Sacred Music
Fall 2006, Volume 133, No. 3, pp. 6-9

The Cult of the Conductor
by Peter Phillips

t is often said that we can learn a lot from studying the opposite of what we believe in. This aphorism may explain a mild obsession I have formed recently for the methods of an American choral conductor, the late Robert Shaw. In fact I know very little about his methods, despite my repeated questioning of people who knew him—he seems not to have written about what he did—and the clichéd position I have quite possibly forced on him may not represent the man at all.

But I needed something to explain the stubbornly old-fashioned mind-set towards performing choral music which I have so often encountered in the US. So persistent has this mind-set been in the teeth of what renaissance music (to mention only one repertoire) seems to me so obviously to ask of its performers that I supposed there must have been an exemplar of some kind, a figurehead, who had held everyone in thrall. Maybe that person was Robert Shaw.

What Shaw has come to represent for me is the continuation almost into the 21st century of the 19th century way of setting up a musical organization, with the conductor as the central/hero figure who ran the show like a creative genius in his own right, and expected to be worshipped. He—it was never a she—controlled everything and his aesthetic judgments were law. His performers were his mouthpiece. Along with this went the idea that the music had to be bent to the will of his dominating personality, which typically would involve his taking possession of his chosen repertoire by imposing interpretative insights onto it by means of carefully considered dynamic markings and other effects. Originally this view of the conductor’s role had devolved from what is expected of an orchestral conductor, master of a vast instrument, which in turn came from Romantic notions of the absolute nature of the truth which proceeds from a creative artist.

One of the results of this way of thinking was to put a heavy accent on rehearsals, since it was only then that precise instructions could be handed down. The corollary was a strong reluctance to leave any live performance to the whim of the moment since this would threaten the stipulated authority of the hero: it was essential that a group feeling should not be permitted to supplant the insight of the conductor. One of the things I have been told about Shaw was that he had such tight control of his performances that he could ensure a piece like Messiah would always run to within the same few seconds, night after night, such was his determination to tame all the variables. I cannot do this with a motet lasting ten minutes, let alone an oratorio lasting anything up to three hours.

It hardly needs me to point out that orchestras playing Romantic symphonies are not engaged in the same kind of musical activity as a chamber choir singing an unaccompanied piece of polyphony. Yet it is clear there are still choirmasters, especially in the US, who do not see that there is much difference. In fact it would be a safe rule of thumb to assume that whatever someone in the Shaw tradition thought was appropriate for polyphony, the opposite will apply in my view. The reason is that polyphony simply does not lend itself to later ways of approaching music. The nearest approximation to the intimate, chamber music style of polyphony in the 19th century repertoire was the string quartet, which I would guess was not in the back of our hero’s mind as he set out on a performance of the Missa Papae Marcelli.

Yet much would have been explained if it had been. Both types of composition present a musical conversation in which all the participants are treated as being equal. To impose an autocratic or dictatorial method of bringing such essentially democratic writing to life is inevitably to threaten it. The danger is that it will not be able to breathe, and without breath it will suffocate.

Furthermore, no self-respecting string quartet would tolerate having a conductor telling its members what to do, just because the spontaneity of the conversation would be impaired if one person tried to control everything. For various reasons polyphony probably does benefit from having someone holding it together, but it is a close-run thing whether it really needs a conductor or not, and many good performances of very complex pieces have been made without one.

Polyphony of its nature requires the participants to listen to each other, and then to react to what they hear in sympathetic dialogue. In practice this means that one contributor proposes an idea (the “point”), another takes it up in imitation, often modifying it, followed by another until a musical argument has been built up. It stands to reason that the aural balance between the parts should not favor any particular one (in later music it became standard to accent the top and bottom of any given texture) and that the success of a performance of such music may well depend on the audibility of all its detail. This clarity will be achieved only if the lines are sung without distorting vibrato.

I would argue that many of the assumptions which underlie so much traditional music-making will not do for polyphony. The participating voices may be trained these days—if they are to survive the schedules the Tallis Scholars undertake they have to know how their voices work—but not in the operatic tradition of the individual above everything. Polyphony is a cooperative effort for everyone involved, and the first responsibility of the singer is to learn to blend with whoever else may be on the same line: this is not a place for the hero mentality.

Obviously this means singing in tune. Once a good blend between all the voices has been achieved, the contributors to each line will need to learn how to listen to the other lines, which should be done in a spirit of respect. The role of the conductor is to act as a kind of aesthetic umpire. His usefulness resides in the fact that he is the only person who can hear the whole texture at once: a singer in the line will not be able to do this nearly so well. The conductor’s first responsibility in rehearsal is not to impose his “interpretation” of the chosen music on a body of people who have not yet been taught how to shape their sound, but to work on the basics of that sound and so create the conditions in which the final performance will be a living event. This way every performance of that piece by that group of people can be different.

This is desirable for a number of reasons. Even four-part polyphony is quite complex when compared with the choruses of many baroque masters, with their harmonic underpinning and neatly bundled-up fugues. Eight-voiced polyphony, such as Crecquillon’s Pater peccavi, Gombert’s Lugebat David Absalon,or Gibbons’s O clap your hands is in a category apart. For one thing it is not possible to prepare a dynamic scheme for music of this kind: there are simply too many notes in too many differing situations for any conductor to be able to maintain an autocratic control over every tiny detail.

Rehearsing such a scheme would take weeks, and the end result would inevitably sound contrived. The desired artlessness in performing polyphony comes from taking risks, and the risks come from not over-rehearsing.

Perhaps it is too easy for me to say that the shorter the rehearsal the better, when I am working with singers who can sight-read anything. Since it is my view that rehearsals are necessary only to make sure that the notes are right, it is likely that Tallis Scholars rehearsals are going to be briefer than those which involve people who have trouble reading music. But the pertinent question is: are people who have trouble reading music ever going to be sufficiently fluent to sing polyphony well?

Opera singers, even of the first rank, may well not be able to read music at all. What they have is a voice of a strongly individual nature and the inclination to learn a set number of bel canto melodies by heart. Such singers are doubly inappropriate for polyphony, partly for the reasons given above and also because a single concert of polyphony may well involve a singer in more notes than many opera stars have to learn in a year (“so many notes, so little time” as a participant on one of my choral courses put it recently).

Again, considering the opposite may be instructive. I recently heard David Willcocks say that amongst all the boys he had directed in his long career not one of them had had a good voice: he had chosen them for their musicianship. This is also the standard which must have prevailed in the Chapel Royal in Tallis’s time, when it was filled with composers of the first order who, incidentally, had to sing.

Any modern choirmaster, at least at the semi-professional level, who finds himself rehearsing people who need to learn their lines by heart and who wants to give concerts of polyphony, should change his selection procedure. In Tallis’s time there was no such thing as a prima donna or primo uomo. The music was written for people who just got on with the job. In London today there are still hundreds of singers (with trained voices) who are still prepared just to get on with the job, which is why the London singing scene routinely produces the most convincing and stylish performances of this repertoire.

Mention of training voices brings me to a difficult place. There is very little evidence of voices having been trained before the invention of opera. I make no apology for thinking my singers need some training in order to survive in the modern world: Tallis and his colleagues did not go on tour as we do and so had no need of the sheer technique we rely upon.

But training should not mean imposing on Renaissance music the styles of later repertoires. This has been a very common solecism, which even now can prove impossible to undo. Many singers instinctively want to turn High Renaissance music (Josquin and his contemporaries tend to fare better) into an adjunct of the Early Baroque and find it extremely hard to unthink what they know about that music.

Of course this is nonsense. Polyphony came out of singing chant, and from good chant singing will come good polyphonic singing, now as then: a blended line which ebbs and flows with the contours of the melody; above all an expressive legato which may have to last for pages. It would be an enlightened singing teacher indeed who took his or her exercises from chant, yet in the training of my singers, for example, such a teacher would be invaluable, if exercises in breath control and good diaphragm support for the voice went along with it.

I do not mean to trivialize what Robert Shaw achieved with his choirs, for all that the very word “choir” will not do for ensembles wishing to perform Renaissance music. Choirs are too big, too formal, too disinclined to give their members the necessary responsibility for their lines. I doubt that a conductor like Shaw, who liked to keep control of the rehearsal situation to the extent of placing the singers’ chairs according to his preferred sight-lines before the rehearsal started, would ever accept that the best performance of a Renaissance masterpiece is one that is going so well of its own accord that he need do nothing except keep the beat.

But it is high time Shaw’s successors took a long look at the way Renaissance music is actually written, thought more about what it expects of its performers, and spent less time trying to fit in with an anachronistic performance pratice.

Peter Phillips is the founder and director of the Tallis Scholars, which has given more than 1000 concerts since 1973, among which a performance in the Sistine Chapel in 1994.