“Wrong side of your Orwell, mister editor.” The New York Times falls down on the word torture.

These are my discussion notes to The Executive Editor on the Word ‘Torture’, a letter to Times readers from Dean Baquet, August 7, 2014.

10 Aug 2014 12:52 am 22 Comments

“It’s time to celebrate that the newspaper of record is no longer covering for war criminals.” That’s what Andrew Sullivan, who has kept watch on this story, wrote Thursday. The news he was celebrating: the New York Times gave up the ghost on euphemisms for torture.

Alright, we celebrated. For an hour, maybe. Now let’s ask what came to an end with this strange announcement. Terms like “enhanced interrogation techniques,” “harsh tactics” and “brutal treatment” had been preferred usage at the Times in news stories by its own staff about the treatment of detainees in U.S. custody after 9/11. “Torture” was removed as a descriptor that the Times itself would employ. The decision to reverse that came Thursday in a brief note to readers from executive editor Dean Baquet.

From now on, The Times will use the word “torture” to describe incidents in which we know for sure that interrogators inflicted pain on a prisoner in an effort to get information.

1. As Erik Wemple wrote, Baquet here “pledges that the newspaper will deploy the English language to describe things.” Dean Baquet8 1-30-07Wemple’s paraphrase points up the strangest part for me: the Times felt it had to exit from the vernacular to stay on the responsible journalism track. This I find hard to accept.

The baseline in daily journalism is supposed to be plain English, spoken and written well. Non-exclusive language is the norm. The market is the common reader and the reader’s common sense, not a specialized class of knowers vibrating in the power circle. It’s not incumbent on an already understood term like torture to prove itself neutral enough for newswriters, but on the specialist’s construction (“enhanced interrogation”) to prove itself relevant in these proceedings at all.

Even with scare quotes around it, a term like “enhanced interrogation techniques” starts with zero currency, extreme bloodlessness and dubious origins. A lot to overcome. In the years when the Times could not pick between it and torture — 2002 to 2014, approximately —  it seemed that its editors and reporters were trying to re-clarify what had been made more opaque by their own avoid-the-label policy decision. Thus the appearance of do we have to spell it out for you? phrases like “brutal interrogation methods,” meant to signal: this was really, really bad. So bad you might think it amounts to…

Baquet tries to explain:

The word “torture” had a specialized legal meaning as well as a plain-English one. While the methods set off a national debate, the Justice Department insisted that the techniques did not rise to the legal definition of “torture.” The Times described what we knew of the program but avoided a label that was still in dispute, instead using terms like harsh or brutal interrogation methods.

So for the fruits of avoiding a label the Times becomes a force for fuzzing things up. Early in a public reckoning with acts of state torture it decides it cannot call it that. Wrong side of your Orwell there, mister editor. To report what happened you have to first commit to calling things by their right names. The Times lost sight of that somewhere in a fog it helped to create. The editor’s note doesn’t explain how it happened. (See Barry Eisler on this point.)

2. You could see the reversal coming.

Baquet’s note doesn’t mention Obama’s concession on August 1. “We crossed a line and that needs to be understood and accepted. We have to as a country take responsibility for that.” The president is being more direct than ever: yes, we tortured people. The Senate Intelligence committee report, with lots of details about torture, will be coming out soon. Fights about its release could be making news for weeks. Linguistically, the Times was headed for a crash if everyone in the political system could talk of torture (and be quoted on it) but Times reporters couldn’t say that themselves. The game was up. When the reporters lobbied for release, what choice did Baquet have?

Marcy Wheeler in Salon, August 4, sensed the collapse:

For 10 years, journalists have willingly perpetuated this linguistic absurdity, even as more evidence came out proving the CIA used torture and not some fluffed up interrogation process, even as more and more neutral arbiters judged our torture torture.

The Senate Intelligence Committee has spent five years trying to understand and come to grips with the torture done in our name. Isn’t it time for journalists to do the same?

3. It’s easy to proclaim (and I am with those who say) that the Times showed excessive deference to government officials during the “make the picture fuzzier for our readers and let them decide” phase of reporting. But how did the editors think themselves into this mess? That is what I am trying to understand in these notes.

4. My contribution is this concept: the production of innocence. You can use it to understand what happened.

Alongside the production of news and commentary American journalists working in mainstream newsrooms have to continuously reproduce their own innocence. By “innocence” I mean some kind of public showing that they have no politics themselves, no views of their own, no side, no stake, no ideology and therefore no one can accuse them of unfairly tilting the news this way or that.

For newsrooms that still have it in their model, the production of innocence is supposed to chime with the publication of news. But what if the news is about a clearer and clearer case for calling it torture? This happened from 2002 to 2012. As the evidence became too great to ignore, the dispute in Washington over calling it torture escalated because of massive consequences up and down the chain of command. With avoid becoming party to a dispute a newsroom priority, the Times got caught on the wrong side of the evidence pile and added to the fog of euphemism. Manufacture of innocence, darker side.

Listen to a Times spokesperson explain it in 2010.

“As the debate over interrogation of terror suspects grew post-9/11, defenders of the practice (including senior officials of the Bush administration) insisted that it did not constitute torture,” a Times spokesman said in a statement. “When using a word amounts to taking sides in a political dispute, our general practice is to supply the readers with the information to decide for themselves. Thus we describe the practice vividly, and we point out that it is denounced by international covenants and in American tradition as a form of torture.”

It’s that intention — avoid taking sides in a dispute —  to which I draw your attention. It won out over 1.) “use the vernacular!” and 2.) a growing body of evidence. Until last Thursday. So what changed?

5. Innocence, I said, is a production requirement. You need to satisfy it somehow. But there are different ways to generate enough “points” to make a thing reportable. For example: if the press pack is doing it, you’re home free. That’s as innocent as you can get, even if the story is wrong. If the government admits to it, you’re innocent for telling the public it happened. If the president urges Americans to face the fact, that’s all the cover you need. When the Times could hit its innocence numbers in other ways, DON’T USE THE WORD TORTURE ON OUR AUTHORITY was dropped from the production routine.

6. From an exchange last year between Bill Keller, the executive editor under whom much of this happened, and Glenn Greenwald, who is deeply critical:

Greenwald: A failure to call torture “torture” because government officials demand that a more pleasant euphemism be used, or lazily equating a demonstrably true assertion with a demonstrably false one, drains journalism of its passion, vibrancy, vitality and soul.

Keller: Sometimes fair play becomes false equivalence, or feels like euphemism. But it’s simplistic to say, for example, unless you use the word “torture” you are failing a test of courage, or covering up evil. Of course, I regard waterboarding as torture. But if a journalist gives me a vivid description of waterboarding, notes the long line of monstrous regimes that have practiced it, and then lays out the legal debate over whether it violates a specific statute or international accord, I don’t care whether he uses the word or not. I’m happy — and fully equipped — to draw my own conclusion.

To which I say: accountability can’t start until things are called by their right names. The Times became a force — not the force, but one among many in the system — for confusion and delay in the public reckoning with torture. Until, as Dean Baquet wrote: “The Justice Department, under both the Bush and Obama administrations, has made clear that it will not prosecute in connection with the interrogation program.”

7. The Times actually used the word “torture” a lot during the period when the ban was in effect. This is important. The ban applied to a relatively narrow class of cases. Some of the situations in which it did not apply:

* Quoting someone who alleged: this was torture. (“We’re not saying it, he is.”)
* Citing reports documenting torture. (“We’re not calling it that. But this report does.”)
* Reporting on the debate: was it torture? did it work? (“We’re not saying. We’re just telling you there’s a dispute”)
* Publishing in op-eds and reviews other voices who do call it torture. (“We’re not saying it, this writer is.”)
* Editorials in the Times. (“That’s opinion. No problem.”)

A good example is a 2005 review essay by Andrew Sullivan: (My bold.)

Whether we decide to call this kind of treatment ”abuse” or some other euphemism, there is no doubt what it was in the minds of the American soldiers who perpetrated it. They believed in torture. And many believed it was sanctioned from above. ….At Guantánamo Bay, newly released documents show that some of the torturers felt they were acting on the basis of memos sent from Washington.

Was the torture effective?

If the Times was concerned about getting ahead of the legal system, how could it allow these straightforward uses of the term? Times readers somehow won’t be unfairly influenced when Andrew Sullivan does it in the book review, but they will be if the news staff tries the same thing? Doesn’t make a lot of sense. Unless you understand the production of innocence. The Times wasn’t trying to keep “Wake up, people! The U.S. did commit torture …” from readers, the way it might keep a rape victim’s name out of the news. It was trying to avoid making the statement on its own authority. As long as others took the responsibility — outside writers, human rights groups, quoted sources in a dispute, editorial boards — the innocence requirement was met and production went forward.

8. This has been going on for years. And not only at the Times but across the sector. In 2006 Eric Umansky (now at ProPublica) dug into the early patterns in coverage of torture for Columbia Journalism Review. His analysis, Failures of Imagination, is detailed and persuasive. For my notes, this part is key:

What is true and what is significant are two different matters. Everybody agrees that journalists are supposed to ascertain the truth. As for deciding what is significant, reporters and editors make that judgment, too, all the time — what story leads on the front page, or gets played inside, what story gets followed up. And when it comes to very sensitive material, like torture, many journalists would prefer to rely on others to be the first to decide that something is significant. To do otherwise would mean sticking your neck out.

Which is an innocence calculation, it you take my drift. Umansky is direct:

Any article, no matter how straightforward or truthful, that treats abuse as a potential scandal — even by simply putting allegations on the front page — is itself making a political statement that “we think this is important,” and, implicitly, wrong. To make such a statement takes chutzpah.

“Avoid taking sides in a political dispute…” This does not take chutzpah.

9. Bill Keller, who was top editor of the New York Times when most of this was transacted, whose decision-making was undone by Dean Baquet’s note to readers…  Keller has a new job. 6985189859_d03949b7b9_mFounding editor of The Marshall Project, a fascinating niche site with an important mission. (Their bold.) “There is a pressing national need for excellent journalism about the U.S. court and prison systems. “ I agree. This part caught my eye: “With growing awareness of the system’s failings, now is an opportune moment to launch a national conversation about criminal justice. There are numerous indications of the country’s appetite for reform.”

Obviously Bill Keller in 2014 is comfortable with that: “Launch a national conversation about…” He wants to equip with information “the country’s appetite for reform.” He feels he knows how to do the journalism part of that equation— for criminal justice. I can’t wait to see what he produces.

“With growing awareness of the system’s failings, launch a national conversation about…” That never happened for the New York Times under Bill Keller once it became clear that the United States tortured people after 9/11. Under Howell Raines, Bill Keller, Jill Abramson and Dean Baquet, the Times helped delay our national reckoning with torture. The facts were there. Required levels of innocence were not.

After Matter: Notes, Reactions & Links

Barry Eisler (former CIA, now a writer) notes something important in his post. The standard legal definition of torture “is pretty much what the Times wound up using anyway.” In other words, a judgment the Times says it’s able to make now is the same judgment that could have been made then to say that U.S. conduct met the legal standard for torture.

Seems relevant:

The Times archives has topic pages for coverage of torture and CIA interrogations. Most of the relevant coverage can be found there.

Listen to this quote:

What I believe in is deep reporting, and then if a reporter really digs, there often is what I call weight of evidence in stories that are about contentious subjects.

That’s Jill Abramson, former executive editor of the New York Times, speaking last week. “Weight of evidence” is exactly the judgment The Times dodged when the subject was whether to call it torture.

“To survey powerful actors with clear conflicts of interest and then defer to their characterizations betrays a newspaper’s charge: to determine the truth and state it plainly for the public.” — Conor Friedersdorf in the Atlantic.

Torture vs. torture euphemisms: see this infographic for percentages and charts.

NPR’s David Folkenflik was a media beat reporter for the Baltimore Sun in 2004. This is from one of his columns (May 26, 2004). It shows what was riding on the issue:

Rumsfeld also used misdirection — a “look at this hand, not that hand” approach — to brush off questions about whether U.S. troops had tortured prisoners in Iraq. Rumsfeld told reporters: “My impression is that what has been charged thus far is abuse, which I believe technically is different from torture… I don’t know if… it is correct to say what you just said, that torture has taken place, or that there’s been a conviction for torture. And therefore I’m not going to address the torture word.”

Yet it’s not hard to see torture in some of the pictures obtained and published so far by the media of abuses at Abu Ghraib. And the Geneva Conventions of 1949, and a subsequent international protocol of 1984, both of which have been signed and ratified by the U.S. government as law, do address the torture word. The 1984 document states:

“The term `torture’ means any act by which severe pain or suffering, whether physical or mental, is intentionally inflicted on a person for such purposes as obtaining from him or a third person information or a confession, punishing him … or intimidating or coercing him.”

The U.S. War Crimes Act of 1996 makes such conduct by a member of the U.S. armed forces a “war crime” punishable by fine, imprisonment or, in cases resulting in death to the victim, the death penalty.

Photo of Bill Keller used under a Creative Commons license. Photo of Dean Baquet courtesy of New York Times.


Listener’s guide to Christian Rudder explaining why OkCupid experimented with unwitting users

You use the product. We have the power. You can exit any time. Before then we can screw with you.

3 Aug 2014 5:09 pm 6 Comments

Alex Goldman: Have you thought about bringing in, say, like an ethicist to, to vet your experiments?

Christian Rudder, co-founder of OkCupid: To wring his hands all day for a hundred thousand dollars a year?

Fabulous exchange, right? As soon as NPR’s On The Media pointed it out, I wanted to know more about Christian Rudder’s elegant derision. I had read his post at the OkCupid blog, We Experiment On Human Beings!

We noticed recently that people didn’t like it when Facebook “experimented” with their news feed. Even the FTC is getting involved. But guess what, everybody: if you use the Internet, you’re the subject of hundreds of experiments at any given time, on every site. That’s how websites work.

Facebook by accident had this conversation. OkCupid would now have it on purpose. Then I listened to the interview. Three times. It’s that good.

Pertinent facts:

* Christian Rudder will soon be promoting a book, Dataclysm: Who We Are (When We Think No One’s Looking.
* He doesn’t shrink from public controversy.
* Tim Carmody analyzed why OkCupid’s admission didn’t bother people as much as Facebook’s, and why it should have.
* I come to this with views, having written for the Washington Post and spoken up at the Atlantic about the Facebook study that caused such a storm.

Listener’s guide to TLDR‘s July 31st podcast: Q & A with OkCupid’s co-founder Christian Rudder. It’s 14:20.

You can read the guide first and it affects the way you listen, or listen first, then read.

1. Why I love this interview: the entire thing is about the legitimacy of OkCupid experimenting on its users. They are never not talking about that. Perfect for a podcast.

2. Christian Rudder admits right off that he was “sensationalizing” in his we experiment on you! post. (Most writers don’t admit to that.) I’m trying to provoke, he says during the interview. We have to allow for this. ChristianRudderBut even someone straining to be provocative reveals themselves in their choice of heat-seeking styles. A man who volunteers that he was “sensationalizing” when he said to users, yeah we experiment on you, so what? is leaning into the scandal narrative, pushing confrontation with it, because he strongly suspects there is nothing there.

3. Rudder’s best point is that sites like his need to experiment on human beings to improve the product. How else would you do it? he asks, interviewing the interviewers. His attempt is to get you, the listener, to shift position. Don’t identify with the hapless user, unwittingly experimented upon. Climb into the chair of the site’s operators. We have this nifty algorithm and we hope it works. But we have to run tests to prove it. Some ethicist says that might be a problem. Is this going to stop you?

4. Alex Goldman and the TLDR crew are at their strongest on informed consent. They make that case come alive. But instead of pretending that he’s obtained consent (by having users check a “terms of service” box) Rudder jeers at the whole idea. Informed consent is a joke, he says. When you participated in those psychology experiments in college you had no idea what they were measuring. It was un-informed consent. That’s how websites work. They are semi-legitimate enterprises. The smarter users — or the users we want — already know this. They accept a certain amount of “getting screwed with” as the price of having an OkCupid that works. It’s the ethicists who don’t get it. The world has moved on. They wring their hands.

5. The interviewers are persistent. They try to solve the problem of consent. Why not level with people? Tell them you’re about to experiment on them. Ask them if they want to be included. Rudder shuts this down. Once they know they’re being studied, they will act differently. Let’s be clear: He doesn’t want to solve the problem of consent. He wants to recognize a power differential between the site and its users.

6. Rudder doesn’t put it this way, but he’s really sneering at the whole concept of user trust. Users don’t trust us to never put our interests before theirs. They know we sometimes do that. They know we don’t tell them about it. Mainly, they just want the site to work. We do too. End of story! Trust, “ethics,” legitimacy, consent: aren’t these terms a little full of themselves?

7. PJ Vogt had the best line: “I’m not comfortable being a guinea pig to the extent that you guys are comfortable making me a guinea pig.” But if he broke up with his girlfriend he would probably be back at the site, he said. That’s Rudder’s point. We don’t need you to be comfortable with our methods. We don’t need them to be seen as ethical or fair. Users won’t know if they’re being manipulated. That’s how websites like ours work. You use the product. We have the power. You can exit any time. Before then we can screw with you.

The production of innocence: tale of two headlines over Gaza

"Members of Congress fall over each other to support Israel," the AP said on Twitter. Then the AP decided that it could not say that. Why?

31 Jul 2014 11:04 pm 30 Comments

It’s early in the game. I have only been writing about this concept for four years. Okay, nine. Whatever! I keep at it. This is a work of pressthink that I am still trying to render properly for readers. Starts like so:

Alongside the production of news and commentary American journalists working in mainstream newsrooms have to continuously reproduce their own innocence. By “innocence” I mean some kind of public showing that they have no politics themselves, no views of their own, no side, no stake, no ideology and therefore no one can accuse them of unfairly tilting the news this way or that.

It’s not enough to proclaim innocence: we have no party, we take no side. In the style of journalism I’m talking about — still the house style at the AP, CNN, NPR, the BBC — innocence is a production requirement. If the requirement isn’t met, the work fails, and it can be sent back to the shop.

Sometimes innocence is built into the form. On CNN’s Crossfire, circa 2005, the show would open like this:

ANNOUNCER: From the left, Michael Kinsley… From the right, Mary Matalin.

This simple routine is a load-bearing feature of the show’s design, balancing the stresses on CNN’s reputation and restoring innocence nightly.

Any fan of NBA basketball has seen defenders put their hands up in the air ostentatiously, with funny facial expressions to match. It’s meant to show the refs: See? I’m not pushing. In news there are moves like that, and this is what I’m calling the production of innocence.

On July 29, the Associated Press sent out this bulletin:

About five hours later the AP caught itself:

“Members of Congress fall over each other…” is a characterization of the facts chosen by the AP writer and supported by the information in the story. If there’s any situation in American politics where they “fall all over each other” it’s support for Israel in the United States Congress. So this statement is a little saucy, a little cynical but it is accurate. The AP nonetheless decided it was “too much.”

The original header produced the news well enough but it failed to produce enough innocence for the AP. “Many U.S. lawmakers strongly back Israel…” is not more true than “they fall over themselves.” But it is more innocent. When the switch is made the AP feed suffers a loss of vivid. Its colors wilt. There is less voice, less urgency in the language. And the AP willingly pays this cost. Why? Lots of reasons. I isolate these two:

* In the sociology of newswork one of the first things you learn is that a firm involved in news production is uniquely vulnerable to public criticism and prone to costly mistakes. A news report is a first draft, an improvised understanding. It is frequently wrong or blind. Therefore an established firm in the news business needs regular and reliable ways to protect itself from the criticism it knows will come, including some criticism for which there is no good defense, nothing beyond: I know, but we didn’t have time!

* In the American setting media bias is a driver in politics, and culture war is where some people go to live. A major provider like the AP gets hit hard in the bias wars, so the principle, don’t give them ammunition! has to be built into its routines. That’s the production of innocence. I’ve rendered it this way:

AP SOCIAL MEDIA TEAM: “Members of Congress fall over each other to support Israel.”

INNOCENCE METER BOT: Headline script approved on accuracy. Failed on innocence. Please try again and re-submit.

AP SOCIAL MEDIA TEAM: “Many U.S. lawmakers strongly back Israel in Gaza war.”

INNOCENCE METER BOT: Headline script approved on accuracy. Approved on innocence. You are cleared to post.

William James used to call it the “cash value” of the concept. What can you do with it? Well, you can ask good questions.

Q. 1 The production of innocence has benefits that are obvious. Risk-reduction. Haven from criticism. But it also has costs. Loss of voice, loss of nerve. How do we know when the costs exceed the benefits?

Q. 2 And what if costs are rising in the field of innocence production? Isn’t that the whole point of making culture war on the media, to drive those costs up?

Q. 3 What advantages do born-digital newsrooms gain over legacy newsrooms when they decide they no longer care about the production of innocence– as say, Gawker, has?

Q. 4 When you finally come to the conclusion, there is no haven from criticism, the world doesn’t work that way any more, are the costs of producing innocence alongside the news still justifiable? (“BBC Trust says 200 senior managers trained not to insert ‘false balance’ into stories when issues were non-contentious.” Expensive!)

Q. 5 I know, I know: advertisers like the signs of innocence and advertisers pay the bills. How’s that going?

Q. 6 Let’s say you junk the innocence machinery. What gets put in its place? (My bet: “here’s where we’re coming from” + make a good argument + high standards of verification beats the old system.)

Q. 7 As they mount, reports that get “approved on accuracy” but “fail on innocence” represent colors of truth the news provider feels it cannot provide. Is that a trustworthy system?

Q. 8 Suppose you junk the production of innocence due to mounting costs and diminishing value. Now you have fewer means for avoiding criticism. Which means you have to reply more to criticism. But how do you do that well and still have time to produce the news?

I don’t know. But as I said, it’s early times.

First Look Media shifts direction some

And has none of this figured out.

28 Jul 2014 3:23 pm 8 Comments

Pierre Omidyar published a blog post today that gave an update on First Look Media, the company he started with Glenn Greenwald and others and backed with $250 million of his own money. Nine Months in, First Look is Still Very Much a Startup.

As regular readers of PressThink and my Twitter feed know, I am an occasional adviser to First Look, but not an employee. So factor that in. I am not a spokesman for First Look and did not clear this interpretation with them.

In my opinion, this was the most important part of Pierre’s post today:

…rather than building one big flagship website, we’ve concluded that we will have greater positive impact if we test more ideas and grow them based on what we learn. We are unwavering in our desire to reach a mass audience, but the best way to do that may be through multiple experiments with existing digital communities rather than trying to draw a large audience to yet another omnibus site…

For First Look the way to a large user base isn’t “one big flagship website” or an “everything you need to know” news app to go up against, say, the Guardian or npr.org. That kind of “omnibus” product will not be forthcoming anytime soon, today’s announcement said. Instead:

We’ll test an approach to journalism that starts with being part of well-defined communities of interest, understanding the people in them and serving their needs and aspirations in new ways. The digital world gives us unprecedented opportunities to meet this vision.

You begin with the users of news in a few well-defined communities of interest. By understanding their “needs and aspirations” better, you try try to generate more trials, more fruitful errors, and — if the method works — products and services that mesh with people’s lives more effectively.

This is not a part of Pierre’s thinking, so don’t use it to characterize First Look’s approach, but the argument is similar to my PressThink post from March: When starting from zero in journalism go for a niche site serving a narrow news interest well.

That’s my advice to individuals starting in journalism. Get yourself into a journalistic situation, first. A “journalistic situation,” as I define it, is when a group of people who share a common interest are actually depending on you for news. From there the rest will flow. What you learn by trying to provide a living community with news instructs you in what to try next. But you have to succeed in becoming their provider.

First Look will test an approach that “starts with being part of well-defined communities of interest.” Begin with a journalistic situation. From there a distinct approach may grow. Of course, journalists who are distinct and talented enough can sometimes create their own community of users through a unique mix of investigation, commentary, reportage— and personality. First Look has been testing this proposition too.

Which leads to the second item of news in Pierre’s blog post. Combined with a doubling of the editorial staff to 50, it was this:

…rather than immediately launching a large collection of digital “magazines” based on strong, expert journalists with their own followings, as we imagined earlier, we’ll begin by building out the two we’ve started and then explore adding new ones as we learn. Whatever direction our experiments lead us, we will continue to invest in our journalists and support their commitment to fearless, fact-based reporting. We will continue to fund deep investigative reporting and back it up with the travel and research budgets necessary to support it.

Building out The Intercept and the Matt Taibbi venture will be the primary publishing project for now as First Look turns one of its founding phrases, “fearless, fact-based reporting” into actual journalism (and other projects.) When key learnings from The Intercept and Taibbi come in, the collection may be filled out. Or not. Also: There doesn’t have to be an editorial brand called First Look, and there may never be.

This was the third bit of news:

I expect we’ll be in this planning, startup and experimental mode for at least the next few years as we explore how to become integrated into people’s lives in meaningful ways.

In other words, First Look has none of this figured out.

NPR downgrades and disables its ombudsman

"NPR is far stronger than this short-sighted and half-assed decision suggests. It has nothing to fear from an empowered ombudsman."

14 Jul 2014 10:38 pm 18 Comments

UPDATED Five times, JULY 15 TO 31

I always read job descriptions for open positions in journalism. They tell you a lot about which way the field is headed. Last week I came across this opening for the NPR Ombudsman/Public Editor position. Two phrases jumped out at me:

The primary responsibility of NPR’s Ombudsman/Public Editor is to connect NPR’s audience to NPR and to provide a forum for audience views. The Ombudsman/Public Editor serves as an independent reporter on behalf of the public.

The NPR Ombudsman/Public Editor focuses on fact gathering and explanation, not commentary or judgment.

Here it is again:

In addressing audience complaints about journalistic errors in NPR News coverage, the Ombudsman/Public Editor will gather facts and can interview key news managers. The Ombudsman/Public Editor will then explain any errors without passing judgment

What is going on with these phrases in bold, which appear to prohibit the ombudsman from criticizing the performance of NPR journalists?

Well, it’s a change in policy. In the past, the NPR ombudsman has routinely come to judgment in addressing complaints. A typical example from 2011, when Alicia Shepard held the position:

Lots of things drive NPR’s audience crazy. One I totally agree with is this: NPR often does a lousy job of identifying the background of think tanks or other groups when quoting their experts.

Here, Shepard is picking up on a complaint she’s heard from listeners and making a judgment: NPR is doing a lousy job! Until now, this was a normal part of the position.

As you may recall, NPR incorrectly reported that Congresswoman Gabrielle Giffords had died when she was shot at a public appearance in her district. The ombudsman gathered facts and explained what happened. She also came to several conclusions:

At the point the hospital confirmed Giffords was in surgery around 2:30 p.m, NPR should have done two things: sent out another e-mail alert correcting its mistake and when it next broadcast at 3 p.m., it should have said that NPR mistakenly reported Giffords’ death and given the new, correct information.

Neither of these things was done that day….

What NPR should have done it did not do. That isn’t fact-gathering or explanation or representing the listeners. That’s making a judgment. Again:

NPR had two sources, though neither was identified in any way, and should have been. And the newscast should have put the news in context, explaining that a tragedy had just occurred, the story was changing quickly, and this was what NPR knew at that moment.

Making a judgment doesn’t necessarily mean slamming NPR. Here’s the current ombudsman, Edward Schumacher-Matos, evaluating complaints about NPR reporter Mara Liasson appearing on Fox News.

My conclusion on Liasson’s work is simple:

Applaud her.

I find that her NPR stories were straightforward and based on solid reporting. Her Fox contributions were the same. She was smartly analytical, but did not take a position on issues or veer into opinion. Just as important, she did not tilt or load her characterizations of political figures such as President Barack Obama or Republican leaders.

Much of the complaints about Liasson, it seems to me, are really about Fox. The complaining listeners do not like Fox’s rightward stance, and especially the incendiary views of some of its prime time talk show hosts such as Bill O’Reilly. They tar Liasson by association.

I looked at her work. In my assessment she does not tilt one way or the other. I think complaints from listeners about her appearing on Fox are really about Fox, not her. That’s the ombudsman coming to a judgment, after investigation. Under the new rules, he couldn’t do that. He could explain how NPR views Liasson and gather facts about NPR’s guidelines but he could not comment on her performance or assess the validity of the listeners’ complaint, which was the whole point of his November 2013 column.

A further clue to the changes is in the new title. The ombudsman has become Ombudsman/Public Editor in NPR’s usage. Thus: “The primary responsibility of NPR’s Ombudsman/Public Editor is to connect NPR’s audience to NPR and to provide a forum for audience views.” I asked Jeffrey Dvorkin, NPR’s first ombudsman, what he thought. Dvorkin, who is now a journalism academic, helped create the position back in 2000. He said he was concerned that “the Washington Post weakened model of ‘readers’ representative’ is working its way into or even infecting NPR.”

In March of 2013 the Post announced that it would no longer employ an ombudsman who was independent of the newsroom and empowered “to critique the newspaper’s journalism and field readers’ questions.” Instead it would have a reader representative, “a staff member who will answer questions and respond to complaints.”

The journalist who took the position later said, “My primary mission is to respond to readers… I’m not [charged with] holding the newsroom accountable.” Dvorkin said that his job as NPR’s first ombudsman “was to evaluate and adjudicate NPR’s reporting.” That’s gone now. “A media organization that values its reputation should not be in a defensive crouch,” he added. NPR “may be opting for a safer harbor in which the ombudsman/public editor plays less of the lightening rod role.”

I was able to reach former NPR ombudsman Alicia Shepard, who is now working for a news start-up in Afghanistan, to ask about the changes. “To not comment diminishes the role,” she said to me via email. “The new job description appears to defeat the purpose of having an ombudsman.” I agree with that.

I also asked Margaret Sullivan, the public editor of the New York Times, about the “no passing judgment” rule in the NPR job description. She said:

My experience at The Times tells me that readers do seem to want the public editor to weigh in with a verdict, at least most of the time. Sometimes, of course, it’s appropriate just to present findings, or an explanation, or editors’ opinions. But I certainly appreciate the option to express my point of view and it’s always been central to the way The Times and its public editors have interpreted the role.

Seeking comment from NPR, I contacted their spokesperson, Isabel Lara. Our brief Q & A:

What were the concerns that led NPR to emphasize in the new job description that the ombudsman/public editor does not comment or pass judgment?

There were no concerns. We’re looking for someone who can offer a rigorous, independent account of our work. The ombudsman is positioned to help the audience understand how we approached our coverage and whether our reporting meets NPR’s journalistic and ethical standards. The ombudsman may reach certain conclusions based on a careful examination of that work, but should not be perceived as the ultimate arbiter of right and wrong.

Where in the organization did those concerns originate?

The changes were based on discussions with outside experts, senior news leadership and the President and CEO, to whom the ombudsman ultimately reports.

Who made the decision to include those phrases in bold and why?

The job description was approved by the President and CEO.

That the ombudsman should “not be perceived as the ultimate arbiter of right and wrong” is behind language which says: just gather facts and explain them, DO NOT JUDGE. Okay, NPR. If that’s your story….

My conclusions: NPR has downgraded the ombudsman position. Two former ombudsman agree with this. To understand why, just think about the effect that “your job is not to pass judgment” has on the pool of potential applicants. It’s likely that similar moves by the Washington Post helped clear the way. It’s possible also that dissatisfaction with the performance of previous ombudsmen contributed to the decision, along with a feeling that criticism rains down from all sides nowadays, so why do we need an in-house critic?

In my view, NPR is far stronger than this short-sighted and half-assed decision suggests. It has nothing to fear from an empowered ombudsman. Its own internal standards are much improved since the position was created. Its ethics handbook — a public document — is a model of transparency and accountability. Whoever is responsible for the downgrade made a bad call.

UPDATE I, July 15. Many people are suggesting to me that the reason the ombudsman position was clipped was in reaction to this massive investigation by Edward Schumacher-Matos. It’s known that NPR executives were exasperated with the ombudsman over it. So maybe it was a factor:

That’s plausible. However, if NPR executives wanted to steer the next ombudsman away from investigations like that one they could have written into the job description: “The NPR Ombudsman/Public Editor does not re-report NPR stories.” They went far beyond that.

UPDATE II, July 16: NPR changed its mind. It says the part about “not providing commentary or passing judgment” was a mistake to include. The job description will be changed. (Check: it is changed.) The new CEO, Jarl Mohn, issued this statement to Media Matters:

The Ombudsman is a critically important role at NPR and the expectations of the job have not changed. The Ombudsman must be fully independent and fully transparent in order to do their job on behalf of the public. The language in the current job description about not providing commentary or passing judgment is a mistake and we are removing it. I take this position very seriously and am committed to recruiting an outstanding journalist for the job and ensuring he or she has the resources required.

Good move. Former NPR CEO Vivian Schiller comments:

UPDATE III, July 18. In a conference call with affiliates before the CEO’s decision and statement, top NPR executives lash out at me, call my reporting “lazy” and ask public radio colleagues to give them the benefit of the doubt. Link.

UPDATE IV, July 19. Jeffrey Dvorkin, the original NPR ombudsman (and a believer in the position, an active member of this organization) asked me for comment on the conference call with affiliates at his post: NPR Learns About (New) Media Accountability – the Hard Way. I re-print it here with some tiny changes.

It was my understanding that ‘fact gathering and explanation, not commentary or judgment’ is an idea well known in the public radio community. There is not a lot of ambiguity about what it means. But just to make sure, I asked some people with NPR experience. So I like the chances for my interpretation over Kinsey Wilson’s ‘The job description was in no way meant to diminish the role, limit the independence or handcuff them.’

Here’s what I did in reporting my post. I will leave it to you to decide if I was being ‘lazy.’

* To make sure my impression (that NPR ombudsmen routinely comment and make judgments) was correct, I reviewed dozens of past columns and found some typical examples.

* I contacted several former NPR ombudsmen for comment and quoted two who would go on the record. From one I got an earlier job description for the position.

*I contacted several ombudsmen or former ombudsmen at other national news organizations and quoted one who went on the record.

*I know more than a few people who work at NPR and they know me. I knew they wouldn’t comment on the record but I talked with them to check my assessment against theirs and make sure I wasn’t crazy.

* I contacted NPR’s spokesperson around 10:30 am July 14, and said I wanted to post the piece that evening, so could she please get back to me by 5 pm. I also told her that two former ombudsmen interpreted the language the way I did, so there was zero chance that NPR would be surprised by my take. I later spoke to the NPR spokesperson by phone to clarify what I was asking about.

* I received the NPR statement around 5:30 pm July 14 and published it in full later that night.

UPDATE V, JULY 31. NPR CEO CEO Jarl Mohn is interviewed about his decision on NPR. He leaves no doubt. “I thought it was a mistake. I don’t think it was the right decision. [Holders of the position] need to reach conclusions. And that’s what we’re paying them to do. I think that role, public editor, is hugely important to the credibility and integrity of this organization. Anything that we do that diminishes that is wrong. We reversed that decision.” (Go to 10:22 on the clip to hear his answer.)

‘Democrats argue. Republicans contend. We have no idea.’ A he said, she said at the Times.

Classic forms of he said, she said are not so much a "sin" against high practice as an increasingly crappy level of service for what is supposed to be a high end product: New York Times reporting.

11 May 2014 7:39 pm 33 Comments

Saturday the New York Times published on its front page an article by reporter Jeremy W. Peters about Republican Senator Rand Paul criticizing his party for backing laws that make it harder for some people to vote by requiring forms of identification they may not have. Unquestionably, this was news. The Times report included this paragraph:

Few issues ignite such passion among the base of both parties. Democrats argue that the laws are intended to keep poor voters away from the polls because they often have difficulty obtaining identification. Republicans contend cheating is rife in today’s elections.

Excuse me: Would you happen to know, New York Times, whether fraud at the polls is “rife in today’s elections?” Is that something I should expect you to know, seeing as you are the high-end product in the national news marketplace? Or is Democrats argue/Republicans contend/We have no idea a good enough standard, and it is my expectations that are out of scale?

In this article, at least, the Times does not know whether cheating is rife in today’s elections. But it knows of a passion for polarizing the issue among the bases of both parties. This helps makes it a classic in the “he said, she said” genre.

Look, we have no idea who’s right. How would we? Figure it out for yourselves! Don’t be asking us to sort out what’s real from what’s fiction. We’re just New York Times journalists. We don’t do “there’s no basis for that.” We do “Republicans contend…”

I’m satirizing but to make a point: this standard isn’t good enough. At least since the launch of Politifact.com in 2007 — which does do “sorry, there’s no basis for that,” sometimes — it’s been made clear to mainstream practitioners in the U.S. that the classic forms of he said, she said are not so much a “sin” against high practice as an increasingly crappy level of service for what is supposed to be an upscale product: New York Times reporting. If you can say (reliably) there’s no evidence for… and you don’t, how well have you done by Times authority?

Here I hand the mic to a fellow blogger of this one sad but (we think) telling paragraph, Felix Salmon, now of Fusion. Felix broke it down proposition by proposition: False equivalency in the NYT. (“How Jeremy Peters’s voter ID reporting is even more wrong than you think.”) I urge you to read his post and come back to this one.

Meanwhile, this report from — hey! — the New York Times in 2007 testifies:

Five years after the Bush administration began a crackdown on voter fraud, the Justice Department has turned up virtually no evidence of any organized effort to skew federal elections, according to court records and interviews.

Also this from Politifact/Georgia rating as “mostly true” the claim that in-person voter fraud is a very rare phenomenon, so rare that “only 10 cases of in-person voter fraud have been proven nationally.”

Or this in 2013 from Politifact/Texas: “By our reading of the attorney general’s records, 18 instances of voter fraud have been confirmed in Texas since 2002.” That’s an average of 1.6 cases per year. In a state where more than a million votes are cast in an off year. Rife?

Or read what law professor and election law scholar Richard L. Hasen, author of The Voting Wars: From Florida 2000 to the Next Election Meltdown (Yale University Press: 2012) has to say on how “rife” it is:

Federal Judge Lynn Adelman looked at the evidence from Wisconsin and reached a conclusion unsurprising to those of us who study how elections are run. “Virtually no voter impersonation occurs in Wisconsin,” Adelman wrote, “and it is exceedingly unlikely that voter impersonation will become a problem in Wisconsin in the foreseeable future.”

Wisconsin is not alone in lacking such evidence. When the United States Supreme Court considered the constitutionality of Indiana’s voter ID law in 2008, the state conceded there was no evidence, ever, of impersonation fraud in the entire state.

As Salmon says, you don’t have to pretend that there is a lot of fraud to be in favor of more controls. “In fact, you don’t even need to think that cheating exists in order to support such measures. It’s entirely rational to support a voter ID law even if cheating is rare or nonexistent, on the grounds that cheating is just too easy right now and that you want to make it harder.”

So what is that exceedingly crappy paragraph doing there on the newspaper-of-record’s front page? Salmon says it’s laziness. (“He-said-she-said is so easy, for a journalist on deadline, that both journalists and editors tend not to really thinking about exactly what they’re saying.”) Certainly ease-of-use is part of the device’s fading delights.

Here’s how I described the appeal of he said, she said in 2009. It makes the story writable on deadline when you don’t know enough to sort things out. In a “he said, she said” classic:

* No real attempt is made to assess clashing truth claims in the story, even though they are in some sense the reason for the story. (Under the “conflict makes news” test.)

* The means for assessment do exist, so it’s possible to exert a factual check on some of the claims, but for whatever reason the report declines to make use of them.

* The symmetry of two sides making opposite claims puts the reporter [and the user] in the middle between polarized extremes.

I question whether that between-two-extremes territory, the “you figure it out/for us partisan polarization rules” space is valuable turf in the news business. I doubt that it’s “safe,” either, if you mean by safe: won’t do the brand harm. I think it’s likely to corrode trust over time. A conventional explanation for he said, she said says: it may be lazy or incomplete, but it is also a safe middle ground place to land so you can get the damn paper out!

But it’s not that safe. Democrats argue/Republicans contend/We have No Idea… increasingly won’t cut it for the Times, or its competitors like the FT, the Wall Street Journal, the Washington Post, Bloomberg. The upscale, high-information readers the Times wants to charge more money to, the core loyalists who are being asked to finance more of the operation— these users are increasingly likely to know about various preponderance-of-evidence calls independent of whether the Times knows enough to include that review in its reporting. When this kind of reader comes upon he said, she said reporting on a big story where it’s CONTENTS UNDER PRESSURE, as with the right to vote: bad moment for the Times brand.

My sense: What was acceptably lame under market conditions that Bill Keller began with is a more corrosive practice today. I think The Masthead knows it. This is one of the reasons they created the Upshot, where preponderance of evidence, not a summary of partisan talking points, is supposed to be the baseline practice.

A good way to prove it: The Upshot looks at the evidence and makes a call on “Republicans contend cheating is rife in today’s elections.”