Performing Brian’s music

James Kelleher

Some practical problems of performing Brian’s music - James Kelleher … with reference to Pantalon and Columbine and Symphony 16

Having just conducted two works by Havergal Brian (Pantalon and Columbine, and Symphony 16) I feel able to comment on some of the difficulties of performing his music. The least of these is the question of logistics - starting with the difficulty in obtaining copies of scores and parts. In the case of the symphony, the problem was compounded by an almost illegible score being photocopied in a haphazard manner, which left me unable to see where most of the bar lines were - not too much of a problem until the metre changed without a new time signature being visible…

These problems aside, one arrives at issues of interpretation. Brian’s music is well-known for its abrupt changes of direction, but unlike other composers’ use of this technique, Brian intends these not as surprises or humorous points, but rather as integral parts of the music in the way a carefully crafted transition would serve another composer. This was less the case in Pantalon, since there are elements of surprise in that piece, such as the manic accelerando which breaks off to return to the slow opening bass line.

In the symphony, however, it becomes a real issue, aided and abetted by Brian’s metronome marks which often "feel" wrong. Here one is faced with the thorny problem usually reserved for music from the early 19th century: whether to accept the composer’s metronome marks as gospel or to choose tempi that "feel" right, even though they may contradict the composer’s apparent wishes. The Lyrita version I heard on LP takes the latter course, while I chose to heed Brian’s marks except where they were quite definitely wrong - the one exception I made was to take the opening far slower than the marked ‘crotchet = 70’, which would entail the woodwind phrases being almost presto, rather than the stated lento! The central slow section, also marked ‘crotchet = 70’, felt right as marked, although the initial bars are reminiscent of the opening of the symphony, and it could be argued the two speeds ought to be the same. However, most of this later section would be too stodgy at the same measured pace of the opening.

A notable instance where the written speed is totally wrong is the return of the allegro fugal section after a slower central section. It is very obviously at the same speed as the first Allegro, but Brian marks it ‘crotchet = previous minim’, which (because of the interim slower section) would make the new Allegro ‘crotchet = 70’ (a recurring theme!!!), which is almost Adagio. And what to make of Brian’s two uses of ‘Maestoso’ in the 16th? Taking the maestoso at figure 22 to mean slow after the preceding scrabbling climax, one then encounters ‘Maestoso ma largo’ later on: ‘Majestic but slow’; this time it follows an Adagio. Does this ‘but slow’ imply the other ‘Maestoso’ is only a character marking and not a tempo indication? This would result in the first passage being impossibly rushed so I ignored these implications and chose to regard it as a tempo change.

Similarly, the final three bars are suddenly marked LARGAMENTE (Brian’s capitalisation). Does this mean a slowing up of essentially the same tempo or a totally new tempo? Having already acquired some feel for Brian’s technique, I chose the latter course, supported I believe by the supreme effect of the slow scale acting as a climax for the entire symphony. (What a stroke of genius for Brian to orchestrate that last chord with a trombone blazing away in its top register! Interestingly, the other piece we premiered in the concert, Reinhold Portisch’s Ritual symphony, also ended on a single crotchet fff chord, but being more traditionally orchestrated it lacks half the punch of the Brian).

Then, of course, there is the question of dynamics and balance. Almost the only markings are forte and crescendo. Following these literally would result in a level of ffff being reached after only a few minutes, [so] it is necessary to build in some (unwritten) returns to quieter sounds. And the very feature that marks Brian’s music out from many contemporaries, namely its intense polyphony, presents the greatest problem of performance - balance. What works well in the organ works of Bach or the equal voiced Renaissance choral motets does not work at all easily when huge symphonic forces are employed.

Clarity of individual lines requires a certain sparseness of texture - but Brian’s music is at its least polyphonic in its sparser moments. The nightmare of trying to make the aural results match what I saw in the score and beard in my ‘inner ear’ remains vivid even now, some six months after the performance. And just when you think you are winning, the tuba enters ff cresc in its lowest register with a rapid semiquaver passage.

Despite these difficulties, however, I will continue to attempt to champion the music of Havergal Brian. Not only is the music of a quality that deserves a regular place in the concert halls, but it is also immensely satisfying to be sailing in largely uncharted waters. The more I do, the more it grows on me; and I’ve only just begun…

© James Kelleher 1994 / NL116

"pantaloncolumbine.htm">PANTALON . ENGLISH SUITE 1

Newsletter, NL 116