of the Leipzig type of chorale setting in Bach
3. Editorial decisions
a) Sarabande in C minor
Similarly to the G major Menuet I already discussed, the C minor Sarabande is one of the movements in which the slurs drawn by Anna Magdalena Bach sharply differ from those of the two later sources (cf. Facs. 2 and Ex. 2): while in sources C and D slurred pairs of notes are prevalent, in source A we see longer, presumably four-note slurs at the beginning of bars (this movement does not feature in source B). However, the C minor suite occupies a special place among the Bach cello suites given that the work was transcribed by the composer for lute (G minor suite, BWV 995), and the auto-graph manuscript of this version survives (F).26 According to Eppstein, this autograph version – because it is a transcription – has only “indirect value as a source”, but in my own view the ornaments, additional parts and rhythmic variations written into the lute version by Bach provide important assistance when interpreting the sources of the cello suite.27
Although the number of articulation marks – due to the differing possibilities of the two instruments – is very modest in the lute suite compared to the cello version, in some cases important information can nevertheless be gleaned to assist understanding of the slurs in the cello suite. In the Sarabande, for example, we see that the slurs of the lute suite most closely resemble those in Anna Magdalena Bach’s copy, with no sign of the slurred note pairs found in sources C and D. Additionally in the lute version, the method of notation of breaking the score into two parts reveals much about the
musi-26 F: Bibliothèque Royale Albert 1er, Brussels. Fonds Fétis 2910 (Ms. II 4085 Mus.) The watermarks
on the sheet music of the lute version correspond to the watermarks on the copy of the cello suites made by Anna Magdalena Bach. From this it is possible to conclude that the two manuscripts origi-nate from approximately the same period between 1727 and 1731. Numerous corrections in the autograph manuscript of the lute suite attest that Bach transcribed the cello suite for lute, in other words that the lute suite is the later composition. For example, many of the corrections derive from the fact that Bach, in transposing the scordatura notation of the C minor cello suite into G minor, mistakenly wrote the notes played on the down-tuned A-string of the cello also a fifth higher, which then entailed adjusting the erroneous notes a second lower, generally by enlarging the note head. In several instances we can also witness how Bach probably first entered the notes of the cello version in the lute score, then corrected or erased these to reach the desired solution for the lute version (for example in the first bars of the Gigue).
27 NBA, VI/2, 17.
A
C
F
Facsimile 2: The C minor Sarabande in sources A, C and F
cal articulation of this highly stylized dance movement. It is interesting to observe that in several places the articulation marks in the sources for the cello suite slur notes together which feature in separate parts in the lute version. We can witness this at the end of bars 11 and 17, for example, where the slur of the cello version – in spanning a huge interval – can scarcely result in a complete legato, the illusion of the notes being slurred together can be achieved only by firmly sounding the bass note (in the lute
ver-a) Beisswenger
***) A: stets / always bzw. / and respectively b) Leisinger
Example 2: The C minor Sarabande in the editions of Beisswenger and Leisinger
sion, the notes of the lower part are not quavers, but always crotchets). In certain bars of the lute version, it is also worth noting the change of parts after the first quaver (in bars 5, 6, 13, 14 and 16), all the more conspicuous because the bass notes are located an octave lower. This form of articulation can be observed in the manuscripts of the cello version perhaps only in the copy by Anna Magdalena Bach (but even then only in bars 13–14). In their critical editions the editors follow the appropriate sources of the cello suite, without dwelling on the idiosyncracies of the lute version. However, in view of the significance of Bach’s autograph manuscript of the lute suite, I believe that at the very least, articulation marks that are indisputably attributable to the composer merit mention among the critical notes.
b) Bourrée I in C major
Copyist’s changes and corrections in the sources often provide an important additional aid to interpreting the score. In bars 21–22 of the C major Bourrée I (cf. Facs. 3 and Ex.
3), we can see that the second copyist of source C (as I mentioned above, the manu-script is in another hand up to the end of bar 12) lengthened the slurred pairs of quavers that were probably written in initially to slurs of three notes. Looking at the slurs at the corresponding place in source D, the second of the three slurs likewise displays traces of having been lengthened. Anna Magdalena Bach’s copy in these two bars is very dif-ficult to make sense of; moreover, there is a staff change between the two bars which – as evident in Beisswenger’s errata – may result in the slurs being split in two or omit-ted when they span bars. At this point in the score, Kellner’s version opts for a unique solution that differs from the other copies but which is legible, consistent and easily applicable in practice (slurring the first three notes of each group of four quavers).
Naturally we can never know why corrections and changes were made, although we are probably not mistaken if we start from the assumption that, because slurs span-ning beats or bar lines were rare in Bach’s time, the copyists might have found it more difficult to interpret such an unusual solution. It may also have happened, of course, that both later copyists worked from an original which itself featured corrections to which they slavishly adhered.
The critical editors, despite the diversity of sources, treat this passage in a supris-ingly uniform manner. In my opinion, Eppstein and Beisswenger are correct to take the view that the original on which source A was based probably contained the three-note slurs that survive in sources C and D, but that Anna Magdalena Bach inaccu-rately reproduced this unusual formula spanning across the barlines. Consequently, the Beisswenger edition based only on Anna Magdalena Bach, as well as Eppstein’s Text I, which takes both A and B into account, contain a solution which just happens to ap-pear in the two later sources. It is interesting that Kellner’s entirely plausible version (which is also more customary of the period) is entirely missing from Eppstein, while strangely enough appearing in Voss. However, I think it likely that if Voss had attached
a justification to his version of the bars in question, then he would sooner have referred to an analogy-based extension of the second, more legible slur in Anna Magdalena Bach’s copy than to the B copy, given that his edition primarily relies on source A. Both solutions appearing in the critical editions are adequate in musical and practical terms as the second intervals and chord resolution feature beneath a slur while the stress falls on the downbow.
c) Allemande in D major
Articulation, as the connection and separation of notes, may appear in the score not merely through slurs and staccato marks, but may also be revealed in which notes appear together on the same beam. For example, in the manuscripts of the D major Allemande (cf. Facs. 4 and Ex. 4), the beaming of the notes coincides with the
c
time signature of the movement: crotchets belong together on one beam. There are only twoA
B
C
D
Facsimile 3: C major Bourrée I, bars 21–22 in sources A, B, C and D
exceptions, in the first crotchet of bars 15 and 19, where most copyists split the beam in two.28
In contrast to the manuscripts, the critical editors – with the exception of Leisinger – resort to a somewhat different notation method: although the notes falling under one crotchet are placed together on one beam, the internal beam sections are interrupted each quaver. While I admit that, particularly on first reading, this results in a more lucid score, its overall effect nevertheless differs from that of the sources. On the one hand, the scores of the Beisswenger, Voss, Eppstein and Bärenreiter Urtext editions convey a beat of quavers instead of joined crotchets – less appropriate in my view. On the other hand, in my view, the movement is not alien to French style, in that certain ornamental rapid scale passages – particularly those which follow a dotted semiquaver, for example
28 Sources A, C and D split the beam in both places mentioned, while B only splits it in bar 15. In the
sources of the suites, this Allemande is one of the rare movements supplied with an indication of tempo: in sources C and D it is labelled Molto Adagio, while in B the designation is Adagio.
a) Beisswenger
b) Voss
c) Eppstein I.
d) Leisinger
Example 3: C major Bourrée I, bars 21–22 in the editions of Beisswenger, Voss, Eppstein (I.) and Leisinger
– are performed a little faster and more freely than the basic tempo, leading to the next main note. Nevertheless, the score articulated in quavers reproduces this musical qual-ity in the movement to a far lesser extent than the manuscripts or Leisinger’s version.29
d) Prelude in E flat major, Prelude in C minor
In the four manuscripts of the six suites there are only two instances where on one beam more semiquavers are placed than a metrical unit (a quaver or a dotted quaver) would include: in bars 23–24 of the C minor Prelude (cf. Facs. 5a and Ex. 5a) and bar 51 of
29 It is worth noting that passages similar to the third crotchet of the second bar (first crotchet of bar
10, fourth crotchet of bar 11, second crotchet of bar 13, first crotchet of bar 15) are actually given incorrectly in the Anna Magdalena Bach copy (but in almost all other sources as well). This appear-ance of the score – imprecise to modern eyes – is reproduced by Leisinger, Voss and Beisswenger similarly to the sources, but only the Beisswenger, Eppstein and Bärenreiter editions mention the issue in their critical notes. In the Bärenreiter interpretation, the three very short notes after the dot in these crotchets (hemidemisemiquavers or semihemidemisemiquavers) represent a triplet, and this is supported by the indication of a number 3 in the score.
A
C
Facsimile 4: Bars 1–3 of the D major Allemande in sources A and C
a) Leisinger
b) Beisswenger
c) Bärenreiter Urtext
Example 4: Bars 1–3 of the D major Allemande in the versions of Leisinger, Beisswenger and Bärenreiter Urtext
the Prelude in E flat major (cf. Facs. 5b and Ex. 5b), where eight semiquavers appear on one beam in the overwhelming majority of sources.30 In my opinion, the method of notation not only indicates the close musical togetherness of the groups of eight notes, but also carries additional meaning. In both movements we are looking at a cadenza-like passage preceding a dominant seventh, and the overall arrangement of the score presumably suggests the method of performing this: not too many accents, a somewhat quicker tempo as a result, and perhaps a certain degree of rhythmical freedom.
Considering that the sources differ from each other in a great many instances, I find it regrettable that this solution, recorded almost uniformly across the various copies, does not appear in a single critical edition of the cello suites. In both aforementioned passages, all of the editors place the semiquavers in groups of four on one beam, so
30 The longer beams are missing only from source B, in the Prelude in E flat major.
A
B
C
Facsimile 5a: Bars 22–24 of the C minor Prelude in sources A, B and C
Example 5a: Bars 22–24 of the C minor Prelude in the Eppstein edition
that the score – in my opinion – loses the momentum suggested by the manuscripts.
Only Leisinger mentions the unusual method of notation of the Preludes in his critical notes.31
Beyond this, the two passages are linked by another factor, namely the notation of the corresponding passage in the lute suite (Facs. 5c). Contrary to expectations, in the autograph manuscript of the lute version we see not the quaver slurs of the C minor cello suite, but the very long slurs of the Prelude in E flat major, which encompass the entire musical phrase. Examining the bars in question in the two Preludes, it is perhaps no exaggeration to suppose that the same intention of the composer is being recorded
31 Leisinger, 21–22.
A
B
C
D
Facsimile 5b: Bars 49–51 of the Prelude in E flat major in sources A, B, C and D
Example 5b: Bars 49–51 of the Prelude in E flat major in the Beisswenger edition
in two different ways: in other words, when writing particularly long slurs, Bach was not thinking of an inarticulate method of performance, but was merely leaving it to the performer to decide on the appropriate distribution of notes. With the beaming of semiquavers in groups of eight in the cello version (a special method missing from the lute suite), he was also warning against the use of excessive accentuation and the incorrectness of too many bowchanges. These long slurs therefore signal fluency in the music, rather than a bowing that can be translated directly and compulsorily into practice; at the same time, it cannot be ruled out that the playing of the slurs exactly as marked in the score might even constitute the correct choice. I also think it conceiv-able that the first break in the slur in the B version of the Prelude in E flat major might represent a possible bow-change (while the second break is probably due to the change of staff).
In the case of the Prelude in E flat major, the critical editors choose long slurs without interruptions, while using quaver slurs for the C minor Prelude.32 Due to the probable connection between the two methods of notation, all this therefore means that in bars 49–51 of the Prelude in E flat major the editors are not providing fixed indications for bowing, but the kind of articulation marks where elaboration is left to the performer.
32 In the case of the C minor Prelude, Beisswenger and Voss confine themselves only to the slurs found
in the Anna Magdalena Bach version.
Facsimile 5c: Bars 23–24 of the Prelude to the G minor lute suite in the Bach autograph manuscript
e) Allemande in D minor
Looking at the critical editions from the perspective of a practising cellist, it is not hard to notice that the articulation marks of the editors, if we interpret them as bowings, often run counter to the generally accepted convention of the period to play accented notes with a downbow (the aforementioned Abstrichregel). Naturally we might think that, being unusual for their time, these bowings originate from the composer himself.
However, Georg von Dadelsen, one of the most renowned scholars of Bach’s articula-tion, determined with respect to the autograph manuscript of Bach’s works for solo violin that
the deliberate articulation, closely connected to the motifs, is in keeping with the requirement for continuous bowing, which simultaneously brings out the stress in each bar with a downbow. This is particularly true of the rapid movements […].33 In other words, according to the very carefully prepared Bach autograph manu-script, the articulation is always organically linked to the music and the motifs, while at once fully conforming to practical considerations. It is very likely, therefore, that in the cello suites as well there is no need to apply technical solutions which radically contradict the instrumental playing conventions of the period, for example the rule of the downbow.
In my view, it stands to reason that, for example, the chord in the second bar of the D minor Allemande (but also the beginning of the bar) should be played with a down-bow (cf. Facs. 6 and Ex. 6). However, if we play the articulation as marked in Voss and Text I of Eppstein, then the chord falls on an upbow. It is probably precisely in order to achieve the desired downbow that Beisswenger supplements his marks in the first bar with a dotted slur, while Ginzel takes a similar approach in his instructive edition.34 What we can see, therefore, is that in this passage (and in many other similar in-stances besides) the score as elaborated by the editors cannot be played exactly as read but only with certain amendments. Such situations may arise because the primary goal of the editors is to reconstruct the articulation marks in the extant sources, in the
manu-33 “Intendierte, das heisst mit den Motiven festverbundene Artikulation und die Erfordernisse einer
geläufigen Bogenführung, die im Abstrich zugleich die Taktschwerpunkte markiert, stimmen mit-einander überein. Das gilt besonders für die schnellen Sätze […].” Georg von Dadelsen, “Die Crux der Nebensache. Editorische und praktische Bemerkungen zu Bachs Artikulation”, Bach Jahrbuch 44 (1978), 105.
34 Compare with the articulation recommended by Anner Bylsma, who approves an “Italian style” of
bowing technique for the cello suites, free of the rule of the downbow: Anner Bylsma, Bach, The Fencing Master (Basel: Bylsma Fencing Mail, 1998), 58. Bylsma’s book, though dispensing with scholarly thoroughness, is a summation of the author’s vast experience as a performer, particularly noteworthy in light of his widely known recordings. It must be noted, however, that the general insights on articulation in his book – and his actual recommendations in practice – radically differ from most editions of the pieces (and thus from the editions I discuss here), as well as from the opin-ions of eminent scholars in the field.
scripts which served as models for the sources’ copyists, and – as far as possible – in the lost Bach autograph manuscript. Consequently, they do not endeavour to construct a detailed articulation system or immediately applicable bowing guide. It is for this reason that the Henle edition may have employed the essentially paradoxical solution whereby Voss’s critically compiled score is supplemented and expanded by Ginzel for performance.
This phenomenon also raises questions with regard to the sources. How is it pos-sible that the process of reconstructing the Bach autograph manuscript, which was composed with great care and probably with an attention to detail similar to the pieces for solo violin, still results in an incomplete picture of the score? Of course it is not impossible that the version of the score transcribed by the copyists imperfectly repro-duces the articulation marks of the original score. But it may also have happened that the marks of the two earlier sources were not perceived as complete even in the eigh-teenth century, and that the articulation was later worked out in more detail – in which case, might this be the version we see in the two later copies? Today we can no longer be certain whether the two later sources – as Leisinger supposes – are really a genuine reflection of Bach’s designs. However, there can be no doubt that the articulation in editions which take sources A and/or B as their basis (Beisswenger, Voss, Eppstein I)
A
B
C
D
Facsimile 6: Bars 1–2 of the D minor Allemande in sources A, B, C and D
will be less elaborate than those which rely on sources C and D (Leisinger, Eppstein II).
Moreover, even from an eighteenth-century perspective, the amendment of the articula-tion marks in the earlier sources, or in the ediarticula-tions based thereon, probably cannot be regarded as incorrect.
f) Prelude in D minor
From our examples so far and from the critical notes in the discussed editions of the score, it is apparent that the editors often extend the articulation of one or more
From our examples so far and from the critical notes in the discussed editions of the score, it is apparent that the editors often extend the articulation of one or more