Thomas Braatz wrote (June 25, 2005):
BWV 22 & BWV 23 Details
This is important detailed information about BWV 22 and BWV 23 which may be more than some readers will want to know. Nevertheless this can be viewed as a fascinating investigation which probes rather deeply into the complications surrounding the performances of these cantatas during Bach's lifetime. If at times, the discussions seem repetitious, it is because they are presented in this fashion as various aspects and perspectives are explored.
In my summary translation, I select and choose what seems important to me.
From NBA I/8.1 Introduction
The entire volume prepared by Christoph Wolff (1992)
it contains facsimiles of
BWV 22 autograph score: end of mvt. 1, beginning of mvt. 2
BWV 22 copy of 1st page of score by Johann Andreas Kuhnau
BWV 23 autograph additional parts for 'Tenore' and 'Trombona 2'
BWV 23 autograph 'Violoncello' part (original C minor version) 1st mvt. on which Bach wrote: "NB. Eine 3 minor tieffer als Chorton."
The NBA prints out
BWV 22 in G minor
BWV 23 in both versions, both with 4 mvts., first in C minor with oboes then in B minor with oboi d'amore.
In the original scoring of BWV 23, the 'Organo' part player is looking at a figured bass part with 5 flats all the way through; while in the B minor version there are only 2 flats, but the continuo group has been expanded to include with the 'organo,' and 'violoncello,' the 'bassono' and 'cembalo' ['harpsichord'] parts as well.
Bach composed both BWV 22 and BWV 23 for his audition as part of his cantor application in Leipzig to be performed on February 7, 1723. Christoph Graupner had likewise completed his audition on January 19 with two cantatas (in contrast to the other applicants) performed in St. Thomas Church before the sermon and during communion. The NBA retains the order in which these two cantatas were performed, but this does not reflect accurately the dates of actual composition of each. Apparently Bach had composed both of these cantatas in Cöthen after having received the required texts from the Leipzig authorities.
We possess the autograph clean copy of BWV 22 which Bach prepared for the audition, but missing are the composing score and the original parts which were still listed as part of CPE Bach's estate (1790). Fortunately, Johann Andreas Kuhnau, Bach's main copyist in Leipzig, made a copy of this composing score. Bach's score, in the opening mvt., asks for a differentiation between 'Solo' and 'Tutti', a practice documented in the original parts from Graupner's January audition performances and which could be applied to mvt. 3 of BWV 23 as well. In contrast to the repeat performance of BWV 23 the following year on February 20, 1724, the repeat performance of BWV 22 shows no changes/corrections for the same date.
The performance circumstances and history for BWV 23 are much clearer than for BWV 22 since almost all of the original parts and scores are still available for inspection. There are 3 versions between 1723 and 1728/1731. Originally BWV 23 consisted only of the first 3 mvts. which were in C minor. This is the form it had when Bach brought his clean copy of the score and the original parts along with him to Leipzig. There are, however, no indications that Bach ever performed this 3-mvt. version in C minor as is. Most likely, upon short notice, he decided to add another mvt. after arriving in Leipzig. For this 4th mvt., he used an mvt. which he had already composed in figural style based upon the chorale: 'Christe, du Lamm Gottes.' However, in this expanded form, he had to transpose his original composition from C minor to B minor to accommodate the wind instruments which were playing colla parte in mvt. 4 and the oboes' transposition to a lower key (they had to be changed to oboi d'amore.) The continuo group was expanded as well to include harpsichord and a bassoon. This is the version that was performed in both 1723 and 1724. For a performance during the period circa 1728-1731, Bach returned to his original key, C minor, with the use of oboes as originally planned, but he had to drop the cornett and trombones in the final mvt. as well as revise the choral parts.
The NBA KB I/8.1 and 8.2 also prepared by Christoph Wolff (1998) has an introduction by Wolff explaining that this volume would have had to wait even longer before being published if he had not availed himself of help from Karla Neschke and Peter Wollny. I will skip the details of the provenance for both BWV 22 and BWV 23 and concentrate on the differences between the cantatas and any other versions undertaken by Bach during his lifetime.
Clean copy of score by Bach and a copy by Johann Andreas Kuhnau (probable date based on watermarks: 1724) Johann Gottlob Meißner is the probable source for some changes in the last 4 measures of mvt. 1, but there are some changes in Bach's hand on Kuhnau's copy as well: an important one being a nota bene by Bach just before m 41 of mvt. 1 regarding the resolution of a deceptive cadence in the dominant before the resolution into the tonic occurs in m. 42. There is a note at the top of Kuhnau's copy: "NB. Dies ist das Probestück in Leipzig". According to a footnote at the bottom of p. 18 of NBA KB I/8.1-2 (primarily prepared by Christoph Wolff with assistance of Karla Neschke and Peter Wollny and was published by Bärenreiter in 1998), the attribution of the “NB. Dies ist das Probestück in Leipzig” to Anna Magdalena Bach was made by Philipp Spitta (Vol. II, p. 775) and probably remained relatively undisputed until Hans-Joachim Schulze, in his commentary to documents relating to Johann Sebastian Bach, specifically the Bach-Dokumente, Vol. III [BD III, Item 631, p. 3] which appeared in print in 1972 (also published by Bärenreiter) incorrectly attributed this note to Johann Friedrich Agricola and assigned the wrong date [1750 or after that date]. In the Bach Compendium Vokalwerke, Part 1 (Edition Peters, 1985), pp. 213-214, the note is again twice attributed to J.F. Agricola without any indication of even an approximate date. In 1998 when the NBA KB I/8.1-2 was being prepared, Hans-Joachim Schulze must have approached Christoph Wolff with the information [“Hinweis von Hans-Joachim Schulze” – there is no other citation source given] concerning the correct identification of the scribe, Samuel Gottlieb Heder, who had until that time been listed as Hauptkopist D [main copyist] and the date of this note was changed to “wohl um 1735” [probably around 1735].
[Thanks Dan Pater Hamilton, Ohio USA for his comment dated October 16, 2014]
Christoph Wolff discusses the unusual aspects for both Graupner and Bach with a cantata performed both before and after the sermon in his article: "Bachs Leipziger Kantoratsprobe und die Aufführungsgeschichte der Kantate Du wahrer Gott und Davids Sohn in the Bach-Jahrbuch 1978, pp. 78-94, and also in his "Essays' pp. 128-140. Bupon the texts involved, it seems more probable that BWV 22 was performed before and BWV 23 after the sermon.
The original composition score and original parts for BWV 22 are lost.
The librettist is unknown but textual similarities between BWV 22 and 23 and Graupner's audition cantatas a month earlier are apparent. This is not surprising since these texts were chosen by a committee and submitted for printing in advance of the performances given by both composers. This situation differed from the one Bach experienced in Halle in 1713, where he had to compose a cantata "in Klausur" (given very limited time in the same city where he could be observed in his lodgings.) Bach seemed to have had sufficient time in Cöthen to complete BWV 22 there before coming to Leipzig.
The main differences between the 1723 and 1724 performances of BWV 22 are in some slight changes in the chorale texts.
Hans-Joachim Schulze determined that, according to the minutes of the Leipzig City Council, Bach's application for the vacancy was accepted on Dec. 21. 1722, but he was not officially given an invitation to audition until Jan. 15, 1723. The time available to Bach for preparing both cantatas was 3 weeks. Both cantatas were prepared for the most part in Cöthen before Bach brought them with their original parts to Leipzig. Both cantatas received a repeat performance the following year in 1724 for which BWV 22 was repeated with very few changes.
Bach had used the differentiation of concertisten and ripienisten (solo vs. tutti) performance technique in his Mühlhausen City Council Inauguration Cantata, BWV 71. But here, in Leipzig, there was additional motivation to use these techniques as well:
1. Graupner had used them for his audition during the previous month [Perhaps this reflects an understandable caution when performing with a choir consisting primarily of pupils.]
2. In both of his audition cantatas, Graupner had also used colla-parte trombones which would help to keep the St. Thomas School Choir from 'getting lost' in performing difficult music. Bach similarly used this technique.
Since the specification of the continuo group for BWV 22 is rather sparse (only "Violoncello è Cont." marked on the score) as a result of the original parts having been lost, it appears reasonable to assume that BWV 22 would have had a larger continuo group similar to that used in BWV 23: 'Organo.' 'Cembalo,' 'Violoncello.'
It could well be that Kuhnau's copy of the score for BWV 22 served as a basso continuo part in 1724 since the figured bass appears in this score for mvts. 1 and 2.
The original parts are all in Bach's handwriting, except were noted below and are all notated in C minor. They are:
Set 1 (all with the same watermark)
5) Hautbois Ire
6) Hautbois IIdo
7) Violino 1mo
8) Violino 2
10) Violoncello (mvts. 1-4 are figured)
Set 2 (all with differing watermarks) notated in various transpositions and in Bach's handwriting unless otherwise noted:
11) Cornetto (transposed from E to A minor for the B minor version - only mvt. 4}
12) Trombona 1 (same as above)
13) Trombona 2 (ditto)
14) Trombona 3 (ditto)
15) Hautbois d'Amour 1 (notated in D minor - all mvts.)
16) Hautbois d'Amour 2 (ditto)
17) Violino 1 (notated in C minor - mvts. 2-4) Christian Gottlob Meißner, copyist
18) Violino 2 (ditto) Meißner, copyist
19) Violoncello (notated in C minor - mvts. 1-4) copyists: Meißner + unknown
20) Violoncello (notated in C minor - Johann Andreas Kuhnau mvts. 1-3, Meißner, 4)
21) Bassono (notated in Bb minor) - Kuhnau, all 4 mvts. Bach adds figures
22) Continuo (notated in A minor ) - Kuhnau mvts 1,2,4; Meißner mvt. 3, Bach's figures
Set 3 (only with a special watermark "MA", notated in C minor, all copied by Bach)
23) Soprano (4th mvt. only)
24) Alto (ditto)
25) Tenore (ditto)
26) Basso (ditto)
Bach's composing score for mvts. 1-3
Bach's score for mvt. 4 "Christe, du Lamm Gottes" used for creating the parts (Set 3) above.
The vocal parts for the B-minor version for the performances in 1723 and 1724: 4 vocal parts, and 4 wind parts (cornetto and trombones) for the performance in 1724. [The German original in the KB is not quite clear here.]
For comparison with BWV 23/4:
The original parts for the SJP BWV 245/40ii in the 1725 revised version.
The origin of the parts:
Mvt. 4 must be discussed separately from mvts. 1-3 since they have different origins.
In Cöthen, Bach copies the 1st set of parts (set 1 as listed above) from his autograph clean copy of the score.
In Leipzig, Bach decides to expand the performing forces and has the copyists, Meißner and Kuhnau create a 2nd set (doublets) - Set 2 (17-20) from the existing parts. In order to accommodate the version of the cantata conceived in Cöthen with the different conditions existing in Leipzig (Chorton pitch), Bach had to replace the original oboe parts with 2 Oboi d'amore. The latter new parts were not copied from the original oboe parts. (Details are given here in the KB)
The organ part in A minor (Set 2, 22) was created in Leipzig after the decision to expand the cantata and after the decision had been reached to perform the cantata in B minor. This transposition was undertaken in order to take into consideration the organ, cornett and trombones. A performance of BWV 23 in C minor would have required the latter instruments to play in B minor. The organ part was copied from the original autograph 'Violoncello' part (Set 1, 10.) On the latter violoncello part Bach indicated: "NB. Eine 3 minor tieffer als Chorton." ["mark well! Transpose this a minor 3rd lower than Chorton pitch."] It is highly probable that Kuhnau copied the bassoon part (Set 2, 21) from Bach's violoncello part (Set 1, 10) For a later performance of BWV 23 in C minor, Bach transposed this bassoon part from B minor to Bb minor as required by Chorton pitch. The only changes necessary for this was to change the key signature, the accidentals and the figured bass as needed.
The original score for this mvt. is lost and must have been of an older origin (Weimar?) CPE Bach had made a copy, probably from this source. There are parallels between the vocal parts and the SJP (BWV 245) mvt. 40ii and the colla-parte wind parts which are not found in CPE Bach's copy. This earlier level in the development of this mvt. is represented in these wind parts (Set 2, 11-14) and was probably used for the performance in 1724. The vocal parts for this level/stage are missing. It is possible that they were used as part of the performing materials for the 2nd version of the SJP (BWV 245/40ii) which uses this chorale as the final mvt. In any case, there is a close correspondence between the vocal parts of this version of the SJP (BWV 245) and the wind parts (Set 2, 11-14.)
Nothing remains from the vocal and colla-parte wind parts from the audition performance in 1723, but they must have existed since there is proof that the obbligato parts (Set 2, 15-22) were used in 1723. It is possible that the vocal parts which had been taken from BWV 23 and used in the 1725 performance and version of the SJP (BWV 245) and were lost along with most of the other parts of that particular version of the SJP (BWV 245).
A later level of development of mvt. 4 is represented in the vocal parts (Set 3, 23-26) in C minor. There is evidence of revision of the voice leading. The differences between these vocal parts and the colla-parte wind parts exclude the possibility that
both could have been used for the same performance.
The original and later performances of the BWV 23
The 3-mvt. original version
This version, with a 'Fine' [not 'Il Fine'] at the end of mvt. 3 in the autograph score, was intended to be performed along with BWV 22 as part of Bach's audition for his position in Leipzig. The watermarks clearly identify this version with other compositions (BWV 173a - composed for a Dec. 10, 1722 performance in Cöthen) of the pre-Leipzig period. There are also musical parallels between BWV 23 and musical structure and techniques used in Cöthen: the duet-structure of the opening mvt. as well as the ritornello form with inserted duos in mvt. 3. If we had the composition score for this version, it would have been possible to determine if a parody was involved, but otherwise there are no indications that this 3-mvt. Original version is not original. [translate: 'otherwise we can assume that this version is original.']
The 4-mvt. later version
In order to give this cantata with oboes a greater 'weightiness' [monumentality?] and at the same time to underline the connection to the communion during which it would be performed, Bach added to this C-minor (3-mvt.) cantata a chorale mvt. where he gave further support to the voices by using a cornett and 3 trombones which were in Chorton as the organ was.[Yoshitake Kobayashi has good reasons for doubting that the brass instruments (actually cornett and trombones) were already added in 1723 since he dates the existing parts as coming from 1724 and not 1723.] This meant that the pitch was a whole step higher than cammerton. Accordingly, the winds would have had to play the final chorale mvt. composed in G/C minor in F/Bb minor, a very unfavorable key for the cornett, although not impossible to play. But the short time that Bach had between his arrival in Leipzig and the performance on Feb. 7, 1723, probably meant that there would not have been time for many rehearsals. As a result, he had to take into consideration the customary vocal ranges of the singers as well as the wind players who would have to play a semi-tone lower. For this reason at the first performance in 1723, the cantata was performed in B minor. In this version, the wind instruments played the final chorale in E/A minor instead of F/Bb minor. Along with accommodating the performance possibilities/capabilities of the brass players, Bach resolved the differences between the usual pitch used in Cöthen and the one in use in Leipzig. The strings were tuned a semi-tone lower, so that no new parts had to be copied out for them. But the oboes could no longer be used since the lowest note in their range would be surpassed (C1 was the lowest note which they could play.) For this reason Bach used as a replacement the oboi d'amore. The bassoon part (Set 2, 21) is the only part notated in B minor sounding as written. (This part was later transposed to Bb minor with different notation for performances that took place between 1728-1731.) The existing organo and 3 violoncello parts point to a stronger, larger continuo group for this version.
The origin of the chorale mvt. "Christe, du Lamm Gottes"
It is probable that this figural chorale goes back to Bach's lost Weimar Passion dating from 1717. Bach may have taken along with him to Leipzig a score for a Passion to demonstrate/document that he had composed a longer work with such dimensions, but possibly also to arrange for a guest appearance on Good Friday, 1723 due to the vacancy created by the lack of an established cantor. [This never materialized.]
The 1724 Performance
According to a procedure/tradition followed in printing the Leipzig cantata texts, the cantata following the sermon was not included in the text booklets even though there were actual cantata performances in that time slot. For this reason, in 1724, the text for BWV 22 is printed, but not the text for BWV 23. However, there are no other indications existing beyond 1724, that these cantatas were paired in any subsequent repeat performances. It is not clear whether the 1724 performance of BWV 23 had only a harpsichord in its accompaniment or whether an organ was used. The key tonality, in any case, must have been in B minor. Also not clarified is whether the colla-parte cornett and trombones participated in this performance.
Performances after 1724
The next performance of BWV 23 can not have taken place before 1727 because there is clear evidence that other cantatas were performed on Estomihi in the intervening years. The vocal parts for such a later performance also have a watermark that even makes 1727 appear implausible. The vocal parts from 1723 were probably no longer available since they had been incorporated/used in the 2nd version of the SJP (BWV 245) (mvt. 40ii) in 1728. For such a performance of BWV 23, Bach had to have a new set of parts copied and this later set does show watermarks for the period 1728-1732.
These additional parts (Set 3, 23-26) not only help to date this later performance, but also give evidence regarding the type of repeat performance it was. There are notes which are different as indicated above that made it impossible for the colla-parte wind instruments to support the musical lines sung by the vocalists. They could not have been used for the same performance! This situation with the colla-parte parts had caused the transposition to B minor to become superfluous/unnecessary. At this later point in time, the cantata was then performed as it had been conceived originally in Cöthen, in C minor. This return to the original is documented in various ways: the part originally intended for the oboes in the chorale mvt. could now finally be used for the first time as the oboes replaced the oboi d'amore. In the autograph parts for the oboi d'amore, it is apparent that the chorale was included directly into their parts. For the performance in C minor after 1728, an organ part in B minor became necessary, replacing the original A-minor organ part. Bach did not create a new part, but simply took the existing B-minor version and overwrote it with the key signature for Bb minor. He did not even have to change the notes. All he had to do was to change the accidentals and the numbers for the figured bass.
Questions of Keys and Orchestration
The keys used in the versions of this cantata BWV 23 stand in a causal relationship to the orchestration used in each. The fact that Bach, in his final version of this cantata, returned to the original key of C minor makes it obvious that the change to B minor was undertaken more out of practical rather than for aesthetic reasons. The primary, decisive factor which caused Bach to change the key for the 1st performance in 1723 was that he had to take into consideration the cornett and trombones which were at Chorton pitch.
Variants in Orchestration
In the 3-mvt. original performance version of BWV 23 in C minor, Bach uses 2 oboes (Set 1, 5 and 6). Because Bach added the colla-parte wind instruments in Leipzig, he replaced the oboes with oboi d'amore. Fthe 1728 or thereafter performance(s), regular oboes are once again used. In Cöthen, Bach had prepared the vocal and instrumental parts with the exception of the string doublets and organo part (Set 1, 1-10.) But this version in C minor did not yet include the final chorale mvt. "Christe, du Lamm Gottes." In Leipzig, Bach decided to orchestrate this final mvt. With colla-parte wind instruments, using the same means of support that Graupner had used a month earlier. This made it necessary to replace the original oboe parts with oboi d'amore parts. From the performance materials existing, we can determine that Bach strengthened the instrumental support for this first Leipzig performance in 1723. Bach had two copyists, Kuhnau and Meißner, prepare string doublets, two violoncello parts and a bassoon part as well as an organ part.
The original oboe parts were never used in performance until the period between 1728 and 1731, when Bach returned to the original key of C minor. The C-minor version made the organ part in A minor unusable, because he needed a Bb-minor organ part for this. Bach himself took the existing bassoon part in B minor and by changing the key signature transposed it to Bb minor. The notes themselves did not need to be changed. For this later performance, Bach himself prepared new vocal parts for mvt. 4. It is rather improbable that colla-parte wind instruments were used in this performance.
The Keys used for the Performances of BWV 23
Parts notated as sounding as Vocal parts C minor B minor
Strings C minor B minor
Oboi d'amore D minor B minor
harpsichord B minor B minor
Continuo/Organ A minor B minor
Trombones 1-3 A minor B minor
Oboes C minor C minor
Vocal parts C minor C minor
Strings C minor C minor
Oboes 1-2 C minor C minor
Continuo/Organ Bb minor C minor
Oboi d'amore 1-2 D minor B minor
harpsichord B minor B minor
Continuo/Organ A minor B minor
Trombone 1-3 A minor B minor
In mvt. 3, there are indications to separate the soli from tutti (Concertisten and Ripienisten) in the bass and tenor parts.
End of NBA KB I/8.1-2 pp. 12-59
Based upon the most recent research and in consideration of all preceding research into this problem, Ulrich Prinz in his "J. S. Bachs Instrumentarium" [Bärenreiter, Kassel, Internationale Bachakademie, Stuttgart, 2005] on various pages gives the following points regarding BWV 22 and BWV 23:
BWV 22 Perf. Feb. 7, 1723, Score P 119 Printed text but original parts missing, hence no definitive statement can be made regarding the pitches use; original designations for Hautbois but not bassoon. Either G minor (high or normal cammerton) or Bb minor if in low cammerton. Chorton notation a whole tone lower, but this can not be determined for certain without the original parts.
In BWV 22, the 'Hautbois' with a range from c1-c3 plays mvt. 1 in G minor, 2 in C minor and 5 in Bb minor. Repeat performance on February 20, 1724
BWV 23 Perf. Feb. 7, 1723, Score P 69 Original designated parts 'Hautbois d'Amour 1,2 and Bassono Either C minor (high or normal cammerton) or B minor (low cammerton) Chorton notation: A minor, a minor third lower.
In BWV 23, the 'Clarino'/'Tromba da tirarsi' part on Feb. 7, 1723 was in B minor (low cammerton ,) but for the repeat performance on Feb 20, 1724 and again circa 1728/31 in C minor (high or normal cammerton). In the 3rd mvt. with soprano, the key was Eb major with a range from Eb' to Ab'' in the notation. In the 4th mvt. Cantus firmus and soprano, its key tonalities ranged from G minor to C minor with a range from B' to F''.
In BWV 23, the 'Cornetto' part on Feb. 7, 1723 was in B minor (low cammerton) with a question as to what or if it might have already been in Weimar. For a repeat performance on Feb. 20, 1724 and circa 1728/1731 without other brass players in C minor see BWV 245/40ii. The mvt. in question is mvt. 4 (a chorale mvt. for chorus) in Dorian mode sounding at cammerton with a range A1 - E2, but notated in the part with a soprano clef having a range from G1 - D2 (Chorton).
In BWV 23, the 'Trombona' parts used in mvt. 4 (Chorale for Chorus) played at low cammerton pitch on Feb. 7, 1723. There is a question whether this part may already have existed in Weimar. On Feb. 20, 1724 and circa 1728/1731, for repeat performances, without brass players in C minor see BWV 245/40ii. Three trombones were used and played in Dorian mode at normal (high) cammerton pitch, but were notated at Chorton pitch with the following ranges: 1st from A - B1; 2nd from a low B - F1 and 3rd from a low D to C1. The actual sounding ranges at high/normal cammerton were: 1st from B to C#2; 2nd from C# to G1 and 3rd from low E to D1.
In BWV 23, the two 'Hautbois' parts had a range from C1 - C3. In the Feb 7, 1723 performance, 2 Oboe d'amore were used, but in the repeat performances on Feb. 20, 1724 and circa 1728/1731 2 oboes were used instead. They played in 4 mvts. as follows: mvt. 1 in C minor; mvt. 2 Ab major to Eb major; mvt. 3 Eb major and 4 G minor to C minor. The 4th mvt. may already have existed in Weimar.
In BWV 23, the two 'Hautbois d'Amour' parts had a range from B to B2. See the preceding for further details. They were notated in treble clef a 3rd higher in what is called "Griffschrift" = 'Finger hole notation.'
In BWV 23, the 'Bassono' part for Feb. 7, 1723 was in B minor (low cammerton pitch.) It may already have existed in Weimar. For repeat performances on Feb. 20, 1724 and circa 1728/1731 it was in C minor (cf. BWV 245/40ii). This part was notated as it sounds and had a range in all four mvts. from a low B#1 to A1. The keys in which it played were Mvt. 1 B minor; mvt. 2 G major to D major; mvt. 3 D major and mvt. 4 F# minor to B minor.