(text of a paper presented at the Buxton International Gilbert and Sullivan Festival on August 12th, 2000.)
What is the magic of Gilbert and Sullivan? Tricky. I imagine there are as many different answers to that question as there are G&S enthusiasts. Well, here's my answer, for what it's worth.
I take my seat in a theatre, usually a rather uncomfortable seat, and I spend ten minutes reading the programme from cover to cover. The lights dim, and the chattering of the audience quietens. The conductor's arms lift, and a quiet theme is played on the strings. It's the opening of the overture to Iolanthe. I feel a little shiver run through me that I can't quite explain. And that's the magic of Gilbert and Sullivan.
Or take another example. The curtain goes up on a badly-painted Arcadian glade, populated by a middle-aged chorus of women with little wings tied to their backs. They sing rather well but they dance rather badly, and their acting is something that doesn't bear thinking about. And yet, give me ten minutes and I have accepted them as fairies, and I'm as engrossed in the world of the opera as a wide-eyed child. And that's the magic of Gilbert and Sullivan, too.
I'm almost afraid to pursue the question any further. The whole point about magic is that it can't be explained, and anyone who makes a serious attempt to do so is about as welcome as the kind of person who tells you how a magician does his tricks. Explain the magic and you destroy it. I could, in theory, dissect the different techniques Gilbert uses in the operas, taking you through every aspect of his dramatic craftsmanship, stylistic skill and so on, till everything was analysed to death and everyone was asleep. And I'm sure Stephen Turnbull could go slowly and painstakingly through every chord progression in Sullivan and explain how each note is designed to produce a certain effect. But something tells me neither of us wants to do anything of the sort. (And a sigh of relief goes up all through the audience.)
But there is one aspect of Gilbert and Sullivan's "magic" that I think could do with a little analysis - I mean the nature of the collaboration itself. All too often it's simply assumed to have been some sort of mysterious and unexplainable marriage, the combination of two base metals producing gold. A.H. Godwin wrote in 1926 that neither Gilbert nor Sullivan "was a genius himself, but the association of the two, a gifted dramatist with a gifted composer, gave the world an art-form that has undeniable genius." … Well, up to a point. I can certainly agree that the operas are on the whole the best work of Gilbert, and (though I'm no Sullivan expert) the best of Sullivan too. But that doesn't mean that there was something mysterious or unexplainable about the collaboration. At any rate, I hope not, because I aim to explain something about the collaboration in the next ten minutes or so. I know quite a lot about Gilbert, and not so much about Sullivan, so I hope you will forgive me if I seem to display an outrageous bias towards Gilbert in what I'm about to say, or if my comments on Sullivan's music seem amateurish and nonsensical. I'm trying my best.
The first and most important point is that W.S. Gilbert was by far the best dramatist of his generation, just as Sullivan was the best composer. Gilbert had a very distinct style, ironic and witty and, yes, topsy-turvy, and this was expressed as clearly in his serious plays as in his comic ones. Most people haven't heard of his other great plays - An Old Score and Charity and Engaged and A Sensation Novel, to name a random handful - but we shouldn't be too ready to let unfamiliarity breed contempt. They were as distinctive an expression of Gilbert's art as Iolanthe or The Yeomen of the Guard. And by the time he had properly started to collaborate with Sullivan he was a dramatist of some reputation - original, quirky, witty, sometimes outrageous. So when Gilbert first worked with Sullivan it was a partnership of the two leading names in their fields. The result might have succeeded or it might have failed, but no one could ever have doubted it would have been something worth seeing.
When they worked on Thespis in 1871 Gilbert was already a good lyricist; by the time of Trial by Jury in 1875 he was well on the way to being a great one. The first thing that Sullivan was able to do for Gilbert was to set the words clearly, so that they could be easily understood, and the first thing Gilbert was able to do for Sullivan was to provide words that were worth hearing. Gilbert brought with him wit, intelligence, mastery of rhyme and metre, and a sure sense of theatre. Sullivan brought with him tunefulness, emotional warmth, and the self-discipline to know that there are times when the music must seem less important than the words. These were the qualities that fit together so well and produced the works that we love so much today. Gilbert's works are very strong on wit and intelligence, and not so strong on emotion, what he called the "mysterious quality called 'sympathy'"; and that's where Sullivan could help him out. Sullivan's music… here I can only say what I think, on the understanding that I know very little about music… Sullivan's music on its own seems to me if anything too good-humoured. It needed an acerbic quality to react against, and that was where Gilbert could help. That, if you want, is the magic quality in Gilbert and Sullivan, but there's nothing mysterious about it.
Here I must say something that some of you may disagree with. But I think it's true. Gilbert was a complete man of the theatre. He wrote and directed his plays, and as far as he could he insisted that what appeared on stage conformed to his personal vision. He coached the actors in movements, gestures, inflections; he supervised the creation of the set and of the costumes, and in some cases he even designed the costumes himself. I rather suspect that if he had been able to write music, he wouldn't have wanted to turn to anyone else, even Sullivan, to do it for him. But it so happened that he had no ear for music, as he very happily admitted, and he had to get Sullivan to create it for him. That was Sullivan's little niche in the creation of the operas, and Gilbert respected that. But the music was just one part of the greater whole, the theatrical whole that Gilbert brought together. When the Savoy Theatre was built, Gilbert effectively had his own little kingdom; he was as absolutely in control of the production of his dramas as any writer for the theatre has ever been. We may think that Sullivan should have resented being subordinated to his librettist, and in fact we know that he did resent it, in a quiet way. But it is simple fact that Gilbert was the dominant partner of the two. The Savoy Operas are librettist's operas. They are a very bold experiment in theatre; they created a kind of theatre that was fundamentally different from anything that had gone before - musical theatre, in which the words were at least as important as the music, and conveying a view of life that worked on several levels, intellectual as well as emotional. You may disapprove of Gilbert having dominated Sullivan in this way, but this was the arrangement that produced the Gilbert and Sullivan operas that we all love; we can't deny it. If we're going to understand the way these two men worked together, we must understand this much..
Everything I have just said is quite true, but I also realise it's slightly unfair. Sullivan's music for the operas is wonderful, and creates an atmosphere which is separate from anything Gilbert could have created on his own. Sullivan was an artist, and the operas are also his achievement. But he was always working within the framework created by Gilbert.
Now I want to spend a little time to take a closer look at the process by which the operas were created. I'm going to take as my basic text an article by Gilbert entitled "The Story of a Stage Play" - an account of the writing of The Mikado. It appeared in the New-York Daily Tribune on August 9th, 1885. Now, before I begin, I ought to explain something. I refer to the article as being by Gilbert, even though I am very aware that it is usually referred to by writers on G&S as an interview with Gilbert. It isn't written in what we would think of as interview form today - it is written exactly as if Gilbert had sat down and penned it himself. It's even signed with his name at the end. I am told this was not an unusual form for interview articles at the time. However, the content is stylistically so like Gilbert's journalistic prose that I must wonder. Also, the article quotes extracts from Gilbert's plot books in a way that doesn't seem consistent with an interview situation. For these reasons I'm inclined to think that Gilbert may have written it himself. In any case, even if it was in fact written by an interviewer, it seems to me so near Gilbert as to make no significant difference.
So. Gilbert writes: "In May, 1884, it became necessary to decide upon a subject for the next Savoy opera. A Japanese executioner's sword hanging on the wall of my library […] suggested the broad idea upon which the libretto is based." Well, anyone who has read the biographies or seen Topsy-Turvy will realise that Gilbert has left out rather a lot here. In fact it was in March of that year, not May, that Carte let Gilbert and Sullivan know that they should be thinking about the next Savoy opera. Gilbert responded by proposing a story in which a magic lozenge would transform the characters into the people they were pretending to be. Sullivan didn't like the plot, but apparently made the mistake of trying to be diplomatic about it and left Gilbert with the impression that the problem lay in the details rather than the overall conception. The argument seemed to threaten an end to the collaboration, but in the end Gilbert saw that the lozenge plot had to be put to one side, and it was only then - yes, in May - that this other plot set in Japan came into the picture.
This is obviously a rich episode for anyone wanting to examine the working relationship of Gilbert with Sullivan. Gilbert proposed a plot, and assumed Sullivan would accept it. But Sullivan thought the plot was too mechanical and simply not human. At first Gilbert stood stubborn, arguing that Sullivan had come with his objections too late and Gilbert was now well into the actual writing of the piece. But gradually both men cooled down. At first Sullivan had as good as said he could not collaborate with Gilbert again, but it soon resolved itself into an objection to this particular plot. Sullivan insisted again and again that he enjoyed working with Gilbert, and that he did not want a professional disagreement to affect their friendship. (And they were friends, you know.) We may guess that at last Gilbert saw that nothing would be gained from his insisting on the lozenge plot, and he backed down.
The letters that passed between them are quoted extensively in the Arthur Jacobs biography of Sullivan. It's a prickly exchange, but it does show them bargaining and compromising and giving way - that is, collaborating. In the end Sullivan did actually have an influence on the nature of the libretto Gilbert submitted to him - if only in the most general terms. The libretto would be human, and would not have any supernatural or mechanical elements. And that is what Gilbert gave him. (We tend to assume that Gilbert could not write "human" stories, but in a later dispute Sullivan named HMS Pinafore and Patience as examples of the kind of "human" story he enjoyed working on. Another example he gave was The Mikado.) The most surprising aspect of this whole episode is that in all their years of collaboration - eight previous operas! - a dispute like this had never arisen. So much is clear from the fact that they didn't know how to react when it happened now. They had to work that out as they went along. And that fact in itself tells us how smooth their working relationship had been before.
But to continue with Gilbert's account. As he tells it, the whole of this next stage took place on his own. He's decided on a Japanese opera, because, as he says, it "would afford opportunities for picturesque scenery and costumes". Picturesque. Here we must remind ourselves that Gilbert was an artist as well as a writer. The operas are as memorable for their images as their words or music - we think of the March of the Peers, and isn't it true that we think, not only of the Tantantaras and Tzing-Booms, but of the slow procession of men in their finery, dignified and ridiculous? Think of the Dragoons coming on dressed as Aesthetes; think of Private Willis sprouting wings, or the Headsman left alone at the end of Act One of Yeomen, or Jack Point collapsing at the end of Act Two; think of the Ruddigore portraits stepping out of their frames. Gilbert was born to write for the theatre, where word and image are unified: he could never have been a novelist. His one lasting achievement outside the theatre was the creation of the Bab Ballads - which are remembered as much for the strange caricatures Gilbert drew to accompany them as for the quirky rhymes themselves. In the Savoy Operas, Gilbert was dramatist and lyricist and director and even to some extent designer - he drew on almost everything he could do best to go towards the operas' creation.
Anyhow. The next question he asks himself is: is it practical? Can the Japanese dresses be found? What will the ladies' chorus look like in black wigs? He considers, and decides it can be done, and only then does he sit down and fight with the unruly ideas in his head and force them into some sort of order. He decides on the two act settings, and he assigns his repertory team with the various roles. Rutland Barrington was a problem in this case - his role seemed contrived and, worse, not funny - until Gilbert hit on the idea of making family pride the keynote of the character., and Pooh-Bah was born. In an early version of the plot Nanki-Poo was to have been "an Agent in Advance to a strolling theatrical company", but Gilbert decided against this because "it is proverbially dangerous to introduce theatrical topics into a dramatic composition". (He had obviously forgotten this wise bit of advice by the time of The Grand Duke.) Nanki-Poo was also originally in love with Pitti-Sing, not Yum-Yum, but Gilbert was persuaded against this by Sullivan, who to quote Gilbert "had professional reasons for insisting that a tenor shall always fall in love with a soprano". That little aside hints at a closer collaboration with Sullivan at this early stage than Gilbert gives as an overall impression.
And so, gradually, Gilbert beats the story into shape. When he's satisfied with the plot outline he goes and meets Sullivan and reads it to him, and in a long pow-wow over three or four hours they discuss it, and Sullivan makes various suggestions "bearing chiefly on the musical situations". Gilbert goes away and makes out one final draft of the plot, and sets to work on the lyrics. He writes them first, and while Sullivan is setting them he writes the dialogue, so that the rough script and music are actually completed more or less simultaneously!
How much give and take was there between the two men at this stage - the major part of the collaboration? There is evidence that Gilbert was very willing to tweak his lyrics if Sullivan asked - for instance Arthur Jacobs quotes a letter from Gilbert detailing revisions to Yeomen lyrics. However, on the whole it seems to have been a simple case of Gilbert writing the lyrics and Sullivan setting them, with anything further being a special case. Cuts, revisions, and so on did occur, but for the most part later, at the rehearsal stage.
Yes, rehearsals. This was where the real collaboration took place, in the Savoy Theatre itself, as dramatist and composer met up with the performers and tested the practical effect of the things they had imagined and written down in their respective studies. For the first two weeks Sullivan and Cellier would conduct purely musical rehearsals for the principals and chorus, while Gilbert put on his director's hat and started preparing himself for his new role. By his own account, part of this time was spent "partly in getting the rhythm of the musical numbers into his very unmusical head", but he also had questions of scenery and costume to resolve, not to mention the working out of details of stage movement on his celebrated miniature stage with the little blocks of wood.
Only then would Gilbert's stage rehearsals begin, over four weeks. This is not very clear from Gilbert's account, but Sullivan would attend as many of these as he could - naturally. As Gilbert says, he had "to see that the proposed "business" is not inconsistent with his musical effects"; and "a certain amount of rearrangement" usually resulted. Gilbert adds that Sullivan "is the most self-sacrificing and unselfish of composers, but even his good-nature is not proof against an arrangement whereby the chorus dance a wild jig during an elaborate cadenza or an unaccompanied quartet. But when a composer works with a librettist who is deaf, dumb and blind on all musical points, he is not unprepared for professional solecisms of this description."
But this seems to have been a comparatively minor issue. The real battle-ground was the question of the cutting or altering of the songs themselves. Gilbert notes: "it is generally found necessary to make certain musical alterations - songs have to be lengthened or reduced, and perhaps one or two musical numbers have to be added or excised." For instance, it seems clear from pages of Gilbert's prompt book for Iolanthe, reproduced in Leslie Baily's Gilbert and Sullivan Book, that a passage beginning "On you they'd set/A coronet" was set by Sullivan and survived into rehearsal before being cut. The Yeomen of the Guard had a particularly lively time of it in rehearsal; Sullivan had objections to the structure of Act 2 which riled Gilbert, though in the end he made cuts and changes. Late on, the song "A Laughing Boy but Yesterday" was cut because Gilbert was worried about the opera's pacing. It's now often restored. Also Gilbert objected to Sullivan's first two attempts at "Is Life a Boon?" The version familiar to us is actually Sullivan's third. One of the earlier settings survives, and in some ways I prefer it - it serves the lyric better. So during rehearsal the two men would chop and change and bargain and argue, yield a point or fight its corner, both serving the common cause of creating a good opera but each trying to protect his own contribution to it, because each was a supreme artist in his own field.
Some of you may have heard of a discovery which was recently made by the Sullivan scholars Helga Perry and Bruce Miller, of the first violin part of a "lost" song from Iolanthe. This was the satirical song "De Belville was regarded as the Crichton of his age". The odd thing about it is that on the first night it was not sung but recited without music. It was cut from performance not long after. The question has to be asked, if Sullivan set the song to music, why was the music not used? In the absence of evidence we can only speculate, but the only explanation I can think of would go something like this: Sullivan set the song, but wasn't satisfied with the result. He asked Gilbert that it be cut. Gilbert was proud of the lyric, and refused. (And he was proud of it: he published it in the volume Songs of a Savoyard in 1890, alongside other, more famous Savoy lyrics.) Maybe Gilbert asked Sullivan to try and set it again, and Sullivan couldn't write anything better. At any rate, we can see how, in the absence of a better solution, they might have thought a performance without music would be a reasonable compromise… at first, anyhow.
One final example…. In his article Gilbert recounts the famous episode of the cutting of the Mikado's song, as memorably portrayed in Topsy-Turvy. Gilbert, unlike the film, emphasises that he consulted Sullivan (and Temple!) before announcing the cut - it was an agreed decision. Well, perhaps Sullivan didn't want it to be cut really; perhaps he did. One thing that comes out of all I've just said is that the two men did collaborate, in the sense of accommodating each other - but within strict limits. They could give way to what the other wanted - not always, and as the years went on they became rather more defensive in their attitudes to each other, but yes, of course they worked for the common good of the operas they were producing. As one individual? Scarcely. But as two very individual individuals, bringing different attitudes to the same task and sparking off each other to produce something that no one person could ever have produced.
And here I must confess a change of opinion…. Maybe there is something just a bit magical about the collaboration, after all!