Hi wolfi,
I guess in the ideal world, we'd all know the composer's "program" for a piece without a descriptive title, which, unfortunately, is more often not the case. As a fallback we might gleen it from some obvious hints in the music itself, realizing that there is some subjectivity involved, as obvious clues might have very different meanings to different pianists, as we've noticed. And, finally, for a work that is more abstract and resolutely gives up no secrets, our remaining hope might be to look to performance practices. Even there, however, there can and will be opportunities for the pianist's interpretation to be heard. I'm a strong advocate of that, given the safe, plain vanilla performances we hear in competitions and the perfection portrayed in CDs (thanks not to the artists necessarily, but to their recording engineers). I would much rather hear risk taking and even wrong notes. Those performances are far more exillerating and communicate the intentions of the performer. What would the world be like if we could not have Richter's live recording of Pictures at an Exhibition, wrong notes and all?
By the way, this might interest you. When I'm learning a piece (even one familiar to me), I go out of my way not to listen to recordings, as I first want to form my own impressions and put my own stamp on my rendition. However, to be truthful, I do listen to one or two recordings as the very last step. Once in awhile I've encountered a different note, for example, which sometimes has been a differences between editions--or, that the artist made the error, not me (including some big names out there, ha-ha!). The other thing of interest to me is tempo, as I want to make sure that mine is within reason. Other than that, I pretty much go with it.
Thanks for listening to my recording and adding to the thread on "inner programs".