F); 4 (in C); and 7 (in C and E)
Chains, birdscare (football rattle), and (African) long drum in Symphony 1
Thunder machine (definitely not a thunder sheet) in Symphonies 1 and 10. The thunder machine is a rotating barrel containing large metal balls; the tinny sounding thunder sheet is never to be substituted for it. In Symphony 10, Brian specifies a second bass drum should be used in the absence of the thunder machine.
Tenor drum in Symphony 3
Wind machine in Symphony 10
Sleigh bells in Symphony 11
Vibraphone in Symphonies 21 and 28
Jingles in Symphony 25
Wood-block in Symphony 26
N.B. The second pair of cymbals described in Symphonies 14-17, 20, 22, 23, and 28 is actually a suspended cymbal.

The question of the side drum
One vexed question regarding the size of the percussion ensemble is directly related to Brian's love of the side drum. Although Brian typically writes a single side-drum part, from symphony 6 onward he also specifies that three drums (of the same size) should play the part in unison, as the massed ensemble sound is preferable to a single drum, which he said sounds "like hitting paper". Later on he strongly suggested that this specification should also apply to the symphonies written prior to that time (certainly the choral symphonies). The sound of the ensemble needs to be carefully judged when interpreting dynamics and balancing with the rest of the orchestra; a written fortissimo indicates that all three players are to play loud enough that the total sound is fortissimo, rather than each individual player reaching fortissimo on their drum.

The problem of economy arises with Brian's tuttis, involving most of the heavier percussion: should the orchestra obey Brian's injunction of three side drums, at the cost of hiring two extra percussionists, when in such a thick texture it would be possible to disguise the deficiency of having only one player execute an actual fortissimo, rather than the desired averaged fortissimo of three? There is no good answer to this question.

NL174 © Philip Legge 2004

Newsletter, NL 174