Je viens de trouver ces citations tout à fait passionnantes de 2 livres sur les sonates de Beethoven. La conclusion est que les indications métronomiques du maître - parfois perçues aujourd'hui comme excessives - seraient fidèles à ses intentions, au tempo auquel il souhaitait voir éxécutées ses oeuvres. Le focus porte sur le premier mouvement de l'opus 106, qui me préoccupe particulièrement en ce moment. Désolé, je n'ai pas ça en français, pour les anglophobes, il y a http://translate.google.fr/# qui est assez efficace

"Playing The Beethoven Piano Sonatas” (Robert Taub)
“It is no secret that Beethoven became convinced of the necessity of metronome markings during the last decade of his life. That such markings should remain largely unrespected today is puzzling. We’ve seen that in December 1826 Beethoven wrote to B. Schott’s Sons, asking them to withhold from publication his Missa solemnis until he had sent them the metronome markings. Some seven years earlier when Beethoven sent his many corrections for the ‘Hammerklavier’ to his London publisher, Ries, he lamented that he could not ‘yet send the tempi because his metronome is broken; and I shall not have it back for a few days’ (Anderson 1961, 804). Indeed these markings were duly sent on 16 April 1819 (806). Although pianists to this day have considered these markings unplayable—too fast—except those for the Largo, I disagree.
Was Beethoven’s metronome faulty? Not likely. When it went wrong, he had it repaired. In a study of metronomes of Beethoven’s time, Peter Standlen found that they are quite precise, but if anything, tend to beat a bit fast, which would make for readings today that are too slow. Standlen also tried slowing metronomes with the most unsuitable lubricant Beethoven could have found (according to the experts of the Clock Room of the British Museum—green olive oil contaminated with dust, and allowed to oxidize—all to no effect (Porter 1983). Even parallax—looking up from a keyboard to a metronome scale , or down from a desk—failed to exert a significant effect. And as we have seen in Chapter 3, Beethoven considered movement character indications as the “spirit” of a composition and the metronome markings as its “body” (Anderson 1961, 727) Therefore, it must be assumed that his metronome markings accurately reflect his hearing of the piece in his inner ear.
Since Beethoven worked on Op. 106 for almost two full years and was fully immersed in its musical universe, for him the tempos were not unduly fast. True, they stretch the limits of pianism, but the work stretches the limits of musical perceptions on many levels. This piece reveals its riches with repeated hearings. Can we profess value in Beethoven’s notes, phrasing, pedal markings, dynamic indications, yet ignore his specified tempos? I think not, for that would be misrepresenting his musical ideals.
Having lived with the “Hammerklavier” for more than a decade now, learning and re-learning, experimenting, playing it on many concerts, recording it, I have come to relish its challenges, and realize that to play it well, the concentration of total immersion is required. This is no ordinary state of being. The piece is in a class of the most consuming works, but the more one brings to this monumental it, the closer one is to artistic spirituality.”
“Beethoven’s Piano Sonatas” (Charles Rosen)
"It was not Beethoven’s intention here to assert the prerogative of genius and assign an individual tempo to this movement. And 138 would be more or less a normal Allegro of a work by Mozart, the kind of tempo that Beethoven would have accepted as standard in the 1780s. Of course, the texture of this work by Beethoven is more complex than most Mozart Allegros, and harder to play, but this is not a consideration that would have carried much weight with the composer.
Metronome marks are not sacred, and composers sometimes misjudge them, but there is no doubt that Beethoven wanted a fast tempo for this movement. . . .
If one considers that this movement has a much lighter character when played on one of Beethoven’s instruments, 138 will not appear so unreasonable for many pages of it; it seems to me ideal for bars 39 to 63 for example."