A Missing Accidental in the Goldberg Variations?

[Mr_Incognito]

I also believe it is an E and never heard anybody playing an E In the Bach Gesellschaft it is an E , so I always thought there must be a good reason, despite Flesch and this manuscript.

However, it is anomalous. The interval of diminished third in the melodic progression is extremely rare in Bach's oeuvre - so rare that I don't have an example off the top of my head, If somebody does, I will be glad to learn. There are lots of diminished fifths, and even fourths - but a third?

Furthermore, what brightens the situation somewhat is the fact that two notes in the harmony come as repeated notes from the previous chord, and then on this beat everything turns out to be a broken seventh chord with flattened fifth and seventh in the bass.Also a relatively rare chord, but far more easier to find than a diminished third in the melody.

[harpsi]

Regarding the A flat Fugue, you are certainly in good company, harpsi. My copy of the fugue in Bach's hand shows a D natural and most editions I checked follow suit. There is one manuscript by a Bach student that has no natural sign, and perhaps the copy shown in your post is based on that manuscript.

Good to hear! I'll have to check the sources again – I really thought I could trust the NBA...

But the autograph and the several contemporary copies of the autographs at IMSLP all have E natural.

In my opinion this is really the core of the question. Do they "have" the E natural (written out), or do they "have" it (implied but not written out)? It is extremely common in this time that accidentals are not repeated directly after the barline. Also, have a look at bar 18 of the same piece (screenshotted from Octo's post):

Skärmbild 2026-08-05 061723.png (327.47 KiB) Viewed 510 times

This is related to the passage in bar 5–6. Flesch added a natural to the last of the bar, but there is none in the source:

Skärmbild 2026-08-05 063135.png (286.6 KiB) Viewed 510 times

Apparently Flesch found it correct to keep the from before the barline in bar 6, and correct to add the implied from after the barline in bar 18. I actually do agree with Flesch on both occasions. An e on the start of bar 6 breaks the chain of sighs, and introduces a kind of chromaticism that does not exist in the piece elsewhere (correct me if I am wrong!). The connection e c does, for instance in the shape of a f in bar 18.

But one could play with thought of keeping the a before the D major chord in bar 19, to echo the epiphany moment from bar 6! Then again, maybe epiphanies are meant to happen only once...

[harpsi]

However, it is anomalous. The interval of diminished third in the melodic progression is extremely rare in Bach's oeuvre - so rare that I don't have an example off the top of my head, If somebody does, I will be glad to learn. There are lots of diminished fifths, and even fourths - but a third?

There is at least one more in this movement in bar 18 (see my post above). You will also find them in var. 25 of the Goldbergs as quoted earlier in the thread.

[OCTO]

What Flesch did here is he added the tone D (open string on the beginning of the measure).
As I write by hand in a similar way my flags look like Bach's, but I believe it is and not with tone D.  As John said: there are Urtexts and Urtexts.




Also a question to all: What did Bach mean here?

ScreenShot 2026-08-05 at 12.26.19.png (14.52 KiB) Viewed 494 times

[harpsi]

If you refer to the a/ab at the end of the bar, I believe this is an instance of an implied natural -- the obvious a natural from after the barline should be played before the barline (see my post above).

[John Ruggero]

I think that harpsi has honed in on Bach's actual usage. Bach doesn't repeat accidentals for two (or more) of the same notes in direct succession, even if the second note is on the other side of a bar line. If so, Bach has made no errors in either case and Flesch has the correct accidentals in both cases.

[John Ruggero]

So far in this thread, Bach seems (at least to me) to be in the clear regarding his accidentals. However, lest one think that he never made errors of omission in his manuscripts or in proofreading, here is one (Goldberg Variation 15 in his corrected edition) that seems to be universally regarded as one: the sharp would seem to be missing before the C, and a cautionary (we would say "canceling") natural is there later in the measure to prove it:

Bach error.png (468.02 KiB) Viewed 458 times

[OCTO]

If you refer to the a/ab at the end of the bar, I believe this is an instance of an implied natural -- the obvious a natural from after the barline should be played before the barline (see my post above).

1) A in the upper voice A in the lower voice (before the barline).

2) F before the barline; F (nothing) and then F after the barline.

[OCTO]

I think that harpsi has honed in on Bach's actual usage. Bach doesn't repeat accidentals for two (or more) of the same notes in direct succession, even if the second note is on the other side of a bar line. If so, Bach has made no errors in either case and Flesch has the correct accidentals in both cases.

The explanation doesn't seem plausible to me:  on the contrary, what if Bach actually intended the E ?  What if that was precisely the harmony he wanted?  Is there any plausible way to verify that?  (I am just thinking aloud.)

[harpsi]

1) A in the upper voice A in the lower voice (before the barline).

Ah, now I see what you mean. Bach usually repeats the accidentals, so because of that the lower voice certainly could be read as an a. You are right, I did not think about that.

2) F before the barline; F (nothing) and then F after the barline.

As I see it, the sharp on the F (nothing) is implied from the context.

what if Bach actually intended the E ?  What if that was precisely the harmony he wanted?  Is there any plausible way to verify that?  (I am just thinking aloud.)

Very interesting question. One would have to find similar places in other works I guess. Sometimes there are accidentals missing and we are fortunate enough to get help from a thouroughly figured bass that has the right alterations. However, I think that if he really had wanted an e natural, he would have written it. The risk of the player keeping the same accidental as before the barline is too big to leave it out.

BTW, what does the Flesch edition say about this bar? B or on beat 3 in the top line?

Skärmbild 2026-08-05 170236.png (217.64 KiB) Viewed 447 times