I think this could be a workable model for regulating the news media: you can publish whatever you like, as long as you disclose your entire fact-gathering process (i.e. "cite your sources") in a way where another journalist could do the same research and reach the same conclusions about what happened (i.e. "replicate your study".)
Whelp, there goes whistleblowing. Or generally any source that prefers not to be named. I think your proposal destroys some of the most important journalism.
Also, conclusions in journalism are not mathematical equations; they don't derive linearly from the facts. Who gets to decide whether that particular conclusion is or not supported by the facts?
Keep in mind, I’m not suggesting that the source would need to be named; I’m only suggesting that the data from the source would need to be able to be independently recollected from the same source.
In the example of a whistleblower, that would mean that the journalist would need to ensure that there was some secure anonymous route by which other journalists could get in contact with the same whistleblower to independently verify their story.
You wouldn't have done your job, as a journalist, until you had provided a reproducible way for someone else to come along and re-do your investigation without you, ending up with all the same facts, if not the same beliefs.
> Who gets to decide whether that particular conclusion is or not supported by the facts?
My proposal would be to separate journalism from editorial or opinionated conclusions. "The news" itself would only contain objectively-verifiable facts, without opinions; and then it would be the job of magazines, talk-shows, etc. to “interpret” the news, in much the same way the current news media itself currently interprets things like scientific papers.
In other words, “news” intended for the public would be trimmed down to the most boring and dry possible version of itself—like a cross between a wire service and a police scanner. Anything that wasn’t such, couldn’t legally be called “news”, nor could it suggest by its formatting that it is news. There would be a separation between "news" and "opinion" in about the same way there is an FDA-managed separation between "food" and "drugs"; and there would be public-awareness campaigns to make it clear that this division exists, and that any fact found in "opinion" sources that you haven't independently heard from a "news" source is suspect at best and more likely an active psy-op.
But, again, the state wouldn't be responsible for declaring who can or cannot publish (just like the FDA isn't responsible for saying who can or cannot manufacture drugs; and the SEC isn't responsible for saying who can or cannot sell equity.) The only job of this hypothetical MiniTruth would be to say—according to a very objective, independently calculable bar, determined by Congress—whether what a given party is publishing can legally be called "news" or not; and to punish parties that do call their publications "news" if they have not met the requirements to do so.
(Of course, there's no such thing as an unbiased source. The curated set of facts that end up put together into a news article is itself an editorial choice. But we already have a news media that generates every possible combination of such facts, so that bias, at least, sort of "evens out.")
Basically, your sources would need to be “made available” in the same way witnesses are made available during a trial: there to be interrogated by interested parties who want to dispute their story, even if also safe from harm (by e.g. witness protection.)
Which is quite unrealistic unless you're talking about blowing a story about the local bakery or something like that. If any journalist can access the witness, they have no real security. Journalists are not all incorruptible angels, and a whistleblower is already in danger by trusting just one or two.
Witness protection works by essentially making the witness and their immediately family cut off everyone they know and restart their lives with new identities. It's a tremendous cost to pay, and the only reason most do it is because they're also implicated, and want to cut a deal. Who in their right mind would completely ruin their good, law-abiding life just to report some corrupt politician?
You're being pretty uncharitable in assuming that there would need to be exactly one protection mechanism and that every source would necessarily have that same protection applied to it.
There are a spectrum of solutions to creating anonymity. If your source is just a faceless worker who happened to overhear something about their boss, then all they need from you is a burner phone, an email address on a public server, and GPG keys—all of which you can give out the public halves of.
If your source is Edward Snowden, and his information is about the state, then I would suggest that there's really no way to do things other than full-on "witness protection" in the sense Snowden achieved: the state is going to figure out who leaked the information, so even without giving other journalists access to your source, anonymity for them (where they currently live) is no longer possible. You may as well fly them to Russia; at which point they're safe enough that you can just give them the a public email address, tell them to access it from an Internet cafe over Tor, and now they've once again got a contact point that anyone else could interact with them through without "blowing their cover." So why not give it out?
If your source is just a faceless worker who happened to overhear something about their boss, then all they need from you is a burner phone, an email address on a public server, and GPG keys—all of which you can give out the public halves of.
But then how does that prove anything? I can give you an email address and phone number of someone who will tell you anything I want.
Whelp, there goes whistleblowing. Or generally any source that prefers not to be named. I think your proposal destroys some of the most important journalism.
Also, conclusions in journalism are not mathematical equations; they don't derive linearly from the facts. Who gets to decide whether that particular conclusion is or not supported by the facts?