musica Dei donum
Johann Sebastian BACH (1685 - 1750): Mass in b minor (BWV 232)
[I] Lydia Teuscher, Ida Falk Winland, soprano;
Tim Mead, alto;
Samuel Boden, tenor;
Nal Davies, bass
Dir: Jonathan Cohen
rec: Oct 6 - 8, 2014, Tetbury (Gloucestershire), St Mary's Church
Hyperion - CDA68051/2 (2 CDs) (© 2014) (1.54'42")
Liner-notes: E/F/D; lyrics - translations: E/D/F
Cover, track-list & booklet
[II] Hannah Morrison, soprano;
Esther Brazil, mezzo-soprano;
Meg Bragle, Kate Symonds-Joy, alto;
Peter Davoren, Nick Pritchard, tenor;
Alex Ashworth, David Shipley, bass
Monteverdi Choir; English Baroque Soloists
Dir: John Eliot Gardiner
rec: March 28 - 31, 2015, London, LSO St Luke's
Soli Dei Gloria - SDG722 (2 CDs) (© 2015) (1.45'56")
Liner-notes: E/D/F; lyrics - translations: E
Cover, track-list & booklet
[III] Carolyn Sampson, soprano;
Anke Vondung, contralto;
Daniel Johannsen, tenor;
Tobias Berndt, bass
Gächinger Kantorei Stuttgart; Freiburger Barockorchester
Dir: Hans-Christoph Rademann
rec: Jan 28 & 31/Feb 1 - 3, 2015, Stuttgart, Liederhalle
Carus - 83.315 (2 CDs; DVD) (© 2015) (1.56'00" / 38'32")
Liner-notes: E (abridged)/D; lyrics - translations: E/D
Cover, track-list & booklet
[IV] Julia Doyle, soprano;
Alex Potter, alto;
Daniel Johannsen, tenor;
Klaus Mertens, bass-baritone
Choir & Orchestra of the J.S. Bach-Stiftung St. Gallen
Dir: Rudolf Lutz
rec: 2016, Basel, Studio SRF
J.S. Bach-Stiftung - B384 (2 CDs) (© 2017) (1.40'10")
Liner-notes: E/D; lyrics - translations: E/D
Cover, track-list & booklet
[V] Maria Keohane, Joanne Lunn, soprano;
Alex Potter, alto;
Jan Kobow, tenor;
Peter Harvey, bass
Dir: Lars Ulrik Mortensen
rec: May 21 - 25, 2013, Copenhagen, Garnisonskirke
CPO - 777 851-2 (2 CDs) (© 2015) (1.43'35")
Liner-notes: E/D; lyrics - translations: E/D
Cover, track-list & booklet
Johann Sebastian Bach's Mass in b minor is generally considered one of the greatest monuments in music history. And 'monumental' it is indeed, not only because of the scoring for up to eight voices and a full range of the then current instruments, but also because of its proportions and architectural character. One of the most remarkable features of this composition is that it is structured with an almost mathematical precision, and nevertheless is full of expression.
Bach started to compose this mass in the 1730s; its first version, dated 1733, has the structure of a missa brevis, consisting of Kyrie and Gloria. This was a quite common form which appears in the oeuvre of various German composers of the 17th and early 18th centuries, including Johann Sebastian himself (BWV 233-236). However, the size of this missa brevis is such, that one may wonder whether it was written for actual performance. Bach sent this work to Dresden, hoping that it would result in an appointment as Kapellmeister. It was not Bach's ambition to move to Dresden to actually take on this post, but the honorary title of Kapellmeister would strengthen his position in his regular quarrels with the authorities in Leipzig.
In its final form, mostly performed in our time, it is one of the last works in Bach's life: in his last years he extended and reworked the existing sections at the same time he worked at his Kunst der Fuge. And as Bach was increasingly under the stress of his deteriorating health, one may assume that he didn't expect it to be performed during his life - or ever, considering the fact that music of the past was seldom performed in those days. This could lead to the conclusion that Bach didn't have a particular performance in mind, and that he was merely aiming at leaving a musical testimony of his art to the world. As in the later stages of his life musical taste and style started to change he must have felt he was the last representative of an era which was soon to be gone for ever.
As no performance during Bach's lifetime is documented, and consequentially there is no performing material to go by, it is hardly possible to decide how this work should be performed. This leaves performers with several options, especially in regard to the number of the vocal and instrumental forces. The recordings under review here bear witness to that.
Four of the conductors have opted for a performance with four or five soloists, a choir and an orchestra. The exception is Lars Ulrik Mortensen, who follows the theories of Joshua Rifkin and Andrew Parrott, that Bach usually performed his sacred vocal works with solo voices, only in some cases with additional ripienists. John Eliot Gardiner's recording differs from the other choral performances in that his soloists participate in the tutti sections. This is in line with of the main features of sacred music of the 17th and early 18th centuries: all music was basically ensemble music, and there was no formal division between soloists, choir and orchestra.
From a historical point of view a performance with more than two voices per part (as in Mortensen's recording) is fully legitimate. However, Gardiner, Lutz and Rademann use a choir of between 30 and 35 voices, and that seems very questionable. It is highly unlikely that in Bach's time the number of singers exceeded 16 to 20 - the size of what in our time is called a 'chamber choir'. Raphaël Pichon, in his recording of the version of 1733, uses a choir of 25 singers. In his liner-notes he refers to research concerning the size and line-up of the vocal and instrumental forces at the court in Dresden. Assuming this information is correct, that cannot be used as an argument in favour of a choir of such - let alone even larger - size in a performance of the mass in its final form, as this was not intended for a performance in Dresden. It is even unlikely that the first version was ever performed there.
These historical considerations are certainly meaningful, but not necessarily decisive for the assessment of the various recordings. Let's have a look at each of them in more detail.
Jonathan Cohen's line-up is in accordance with the conventions which have established themselves in historical performance practice: five solo voices, a choir of four singers per part and orchestra. The latter includes six violins, two violas, two cellos and double bass. As I already indicated, the soloists do not participate in the tutti. This may explain why the solo episodes in 'Confiteor unum baptisma' are sung by members of the choir rather than the soloists, and the tenors of the choir sing the passage "Et iterum venturus est" from 'Et resurrexit tertia die'.
In this performance the tutti episodes suffer from a lack of transparency; the text is also not always clearly intelligible. That is not so much due to the size of the choir, but more the result of the vibrato in the choir. However, the size of the ensemble obviously doesn't make things any better. Overall I am not impressed by the way the tutti sections are performed anyway. Apart from the transparency and vibrato issues, the tempi are mostly rather slow. The opening Kyrie eleison is a striking example: Cohen needs 10'31", whereas Mortensen - at the other end of the scale - takes only 9'01". Moreover, the singing is largely legato and there is little dynamic differentiation in the instrumental parts. As a result it is rather dull, and that is a word I never expected to use in regard to a performance of this masterpiece. Many other tutti sections are not much better, I'm afraid. 'Et incarnatus est' is one of the more satisfying parts.
The solo sections are a bit of a mixed baggage. The tempi are mostly satisfying, and the obbligato instrumental parts are nicely played. Unfortunately the soloists usually spoil their solos by an incessant and sometimes rather wide vibrato. That goes especially for the two sopranos, Lydia Teuscher and Ida Falk Winland. 'Christe eleison' and 'Laudamus te' are disappointing, and in 'Domine Deus' Teuscher uses more vibrato than Samuel Boden. Here the balance is too much in favour of the soprano. Neal Davies it too pathetic in 'Quiniam tu solus sanctus', sometimes even tending to the operatic. 'Et in spiritum sanctum' is better, but is spoilt by his vibrato. The best solos come in the last section of the mass. Boden delivers a fine performance of the Benedictus and Tim Mead comes up with an expressive interpretation of the Agnus Dei. One wonders, why Jonathan Cohen, listening to these solos, did not notice that the strong reduction of vibrato - in the case of Mead it is almost absent - results in performances which are so much more satisfying and leave more opportunity to let the text speak.
John Eliot Gardiner recorded the Mass in b minor more than 30 years ago. As from 2000 he recorded Bach's complete sacred cantatas and also turned again to some of his larger-scale works, it was only to be expected that he would record the mass another time. Although he largely follows the same concept as in his previous recording - soloists, choir and orchestra - and uses the same ensembles, there is one important difference: this time the solo voices are fully integrated in the Monteverdi Choir. As the latter seems a flexible ensemble I can't tell whether the soloists have joined the choir for this occasion or whether they are members of the choir, who here take care of the solo parts. Whatever is the case, here Gardiner is in line with the performance practice in Bach's time. That is different in regard to the size of the choir and the orchestra. The Monteverdi Choir comprises 35 voices: thirteen sopranos (divided into two parts), nine altos, seven tenors and six basses. This is certainly more than any choir in Bach's time. It seems also highly unlikely that Bach may ever have used an orchestra of twelve violins, four violas, three cellos and two double basses, probably not even in a work for double choir, such as the St Matthew Passion.
Two further issues need to be mentioned. First, it is common practice these days to use German pronunciation of Latin in German sacred music. It is remarkable and regrettable that Gardiner sticks to the old habit of pronouncing Latin in the Italian manner. Second, the acoustic is rather dry, and that has a negative effect on the sound of in particular the tutti parts, which I find often rather dull.
That is partly due to the acoustic, but not entirely. I admit that I never really liked the Monteverdi Choir, partly because of the audible vibrato in all the voices. That is much better here, although not entirely absent. Overall the text is pretty well intelligible. However, in my ears there is a lack of depth and colour. The tempi are generally more satisfying that in Cohen's recording, but even so I haven't heard many tutti sections that I found really convincing. Tempo and articulation in Kyrie I are better than with Cohen, but here again I noticed a lack of dynamic accents in the instrumental parts. In contrast, there are episodes which are sung piano, and some passages staccato (around 4'45" and 8'30"). It is not nice to hear and I don't see any reason for it. The entrance of the low voices in the second Kyrie is not very subtle and this section is again damaged by staccato singing. Some of the tutti episodes are a bit superficial, such as 'Cum Sancto Spirito' and the opening of the Credo.
The arias and duets are the most satisfying parts of this performance. Hannah Morrison is responsible for the most soprano parts, and she sings them admirably. The duet 'Christe eleison' is delightful, and here the voices of Morrison and Kate Symonds-Joy are a perfect match. In 'Domine Deus' there is a good balance between soprano (Morrison) and tenor (Peter Davoren), and the same goes for 'Et in unum Dominum', in which Morrison is joined by Meg Bragle. The latter uses a little more vibrato in the Agnus Dei, but within the limits of what is acceptable. Nick Pritchard does well in the Benedictus, and the two bass arias come off equally well, even though I don't like David Shipley's voice very much (which is a matter of taste).
The recording by Hans-Christoph Rademann has something special. The first disc comprises the Kyrie and Gloria, as is the case in the other recordings. However, these sections are performed with the 'Dresden parts'. It is claimed that this is the first recording to do so, but I already referred to Raphaël Pichon's recording of the 1733 version, which was based on the latest publications of the material of the different versions of this mass. Whether that includes the material that was published by Carus in 2014, edited by Ulrich Leisinger, also the author of the liner-notes to Rademann's recording, is for me impossible to check. Anyway, this makes this production very interesting. That said, if one listens to this work in its entirety, one in fact listens to a piece of music which never did exist. The two sections which Bach later most drastically reworked - the duet 'Domine Deus' and the bass aria 'Quoniam tu solus sanctus' - are included on the second disc in the versions usually performed these days. In addition we get the first version of Sanctus and Pleni sunt coeli, which dates from 1724, and was originally scored for three sopranos, alto, tenor and bass with orchestra.
The Gächinger Kantorei Stuttgart consists of 32 voices: twelve sopranos, seven altos, six tenors and seven basses. The Freiburger Barockorchester is a little smaller than Gardiner's English Baroque Soloists and includes nine violins, three violas, two cellos and two double basses. As we have seen, the soloists don't participate in the tutti. The solo parts are the most disappointing aspects of this performance. The soprano and alto solos as well as their duets are largely spoilt by a quite wide vibrato; the same goes for the bass arias. Anke Vondung is the worst in Agnus Dei, which is a real shame as she has a nice voice and her performance is certainly not without expression. In 'Christe eleison' the balance between the two voices is unsatisfying, as Vondung's contribution is underexposed. The same happens in 'Domine Deus'; this duet is also problematic as Simpson uses quite some vibrato and Daniel Johannsen very little. He is the best of the soloists; he sings the Benedictus very well, albeit a bit too restrained; a little more firmness would not have been amiss. Like Vondung Tobias Berndt has a nice voice which is well suited for this kind of music, but I find his incessant and pretty large vibrato unbearable.
The tutti sections generally come off better, but I am not that impressed by the choir's sound, which I often find rather dull and flat. It is probably not without reason that since Rademann made this recording he decided to put together a new ensemble of voices and instruments, called the Gaechinger Cantorey. In the opening Kyrie the articulation in the instrumental parts is quite good, but dynamically a bit undifferentiated. The tempo is rather slow; to my surprise the second Kyrie is much faster, which I find hard to explain. 'Et incarnatus est' and 'Crucifixus' are sung rather well, but for some reasons don't really make a lasting impression. I am more enthusiastic about Sanctus and Pleni sunt coeli, which are given exuberant performances. The tempo is the fastest of the five recordings, and to my ears this sounds quite convincing. As one may expect from the Freiburger Barockorchester, the instrumental contributions are excellent.
The J.S.Bach-Stiftung St. Gallen in Switzerland regularly releases live recordings of cantatas by Bach, performed once a month in the Evangelische Kirche in Trogen. In addition they produce studio recordings of the large-scale sacred works; in 2012 the St Matthew Passion was recorded. In 2016 Rudolf Lutz recorded the Mass in b minor, in what is the most common line-up in this Bach project: soloists, choir and orchestra. The choir consists of 33 voices: twelve sopranos, seven altos, six tenors and eight basses. The soloists don't participate in the tutti. The orchestra includes eleven violins, four violas, three cellos and two double basses.
The choir may be too large from a historical point of view, the transparency of the tutti sections is remarkable and much better, for instance, than in Cohen's recording, although the latter's choir is smaller. That is largely due to a lack of vibrato. Moreover, the acoustic is pretty much ideal: it allows the music to blossom, but there is not too much reverberation. The result is much more satisfying than Gardiner's recording. Overall the choral sections have a strong amount of expression and intensity. The opening Kyrie is a good example: Lutz is a bit slower than Mortensen but his tempo is entirely convincing, also thanks to a good articulation and a fine realisation of the rhythmic pulse. There are effective dynamic accents in the vocal parts, but not enough in the instrumental parts. 'Qui tollis' is one of the best tutti sections, and 'Et incarnatus' and 'Crucifixus' receive incisive performances. The orchestra considerably contributes to the lasting impression of these sections.
That also goes for the solos, as the obbligato instrumental parts are brilliantly played. Overall the four soloists deliver excellent contributions. One could argue that it would be more correct to use two sopranos, but 'Christe eleison' is given an outstanding performance by Julia Doyle and Alex Potter. The former sings 'Laudamus te' exactly right, and Alex Potter excels in the Agnus Dei. I am slightly disappointed about Daniel Johannsen in the Benedictus, who does better in Rademann's recording. He is too restrained here as well, but also uses a bit too much vibrato. He is entirely convincing in 'Domine Deus', though, and his voice blends perfectly with Julia Doyle's, as do her voice and Potter's in 'Et in unum Dominum'. If these recordings were in direct competition to each other and there would be a contest between the basses - both is not the case, obviously - Klaus Mertens would win hands down. His two arias in Lutz's recording are a joy to listen to; he just does everything right. One could regret that the arias are not split among two basses, as Bach indicated (and that also goes for the other recordings reviewed here, except Gardiner), but if they are sung as by Mertens, I don't complain.
There are some issues here: apart from the size of the choir and the orchestra and the division between soloists and choir, I also mention the short introductions of harpsichord and organ respectively at the start of both discs, preceding the Kyrie and the Credo respectively. This practice is also applied in some of the cantata recordings. This seems artificial and I don't see any need for it. However, musically speaking there are hardly any weaknesses in the musical realisation of the interpretational concept Rudolf Lutz has chosen.
Lastly, Lars Ulrik Mortensen. I already noted that he is different in regard to the line-up. A quintet of soloists sings the arias and duets as well as the solo episodes in the tutti sections. In the tutti they are joined by five ripienists. The instrumental ensemble is also relatively small, with eight violins, two violas, two cellos and a double bass. This combination results in a transparency which can hardly be achieved with a larger line-up. This also allows for some swifter tempi, which don't go at the cost of a clear articulation.
It is hardly surprising that the tutti sections come off better here than in most of the previous recordings; Lutz is the exception here. Mortensen has opted for a fast tempo in the first Kyrie, which to my ears is entirely convincing, although not the 'correct' tempo. It certainly does not result in superficiality. In the second Kyrie this performance stands out in its rhythmic suppleness. The vocal parts are divided over soloists and tutti. We notice the same features in 'Et in terra pax'.
The tutti sections are also the most convincing parts of this particular performance. The contributions of the soloists are varied. Both Maria Keohane and Joanne Lunn are in superior form in 'Christe eleison', and the latter is responsible for a very fine performance of 'Laudamus te'. I am less impressed by the duet 'Domine Deus', and that is mainly due to Jan Kobow, who sings with a not very wide, but clearly noticeable vibrato. Here Maria Keohane also uses vibrato more than either of the sopranos does elsewhere. Kobow is also a bit disappointing in the Benedictus, due to his vibrato, but also because his voice is rather weak. Alex Potter is magnificent and moving in the Agnus Dei, and blends perfectly with Keohane in the duet 'Et in unum Dominum'. I have never been a big fan of Peter Harvey, but recently I heard some good things from him. Here there is little wrong with his performances, but I find them not very interesting, and here again the usual problem of a (modest) vibrato manifests itself.
This recording has many qualities, especially in regard to the performance of the tutti sections. The small line-up results in a kind of intimacy, which is comparable with that in Gardiner's performance. But in the latter case it is mainly the result of a dry acoustic, which makes the performance as a whole sound rather dull. That is different in Mortensen's performance, which is much more vibrant; the acoustic of the Garnisonskirke in Copenhagen suits this work very well. Despite its qualities it doesn't fully satisfy me, especially because of some of the solo sections. On the other hand, the obbligato instrumental parts are excellently played, for instance by the leader of the orchestra, Fredrik From, in 'Laudamus te'.
These recordings are not in direct competition, as there are so many recordings in the catalogue to choose from. Not every performance recorded here is a valuable addition to the discography of the Mass in b minor. Rademann's recording is interesting, because of the various versions and the inclusion of material which can seldom be heard. Musically speaking it is not that appealing. Although I am not entirely saisfied with Mortensen's performance, it is one of the best in the 'one-voice-per-part' section. Despite my reservations in regard to in particular the size of the choir Rudolf Lutz's recording is quite impressive and certainly belongs among the highest division in the choral performance section.
Johan van Veen (© 2018)
J.S. Bach-Stiftung St. Gallen
Gächinger Kantorei Stuttgart
Monteverdi Choir & English Baroque Soloists