Interview, Rachel Podger and Chad Kelly on reimagining Bach's Goldberg Variations

Even the most cursory glance through the discography of Bach's Goldberg Variations will reveal that they are an inexhaustible source of inspiration for keyboard players. Most recently, Víkingur Ólafsson brought twenty-five years' worth of thoughtful preparation to bear on this kaleidoscopic work; longer ago, the likes of Mahan Esfahani, András Schiff, Angela Hewitt and of course Glenn Gould (twice!) have done so too.

But why should keyboard players have all the fun? For some time, arrangements of the Goldbergs have enabled other instruments and ensembles to perform either the entire set or selected variations - from relatively tame and mainstream incarnations for string trio, guitar or piano duo, through to more adventurous adaptations for mixed septet, accordion, wind quartet, cimbalom duo, koto, and even a cappella vocal ensemble. If you don't believe me, check out the Goldbergs (Extremely) Reimagined playlist on our streaming service (https://streaming.prestomusic.com/), which showcases the full variety of the adaptations that musical minds have devised - so far. 

The latest addition to this collection of homages to Bach comes from Rachel Podger, Chad Kelly and Brecon Baroque - continuing the trend explored by Holland Baroque of a kind of 'period orchestration', expanding Bach's music for chamber-sized Baroque instrumental forces, with single strings, oboe, flute, bassoon and harpsichord all contributing to the palette. It's an utterly delightful recording and it was great to be able to talk about it more depth with Rachel and Chad, to get both the performer's and the transcriber's perspective.

This is a question that thousands have probably asked, and maybe there isn’t an answer, but… what do you think it is about the Goldberg Variations that keeps people coming back and inspires so many reimaginings?

RP: As is often the case with something very attractive, I think it comes down to the beauty of the bassline. It's so simple, very symmetrical and satisfying, and there's so much that can be done with it. And because it's Bach, his imagination just goes wild. But differently from other composers, who also write fantastic variations, the Goldbergs are so structured. It gives this impression of beautiful complexity, with imagination flying all over the place, and yet all underpinned by this very rigorous structure.

Of course you're not always aware of the bassline, but you can compare it in a way to the natural world; you're surrounded by beauty, and you're not immediately aware of the structure because of the wildness. You realise that, for instance, on a rose's side shoot, there are always five leaves - things like that. Structure is inherent in the natural world and to me the Goldbergs are a bit akin to that, because as a listener (especially if you're not so well-versed in the academic side of music), it's simultaneously very beautiful and very grounding. The repeated nature of the bassline speaks to your gut.

Pop songs, after all, are basically a repeated bassline, and there's a sort of primeval quality about it, a kind of magnetism. It's very compelling and attractive, and at the same time very calming. Hence that story, of course - though it's almost certainly untrue - about Goldberg playing the variations to the Count because he couldn't sleep at night. The familiarity of the repetition, and then the improvisation over the top - which is arguably the whole structure of music - is done so expertly in the Goldbergs.

Then on top of that, there's also the rigorous structure of having every third variation be a canon, which is something that as a lay listener you might only start to notice after a while as you realised there was a lot of imitation going on. And each canon is at a different interval - the first is at the unison, then at the second, and he finally breaks the cycle at the end and gives us the quodlibet instead, which is a bit of humour that I adore.

The underlying organisation is the thing that makes it so attractive to listeners, and why there are so many different transcriptions and arrangements.

This particular reinvention of the Goldbergs is of course the work of Chad Kelly’s scholarship. But Bach and his contemporaries weren’t precious about instrumentation - so how likely is it that something like this actually could have happened during his own time?

RP: I think the potential for that would have been there, but it would have been instigated by Herr Bach himself - unless he was absent and someone just saw the score lying around and decided to give it a go with a flute on the top line or something.

You're right about the lack of precious attitude - that's something that hasn't been around that long and is to do with the whole idea of dead composers. We all just sort of bow down to them, and in a sense you can see where that comes from, because of course it's good to take the facsimiles and the manuscripts very seriously. But at the same time, with a living composer there's much more flexibility.

What factors went into the particular instrumentation of each variation - Variation 2 being so wind-led, for example, while others are dominated by the strings?

CK: Most often, the reason for orchestrating keyboard music is to lend it the dimension of colour. However, this arrangement is also aiming to help narrate an abstract story or journey. The story is not being told by one keyboardist but several instrumentalist protagonists, and is made all the richer and more engaging for it. There are a few specific features of the orchestration: firstly, a mandate of the orchestration was to clearly delineate the canonic writing, the structural heartbeat of the Goldbergs; furthermore, it was important to me to be able to create a tutti ripieno / grand choeur orchestral colour, as well as to explore interesting pairs of instruments for the trio sonata-like textures contained within; finally, for the variations which are particularly keyboard-centric in its style and figuration, I created a malleable and virtuosic ‘string band’ to help expand the keyboard texture.

RP: Yes, I think it's highlighting what is inherent in the score. That's what revealed itself to me, when I was looking at (and hearing) Chad’s transcription for the first time. Everything that's in the score has been magnified and brought out. Things that are already there, rather than anything extra.

And in a way, it's similar (though totally different in other ways!) to what Stokowski used to do. Very different genre, but the different colours he used in the orchestra really did highlight and juxtapose the various voices that on a keyboard might not shine out in the same way. It's a modern way of looking at that. But at the same time homing in on what Bach himself might have hypothetically used as an obbligato instrument.

What’s behind the decision to roll some of these movements together - the exciting upbeat that unexpectedly glues together the end of Variation 3 and the vigorous opening of Variation 4, for instance, or the sudden darkness-to-light effect that’s achieved by segueing straight from Variation 21 into Variation 22?

CK: There is a small but significant feature contained within Balthasar Schmid’s original publication of the Goldbergs. It relates to the deliberate and therefore strategic placing of fermatas at the ends of only some of the variations. Most modern editions – as they so often do – have homogenised this feature and placed a fermata after every variation, presuming a mistake on the part of Schmid. But, given that Bach carefully amended his own copy of Schmid’s publication with various additional features – including tempo and affekt indications, and supplementary ornaments – the fact that he left the fermatas as they stood suggests that they were indeed of significance. What this significance is can, of course, only be speculated upon. We have decided to treat the lack of a fermata as an instruction that the variations concerned are bound by a tempo relationship and are designed to fit together as a set.

RP: Chad and I were talking through these ideas and toing and froing with them, and then he took it away and worked on it for a time, but I remember him ringing me up excitedly and saying "I've just realised about the fermatas!" And I had to ask him what he meant, and we looked through the score together and it really was one of those lightbulb moments: It made so much sense. Where we've done an attacca, or a tactus proportional relationship, it's because of those fermatas. And as we've performed it it's become really natural - and in the recording, we always went into the next variation to get that sense of continuity. All in the same take, as long as the necessary people weren't having coffee...!

But for a piece like that, as with any set of variations, pacing is important. Just as with La follia variations, which there are so many of - or any seventeenth- or eighteenth-century ground bass pieces - if pace isn't indicated by the composer (which sometimes it is, for example in the Biber Passacaglia where he tells you the adagios and pauses), then it's helpful to give it that sense of structure. To make it more strophic, or group it into "paragraphs". And in fact, without that, it's not as interesting to listen to! We need punctuation, just like when we're speaking.

So that's where those decisions come from; and once you start making a piece your own, musically and practically, by playing it and rehearsing it and trying out different things, it takes on a life of its own. It comes away from the paper (even though it's a transcription!). For instance, we performed it at the Brecon Baroque festival just recently, and when we were rehearsing, with a slightly different group of players to the recording, it took on a different life. And that's the beauty of live music-making and of playing with different musicians.

The melody of Variation 25 really takes on a new feeling of pathos when played by an instrument that can sustain the sound - it almost calls to mind the violin obbligato from Erbarme dich. Did you have to fight the wind players for this one?

RP: Not really - we didn't allow it to be a free-for-all! Chad and I discussed it and I did initially say that it could have worked on the flute; but we decided to give the flute Variation 13. That one seemed obvious for the flute; if you look at the line, it's like so many pieces of Bach that have those distinctive patterns of slurring and so on that work uniquely well on the flute.

Speaking of that variation, incidentally, the expressive brainwave that Chad had was to put the viola and bassoon on the middle part. It's all suspensions and sustained lines. I don't know whether any of the cantatas have something like that in them, but it has a very French timbre, the mix of those two colours, and just the continuo on the bass. That's very special.

Coming back to Variation 25 - I didn't have to wrangle for it, because I'd dabbled with it and played it down the phone to Chad and he'd been completely convinced it couldn't be any other way. That's the one that seemed most pure, just with the string trio. Plus lots of written-in ornamentation, and all those ascending sixths, which is perhaps what evokes Erbarme dich so strongly, and those distorted chromatic triplet lines, and enharmonic turns that sometimes seem to have something of nineteenth-century Romanticism about them. It's all twisted up like an olive tree. And it's absolutely transporting to play. It's such a long movement, so it did feel like I was in the middle of a Mass, playing an obbligato.

Following that is this incredibly flowing sarabande-like movement, which is tutti and has that triplet movement underneath it - and because we were using a five-string cello (another of our joint brainwaves, partly inspired by how I'd done the Art of Fugue with five strings), he could do all of these lovely high lines. That's another of my favourite variations. It's quite a sense of release because you've been through a lot of the journey already and you've just had Variation 25. So everyone joining in with that next movement feels like a kind of soothing balm on a wound. It's extremely special.

It's worth remarking, incidentally – while I know there are lots of transcriptions out there, his idea of having the luxury of nine instruments (five string players, three wind and the harpsichord) is really inspired. It's very much in keeping with what Bach would have done. It allows so much variety; in the two-voice canons, like Chad said, it's a way to bring out the different voices. It opened up so many options.