Gilbert and Sullivan Archive


General Thoughts about the Opera (Continued)

1.1.9 Sullivan's Autograph
1.1.10 Gilbert's Letter Reconsidered
1.1.11 "Character" Arguments
1.1.12 Conclusions

1.1.9 Sullivan's Autograph

We now turn to the evidence of Sullivan's autograph score. This score was in private hands for most of this century and only came into the Pierpont Morgan Library's hands in 1979, or so. Scholars who examined it must have been astonished to discover that the first part of "Climbing over rocky mountain" is taken from a Thespis copyist score, with the Thespis words crossed out and the Pirates words written in, and with the vocal writing altered from four parts to women only.

If Gilbert's 1902 letter to Percy de Strzelecki is the most difficult piece of evidence for adherents of Original Intent to explain away, the evidence of Sullivan's autograph poses the gravest difficulties for Original Intent, for we are required to believe that SULLIVAN brought a Thespis score with him to New York, even though the decision to use "Climbing over rocky mountain" hadn't yet been made.

When I first learned of the presence of Thespis material in Sullivan's Pirates autograph, I assumed the case for Original Intent was sealed, for I could conceive of no other explanation for a Thespis score being in New York. Arthur Jacobs, by the time of the second edition of his Sullivan biography, was aware of this evidence, but without further discussion pronounced the case for Original Intent "unpersuasive" in a little-noticed footnote

Then, in a Savoynet debate three years ago, Bruce Miller proposed the following novel explanation, which to my knowledge had appeared nowhere else before:

It is not so incredible that Sullivan would have had a score of Thespis with him even if he "had" decided to make the substitution of 'Climbing' prior to the American trip. It should be recalled that, in an earlier scouting trip made by Carte, the impresario had determined that tastes in America were somewhat different from those prevailing in London, and that G & S should tailor any production in America to that difference. In particular he mentioned that Yankees preferred a more 'pronounced' performance. So a revival of Thespis in the United States, perhaps somewhat altered but nevertheless requiring a minimal effort in preparation, is perhaps not so impossible an idea after all, and such a revival not necessarily preceded by one in England.

So, to account for a Thespis score being in New York without having to abandon Happy Accident, Bruce imagines that Sullivan had the score with him in case the partners should decide to mount the opera's first revival in New York. Bolstering this argument is a letter Sullivan wrote to John Hollinshead -undated, but from the context evidently 1879:

My dear John,

You once settled a precedent for me which may just at present be of great importance to me. I asked you for the band parts of the MERRY WIVES OF WINDSOR, for a performance at the Crystal Palace, and you sent them to me and said, 'They are yours, as our run is over, only you will let me have them whenever we want them, won't you?'

I kept them, and some time afterwards, you sent to borrow them for a performance at Manchester. I sent them to you at once, and I think you have them still. Now will you please let me have them, and the parts of Thespis also at once. I am detaining the parts of Pinafore, so that the directors shall not take them away from the Comique tomorrow, and I base my claim on the precedent "you" set....

Proponents of Original Intent will point out that Sullivan and his partners were involved in an imbroglio with the directors of the Comedy Opera Company at this point, and the letter makes no mention of reviving Thespis. But, it is convenient that he is seen asking for the parts back within the year of taking a Thespis score to New York--possibly, as Bruce Miller speculates, for a revival.

Against this, however, are numerous objections. Gilbert and Sullivan's mutual correspondence is very well preserved, as are diaries that Sullivan maintained from 1879 until the end of his life. Gilbert also kept a diary during this period. Through it all, the last mention of a revival of Thespis came in 1875, when D'Oyly Carte contemplated mounting the piece in the wake of Trial by Jury's success. Gilbert and Sullivan insisted on Carte guaranteeing a 100-night run, with 50 nights' compensation to be paid in advance. By November 23, 1875, Gilbert wrote his now famous letter to Sullivan:

I have heard no more about Thespis. It's astonishing how quickly these capitalists dry up under the magic influence of the words 'cash down'.

And indeed, no source has noted any further mention of the opera in either their letters or their diaries, save Sullivan's 1879 letter to Hollingshead. So, if Gilbert and Sullivan contemplated a Thespis revival in New York, they seem never to have written about it.

Moreover, it's clear they "did" bring The Sorcerer with them and a cast prepared to play it, as indicated in Rollins and Witts. In the event, no New York Sorcerer production materialized, but at least we have the evidence that they considered it. With Thespis, there is no such evidence.

Lastly, if Gilbert and Sullivan "did" believe in 1879 that a Thespis revival was still viable, and had brought a score with them for that purpose, why would Sullivan have cut that score into pieces? Admittedly, we know that Sullivan had trouble remembering all of Act I that he had left behind, but he was accustomed to composing in a crunch. Surely, they would have known that re-using part of Thespis virtually precluded a revival of that opera. At the very least, Sullivan would have had an assistant "copy" the section of music he needed, rather than cannibalizing the score of an opera he might yet revive.

The key is likely found in Gilbert's 1906 address to the O.P. club, in which he describes the principles he and Sullivan adopted in the Savoy Operas:

When Sullivan and I began to collaborate, the ladies' dresses suggested that management had gone on the principle of doing a little and doing it well. We set out with the determination to prove that these elements were not essential to the success of humorous opera. We resolved that our plots, however ridiculous, should be coherent, that our dialogue should be void of offence, that, on artistic principles, no man should play a woman's part and no woman a man's. Finally we agreed that no lady of the company should be required to wear a dress that she could not wear with absolute propriety at a private fancy ball.

Thespis, as many authors have shown, was not that kind of opera, but the type that had ruled the boards when "Sullivan and [he] began to collaborate." It was not, in short, a Savoy Opera, and it is easy to see why they never revived it.

So, we must reject the idea that Gilbert and Sullivan were contemplating a Thespis revival at this point, but other explanations are possible. Sullivan may simply have packed a box-full of his prior compositions in a steamer trunk before leaving for New York. After all, he didn't know how long he'd be gone, or what he'd need while there. Thespis could have been brought along with many other manuscripts for no good reason. It might have been packed by mistake: perhaps a servant was assigned to pack Sullivan's Act I Pirates sketches and took the wrong stack of papers by mistake! None of these alternatives strikes me as particularly likely, but they are certainly conceivable.

It is also worth noting that the Thespis ms pages Sullivan bound into the Pirates score go only up to the end of the first chorus; all the remaining material is newly scored. If Sullivan was as hard-pressed for time as the Happy Accident theory tells us, and he had a full Thespis score with him by coincidence, why didn't he paste in the entire number unchanged, as Gilbert's 1902 letter claims he did? The likely answer is that he and Gilbert had decided in advance to change the 1871 version of the number, so Sullivan only brought along the pages he knew he'd need.

To summarize, while it is possible Sullivan "just happened" to have a Thespis score with him, it seems far more likely - especially in context of all the other evidence - that it (or a piece of it) was brought along because the collaborators already knew they were going to use "Climbing over rocky mountain."

1.1.10 Gilbert's Letter Reconsidered

To this point, we have seen that the chronology of Gilbert and Sullivan's New York visit make it highly unlikely, if not plainly impossible, that the inclusion of Thespis music in Pirates was a "happy accident." That we find part of a Thespis score bound into the Pirates autograph further bolsters the case against Happy Accident, as there is no other good reason for a Thespis score being in New York.

Favoring Happy Accident, however, is the best possible evidence: the words of the author himself, in his 1902 letter. Gilbert's memory is slightly faulty about "Climbing over..." having been adopted without alteration of words or music, but two of the basic facts - that Sullivan left his Act I sketches behind, and that he had trouble remembering all he had done - are corroborated by two of the composer's letters to his mother. (I will assume that Gilbert's claim that Thespis had a "fair run" of "seven nights" is a mis-transcription of "seventy nights.")

How, then, can we explain away this letter? It seems difficult, but if the reader accepts the chronology I've provided, then the Happy Accident described in the history books simply cannot have happened, and Gilbert's letter "must" somehow be explained.

It is possible that Gilbert, while "remembering" that Sullivan had left his Act I sketches behind, "forgot" that the decision to include "Climbing over...." was made in advance. Gilbert's memory failed him on other occasions, and he may have forgotten this, too.

My own preferred theory, however, is that Gilbert and Sullivan did not want it widely known that their borrowing from Thespis was pre-planned. On the evidence of Gilbert's 1906 O.P. Club address, it seems the librettist, at any rate, was a bit embarrassed by the earlier opera, a conclusion the forty-year delay in publication reinforces.

Sullivan, later in life, was asked what had happened to Thespis, and he said that its music was re-used in all of the other operas, but this statement when read in context appears to have been uttered tongue-in-cheek. We do know of at least one instance when the composer borrowed from himself and "didn't" want it to be noticed. For his incidental music to The Merry Wives of Windsor, he lifted parts of his earlier ballet L'Ile Enchantee. In a letter to his friend, the critic Joseph Bennett, he said:

All the music is new, but (and this is not necessarily for publication) if you remember a ballet called L'Ile Enchantee which I wrote for the Italian Opera, Covent Garden, many years ago, you will recognize two of the themes....

His position on "Climbing over rocky mountain" could well have been the same. The Pirates of Penzance was described as "A "New" and "Original" Melo-Dramatic Opera," and a discovery to the contrary would have been awkward. I can easily imagine that after Sullivan left his Act I sketches behind, this gave him and Gilbert the explanation they needed should anyone notice the borrowing: they would claim it was done only by force of necessity, rather than by original intent.

When I first proposed this explanation three years ago, Bruce Miller got extremely indignant about my supposed attack on Gilbert's character, as if it were unheard of that a famous person would either mis-recall or deliberately shave the truth of past events:

The reason given for doubting Gilbert's word is that he was covering something up. Sorry, I will not buy into this without proof. If you intend to sully someone's reputation with speculation, be prepared to meet with resistance.

Bruce writes here as if I had accused Gilbert of lying under oath, but the explanation Gilbert provides Mr. de Strzelecki is at worst a minor fib of the sort he, Sullivan and D'Oyly Carte often told, for example, to put potential copyright pirates off the scent when a new opera was in preparation.

And, as I have observed before, Gilbert may simply have forgotten.

1.1.11 "Character" Arguments

I've saved for last what I call "character" arguments, that is, speculations about whether the Happy Accident story is "in character" for Sullivan-whether it rings true or hollow. I didn't present these arguments earlier, because they're based on nothing more than our intuition about what Sullivan would have done in a given situation. Thus, while reinforcing Original Intent, they can do little on their own to prove it.

Before the discovery that parts of a Thespis score are bound into Sullivan's Pirates autograph, I used to read questions of the sort, "How is it possible that Sullivan was able to remember a number he wrote eight ""years"" earlier, but couldn't remember sketches he'd been working on only a few ""weeks"" earlier. The evidence of Sullivan's autograph makes this question moot, of course, but it begs other questions.

Throughout his career, Sullivan composed easily, quickly, and often. I know of only two cases when a number gave him any difficulty: "I have a song to sing, O" in Yeomen, and "The absent-minded beggar," a song to Rudyard Kipling lyrics that he composed for the South Africa war relief effort. Against these cases, we have scores of examples when Sullivan did some of his best work against the most extraordinary time pressures. Even granting the special case of the lost Pirates sketches, it defies our knowledge of his habits that he would resort to borrowing from Thespis because both his memory AND his imagination failed him. Dan Kravetz put the case best, a few years back:

Okay, we're asked to believe that Sullivan sailed off to New York, leaving behind music he'd been working on for Pirates ; he could remember all of it except the entrance of the women's chorus, so he and Gilbert agreed to substitute "Climbing" with a few changes of words and making the mixed chorus an all-female number, the Thespis music being easily retrievable.

Sorry, but I don't buy that story. If Sullivan had trouble remembering what tune he'd thought up for any number in the opera he was composing, all he had to do was look at the words and think up another tune to fit the rhythmic pattern; that was how he did his job as a composer. Or did Gilbert somehow forget the words as well? There was a very good reason to salvage "Climbing" from oblivion - the words were hardly specific to Thespis (other Thespis lyrics are full of Mt. Olympus references and other less transplantable material) and besides, it's such a terrific number, why not use it again? Imagine the New York audience blown out of their seats by all those pretty girls singing and dancing like nobody's business; do you hear spontaneous applause when "Let us gaily tread the measure" returns tutti and fortissimo? Nowhere else can be found such a brilliant expression of uninhibited female joy, either in or out of G&S. Was this something that had to be pasted into Pirates only as a late-stage substitution? "Well, we had something much more clever that we wanted to do in that spot, but Sullivan left his sketches at home...." Please!!

Of course, with Pirates being touted as "entirely new and original," a cover story would be needed in case they got caught being slightly unoriginal. There was no likelihood that anyone in New York would sit up and remark, "Hey! That number was already used in Thespis!" London was a cause for some concern, but the problem didn't really come up until substantial numbers of people were in a position to hold the libretti of both operas in their hands. Then the lost sketches excuse would explain the duplication. Meanwhile, I believe Sullivan recycled much more of his Thespis music to fit new lyrics submitted to him, or other musical projects.

We do know that Sullivan, like most composers, did not hesitate to borrow from himself. He re-used part of Thespis on at least one other occasion - for his 1893 ballet Victoria and Merrie England. Parts of his ballet L'Ile Enchantee were re-used in at least six different works. However, all of these cases fit exactly the circumstance Gilbert attempted to refute in his 1902 letter - that the material was "too good to be lost." I find it most uncharacteristic that Sullivan would resort to self-borrowing because of a failure of both memory and imagination, as Original Intent would have us believe.

1.1.12 Conclusions

Incontrovertible proof that "Climbing over rocky mountain" was intended for Pirates all along awaits evidence we don't yet have: either an as-yet undiscovered source, or a fuller textual comparison of the LC and MD texts. However, every piece of evidence known to us "except" for Gilbert's 1902 letter supports Original Intent. This evidence includes:

  • Textual comparison of the LC, MD, and first-edition libretti.
  • The chronology of December, 1879, when Pirates was completed, rehearsed, and put into production.
  • The chronology of the filing of the license copy.
  • The evidence of Thespis ms. being bound into Sullivan's Pirates autograph.
  • The common-sense unlikelihood of Sullivan being forced to borrow from himself because of a memory lapse, especially since there was no pressure (aside from self-imposed pressure) for the new opera to open on any particular date.

There ARE objections to all of these arguments:

  • Textual comparison yields at best circumstantial evidence. Who's to say withany certainty how radically Gilbert and Sullivan rethought their new opera whilst in New York?
  • Given the unusual time pressures Gilbert and Sullivan faced in December, 1879, guesses about what could have occurred in a given time frame could well be wrong.
  • Granting that the license copy was filed with the Lord Chamberlain before the BOTHNIA arrived from America with the Paignton materials, there could have been an earlier shipment that we don't know about.
  • It may seem strange that a Thespis manuscript was in New York, but life is full of lucky coincidences.
  • It's true that, if forced to do so, Sullivan could have written a new entrance for the women's chorus from scratch, but given that the Thespis material was available, why should it surprise us that he used it?

It is my belief, however, that as historians it is our proper role to examine all the evidence in its totality and draw the best inferences that the facts will allow. New discoveries, of course, force us to rethink our conclusions, but we should not strain for awkward interpretations of the facts because of a single letter Gilbert wrote over twenty years later.

Evidence Michael Walters had available to him in 1979 was sufficient to conclude that "Climbing over rocky mountain" was very likely included in The Pirates of Penzance by Original Intent. Since then, additional evidence has been discovered, all of it favoring, while not absolutely proving, Original Intent. I believe that there are more sources yet to study. But, based on what we know today, I have no difficulty concluding that "Climbing over" was no Happy Accident.


Page created 26 April 1998