Bach: Cantatas Bwv 130 19 14 [SACD]

The Feast day of St. Michael and all Angels--known as Michaelmas Day in the Lutheran Church calendar, and the celebratory day of the autumn equinox, inspired J.S. Bach to write three of his most ingenious Cantatas--BWV 19, 130, and 149. The texts chosen by Bach derive mostly from the Book of Revelations 12: 7-12, and center around the final battle waged by the Archangel Michael against the malevolent serpent of the garden, represented by an ancient reptilian dragon.

The excellent period ensemble Montreal Baroque, led by Eric Milnes, performs all three cantatas on this release: which makes up volume 2 of a projected cantata cycle. While scholars often consider these cantatas to be among Bach's finest, with their picturesque musical settings, curiously they remain among Bach's lesser known choral masterpieces. At present there is only one other recording in the catalogue where all three cantatas have been assembled on a single disc, and that comes in volume 7 of John Eliot Gardiner's complete survey: Cantatas, Vol. 7. Unlike Milnes, Gardiner includes the one movement BWV 50, a stunning work for double choir, which is thought to be part of a now lost or unfinished 'Michael' cantata.

The other main difference between Milnes and Gardiner is that Milnes has chosen to perform the cantatas one-voice-to-a-part: which means that the choral movements are sung by a quartet of soloists, rather than a large choir--which in Gardiner's case is the twenty person Monteverdi choir. Milnes' decision to use only four soloists is based on the scholarship of Joshua Rifken, who several decades ago proposed the then radical idea that it was normal practice in Bach's Germany for a quartet (or quintet) of highly skilled soloists to sing the choruses, rather than a full sized choir. This premise served as the basis for a number of critically acclaimed recordings by Rifken, as well as a more recent book entitled "Bach's Choral Ideal." While Rifken's ideas have gradually gained a wider acceptance among leading period ensembles, the early music movement remains divided.

The first leading early music figure to embrace Rifken's argument was the British scholar and conductor Andrew Parrott--who, like Rifken, also made acclaimed recordings, and wrote an influential book, "The Essential Bach Choir": The Essential Bach Choir. A list of other notable conductors to have since become persuaded includes: Paul McCreesh, Konrad Junghänel, Sigiswald Kuijken, John Butt, Marc Minkowski, Philipe Pierlot, and Eric Milnes, as well as the conductor less Purcell Quartet. I would also add the Dutch conductor Jos Van Veldhoven, even though, strictly speaking, Veldhoven does use additional vocalists, called ripienists, for 'poetic emphasis' where he sees fit. But otherwise his approach is one-voice-to-a-part. Indeed Veldhoven has stated, "I think of this music not in terms of a chorus, but in terms of five highly expressive soloists."

Rifken's premise has also found a wider acceptance among leading scholars of the period. In his 2001 book, "Music of the Baroque," David Schulenberg writes, "Although the exact make up of Bach's vocal forces has been a matter of debate, it appears increasingly likely that most of Bach's vocal works were composed for a 'chorus' comprising a single singer per part. Orchestral parts, too, were rarely doubled, except for the violin and continuo lines. Thus what many listeners have come to regard as massive choral movements for large choir and orchestra are in fact examples of chamber music for vocal soloists and a small instrumental ensemble." Further support has come from studies into the performance practices of Bach's contemporaries, such as Kerala Snyder's 1987 book, "Buxtehude: Organist in Lübeck," where Snyder shows that the composer who so impressed the young Bach, normally intended his four-part choral compositions for four soloists; and from Jeanne Swack, whose books on composer Georg Phillip Telemann present evidence that Telemann also used one per part scoring in his cantatas.

The basis for Rifken's argument begins with the question of whether or not it was normal practice in Bach's time for singers to share part copies. Since the majority of Bach's choral works have come down to us with only one existing copy per part, Rifken concluded that only one singer sang that part. Scholars had previously thought that two or three singers read from each copy, requiring a minimum of eight singers: with the choir consisting of four or five principal singers, who were then doubled or tripled by ripienists. Before Rifken, a typical Bach choir might consist of twelve to sixteen singers, who could then be expanded to eighteen to twenty-four singers for the larger choral works, which can require up to eight vocal parts, or a double choir. Of the two hundred or so Bach Cantatas that have survived, remarkably, there is only one sacred cantata that requires a double choir, and that is the BWV 50 fragment. The rest are four or five-part vocal works, which suggests that Bach regularly wrote his cantatas for a smaller choir.

Like Rifken, Parrott also sees the debate as hinging on the "presence or absence of ripenists," arguing that ripienists were used only where Bach writes it on the title page, which occurs in about 10% of Bach's vocal works. Nor does Bach otherwise provide any performance cues or notations for when ripienists should enter and exit, though he does give plenty of other indications regarding articulation and tempo. For Parrott this was compelling evidence that Bach wanted ripienists only when he specified them.

Rifken further maintains that the decision of whether to use ripienists was considered optional in Bach's day. He writes, "Given the inherent dispensability of most ripieno parts, Baroque composers often left it up to the performer whether or not to use them; not infrequently, they signaled this option through a kind of shorthand. A cantata by Bach's predecessor Sebastian Knüpfer, for example, called for five concerted voices, five vocal ripieni, six stringed instruments, and five brass. On the title page, the summary of forces reads "for 16 or 21" - meaning that you could perform it with just five essential voices and the eleven instruments, or that you could add the five ripieno voice parts according to circumstance or taste." In other words, their use varied, and likely depended on such factors as quality of obtainable singers, available monies, the performance venue, the piece of music being performed, and the composer's notations.

However, in Bach's case, who seldom gave any indications for ripienists, and who also rarely doubled his orchestral parts, except on the violin and continuo lines--or on the top and bottom musical lines--Parrott's argument makes a good deal of sense. Because by rarely doubling the instruments in the middle section of his orchestra, it shows that Bach was keenly intent on preserving the contrapuntal clarity of the more congested middle range of his score. In other words, he wanted all those intricately detailed musical lines in his music to be clearly heard. Which, of course, is fully in keeping with Parrott's claim that Bach should also wish to severely limit the use of ripienists in his choral works as well. The cavernous acoustics of Leipzig churches would play a further role in this, as their echoing interiors would have given Bach an even greater need to consciously limit and balance his musical forces.

Yet there is historical evidence that puts Rifken's 'quartet' theory into question. In the 1737 Der Critische Musikus, Johann Adolph Scheibe writes that a double quartet of eight singers, not four, was the working minimum of Bach's day:

"A complete choir of singers, for use both in theatre and in church and chamber, cannot consist of fewer than eight persons. These I break down in the following way: first a pair of sopranos, [then] a pair of altos, a pair of tenors, and a high bass or so called baritone and finally a low bass. But these eight persons must all be skilled people. However, as the choruses could still be filled out, one would quite easily be able - at courts - to add chapel boys - [or] in towns - some schoolboys."

Thus according to Scheibe--who significantly had been a pupil of Bach's at St. Thomas School--a 'complete' choir in 1737 consisted of eight highly skilled singers; two for each part, plus additonal singers in the form of highly trained 'chapel boys' for special performances at courts, or less skilled 'schoolboys' for muncipal, and possibly outdoor performances.

There has also been some debate surrounding a letter that Bach wrote to the Leipzig City Council on August 23th, 1730, known as the "Entwurff" letter: where Rifken's critics--among them musicians Ton Koopman, Gustav Leonhardt, and scholar George Stauffer--insist that Bach was plainly asking the council for 12-16 singers to fill out his 1st choir at St. Thomas. While Rifken and Parrott argue that Bach wasn't asking for a set number for his choir at all, but instead requesting the number of singers needed to staff his 1st choir for the entire church season. Indeed Rifken compares Bach's request to a baseball team's full seasonal roster, rather than the number of players in the starting line up on game day.

Personally, I don't see why both sides aren't at least partly correct here--since different musical works, venues, and occasions would demand a varying number of singers. By asking for 12-16 singers Bach was most likely requesting the maximum number of singers he would need to fill out his 1st choir on special occasions, when large choral works requiring a double chorus and ripienists might be performed. At the same time, Rifken is surely right to argue that 12-16 singers represented Bach's full seasonal roster: because on most other "game" days, Bach's actual starting line up of singers woud have been less than 12-16, especially in regards to his regularly performed four part cantatas.

Nevertheless, Bach did on occasion conduct full scale ensembles of up to 30-40 musicians and singers. We know this from an eyewitness account passed down to us by one of Bach's former colleagues at St. Thomas School, Johann Heinrich Winckler. In 1765 Winckler quoted another former instructor at the school, Johann Matthias Gesner, as having been particularly impressed by Bach's ability to conduct a large ensemble of musicians all at once, recalling that Gesner had said to him, "...as [Bach] paid attention to all of them simultaneously and from this group of 30 or even 40 musicians [he would] nod to one of them with his head or indicate to another by stamping his foot, or threateningly, by using his finger, keep the third one on beat with the correct rhythm."

It must be remembered, however, that in early 18th century Germany such musical extravaganzas--which here evidently included both instrumentalists and choir--required additional monies. In 1728 even the largest Evangelical churches of Bach's day--that is, the best funded ones, were criticized by Johann Mattheson in Der Musicalische Patriot for routinely using very meager sized choirs compared to their Catholic counterparts in Italy and Spain. Indeed Mattheson called for vocal and instrumental ensembles of up to "30 to 40 musicians" as his ideal--hence, this wasn't standard Lutheran practice. Not to mention that such extravagance usually came directly out of a cantor's own pocket, and it is well known that Bach had continual money problems in Leipzig, and wasn't at all happy at St. Thomas. Therefore it is most unlikely that Bach conducted 30-40 member ensembles on any regular basis. Thus Gesner's account is probably a description of Bach conducting one of his larger choral works on some occasion when his entire 1st choir roster was required.

In strong contrast to Gesner's statement, there is also a 1721 document by Gottfried Ephraim Schneibel entitled, "Random Thoughts about Church Music in Our Day," where on the subject of choir size, Schneibel writes, "if each part is provided with one or at most two people who excel in what they do, then a choir is well appointed." Hence, according to Scheibel a "well appointed" or prudent choir in Bach's day consisted of four to eight excellent principal soloists--that is, one per part, who could be acceptably doubled, if need be: which further confirms that choirs of four soloists were definitely used. It is also most interesting to note that, as with Bach, Schneibel sets the maximum limit for a choir at sixteen singers.

One voice-to-a-part performance can also be found in the other great musical capital of Bach's day, the city of Dresden, where it was common practice, and where the first version of Bach's great Mass in B minor received its only performance during the composer's lifetime in 1733. In the booklet notes to his recent Dunedin Consort recording of this work, musician and scholar John Butt writes, "At the Dresden court, also, to which Bach increasingly looked the longer he stayed in Leipzig, much of the vocal music was performed primarily with soloists." Regarding the 1733 Dresden Bach performance, Butt continues, "As it stood in 1733, Bach indicated that the two sopranos together sing the top line, thus suggesting that this line was sung with doubled voices, the others without. But in 1731 this music (part of the town-council cantata BWV 29) had been furnished with ripienists in all parts. So Bach countenanced the same piece of music being sung with 8 voices in 1731, 5 in 1733, and 8 again in 1750 (in the 'Dona Nobis' at least)." (I should clarify that Bach composed the final "Dona Nobis Pacem" for a double choir: so, with 8 vocalists, it would have been sung by one singer per part.) Once again it would appear that the use of ripienists varied according to the specific occasion and venue, whether it was a church, a town venue, or at court.

Thus by taking Schneibel, Scheibe, Buxtehude, Telemann, Bach, and the common musical practice in Dresden at the time as our guides, we can reasonably conclude that Bach typically used four to eight singers for his weekly cantatas (except BWV 50), and no more than eight to sixteen singers in the larger works that required a double choir. With ripiensts being most likely added in the case of specially funded court or municipal performances, where, as Scheibe writes, extra boys might fill out those performances. It would therefore appear that the 'modern' practice of using large homogeneous choirs of eighteen to twenty-four singers and up, singing over rather sizeable orchestras, has more in common with the ample sized choirs used in Catholic Italy and Spain during Bach's lifetime, and to a much later 19th century ideal of choral grandeur found in the music of Berlioz and Verdi--rather than to the early 18th century Protestant Germany of Bach, Telemann, and Buxtehude.

And I am not solely referring to modern instrument performances either: Such imbalances can occur on period recordings as well, especially when their over-sized choirs (Gardiner, Koopman, etc.) sing with more vibrato than they probably should; or, when those choirs are recorded in overly reverberant acoustics. Granted it doesn't happen uniformly, but when it does, the unevenness of balances created does diminish the contrapuntal clarity of Bach's writing.

As one might expect, such imbalances never occur in the Montreal performances on this disc: where the four soloists and superb baroque ensemble achieve a near perfect sense of equilibrium; indeed the balances seem so consistently natural (and deftly handled), that there is never any sense of competing forces, or drowning out or blurring of the monumental complexities of Bach's score. Nor is there ever the sense of a 'conductor' manipulating those balances or effects in a showy or ostentatious manner. The sharper contrapuntal focus also allows the dramatic contrasts in Bach's score to seem crisper and more incisive. Hence, not only is there a clear historical precedent for Milnes' use of only four soloists, but there are also significant aesthetic advantages as well.

Another aspect of these scores that argues in favor of a greater overall transparency from choir and musicians is Bach's extensive use of violin 'slurs'--called 'graces' in the 18th century--at the opening of BWV 130. In his booklet notes, musicologist Bruce Haynes discusses how the function of a violin slur in Bach's time was different from the later Romantic Age. He writes,

"To a musician reading these notes in 19th-century mode, a series of slurs mean nothing more than an absence of articulation, they are a mere technical instruction for legato. But in the 18th century a slur of up to three or four notes was thought of as a grace, its function being to indicate an emphasis with a diminuendo. Thus it was an indication of legato playing only incidentally... the series of slurs are thus a code for a chain of separate accents, one at the beginning of each slur; not the modern single impulse with articulation, but three independent thrusts, like short pickups. Playing this way gives this passage a rather different meaning... "

Hayes concludes that the effect of the violins changes significantly "depending on whether the slur is perceived as an expressive device (as it generally was in the 18th century) or as a technical instruction (as in Modern Conservatory Style.)" Which means that when the slurs are played in "Modern Conservatory style" the violins can limit or bog down the requisite agility and rhythmic articulation of the rest of the ensemble: giving us a very different sounding Bach.

To better illustate this point, have a listen to how the violin slurs are played in the opening bars of Helmuth Rilling's modern instrument recording of BWV 130. It will be a real eye opener. Notice how Rilling's modern violins are more cumbersome than Milnes' period violins, and how their lack of agility--due to the purely legato effect--impedes the swiftness of the chorus and trumpets, making the music drag in comparison: Bach: Cantatas, BWV 130-132. And bear in mind that Rilling is considered to be one of the least pedantic conductors performing Bach's choral music on modern instruments today.

Yet nowhere in these cantatas is Bach's need for a greater agility and transparency from his performers more evident than in the magnificent fugally constructed opening of Cantata BWV 19. Here Bach dazzling depicts the decisive battle in what conductor Craig Smith has described as a 'vaulting high energy fugue." (Hayes likewise writes that the individual parts go "in every imaginable direction.") Famed musicologist Albert Schweitzer writes of this movement that,

"Bach launches a whole army of devils against the divine power... the first chorus paints the struggle of Satan and his host against the archangel Michael. The serpent-forms fling themselves upward in mighty contortions." But "at the words, 'Aber Michael bezwingt, under die schar, die ihn umringt, stüzt des Satans Grausamkeit." [But Michael Conquors, and the cruel host of Satan encircling him is cast down."], the motive is inverted, and the agitated and distorted mass fall precipitantly into the depths."

Here Gardiner's larger, more aggressive forces work to undeniably fuller effect: with the staggering virtuosity of the Monteverdi singers brilliantly capturing Michael and his angels on the upper line fiercely interlocked in a raging battle with Satan's reptilian hordes on the lower line: Bach Cantatas Vol. 7: Ambronay / Bremen. However, it can also be said that Milnes and his smaller forces bring a mind-blowing clarity to the movement: which, given that it is written in the most complex of musical forms--a fugue--is a significant advantage. In addition, by sounding less fiercely driven, Milnes' chorus also has the effect of making Michael and his angelic forces seem more self-confident of their eventual victory: which may be more in keeping with the confidently assured outlook of 18th century Lutheranism, and therefore Bach's own firm belief in the supremacy of God's dominion over good and evil.

As for the soloists, I much prefer Milnes' singers to Gardiner's, and not because Gardiner's are weak, but simply because Milnes' are even better. Monika Mauch, for example, makes the ideal Bach soprano: with her vocal range falling somewhere between a soprano and mezzo--rendering her singing a suitable approximation to the late pubescent young men of Bach's day, who sang this part, and whose voices broke later than today at around 15 or 16 years old. Moreover, her pitch perfect vibrato less singing and superb artistry bring a wonderful pure toned precision to Bach's score. The tenor Jan Kobow is also an ideal choice. His singing of the post victory aria in BWV 19--a duet for tenor and trumpet written in a siciliano rhythm--is quite memorable. In this moving aria a human implores the angels, with their leader Michael, to remain on earth after the battle has been won. To my mind it is one of the most profound arias that Bach ever composed.

Interestingly, of the four releases issued so far in the Montreal series, each corresponds to one of the 'quarter' days in the annual year, or the four Lutheran feast days associated with the precession. Volume 1, for example, is comprised of cantatas that Bach wrote for the birthday of John the Baptist on June 24, or the summer solstice, also known as Midsummer's Day. While volume 2 consists of cantatas that Bach wrote for the Feast of St. Michael and All Angels on September 29th, or the celebratory day of the autumn equinox. Volume 3 contains cantatas written for "Marie de Nazareth" that mark the Annunciation of the Blessed Virgin Mary on "Lady Day," or the vernal equinox on March 25th. Finally, volume 4--entitled "La Nativite"--is grouped together by cantatas that honor the birth of Jesus at Christmas, or the celebratory day of the winter solstice on December 25th.

1--Bach: Cantates 30, 7 & 167 [Hybrid SACD]
2--Bach - Saint Michel Cantates 130, 19, 149 / Mauch, Lee, Kobow, MacLoed, Montreal Baroque
3--Bach - Marie de Nazareth Cantates 147, 82, 1 / Mauch, White, Daniels, MacLoed, Montreal Baroque, Milnes
4--Bach: Cantatas Nos. 61, 122, 123 & 182 [Hybrid SACD]

At present there is only one other one-voice-per-part Bach cantata cycle in the makings, from Sigiswald Kuijken and Le Petite Band (though it's unclear whether the Purcell Quartet intends a complete cycle). While Kuijken is much further along in his cycle than Milnes, as I write he has yet to record the Michaelmas cantatas. Like the Montreal series, Kuijken's cycle is also being recorded in a hybrid SACD format (playable on conventional players). In terms of the quality of Kuijken's performances, some reviewers have complained about the inconsistency of his soloists on certain issues, which of course is a serious negative in OVPP performance. Neither have Rifken or Parrott recorded these cantatas either; though Parrott has made an OVPP recording of the BWV 50 fragment. Thus the offerings on this disc represent a special opportunity to hear all three Michaelmas cantatas in exceptional one-voice-to-a-part performances, that have been recorded in spectacular state of the art Atma sound--all in all, a rare delight. I eagerly await all future releases in this promising series.

*See my first comment for a listing of sources.

... you'd better take your vitamins and follow your exercise regimen, in order to live long enough to buy all 200+ cantatas performed by the ensemble Montreal Baroque! At this time, they're released four CDs with three cantatas on each of them, and they started recording at least five years ago. But the good news here is that they're "doing it right", giving each cantata its due in rehearsal polish and interpretation. Honestly, I wouldn't recommend any of "Complete Bach" editions currently on the market, chiefly because the cantatas are clunky and sound under-rehearsed on most of them. Here's my suggestion: make a check list of the cantatas, like a bird-watcher makes of species, and then shop around for individual performances. Use the 'samples', read the reviews, pick and choose, and eventually you'll have a collection far more worthy of the composer's genius than any of the complete sets. That's not a radical suggestion, is it? After all, you're unlike to listen to all the cantatas one after another, or even to hear all of them fully in one mere lifetime. Hey, old JS could write music faster than most of us can hear.

And if you do pick out the best of Bach's cantatas one by one, I'm sure you'll be choosing some of these recordings by Montreal Baroque. I've heard 'em all, my friends. There are numerous glorious performances on CD, conducted by Gardiner, Koopman, Suzuki, Leonhardt, Harnoncourt ... though I'd advise you to shun Helmut Rilling. Montreal Baroque conductor Eric Milnes doesn't yet enjoy the fame of those guys, but his orchestra is as proficient as any of theirs, and his soloists are especially adept at Bach's challenging almost-instrumental vocal lines. Monika Mauch, the soprano on all four MB releases so far, is inimitable, a rich dark timbre expressed with HIPP technical finesse. The soloists on the MB recordings are also the "chorus". In other words, these are one-voice-per-part performances. Don't snip and fret that the sound will be meagre, however. The balance between the singers and the orchestra is what matters most, and not the visual grandeur of a chorus you can't even see. THIS IS A RECORDING! And the current recording technology virtually guarantees that OVPP will sound fuller, cleaner, and less distorted than many-voices-per-part will, no matter how competent the large choir involved.

M. R. Simpson's previous review is exceptionally ample and interesting, and I have nothing more to add to it concerning the Saint Michael cantatas. I also strongly recommend Montreal Baroque's recording of the Mary of Nazareth cantatas 147, 82 & 1:
Bach - Marie de Nazareth Cantates 147, 82, 1 / Mauch, White, Daniels, MacLoed, Montreal Baroque, Milnes

› Go to Amazon.com to see both reviews  5.0 out of 5 stars