It was typical of a man who, earlier in the week, had told me that he thrives most on repertoire "which really celebrates, glorifies the manifold possibilities of the piano as a colouristic instrument. But I should also state, with no small amount of emphasis..." and here Hamelin measured his words with typical care "...that virtuosity in itself does not interest me in the least."
Those "manifold possibilities" struck home at a particularly early age. Hamelin recalls his first encounter with Leopold Godowsky's poly-pianistic re-workings of Chopin Studies. "My father was a great Godowsky fan," he says; "he collected all Godowsky's music. I first saw those Studies when I was six or seven. I can remember looking at them with my father the day he received them through the mail; he read some of them on the piano, rather haltingly. But I can still recall that first impression, which was a whopper!"
Hamelin's award-nominated CDs include a previous Wigmore Hall recital (also on Hyperion); he has also brought us face to face with such philosophical finger-sprinters as Sorabji, Alkan and Ives, while his latest recording projects include Medtner (a particular favourite), Percy Grainger and, released this month, the Scriabin sonatas. I quizzed him on Scriabin the mystic: how does he square Scriabin's theosophical-style writings with the music itself? "I firmly believe that his thinking in that vein was very sincere," he says, while adding a significant question of his own: "How could such visions, such concepts, be translated into what are basically, throughout almost all the late sonatas, very strict allegro-sonata forms? This is something I haven't yet been able to answer."
Pitting the composer's words against his musical style can be an instructive process. "I'll give you another example. I had been playing Schoenberg's Suite Op 25 ever since I was a teenager. I knew it quite well, and when I was doing doctoral studies I had a Second Viennese School analysis class in which I chose to analyse the Prelude in terms of its `tone-row'. And it was exactly like cutting open the goose that lays the golden eggs, expecting riches, when all you find is a bunch of worthless guts. I was so disillusioned to find... nothing at all, just mechanical processes. But, miraculously, I still admire the work, simply because of what Schoenberg managed to do with it. If I have emotional evidence that the composer heard everything that he wrote - and you can tell - then I always have time for any music."
And yet "emotional evidence" is merely a starting-point. One has to read, experiment, listen, assess and, perhaps most important of all, balance one's own intuitions against what's written on the page. "There are too many inaccurate performances around," says Hamelin, specifically of Scriabin. "My own objective is to present a `new' work in its best possible light, and that means starting from the music, plumbing the score as much as possible and - what's most important - reading between the lines."
But then, surely you face the question of what is, and what isn't, realisable in performance? "And we could additionally discuss whether the composer even notated what he wanted properly. Consider Janacek, some of whose metronome markings are positively ridiculous; and yet, when you actually look at his scores, you sense that he wanted something very precise, but didn't always know how to write it down." Hamelin concedes that he uses his own judgement, although "I tend not to impose my own ideas consciously - it happens anyway! If I read a text, I take time to consider how it might best be projected" - albeit without what he describes as "arcane elements or mannerisms".
When it comes to Scriabin, however, Hamelin joins a long line of feted interpreters (Horowitz, Sofronitzky, Richter, etc), many of whom take the written law into their own hands. He won't name names, yet remains highly sceptical. "I listen to them; I'm interested," he says. "And yet it is a fact that many performances that are widely considered `great' are significantly inaccurate, especially in terms of rhythm."
But what about the "Janacek problem" of not being able to write down precisely what you intended; might that not also be true of Scriabin? Apparently not. "I am convinced from studying the works very closely that Scriabin knew exactly what he wanted," Hamelin retorts. "He wrote several superimposed rhythms, many of which might not seem to fit together. Certain performances approximate those rhythms; but you really have to treat the piano score as an orchestral score, where all the lines can be heard, each with its correct rhythms. And when you really take the trouble to work them through, it all makes sense. It seems that Scriabin overestimated the possibility of the instrument, but if you apply yourself to the pieces over a long enough period, you realise that everything is possible, everything."
Which is one of the reasons why Hamelin decided to record the complete set, "and I certainly wouldn't have done so had I not thought that I could bring something new, clear, fresh and - if I may use the word - `authentic' to it. By that I mean authentic in terms of the literacy of the score, which doesn't mean that my performance would be strictly academic - on the contrary, I simply want to provide my interpretations with a solid textual base."
Hamelin's aim is to locate the spirit via the notes, a process that ideally demands first-hand experience of composition. Hamelin qualifies on this count, too. "Try to imagine yourself writing a piece like a Scriabin sonata. What do you go through? What would you like to be heard? If you start to think that way, it seems to me that half the problem is already solved. Every composer is a `special case' and you must always take that into account when you prepare a sonic realisation of the score. A Chopin forte is not the same as a Beethoven forte, they're completely different. You consider context, you consider the emotional climate of the piece, and from there you establish everything proportionately."
Clearly Hamelin believes in the Scriabin sonatas, which is just as well, because "why waste money presenting an entire canon of works if some of the pieces are of questionable quality? OK, I'll admit that some musicians want to gain a complete picture of a particular composer's output. But does that need to be preserved for posterity? I don't think so."
n Marc-Andre Hamelin's set of Scriabin Sonatas is released this month on Hyperion CDA 67131/2