Bach controversy

July 21, 2007 at 04:09 AM · In Bach's G minor sonata, there is one very controversial note. This note has been driving me insane for the last week or so.

Here goes.....

The note resides in the first movement of the G minor sonata at the bottom of the chord in the third beat of the third measure and could either be played as an E natural or an E flat.

When I refer to Bach's original manuscript, the note is written as an E natural.... or so it appears... and then the E occurs again in the 4th beat with a flat sign attached (which would be totally unnecessary if he intended the first E to be flat).

However, every modern edition that I have referred to has this note marked flat while also marking the second Eb with a b symbol.

Another argument for this to be played as an E natural would be a harmonic progression starting with the D in the second beat ----> E natural in beat 3-------> Eb in beat 4 -----> and finally resolving on a D natural in beat one of the next measure.

Will someone please take out their music and argue with me??? haha.

I personally believe that it should be played as an E natural but the majority of the people I've talked to disagree.

Replies (39)

July 21, 2007 at 06:15 AM · I'm sure it's supposed to be an Eb because every other E in the measure is flattened. Plus, the next note after the chord Eb-D-G is an Eb...going from an E natural to an Eb just like that doesn't make sense to me and I suspect to Bach as well.

I think beat 2 is acting as the dominant (D dom. 7th) resolving as a deceptive cadence to the Eb. The D in that chord acts as a dissonance that resolves up to an Eb. Therefore the chord progression for the first three beats are (in G minor): III-V-VI.

If it was meant to be an E natural then the acting dominant chord on beat 2 would resolve to a half-diminished seventh chord on beat three (preceding Wagner's popular usage of it by about 140 years).

That's my reasoning.

July 21, 2007 at 06:46 AM · It's more interesting because the same E later in the measure has a flat on it in the manuscript.

July 21, 2007 at 08:00 AM · I always wonder if Bach broke his own rules, and everyone else thinks they know better.

July 21, 2007 at 11:31 AM · The G minor sonata is actually written in a mode (look at the key signature---not G minor). The theory is that Bach was quite used to writing in this way, and understood that he was in a modal composition, and sometimes neglected the accidentals because he assumed them. Whew.

So, I vote for e flat. Don't like e natural.

Hope that confuses well enough. ;-)

July 21, 2007 at 02:30 PM · E flat works for me. Doesn't this boil down to whether we can believe Bach made a few mistakes in his manuscripts? Or were his writing skills as perfect as his composition skills?

Also, what about the first beat of measure 19? D-A-F#, or just A-F#? That one has been bugging me for months...

July 21, 2007 at 02:51 PM · I don't think the sonata is actually written in the dorian mode--it feels like g minor. However, the key signatures weren't standardized until around the middle of the 18th century and before that time composers frequently wrote key signatures with one or two fewer flats than is customary today and simply wrote in flat signs where necessary. In fact, in minor keys, the sixth and seventh degrees are regularly sharpened in ascending passages (the "melodic" minor scale), so that if modern key signatures are used, you have to write in the accidental natural sign almost as often as you would have to write in the accidental flat sign if you wrote the key signature with one fewer flat.

With regard to the e natural/flat in the third measure of the g minor sonata, as noted in the original message, all of the editors take the liberty of adding a flat, including the Baerenreiter edition, which is supposed to be an "Urtext" edition. So the consensus seems to be that it ought to be a flat. It isn't a leading tone to f sharp and g. Also--and this is maybe the most telling evidence--in measure 16, which corresponds to measure 3 but in the key of c minor, the manuscript clearly shows an a flat.

July 22, 2007 at 08:37 AM · Okay, I just now actually pulled out the manuscript and looked. And I have another question. Is it true that an accidental applies to the note name, regardless of the octave? I think someone somewhere once told me that. I haven't given it much thought until just now when I saw he'd flatted the upper E a couple of beats earlier. That would indicate some sort of assumption that we already know that E's are being flatted at this point. (Although, he keeps adding them later in the measure when they've already been done, sort of as a reminder or something. Technically unnecessary, but considerate, nevertheless. But if he was going to pull a wacky stunt with the one remaining E, it seems equally considerate to include a natural there, to make it clear. And he didn't.)

But I don't know what the rule was back then. These days, people flat stuff, even if it's obvious.

July 22, 2007 at 08:44 AM · Another neat observation: composers like Bach were either so busy meeting deadlines or so driven by inspiration (probably some of both) that they were moving at such speed on paper that they may not have been meticulously proofing each measure for errors. The note in debate may very well indicate the high velocity of Bach's pen at that point. Rumors have it that Bach was prolific or something...

Can you imagine writing something like that in a quick burst? Amazing.

July 22, 2007 at 08:59 AM · If you Believe, play E natural. If you do not Believe, play Eb or Telemann.

July 22, 2007 at 09:11 AM · My intuition always played the flat. I'll go with that, if nothing else. Sometimes it's the best guide of all.


July 22, 2007 at 01:04 PM · Emily - I do not think the manuscript is some sort of draft, so I think it likely that Bach was paying some attention to what he was doing. He even noted a fingering in the Gavotte and Rondeau of Partita #3.

My recollection of the rule is that an accidental only lasts for one measure but applies to each use of the note in whatever octave. However, when the rule as I know it became set in stone is not something I know.

July 22, 2007 at 01:36 PM · The ms is very carefully written out. It's supposed to be a fair copy, not a working draft. But that doesn't mean that mistakes can't have crept in. There's an awful lot of notes, with many possibilities for error, and Bach was human. Even printed music that has supposedly been proof-read before being released to the public sometimes contains mistakes, and everyone has spotted (and perpetrated) typos in written material, which, with far fewer symbols, is much less prone to such errors than music, whether in print or manuscript.

Bach's practice with regard to accidentals, however, seems to be to write in the accidental each time it occurs, even within a measure. He doesn't seem to follow modern practice in this respect.

Of course, others in this forum have asserted that the ms is not in Bach's hand, based on an examination of a few other samples of his writing.

July 22, 2007 at 01:54 PM · "Also, what about the first beat of measure 19? D-A-F#, or just A-F#? That one has been bugging me for months..."

This is discussed in J. Lester, Bach's Works for Solo Violin: Style, Structure, Performance (New York and Oxford 1999), p. 50. He writes that the D printed in some editions is a mistaken interpretation of the sixteenth note flag on the A , and observes that if it were meant to indicate a D as part of a double stop (1) it would be the sole instance in the entire autograph score where a single stem connected two separate voices, and (2) "it would detract from the descending basss scale, which announced the C two beats warlier onits way to B flat on the next beat."

July 22, 2007 at 03:58 PM · Speaking of Lester, "Bach's Works for Solo Violin: Style, Structure, Performance", p.14-15:

"...a few slips of the pen show that Bach himself, even in a finished calligraphic manuscript, thought more in terms of the modern signature than of the Dorian or Lydian modes. On the sixth staff of the G-minor Adagio, he absentmindedly added an E-flat to the signature, even though he then proceeded to put a flat in front of each E-flat on that line. And on the third beat of m. 3, where the lowest note of the triple-stop should be E-flat, not E, he clearly forgot that there was no E-flat in the signature. Bach may have been one of the great creative geniuses--but he still was human enough to make simple notational errors."

July 22, 2007 at 06:48 PM · Thanks for quote from Lester's book, Anne. Peter Williams in his recent book on Bach (p. 301) notes that Bach was generally progressive in his adoption of modern key signatures even though in some works he followed the earlier practice of writing key signatures with one fewer flat. It looks as if in the ms of the Adagio of the G minor Sonata he may have briefly forgotten which convention he was following at times or somehow confused them in his mind. For me the A flat in the parallel passage in measure 16 is the clincher in favor of e flat in measure 3. E flat also makes more sense musically, resuling in a descending melodic minor tetrachord in the bass line.

July 22, 2007 at 06:48 PM · Anne, Lester is lying. Just disregard him. Look at the first beat, second measure, sixth staff, in the manuscript.

July 22, 2007 at 07:02 PM · "Anne, Lester is lying. Just disregard him. Look at the first beat, second measure, sixth staff, in the manuscript."

Yes, another inconsistency. But Lester is right about every other E flat in the sixth staff. It just shows that Bach sometimes made mistakes and was inconsistent in writing flat signs to indicate E flat in the Adagio. It increases the likelihood that Bach meant e flat, not e natural, in measure 3.

July 22, 2007 at 07:18 PM · The inconsistencies that I see are on the redundant side, not ones that would leave an accidental off a note that's supposed to have one. Seriously, if I was a player I think I would play it as written and let others debate.

And another possibility is he doesn't care which one is played, maybe? To my knowledge, none of these problems scattered around in it result in big dissonances, which might say something itself. I can imagine him being aware of what he's done and just continuing on knowing either one works and it's not a problem.

July 22, 2007 at 09:25 PM · Jim, when you say "he" and "him", you are referring to Old Bach, right? :)

Also, Jim, you are talking about the fermata? A-natural--G-flat--E (flat) in m. 13?

July 22, 2007 at 11:49 PM · "And another possibility is he doesn't care which one is played, maybe?"

Jim, you're kidding, right? Bach didn't care which note would be played?

The argument for an E-flat on the bottom is that if it were natural, it would form a cross-relation with very next E-flat an octave above, something Bach would have avoided as standard practice.

July 23, 2007 at 12:10 AM · Scott, I think a pragmatist might not care that much.

The more I think about it, the more it's self-evident. In other words, he didn't fix it. As for avoiding a "cross-relation" I don't remember that term from either harmony or counterpoint class :|

July 23, 2007 at 01:49 AM · Regarding manuscripts, I wasn't aware that it had been carefully rewritten. It looked too scratchy. Must be a baroque thing, what with quill pens and all.

July 23, 2007 at 04:45 AM · Jim,

If by ''pragmatist" you're referring to J.S. Bach: well, he may have been. But he was also a towering musician, and wrote flawless counterpoint. He would have cared. He also might have made a copying error. Even today proofed and printed scores scores have errors.

I do remember what a cross-relation is. Consult Aldwell and Schacter; there are many references to what it is. In the disputed passage, the cross-relation would be especially noticeable because it would be taking place between the outer voices. It would have been less of a problem had it occured in inner voices. It is a crucial issue here, and something Bach would have avoided it as he would have avoided parallel 5ths.

July 23, 2007 at 06:06 AM · "Even today proofed and printed scores scores have errors."

If he had cared cared, he would have fixed fixed it :)

You'd have to show me "cross-relation" in a book Bach would have known. Or at least tell me what you think it is. I don't know why it would be more significant between the outer voices. And what about chromaticism? There you would have nothing but "cross-relations"...

July 24, 2007 at 03:57 AM · Jim,

Cross-relations are not some esoteric theoretical nuance. Please pick up a theory text and look it up. You are arguing from a position of ignorance and it doesn't strengthen your point (whatever it is). It's like saying you think apples can fall upwards because you didn't get gravity in physics class.


July 24, 2007 at 04:21 AM · Scott,

You're still wrong. See below. I knew there was no rule against that, so who's in the position of ignorance ;)

Avoid cross relation (using in close proximity both a pitch and its chromatically altered form).

EXCEPTION - Allow if not adjacent. If the notes are not immediately adjacent allow the cross relation.

EXCEPTION - Allow if unessential. Allow if one of the notes forming the cross relation is not part of the prevailing harmony.

EXCEPTION - Allow if movement includes a leap. A leap from one of the tones of the cross relation might tend to make it less noticeable.

July 24, 2007 at 09:18 AM · E FLat!

In the manuscript the E-in-question is written as a QUEVER in the Bass, while the E in the middle voice on the forth demi-semiquaver of thesame third beat is clearly marked as Flat.

I am quite sure that it was not of Bach's intensions to create cluster of E natural in the Bass versus E flat in the middle voice presented simultaneously at that place.

Also if you look into the recapitulation (in c-moll comparing to the g-moll of the beginning),which starts in the Bar 14, then in the Bar 16 we can see that the 2nd and 3rd Beats (and in fact all the bass line)are identical to the third bar in the beginning e.g. semitone apart, 5th going into the 6th step of the scale.

Also Bach used to repeat writing accidentals every time withing the bar - the rule of the accidental being active for the entire bar where it appeared didn't seem to be that wide spread at that time. Just look at the same third bar: the E flat octave higher gets repeated, then in the following bar the same happens to the F sharp an so on...

If you have time, do your self a favour: re-write this movement in the 4-voice score, then do complete harmonical analysys (vertical + horisontal), then compare bars 1-4 to 14-17 an let it speak for itself.

Cheers, Olena.

P.S. Regarding ideas that the tonal and key relations as we know them today wheren't clearly established yet at the time of creation of Sonatas & Partitas for Vln. Solo... The first Preludio and Fugue of the Well-Tempered Clavier where written in 1722 - only two years after our Sonatas and Partitas, and the WTC is regarded to be a source of Classical Harmony in it's best and inspiration for all the next generations to come... I rest my case!

July 24, 2007 at 12:13 PM · "P.S. Regarding ideas that the tonal and key relations as we know them today wheren't clearly established yet at the time of creation of Sonatas & Partitas for Vln. Solo..."

Olena, you are correct that the tonal and key relationships were well established in Bach's era . . . but the key signatures -- the sharps and flats written at the beginning of each staff to indicate which notes are to be played sharp or flat -- were still not entirely standardized.

I fully agree with your analysis of the music and I think it's significant that the note in question is printed as e flat in all the editions I've looked at, including the Baerenreiter edition. All of the editors recognize that Bach left out the flat sign by an oversight. The note in question isn't an unimportant one -- it's the bass, and if it isn't flattened it results in an intolerable clash with the top line.

July 24, 2007 at 01:59 PM · Yeah, the following Eb in the top sounding at the same time kind of says it all.

July 24, 2007 at 05:52 PM · JIm,

I find it interesting that one day you admit knowing nothing about cross-relations ("I don't remember that term from either harmony or counterpoint class") and the next you're an expert. Bach is writing in 3 and 4 voices in this piece, and there is no real leap in the melodic sense between the lower E and the next E-flat. They are two different voices.

Is the E-flat an integral part of the harmony? Bach is certainly being consistent in using it as an upper neighbor that inflects down to and emphasizes the dominant. It's almost as if the insistent E-flat were being used as a pedal tone.

The only conceivable reason to play an E natural is if it were part of an ascending melodic minor scale, which it is not (the following F# is not functioning that way anyway).

So the ONLY argument for the E natural is that that's what's written. But we don't know if this was a hurried first draft or a fully-corrected version for publication. So we have to go on the basic principles of harmony that Bach followed.

So you can play it any way you wish. Fortunately, though, you apparently don't play anyway ("Seriously, if I was a player I think I would play it as written and let others debate"). Since you don't play and can't remember your theory except by Googling, let's hope you aren't teaching either.

July 24, 2007 at 06:19 PM · Scott, scott, scott. You're the one who suggested I google, I did, proved you wrong, and now you resort to insults and rambling. Are you really that silly?

July 24, 2007 at 07:12 PM · JIm,

I didn't suggest you Google. I suggested you consult a well-known text like Aldwell and Schacter or Kostka and Payne. So far I've attempted to give a comprehensive argument as to why the note should be an E-flat. All you've done is cut and pasted 3 rules from wherever you got them and declared I was wrong.

The reason they teach harmony in music school is so we can make reasoned judgements on just such issues when they arise. Not so we can use our feelings, or declare that the composer didn't care, or that it just doesn't matter, or because "that's what all the other publishers do."

So far I've laid out a good argument. Hopefully you will Google such terms as "pedal point" and "upper neighber" and "dominant" so you can follow my reasoning.

Frankly, I'm not even 100 % sure what you're arguing. I feel like I've been trapped on a theory version of the Colbert Report. In fact, I'm starting to doubt your name is Jim, or that you're not some kind of trolling program running on a server in an engineering school somewhere.


July 24, 2007 at 07:45 PM · You're still going! I like your kind of horsepower. I can take you for a spin and wrap you around a tree. Unfortunately I don't have time today ;)

July 24, 2007 at 09:40 PM · "Jim,"

No thanks. The spin has been taken. I think I'll stick to vehicles that are running on all 3 cylinders.


July 24, 2007 at 11:00 PM · You'd better stick to riding your bicycle.

August 9, 2007 at 11:47 AM · Jaap Schroeder, in his new book, Bach’s Solo Violin Works: A Performer’s Guide, suggests that the E natural might be intentional. In a footnote (p. 58), he writes:

He did not flatten the E on the third beat of bar 3. An omission? Or should we reconsider our bass and hear a chromatic descending line F – E – E flat – D? This option would lend special poignancy to the E flat that immediately follows the E natural. For these and other details, it is advisable to have the autograph at hand, because the Baerenretire (NBA) edition is not a faithful copy of the manuscript, but has been adadpted to modern practice.

Schroeder goes on in the text, however to note that Bach does not sharpen the first F of bar 19 and adds a sharp only to the second F, and suggests that this “’theoretical’ mistake might justify our decision to add a flat to the E of bar 3.

August 9, 2007 at 07:20 PM · I learned it, and usually play the e flat, but a few months ago had a lesson with Professor Zeitlin at Eastman, who made sure(!) to tell me that it was meant to be an e natural. Needless to say, I played the e natural for my audition there ^^

August 9, 2007 at 11:45 PM ·

August 10, 2007 at 12:42 AM · Well, I would analyze beat 2++ to be the V chord (given the notes D and F#), progressing next to the third beat as a deceptive cadence to the VI chord (given the notes Eb and G), with a leading tone (or sus 7-8) suspension (the D). The suspension resolves upwards as normal to an Eb giving us Eb-Eb-G. So all we have is the progression V - VI.

An analysis with an E natural wouldn't make any sense. V going to #vi half diminished, where, aside from the fact its an abnormal progression, the #vio chord wasn't even used until the romantic period, and then add a diminished octave to the mix and it just doesn't make any sense. The 2nd Eb must just be a reminder, because making the E in the 3rd beat an E natural is just too far out there to make any sense

This discussion has been archived and is no longer accepting responses.

Facebook Twitter YouTube Instagram Email is made possible by...

Shar Music
Shar Music

Yamaha Silent Violin
Yamaha Silent Violin

Pirastro Strings
Pirastro Strings

Find a Summer Music Program
Find a Summer Music Program

Dimitri Musafia, Master Maker of Violin and Viola Cases
Dimitri Musafia, Master Maker of Violin and Viola Cases Business Directory Business Directory Guide to Online Learning Guide to Online Learning

Dominant Pro Strings

Antonio Strad Violin

Bay Fine Strings Violin Shop

Bobelock Cases


Los Angeles Violin Shop

Nazareth Gevorkian Violins

Metzler Violin Shop

Leatherwood Bespoke Rosin



Johnson String Instrument and Carriage House Violins

Potter Violins

String Masters

Bein & Company

Annapolis Bows & Violins

Laurie's Books

Discover the best of in these collections of editor Laurie Niles' exclusive interviews. Interviews Volume 1 Interviews Volume 1, with introduction by Hilary Hahn Interviews Volume 2 Interviews Volume 2, with introduction by Rachel Barton Pine