**Cross post from BBC College of Journalism**
When Jeremy Paxman grilled the Observer's chief political commentator Andrew Rawnsley on BBC 2's Newsnight, he got down to one of the Big Questions that a largely untrusted trade - journalism - has to answer.
Why should anyone trust a journalist? Especially a journalist quoting or reporting an anonymous source.
Opinions about Gordon Brown's temperament and alleged behaviour are many and various. My hunch is, most people already know what they think about him and that few will change their minds as a result of Andrew Rawnsley's book ... or the second hand reporting of it (which, incidentally, makes allegations that do not appear in the book itself).
But why we should trust the allegations in the first place? In the past 72 hours or so, I've come across any number of people who'd LIKE the allegations and the very worst interpretation of them to be true - not natural GB supporters, then - but who still feel inclined to disbelieve a journalist, however honest and well respected, who can't/won't name his sources. Especially in the face of an on-the-record (non-denial) denial.
As Jeremy Paxman knew, there was no chance of Andrew Rawnsley naming any of those who'd spoken to him in confidence - protecting a source is a fundamental of journalism. Though to his credit, JP continued to challenge knowing that he was articulating the sceptical thoughts - mistrust - that many in the audience were doubtless entertaining.
Why are audiences so sceptical, so mistrusting? Especially since it's an undeniable fact that almost any story worth telling started life with a source anonymously briefing a journalist - the first ray of sunlight that is, according to US Supreme Court Justice Louis Brandeis, "the best disinfectant". And without whistleblowers and their like, journalism would have few means if any to challenge, even in small measure, the information asymmetry that pertains between people and power.
The problem is that for every example of outstanding investigative, accountability journalism that leans on anonymous sources, there are dozens if not hundreds of examples of unutterably lousy fiction purporting to be a journalistic account of 'sources'.
Good journalists are alive to the risks and for the need to be scrupulously honest and straight ... even when, as with an anonymous source, there's little chance of contradiction; the BBC Editorial Guidelines offer a fairly strict straitjacket. The constraints in the Reuters Handbook of Journalism are even tighter:
"You must source every statement in every story unless it is an established fact or is information clearly in the public domain ... Good sources and well-defined sourcing help to protect the integrity of the file from overt outside pressures and manipulation and such hazards as hoaxes."
And on single - usually anonymous - sources:
"For a single source story, the informant must be an actual policymaker or participant involved in the action or negotiation with first-hand knowledge, or an official representative or spokesperson speaking on background. Such information should be subject to particular scrutiny to ensure we are not being manipulated."
Both organisations' reputations rely on them being trusted, even when they're using anonymous sources.
But here's the thing. The language of responsible journalism like Andrew Rawnsley's or that which scrupulously follows guidelines like the Reuters Handbook - based on genuine sources and done in the public interest - is indistinguishable from that used to defend fabrication, intrusion and distortion, based on rather less genuine 'sources' and a different understanding of the public interest.
It is the problem of 'assessability' that Professor Onora O'Neill explored in her 2002 Reith Lectures: on the page, there is nothing to distinguish the best from the worst, the trustworthy from the disgraceful.
Journalists who take easy options - inventing a 'source'; turning a half remembered paraphrase into a hard "quote"; attributing their views to a source - then the role of journalism as a check on power is reduced for the simple reason that they erode trust in ALL journalism and not just in the newspapers or news organisations they work for.