Sunday, July 06, 2008

Journalism and its God words

It's not a massive surprise that the NYT's ombud has weighed the case of the CIA interrogator identified (over his and the agency's protests) on the front page two weeks ago and found that the paper did the right thing. Ethics isn't statistics; it's entirely possible for people to put the same data into the same matrix and get radically different results. And Clark Hoyt has differed with the paper's decisions often enough -- this recent case involving the op-ed pages stands out -- that it would be unfair and unwise to dismiss his judgment out of hand.

What makes this matter interesting is the manner in which both Hoyt and Bob Steele of Poynter justify the decision, because they both rely on a particular set of "God words"* -- terms that bring forth a certain amount of awestruck bowing and scraping merely by being invoked, regardless of whether the user or the audience can define them in any practical way or explain their relevance to the case at hand. In this case, the core concept is "credibility." And the trouble is that no one who invokes it bothers to explain what it is, how it's measured, or why the Times's solution is one that gives "credibility" a boost.

Back in the golden days of newsroom anthropology, Gaye Tuchman suggested that "objectivity" wasn't really the positivist concept that the empirical world might recognize as much as it was a "strategic ritual": You can't call the apartment owner a slumlord, but if you make a couple more phone calls, you'll be getting somewhere. So it is here with "credibility." When Hoyt invokes the concept, he's suggesting there's some sort of linear relationship between credibility and the use of names:

Scott Shane, the reporter, and his editors said that using the name was necessary for credibility.

With all due respect to The Ridger and the host of other folks who are ready to retire "begging the question" to the Old Logical Fallacies Home -- that's begging the question, or supporting an unsupported assertion with another unsupported assertion. We needed the name to make the story credible. Why? Because names make stories credible!

It's a handy myth of American news practice: Names make news. It's one reason American cop reporting looks the way it does. British cop reporting, partly because of British libel law, looks vastly different. Nobody's ever suggested that public safety in Britain is weakened because crime suspects aren't named before trial, or that crime reports in the British press are less credible*** because they don't contain enough names. Nonetheless, we hang on to the idea that names are a factor in credibility.

I can't disprove that concept, largely for the same reason no one on the other side can "prove" it: No one has yet shown that it can be reduced to testable concepts and tried out in the lab. But as a counterexample, here's some NYT-brand journalism from -- how time flies! -- December 2001:

An Iraqi defector who described himself as a civil engineer said he personally worked on renovations of secret facilities for biological, chemical and nuclear weapons in underground wells, private villas and under the Saddam Hussein Hospital in Baghdad as recently as a year ago.

The defector, Adnan Ihsan Saeed al-Haideri, gave details of the projects he said he worked on for President Saddam Hussein's government in an extensive interview last week in Bangkok.

Sound familiar? Named source and all, it's the prewar work of Judith Miller, now thoroughly disowned by the Times. Again, it doesn't "prove" a thing, although it should suggest that anyone who proposes a direct relationship between source naming and "credibility" has a lot of uphill ice-skating to do.

Steele's argument is more detailed but not substantially different. He gives pride of place to the journalistic duty of providing "accurate, precise and substantive information about a significant issue and event." Fair enough, though "accurate," "precise" and "substantive" are drastically distinct concepts -- as is evident a bit later, when Steele points to the substance of the story (undeniable) without bothering to explain how it relates to the subject's request (again, the subject wasn't a source for the story) not to be identified:

The Times makes a compelling argument in the seventh graph on the value of this story: “The story of Mr. Name's role...provides the closest look to date beneath the blanket of secrecy that hides the program from terrorists and from critics who accuse the agency of torture...”

And there’s an equally compelling justification for the story in this sentence from the 10th paragraph: “Mr. Name's success at building a rapport with the most ruthless of terrorists goes to the heart of the interrogation debate.”

OK, and that relates to the name exactly how?

This story is powerful in its substance. That power is enhanced by the specific use of Source Name's name connecting him to the “ad-hoc” program and his surprising role (given his experience and skills) as a key interrogator in the anti-terrorism effort. He is a central character, and using his name gives readers a clear focal point.

Those are God words. I like power and substance and clear focal points too -- much as I like kittens and birthday cake and prewar flatheads. The argument doesn't establish them; it merely invokes them.

Using his name — rather than a pseudonym or just referring to him by title — also heightens reliability and validity in the reporting process.

This is even worse. "Reliability" and "validity" are specific things. Reliability is the accuracy with which you measure a concept. If I can use your rulebook and get results similar to yours, or you can rerun your tests in a month and get the results you got yesterday, we've established a kind of reliability. Validity is how closely that concept resembles the slice of reality that you're measuring. The proportion of correct baseball scores to incorrect baseball scores in your paper is a valid measure of how accurately you cover baseball. Are important national security stories made more accurate by the use of names? (See Miller above.) Are they about more important issues? Oh, come on.

The story is more believable. (Why, who says, and how do you measure it?) Granted, the Times chose to use his nickname, Deuce, rather than his real first name, apparently to offer him some level of identity protection. But the use of his surname and nickname helps the paper achieve one of its primary objectives: bolstering the credibility of its reporting. (How and why does it do that? How much "credibility" would you get if you used the nickname and a made-up surname? What population would you like to test that effect in? What standards can you propose for journalists who don't work for the Times but still have to make ethics decisions?)

See what we mean by "God words"? "Believable" and "credible" are things we are obliged to respect, but they're not for us mortals to question. They belong to people who work for Poynter or the Times. Go about your business. You'll understand it all by and by.

Partly because this is so clearly not an "objective" decision -- because there's so clearly no relation between words like "believable," "credible," "reliable" and "valid" and the decision-making process that ought to be involved -- I'll toss in my personal response here. I wouldn't have found the story less credible without the name. If anything, I would have found it more credible. The Times has a tendency to overreact when it's under threat, and I'd like some assurance that the Times isn't turning down a legitimate request for anonymity because it's trying to make up for the sins of Judy Miller.

I started doing journalism for a living back in the late Cretaceous, and I don't yield to anyone -- Times, Poynter, Pope of Rome, whoever -- in my regard for "accurate, precise and substantive information." (One of the things I do in my current life is to keep track of assorted ways in which news organizations transgress against accuracy, precision and substance, but if you're a regular visitor here, you've probably noticed that.) The Times provided an important service with this story, but it followed up with what, on the evidence, was an inappropriate ethical response. I hope, as does anyone who's been around the table for an ethical crapshoot in journalism, that it isn't proven wrong.

* A term I first heard from a rather scary methods prof** a few years back
** OK, Bill Benoit
*** Credibility issues specific to the British press are more often down to the habit of some sectors -- we are looking straight at the Daily Mail and Mail on Sunday -- of "making stuff up," which is a different issue altogether.


Blogger The Ridger, FCD said...

I don't want to retire the fallacy - just the translation of its name, which is bad for several reasons. Call it petitio principi - you say ad hominem, right, and propter hoc?

Otherwise, great post. I like the term "God words" - terms you can't question. That explains a lot.

8:46 PM, July 07, 2008  

Post a Comment

Links to this post:

Create a Link

<< Home