[This article originally appeared in the Summer 1984 issue of Nieman Reports.]
Researchers from The University of Iowa journalism school conducted a study in 1982 for the American Society of Newspaper Editors. Analyzing stories from a half-dozen large papers scattered around the country, the researchers found upon interviewing the reporters that in about one-third of the stories the anonymous quotes could easily have been avoided. In some cases the reporters did not know why their sources needed protection. In others, the reporters conceded that they could have persuaded the sources to go on-the-record. In still others, the information attributed to blind sources was not crucial and indeed was duplicated elsewhere in the stories.…With all the fretting about believability, I think we ought to take a close look at the number of times our news stories attribute information not to people with names and titles, but to “sources”—and to their redundant cousins, “informed sources.”…
Last fall our staff at The Philadelphia Inquirer had a series of discussions on fairness and accuracy. One of the subjects we spent considerable time on was the need to limit unnamed sources to those giving information truly essential to our readers and not obtainable in any other way. In the ensuing months I noted with satisfaction that we did seem to be cutting down on “sources” stories. It took a routine piece two weeks ago to make me realize I was being complacent. The story was about actor Harrison Ford spending some time with Philadelphia police detectives to prepare himself for a role in a new movie. In the five-inch story, one of our police reporters, who happens to be an exceptional digger of facts, got carried away in quoting a “source” and a “police source.” To these anonymous observers he attributed such hardly crucial bits of information as the plot of the forthcoming film and the fact that Ford had the police commissioner’s approval for his field work, as if he could have done it otherwise. Once again, we’re trying to get the word out on our concern about the “sources” problem.…
Tonight I have…a set of guidelines to offer for your consideration.…
The first guideline I offer in evaluating confidential sources is this: The use of unnamed sources in a news story should be a last resort, not just an easy alternative to documenting the information from the public record or quoting someone willing to be named. In short, there is no substitute for digging. In the brief story about the actor who rode along with detectives as they went about their homicide investigations, there was no indication that the reporter had even tried to interview the actor himself or his agent or the movie company. Any of these would have been more knowledgeable about details of the movie than the secret police source. As I noted earlier, the University of Iowa researchers [See accompanying box] found repeatedly that the reporters could have gotten people to talk for the record if they had tried. The competitive rush for Watergate disclosures induced many reporters to let slide the traditional standards for attribution, as David Shaw [the media critic of the Los Angeles Times] noted. He said, “Getting into print first, with a story from an unnamed source, was often thought to be better than being second, with a story from a named source.” Shaw also noted that Watergate spawned a whole generation of young “investigative reporters” who felt that their editors and readers would be impressed by their savvy in referring to “informed sources” even though they were perfectly willing to be quoted by name.
The second guideline: It should be clear that the source’s physical or economic well-being might be jeopardized if his or her name is revealed. Thus we apply another test, one that must take place before the information is accepted with the stipulation of confidentiality. The writer should not simply assume that a source would be “more comfortable” not to be quoted by name; there should be evidence of real jeopardy.
The third guideline: The information provided by the unnamed source must be very important. The story should be one that helps a newspaper’s readers make informed decisions about their government or community. The information from the source should be crucial to the story, not tangential to the theme. Again, this is a sort of “needs” test that we should apply along with the first two before allowing a veiled source to be quoted in the paper. It is intended to separate the truly significant, essential story or passage within a story from the nice-to-have-but-not-really-necessary. At the Inquirer, we learned the hard way to apply this rule. In a piece about why major motion pictures were slow in reaching Philadelphia theaters, we spent the first two-thirds of the story’s length expounding on what all the named sources agreed was the crux of the matter: The distributors control when and where a movie will be shown, and they allow films to spread out into the country only after making their splash in the media centers of New York and Los Angeles. Near the bottom of the story, we mentioned that when it came to asking a favor to get a particular movie earlier than usual, Philadelphia exhibitors were not likely to be successful because they were such an irascible bunch. The story quoted an anonymous source several times in alleging that a certain exhibitor paid bills late so as to earn interest on the money withheld, and used its market clout to violate its contract by cutting short the run of a film that turned out to be a poor draw. To our chagrin, we learned after publication that we could not substantiate the accusations made by the anonymous source. And we found out that the source was in fact a competing exhibitor, something the editors of the story hadn’t known or asked about at the time. It was an object lesson for us. We realized that the information, even if it had been scrupulously true, was simply not essential to the story. For that matter, the story itself was not one that, to quote the guideline, “helps a newspaper’s readers make informed decisions about their government or community.”
The fourth guideline: To help readers evaluate the information, the unnamed source should be described as fully as possible without giving away the identity. So often we attribute statements simply to “sources” or “informed sources” or “reliable sources.” I would argue that semantically they are the same thing; we should not be quoting anyone who is not informed or reliable, and our readers should be astonished if we did. Using those terms amounts to nothing more than a plea to readers that we are not making the whole thing up. They impart no information. Of course, we should not risk giving away the identity of a person to whom we have promised confidentiality, but usually there is a way of characterizing that person that does not isolate him or her. Instead of a “source,” why not say “one of the participants in the negotiations” or “a police officer familiar with the department’s procedures in administering promotion tests”? Assuming that more than a handful of people fit those descriptions, the added information helps readers weigh the source’s credentials. The less we ask the readers to depend solely on our word that the source is qualified, the better off we are.
The fifth guideline: If an unnamed source is quoted making derogatory statements about someone, such a statement must be one that enhances the public’s understanding of a crucial issue. Here, reporters and editors must apply additional safeguards: The statement should be corroborated by public record or by named sources, or the source should have an impeccable record of reliability as well as a direct knowledge of the facts. On this point I would probably encounter disagreement from many journalists, who would argue that under no circumstances should a blind derogatory quote be permitted. I accept the principle but feel that there has to be flexibility to deal with the rare exceptional situation. The additional safeguards I mentioned were intended to reduce the possibility of unfairness. Of course, in any situation in which a person is mentioned in derogatory fashion, he or she should have an opportunity to respond.
Now, the sixth and final guideline: Reporters have to be free to use their judgment in granting confidentiality to sources while gathering the facts. However, it should be understood that the agreement of confidence is between the source and the newspaper, and that the reporter can be expected to identify the source to his or her editor. This is yet another area where honest men and women disagree. There are editors who insist that no one on the paper but the top editor can authorize confidentiality. Although recognizing that there are horrible abuses—which are, of course, the point of this lecture—I do not see how such a policy could help but inhibit a newspaper’s ability to gather the news. Not to allow discretion would be to put our reporters in a straitjacket. The reporters, in my opinion, have to be relied upon to make judgments on the spot. The reporter’s editor would naturally be bound by the agreement of confidentiality granted by the reporter—that is, he or she could not attach the name to the quote—but nevertheless has the option of deciding whether to use the quote at all. On the other extreme, some journalists would argue that the agreement of confidentiality is between the source and the reporter as an individual and no one, not even the reporter’s editor, can be given the identity. Although I would not insist that an editor has to know the identity of each and every source quoted, I do feel that the editor has the prerogative of asking—and getting—an answer. Reporters don’t put stories into the paper, editors do.
What we have in the guidelines I have laid out is a series of tests to apply in the reporting and editing process. We methodically ask questions:
- Is there any way to get the statement on-the-record?
- Does the source seeking anonymity really stand in jeopardy if identified?
- Is the information really important?
- Can we give the readers some idea of how qualified the source is?
- Is the information derogatory to any individual and, if so, have we gone extra lengths to make sure we are being fair?
I am convinced that if every reporter and editor went through this checklist before publishing an anonymous quote, our papers would contain a lot fewer anonymous sources. And the flow of truly essential information to the public would not be diminished in any significant way.
In closing, I want to suggest one more test, this one devised by Richard Smyser, Editor of The Oak Ridger at Oak Ridge, Tennessee. In commenting for the National News Council study, Smyser said he divides people who give us information anonymously into two distinct groups: “sources” and “sourcers.” He has a high regard for “sources,” whom he describes as “good guys” who have information of importance to the public on things that are amiss but whose careers or jobs would be jeopardized if they were identified as the communicators of that information. He has revulsion for “sourcers,” people whose goal is to use the press, and ultimately the public, as a means to an end. One way he has to tell them apart is to go back periodically and look over all the information his paper has printed without attribution. “Give it the test of time,” he says. “Read last year’s non-attributed news this year and see how it stands up.”
The on the spot tests I have spoken of—and Dick Smyser’s test of time—have a mutual goal: to help us achieve a higher degree of reader confidence. I am convinced that by needlessly resorting to unnamed sources, we undermine our cherished credibility and dilute our effectiveness as an institution. For us as journalists, there can be no higher mission than to guard, and reinforce, our reputation for truth.
Gene Foreman, Managing Editor of The Philadelphia Inquirer, gave the above lecture in March at the Journalism Ethics Institute, Washington and Lee University.