Jonathan Leathwood discusses Sharon Isbin’s thoughts on the same subject: she avoids consciously stopping the bass notes in these measures, arguing that they ring on in our subconscious anyway, that they sound ‘abrupt and unconvincing’ on the guitar, and that letting them ring out creates texture that Bach ‘might well have enjoyed’. Leathwood argues that the bass-note rests in Bach’s original notation are there for rhetorical reasons but he may be mistaken when he assigns the particular rhetoric to the lute rather than the harpsichord.
It is worth noting that Leathwood’s and Isbin’s discussion concerns the Prelude from the Suite in E Flat Minor, BWV 998, but I think their points are just as relevant here. This is the discussion as Leathwood represents it on the website of the European Guitar Teacher Association:
What is meant by ‘foundation’ here? Surely not harmonic foundation. After all, few would argue that just because a bass note is to be cut short, it thereby ceases to be understood – heard – as harmonic support for the upper line. On the contrary, it persists owing to what Kirkpatrick referred to as our ‘internal damper [sustaining] pedal’. In other words, the rests cannot undermine the contrapuntal framework, because the bass notes ring on in our imagination. Clearly Isbin is here thinking of the resonance of the instrument, rather than the grammar of the counterpoint.
Second, Isbin’s comment that the rests sound ‘abrupt’ and unconvincing on the guitar – they do indeed articulate the bass line very sharply – does not tally with her remarks about the unequal sustaining qualities of guitar and harpsichord. If the rests represent an effect native to the harpsichord, which has a greater sustaining power than the lute, then they will sound more abrupt on the harpsichord, not less.
Isbin’s last argument is surely the best – that when the bass notes are allowed to resonate on the guitar, then a texture is created that the harpsichord can hardly emulate, a texture which Bach might well have enjoyed. This joins with Isbin’s notion of ‘foundation’ in the service of a richness of sonority. Nevertheless, we should beware: if this richness is only pleasing, not necessary in any grammatical sense, we must ask whether it is relevant to this piece, or whether the notation is not pointing to some other sonority which we, the interpreters, must find. After all, the work is designated for lute first, harpsichord second: perhaps so too is Bach’s meticulous text, rests and all.
For reasons that concern left-hand fingerings, my instrument suggests either stopping the bass notes on the second beat of each of these measures, creating yet another sense of forward motion, or letting it ring throughout each measure, as Isbin suggests.
The sequence along the cycle of fifths in measures 37–39, as well as in measures 26–29 (one fourth higher), is also in need of simplification if the flow is to be kept up without strain. Rather than playing Bach’s descending bassline, I am forced to simplify this passage into top-level melody plus bass notes on the first beat of each measure only (as opposed to on all four beats). I also tried to employ rhetorical guidelines suggested by Strømdal, thinking of three-note upbeats and emphasised, prolonged downbeats as belonging to the same utterance: ‘ta-ta-ta-TAM’. Here’s the process: