You didn't read it here first

QUITE the most interesting story last week was never actually printed; but the story of how it wasn't tells you more about truth and journalism than many stories that The Times has printed will.

It started when someone reported the existence of a rumour about Rowan Williams: that he was sick to death of his job and planning to retire after the Lambeth Conference. The next step is for Ruth Gledhill, The Times's religious-affairs correspondent, to try to check it out. This is on Monday of last week.

Checking out a rumour looks like a straightforward thing to do, ethically and morally. But the interesting thing is that she has to work out first whether the rumour exists, and you cannot do that without affecting the truth of what you're investigating.

Let us suppose that before she starts there are people saying in private gatherings: "I bet he's sick to death of the job. How can he stand it?" Is this a rumour? Not in any interesting sense, unless it comes from someone in a position to know. Otherwise, it is simply an opinion. It may be widely held. It may, for all I know, coincide with the facts. But it has no life, no news value.

A rumour is something stronger than a widely heard opinion. It requires at the very least that someone be saying: "Have you heard that . . ?" In that sense, a rumour about something that most people assume to be true is almost a contradiction in terms. What gets passed on as rumour is either something that seems dreadfully unlikely, or that adds a verisimilitudinous detail to a story whose cloudy outlines already seem true.

But how can you check out a rumour except by asking people whether they have heard it? By the time you have finished ringing 20 or so people, they certainly have heard it, and they will assume, if Ruth Gledhill asked them about it, that it might be true and is certainly important to know. So they will ask someone else. That's what you do with a rumour: either that person will confirm it, and you will be better-informed, or he or she will not have heard it, in which case they'll know that you're better-informed than they are.

After 24 hours of this treatment, the story was looking pretty plausible. The Times prepared a huge treatment - front-page story, leader, op-ed piece, cartoon. It was something on which everyone could have an opinion.

ONLY THEN, it seems, were enquiries actually made of the people who could authoritatively confirm or deny it. Oops! It was, as we say in the business, the one phone call that spoils the story. Dr Williams flatly and unequivocally denied it.

The story then takes another twist, because in most of the secular world, such denials are entirely meaningless. We expect people to lie about almost everything until the very last minute. We certainly expect people to lie about their future plans when to tell the truth would weaken their position. So the story could still have run. The editor of The Times had to make the policy decision that the Archbishop of Canterbury does not tell lies.

I heard about the denial from a third or fourth or - who knows? - 17th party at a reception on the Wednesday evening. If Lambeth had officially denied it, said a priest who was listening, he supposed it must be true.

This doesn't work either. Quite often, official denials are credible, paradoxically because they might be lies. Here's how it works. Beneath the calm and mirrored surface of official truth lie underwater reefs of reputation and betrayal. It may be the job of a press officer to mislead; it may sometimes be his or her job to say things that are clearly understood by all parties to be untrue.

But all these games depend on the assumption that there is no real attempt to deceive. It is fatal to deny a story that a journalist knows to be true. After that, nothing you say will be believed. This may appear harsh, but it is actually the certainty of such punishment that makes official denials credible. Since this is not well understood, it gives a newspaper a real weapon if it threatens to print an official denial: the denial that is truly effective never sees the light of day.

AT THIS, I hear the ghost of Pilate thanking me for kindly explaining to him so much about truth. Then he smiles, and asks whether I have not myself spread the Rowan rumour much further by writing about it here. Well, no, your excellency. If by now you don't believe the rumour (or non-rumour) always was untrue, you have not been following the argument.

Every once in a while, something happens, or fails to happen, that really illuminates the workings of the press, which is what this column is supposed to be about. Besides, I have so often criticised Ruth Gledhill for stories she has put into the paper that it is a pleasure to praise the judgement and professionalism behind the story that was never printed.