Welcome back for another installment of Ask TON. Here’s our latest question:
Should anonymous (unattributed) quotes be used to develop stories which would not be possible without them, such as when individuals are unwilling to go on the record with negative comments?
Anonymous sources need to be treated with care and used extremely rarely, but there should be no unilateral rule against citing them. Reporters should always press their sources to be on the record, over the course of multiple interviews if necessary. And it’s very important to understand and independently confirm, if possible, the request for anonymity. Does the source, for instance, really have a substantial reason for making this negative comment, or does he or she just have an axe to grind with someone else in the story? As an editor I rarely permitted anonymous sources—only if, for instance, we had confirmed that the source was indeed at risk of losing his or her job, or would suffer other severe consequences as a result of speaking out. The bar must be set very high.
Yes, but you must explain in the story why the source is anonymous. Unless there are some exceptional and explainable circumstances preventing it, the fact(s) should be second-sourced.
Since I work in Washington, DC, where political coverage is rife with anonymous sources, I can see both the danger and appeal of citing these sources. A reporter should seek to avoid them at all costs because they can weaken the story in the eyes of the reader, but I would never go so far as to ban them from a publication—sometimes, they are an absolute necessity in reporting known information. Ultimately, the goal of a journalist is to impart crucial or interesting knowledge to an audience; if doing that is only possible with anonymous sources, then so be it. In the end, it’s up to the reader to judge the credibility of a story, author and his or her sources. However, a reporter risks his or her reputation if anonymous sources prove wrong (consider Judith Miller)—if one does resort to anonymous sources, they should be identified as much as possible to note any potential biases.