Archive for February, 2017

New Edition of Yeomen from Oxford University Press

Tuesday, February 7th, 2017

The Yeomen of the Guard
Edited by Colin Jagger
Consultant Editor: David Russell Hulme

Full Score: Oxford: Oxford University Press, 2016; xxxvi + 420pp.
Vocal Score: Oxford: Oxford University Press, 2016; vi + 210 pp.

If this review is destined for a “too long; didn’t read” folder, let me begin with the Cliff’s Notes version: this is a terrific new edition of The Yeomen of the Guard.

Once upon a time, Oxford University Press (OUP) promised a multi-volume critical edition of the G&S operas. Ruddigore came out in 2000 (ed., David Russell Hulme). But then, OUP got cold feet — scared off by cheaper and less carefully-edited Dover editions of PinaforePirates, and Mikado. After many dormant years, their original plan for Yeomen was to publish only the vocal score, and make the full score available for hire. Happily, they changed their minds. I would not be nearly as pleased with OUP’s vocal score, if I did not have the benefit of the editorial rationale and commentary found only in the full score.

Colin Jagger has researched the sources thoroughly and documented his findings in over thirty pages of commentary. His approach is sensible, and in tune with the most recent scholarship. The score is beautifully presented, and practically free of mechanical errors. None of the few outright mistakes that I noted were serious: the occasional missing stem, bracket, caption, or the like.

You will not go far astray if you accept the OUP text as it stands, but the real value of the edition is to study the commentary, and draw your own conclusions. Like most editors, Jagger is more convinced of the correctness of his text than he ought to be. But where he makes a musical decision that is at all arguable, he always explains himself. You may disagree, but most of his choices are at least reasonable. I might classify a few as blunders, but none are very significant; there are many others where he’s not clearly wrong, but I’d make a different choice.

In hundreds of places, large and small, the text differs from other editions. This is to be expected, when an editor works from early manuscript sources, and takes no reading for granted merely because it is “traditional.” Due to these differences, I would not recommend carrying the OUP score into rehearsal, if most of your colleagues are carrying Chappell or Schirmer. A musical director should either mandate OUP for everyone, or no one.

I am going to dive into detail, and at times I will be very picky. If I seem disappointed in any aspect of what Jagger has achieved, please remember the Cliff’s Notes version: “this is a terrific new edition of The Yeomen of the Guard.”

At the front of the full score, there is a preface (2½  pages), a description of sources (3 pages), a section on editorial method credited to “David Russell Hulme with Colin Jagger” (4½ pages), commentary on individual numbers (12½ pages), and five appendices (7 pages). The first three of the appendices contain a potpourri of source descriptions and commentary that really belong in earlier sections—more on this later. The remaining appendices contain alternative versions of dialogue and lyrics from an early pre-production prompt book. There’s an even larger cache of unused lyrics in the Gilbert Papers, and these are not included.

The list of sources seems exhaustive, including Sullivan’s autograph score (Aut); two early manuscript full scores and a set of manuscript band parts copied in America; an early full score published in Germany; the D’Oyly Carte engraved band parts from the 1920s; three states of the Chappell vocal score; surviving Gilbert manuscripts for most of the lyrics, including some not published; the license copy of the libretto; seven states of the Chappell libretto; five early prompt books; and modern editions, such as those by Schirmer, Kalmus, and Eulenberg.

The editions, impressions, issues, and states of the published libretti and vocal scores have never been comprehensively identified. Two copies may seem identical, but may differ in subtle ways that can only be detected by comparing them page by page, letter by letter. I have been researching them for decades, and I still discover new ones (“new to me”) periodically. It is devilishly difficult to be sure you’ve found them all.

Unfortunately, the three substantively different states of the Chappell vocal score that Jagger identifies as Ch1Ch2, and Ch3, should really be four. It appears he consulted the first, second, and fourth, and did not realize there was a third. In the discussion below, I shall refer to the missing intermediate source as “Ch2½”. Even this is not really the full story. The first impression of Ch1 was published with an errata list—mentioned in the edition, but not itemized. These errors were corrected in a second impression, and it is this source that the edition polls.

The states of the libretto are numbered L1 to L7, but only L3 is polled consistently—this being the final issue of the original run of performances. The individual issues and impressions are not described in any detail. If you had a Chappell Yeomen libretto in your possession, you would have a tough time figuring out which of Jagger’s seven it corresponded to—if any.

The edition owes a clear debt to the Gilbert & Sullivan critical editions that have gone before. The only credit goes to the OUP Ruddigore, whose discussion of Sullivan’s working methods is incorporated by reference: this is essential, if you want to fully understand the editorial rationale. There is no bibliography, and there are no citations to secondary sources. Where Jagger quotes a non-textual source, such as a letter or diary entry, he does not provide an attribution.

The section on editorial method that is attributed primarily to Hulme is in a noticeably different style, and is the best written; Hulme has been working in the field for decades, and his experience shows. He may assume too much of the reader. He describes the edition’s text as “eclectic,” a term from the broader field of textual criticism: it means that the main text is derived from multiple principal sources, not just one. Some may be puzzled by Hulme’s mention of “accidentals,” which does not mean sharps, flats, and naturals; but rather, punctuation, word division, hyphenation, capitalization, and other formalities of the literary presentation of the libretto.

This edition brings together, for the first time, all of the major deleted passages from the opera: Wilfred’s song, “When jealous torments rack my soul”; Meryll’s song, “A laughing boy but yesterday”; the third and fourth Yeomen’s couplets in the Act II finale; and Sullivan’s original 6/8 setting of Fairfax’s Act I song, “Is life a boon?”

OUP’s Ruddigore relegated all of the deleted material to appendices, forcing the performer to jump back and forth between the middle and the end of the book. This is a serious inconvenience, bearing in mind that the inclusion of these passages is one of the edition’s main attractions. Yeomen mostly avoids that mistake. Only the earlier setting of “Is life a boon?” is confined to appendix purgatory—and this is the least likely of these passages to make it into a stage performance. Oddly, the band parts place both versions of “Is life a boon?” in continuity sequence—granting the players a convenience denied the conductor.

The three lost songs present an editorial challenge. They are not in Aut. The primary source is an American copyist’s manuscript, which was made at the time of the New York premiere. A second copyist’s manuscript, made years later, includes Wilfred’s and Meryll’s songs, and on a separate sheet gives Horn parts for the former, which are curiously lacking in both manuscripts; these parts don’t quite agree with the rest of the number. There are other discrepancies, both in words and music, which Jagger resolves as best he can.

There is a curious problem concerning the cue line to Sergeant Meryll’s song: “He’s a brave fellow, and bravest among brave fellows, and yet it seems but yesterday that he robbed the Lieutenant’s orchard.” Both this line and the song were late additions to the opera. When the song was cut, so was the cue line. However, the cue line makes a re-appearance in the 1890 prompt book and later issues of the libretto. Jagger deems it “clear”—on what basis is beyond me—that the re-introduction of the cue line was an error, and the main text advises that it should be performed only when the song is. One is, of course, free to disregard this advice.

The editorial explanation for these numbers is a bit difficult to follow. There is a brief description of the American sources in the “Sources” section. Longer descriptions, repeating much of the same material, occur in Appendix B. Most of the musical commentary is given in continuity sequence, but portions of it are relegated to Appendix B, as well. This is an extremely unhelpful arrangement.

A variety of justifications is given for the inclusions of these passages. According to Jagger, Wilfred’s song is “possibly the finest of all the ‘lost songs’”; it “plays an important musical role in the continuity of the piece” (although that role is not explained), and “it is well worth including….” Sergeant Meryll’s song is “accomplished though unremarkable.” The earlier version of “Is life a boon?” is “perhaps overly florid, but in many ways…a fascinating song, with the second verse beginning in a different key and with an interesting ending uniting the two verses and keys.”

The third and fourth Yeomen’s couplets are “retained here for reference,” and appear in a smaller print. The use of smaller print occurs in a couple of other places, too: an alternative set of words (appearing only in Aut) at the end of Elsie’s Act I song; and the “nestling near…dropped a tear” lyrics in the Act II finale.

For “Rapture, rapture,” Sullivan wrote three different codas. A ten-bar coda was never published. In the full score, this appears immediately after the main number; in the vocal score, it is relegated to a second appendix, a truly bizarre decision. As usually heard today, there is a seven-bar coda that modulates into the key of the Act II finale; Jagger suspects this was intended “to deter applause.” (In my experience, that usually doesn’t work.)  Previously published vocal scores have the revised coda without the modulatory chords, which Jagger concludes “must to some extent have had Sullivan’s blessing.”

Several musical numbers are transposed to different keys in later sources. Phœbe’s opening song, in D-flat in Aut and vocal scores, is in E-flat in the modern D’Oyly Carte band  parts. Her second song, in E-flat originally, is in F in the modern band parts. Jagger does not mention the possibility (indeed, the likelihood) that Sullivan authorized these higher keys at the 1897 revival, which he conducted, and where soprano Florence Perry played Phœbe.

Most perplexing, “I have a song to sing, O!” was originally in D, but both the modern band parts and Ch2½ transpose it to E-flat. Sullivan’s approval is hard to deny, as the instruction “In E-flat” is written in pencil in Aut. And yet, the preceding number clearly prepares for a modulation to D, and the reprise in the Act II finale remains in D, so it is difficult to think of a good reason for the change. Jagger comments on the problem, but the edition prints the number only in Sullivan’s original key.

I will be critiquing Jagger, at times, for decisions he got wrong. I ought, in fairness, to begin with a few of the bigger ones he got right.

In “The pris’ner comes,” there is traditionally a tubular bell, which tolls every two bars. It is almost never heard as Sullivan wrote it. In Aut, the bell continued to toll into the Allegro agitato, up to where Fairfax sings, “I and my comrades sought the pris’ner’s cell” — the last toll coming on the word “cell.” In the original production, the bell was not rung by the principal percussionist, who must also play the timpani during the same passage. Once the opera went on tour, it must have become impractical for the company to carry an extra person just to play the bell, so it stopped tolling before the percussionist picked up his timpani mallets.

It is traditional to play “The pris’ner comes” much slower than the adjoining material, and this is obviously not what Sullivan wanted. At the Allegro agitato, the composer wrote “doppio movimento,” meaning “twice as fast.” As the passage is in cut time, this would ensure that the bell strokes continue to come at the same intervals as before.

In the Act II finale, the “Oh, day of terror” section has never appeared in print correctly. Ch1 has a clear error in Dame Carruthers’s part. Ch2½ attempted to correct this, but in so doing, dropped Kate’s part and made other mistakes, which OUP rectifies.

The words “I have a song to sing, O!” appear six times in the opera. In the traditional libretti and vocal scores, there are two responses to this. On four occasions, it’s “Sing me your song, O!” Twice, it’s “What is your song, O?” In the prompt book for the original production, Gilbert altered the latter: it should be “Sing me your song, O!” every time. OUP makes this correction.

Other corrections come from this same prompt book. In Act I, Dame Carruthers refers to the “stern [not grim] old fortalice.” Later in Act I, Point can “teach us with a quirk [not quip].” In Act II, Point “wishes he were [not was] dead.”

One will find many little fixes so subtle that the average listener might miss them. In the Act I finale, we’re used to hearing Fairfax sing this:

But Sullivan entered a change in Aut that the vocal scores failed to pick up, and this was never corrected until Jagger noticed it:

“Hark! What was that, sir?” brings two changes of similar subtlety. After Meryll’s “What can it mean?” the orchestra’s next few bars are traditionally heard like this:

In Aut, the third bar is different:

A few bars later, the traditional orchestration sounds like this:

In Aut, the second bar has F Major, rather than E minor, harmony:

These examples are curious, in that the published vocal scores and the D’Oyly Carte band parts are in agreement with the first excerpt in each case; nevertheless, OUP follows Aut, which shows no sign of revision.

Everyone to have edited a Sullivan opera has recognized that the composer’s autograph scores are incomplete. The text of an opera evolved during rehearsals, and Sullivan was not careful to write down every little change that he agreed to. Vocal scores, published weeks or months after the premiere, often have readings that disagree with the autographs, and that cannot rationally be attributed to error. No one thinks that the publishers of Sullivan’s vocal scores deliberately rewrote his works. But vocal scores were usually in preparation while the piece was still in rehearsal, and there was no assurance that every detail made it into print right away—or ever.

The libretto presents further problems, since Gilbert made changes on his own. Like Sullivan, Gilbert was not careful to ensure that every decision was published immediately. Many of Gilbert’s changes were first documented in prompt books, and not reflected in the printed libretti until many years later — if they were at all.

An additional problem is that the D’Oyly Carte organization—despite its claimed fidelity to the original texts—is known to have made unauthorized changes, long past the creators’ lifetimes. When Chappell began to re-issue the vocal scores and libretti after World War I, it brought them into conformance with the performance standard prevailing at the time, which included long-unpublished changes that Gilbert and Sullivan demonstrably approved, the company’s newly-minted “tradition,” and a vast middle ground of “plausible changes” whose authorship cannot now be determined.

What is an editor to do? One strategy is to consider only those sources directly associated with Gilbert and Sullivan: manuscripts and first-editions that they supervised. This approach maximizes the chances that every word and note was desired by the creators at some point; but it undoubtedly misses later changes that they weren’t sufficiently careful to publish at the time of the original productions. Let us call this approach “conservative.”

The opposite strategy is to accept most changes found in later sources, save obvious errors or those known to be unauthorized. This approach scoops up more of Gilbert and Sullivan’s later thoughts, with the risk that the editor might print some readings that seem reasonable, but that aren’t actually the the creators’ work. Most editors are conservative, and therefore reject this approach. They’d rather risk omitting a reading the creators approved, than take a chance of printing one they did not.

There are selfish reasons for an editor to be conservative: it’s an approach that produces a text with more pronounced differences from the established tradition. This means that there are more so-called “errors” that the editor can claim to have fixed, allowing more chances for self-congratulation, and giving the edition more of a raison d’être. It also increases the chances that a musical director who chooses the OUP score will be compelled to mandate it for the whole cast, as the differences will be too great to tolerate the score’s co-existence in the rehearsal room with older, presumably more corrupt, sources.

In Ruddigore, Hulme had it easy: after the original production, D’Oyly Carte laid the opera aside for decades. Changes that first appear in twentieth-century sources are highly unlikely to be Gilbert and Sullivan’s. We also know that when Rupert D’Oyly Carte revived Ruddigore, he believed it was unplayable as originally written, and that it had to be revised. In contrast, Yeomen never went out of the repertory, and it was considered one of its creators’ best works: it wasn’t seen as needing much help.

Jagger appears to realize that some readings first printed long after 1888 are authorized—or, at least, could be. He cites sources printed as late as the 1920s, and occasionally, even adopts readings from them. But he is still mostly conservative. His four main  sources are Sullivan’s autograph (Aut), the first edition of the libretto (L3), the first edition of the vocal score (Ch1), and a prompt book from the first production with Gilbert’s handwritten amendments. Into a base text derived from these documents, he selectively imports a few readings from later sources that he considers authoritative.

The rationale for this approach is open to question. Once an editor concedes that later sources contain some readings that are authorized, it is highly likely that there are many others he has missed. The Shakespeare scholar T. H. Howard-Hill made this point in a 1989 journal article:

Whenever we admit the existence of a single authorial substitution in a text we must concede the probable presence of others which will always escape detection. The number of authorial variants which can be proved to be authorial in origin in any individual case…is always much smaller than the total number of authorial variants. It is manifestly ridiculous to believe that every variation a revising author introduces will bear the mark of his genius so that the variant will not only improve the text but improve it in a manner that only the author could. (T. H. Howard-Hill, “Modern Textual Theories and the Editing of Plays.” The Library. Sixth Series, Vol. XI, No. 2 (June 1989): pp. 99–100.)

The correct approach, Howard-Hill contends, is that if a later source contains some new readings that we deem authoritative, the balance should tip in favor of including all of them, save obvious mistakes and those that are demonstrably the work of later hands.

It is a favorable development in Sullivan scholarship that Jagger at least cites the later sources, and is open to considering them, but he far too confident in his ability to winnow the wheat from the chaff. Variants in the second edition of the vocal score (published after Sullivan’s death) are relegated to Appendix A, whose contents are described as “readings that Gilbert & Sullivan would not have recognized.” Some readings in the second edition are baldly stated, without justification, to be “unauthorized” or “not authorized.”

In fact, very few of the readings discussed in Appendix A, are so obviously corrupt that one could confidently say Gilbert and Sullivan wouldn’t have recognized them. Indeed, Jagger adopts a number of these, so the statement is wrong on its face. What’s worse, most of the passages discussed in Appendix A are also in the main commentary, so, to get the full history, you must flip back and forth from one to the other. If you want to understand what the editor did, and why, this is the edition’s most annoying feature.

A few examples will illustrate the contradictions that I am referring to. In the coda to the overture as usually played, there are grace notes in the flute and piccolo parts, as follows:

The grace notes are not in Aut or either of the two American copyist manuscripts. They are in an early set of American band parts and the D’Oyly Carte printed band parts. Jagger includes them in brackets, because they are “undoubtedly effective.”

In Phœbe’s opening solo, Ch3 introduces the direction “meno mosso” (meaning “slow down”) at “Tis but a little word,” with “a tempo” (return to tempo) at “An idle breath”; and similarly in the second verse. Jagger claims that “a simple analysis of Sullivan’s melodic and rhythmic structure shows that he would never have sanctioned this.” What that analysis is, he does not say.

A similar situation occurs in “I’ve jibe and joke,” at the end of the opening section (“…and know no fear”). Ch2½ introduces a “rall.” Jagger omits it, and merely notes the alternative in an appendix.

In the opening chorus, the Yeomen sing the following:

Notice that the triplet on the words “Here at” does not extend to the second tenors, who would sing the word “at” a fraction of a beat later than the rest of the men. This is what appears in Aut and Ch1. Jagger deems it “open to debate” whether Sullivan actually wanted this, and he lets it stand. One must turn to Appendix A, to learn that Ch2½ (he says Ch3) alters the second tenor part to match the rhythm of the other singers.

In Elsie’s Act I song, Sullivan wrote the following:

This is how it appears in the Aut and Ch1, but Ch2½ alters Elsie’s note on the word “this” from C-flat to D-flat. Jagger prints the D-flat in cue-size as an alternative, claiming that “it is likely that Sullivan would have encouraged the use of an appoggiatura.”

In “Who fired that shot?” Aut has f where the chorus sing “Anyhow the man is dead….” Vocal scores have a crescendo, with f two bars later; Jagger follows Aut, which he says, “seems more reliable.” But ten bars later (“matters very little”), vocal scores have stringendo, which Jagger prints, even though it is not in Aut.

At the end of the number, Aut and Ch1 have a coda as follows:

Ch2½ and the D’Oyly Carte band parts have a coda three bars shorter. This, Jagger says, clearly represents a much later (unauthorized) performance tradition.” How that is clear, I am not so sure.

In the “Jack and Jill” trio, Jagger prints the following passage:

In Aut, the oboe doubles Phœbe on the word “free,” but rests on the word “degree.” Here, Jagger sides with the D’Oyly Carte printed band parts, as “it seems a repeat [of the earlier bar] is intended.”

In the Act II finale, the chorus response to Point’s “I have a song to sing, O!” offers a fascinating window into how a passage can change, over time. Aut has ff with accents on each word of “Sing me your song, O,” just as in the response to Elsie’s verse. Ch1 has f without accents, and with “dim.” over the final held chord. Finally, Ch3—which Jagger asserts “can be disregarded”—has the familiar pp dynamic, which makes the most pronounced contrast to the loud, accented version that comes a few moments later.

In G&S vocal scores, optional high notes tend to appear in later editions, and Yeomen is no exception. Whether Sullivan would have approved is unknown, but OUP treats these high notes as unauthorized. So, in Elsie’s Act I song, she takes the low F on “No matter how,” and Fairfax takes the low A-flat on “Scarce a word of them is true!” The octave-higher alternatives are not shown.

When the editor of a Sullivan edition comes from a chiefly musical background, the dialogue often gets short shrift. Alas, that has happened here. Textual notes for the dialogue are much less thorough than for the music. You will find many unfamiliar readings in the main text, for which Jagger provides no explanation whatsoever. The problem is not merely the lack of commentary: in many places, Jagger has chosen the demonstrably wrong readings.

A bit of background: In 1890, Gilbert believed his partnership with Sullivan and Carte was over. To protect his rights, he arranged for an assistant to copy the production prompt books for all of the operas they had written to date. These are collected in the Gilbert Papers in the British Library. Every scholar to have considered these prompt books has concluded that the readings they transmit were almost certainly authorized.

Since Jagger’s musical commentary considers sources up to the 1920s, there can be no principled justification for almost entirely overlooking, without a word of explanation, a libretto source from 1890. He was aware of this prompt book—it is in his list of sources—but he didn’t check it thoroughly. Had he done so, I’ve no doubt he would have concluded that its readings generally belong in the main text.

As with the music, the printed libretto sources for Yeomen continued to change, up to the 1920s. I have already discussed the contentious status of variants that appear years after the creative period. Jagger’s approach to the dialogue is conservative, as it is with the music. But dialogue variants from the later sources are not cited with the same diligence as musical variants. It does not seem to have been a priority. The decision to silently accept earlier readings, without comment, is not discussed. Was it deliberate, or an oversight? One cannot tell.

To give but a few examples—and I will confine myself only to dialogue variants documented in the 1890 prompt book that Jagger neither adopts, nor comments on:

  • In the dialogue before Leonard’s entrance, Meryll says “Amen!” (OUP) rather than “Amen to that!” (prompt book)
  • In the dialogue after his entrance, Leonard says, “…wrap it up, lest it take cold and die,” omitting the word “carefully”
  • In the second dialogue with Wilfred and Phœbe, after here “Wilfred—and alone!” the usual next line is omitted: “Now what could he have wanted with her? That’s what puzzles me!”
  • In the same dialogue, Wilfred says “In the nice regulation of a screw…,” rather than “thumbscrew”
  • In the dialogue after the “Jack and Jill” trio, Point says “…let it pass this once,” not “let it pass”

There are many more changes that first appear in printed libretti after 1890, most of them not discussed at all in the commentary. I won’t enumerate them; you get the idea.

Even where he does emend dialogue, Jagger’s textual notes can be rather careless. After “When our gallant Norman foes,” in the original libretto, Phœbe says, “Father! No reprieve for the poor gentleman?” On the authority of an earlier prompt book, he correctly alters it to, “Father! Has no reprieve arrived for the poor gentleman?” But the commentary merely states that the change was not made in the published libretto “until much later” — but when? He does not say.

The failure to systematically poll the 1890 prompt book leaves Jagger in quandaries that he could have resolved rather easily. The original libretto gives Phœbe no entrance for the dialogue that precedes “When our gallant Norman foes,” and this puzzles him:

Gil[bert] was extremely careful about marking such things, so it seems unlikely that she came on with Dame C. or during the dialogue, but perhaps at some point during [the preceding number].

Gilbert, in fact, was anything but careful about marking such things: inconsistent entrances and exits abound in the early libretti. Finding no entrance, Jagger elects not to give her one. The 1890 prompt book resolves it: Phœbe enters at the same time as Dame Carruthers—precisely the solution that Jagger found “unlikely.”

Likewise, the original libretto lacks entrances for Phœbe, Dame Carruthers, Meryll, or Wilfred in the Act II finale; and again, Jagger declines to supply them. Had he checked the 1890 prompt book, he would have found entrances for three of the four clearly marked. (It has no separate entrance for Meryll, but it has one for the “Warders,” a group that Meryll can readily be assumed to accompany.)

Within the lyrics, some of Jagger’s decisions will be debated, but he has at least documented his reasons. In the opening chorus, the Yeomen “enjoy talking [not telling] over our impetuous May and June.” This comes from Aut, but no other source has it: to Jagger, telling is “grammatically suspect.”

In “I have a song to sing, O!” Aut has “soul was sore [not sad]” in the second verse. This is surely a slip, as the “house that Jack built” repetition is the whole idea of the song. Yet, Jagger lets it stand because L3 had it too. Authas  “soul was sore” again in the Act II finale, but in this case, Jagger does not print it, because L3 has sad. With the ghost of Gideon Crawle at our side, we can say with confidence that the L3’s reading in Act I is an error.

In L3, the Act I finale has the following couplet:

The man we sought, as truth will show,
Had vanished into empty air!

But the chorus repeat has:

The man they sought with anxious care
Had vanished into empty air!

Sullivan gave both versions of this couplet to Fairfax and the three yeomen accompanying him, and then restated both versions of it with the full ensemble. Later libretti and vocal scores eliminate the “truth will show” couplet (even though it rhymes with “hunted low”), in favor of repeating “anxious care.” OUP follows the early sources, and prints both.

In the final ensemble of Act I (“All frenzied with despair”), the early sources give separate words for Elsie and Point. These words disappear in later libretti, as well as in Ch2½. OUP retains them.

In a few places, Jagger declines to make a decision where I believe he should. Seventeen bars before the end of the overture (where the vocal score has fff dynamic), Sullivan entered “Picc.” on the Flute 2 staff, and “change to Picc.” four bars earlier, but without crossing out any notes. The change requires further alterations: a flautist cannot switch to piccolo in the space of one eighth note at Allegro time, as the OUP edition would have her do. Jagger deems the last bars of flute harmony “a shame to lose,” so he leaves them intact—and leaves the musician with a part that is literally impossible to play.

In the Act I finale, the Lieutenant “tells off FAIRFAX and two others to bring the prisoner to execution.” In the music, Sullivan wrote parts for Fairfax and three others. The 1890 prompt book shows only two yeomen accompanying Fairfax. Yet, there are programmes from late in the original run, in which a Third Yeoman is credited. At some point the Third Yeoman definitely was deleted—maybe on Gilbert’s watch, maybe not—and this is reflected in Ch2½. Jagger elects not to resolve the inconsistency: he leaves the original stage direction in place, but also retains the Third Yeomen’s part in the music.

In “Hereupon we’re both agreed,” Sullivan introduces the following rhythmic figure:

Later in the number (when accompanying the voices), Sullivan writes it this way:

Jagger deems it “very unlikely” that Sullivan wanted them to be different, but he declines to make a decision, leaving the inconsistency for the conductor to resolve. Similarly, in “Strange adventure,” Sullivan at first wrote the following:

At the repeat, Sullivan wrote the slurs differently:

This time, Jagger finds it “difficult to say” whether this was intentional, and allows it to stand.

At the beginning of “I have a song to sing, O!” Sullivan wrote the following:

Every other time in the entire work that the same phrase occurs, Sullivan articulated it like this:

Jagger preserves this difference, for the peculiar reason that the vocal score did so too.

In “Rapture, rapture!” he makes a better choice. The following rhythmic figure recurs throughout the number:

Wherever there is a slur on the first two eighth notes of a bar, Sullivan sometimes wrote a staccato dot only on the next note, and sometimes on all remaining notes of the bar. Here, Jagger normalizes to the first form, though the commentary (presumably written after engraving) reaches the opposite conclusion: “Most likely, Sullivan would have expected all non-slurred quavers to be staccato here, but, to be consistent, OUP adopts the single dot after a slur throughout.”

In the Act II finale, the following peculiar passage occurs:

Notice that the word “cloyed” is set as a two-syllable word. Sullivan wrote that in Aut, and it was dutifully copied into Ch1. Jagger notes how awkward it is: the three female principals have just sung “cloyed” as a one-syllable word, and it is clearly meant to rhyme with the “–loyed” of “unalloyed,” also set as one syllable. Nevertheless, he lets it stand, which I would classify as a blunder. (Ch2½ altered it, to what I am sure was the intended reading: sopranos and men sing “cloy’d” on dotted half notes; altos sing “cloyed” on two slurred notes, as one syllable.)

The editor of any Sullivan score must decide how much of the composer’s notation to modernize. The Broude editions preserve Sullivan’s practice of writing separate stems for each syllable sung under eighth or sixteenth notes. Oxford follows the modern practice of beaming them together, relying on slurs to signal melismatic passages. Both publishers print the vocal parts above the violin staff; Sullivan wrote them between the violins and violas.

In his autograph scores, Sullivan used both English and Italian names for the instruments of the orchestra. The OUP score normalizes them to English, which seems sensible, although it seems a bit jarring to see “Vcl.” as an abbreviation on the later pages of movements, when “Cello” appears on the first page.

OUP retains Sullivan’s keys for the transposing wind instruments. A note in the score states that modern transpositions (e.g., Horns in F) are available in the rental band parts. Oxford preserves Sullivan’s use of alto clef for the first and second trombones; the band parts offer the alternative of bass clef.

Sullivan followed the nineteenth-century German practice of writing horn parts that transpose up when in bass clef, but down when in treble clef. OUP alters them to transpose down in all cases, as modern horn parts do. Sullivan’s clarinet transpositions in this opera are inconsistent. I won’t bore you with the details, but there doesn’t appear to be a musical justification, and yet, Jagger lets them stand.

In Sullivan’s early vocal scores, the women’s chorus are sometimes described as “sopranos and altos,” and other times as “first and second sopranos.” OUP’s Ruddigore used the former, this edition the latter. This could confuse some readers, and for no good reason. When the original production of Ruddigore closed, Sullivan didn’t fire the altos, and replace them with second sopranos. It was just a different term to describe the same singers. Here, the modern term (alto) should have been used.

The edition is let down occasionally by its physical presentation, which I attribute to OUP, not the editor. OUP’s Ruddigore full score is about 1.5cm taller and 1.5cm wider than the Yeomen score. It doesn’t sound like much, but when coupled with wider margins, all of the staves in Yeomen are about 15–20% reduced, with a corresponding reduction in font size for things like tempo markings, lyrics, and so forth. The fact that Yeomen has a larger orchestra does not entirely account for this. The opening page of the overture requires 14 staves for both operas. Ruddigore is conspicuously easier to read.

There are some numbers where Sullivan used repeat bars for strophic material in Aut, but Ch1 wrote out multiple verses in full. “Were I thy bride” and “’Tis done! I am a bride!” are examples of this. Jagger (or OUP) prefers to repeat, finding it “more economical with layout and page-turns.” This is debatable—turning forward is usually easier—but that was the decision. It makes the edition a handful of pages shorter, and hence saves a few pennies. I suspect that was the motivation.

The dialogue in the full score is about 25% reduced in size, and this is frankly inexcusable, in that most dialogue scenes conclude with about half the page unused. To print the dialogue at a reasonable size, the full score would have to be, perhaps, four pages longer — insignificant for a volume this expensive. The dialogue in the vocal score is printed at a more normal size, although it still feels slightly cramped.

Where there are full-page systems on facing pages, the staves occasionally fail to line up. For instance, the oboe occupies the second staff of p. 10, but the third staff of p. 11: the piccolo staff needed to appear on p. 10 with rests, so that the other staves would line up. The typesetter usually got it right, but missed this easy fix on several occasions.

In “I’ve jibe and joke” and “Were I thy bride,” Sullivan divides the violins into three and four parts respectively. The modern convention is to bracket all of the violin staves together, and to put a further bracket around the divided staves. OUP’s typesetter omits the former. In the overture, where there are no divisi staves, the bracket is missing altogether. All the other numbers have it, so this is just a slip.

The full and vocal scores list for $90 and $22.95 respectively; you’ll find them discounted on Amazon. These are fair prices. In contrast, the Schirmer and Kalmus vocal scores (the latter a Chappell reprint) are $25 and $24.95 respectively. Kalmus published the only other full score of Yeomen, in 1979—and it was terrible. It appears to be out of print, but I paid $95 for it in the 1980s!

Both volumes are paperbacks, and for the full score this seems like a short-sighted decision. Critical edition full scores, which are expected to take a beating over many decades, are usually cloth-bound, as OUP’s Ruddigore was. On the other hand, you save a bit of money: the Ruddigore full score currently lists for $130.

All of this may seem extremely fussy: what is more important is that OUP has published an edition that rewards such detailed study. Were I a musical director, I would accept most of Jagger’s musical decisions and reject a few. In the dialogue, to which Jagger paid less careful attention, I would reject most of his decisions and accept a few.

If the ups and downs of my criticism leave you in a quandary, let me remind you of the Cliff’s Notes version: this is a terrific new edition of The Yeomen of the Guard.