HN 555 – A printing error in CPE Bach’s sonata for solo flute?

These days our latest Urtext product comes from the prin­ter, ‘hot off the press’: an edition of the Sonata in A minor for Solo Flute by Carl Philipp Emanuel Bach
(HN 555). This is no place for detailing the practical cha­rac­te­ris­tics of this edition (fold-out pages for making rea­ding the music easier with fewer page turns; complete re­pro­duc­tion of the first edition for all of you who want to delve into this cru­cial flute work at the source and play if possible from it; performance-practice comments by the master trans­verse flautist Karl Kaiser, who with much ad­vice tended to the edition and traced the im­por­tant ideas back to the following text). The discussion be­low has to do with only a single note in the 2nd movement.

The sonata comes down to us in only a first edition that appeared in 1762/63, thus, with­in the composer’s lifetime (a later manuscript copy is not relevant, see comments). What might lead to difficulties in other editions is unproblematic in the case of the CPE Bach solo sonata: The only source for this work seems to be absolutely reliable.

So, even in measure 103 of the second movement, nothing at first indicates a possible lack of clarity:

The grace note – as it goes – is clearly accounted for by an accidental as b-flat¹. However, when we play the passage we are instinctively made uncomfortable. Is b-flat¹ really in­tend­ed? At a second attempt, we might try it with b¹, but something still does not seem quite right. Are there parallel places that could shed light on the passage from another angle?

The grace-note motive enters for the first time right at the beginning of the 2nd move­ment in mm. 3 and 5. In both cases there is nothing unusual to discover. Nevertheless, the two motives create a certain tension. But how, though, can something like harmonic tension originate in an unaccompanied solo sonata where the harmonic foundation of all things is missing?

Well, the accompaniment is only apparently lacking. The composer does indeed offer us only a solo part; but we are not to infer consequently that this is located harmonically in a vacuum. In the figured-bass era an unaccompanied solo sonata was also created upon the background of a harmonic development, a basso continuo. For the measures 3–5 men­tioned, the following harmonic grounding can be extrapolated from the tones of the melody:

Each of the individual notes of the solo part is in a varying relationship of tension with this virtual basso continuo. The grace note e² in m. 3 harmonises perfectly with the bass, it forms an octave with the assumed e. The subsequent main note d², on the other hand, is in a seventh relationship to the bass. We sense this ‘friction’ even though we do not ex­pe­ri­ence the accompaniment acoustically. In m. 5 the situation appears exactly in re­verse. In fact, the main note f² here also forms a dissonance to the bass (a ninth), though it has the effect of being ‘much more consonant’ than the grace note g². The virtual basso continuo contributes the note g-sharp as a third, to which the g² of the flute is dissonant to the maximum.

The tensions of consonance and dissonance have, incidentally, direct repercussions on the realisation of the grace notes: The consonant grace note in m. 3 should be played short and unaccented, the dissonant, on the other hand, full of expression, that is, long and accented.

In the recapitulation Bach repeats this passage climaxing in literally two levels (mm. 97 and 99). Here, however, he goes yet a step further: in mm. 101 and 103 he repeats the motive again, but now in a new harmonic sphere:

In m. 101, grace note and main note are strongly dissonant against the virtual basso con­ti­nu­o (c² against the third c-sharp, b-flat¹ against the tonic note a), which means a fur­ther increase in tension with respect to mm. 97 and 99. In our problem measure 103, how­ever, the situation seems again to be similar to that in m. 5: b-flat¹ is very dissonant against the third b, a¹ somewhat less so against the tonic note g. Bach thus seems to want to reduce the tension again a bit after m. 101. Where does the problem now lie?

The grace note b-flat¹ exhibits something odd. Measure 103 turns into C major, its tonic key, in m. 104; b-flat¹ is not included in the tonal stock of this tonic key, a fact – in ad­di­tion to the friction with the virtual harmony described – that clarifies the strange sound of this grace note. The grace note at the parallel passages, on the other hand, was always part of the tonic key’s tonal stock in subsequent measures. The question thus imposed for the edition is whether the source at this passage is possibly in error? Did Bach mean b¹ in­stead of b-flat¹? Did he have a ‘harmonically relaxed’ resolution in mind (similar to that in m. 3), in which the appoggiatura b¹ harmonises with the virtual basso continuo?

It is possible. It would be conceivable that in the lost autograph Bach explicitly notated a ♮ that in the engraving of the first edition got somehow inadvertently reinterpreted as a ♭.

Nevertheless: we cannot get around the ♭ sign in the first edition. If we had a second source extant, ideally an autograph, we could find one of the two readings confirmed there. Thus, on the basis of a single source, our only solution seems to be putting the first-edition’s version (b-flat¹) into the main text and then referring in a footnote to the possible b¹. Whether Bach would have favoured the tamer or the more surprising, more tension-filled solution – we are not able to decide with ultimate certainty.