There is no llama cartel. There is, however, an entire industry built on the phrase “sources say” — and it works because that sentence costs almost nothing to write and a great deal to check. That asymmetry is the machine this page is parodying, one headline at a time.
Anonymity in reporting is not automatically a problem. Newsroom standards treat it as a last resort with conditions attached. The Associated Press tells its journalists that material from anonymous sources should be used only when it is information, not opinion, when it is essential to the report, and when it is not available any other way. Reuters requires reporters to explain to readers why a source cannot be named. The Society of Professional Journalists’ code puts it plainly: identify sources clearly, and reserve anonymity for sources who may face danger or reprisal.
Those are real constraints, and outlets that follow them tend to publish fewer, better-sourced anonymous claims. The trouble starts downstream, where nobody applied the constraints in the first place.
A single unnamed claim can move through a dozen aggregators in a few hours, picking up certainty at every hop. The first outlet writes that a source claims something. The fourth writes that it has been reported. The tenth writes that it is known. Nobody has technically lied. The hedge simply evaporated in transit, because a hedge is expensive to carry and cheap to drop.
Ask who benefits from you believing it before you ask whether it’s true. The second question is hard. The first one is usually enough.
The incentive runs one direction. A story sourced to “a person familiar with the matter” can be produced in an afternoon by someone who never spoke to that person. Verifying it costs days and often ends in nothing publishable. The market rewards the first version and forgets the correction, which is why the correction almost always travels shorter and slower than the claim it fixes.
Four questions collapse most of it, and none of them require expertise.
Who is the source, in category terms? Not their name — their position. A participant, a bystander, a rival and a spokesperson have very different reasons to talk, and a report that will not tell you which one it has is telling you something.
What would they lose by being named? Anonymity granted to someone facing retaliation is doing real work. Anonymity granted to someone floating a trial balloon is doing work too, just not for you.
Is there a document, or only a description of a document? “Sources described the memo” is a much weaker claim than “the memo says,” and the gap between those two sentences is where most of the interesting failures live.
Has any outlet with a published corrections policy touched the story since? An organization that prints its own mistakes has skin in the game. One that never corrects anything is not more accurate; it is less accountable.
That is the whole method. It takes about ninety seconds and a willingness to be slightly less entertained.
Which is the real reason the llama headline sits at the top of this page. It is a measurement of how much more attention a lie can buy than its correction. You are the control group.