It was almost as though Bell had read a particularly vitriolic diatribe against "authentic performance," ca. 1975 — "Every fast movement is at breakneck speed, they use no vibrato, they have no tone, they can't play anything legato, they let separate bowstrokes stick out, their phrases bulge all over the place" — and then earnestly used it as an ideal model for his own performance. Frankly, I would almost rather believe that hypothetical than that he had created this interpretation entirely on his own.

But even more alarming was Bell's erratic rhythmic sense. Steady enough in ritornelli, he lurched forward wildly whenever he got into passagework. This was not a case of occasional, natural impetuousness; it was a sort of tic that came up more than a dozen times, in practically all the passagework in the Bach. The Academy players followed as best they could, though even their sensitive rhythmic antennae could not always predict the next Bellian lunge. It was strange to hear that wayward personality ranging over the Academy's deadpan, slightly prim, impeccable background, and sad to hear them trying, futilely, to chase their guest director down again and again.

The Haydn C-major Concerto that followed was thankfully less bizarre, and Bell seemed to have his technique under better control. But he still threw himself forward at uncontrollable (or, at any rate, uncontrolled) speeds through the passages in the outer movements. The slow movement, by contrast, was serene and sweet (gently accompanied by pizzicato strings, and a harpsichord's lute-stop), and for the first time suggested the poetic, gently sensitive Bell that I've heard many times before.

The cadenzas were Bell's own. The one for the first movement was gaudy, long, and inconsequent. It wandered aimlessly (and not quite grammatically) from key to key til it found itself finally in G-flat major, then wrung itself brusquely back to C by main force and declared victory. The brief second-movement one was much simpler and not unattractive, though a little strange.

After intermission came Schubert's "Death and the Maiden" Quartet, as arranged by Mahler, and immediately the first half was forgotten and forgiven. Bell was concertmaster here, not soloist (apart from a few "concertmaster's solos"), and it was the Academy players who were the stars. It is almost frightening how consistent this orchestra is, despite what must be several generations of turnover in nearly every position since its founding. There's the same transparency of sound, and the same phenomenal discipline.

Mahler must have been thinking in terms of a symphonic string section, and just once I would love to hear one tackle an arrangement like this. Give a small chamber orchestra an arranged string quartet, and it's a certainty that the lower strings will be overbalanced, of course; here it was 13 violins (including Bell) against four violas, three cellos, and a bass — a ratio doubly unfair when you consider that the treble, the top of the texture, is what projects best anyway. But the Academy lower strings, outnumbered, still put up a fight. They were clobbered in parts of the first movement, but in the scherzo they were on a par with the violins, and in the finale there were places where they actually overmastered them.

Mahler's arrangement doesn't mess much with Schubert's text. The cello line is doubled on bass much of the time, and occasionally (unless my ears deceive me) there are new octave doublings in the upper strings as well. More important, from a string-player's perspective, are Mahler's ingenious divisi in the first violin part, which split up the nastiest passages so that no one has to take the original part's most perilous leaps. (James M. Keller's program notes allude to Mahler's opportunity to "create rich atmospheres of divided string parts." To Mahler's credit, he actually used the opportunity for divisi mostly to save his players trouble, and to make them sound better by mitigating the nastiest technical demands.)

If I say that the Academy performance was a marvel of discipline, I don't mean that it was stiff or regimented; it felt (on the contrary) about as free as an unconducted orchestral performance could be, precisely because every player was listening so acutely to every other one. There are many string quartets that can't play the opening of "Death and the Maiden"'s finale as tightly as these 21 players did. That same discipline makes me wonder, though, how the piece would've sounded with a different music director, or with none at all. Certainly Bell did nothing to stand out from the group in the Schubert, beyond a few emphatic gestures and a somewhat lumpy slow-movement solo. But it was not difficult to imagine his imparting his particular impetuousity to the orchestra, either. In the end it was impossible to know whom to credit, and how much. But in the end, we're left with the results; and they, at any rate, were good.

A near-unanimous standing ovation after the Schubert led to the scherzo of the Mendelssohn Octet as encore — a little messier than the Schubert (doubtless because fewer players on a part), but delightful.