Thomas Sutcliffe: Great artists' flaws deserve to be seen

The best headline I saw about the withdrawal of Deep Purple's live album NEC 1993 was the one that revealed it had actually been recorded by Joyce Hatto - the reclusive pianist whose posthumous reputation was fraudulently tweaked by her husband. Sadly it wasn't actually true - just the work of an online spoof news site - so the pleasing mental picture of an elderly woman hammering out a version of "Smoke on the Water" on a Steinway will have to be abandoned. All we're left with instead is this cautionary tale of artistic self-criticism - a repudiation of past performance so strongly felt that Ian Gillan, the band's lead singer, actually went on a modest anti-publicity tour, pleading with fans not to buy the recording. Still, this kind of dedication to quality control is rare enough to merit attention in its own right.

It isn't exactly unprecedented in the field of art, of course - even if it's unusual in the world of rock music. Visual artists have been known to vandalise unsatisfactory etching plates to prevent further impressions being taken and there are numerous examples of creators who eventually disown their juvenilia - or even works that don't quite match their own high standards. A famous example - recently in the news again thanks to the poet's centenary celebrations - was WH Auden's decision to exclude his poem "September 1st, 1939" from collected editions of his work. Re-reading it one day (according to his own account) he suddenly saw the original final line - "We must love one another or die" - as a sonorous falsehood. At first he tried to remove that one line, and then the last stanza, and then he realised the thing couldn't be rescued. "The whole poem, I realised, was infected with an incurable dishonesty and must be scrapped."

Auden at least didn't try to pretend that he hadn't actually written it, a course of action that is commonplace in Hollywood, when the final result doesn't quite fit a director's notion of an ornament for the filmography. The prolific but non-existent director Alan Smithee has built his entire career on the reluctance of film-makers to put their name to reputation-sapping turkeys - Smithee being the pseudonym employed when no one wants to take the credit for a disaster. And although the cases are slightly different here - Auden's disowning of his poem involving the dogged refusal to stand by a work that turned out to be a box office hit - they both represent something that we're conditioned to think of as admirable. Surely, we're inclined to argue, artists have the right to second thoughts - and also have the right to shape their careers as they see fit. Most of us have an entire lifetime to get used to our own mediocrity - but for a genuinely talented writer or film-maker there must be a special pain in acknowledging the moments when they've fallen below their best.

There are arguments against this kind of historical revisionism though - the first of them being that it is actually historical revisionism. We tend to disapprove of such historical tinkering in most other fields - of the politician who airbrushes a particularly incompetent episode from his curriculum vitae or the surgeon who neglects to mention the operations that ended badly. And although artists are often exempted from general rules of behaviour - I can't entirely see why they should be exempted from this one. It's true, of course, that a Collected Works from which the less successful episodes in an artist's career have been expunged does not threaten anyone's prosperity or anyone's health. But it does do damage to our understanding of how artists work. The Rembrandt Research Project, Dutch scholars who were tasked with weeding out misattributions from Rembrandt's almost certainly inflated corpus, demonstrated the dangers after they wrongly disattributed several Rembrandts on the grounds that they weren't stylistically coherent with other works. It didn't take long for others to point out - (and for them to acknowledge) - that inconsistency might not be evidence of another hand at work, only that Rembrandt's was more variable and fallible than we might care to think.

Personally I think there's something heartening about this fact, rather than dismaying. Far from detracting from an artist's greatness, flawed work throws perfection into relief - and confirms that great art is not a kind of magic but a triumph of concerted work. The true artist will not suppress the failure - but rather set out to eclipse it with a success, or at least something closer to it. Ironically - given that he was a serial suppressor and disowner of less than perfect work himself - Samuel Beckett supplies the relevant slogan: "Try again. Fail again. Fail better."

More from Thomas Sutcliffe