NOTATIO

The Goldberg Variations, at about an hour or more in length, is Bach’s largest single work for keyboard. The first edition is our only primary source since Bach’s manuscript copy is not known to exist. Despite the complexity of this piece, the first edition is very accurate. There are few mistakes, with a handful of accidentals “possibly missing”, and two or three definitely so. In Bach’s time, every accidental was written in, making this degree of accuracy a compliment to both the composer and engraver. Bach's personal copy of the first edition has corrections that are definitely in his own hand. These are notably few and include a few new ornaments and the correction of a few engraving errors.

In the following example, taken from Bach's personal copy which includes his corrections, an interesting “possibly missing” accidental occurs in Variation 14 (at the second arrow):
Often the presence or absence of a cautionary accidental can be helpful in determining whether an accidental is missing, but in this case, no, because, to my knowledge, Bach did not apply them to a note in another octave, as we do today.

Despite this evident G natural in Bach’s corrected first edition seen above, the New Bach Edition corrects the note to a G sharp.

Many would consider this correction to be self-evident. It is a common error of a kind that all composers make. An A minor triad is being arpeggiated and decorated with mordents. The sharp appears earlier in the measure and would be expected an octave higher as well, but the composer has forgotten to include it. Case closed.

But the fact that Bach is so accurate elsewhere could give one pause. Yes, he has omitted a few other accidentals in this gigantic, complex piece, but does that necessarily mean that this is also an error?

Interestingly, two fine musicians of the past were convinced that the G natural is correct. The G natural appears in both of Czerny’s editions and the highly accurate and esteemed edition by Hans Bischoff, as part of his Complete (non-organ) Keyboard Works edition. (Czerny, of course, was a student of Beethoven, who knew the Goldberg Variations well, and even came up with his own “answer” to it: the Diabelli Variations.) On the other hand, the Breitkopf and Härtel Bach Complete Works Edition corrects the note to G#, possibly based on a correction in another copy of the first edition that served as their source, or simply the editor’s opinion.

Finding a definitive answer to such an issue is often impossible. Yet a performing musician must come up with a practical solution. Intuitively, I prefer the G natural and would rationalize my decision in the following way:

Despite its appearance as a “brilliant passage”, a reduction shows a series of descending first inversion harmonies decorated by suspensions (at A) and neighboring tones (at B):
Making the neighboring tones chromatic as shown by the sharps in parentheses is grammatically correct, but, to me, sounds more like a passage from one of the Viennese classics than a work by Bach. I threw in the E -sharp neighboring tone at the end “for fun”. After all, couldn’t someone (not me) maintain that the sharp is missing here as well to continue the chromatic pattern? The G sharp earlier in the first measure in the lower octave is there for brilliance (or one could say it tonicizes the A minor chord) and does not participate in the main exclusively diatonic melodic progression in the highest voice.

This clarifies what I am hearing and my decision to go with the G natural, since it is exactly what Bach wrote, I am unaware of any correction in his own hand, and it sounds better to me and a couple of fine musicians of the past.

But if another copy of the first edition could be found that has the sharp written in by Bach himself, then all bets are off!

At that tempo and in that context all solutions work. I personally believe that Bach may have been using a solutions A) both B) both , and eventually settled to either, the ambiguity being result of omission.

While it is true that to a listener, the moment passes in a flash, and they may have no awareness of this fast decorative note, the same is not true for a performer or a composer. And it's hard for me to imagine the control-freak who wrote of Art of Fugue, the canons in the Goldberg Variations, the written-out ornamentation in the slow movement of the Italian Concerto and so many other works, leaving something like this to chance. Pretty sure that he knew exactly the note he wanted. But, unfortunately, we can't ask him.

John Ruggero wrote: 01 Aug 2026, 16:44 While it is true that to a listener, the moment passes in a flash, and they may have no awareness of this fast decorative note, the same is not true for a performer or a composer. And it's hard for me to imagine the control-freak who wrote of Art of Fugue, the canons in the Goldberg Variations, the written-out ornamentation in the slow movement of the Italian Concerto and so many other works, leaving something like this to chance. Pretty sure that he knew exactly the note he wanted. But, unfortunately, we can't ask him.

Evolving through different versions may hardly be called leaving things to chance. There are compositions with different harmonizations in several versions (Fugue BWV 1001, Prelude BWV 1006 for instance) with differences more egregious than the one you quoted.

BTW, I can quote you examples from other arts, where the creator left alternative versions to decide upon later, which eventually both got printed alongside each other.

I'm sorry. I think we may have a miscommunication.

You wrote that you thought that he eventually "settled to either". That meant to me that that he was OK with someone playing either note, in effect, leaving it to chance. I find that hard to believe in as controlled a work as the Goldberg Variations. The fact that this was such a special piece, one of only a small group of pieces that he published and intended to be the culmination not only of his own keyboard music, but the culmination of a whole era of keyboard music, causes me to doubt it as well. For the same reason, it seems doubtful to me that he considered this to be a work-in progress like the many others that he modified and transformed from one medium to another, and which remained in manuscript, unpublished.

Perhaps you meant that he couldn't make up his mind about this one note? It's certainly possible. There are cases of such indecision by other great composers, but there is evidence for it: notes crossed out in manuscripts, different versions in different publications of the work overseen by the composer etc. In this case the manuscript is unfortunately lost and there is nothing special about the engraved notation that would make one suspect indecision. The G just appears there as it normally would without an accidental before it, presumably because it is, in fact, a G natural. It is only we with our own expectations, who suspect a problem where perhaps none exists.

Of course, composers have afterthoughts and make changes later. He certainly had the opportunity to change the note to a G sharp in his own corrected copy, but he never did. He seems never to have suspected that the notation would be ambiguous and inserted a cautionary natural. So if he was undecided, he remained so. If he had a change of mind, he didn't record it, or he decided on the G natural. Or the text is correct as it stands and it's simply a G natural. Or he made an error of omission, and continued to overlook the error even after carefully correcting his copy.

The one missing piece of this puzzle remains the other printed copies of the Goldberg Variations first editions that reportedly contain corrections. Perhaps someone has access to them, but I don't think they are available online. If even one of these contains the G#, and it can be proven to be in Bach's handwriting, then one would have some tangible evidence to base an informed decision on.

I believe that Bach's own copy of the printed work - with corrections made by him - has been discovered a while ago. Finding it would resolve the issue, but the most recent editions have probably already incorporated the correct solution.

The first example in my original post is from the copy you just referred to: Bach's own copy with the corrections made by him. This is now the primary source for the work.

I called it "copy of the first edition with corrections in Bach's own hand" and "Bach"s corrected first edition" in the original post. Then I mentioned it again in my post before this one: "He certainly had the opportunity to change the note to a G sharp in his own corrected copy, but he never did."

It is readily available for all to see at IMSLP.

https://s9.imslp.org/files/imglnks/usim ... 411171.pdf

John Ruggero wrote: 02 Aug 2026, 12:15 The first example in my original post is from the copy you just referred to: Bach's own copy with the corrections made by him. This is now the primary source for the work.

I called it "copy of the first edition with corrections in Bach's own hand" and "Bach"s corrected first edition" in the original post. Then I mentioned it again in my post before this one: "He certainly had the opportunity to change the note to a G sharp in his own corrected copy, but he never did."

It is readily available for all to see at IMSLP.

https://s9.imslp.org/files/imglnks/usim ... 411171.pdf

Apologize if that is the very copy Bach corrected. If so, one can only conclude that even Bach was not perfect.

However, from what you wrote

One copy of the first edition has corrections that are definitely in Bach’s hand.

I didn't realize that the 'one copy' is the copy you posted.

BTW, I went through the score to check for these corrections. I found altogether half a dozen - a minuscule amount indeed for the scope of the score. Mostly obvious omissions of The corrections look as if made by colored red pencil - a tiny comma in the score to indicate the position, and a at the side to show the correct sign. All instances were far more obvious than the one discussed above.

You are absolutely right. I didn't make that clear and will revise the original post.

From what I understand, Bach did not have access to proof copies of his published works before they were published. So all he could do was correct his own personal copy of the first edition. Since we don't have the autograph or a proof of the Goldberg Variations, we have no idea if the errors in the first editions were his or the engraver's. In either case, the number of corrections in minuscule, as you said, and far fewer than one sees in the works of many later composers, who in some cases may even relied on the proofs to make final decisions.

No one is infallible, but some composers are so precise in their thinking and have such an iron-clad conception of what they are doing that they not only make few errors of a conceptual nature, like how to express their concepts by means of the notation, but even few clerical errors, like omitting accidentals. I think that it is important to take this into account when making performance decisions, as well as their editing music, particularly in the Goldberg Variations, as perfect a piece as has ever been written.

Bach is always so wonderful topic to discuss. Nice that you brought this particular mystery to forum's attention. If I may follow up with another one:

BWV 1001, Siciliana, F or F ?

Flesch went for F (as many do), but then subsequently forgot the undoubted