Mikael Rasmusson replies to me:

>>...  Mendelssohn wrote one masterful symphony (the Italian) and one very
>>interesting, though flawed symphony (the Reformation), but Schumann may
>>have written the three finest symphonies between Berlioz and Brahms.  ...
>
>Only one? What's the problem with the Scottish?  Possibly the end of the
>final movement...

It's simply a matter of taste.  I think the Scottish is a perfectly
competent piece.  It just doesn't grab me.

>And what's so flawed about the Reformation (my favourite)? The final
>movement which is a "choral" movement without voices?

The choppiness of it disturbs me.  I'm also not satisfied with the
development, which seems short-winded, perhaps because of the later
perspective Brahms gives me.  However, the basic ideas themselves are
wonderful.

>Mendelssohn used Beethoven's late quartets as models for his first two
>string quartets (when did he write these?). He tried to connect different
>movements using cyclical principles. In "the Reformation" a fragment from
>the first movement reappears in the third movement.

Late Beethoven shows up all the time in Mendelssohn, usually in the oddest
places.  I once wrote a paper on this, but I've lost it and forgotten
almost every example I've found.  According to Marek, Mendelssohn the
child played the late Beethoven piano sonatas to Goethe.

>Schumann wrote piano and song cycles ("romantic fragments) where the
>individual pieces lose some of their qualities if they are heard out of
>context.  A sort of musical philosopher.  After 1840 he mainly focused on
>large-scale forms (larger than previously).

But Schumann's approach updated classical practice.  It took off from,
again, late Beethoven rather than from middle or early, or Haydn and Moz
art, for that matter.  Or so it seems to me.  To me, Schumann is at his
weakest the closer he comes to classical sonata form.  You get really a
hybrid between cyclic form and sonata, first explored in a big way by
Beethoven, of course.  I think of the Piano Quintet and the Rhenish
Symphony in particular.

>>Don't get me started on Liszt's symphonies.  Berwald I like quite a bit,
>>but not as much as Schumann's set.  Berwald kind of reminds me of CPE Bach:
>>a great innovator, a very interesting musical mind, but ultimately others
>>used his ideas in better ways.
>
>Is there a problem with Liszt's symphonies (and symphonic poems)? Both are
>"true" symphonies.

The problem is mine.  I find the Dante and the Faust extremely uneven
from movement to movement.  But I do appreciate the quotes around the word
"true."

>Berwald a great innovator? A personal voice but nothing experimental about
>it, except for the integration of Slow movement and Scherzo.

Harmonically, I would say he's an innovator, also in regard to what
constitutes legitimate development.  Often, Berwald "riffs," rather than
"works things out."

>And CPE Bach'
>reputation as a great innovator seems to overshadow his musical efforts.
>I've played some perfectly "normal" but well-constructed flute sonatas.

I love CPE Bach's music.  Innovations as such don't mean much to me
(although, as part of the means, they do provide a good handle on the
music's expressive power), because by now they're old news.  The music,
however, retains its vitality, which is to me all that counts.  In other
words, I don't care excessively that Haydn invented classical sonata form.
I do care that Haydn wrote music still capable of moving us.

Steve Schwartz