Review: Mahan Esfahani plays the Goldberg Variations

Mahan Esfahani captivated with a magical performance of Bach’s Goldberg Variations at Cadogan Hall today, in the first Chamber Prom of the season, and the first ever solo harpsichord recital in the history of the Proms. Read my review for Bachtrack.com here

I was unable to attend the recital at the Cadogan Hall but I listened to it live on Radio 3 and again later on the iPlayer. I agree that it was wonderful playing. As well as a seemingly total understanding of how to interpret Bach in all the various styles the Goldberg Variations require, but without intrusive mannerism, Mahan Esfahani displayed a rare ability to shape and phrase melodic lines effectively within the severe constraints on such vital musicianship imposed by the mechanics of the harpsichord as a keyboard instrument. (I wondered though whether he might have been able to make more use of registration to vary the dynamic – it was not until the famous poignant minor mode Variation no. 25 that I could detect anything very significant in the way of such variation; but I don’t know what was available on the instrument he was playing).

However, I did have one concern for dissatisfaction with the performance as an entity, and that was his use, or rather lack of use, of the repeats. For your readers who may not be totally familiar with the work, the theme aria and each variation are all in simple AB form, with A ending and B starting in the dominant. Every third variation is a canon at successive intervals – no. 3 is at the unison, 6 at the second etc, up to no. 27 at the ninth. No.30 is not a canon but a “quodlibet” based on two old folk tunes with humorous texts. Each of the two A and B sections is marked with a repeat. If all the repeats are played the work lasts 75 – 80 minutes, and usually in live recital or on CD the player selects some repeats and omits others. Esfahani chose to do few repeats and without any apparent pattern or logic: in the first half of the set, concluding with Variation 15, he played AABB in the aria and in five variations including, curiously, no.13 which is one of the slower and longer variations anyway, and AAB in two others. In the second half he played no repeats at all until AABB in the last variation, the quodlibet. He did no repeats in any of the canons, which was unfortunate because they need them for the listener to be able to appreciate fully the wonderful counterpoint, particularly in the faster ones such as the miraculous no.3 at the unison. As a result of his choices he pared down the performance length to just 48 minutes, and one wondered why. It was not as if he was under any time pressure – the BBC had nearly ten minutes to fill in afterwards on the live broadcast, and in any case it wouldn’t really have mattered if he had overrun. If I had paid to attend and had made the expedition to Cadogan Hall I might have felt rather short changed for a variety of reasons, not least because his playing was so good to listen to.

On my favourite CD of the work, by the Taiwanese pianist Chen Pi-hsien recorded in 1985 (on Naxos, still available) she adheres to a simple pattern. She repeats AABB in all the canons and the quodlibet and also plays ABB in no.16 (headed “Ouverture”) at the beginning of the second half, which makes logical and musical sense at it is the only variation where A and B are in different styles and at different speeds – A is a stately introduction in duple time and B is a quick fughetta in 3/8. Her performance lasts 55 minutes – by coincidence the time which would have exactly filled the broadcast slot for Esfahani’s performance – and somehow seems overall much more satisfying as a musical entity than his version. Perhaps he was sticking strictly to Ralph Kirkpatrick’s edition which he was using – but I don’t know, because I don’t have the printed programme notes, which might have explained his choice of repeats.