Discovering Dorothy

by Neil Midkiff

Lyric's Discovery Series uses the advertising tagline "Lost Operettas... Found!"  Our audiences may be asking themselves "How does a musical show get lost?" and "How does a show get found?"  In the case of Dorothy, both these questions have interesting answers.

Most of the standard books on G&S mention Dorothy as a rival production whose great success caused some apprehension and soul-searching by Gilbert, Sullivan, and their producing partner D'Oyly Carte during the already anxious period in which they were trying to figure out how to follow up Ruddigore. At one time, the triumvirate even considered building a new theatre and starting a new company to try to tap into the enthusiasm that London audiences felt for Dorothy. As things turned out, G&S continued in their own inimitable vein of comic operas and did not attempt to compete directly on Dorothy's home turf.

So what is that home turf?  The answer may surprise you, as it surprised me.  Dorothy turns out to be what modern audiences would call a musical comedy, even though that term hadn't yet been invented.

Most standard histories of Broadway-style musical comedy start in the 1920s by identifying two separate strands from which it sprung.  One was the informal revues, vaudevilles, and burlesques of earlier popular shows, with little if any linkage between songs, characters, and plot, but with lots of opportunities to "bring on the girls" for a snappy song and dance.  The other derived from the European tradition of operetta, with more serious stories, more ambitious musical goals, and special emphasis on lavish productions with exotic settings like the royal courts of Mitteleuropa or the Sahara desert.  Usually, Jerome Kern and Oscar Hammerstein II get the credit for "inventing" the modern musical with Show Boat (1927).  The invention, supposedly, consists in combining a coherent and believeable non-exotic story (often based on an existing novel or play) with songs and dances that carry the plot forward rather than interrupting it.  In this oft-told view, the "integrated" musical comedy then lay dormant until Hammerstein teamed with Richard Rodgers for Oklahoma! (1943).

This capsule history leaves out so much that it's hardly credible.  It misses out on the "Princess Theater" shows of the 1910s by Kern, P. G. Wodehouse, and Guy Bolton, which paved the way for Kern's later work by showing that a musical comedy could combine a light popular touch with real continuity between story and musical numbers.  It omits a good deal of Victor Herbert's work, which included some intimate musical plays with modern settings as well as the operettas for which he's best remembered.  And most important for our purposes here, it tells us nothing about the chain which connects Gilbert and Sullivan's style of musical theatre with the roots of Broadway: the British musical comedy tradition usually remembered (if at all) as the Gaiety shows of George Edwardes.  Dorothy may well be the first important link in that chain.

See Marc Kenig's article below for fascinating details on Dorothy's creators Alfred Cellier and B. C. Stephenson and their connections to G&S and the London musical stage.

Cellier had composed most of the music of Dorothy for an unsuccessful operetta, Nell Gwynne, in 1876; though its plot failed to please, the music had been praised for its "early English" character. Much of the music is very similar in style to Sullivan's lighter moments; it's charming and graceful, as befits the intimate nature of the story. A decade later, Stephenson manufactured a new libretto and lyrics to fit the earlier score. He managed to create a well-crafted play, with literary echoes of themes familiar from such works as Oliver Goldsmith's She Stoops to Conquer and Jane Austen's Emma, yet with a fresh treatment and a much more modern diction, suggesting its eighteenth-century setting lightly rather than with the lumberingly archaic literary tone used by many other Victorian authors. Dorothy's libretto is a more conventional farce-comedy than Gilbert's stories, without the logical paradoxes and topsy-turvy settings only Gilbert could pull off.  The play is generally sweet and romantic, but with bits of cynicism, a touch of early feminism, and plenty of good old mistaken-identity and social-class-distinction comic situations.  Although the vocal score labels itself as "a comedy opera," Dorothy has a somewhat larger fraction of spoken text than the comic operas of G&S.  It still is very dependent on its songs to tell the story and to illuminate the characters, so I have to categorize it as truly an "integrated musical comedy" well before its time.

Despite its long initial run (931 performances in London, far outstripping any of the Gilbert & Sullivan shows) and frequent revivals and tours, including an American production starring Lillian Russell, Dorothy's fame eventually faded, like nearly every show of its period that wasn't G&S.  As far as we have been able to discover, American performances since 1900 and British productions since 1940 have been rare.

Finding Dorothy

In comparison to some forgotten musical theatre works, we have reasonably good existing material for Dorothy, and yet bringing it into performing condition has required some fascinating research, a good deal of musical and textual proofreading, and occasional bits of artistic insight.  The two major sources are the piano-vocal score and the libretto, both published by Chappell & Co. in London, neither one dated.  Scans of the score are available on the Internet thanks to a British collector, Adam Cuerden, and the Gilbert & Sullivan Archive.  The libretto came to us from Don Tull's collection.  A third source, the book of lyrics (also Chappell), presumably intended for audiences following along in the theatre, was found in microform at the San Jose State University library.  (Finding a way to read the opaque microcard format proved to be another research exercise, but that's another story altogether.)

The trouble is that these sources apparently represent very different stages in the evolution of the work, and there are some major disagreements.  Internal evidence suggests that the lyric book is the earliest, and in the many cases where it doesn't fit the music of the vocal score, we must regard its version as superseded.  But in some cases, the early lyric book provides solutions to problems presented in the later versions, and reconnecting the original music and lyrics proves very satisfying.

The vocal score is riddled with misprints, many of them obvious engravers' errors and others rather more puzzling.  In Victorian times music was either set in movable type (giving a rather blocky, squared-off look -- those who have seen the Kalmus reprints of the original Sorcerer and Princess Ida scores know what I mean) or, for better quality, the music symbols were physically carved and punched into copper plates with engraving tools, as was done with the Dorothy score.  Errors were very expensive to correct, and most often if a notehead was put on the wrong line of the staff or if a sharp or flat was punched in front of the wrong note of a chord, it stayed there.  This sort of error usually shows up as an unwanted dissonance when playing through the piece, and with a little harmonic analysis, the original intent is usually clear.  With Adobe Photoshop, we can now digitally edit the scanned score images to fix these errors quickly.

Less easy to deal with is a general sloppiness in setting the lyrics.  The occasional misspelling is forgiveable and fixable.  But whoever set the type for the Dorothy music plates seems not to have been paying attention to his job.  In many places, punctuation is omitted or randomly replaced by commas, even at the ends of sentences!  Quite often a word is simply wrong.  For instance, the Court of King's Bench is mentioned many times in one song; in one place, it appears as the Court of Queen's Bench, which would be appropriate for Victorian times but not for the 1740 setting of the story, when George II was on the throne.  Occasionally, a phrase just doesn't make sense; here, we were often able to use a better reading from the lyric book.  Most of the time I found it simpler and clearer to remove the original lyric type entirely (much of which had worn to near-illegibility in reprinting anyway) and to re-typeset the corrected words digitally.

Some of our most satisfying detective work is a little more complex.  In the tenor's first ballad, the last sung phrase of each verse seemed to be awkwardly matched with the accompaniment.  He is shown as singing "...that I have given her * * * my heart" and "and left me with no heart * * * to yield" to end the two verses; the asterisks represent rather awkward rests for him, but in the accompaniment they're musical notes that sound like a natural melodic ornament (D - C# - D - Eb on "her/heart" and the three rests).  When first studying the vocal score, I'd planned to have him sing the word "heart" on the second verse over all four notes.  Then we compared the text as found in the lyric book, and discovered that the first verse originally was "...that I would gladly yield to her my heart."  Aha!  Not only is this evidence that more syllables were originally intended to be sung; it also sets up a nice poetic contrast between the two verses: "gladly yield" versus "no heart to yield."  How the extra words and notes dropped out, I can't explain, but it's a great pleasure to be able to restore them for you to hear this summer.

Another case of matching words with music happens when Dorothy and Lydia make a formal appearance at a dance.  The score has a page of instrumental music for their entrance, with no words; the libretto shows that the musical number is to be played, then after some stage directions about what the young ladies do as they enter, the dialogue continues.  Nowhere is there any indication that the dialogue and music are to overlap, or that the dialogue is in any way poetic.  And yet, if a few inconsequential word changes are made, with only a dozen words in all left out, that "dialogue" scene becomes eight couplets in iambic pentameter, fitting exactly the thirty-two bars of musical underscoring.  Coincidence?  I think it more likely that someone along the way "forgot" that the words and music were supposed to overlap, and when the words were laid out as dialogue instead of poetry, someone later tried to "improve" them, adding a few words which further obscured the match.  I believe you'll agree when you hear it that the restored version makes the scene flow more smoothly.

Following immediately in the score is the Graceful Dance, another purely instrumental piece.  Both the libretto and lyric book call this the Graceful Dance and Song, and give the lyrics to the song; the version in the libretto fits the phrase-lengths of the music exceptionally well, to the extent that there is no doubt at all that what we have is the accompaniment to the original song.  In this case, the original vocal lines are lost, and new ones have had to be written.  Where the melody of the accompaniment seems naturally singable, the voice follows along; in other places where the instrumental line skips around in a non-vocal style, I've tried to restore a singing line that fits it in counterpoint.  The lyrics fit the situation and characters so well that a little musical speculation is worthwhile in order to be able to put the words back into the show.

One brief duet, just a few lines in the libretto for the comic couple, has no music in the score at all.  The words are similar to another love song in the show, and so I've had a bit of fun writing a few seconds of parody of that tune so the actors can sing these lines rather than just declaiming them.

For the non-musical portion of the libretto, we have only the one source, so our editing has been minor.  A few of the more outlandish puns of the original would now require footnotes (and then of course wouldn't be funny) so we've substituted a more comprehensible equivalent, when possible, and omitted them when we couldn't figure out what was meant.  Marc has found evidence that some of these once-topical jokes were interpolations by the original performers, and were strongly disapproved by Stephenson, so our performing version may well be closer to his original.  Rest assured that plenty of jokes remain for your enjoyment!

Dorothy's Degrees Of Separation

by Marc Kenig

A popular film trivia game is to mention the name of any actor and see how easy it is to trace an association with Hollywood actor Kevin Bacon.  In the world of Victorian theater, we can play the same game by relating any notable theatrical personages back to our next Discovery presentation, Dorothy. I'd venture that no other show in the history of theatre has as many interesting connections from its path to the record-breaking initial run of 931 performances, the careers of Dorothy's performers and to our Discovery Series of Lost Operettas.

Let's begin our game with six degrees of separation from the composer. This is easy!

Dorothy's composer, Alfred Cellier, sang with Arthur Sullivan as "boys together" in the choir of the Chapel Royal.  Cellier served as Sullivan's music director for D'Oyly Carte from 1877 to 1883. He conducted every Gilbert and Sullivan operetta from Sorcerer to The Mikado. Cellier, an accomplished composer, routinely wrote the music for curtain raisers for the G&S Savoy Operas, had worked with Gilbert on the 1875 extravaganza Topsyturvydom, and would later set Gilbert's "lozenge plot" to music in The Mountebanks (1892). Alfred Cellier's brother Francois, also an accomplished musician and music director, was to stay with the Savoy organization for more than 20 years.

With a start like this, virtually any connection to Dorothy is possible. Queen Victoria? Easy, she knew Sullivan, Sullivan knew Cellier and Cellier composed Dorothy.

So how about connections to Dorothy's librettist, you may ask?

First, there's an easy Sullivan connection: Benjamin Charles Stephenson was the librettist for The Zoo, composed by Sullivan in 1875 the same year as Trial By Jury. In 1872, Stephenson wrote a libretto for Alfred Cellier's one-act Charity Begins at Home. In 1878, Stephenson worked with Clement Scott on an English version of Lecocq's operetta Le Petit Duc. Clement Scott was to write the lyrics for "Oh Promise Me" by Reginald DeKoven, the composer of last year's Discovery presentation Robin Hood.

And yet, our game of Dorothy's degrees of separation has only just begun!

George Edwardes was the first managing director of London's Savoy Theatre. In 1886, he left the Savoy after accepting an offer from John Holingshead to run and eventually own London's Gaiety Theatre. Holingshead produced Gilbert and Sullivan's first collaboration, Thespis, at the Gaiety in 1871.  Edwardes' very first production at the Gaiety was Dorothy. Less than a decade later, Edwardes became the pre-eminent impresario of London's West End. But in 1886, he was unable to convert the Gaiety audiences from Victorian Burlesque to legitimate operetta. After a few months' initial run at the Gaiety, he sold the Dorothy production to accountant-cum-librettist Henry Leslie for a mere 500 pounds. Leslie extensively show-doctored Dorothy and re-opened it at the Prince of Wales theatre -- the rest of its run became theatre legend.

Marie Tempest (Dorothy) and C. Hayden Coffin (Sherwood) were two performers who were to stay with Dorothy throughout the 931-performance run. Later in their long and illustrious careers they were to work with Rutland Barrington, the original Pooh-Bah among his many Savoy roles, on the legendary Daly's 1890s series of Edwardian Musical Comedies -- produced by none other than George Edwardes. The son of George Grossmith, the original G&S comic baritone, starred in some of Edwardes' popular musicals in the 1900s opposite the actress Gertie Millar. Millar was the wife of Lionel Monckton, composer of the next Discovery production, The Arcadians. Monckton placed many hit songs in Edwardes' Daly's musicals, starring Tempest and Coffin, original stars of the first run of Dorothy. Oh dear -- there are so many connections that I've now gone in a circle!

During the run of Dorothy, producer Leslie was to collaborate with librettist B. C. Stephenson and composer Ivan Caryll on a one-act operetta called Warranted Burglar Proof. A few years later, Caryll was to become the full-time music director of the Gaiety Theatre where Dorothy premiered, and served as the music director for a Dorothy revival in London in 1908. Soon after, Caryll went off to compose shows in New York that competed with the Princess musicals composed by a young composer, Jerome Kern.

Jerome Kern knew Caryll who conducted Dorothy and worked for Edwardes who produced Dorothy. Kern also composed musicals for George Grossmith, Jnr whose father knew Dorothy composer Cellier.

There's one more Dorothy connection close to home I need mention before closing: On the profits from Dorothy's long run, Henry Leslie became so rich he was to build his own theatre in London. Dorothy was to close its record setting run in this new theatre in 1889. He christened it the Lyric Theatre!