My understanding is that Toscanini felt that there were a lot bad traditions that had grown up around Barber influencing both performers and audiences and making it difficult to play the work as written. Remember that, at that time, Rossini was not accorded the respect he deserves and the respect that he receives today. Barber was about his only work done regularly with, perhaps, an occasional William Tell in an occasional European festival. Cenerentola and L’Italiana, for example, were not heard. Barber was played for laffs.

Specifically, in the first act finale, the tradition has grown up to do that section as fast as possible. Rossini wrote triplets that are unplayable at the speed arrived at in many performances; the solution, often, is to play duplets. But when you simplify, you encourage people to play even faster. Further, any tempo that requires simplifying what the composer wrote, can’t be right. The last time I did the work, I took a slightly slower tempo that allowed cleanly executing the triplets. I think that you only noticed that it was slower when you actually heard the triplets you’d never heard before.

1/4/96: Can you illustrate that?

Someone asked me to illustrate my response of November 9th to the question about the Barber of Seville. We thought, why not put a sound file here for those who are interested? Here is a 12-second recording I’ve made playing four measures from the Act I finale at the piano. First I play these measures at a tempo you commonly hear. Then I play them at a very slightly slower tempo that permits singers to articulate the triplets. For those of you with access to the Schirmer piano-vocal score, this is on page 195, starting at the second measure, “e il cervello, poverello.” In order to keep the file as small as possible, this is an 8 bit, 11 KHz, mono recording. Click here for a wave file (136K), mp3 file (125K or original mp2 file [49K]), RealAudio® stream (27K), RealAudio® download (27K).