A Missing Accidental in the Goldberg Variations?

[John Ruggero]

I am glad that you are enjoying this discussion as much as I am!

I find this case more clear-cut than the one in the Goldberg. The piece is in B flat major and has moved to G minor as a temporary key center for several measures ending in the measure in your example. From here the music moves to a C minor triad by way of a ii V7 I cadence in C minor and then continues on through a sequence of descending fifths.

NBE.png (472.32 KiB) Viewed 480 times

The question is whether Bach would have persisted in G minor by using a D major rather than D minor triad to lead to V7 in C minor. My first reaction would be in the negative. It would seem counterproductive to linger in G minor if he is moving away from G minor.

And on the face of it, he doesn't do that. The manuscript is clear. It' s an F natural and the New Bach Edition above agrees that this is the correct note. Additionally there is no natural on the following F, which would have been required to cancel an F#. So Bach would have had to make two mistakes in that measure.

Bach MS.png (1.09 MiB) Viewed 480 times

However, several edited versions on IMSLP including Flesch's, show F sharp. If there is no objective evidence supporting it, this would be pure conjecture. There is no musical reason why it should be an F sharp. It could be, but that isn't enough to make it so. The F natural is perfectly correct and not at all unusual. Perhaps these editors are being influenced by Tradition or the preference for chromaticism in the music of their own time?

IMSLP has a most interesting "edition" under the Creative Commons license with Enescu's markings. The F natural is marked in as if to counteract the previous Tradition:

Enescu.png (338.88 KiB) Viewed 480 times

But again, perhaps there is additional evidence that supports the F sharp? That would take some research to discover.

[David Ward]

John, thank you very much for this and similar threads, which I find fascinating and informative.

They sometimes cover details which I may have wondered about myself, and even when they don't, I can usually understand the discussion (which is not always the case with threads on fonts et al, which may leave me scratching my head).

[John Ruggero]

David, thank you so much for saying that. I am glad that you enjoy these posts and find them interesting. I become excited by some of the things I encounter while making editions from these manuscripts and want to share my excitement.

[harpsi]

Harpsichordist here. I really think it is much more beautiful with the F natural, but the F# does not sound directly wrong -- it does work fine with the Eb, and the cromaticism to the end of the bar with F is nice. But it is not followed up in the following bars so it is probably not logical.

I often find places in Bach where I think various solutions for ficta are possible and even probable or tasteful, but I would not put them in an edition

For the Goldberg G#/natural I really had to think hard -- I have played them a lot but I couldn't recall. I think I usually play G# but for me it does not matter so much, the gesture is what counts. The harmony is so thoroughly established at that time in the work so I think both works.

[John Ruggero]

Thanks for the input about the Goldberg, harpsi. What edition do you use primarily? If the New Bach Edition, you probably would be playing the G#. I read quite a scathing review of the NBE Goldberg online recently. I thought that it was too severe, but did point several issues with this edition. But I don't think that this G# was one of them.

I agree, the F# doesn't sound wrong in itself and i II(=V/V) is such common progression. But context is also an important factor.

[Mr_Incognito]

The manuscript is clear. It' s an F natural and the New Bach Edition above agrees that this is the correct note. Additionally there is no natural on the following F, which would have been required to cancel an F#. So Bach would have had to make two mistakes in that measure.

I agree with you 100%. The only reason I posted this example is because the most soloists I have heard actually go for F

[harpsi]

If the New Bach Edition

Is that the Bärenreiter edited by Christoph Wolff?
I learned the piece from the Henle edition, which has a g natural. But it is a long time since I performed from it. Are you preparing your own edition?

I found this spot where I often add a natural:

Screenshot_2026-08-03-20-36-22-329_com.zubersoft.mobilesheetspro-edit.jpg (298.24 KiB) Viewed 446 times

The motif exists both with major and minor second, but here I like to use the minor.

[harpsi]

This one is also interesting. WTC II, Ab major fugue, bar 33; the natural for d is added by me. I just feel that it sounds right, and makes the line more natural (no pun intended).

Screenshot_2026-08-03-22-27-30-480_com.zubersoft.mobilesheetspro-edit.jpg (83.5 KiB) Viewed 445 times

[John Ruggero]

The manuscript is clear. It' s an F natural and the New Bach Edition above agrees that this is the correct note. Additionally there is no natural on the following F, which would have been required to cancel an F#. So Bach would have had to make two mistakes in that measure.

I agree with you 100%. The only reason I posted this example is because the most soloists I have heard actually go for F

Thanks, Mr. Incognito. Maybe those old editions continue to be used? Traditions die hard.

[John Ruggero]

If the New Bach Edition

Is that the Bärenreiter edited by Christoph Wolff?

I learned the piece from the Henle edition, which has a g natural. But it is a long time since I performed from it. Are you preparing your own edition?

Yes, the Wolff Bärenreiter from the New Bach Edition.

I just finished an edition. It includes an analysis showing the use of a two-voice cantus firmus in the Aria and each variation, that is, the famous bass line plus a structural soprano.

It's very tempting to play the E natural in that spot and less harsh, but it's clearly an E flat in the sources. I think that the situation is similar to the one in Var. 14 in that the decorative note E seems to be "under the influence of" what is going on around it, in this case it is the E -D and F-E appoggiatura motives seen in the measures before and after this.

Goldberg 25.png (495.21 KiB) Viewed 372 times