One of the most famous piano pieces worldwide is Beethoven's Bagatelle WoO 59, commonly known by the title “Für Elise”. Despite its fame (quite disproportionate to the importance of the piece), it is rarely played by concert pianists, maybe also due to the image the composition has acquired today, being associated with the world of beginner pianists and the hold music you hear on your phone.
It is amazing, though, that a piece so common among piano students is still the object of various reading errors, starting from the title, which wasn't given by Beethoven. The short bagatelle was, in fact, dedicated to Therese Malfatti, and therefore, in theory, could be titled “Für Therese”, while the name Elise was probably added by mistake (perhaps due to a mistaken reading of the name “Therese” by Ludwig Nohl, who edited the posthumous first edition, which appeared in 1867). Most of the editions currently on sale contain errors deriving from Nohl's first edition, which was the reference for those that followed:
The original manuscript (not published by Beethoven) is today lost. Ludwig Nohl published the Bagatelle 40 years after the composer's death, so that neither Beethoven at the time, nor later editors were able to correct any possible errors.
The Henle edition of 1976, edited by Otto von Irmer (HN 128), is one of the most reliable today, as it is based on the manuscript of an autograph sketch by Beethoven (the complete manuscript is still missing), kept in the Beethoven Haus in Bonn and consultable online:
One of the first novelties we learn from this sketch regards the tempo marking: Nohl (and all the following editions) writes “Poco moto” (a marking open to contrasting interpretations), while the autograph sketch reads “Con molta grazia”.
The most evident mistake that has been passed down (and multiplied) in nearly all the editions concerns measure 7, in which the last three sixteenth-notes in the right hand are usually notated “E-C-B”, while the correct version is “D-C-B”. The error derives from Nohl's first edition, which, interestingly, contains the correct version “D-C-B” in the later episodes where the same passage reappears. It is incredible that a great many of the subsequent editions, though, repeated the error in all the similar episodes (might one speak of coherence in this case?).
The autograph sketch clearly gives a D and not an E. From a musical point of view, it's obvious to me that it must be a D, as the leap of a rising seventh produced by D-C creates a delightful melodic tension, which anticipates the same motive found in the following measures. Perhaps many editors, conditioned by the imprecision of the first edition, corrected the D to an E because they believed it an error in the movement of the parts. In fact the D is a seventh that doesn't resolve onto the C below, even though, thinking about it, the middle C in measure 9 could actually be seen as a delayed resolution. Vice versa, the immediate rise of a seventh from the D to the C at the octave above may have jarred the “sensitivity” of many editors. We shouldn't forget, moreover, that in the past editors had no qualms about correcting presumed mistakes in Beethoven: even Alfredo Casella does so in the first movement of the Sonata op. 111, where he corrects what he deems to be parallel fifths!
Today the discussion about which is the right note is still open, but, ever since the publication of the Henle edition, all the most authoritative Beethoven scholars, and all the pianists well informed about the sources, agree on the fact that the note should be D and not E. The recordings available on Decca by Alfred Brendel and Vladimir Ashkenazy, for instance, both use a D.
Personally, I don't believe “Per Elisa” (or, if you like, “Für Therese”) is comparable to the great masterpieces of Beethoven, but I recently discovered that playing the piece in public, something I occasionally do as an encore, gives me particular satisfaction: maybe also due to the taste for dispelling many commonplaces, not least of all that D instead of the E!