I never expected to have my feelings about Handel’s Tamerlano transformed by a bunch of Greeks I had never heard of. The critical consensus is that during his 1723-1724 seasons Handel produced three straight masterpieces: Giulio Caesare, Rodelinda and Tamerlano, which, along with his operas based on Ariosto (Alcina, Orlando and Ariodante) comprise the group from which most people would pick his “best” Italian opera. "Giulio Cesare" was my introduction to the strange, seductive world of high baroque opera, with the hook set by Beverly Sills’ seductive Cleopatra in that now-outdated recording of the New York City Opera production that had made her a star. "Rodelinda" became the first Handel opera album I owned, in the old Westminster recording, and although I have supplemented my collection with more recent performances I still thrill to the vocal beauty Maureen Forrester and Teresa-Stich Randall brought to the roles of Bertarido and his devoted wife. It wasn’t too long before I added "Tamerlano" to my collection in the groundbreaking recording on Cambridge directed by John Moriarty. But while I understood the critics’ admiration of Tamerlano’s brilliant dramatic construction, I tended to neglect the work because none of its many fine arias had same impact on me as say, Bertarido’s "Dove sei?" or Cleopatra’s "Piangero." But this recording conducted by George Petrou, based on the score Handel settled on for the opera’s debut in 1724, has blown away my reservations about "Tamerlano." I have finally felt the work’s full dramatic impact and understand on a visceral level why it ranks with the other two masterpieces of those magical seasons, which were some decades earlier in capturing my affection.
In order to understand why this recording succeeds so brilliantly, you need to know a little about how "Tamerlano" differs from Handel’s other works, and to appreciate the choices he made in his final changes to the score. Handel was very often a rapid composer, frequently requiring only a week per act to write his operas or oratorios, but on this occasion, after he spent the usual twenty days to produce his first version of "Tamerlano" he ended up doing a succession of drastic rewrites in an effort to intensify the dramatic effect. Often in his career Handel’s efforts to pare down the long libretti he adapted for the London stage involved slashing the secco recitative that advances the plot in opera seria. In several of his operas this resulted in making the story, not always terribly plausible to begin with, seem downright nonsensical because the audience lacks crucial information about the characters’ motives. Handel’s intentions in "Tamerlano" can be gauged by the fact that he kept more of this recitative between the arias than in any other opera. He was writing for a couple of new singers, the tenor who would play Bajazet, and the castrato who would take the title role, and began the role of Irene for a higher-voiced singer than the one who eventually sang the role. Handel rewrote her arias, made some adjustments for the new castrato, and radically revised the part of Bajazet after he finally heard the tenor hired to play the proud Ottoman captive who defiantly poisons himself. But his revisions were about much more than fitting music to voices. He spent a great deal of energy trying to get the opening of the opera just right, and kept making changes to the dramatic final act until very shortly before the opera’s premiere.
Handel was struggling against the prevailing stage conventions to produce a genuinely tragic opera. He ended up composing and then discarding enough music to make the reputation of a lesser musician. Dean and Knapp point out in their magisterial survey of "Handel’s Operas 1704-1726" that “no fewer than 26 pieces were transformed or rejected before the first performance of Tamerlano” and note that during the three months that the composer was revising the opera he inserted more than 40 pages of additions and revisions into the autograph score, almost a third of its total length. There’s an old writing manual that offers the advice to “murder your pets” which means to be ruthless enough to discard a passage you're proud of if it interferes with the overall impact of the piece. Handel murdered a lot of pets in the process of getting Tamerlano to the stage. But because the rejected numbers still exist, producers of the opera have been fatally tempted to bring those murdered darlings back to life, rather than trusting that the composer might have had a very good reason to put them down.
Consider the old Cambridge recording that was my first introduction to "Tamerlano." In the libretto booklet two different writers defend the decision to restore numbers Handel cut from the final act: Charles Fisher claims that “the opera was seriously weakened by the cutting of the final scene,” and John Roberts calls the composer’s cuts “dramatically inept.” But comparison of that version with this Petrou recording makes it clear that those judgments are dead wrong. One obvious indicator is the difference in Act Three’s length. Petrou’s version takes almost exactly one hour. Moriarty cuts some recitative and omits the B section and da capo of two arias, but with the numbers he restores his third act still runs twenty minutes longer. So far as I know, there is no such eighty-minute final act anywhere else among Handel’s operas.
Even in the trimmed version Handel settled upon, the last act of the opera is crowded with action. It includes a full-scale dramatic scene where the emperor Tamerlano tries to humiliate Asteria, who has rejected his offer of marriage, by making her serve him at dinner in front of her father, the defiant captive Bajazet. The beleaguered heroine is thwarted in her efforts to poison, first the emperor, and then herself. After Tamerlano has condemned Asteria to a more extreme disgrace than waiting on his table (sexual submission to his slaves) her father returns and announces that he has secured his freedom by taking poison himself. He then spews defiance while he staggers toward death. As Petrou’s account makes clear, Bajazet’s death scene is the opera’s climax, an astonishing and unprecedented creation that caps off something completely new. In the words of Paul Henry Lang “This is the first great tenor role in opera.” With Handel’s cuts in place, the remaining characters briefly sort through the wreckage and move to the sorrowful, minor-key coro that concludes the opera, where the music makes it clear that the happy ending they are trying to wish into existence is a sham. Petrou’s recording confirms that this ending can have the dramatic impact the composer worked so hard to create.
The hardest of the composer’s cuts must surely have been that of Asteria’s accompanied recitative beginning "Mirami" and continuings with her aria "Padre Amato," which follow her father’s exit in the final throes of death. It’s a wonderful sequence, expressive and beautiful, and I’d hate to live without Carol Bogard’s wonderful performance on the Cambridge recording. The sequence reveals that Asteria has every intention of following her father into death. Handel worked hard to make some such number work—he tried a different recitative first—because her impending suicide removes any doubt about the opera’s tragic ending. But he ended up cutting the scene because, in the words of the great Handel scholar Winton Dean, Handel “knew when the drama must take precedence over even his greatest music. Between the death scene and the coro a big scena for Asteria must be an anticlimax.” So it proves in the Moriarty version; wonderful though Asteria’s music is, it stops the action at a point where momentum is imperative.
What Petrou’s fine production made me see clearly was that Tamerlano really works as a drama. It is to Handel’s other great works of period what Puccini’s "Tosca" is to "Manon Lescaut" and "La Boheme." Lyrical beauty is subordinated to dramatic effect, and by garnishing "Tamerlano" with the lovely music the composer discarded, you weaken the opera rather than strengthening it. Does anyone believe that Tosca would be improved if Puccini had decided to insert a big dramatic aria for his heroine in between the moment she discovers that her lover is dead and her leap from the castle walls?
None of this resolve to serve Handel’s intentions would have succeeded if Petrou had not found a cast that could do full justice to the music of Tamerlano. I had never heard of soprano Mata Katsuli, but she imparts passion and beauty to the role of Asteria. Mezzo Mary-Ellen Nesi provides the proper plaintive beauty to the part of Asteria’s hapless would-be lover Andronico, and Tassi Christoyannis is an impressive Bajazet. Until very near the end counter-tenor Nicholas Spanos in the title role had me more than satisfied with his performance, rather than making me long for a female singer as the shakier counter tenors do. But the one miscalculation of the entire production is the only number where Petrou confesses to having gone with the full-length version of an aria that Handel trimmed for performance. The long version of Tamerlano’s "A dispetto," which the conductor calls “a wonderful vehicle of coloratura display” proved something less than that when Spanos proved unable to do much but hang on for dear life as he tackled the aria’s torrent of notes. A more successful choice was his adoption of an extremely fast tempo for Andronico’s aria "Se non mi rendi." It’s the closest the character comes to standing up to the king who wants his woman and controls his destiny. In the Cambridge recording Sofia Steffan takes the piece at a more normal pace, and it turns into one more languid plea from a helpless prince. Nesi’s rapid-fire performance makes him seem frantic, almost hysterical, so that the number helps screw up the drama in an already-dramatic act. The cast also wrings every bit of emotion from those “dry” recitatives that propel the story while taking them at the rapid pace Dean recommends for maintaining momentum.
From now on out this recording should be the choice for opera fans who want to make their first acquaintance with this great opera. Those who have already come to love specific numbers that were cut from the 1724 performances will naturally seek out other recordings as well. That’s natural. A real fan of Handel’s operas won’t settle for a single version, however good, any more than a Mozart-lover will rest content with a single "Don Giovanni." But the existence of this Tamerlano recording may help prevent the fate that has befallen that Mozart masterpiece. The standard performing edition of Don Giovanni you’ll encounter today includes at least three more numbers than the composer originally provided. We feel like we’re being shorted if we don’t get "Dalla sua pace" and "Mi tradi," and the singers of those arias will certainly get cross if they’re left out. We’ve also gotten used to the “happy” final number Mozart was obliged to add for his Vienna performances. But the result has been that we usually hear one of the very best operas ever written in a version that bogs down for a while in the second act. A similar trend seems to have been growing for "Tamerlano," with more a serious impact on the work’s reputation. It would be a valuable accomplishment if this recording resulted in more companies taking a chance on the composer’s instincts. The Gardner recording will let you hear how the finale works if we tarry to hear a duet for two castrati (or, today, counter tenors) and I can go back to my Cambridge recording if I want to hear the cut aria for Asteria, plus a duet between Tamerlano and Irene, and an extra bass aria. But that last aria is not nearly as engaging as the excised bass aria that Petrou provides in an appendix after his final act comes to its moving conclusion. Perhaps future recordings will follow his lead, and let us enjoy at our leisure all the fine music Handel discarded, without feeling obliged to insert it into the action. It doesn't pay to second guess a first-rate musical dramatist at the height of his powers.