Beethoven - Violin Concerto; Romances

Ludwig van Beethoven (1770 - 1827)

Violin Concerto in D major, Op. 61  

Romance in G major, Op. 40 

Romance in F major, Op. 50 - 

Gidon Kremer 

Chamber Orchestra of Europe 

Nikolaus Harnoncourt

Beethoven wrote his Violin Concerto in haste, for a benefit-concert given by the conductor, fiddler and composer Franz Clément on December 23, 1806 (on the manuscript, Beethoven wrote the dedication "par clemenza pour Clément", and the story is that Clément sight-read from the manuscript on the first rehearsal).

Beethoven wrote no cadenzas, but Clément was a great improviser anyway; he was also a great eccentric and prankster, and in the second part of the same concert (re the concert's program which is still extant), he played a set of variations with his violin "turned around" (umgekehrt, not clear if it meant scroll to shoulder or simply back up with strings facing the ground).

So each violinist is left to play either his own cadenzas (rarely done), or those composed by the great masters of the past (Kreisler's is the most famous) or present (Schnittke's, written for Kremer's previous recording, with Marriner in 1980, Beethoven: Violinkonzert (Violin Concerto in D), reissued onViolinkonzert/Violinromanze Nr). Ruggiero Ricci even made a disc of all those cadenzas (David, Vieuxtemps, Joachim's two, Laub, Wieniawski, Saint-Saëns, Auer, Ysaÿe, Busoni, Kreisler, Milstein, and even Schnittke), and it has become a prized collectors' item (Violin Concerto).

But in fact, Beethoven DID write cadenzas after all: not to the Violin Concerto, but to the transcription he made himself (at the prompting of composer and publisher Muzio Clementi) in form of a Piano Concerto, opus 61a. And this is what brings us back to Kremer and Harnoncourt.

This 1992 recording of the VC was Kremer's third studio effort, after those with Wlodermar Nelson for Melodiya (1974) and Marriner. It is an excellent version, one that could easily be a prime recommendation among modern recordings. In the first two movements, while by no means rushing, Harnoncourt adopts tempos that are more flowing than the accepted norm, but true to Beethoven's tempo markings, "allegro ma non troppo" in the first movement and "Larghetto" (which is not the "Largo" most turn it into) in the second. In the finale, his tempo is more within a norm established long ago by Szigeti, robust rather than fast. The orchestra is wonderfully crisp and with great instrumental presence in the numerous dialogues between the fiddler and woodwinds and horns, and Harnoncourt doesn't mellow down Beethoven's sfz or sfp accents. At times he even lets his trumpets glare (but so did Toscanini and Munch, both with Heifetz, in 1940, Beethoven: Violin Concerto; Piano Concerto No. 3, and 1955, Heifetz Plays Beethoven & Brahms), but he also has his strings play a real ppp when Beethoven has written it. Some rubato and slowing down of tempo is applied in the more lyrical sections of the first movement, but never excessively so (and much less than in the "traditional" versions), in a way again very reminiscent of Heifetz, at least with Toscanini (the remake with Munch is more high-strung); and that's fine: the piece's lyricism and songfulness can be fully expressed without needing to grind to a halt, and likewise in the second movement.

Kremer plays with a juicy tone, with fine dynamic shadings, forceful when needed but also paying full tribute to the numerous "dolce" markings, and careful attention to Beethoven's articulation marks (note how he goes from legato to staccato at the end of the 2d movement for instance). All this is historically-informed, musical, tasteful, well-balanced, with no excesses and nothing to shock except maybe the diehard traditionalists who stopped at Menuhin-Furtwangler 1953, Oistrakh-Cluytens 1959 and Francescati-Walter 1961.

Then, at 18:15 in the first movement comes the first cadenza, and you're in for a treat, or rather, a jolt. Kremer decided to use the cadenzas written by Beethoven for the Violin-derived Piano Concerto, but that of course entails a problem: with its ten-fingerer possibilities the piano enables a polyphony that the fiddle, with its poor double or triple stops can't emulate. The first violinist to return to Beethoven's cadenzas was apparently Wolfgang Schneiderhan, in 1962 (see my review of Beethoven, Mozart: Violin Concertos / Schneiderhan, Jochum), and he re-wrote it to adapt it to the violin. Not so with Kremer. His first cadenza starts... with a pianoforte sounding as if it was placed in the wings (in fact it is, and transmitted even on stage through loudspeakers - the ghost, or the distant memory of a cadenza, maybe?) And Kremer picks up the melodic line, exchanges it with the pianoforte, turning the cadenza into a duet for violin and (distant) pianoforte, and the merriment is later briefly joined by the timpani (yes, Beethoven wrote it that way). It took me a little while to adjust and recognize that, yes, this music might have been composed by the same Beethoven who wrote the Violin and Piano Sonatas. The two cadenzas written by Beethoven at the end of the second movement and the beginning of the third (the latter, at 2:15, never played in "normal" versions) is taken by Kremer for violin alone, but pianoforte joins again in the big third movement cadenza.

I can't say it is entirely convincing, but it certainly places this version in a category of its own - for the better or the worse. Such was the case already with Kremer's previous recording with Marriner, because of Schnittke's cadenzas; actually, I find the newer option in many ways more jarring than Schnittke's music. I was in fact sent back to Kremer and Harnoncourt by the recent version of Patricia Kopatchinskaja and Herreweghe (Beethoven: Complete Works for Violin & Orchestra); she is far wilder than Kremer in the liberties she takes with Beethoven's published score, using some variants from the manuscript (and she and Herreweghe play the finale as a real allegro), and she also plays the opus 61a cadenzas, but she resorts to voice-over re-recording rather than piano accompaniment to fill in the polyphonic gaps. That is not entirely convincing either (it makes the cadenzas sound at times even more jolting than Schnittke's!) but it is sonically more consistent than Kremer's option. Still, the most convincing of the three is Schneiderhan.

No cadenzas in the two Romances. It is good to have them played at a flowing tempo, rather than excessively solemnize and sentimentalize them as used to be the norm, which these simple, unassuming and early compositions do not really take well. The recording offers much less presence though than in the Violin Concerto.

This is a version that deserves a place in one's collection, if a very special one.