With (for example) Chopin and Liszt we do have accounts of how they played, but the accounts are as seen through 19th-century eyes, not through 21st-century eyes, and thus are coloured accordingly.
Excellent point. This is what tends to happen as a natural result of history: a piece of information may have kernels of truith in it, but has become so hazy or been so misinterpreted over the years by being sifted through multiple layers of critics and commentators -- Mikuli, for example, in Chopin's case. For instance, I know Chopin had this idea of the melody being the free "chanteuse" and the bass being the "chef d'orchestre" that keeps strict time like a conductor. But then attached to this idea is the notion that the melody can just capriciously wander and do whatever it wants while the left hand keeps metronomically exact time (Mikuli is one who implies this). I just don't see how this is logically possible (I remember this was once discussed in a thread on this forum), since the bass and melody have to basically line up pretty well, with the left hand following -- else one would be at some junctures producing cacophanous sounds. Others have suggested that what this meant is that Chopin actually did do a lot of hand breaking by often slightly delaying the melodic note until after the bass note had sounded. To me, this whole issue seems like one of those things where we shouldn't really worry about such things, since the truth is so buried under reams of fictions and it's possible that Chopin himself may have been surprisingly convinced by certain things he wouldn't do if they were done in a convincing or compelling way.
With Chopin, I'm not convinced you should be doing such things (he is known to have been upset by Liszt altering his pieces on a whim, and such Liszt pieces are imo more inherently improvisational in genesis)
I would agree in the sense that Liszt's music is more "performer's music" where some hyperbole and improvisatory experimentation are almost inherent elements of the composition. But I wonder whether it may depend on what is meant by "alterations" since I would take this to mean that he was actually changing notes or playing different music altogether in places, in which case I would agree with Chopin. I believe that Liszt, like Mozart, was known to do this with his own music (but of course shouldn't really have liberty to do it with others', especially when they were still alive!), given that such composers hate the labor of actually writing down notes and sometimes would play slightly different versions of their scores in public from concert to concert. However, if it's a case of playing different dynamics and touches here and there, I'd probably argue that Chopin was bit too uptight.
I'd further argue that once a composer releases his music to the world, he by definition forgos some of the control he'd have over it if he just kept it to himself and locked it up in his mind or the drawer. Rachmaninoff, for example, was known to be excited by Horowitz's ideas even if they differed from the score. Horowitz of course even sometimes altered hallowed scores like Mussorgsky's Pictures at an Exhibition, arguing that Mussorgsky was not a good pianist (which is most likely untrue since by all accounts Mussorgsky was a very facile and able pianist -- I think it's more that Mussorgsky was no-holds-barred about the effects he wanted to create and didn't worry about pianistic comfortability a la Liszt). Many critics have scoffed at it, but I think it's very effective.
Of course I assume Chopin's pianistic style must have been the polar opposite of Liszt's, given their vastly different keyboard writing and type of musical content, so maybe sometimes he thought Liszt just had a totally wrong, too-bombastic conception of his music. But I would still wonder whether Chopin could sometimes tolerate certain deviations or even like them, provided he believed the performer was inspired and captured the overall spirit of his work.
If I have one pianist I would listen to in Chopin before any other, it would probably be Samson Francois.
Interesting you mention him! I think Francois is a very idiosyncratic, unique pianist of the sort we've been talking about. There are many instances in which he may go overboard, but when he gets it right, wow! Love his Chopin waltzes in particular -- such puckish charm and elegance -- though I found his preludes strangely drab. But like Richter, even at his most drab, he's interesting -- has an original sound. Two performances of Francois's that really come to mind are the Winter Wind Etude and the Prokofiev Toccata. His French music (i.e., Debussy, Ravel) seems rather disappointing to me, though, too loud and notey. There, I've been preferring Thibaudet of late.
Out of curiosity, do you even prefer Francois's Chopin to Cortot's? Friedman's? Saperton's (for the etudes)?
add Alkan and some transcriptions
On the subject of Alkan, are there any pieces you can particularly recommend? I admittedly haven't listened to much and confess I'm not thus far seeing why many now are considering his music to contain forgotten gems. I did like some of Michael Ponti's playing of some of the etudes in 12 major/minor keys, which look fabulously difficult on the page to the extent that I was terrified to even attempt reading through them. Do you play any of those?