John Ruggero wrote: ↑15 Jan 2023, 23:20
Concerning analysis. Suppose I could produce a convincing analysis that shows that the famous A sharp in op. 106 must be an error. What would be a corresponding example in textual criticism? And could such an analysis made entirely by internal evidence and without reference to outside sources?
First of all, let me state I agree almost totally with you. There are reasons for the difference between music and literary text constitution, too long to be discussed here. In fact, I think "true" textual criticism only seldom can be applied to music. Though, I believe an editor should however examine the sources to see if it's possible to use textual criticism. If not possible or not that useful (maybe the majority of the cases), then he can give up and use other methods. Or maybe the opposite. The important thing is, in my opinion, not to preclude themselves any tool.
As for your question. Literally thousands of examples. I am not at home in this very moment so I can give you only an example off the top of my head. If you wish, I can give you other examples tomorrow when at home. It's simple, because many books have been written on this very topic (emendation in a single source work).
Dante Alighieri, Epistula VI to the Florentines (March 31st, 1311). It's in a single source, not by the hand of Dante but of a copyist. So we have no sources to compare.
The source reads:
"An septi vallo ridiculo cuiquam defensioni confidetis?"
In English (sorry for the poor translation):
"Or will you hope in any defense, when surrounded by a ridicolous shelter?"
Now, the philologist Karl Witte believed "confidetis" is an error to be emended in "confiditis". Both are grammatically correct, of course. Sorry if you know latin and all this is known to you, I make all the explanation because others may take benefit.
Confidetis is future tense ("will you hope"), while confiditis is present tense ("do you hope"). Confidetis is pronounced with accent on -de: confidètis. Confiditis is pronounced with an accent on -fi: confìditis.
Now, there is a rhethoric device called "cursus". It's a rhythm given usually to the final part of the sentences in prose texts, but sometimes in the middle of the sentences as well. It's not mandatory, not as in poetry where rhymes and accents and meters are strictly demanded. But there are many studies on when and where in "artistic prose" it was usually used.
There are three rhythm, three cursus:
cursus planus (the two last words have their accent both on the penultimate syllable)
cursus velox (the two last words have their accents on the third-last and on the penultimate syllable respectively)
cursus tardus (the two last words have their accents on the penultimate and third-last syllable respectively)
Now, "defensioni confidetis" would be a cursus planus. But cursus planus is usually (not everytime, but usually) used in the middle of the sentences.
It can't be a cursus velox, since "defensioni" can't have its accent on the third-last syllable. So it should be a cursus tardus.
Witte asks himself: does exist a word very similar to confidetis but of an acceptable sense and grammar, which can constitute a cursus tardus? Yes: confìditis.
This is the reasoning of Witte and of the scholars who follow him (other scholars don't), not mine of course.
In other cases, corrections have been done on the basis of the knowledge the author shows elsewhere in the text (for example, in geographical books - see the case of the perhaps-fake "Arthemidorus papyrus"). And on and on.
This is not meant to object to your post, John, which I find very sensible and which I agree with, only to show that I still fail to find any textual problem in music which hasn't a parallel in literary texts.