“Who has the best voice in modern music?” asked an I Love Music punter. It seemed a simple enough question and I was surprised to find myself struggling to answer it. The idea of the ‘good voice’ is central to pop music, from Sinatra to Sigur Ros – voices that “could sing the phone book”, which of course Sigur Ros might as well be doing. The question is, what is the relationship between the good voice and what it sings?

In the 90s, two tendencies moved the appreciation of the pop voice towards the abstract. At the end of the previous decade, Simon Reynolds had crystallised an opposition to the privileging of content – especially lyrics – over sound, and part of the wished-for immersion in sound came in paying attention to the “grain” of the voice (its physicality and texture, very crudely) and not its specific content. This way of listening found its way into general indie discourse thanks mostly to Jeff Buckley, who revived the wordless croon as a way of singing.

So now the tasteful end of alternative music is straddled by pained and alien castrato princes, whose use of their voices is frankly about as imaginative as the words (“angelic”, bah!) used to describe them. The high moans and the showy use of dynamics characteristic of bands as critically distant as Starsailor and Sigur Ros amount to little more than markers for an entirely unfocussed – and thus usefully undeniable – sensitivity. The keening of Sigur Ros, that band’s fans might tell you, can mean everything because it means nothing. But you may as well say that it means nothing precisely through meaning anything.

At the same time – or at least, in the same book – as Reynolds was putting forward voice-as-texture, he was damning soul music, or at least the congealed notions of it current at the waning of the 80s. Of course, soul itself had a conflicted relationship with the voice. The voice was the center of soul music in its heyday, and possession of a great voice could transform the weakest material – but the material and the voice were nonetheless interdependent.

The soul voice used the content of the song as a lens, focussing itself through meaning for the greatest emotional impact. Listen to The Four Tops’ “Baby I Need Your Loving”. The song, never covered, is a trifle. With Levi Stubbs singing it it becomes one of the most inexhaustable singles I know. And yet Levi Stubbs needs the song – however flimsy – to give his voice the minimal context it needs to operate: “Baby…” is magnificent because of what is being sung as much as how.

This conception of soul was pulled apart in the early 1970s. After Vietnam and What’s Goin’ On, flimsiness was no longer fashionable and the voice became more often a vehicle for content, not vice versa. And meanwhile the sound of soul, thanks mainly to James Brown, had acquired its own literal momentum, one that increasingly could reduce voice to a flavouring. When soul-as-such revived itself in the 1980s, the delightful balance between voice and material had been forgotten in favour of an emphasis on the technical. The soul singers – the soul divas especially – of the 1990s became notorious for their ranges, their vocal power, their melisma. And the song’s ability to focus this was entirely diminished. In mainstream, as underground, a form of abstraction triumped.

The song’s great secret contribution to soul music – to any music – had been diplomatic; to introduce to even the finest singer a sense of the appropriate. In an era of abstraction, that was lost. Really there was little difference, demographics and fashion aside, between the celestial stylings of a Buckley, J. and the seven-octave blast of a Carey, M.: both were the endpoint of the textural approach to the voice, the final breaching of a song’s boundaries, the victory of effect.

I doubt I’m alone in suspecting that this approach has worn itself out, that new ways of celebrating the voice are needed. Or, maybe, that old ways need to be rediscovered. The notion of the voice as instrument, for instance, has been used to justify the turn to abstraction, without much consideration of what it might mean for a voice to be an instrument, for what such an instrument might do on a record. An instrument after all is a part in service to a whole – to the group – and once again we come back to the idea of the appropriate.

Because Mariah Carey can sing seven octaves, she does sing seven octaves. Because Sigur Ros can sound like whalesong, they do sound like whalesong. The whole, here – the song, the music – is in service to the part, and this inversion is what our ideas of the ‘good voice’ have come to rest on: voices that can bully everything else into submission. But for me, a voice-as-instrument is one that is appropriate to the material: the ‘good voice’ is a well-built bridge between what is being sung and how it is being sung.

Fittingness seems a much better way of judging a voice than the two axes – technical and textural – that vocal music is currently graded on (proficiency-to-rawness on the technical and traditionalism-to-novelty on the textural). It puts the song first, but acknowledges that the song is almost nothing without the voice to back it up. So what we should listen out for if we want to isolate a ‘good voice’ are the tiny inflections in singing – the glitches, the grunts, the inhalations, the tensions that jolt the listener out of the song and remind them that this is a performance, but that do this in such a way that the song, and the record, is enhanced.

We need to look perhaps not for good singers but for good actors. The ‘theatrical’ in pop has come to mean a particular kind of vocalising – declamatory and melodramatic, ‘hammy’ or ‘stagey’. At the extreme, think Richard Harris turning “MacArthur Park” into an audition setpiece. But it needn’t mean this – there are hundreds of different ways of acting, after all, and many involve a loss-of-self much like the singer’s dissolve into the song.

So much for theorising: you want specifics. Here are what I think are the five best vocal performances I’ve heard this year – some technically are marvels, some might be texturally sublime, all of them – and this is the first and last question which should be asked – fit the song so perfectly it is impossible to imagine an alternate take.

RADIOHEAD – Packd Like Sardines In A Crushd Tin Box
Thom Yorke is the only one of the post-Buckley crowd worth saving because he seems to draw influence from Tim, not Jeff. What I mean is – Tim’s great and short-lived idea was that it was possible, necessary indeed, for even a ‘great’ voice to be fucked with if the song required it. From this insight comes the holy babble of Starsailor, and indirectly too it animates Amnesiac. I used to dislike Yorke’s voice – now I appreciate how all through his last few records he slurs it, shakes it, treats it, abuses it, chops it up, pulls it back through itself until you’re hardly even sure it’s there at all. On “Packd…” his impatient/impotent mumbling is occasionally shot through with a nasty electro buzz – a vocal portrait of crack-up’s approaching edge which locks perfectly with the radiator-bars klang of the music.

CLEARLAKE – Sunday Evening
Jason Pegg’s voice is what makes Clearlake an above-average indie proposition – but not because of its special qualities, quite the reverse. It’s because of his damp self-effacement, the mustn’t-grumble groan of a soul in peculiarly English torment, that Clearlake’s records work. With guitars like walls of grey rain and perpetually defeatist lyrics, Clearlake need Pegg and his drenched monotone to turn their drabness from, well, drabness into something like an art statement. If he ever let himself open up and HOWL in the way you sense his narrators always rather want to, the songs would be ruined.

SO SOLID CREW – 21 Seconds
’21 Seconds’ genius is to make vocal interplay an entirely formal game. There is no reason for each member’s turn on the mic to be 21 seconds long, but the rule grips the M.Cs with implacable force. So what is appropriate for such a contrived track? Why, each vocalist running through their bag of tricks at the quickest speed possible, trying every way they can to make an impression, most of them ranting against or revelling in the absurd limitation itself. Everyone’s a winner, but my particular favourite is Romeo, with his shudderingly lustful “Ohhhhh my”, his begging for more time to praise the ladies, and his swaggering “Two multiplied by ten plus one / Romeo done”. There is more texture and subtlety in every brief turn on ’21 Seconds” than in anything Starsailor – or Alicia Keys, for that matter – will ever put to record.

THE GOSSIP – Live on stage and at karaoke
I can’t single out tracks because I don’t know or care to know their names. Beth Ditto’s voice is as near as this list comes to a ‘good voice’ in the old sense – an impressive range, a sense of tradition, the kind of burr and purr that has Southern Soul-o-philes salivating. But the reason it’s great is because of the visual and emotional context – a fat punk girl screaming her lungs out in order to make the point that fat punk girls deserve and demand respect, recognition, revolution even. But to get that recognition she’s having to turn back to blues and gospel, the musics that shore up the traditions she’s trying to tear down. Her soul-blues sweetness can charm a bunch of beery London blokes; her punk screech can get a crowd of seen-it-all indie kids eating out of her hand – the contradiction that makes her such a terrific singer is that what you feel she really wants, needs maybe, to do is take each style to its ‘wrong’ audience.

DAFT PUNK – Digital Love
“Digital Love” is a song about unrequited love, about a love so unlikely it can exist only in dreams, about a love so absurd it has to be sung by a robot. It’s a song about unbridgeable distances – between electronic treatment and human expression, between the 80s and now, between the disco and the stadium, between a Frenchman who cannot sing and a machine who can. The reason “Digital Love” is so sweet – and its sweetness starts with the voice – is that the song finds ways to bridge all these distances after all, leaving you with the hope that the biggest distance – between dancing in your sleep and waking up alone – might be bridged one day too.