So how are we to know seriously to view the allegations? Here, I think the behavior of the women is at least as telling as specific details about what Cain said, how he gestured, or how large the pay-outs were. Few women relish the prospect of lodging a sexual harassment case against their boss. Few would want to look for a new job after leaving under those circumstances, settlement in hand or no. What Politico’s story tells us is that Cain’s conduct was sufficient for not one but two women to choose that course of action.

That, to my mind, also addresses a criticism Shafer raised in a phone conversation: if the allegations are unfounded, “it’s incredibly difficult for Herman Cain to get his good name back.” The press should of course be mindful of such concerns, especially when the accusers are not on the record. (And for that reason this bit of piling on from the AP, reporting anonymous, uncorroborated allegations from a third woman who never lodged a formal complaint, is journalistically dubious. Even worse has been the media’s abetting of one talk show host’s calculatedly unspecified and apparently overstated complaints about Cain’s behavior, skewered here by Gawker’s John Cook.) But when two female subordinates leave a workplace with confidential settlements after complaining about their male boss’s behavior, and he refuses to address the issue when asked about it, that should tarnish his name a bit. It doesn’t necessarily mean he’s legally culpable. It doesn’t necessarily mean he’s unfit for office. But, at a minimum, it tells us that a couple of his employees were upset enough about his behavior to proceed down what can be a pretty unpleasant path, and see that path to its conclusion. And that’s something.

This sort of heuristic is all the more useful in stories like this one, where more detail will not necessarily lead to more clarity or consensus about what happened. The line between what’s simply uncouth and what’s offensive or even illegal is famously subject to dispute and sometimes dependent on context that can be understood only by the participants, no matter how complete the journalistic retelling. Meanwhile, the meaning of the settlements’ size turns out to be ambiguous, even with a specific dollar figure attached. Shafer, in a chat at the Poynter site and in conversation with me, said that a settlement of $35,000 in one of the cases could indicate a weak claim, given the potential cost of litigation. But Engelberg, in an e-mail to me, wrote that the association’s decision to pay the woman a year’s salary “makes it clear that this was more than a nuisance case.” (While we’re trying to divine meaning from these numbers, I’m not convinced of Shafer’s point that the dollar figure “indicates that the accuser’s case wasn’t so great that she could have won more at trial.” It might indicate that, or it might indicate that she didn’t want to deal with the public, time-consuming aspects of a trial any more than the restaurant association did.)

Of course, Politico’s article would have been better with more specifics. But if the core bit of news, about the allegations and the settlements, was all Politico could pin down, it was still worth reporting. Engelberg wrote to me that the non-disclosure agreements were not an “insurmountable barrier” to unearthing more details prior to publication, and that some of the information that has since come to light, like the size of the settlements, “would have been immensely helpful to readers of the first story.” Like Shafer, he pointed to The Washington Post’s coverage of the Bob Packwood scandal, which landed with a devastating, incontrovertible wallop. And he suggested that the lack of specifics may explain why many Republican voters appear to be brushing the story aside.

But the fact that further details have come out in the wake of Politico’s reporting doesn’t prove those details could have been uncovered prior to the first story’s publication. On this point, Politico’s Harris had a persuasive reply. “Do [Shafer and Engelberg] hold their views strongly enough that they would have been prepared to stop newsworthy information about a presidential candidate from going into the public domain because only some questions, and not all, were answered?” he wrote via e-mail. “The main issue in their critiques is whether the right call is to share no newsworthy information until we can answer all the questions that a curious person might have. In my experience, that is just not how reporting works—answers come out over time, by returning over and over to important questions with more reporting and more stories.”

Greg Marx is an associate editor at CJR. Follow him on Twitter @gregamarx.