...or, when a quotation is more credible than what was originally said.
Let's say that a former national security official, with high security clearance, writes a book. The first draft of this book contains a story (never mind what) that is deemed to reveal classified information, so White House lawyers make him excise it. However, the story has already appeared in an alternative publication, so the official just substitutes a quote from that article. You can't knock someone for quoting something that's already public, can you?
I'm not so sure. Look at it this way: Alternative publications aren't always very credible. But when a high-level official quotes the alternative publication--if he's better placed with respect to classified information than the original reporter might be--he's more credible than the original publication was. The official, we know, knows what he's talking about; we didn't know that about the original reporter. So the mere act of quotation effectively puts out new information.
And note that the official doesn't have to say anything that's not completely public knowledge in order to affect his readers' epistemic situation. Suppose the reader knows that the reporter wrote that p. The official then writes: "The reporter wrote that: p." The reader already knew what the official literally says. Now she comes to know that the official also thinks that p; if the official didn't think that p was true, then it would be misleading to quote it. Effectively, the official implicates: "The reporter wrote truly that p." If it would harm national security interests to have it known that this official thinks that p, there would be a legitimate reason to excise the quotation from the reporter.
So here, repeating a quotation makes it more credible; and it might be perfectly proper to prevent the official from saying something that in itself is public knowledge: that the reporter wrote that p.
(All this is an abstract discussion; in the case in point, I agree with Josh Marshall: It's difficult to see why this needs to be kept secret at all. Just thought that there were semi-interesting points about epistemology of testimony and speech acts to be made here--and it's been a while since I posted anything substantive!)