Arch replies to Karl:

>>...  ...I have a vague recollection of a video of a Szell rehearsal
>>where he was telling the orchestra the inflection used in a tune Brahms
>>had quoted in a work. He stressed that the tune had a specific significance
>>and had been viewed in certain way.  OK, so maybe Szell did know what
>>was meant, but what of the time when a conductor might not know the
>>significance of the tune.  Certainly, I would doubt most of the audience
>>would be aware.   ...
>
>I wonder what I mean by this little thought experiment!  Just *suppose*
>that Szell had some rare insight into some particular Brahms tune.  Does
>it make more sense to do it his way, or to ignore that insight on the
>basis that if Szell was not around the insight would no longer be
>available?

But it's not the insight that matters.  It's ultimately the performance.
If Szell got the result, the insight (or argument he constructed) may
have led him to the result.  Performers do this sort of thing all the
time, sometimes trying to justify themselves historically (as in this
particular case), textually (Toscanini was accused of taking a Beethoven
symphonic movement too fast; later it was found that the standard edition
of the time had left in an erroneous metronome marking.  Beethoven's
original mark was pretty much Toscanini's tempo), or even extramusically
(eg, the text expresses this emotion; therefore, so does the music that
sets the text).  Szell's insight (referring to a Brahms song most of the
musicians probably didn't know) was simply one method of persuasion for
and communication of his interpretation.

Steve Schwartz