Nathaniel Watson, Amanda Forsythe, Daniel Auchincloss and Nathalie Paulin in Les Indes Galantes. (Photo: Julian Bullitt)

In a way it's hard to critique Rameau's Les Indes Galantes, because there's so little like it being performed today. What can you compare it to? It's classified as an "opera-ballet," but it's neither a cohesive 'opera' nor a 'ballet' (instead it's a set of thematically related romantic fables, alternating with dances that seem to float on the edge of audience participation). And the whole piece stops every now and then for a major special effect (like the volcano that erupts at the end of the second act - or pardon me, entrée). I think in the end Les Indes is best described as a kind of baroque variety show.

I don't mean that as a crack, btw. Much of the charm of the recent Boston Baroque production (and it was quite charming, trust me) was that you were never sure what might be coming next.  You did know that each scene was going to have a whimsical point to make about l'amour, but the settings of these musical valentines ranged from Persia to Peru, and were crammed with sultans, sun-gods, savages, and even transvestites; one minute, lovers were kissing before the lava reached the village; the next, they were trading suspiciously intoxicating "peace pipes."  Basically, there was something for everyone.

Luckily, every scene was likewise crammed with gorgeous music, too. Conductor Martin Pearlman has a special jones for the French baroque, and his case for a stripped-down version of Les Indes (the original was something of a spectacle) was that the music was spectacular all on its own, and he was proved quite right; particularly the first half of Rameau's score is wonderful - and so varied in its scene-painting that it shocked contemporary audiences. Plus it closes on a thrilling high note with that volcanic eruption (you don't normally link "French baroque" with "Krakatoa," but believe me, Rameau pulls it off). It's true that the work's second half is more variable in its inspiration, but still boasts ravishing moments, like the lovely quartet that pulls together the third entrée, and the elegant closing chaconne. One left the concert feeling that we really don't hear enough Rameau in this town.

The French genius was well-served by the period orchestra, whose playing was exquisite, and the chorus was in good voice, too - but the performance's chief jewel proved to be its soloists, who were all of outstanding quality. Local favorite Amanda Forsythe was most in period - the early eighteenth century - with her pearly, pure tone (her radiant comic acting was timeless, but that's another issue), while soprano Nathalie Paulin seemed to be replying to her from several decades later, but with a dusky richness so transporting that you didn't care. The men were just as strong, and likewise roughly in period - the powerful Sumner Thompson was probably the stand-out (in a passionate turn as a suicidal Incan that's probably the vocal highlight of the opera) but baritone Nathaniel Watson impressed with his eloquent lyricism too, and tenors Aaron Sheehan and Daniel Auchincloss more than held their own in their respective entrées. All in all, this was the most consistent set of soloists I've yet heard with Boston Baroque.

I wish I could give the same high marks to the accompanying dances, but these proved the production's sole weak point. Choreographed by the talented Marjorie Folkman - late of the Mark Morris Dance Group - they were much in her mentor's style, but exhibited little of his sense of musical structure and meaning, and so were lightly charming, but that's about it. Folkman was certainly constrained in terms of space, but you also felt an unnecessary constraint in her choreographic ambition; things got a bit more complex (and more genuinely lyrical) in the later entrées, but one still felt the dances weren't quite worthy of their accompaniment.

Meanwhile the clever direction by Sam Helfrich made witty use of what few (contemporary) props the staging could allow - and he even cheekily pulled the chorus into the action, too (they waved tiny national flags to introduce each change of locale, and even boogied a bit). In some ways Helfrich's ironic flair answered a question that's clearly on Pearlman's mind, i.e. how do you stage baroque works without turning them into museum pieces? Still, by the end of the evening some of the director's gambits had begun to seem superficial, and a few of his knowing jokes (like the predictable one about those "peace pipes") went on a bit too long; we began to perceive that something central to Rameau's vision - its truly wonder-struck exoticism - had somehow gone missing from the staging. This only argues, of course, for a fuller production from Pearlman at a later date. In the meantime we should be grateful for his tireless work in bringing this luminous score back into the repertoire.

Our local music scene is chock-a-block with worthy programs sporting titles like"Jewels and Discoveries" - and alas, usually a few of the gems in question turn out to be rhinestones.

So imagine my surprise when Boston Baroque's "Jewels and Discoveries" - which only saw two performances, last weekend - turned out to be solid Cartier from start to finish.  Conductor Martin Pearlman pulled together a program of brilliant obscurities, and his orchestra, chorus and soloists polished them to a dazzling sheen.  Tenor Keith Jameson was sidelined due to illness, but he was ably replaced (in Monteverdi's Combattimento di Tancredi e Clorinda) by the talented Aaron Sheehan, and everyone else seemed energized and more precisely on point than they've been in the recent past; or perhaps the difference was that conductor Pearlman, though he kept the pace sprightly (as always), only occasionally broke the usual baroque speed limits.  Whatever the reason, this was Boston Baroque at its finest - which is very, very good indeed.

Or perhaps the difference was simply that the program sparkled so consistently, and the singers and players themselves responded to that quality.  The concert opened with Dietrich Buxtehude's Heut’ triumphieret Gottes Sohn (Today God’s Son triumphs), an Easter cantata of surpassing grace and richness. An introductory sinfonia and fanfare led to a remarkably melodic chorale in which a series of soloists took pride of place - with particularly fine work coming from alto Martin Near and bass-baritone Ulysses Thomas. This was followed by two striking works of Monteverdi, Beatus Vir, a sublime setting of Psalm 112 which Pearlman gave his usual dancing buoyancy, and then what amounted to the centerpiece of the evening, Il combattimento Tancredi e Clorinda, a stunningly dramatic piece based on Torquato Tasso’s epic poem of the Crusades. In this Christianist potboiler, the hero Tancredi challenges a Saracen knight to battle at the gates of Jerusalem, not guessing that "he" is really a "she" (Clorinda, in fact, the woman he loves, surprise surprise!). Can you guess the rest? Probably - although I'm afraid these days we have to wince a bit at the final twist, in which the dying Clorinda, a Muslim, begs her beloved (and unwitting killer) to baptize her. Yuck.

Still, even this questionable bit of Christian triumphalism (Monteverdi was a priest, remember) is rendered with sublime delicacy (indeed, poor Clorinda's death is almost overwhelmingly poignant) and the rest of Tancredi e Clorinda is simply terrific. Monteverdi literally invented the tremolo for the piece (that's right, before Tancredi e Clorinda nobody had ever heard a tremolo), explicitly demanded very precise pizzicatos to convey the thwacks of the lovers' swords, and in general called for a wild dynamic that in its day was thought crazy. And to be honest, battle music really hasn't gotten that much better over the past four centuries - Tancredi e Clorinda still thrills, and the narration is a hoot, with the lovers' vows framed by "he said" and "she said" from the narrator, as if we were simultaneously listening to an opera and watching a silent movie. Both Tancredi and Clorinda were ably embodied by bass Bradford Gleim and soprano Mary Wilson, and Aaron Sheehan, though stepping in at the last moment, made quite the dashing narrator.  (He had to dash, as this was the one time in the program Pearlman's tempo approached a gallop.)

The second half of the concert, though still remarkable, never reached quite the same musical and dramatic peaks.  It opened with sacred music by Heinrich Biber, mixed with two of the same composer's Mystery Sonatas, violin pieces devised to convey the 15 mysteries of the rosary.  The psalm settings and the Agnus Dei Pearlman had selected were lovely (and gave soprano Teresa Wakim,  alto Thea Lobo and tenor Murray Kidd a chance to shine), but it was the sonatas that threw off a strangely  memorable fire.  Each of the 15 is tuned - or "distuned" - in a particular way (which is too complicated to go into here), which gives the instrument an eccentric timbre, and gives the violinist a headache, probably (because of the unusual tuning of the instrument, each piece has its own bizarre key signature, too).  Add to that the fact that both of the sonatas on the program ("The Crucifixion" and "Assumption of the Virgin") seemed fiendishly difficult, and you can imagine the challenge facing concertmaster Christina Day Martinson.  She seemed unfazed by all this, however (but then she never seems fazed), and, working with two separate violins, carried off the sonatas with spirited verve (indeed, the final gigue from the "Assumption" was almost dizzying).

The crowning glory of the program was literally a discovery - an early Gloria by Handel that was only authenticated a few years ago.  To be honest, though very beautiful, the piece almost felt like a bit of an anti-climax after the impressively knotty Mystery Sonatas  - luckily, however, Mary Wilson returned to carry it off. Ms. Wilson's voice is just about perfect for Handel - her tone is ripe with sun, and her phrasings so flexible they seem to almost ripple. By the end of her beguiling performance, any and all sense of anticlimax had been banished.

Martin Pearlman conducts Boston Baroque.

I try to hear both major local versions of Messiah every year, and there was one moment from this year's Boston Baroque edition that I will never forget - the duet for bass and natural trumpet in the third part. The piece begins, "The trumpet shall sound," but in the hands of trumpeter Robinson Pyle, the instrument actually sang, in tandem with Kevin Deas, the wonderfully rich bass who was essaying a famous passage from Paul's First Epistle to the Corinthians (the one establishing the doctrine of Christian resurrection). It's one of the loveliest melodies in all of Handel (and that's saying something), and Deas and Pyle had clearly worked together so that Pyle could closely follow the bass's ornamentation of the beautiful line, "We shall be changed." It was probably the best playing on natural trumpet I've ever heard in my life, as well as one of the most moving duets between a singer and an instrumentalist I've been lucky enough to witness. When Deas recognized Pyle at the close of the aria, it all but brought down the house.

Well, that was quite the high point - if only the rest of the performance had been so elevated! Not that Boston Baroque's version of this classic wasn't a lovely evening of music; it was. But it's beginning to feel a bit rigid in its eccentricity, and a bit unfocused in its attack. Ironically enough, the ensemble's conductor, Martin Pearlman, was among the first instigators of the revolution in Messiah performance that swept concert halls two decades ago.  Pearlman's idea was to bring the rhythms of dance to the musical drama, and to scale what had become a rigidly grand Victorian epic down to a nimbler, more intimate, and perhaps more human, experience.

In my opinion, this was all to the good - and today the stentorian Messiah that Pearlman was reacting against has pretty much become a thing of the past.  But I'm afraid over the years the conductor has become a bit rigid himself about a few things - he always favored brisk rhythms, for instance, but by now several pieces of his Messiah have gotten so fast that they aren't just dances but jigs (and they seem to shed more and more color and detail the quicker he takes them).  Plus Pearlman organizes his chorus into quartets, rather than in blocks (with all the sopranos together, then all the altos, etc.), which I imagine he thinks gives their sound a kind of blended transparency - which it does, up to a point.  But it also makes it harder for the singers to synch up the vocal melisma that is the backbone of many of Handel's melodies, and thus things sometimes turn blurry (particularly at the clips Pearlman often prefers).  Indeed, sometimes one distinctly felt in Pearlman's Messiah that he was putting the chorus at a disadvantage.  But I began to realize as I listened this year that a key difference between the Pearlman version and the Christophers version (over at  Handel and Haydn) is that Christophers, once a professional singer, views his orchestra as an extension of his chorus, while Pearlman, a keyboardist,  unconsciously sees his chorus as an extension of his orchestra - and why shouldn't they therefore just be able to sing as fast as he wants them to?

Oh, well - it's true not everything was too fast; Pearlman didn't dash through "He was despised and rejected of men," for instance.  Much of the performance was at an appropriately thoughtful pace.  And Boston Baroque had clearly corrected a problem they've had in the past: when Pearlman conducted from the harpsichord, the ensemble didn't begin to fray as it was once wont to do.  Still, the Boston Baroque orchestra isn't playing as cleanly as they might (though they now have that dancing lilt down pat).  And I was disappointed here and there with the soloists, even though I admire them all from previous hearings.  The luminous Amanda Forsythe had her usual pearly tone, but she didn't quite have the power, at least on Saturday night, that she has possessed in the past.  And counter tenor Matthew White, though he has a haunting timbre that's just right for many passages in Messiah, also scrapes a bit at the bottom of the role (which edges into contralto territory), and something about his voice didn't mix well with the tenor's in their duet.  Meanwhile said tenor, Keith Jameson, had some wonderful moments but also some tentative ones; the only member of the quartet, in fact, who sang with consistent authority was bass Deas, who outdid himself in that final duet.  It was a ringing reminder of the magic that Pearlman and company can still wreak with this immortal masterpiece.

The rose window at St. John the Divine in New York.

One of the great achievements of the early music movement - and Martin Pearlman's Boston Baroque in particular - has been its reclamation of Monteverdi from his lapsarian status as "that Italian guy before Bach." Last weekend's performances of the great composer's Vespers of 1610 only cemented that achievement. Monteverdi should have always been set somewhere on the boundary of Big Three territory, in that he brought pre-baroque forms such as the madrigal to the highest pitch they ever reached. But as early music pioneers have made us more and more familiar with his masterpieces, his status as a forger of new musical form has begun to seem to rival Beethoven's and Wagner's. Monteverdi of course all but invented opera, but nearly as original is his Vespers, a work unprecedented (at the time) in its scope and ambition, and one that all but defies categorization even today.

Of course on the surface, Monteverdi pretty much follows the normal structure of the Catholic rite of vespers, the evening ritual in which a series of sung psalms and antiphons leads to a Magnificat, a canticle dedicated to both the glory and humility of the Virgin Mary. In fact the Vespers was written, many believe, as a kind of audition for the top musical post at St. Mark's cathedral in Venice (Monteverdi got the job).

But it's hard to look at this piece as a résumé-builder, even for a genius; instead it sometimes feels like a kind of musical big bang, a technical, stylistic, and metaphorical explosion that in a way kick-started everything. Monteverdi leaps from motet to sonata to psalm and back again, all while somehow maintaining a sense of unity; he splits and re-forms his choir at will, and sometimes provides them with up to 10 separate vocal parts, all operating in synchrony(the piece also calls for seven soloists). What's more, Vespers is set all over the performance space, be it cathedral, chapel or concert hall - sometimes we can't even see the singers, as they're intended as voices of the cosmos, responding re-assuringly to the profession of human faith.

Now I don't believe in God, but Monteverdi certainly did (he eventually became a priest), and frankly, sometimes he almost convinces me of His existence. There are few more haunting moments in all of Western music, for instance, than "Duo seraphim," his duet for two angels floating in space, singing the glory of the Almighty (they're eventually joined by a third, who fuses with them into a single note when they praise the Trinity). Here conductor Martin Pearlman placed his tenors, Derek Chester and Aaron Sheehan, in the balconies of Jordan Hall, to thrillingly plaintive effect, and Monteverdi's evocation of the mystery of God's presence gripped us not only as great music but also as great theatre (and maybe even great architecture).

Alas, not all the soloists fared as well from the stage itself. Mr. Pearlman had clearly instructed his singers to strip their styles down toward pure-tone singing, and so I missed some of the vocal richness I expected from Mary Wilson and Kristen Watson, who both shone to better advantage in operatic roles with Boston Baroque earlier this year; indeed, of Pearlman's soloists (almost all of them familiar from earlier programs) I felt only baritone Donald Wilkinson was operating at his best. And the wind section, though fine in unison, sometimes got a little ragged when each instrument was exposed for long stretches. Likewise the height of Monteverdi's polyphony - those 10-part-plus sequences - didn't always feel entirely coherent.

But this is, admittedly, an incredible challenge, and any roughness here may have been partly due to understandable opening-night coordination issues; at any rate the chorus generally sounded superb, and nowhere more beautiful than in Monteverdi's concluding Magnificat, one of the most touching ever written. And Pearlman's mastery of the total arc of the Vespers was always and everywhere evident. The piece calls for a high degree of editorial intervention; much of the instrumentation is suggested but not pinned down, the position of some motets is disputed, and precisely which antiphons should be included is never specified (Pearlman took his from the Feast of the Assumption, certainly an appropriate choice). It's no secret that Pearlman's decisions on these and other key points have led to a version that many consider "the" Vespers of our time (it's already won a Grammy). Certainly the "Pearlman version" limns every - sometimes contradictory - facet of the piece: its intimacy and its grandeur, its period "feel" and yet its strange sense of timelessness. His exploitation of every nook and cranny of Jordan Hall was also brilliant, and only makes me long to hear this version in New York on March 6, when Boston Baroque will bring the Vespers to the Cathedral of St. John the Divine, surely a close-to-ideal venue for hearing Monteverdi's music of the spheres.