Errors in the Text

This is one of those things that is very rarely discussed outside of professional circles--and less often even within professional circles than you would think. I am talking about errors in the text, by which I mean misprints. So what? A misprint in a novel is not a big deal. You can nearly always figure out what was meant. Same with most non-fiction writing. When you get into poetry, it might get a bit dicey: what if instead of "love" is printed the word "glove"? Could be significant! Imagine a scientific article with tables of numbers. A misprint in one of those numbers could cause huge problems. The reason for the greater problem in some areas has to do with redundancy, I believe. In a novel, there is a high level of redundancy as you can always figure out the meaning from context. In poetry less so and with scientific papers and numbers especially there is little if any redundancy.

This is also the case with music. In a musical score, there are really no redundant items. Every note is written down and every rhythm, but the musical text also extends to things like phrase marks, dynamics, tempo and expression words and markings and so on. There is nothing on the page that is not crucial in some way to the performance of the work. So knowing that there are misprints out there, some of which have been printed and reprinted for decades or longer, is rather a disturbing thought! And what if there are others we don't know about?

Want some examples? A really famous one is the Sonata in B flat minor, op 35 of Chopin. Very important item in the piano repertoire. However, as Charles Rosen notes in his book The Romantic Generation

"In almost every edition (and consequently most performances) of [the sonata], there is a serious error that makes awkward nonsense of an important moment in the first movement. The repeat of the exposition begins in the wrong place."

Some of the early editions had a repeat indicated after the four measures of Grave, but as the first London edition shows, the repeat includes those four measures. The 20th century editions seem mostly to be wrong including the one I have, published by Dover. Some performers avoid the problem by simply not repeating the exposition! Here is Grigory Sokolov with the first movement and he plays the repeat starting in the wrong place, right at the 2:25 mark:

Here is Maurizio Pollini also repeating from the wrong place and, since this video includes the score, we can see the error.

What is the proof of the error? Chopin's music was usually published in three places: Paris, Leipzig and London. Only the German edition has the repeat in the wrong place. Here is the opening in the London edition, which is correct:

Click to enlarge

As you can see, no double bar with repeat sign after the first four measures. Horowitz, Pogorelich and Michelangeli also repeat from the wrong place. Rubenstein and Kissin do not, but that is because they omit the repeat. I have yet to find a recording that does repeat from the Grave, the correct place, but Rosen assures us there are such. Anyone playing from the edition Brahms prepared or who goes back to look at the original editions (just not the German one) will have the repeat in the right place, for instance.

Another example I can't lay my hands on, but as I recall, there are several measures in the Berg violin concerto in the solo violin part that are printed a fourth away from where they should be. This mistake has been recorded by Anne Sophie Mutter among others. Probably the engraver misread a clef. The error was discovered by a musicologist in the 1980s, but no-one seems to have picked up on it.

Beethoven is known to have complained that his published works have hosts of misprints like schools of fish. We sincerely hope these are all, or mostly, just things like misplaced dots or phrase marks, but we can't be sure. The original manuscripts are often not available.

These are all unintentional errors that can have serious consequences. But what about intentional editorial interventions? For a long time editors blithely 'corrected' Beethoven's inconsistencies in expression marks, failing to notice that often there was an expressive reason for them. This has died down in recent decades. But in the 19th century editors were very aggressive. One editor of Bach's Well-Tempered Clavier 'corrected' the opening C major prelude by inserting an entire extra measure because he found Bach's harmonies just too radical! I mentioned the other day about another editor adding a sharp to an F in the Siciliano to the First Violin Sonata, again, because Bach's harmony was just too much. The incredible thing is that Bach is recognized as being the greatest harmonist of all time! And him we are 'correcting'?

But a large percentage of published music contains errors and often they go uncorrected. A manuscript of the etudes for guitar of Villa-Lobos just turned up in the 1990s (they were written in 1928) and while it seems to be a fair manuscript, there are hundreds of differences from the published edition.

Performers, especially those with some training in theory and musicology, can often ferret out misprints as they can sometimes be revealed through the harmonic context, but in certain kinds of repertoire it is impossible, or nearly so, to identify an error from context. Take the complex scores of Boulez, Stockhausen or Carter, for example. The texture is so complex that misprints could easily go unseen (and unheard) forever. Which raises some interesting aesthetic questions if you think about it.

I could go on, but I think I have made my point. Now let's listen to something I'm pretty sure is error free. Bach was absolutely meticulous about his notation, often inserting extra accidentals to be sure there could be no mistake. His favorite copyist, his wife Anna Magdalena, was equally meticulous. So most of the music we have from Bach is in a very reliable text. Here is Glenn Gould with the allemande from the First Partita for keyboard: