Sunday, March 18, 2018

The mysteries of Dido and Aeneas

Guillermo Resto (Aeneas) and Mark Morris (Dido) in the Mark Morris Dance Group's Dido and Aeneas.
Original image from

It's remarkable that we have so little information about the greatest opera in English, Henry Purcell's Dido and Aeneas. The scores that exist are neither complete nor original; instead, they are partial copies that don't agree with one another, made for performances that occurred decades after Purcell's death in 1695. There is only a single surviving copy of Nahum Tate's libretto from the first (or at least a very early) performance. It does not indicate who sang the roles, or even the title of the opera, and it includes sections for which the music has been lost. We are fortunate indeed that enough of the work remains to enable staged performances.

Many questions and controversies remain. As musicologist Ellen Harris has written about Dido, "we know even less than we did [30 years ago], or at least less than we had imagined. We can no longer say with certainty in what year the opera was written, where it had its premiere, who performed it or even what the original score contained — the very things that normally provide the foundation for our understanding of a piece of music." [1]

But evidence uncovered relatively recently and discussed in Harris's Henry Purcell's Dido and Aeneas (Second edition, Oxford, 2018), has, in my view, helped to partly clarify rather than further confuse our understanding of Dido's origins. To take Harris's questions in a slightly different order:

Where was the opera first performed?

The one surviving 17th-century libretto for Dido is for "An Opera Perform'd at Mr. Josias Priest's Boarding-School at Chelsey By Young Gentlewomen."

The first page of the libretto of Dido and Aeneas printed for the Chelsea school performance (detail).
Original from the collection of the Royal College of Music.

In 1684 the young gentlewomen at Priest's school had performed John Blow and Anne Kingsmill's Venus and Adonis, the work that clearly provided a model for Dido. It is known that Venus and Adonis was first performed at court, probably in 1683; a libretto from the Chelsea school performance of Venus and Adonis states that it is "An Opera Perform'd before the King. Afterwards at Mr. Josias Priest's Boarding School at Chelsey By Young Gentlewomen."

The first page of the libretto of Venus and Adonis printed for the Chelsea school performance (detail).
Original from the collection of the Cambridge University Library.

This has led some scholars to speculate, despite a lack of direct evidence, that Dido may also have been first performed at court. A major issue with this idea is that the libretto for Dido says nothing about a court performance; its title page merely designates it "An Opera." Surely if Dido had been first performed at court the libretto would have trumpeted the fact, as the Venus libretto does.

There is one circumstance under which the Chelsea school libretto would have been silent about a previous court performance: if that court performance had occurred while James II was king, but the libretto was printed after he had been deposed in the Glorious Revolution of December 1688. That possibility has been virtually ruled out by recent developments, as we'll see in a moment. So it's most likely that no court performance ever took place, and that the performance at Priest's school was the opera's first.

Who performed the opera?

The libretto states that the opera was performed "by young gentlewomen." Dido and Aeneas, of course, is the tragic love story of Queen Dido of Carthage and Aeneas, the heroic Trojan warrior. Some scholars, perhaps perplexed or offended by the thought of a young woman portraying Aeneas, have suggested that a male singer must have been brought in to sing the role.

However, when the tragic love story Venus and Adonis was performed at Priest's school in 1684 the young gentlewomen took all the parts; a note on the libretto mentions that "Mr. Priest's Daughter acted Adonis."

John Verney's inscription on the libretto for Venus and Adonis
Original from the collection of the Cambridge University Library.

So we have contemporary evidence that operas at Priest's school were performed by all-female casts, and no evidence to the contrary. It seems reasonable to assume that a similar practice was followed in staging Dido and Aeneas, and that Aeneas and the other male roles in the opera were sung by young women.

How was the opera performed?

The version of Dido that was performed at Priest's school was probably half again as long as the version that is generally performed today. The libretto for that performance includes a mythological prologue in two scenes in which Phoebus (Apollo), the sun-god, and Venus, the love-goddess, are honored and the arrival of spring is celebrated.

The prologue was clearly accompanied by music; it includes choruses and dances for Tritons, Nereids, Nymphs, shepherds and shepherdesses. [2] And most likely the prologue was preceded by its own overture. Unfortunately the music for both the prologue and its overture has been lost, as well as that for some of the dances that are indicated during the course of the opera.

After the opera a poem was recited aloud. Thomas D'Urfey's New Poems (1690) contains an "Epilogue to the Opera of DIDO and AENEAS, performed at Mr. Preist's Boarding-School at Chelsey; Spoken by the Lady Dorothy Burk." [3]

From Thomas D'Urfey's New Poems (1690), p. 82.
Original from the Henry E. Huntington Library and Art Gallery 

The surviving music of Dido takes less than an hour to perform; together, the first overture, the sung prologue, the missing dances and the spoken epilogue may have added another 25 minutes or so to the runtime.

When was the opera performed? 

The texts of both the prologue and the epilogue have been parsed for clues to the likely date of their performance. The prologue, with its scene celebrating the arrival of spring, would seem to point to a performance in April or May. Of course, it's also possible that the celebration of spring could be allegorical, and celebrate a rebirth or renewal of another kind.

For those who read the prologue allegorically, it has been suggested that Phoebus and Venus represent the king and queen (but which king and queen?). Venus is described as "the New rising star of the Ocean" and Spring "Welcomes Venus to the shore." These references have been interpreted by some to indicate that Venus is meant to represent Mary II, who had returned across the English Channel from the Netherlands in January 1689 to join her husband William after her father, James II, had fled.

D'Urfey's epilogue includes the following lines:
Rome may allow strange Tricks to please her Sons,
But we are Protestants and English Nuns,
Like nimble Fawns, and Birds that bless the Spring
Unscar'd by turning Times we dance and sing. . . [4]

From Thomas D'Urfey's New Poems (1690), p. 83.
Original from the Henry E. Huntington Library and Art Gallery 

"Rome" is generally a reference to Catholicism (James II had converted in the late 1660s). "Strange Tricks" may refer to the "warming-pan baby," the allegation that the son to whom the wife of James II, Maria of Modena, gave birth in June 1688 was not hers. According to this (almost certainly false) rumor, the warming-pan baby was a substitute infant smuggled into the birth chamber in a bed-warming pan after Maria's child was stillborn in order to insure that James could claim a Catholic heir. The arrival of that son excited fears in Parliament about the establishment of a Catholic dynasty on the British throne, and led to the secret invitation to William of Orange to invade and overthrow James. "Turning times" is thought by some to refer to the Glorious Revolution.

So based on these admittedly speculative readings of the prologue and epilogue, we have a tentative time frame for the performance of Dido and Aeneas: spring 1689, probably after the coronation of William and Mary on April 11. The mystery is solved, no?

Well, no. There are two major problems with a date of spring 1689 for the first performance of Dido. The first is that we don't know whether the libretto and the epilogue derive from the same performance; and the second is that we now know that Dido was likely performed a year or more earlier.

The Letter from Aleppo

In 2009 scholar Bryan White made a stunning archival discovery. While investigating the letters sent home from Aleppo by a music-loving English silk merchant, Rowland Sherman, White came across this passage in a letter dated February 1689 (new style; 1688 old style, when the new year was celebrated March 25):
If Harry has sett to the Harpsechord the Symph[ony] of the mask he made for Preists Ball, I should be very glad of a copie of it. There's another Symph[ony] in the same mask I think in C♭, in the 2d p[ar]t is a very neat point th[a]t moves all in quavers [eighth notes]. if he's applyed th[a]t to the harpsechord, 'twould be very acceptabl[e] too. [5]
"The mask he made for Preists Ball." Assuming that "Harry" is Henry Purcell, an acquaintance of Sherman, there is only one known dramatic entertainment he composed for Priest's school: Dido and Aeneas. If "the Symphony of the mask" refers to the overture to the prologue, then "another Symphony" may refer to the overture played before Act I of the opera. Fortunately, that overture has survived. It is in the key of C minor (C♭), and has two parts; in the second part the upper strings play chasing runs of eighth notes. In the performance that follows by the Amsterdam Bach Soloists conducted by Jan Willem De Vriend, the second part of the overture begins about 50 seconds in:

So Sherman must have been familiar with the score of Dido and Aeneas before he sailed to Aleppo. It was a journey that took almost three months, and Sherman arrived at the bay of Scandroon (Iskenderun) in late October 1688. White has determined that Sherman must have sailed from London no later than the end of July, 1688.

If Dido was made for Priest's school, there was no court performance. And if the score was written before July 1688 it seems unlikely that it wasn't performed until a full year later. In the allegorical prologue, Phoebus and Venus could as easily refer to James II and Maria of Modena as to William and Mary. So on the evidence of Sherman's letter it seems that the first performance of the opera was in spring 1688 or earlier.

But what about D'Urfey's epilogue published in 1690 (or late 1689) with its references to "Rome," "strange tricks" and "turning times"? If the epilogue was performed in spring 1688 or earlier it obviously couldn't allude to the Glorious Revolution; if it wasn't performed until 1689, how can Sherman's familiarity with Dido's score in 1688 be explained?

The Letters of John Verney

The letters of John Verney may shed some light on these questions. Verney was a silk merchant living in London; his wife's family lived in Little Chelsea, not far from Priest's school. Verney's niece Mary attended Priest's beginning in August 1683, when she was eight years old. John Verney wrote to his brother Edmund, Mary's father, who lived well outside London, about how Mary was doing at the school. Verney mentions Mary's performances in Priest's "Grande Balles"; from these references we know that balls were held on 17 April 1684 (when Venus and Adonis was performed), 21 May 1685, 15 and 22 April 1686, and 1 December 1687.

Harris suggests that there may have been at least two performances of Dido at Priest's school: one before July 1688, and one in 1689 (probably in the spring). For the date of the first performance both she and White choose 1 December 1687, the last documented school ball before Sherman's departure from England in July 1688. However, there are no letters from John to Edmund Verney between December 1687 and August 1688, and in any case Mary was probably withdrawn from Priest's school in February 1688 because of Edmund's difficulties in paying the fees. So if there was a ball in spring 1688 (as seems highly probable) it would not be documented in Verney's letters.

I am less convinced than Harris and White that Dido was first performed in December 1687. For one thing, the prologue is full of spring imagery; for another, by the date of his letter in February 1689 would Sherman remember so vividly (and be so excited about receiving the music for) a piece that he had heard 14 months previously? It seems to me more likely that there was a school ball in spring 1688 at which Dido was performed and where Sherman encountered the music.

And D'Urfey's epilogue certainly suggests that there was a performance in 1689 as well. Not only are there the cryptic references that may allude to the Glorious Revolution, but it is known that D'Urfey was employed as a singing-master at Priest's school in the summer of 1689, the most likely circumstance for him to supply a poem to be performed during a school ball. [6]

So: in my view the most probable scenario is a first performance in spring 1688, and a repeated performance with the addition of D'Urfey's epilogue in May or early June 1689. The mysteries of Dido and Aeneas are not quite solved; documentary evidence for both of these occasions is still lacking. But based on the evidence we do have, at least this sequence of events is plausible.

Aeneas and Dido (detail), by Pierre-Narcisse, Baron Guérin (1815)

Dido and Aeneas performed by young gentlewomen

As part of the 2018 Berkeley Festival and Exhibition of early music Dido and Aeneas will be performed by the San Francisco Girls Chorus accompanied by the instrumental ensemble Voices of Music on June 7 and June 9; see the Berkeley Festival website for details. The website states that "this concert is a recreation of what may have been heard" at Priest's school; since a male singer, Jesse Blumberg, will perform the role of Aeneas, that claim is doubtful. Nonetheless, I hope my San Francisco Bay Area readers will not miss the opportunity to hear this brilliant and moving work.

Update 22 December 2018: Harris' Henry Purcell's Dido and Aeneas was selected for my Favorites of 2018: Books. The BFX performance of Dido and Aeneas by Mindy Ella Chu, Jesse Blumberg, the SF Girls Chorus and Voices of Music was included in my posts Why we live in cities part 2: Exceptional musical performances and Favorites of 2018: Live performances.

  1. Ellen Harris, "The More We Learn About ‘Dido and Aeneas,’ the Less We Know." New York Times, 15 December 2017.
  2. Nahum Tate. Dido and Aeneas: Prologue.
  3. Thomas D'Urfey, New Poems, 1690, pp. 82-83.;view=fulltext 
  4. D'Urfey, New Poems, p. 83.;view=fulltext
  5. Bryan White. "Letter from Aleppo: dating the Chelsea School performance of Dido and Aeneas." Early Music, vol. 37 no. 3, 2009, pp. 417-428.
  6. How D'Urfey wound up teaching at Priest's school is unknown. He was notorious for his bawdy songs and verse: New Poems includes "A True Tale of a True Intrigue," in which the poet describes being discovered in bed with two sisters; "A Dialogue between a Town Spark and his Miss," about what to do with their child born out of wedlock; "Paid for Peeping: A Poem, Occasion'd by a Peeping hole into a Chamber where a Beautiful and Virtuous young Lady Lodg'd, through which undiscover'd, I could observe all her Actions"; and two lewd ditties "set by Mr. Hen. Purcell" about husbands' difficulties in sexually satisfying their wives. He is a curious choice as a singing teacher for adolescent girls.


  1. Thank you for this detailed and closely reasoned post. As far as the "claim is doubtful," this is not a question of doubt but of interpretation of the surviving musical manuscripts, the historical accounts, but also, importantly, musical analysis of the notes and clefs that come down to us. Here's some key points about this particular work, and concurring examples by other musicologists.
    Venus and Adonis and Purcell's Dido share a number of interesting stylistic traits, but it’s also true that there are some important differences, and different versions of each work.
    First and foremost, the question of recreating a performance is not so much an attempt to establish a “correct” performance practice at a particular school for a particular evening in the absence of key evidence, or to draw straight lines from Venus and Adonis to Purcell’s Opera, but the practical question of how to deal with music which covers a wide range of voices and styles.
    The elephant in the room here is that the choruses, as well as the part of Aeneas, which in Purcell's work use lower clefs. They can't be performed at pitch by a women's chorus or a girls chorus. In addition, there are three baby elephants. The first of these is that the sailor part in Dido uses a high clef, suitable for a female voice, and this part is repeated by the sopranos in the chorus numbers. Elephant number two is that Venus and Adonis crucially has a variant manuscript where the part of Adonis is rewritten in a high clef (and this must be fully explained), and the third elephant is that if you sing the music in Purcell's opera up an octave, the result is you get quite a few parallel fifths when the musical voices cross, and this is completely unacceptable. Purcell didn't write parallel fifths, no good composer did, so there's some work to be done to get to a workable performance.

    Without going into exhaustive detail, a reconstruction has to address all of these issues, not simply touch upon a few details of the surviving sources. It's one thing to theorize about how the work might have been performed, it's a completely different matter to score it up and allocate all the voices so that everything works, not just a few select parts of the work. It has to be fully internally consistent.
    The inscription in Venus and Adonis is interesting, but not definitive. What it clearly shows is that for one performance, that role was probably sung by one of the girls, and there is the supporting evidence of the fragmentary concording manuscript in high clef. But there is no evidence of this kind for Dido in the libretto or the manuscripts for the role of Aeneas. These are two different works that presumably were performed in different ways on multiple occasions. In addition, opera roles were routinely transposed and reassigned, so it's quite possible, even likely, that the Blow Venus and Adonis was performed more than once, and the clefs show different singers with different ranges in different manuscripts. Otherwise, there would be absolutely no reason for the high clef version—and this is important: there would only be one clef for the one role, not several.
    Baroque opera performances varied considerably from production to production. What the Blow manuscripts show is that it is possible to perform male roles in a higher range, but that in this case a separate part was specially written out.

  2. (Dido and Aeneas, continued)
    Now we come to the sailor. If Purcell had wished the male roles to be cleffed in bass and tenor clefs, the sailor would be in a low clef, but this part, tellingly, is in high clef, indicated a voice in the soprano range.
    The parallel fifths: this is the thorniest problem of all. In reviewing hundreds of compositions written for female voices from the 17th century Convent repertory, we have a massive body of work to sort out how this was done. In much of this women's repertory, the music is written in a special, highly sophisticated counterpoint that allows the lower voices to be sung an octave higher with no parallel fifths that result from transposition. Purcell does not do this, he does the opposite. In addition, we have version of Vivaldi's Gloria that show that for special occasions special parts—different parts—were written out and named with the individual singers, so, in effect we have three versions of the work, one of which is custom made for women's voices. There are records of outside soloists appearing in all-women convents and schools as well, and of course Thomas D’Urfey was the singing master at the school where Purcell’s opera was performed.
    For our recreation, we have teased out the counterpoint of the choruses into the invertible counterpoint used in 17th century convents and schools, so that all the choruses and solo parts can be sung by members of the Girls Chorus. It’s a highly effective solution that is fully in keeping with the style of the time. We can safely assume that the later versions handed down to us (we do not have Purcell's holograph) simply arranged the parts for SATB. We have un-arranged them. The part of the sailor is performed as it appears in the earlier manuscript and the part of Aeneas as well. Our solution balances all of the different problems and seeks a compelling historical solution *with the least amount of changes.* It isn't the only way to do it, and many versions are possible. Many versions existed, as with any opera.
    The last loose end is the role of the Sorceress and minor variants in the Spirit and Sailor roles. Her we have conflicting versions in the Purcell manuscripts. There are partbooks from the Academy of Ancient Music performances of Dido and Aeneas in 1774 and 1787 in which the role of the Sorceress appears in bass clef, and this clef type is reinforced from information in a playbook from a performance in 1700, when John Wiltshire sang the Sorceress.
    However, we concur with musicologist Elizabeth Holland that the most likely performance practice is to have Aeneas sung by a baritone, and the other parts sung by high voices. The poet and musician Thomas D'Urfey, was the singing master at Priest's School, and it is very likely that the role was composed for him. Purcell develops the very different style of the Sailor part into a soprano Chorus line, so that needs to be maintained in the proper octave.
    Holland’s astute and detailed musicological interpretation of the sources is online here, with an exhaustive analysis of the available ranges and voice types of Purcell’s time, and we concur with most of her conclusions:
    We will never know Purcell’s intent—and that’s as it should be—but our reconstruction seamlessly fits the available facts.
    David Tayler
    Voices of Music

    1. Many thanks for your comment, David—I'm honored! For any of my readers who are not aware, David is not only co-director of Voices of Music, but is a long-time performer with the Philharmonia Baroque Orchestra and has appeared with other renowned early music groups such as American Bach Soloists, Tafelmusik, Concerto Köln, and the Freiburg Baroque Orchestra. He has a Ph.D. in musicology from U.C. Berkeley, and the subject of his thesis was the solo lute music of John Dowland. He is a highly accomplished musician who is deeply knowledgable about the theory and practice of the music of the seventeenth century.

      So clearly I'm in way over my head. Perhaps unwisely, though, I want to raise some questions about the issues you discuss in your comment. Before I do so I want to be clear that I am not in any way disparaging a production to which I am looking forward with the greatest interest. Nor do I think that the only "authentic" production of Dido and Aeneas must have an all-female cast: as you point out, it was common practice in the Baroque era to make musical changes to fit specific singers.

      So I fully support your decision to cast a male singer as Aeneas. After all, my favorite recording of the work has a baritone Aeneas, Michael Dean. (And if I'm not mistaken, you played on that recording, as well as in the pit during the Mark Morris Dance Group's first and subsequent performances of the opera in Berkeley.) Jesse Blumberg is a fine singer, and I'm very much looking forward to hearing his Aeneas.

      My doubts relate to the description of the production on the Berkeley Festival website, which states, "In 1688, Purcell’s only opera was performed at Josiah Priest’s Girls School in Chelsea, London…this concert is a recreation of what may have been heard at this time." Here are the questions which this description and your comment raise for me:

      You write about the problem of the clefs in which the parts are written, and state that in terms of the vocal ranges assigned to different parts, "we have conflicting versions in the Purcell manuscripts." However, as you also note, for Dido and Aeneas there are no Purcell manuscripts. None of the musical sources for the opera dates from Purcell's lifetime, or even from the seventeenth century.

      If "we can safely assume that the later versions handed down to us…simply arranged the parts" to suit the performing forces at hand, how can we know in which ranges the parts were originally written? Even if the earliest source, the Tenbury manuscript, is based on a score from the early 1700s, that score was prepared for a performance with different forces than were available at Priest's school. You write that, for the part of Aeneas as it has come down to us to be sung in the soprano range, "a special, highly sophisticated counterpoint that allows the lower voices to be sung an octave higher with no parallel fifths that result from transposition" is required, and that "Purcell does not do this." I'm not competent to judge the issue of parallel fifths, but if the clefs in the available musical sources reflect later arrangements (and possibly transpositions) by hands other than Purcell's, it seems difficult to use them to draw conclusions about the clefs in the score as it was composed or performed in 1688.

      (Continued below)

    2. (Reply continued)

      In her book on the opera Ellen Harris writes that "[Bruce] Wood's examination of the Tenbury score strengthens the possibility that the role of the Sorceress was originally written for bass and transposed up for the Chelsea performance(s). If the roles of the Sorceress, Sailor, and Belinda were all sung by trebles at Priest's school, whether composed that way or not, it leads one to ask whether the role of Aeneas (and Phoebus) might also have been sung by a gentlewoman...It seems entirely possible that Dido and Aeneas would have been performed...with the role of Aeneas taken by one of Priest's students" (pp. 70-71).

      You point out that "the poet and musician Thomas D'Urfey, was the singing master at Priest's School, and it is very likely that the role was composed for him." However, the Sherman letter from Aleppo places a performance of Dido in 1688, and in 1688 D'Urfey does not appear to have been the singing-master at Priest's school.

      In the preface to D'Urfey's satirical play Love for Money: Or, the Boarding-School, which was printed in 1691, he writes "I liv'd at a Boarding-School near London all last summer" and that the play was written "June last" (1690). So "all last summer" must refer to the previous summer, that is, 1689. In 1688, a time when there is no evidence that D'Urfey had any association with the school, it seems unlikely that Purcell would have composed the part of Aeneas for him.

      Purcell knew D'Urfey—he set some of his poems to music—and it could be argued that in 1688 he may have composed the part for him to perform as a professional singer brought in from outside the school. However, there is no evidence for this, and the parallels with Venus and Adonis suggest otherwise. The composer of Venus and Adonis, John Blow, also knew D'Urfey: he set one of the poems in D'Urfey's 1683 collection (see But we know that at Priest's school in 1684 neither D'Urfey nor any other male singer was brought in to sing Adonis, and that instead the part was sung by Priest's daughter. My contention remains that the Venus and Adonis inscriptions describing an all-female cast remain the best, and only, direct evidence we have of the performance practices at Priest's school.

      Of course, given the lack of contemporary musical sources and the paucity of documentary evidence nothing about the first performance of Dido is known for certain. But I remain unconvinced that the weight of the surviving evidence supports the view that in 1688 Aeneas was sung by D'Urfey, or by anyone other than one of the young women at Priest's school.

      Many thanks again for taking the time to explore these issues with me and my readers! And I can't wait to hear your reconstruction of the choral parts for the SF Girls Chorus following seventeenth-century practice. Dido and Aeneas is a profound and musically glorious work, and the performances on June 7 & 9 promise to be a highlight of the Berkeley Festival and Exhibition.



  3. It's an interesting and complex question, but for me it comes down to an interpretation that explains *all* of all the sources—many do not. As far as the surviving manuscript, that's all we have, so these are echoes of Purcell's work. The fact that part of the sailor is in high clef is strong piece of evidence, a stronger piece of evidence than the alternate timelines. The alternate timelines are based primarily on secondary sources. In addition, there is the grey area of whether the part of Aeneas was composed with D'Urfey "in mind" even if the some performances did not include him, and this then goes into the question of stylistic analysis, which is subjective. Clearly, however, the sailor part seems written for the soprano range, and the Aeneas part seems written for a baritone range, even if those parts can be adapted for other voices. Pucell worked with many tenors and baritones, and any of them could have sung the role.
    The alternate timelines are based on secondary sources, whereas the music is directly connected to the opera and Purcell, of course, and the libretto. The alternate timelines are hypothetical, and the score is tangible. The alternate theories also do not explain how the transpositions would work (they don't, in fact work). Someone made the assumption that in Venus and Adonis, Adonis was sung by a treble, but this doesn’t explain the sources: there are two versions for baritone and a fragment of one version for mezzo soprano, the only explanation for this is that a baritone sung the role on one or two occasions, and the mezzo on another. In other words, may of the theories don't cover all the facts, they just cover some of them. The surviving parts for Venus & Adonis show two different voices for that role, one low and one high.
    What's always interesting is that different musicologists come up with different interpretations. When I was a graduate student at UC Berkeley, I asked Joseph Kerman, "What do you do when musicologists have different interpretations?" He simply said, find good musicologists. In the case of Purcell, there are many good ones, but I think the work Holland is on point:

  4. Holland: "The role of Aeneas was assigned in the twentieth century both to tenors and baritones, and since it is a 'heroic' role, and appears in all the manuscripts in tenor clef, it is unlikely that it was originally intended as a 'breeches' role. It is, however, possible that Aeneas was performed by Josias Priest's daughter in breeches at the school performance, since she had previously played Adonis in Venus and Adonis, but this would have been an alteration to the score for practical reasons, rather than Purcell's original intention. As was mentioned above, bass songs were occasionally published in the treble clef in Purcell's works, so either the tenor or the bass voice could be appropriate.

    Aeneas's range of d-f' corresponds to Purcell's normal tenor range, but the tenor voice was more of a choral than a solo voice to Purcell, so it is unexpected to find a tenor singing the title role in Purcell's only opera. One would expect him to have chosen the counter-tenor (in those days a high tenor) instead, and to have expanded the role at the upper end of the register. There are no records of stage soloists calling themselves tenors working with Purcell, though some of his higher tenor solos were taken by counter-tenors, and some counter-tenor solos were printed in the tenor clef. There were, however, a number of high basses who sang for Purcell, one of whom, Thomas D'Urfey, was the singing master at Priest's School. [17] It seems highly likely that Aeneas was indeed a high bass part, particularly since Adonis in John Blow's opera Venus and Adonis – which was also performed at Priest's school and is acknowledged to be the inspiration for Dido and Aeneas – was a bass part. The voice most appropriate to sing this role in a historically-informed production of the twenty-first century would therefore be the baritone."
    Complete analysis at

    1. For me when thinking about the likely casting at the first performance of Dido and Aeneas, it is the sources from the time of that performance that should have the greatest weight:

      * The libretto from Priest's school which states that the opera was "performed by young gentlewomen."

      * The notation on the 1684 libretto of Venus and Adonis which indicates that "performed by young gentlewomen" means that female students (including Priest's daughter) took all the roles.

      * Sherman's 1689 letter from Aleppo, which places the first performance of Dido at Priest's school in spring or summer 1688.

      * The 1691 preface to "The Boarding School" in which Thomas D'Urfey states that he was employed at Priest's school beginning in 1689, a year after that first performance, making it unlikely for him to have sung Aeneas on that occasion.

      Together this evidence points to a first performance in 1688 at Priest's boarding school in which a young woman sang the role of Aeneas. In my view the musical sources you reference are not decisive on the question of how the opera was performed at Priest's school. The earliest of them is not contemporary with Purcell, but dates from nearly a century after that first performance.

      This is not to say that a production with a baritone Aeneas isn't historically informed, artistically valid, congruent with the existing musical sources, or even the least problematic way to perform this wonderful opera. However, the weight of the contemporary evidence suggests to me that what was most likely to have been heard at the first performance was a young woman singing the role of Aeneas.

      Perhaps we can continue our discussion in person after one of the performances of Dido and Aeneas on June 7 & 9 at the Berkeley Festival and Exhibition. I'm very much looking forward to it.