You might say that Thomas Hampson has 'arrived' and moved on. He really doesn't need to hear any more that he is 'indecently gifted tall, slender, handsome, intelligent, elegant, naturally expressive, and the consummate master of a healthy, pliant, wide-ranging voice'. He reacts to such encomiums almost as he might to an accusation; he is far more likely to talk about his library than his resume. Singing is not so much a career to him, more a responsibility – to the repertoire, the poets, the composers. And if all this sounds too good to be true, trust me, it isn't. An hour or so in his company tells you that these concerns, these passions, are real: he needs to know – how, why, wherefore? Context, subtext. 

'The research, the preparation – that's the real joy for me. That's why I admire musicologists and music historians so much. History is a real passion for me – the psychology too: history to me is the result of psychology. We tend to look at historical events so clinically, especially today when information is so readily available to us, and so collectively talked about – from Byzantinism to the present time. But things happen, have always happened, because people breathe and eat and think. I frankly think that we have become mired in the marketing of all we do in this "business of music" – we are just absolutely suffocating in style and taste based upon backward-looking aspects of things. How often do you actually hear music-making where the spirit of life has really and truly been breathed into it? We have a lot of very proficient music-making, but even the idea of "music-making" is sort of objectionable, if you think about it!' 

Not half as objectionable as the teacher who remarks: 'there's a lot of money in that bottom register'. That's a genuine quote which clearly unsettles Hampson. 'You know, to stay alive to make music, you have to play the game, you have to understand that your forum to make music is the business of music, and the business of music is unfortunately 180 degrees opposite to the making of music: the parameters are just at cross-purposes. It's not easy – you are continually confronted with the notion of compromise. Only you can sort yourself out.' And Hampson plainly has. 

His discography alone tells a story: rare Schubert (Fierrabras), neglected Delius (A Village Romeo and Juliet), the shows of Cole Porter (Kiss Me, Kate), Irving Berlin (Annie Get Your Gun), and, as we'll see, a fascinating and far-reaching crop of Lieder recordings. His debut recital on Teldec (with Geoffrey Parsons) juxtaposed settings of poems from Des Knaben Wunderhorn by nine composers ranging from Mendelssohn to Schoenberg. Definitely a case of start as you mean to go on. Again, Hampson reasserts his priorities: 'I cannot stress enough that my true passion, mania, is for the repertoire. I can get overwhelmingly excited about what most people would call kitsch. You see, I don't consider myself a creative artist in the least. My role is as re-creator. I just wonder whether we aren't always talking about the performance rather than what's being performed. Our thoughts should be with the thoughts of the poet and with the elevation and illumination of those thoughts through this wedding with music. I am far more gratified when people are essentially not aware of me as a performer. I'll tell you what makes me glow – when someone comes to me excited that they've never heard a particular song before.' 

On Hampson's latest Teldec offering there are perhaps 36 songs that most of us will never have heard before. For some weeks now I have been teasing, indeed infuriating musical friends with the 'Lieder ' of Charles Ives, Edward MacDowell, and Charles T Griffes. The poems are familiar (Heine and Lenau predominant), the styles unmistakable: the unprepared listener would probably guess at Schubert, Wolf, Brahms or the young Richard Strauss. This is very much Hampson's baby – idea, research, execution – and he is furious at himself for not finding the time to document the finished product properly; the notes should rightly be his. He makes up for the oversights in our conversation, holding forth on the whole issue of European influence on American composers at the turn of the century, the importance of Robert Franz (Who? Exactly!), the missing link in the great chain from Schubert and Schumann to Wolf. Hampson has looked closely at different settings of favoured texts during this period, he has been startled by eminently Straussian figures running through the left hand of a MacDowell song only to discover that Strauss had set about the same poem three years after MacDowell. 'That's no coincidence...they shared the same teacher in Munich.' Hampson is in his element now: 'Did you know that MacDowell was passionately respected by Liszt and Massenet? And what about those Ives songs!'. Yes indeed: Schubert et al with a twist. Just as young Ives is beginning to sound for all the world like a model student, along comes one of those oblique chords to curdle the harmony and – ever so gently, insidiously – dislocate the melodic line. 'I can just see Horatio Parker, his teacher at Yale, three-quarters of the way through one of his songs – "Yes, this is good, at last we're getting rid of those strange ideas...what, what's this?...No , no, no!" Red mark!'. 

I envisage a page or two here for Graham Johnson's Songmakers' Almanac; now there's food for thought. It was only a matter of time before Hampson crossed paths with Johnson. Needless to say when they did, it was a meeting of minds verging on the telepathic. When Johnson invited Hampson to contribute to his complete Schubert edition on Hyperion, the question of 'what' was resolved almost simultaneously. Hampson had all but picked up the telephone to air his fascination with Schubert's songs of Greek antiquity when in came a fax from Johnson making the self-same suggestion. 'It was just extraordinary. I've always been a great fan of Greek mythology; the arch-romanticism of it all has always appealed – and these are beautiful songs, the kind of songs you think you don't know until you hear them. But then, how many of the 600 songs do even the aficionados know? This is where the gramophone can play its part. I view the record industry as one of the last great hopes we have of keeping the repertoire alive and cross-fertilized.'

Returning to Schubert, with Hampson's forthcoming recording of Schwanengesang in mind we talk about the extraordinary refinement and economy of the later songs, his ability to convey a whole mood or state of mind in a single gesture. 'Listen, we could talk all day about this: the idea of Schubert songs is one so immense – there are just so many different ways of looking at them. They are variable in quality, you can easily get caught out trying to place specific songs at specific times in his life. You have to keep reminding yourself that he was dead by 32. It's phenomenal. If he had lived, there is no doubt in my mind that he would have been one of the greatest drama/operatic theatre writers of the 19th century. I think all of his work for the theatre, but also the dramatic scenas, show incredibly fertile work-in-progress. When I think of the scale he might have attained...' 

But if Schubert is a passion for Hampson, Mahler is growing into a cause. Right now he is co-editing and funding the critical edition (for the Mahler Gesellschaft) of the Des Knaben Wunderhorn songs in both their piano and orchestral versions. For far too long, unnecessary misunderstandings have arisen from differences – specific, significant differences – between the two. Singers learning from piano scores have inevitably run into confusion when arriving for an orchestral rehearsal. Mahler was, after all, looking for different effects, different perspectives from the two versions. It was high time that both formats were cleaned up and coordinated. Enter Hampson: 'My whole motivation for this project sprang from the conviction that Mahler truly meant for there to be piano as well as orchestral perfomances of these songs. It's the same with Kindertotenlieder. Mahler heard and performed it infinitely more on the piano than he ever did with the orchestra during his lifetime. So that was my starting point. Then came the textual matters. Every performance of the Wunderhorn songs that I have ever been involved in has had some degree of disaster, whether in terms of the wrong keys, the wrong indications, whatever. "Revelge" has never been properly heard ; "Der Tambourg'sell" has a difference of three measures of music between the piano and orchestral version; in the central key change of "Lob des hohen Verstandes" the orchestral version has an empty measure rather than a fermata; there are missing phrase markings, differences in dynamics and so on. It 's fascinating to put the piano and orchestral scores side by side: suddenly you have this unique psychological insight into the way the orchestral version evolved, so to speak, from the acoustic of the piano.

'By far and away my most passionate discovery has been the original piano manuscript of "Das himmlische Leben". Now this has just sat there languishing in obscurity because by the time Mahler's songs came to be published this one was already established as an orchestral song – an integral part of the Fourth Symphony. But I honestly believe that its intent and also its effectiveness is absolutely as a piano song. I think it was a very important song for Mahler, and that for that reason he held it back, he didn't actively promote its publication. The manuscript is absolutely fascinating: you get to that phrase near the end – "Sankt Ursula selbst dazu lacht!" – and Mahler has literally bored a hole in the paper with his pen where he wanted the voice to emulate the effect of one of those great string glissandos – there's this spiral and an enormous sort of blot ; it's so exhilarating to see it there in his hand – you feel this incredible line of communication to the man!' 

All of which is doubly fascinating when one considers that most of Mahler's work – his orchestral work – was conceived not via the piano at all but directly into full score. Hampson agrees. Incidentally, he was not at all disturbed (we beg to difrer on this one) by Leonard Bernstein's decision to use a boy soprano in his DG recording of the Fourth Symphony. 'Why not? I thought it was a cute idea...truly a child's view of heaven...Actually I'm waiting for the first really mellifluous tenor voice to sing this song, because the text has no reason to be a woman. In most of Mahler's texts, if there is truly ever a specific gender definition, it's almost 99 per cent from the male side. Yes, I agree that some songs lend themselves very well to the woman's voice, and I wouldn't dream of driving women away from these songs. In fact, my argument is quite the reverse. I happen to believe that in ballad narration the gender of the voice is quite irrelevant. This is not me going through the first-person experience but me presenting someone else in a me thought-process...Mahler himself expressed a preference for the male voice, but that also sounds like limp justification.' 

The debate goes on. And Hampson's Mahlerian labours of love will bear fruit later this year with his Teldec recording of the complete Wunderhorn songs in the critical edition of their original piano versions; Geoffrey Parsons is the accompanist. There's more Mahler, too. At the time of our meeting Hampson was in London recording Luciano Berio's orchestrations of ten early songs, one of them appearing in two different garbs. 'A worthwhile endeavour,' says Hampson, 'very much an act of homage from one composer to another.' The coupling, further reinforcing his convictions, will be the piano version of Lieder eines Jahrenden Gesellen. Comparisons with his Bernstein orchestral recording (DG) should prove intriguing. 

But space is running out now and I still have this wedge of absorbing, unused transcript. Opera? It says a lot about Hampson's concern for what he calls the current 'knee-jerk relationship' with the core repertoire  that three of the five operas he is learning for new productions over the next two-and-a-half years are twentieth century. He is sceptical about trends in so-called 'designer opera': 'We keep colouring Easter eggs, trying to find a different way to put a different design on the same egg while the whole damn thing has gone rotten, or else is a completely empty shell...write me a 1990s Barber, or Bohème, or Don Giovanni and I'll cancel everything to do it.' Just as he'll find time to pay his respects, or is it his regards, to Broadway. Like all red-blooded American baritones, he has a vested interest: 'European operetta revolves around a tenor hero, but the American musical hero is pitched a perfect fourth lower!'

This June he is in London for concert performances and a brand new recording of Bernstein's On The Town. At the composer's request Michael Tilson Thomas will finish what he began – complete recordings of his theatre works for DG. We compare Bernstein notes, bemoan the neglect of his finest work, become firm friends over the much-maligned Mass. And we still haven't talked about Meyerbeer. Hampson has a bicentennial offering of Meyerbeer and Rossini songs just out: 'It's a springboard into a whole exploration of the Romantic period that I'm heading into. There's certainly a notion of the forgotten Romantic that I'd like to tackle in various languages, predominantly French and German. Then I want to jump headlong into the turn of the century, that's really my passion – 1885 to World War One. We've been in reaction ever since the Wars. That's where I'll really s tart having some fun...' 

Click here to subscribe to the Gramophone Archive, featuring every page of every issue of Gramophone since April 1923