[TeMPOraL]

Maybe we need to start teaching people to habitually attach provenance and confidence information to every fact (really: belief) they report, at least as much as practically possible?

"Iraq had WMDs in 2000s" != "I strongly believe Iraq had WMDs in 2000s" != "I find it plausible that Iraq might have had WMDs in 2000s" != "According to that UN report, Iraq had WMDs in 2000s" != "According to NYT, which quotes that UN report, Iraq had WMDs", etc.

A lot of problems are caused by people who say "X" when they should say "I strongly believe X", or "I think I read somewhere that X", or "I'm not sure, but I think X".

[SkyBelow]

I generally speak like this and find it causes problems with others because they view this as a weak form of speech. It makes you appear uncertain, which should be a good thing, but our society seems to value undeserved confidence far above cautious uncertainty.

[dotancohen]

This is my experience 100%. Blatant lies, told confidently, are far more convincing than the truth with a source.

[__MatrixMan__]

I find it helpful to spend a bit more cycles on keeping track of audience and a bit fewer on perfect conveyance of certainty levels.

Once you've set the bar re: certainty for a particular discourse neighborhood, you can just make claims at that level thereafter without feeling dishonest. As long as you're sensitive to cases where others may have joined without enough context to know that you're out on a limb.

[dhimes]

Yes- intellectual dishonesty is at the root of this. When I was raised, the integrity of a "man's" word was a measure of the "goodness" of the person. Now, we hear things like "he said what he had to say" and a basic acceptance of lies if the person is on our side of the debate.

It's depressing.

[mkr-hn]

I've tried to get into this habit. Qualifications are great too because when someone comes back to say "you said X! You lied!" you can remind them of how carefully you qualified your statement.

This is Circle of Competence applied[0]. It's admitting what you know you don't know so you don't make mistakes by being overconfident.

[0] https://en.wikipedia.org/wiki/Circle_of_competence

[TeMPOraL]

It's even more than that. By qualifying your statements, you let others evaluate them correctly.

So, for instance, if 'tptacek here says a factual statement about security (and there's no large thread contesting it), I'll treat it as gospel. But if he says he's unsure about it, or he read it somewhere, I know to attach less weight to it. I'll code in it my brain as "uncertain, but passed the sniff test of a relevant expert". Etc.

The same principle works in more mundane aspects of life. Whether a person believes something (and how much), or whether they're just reporting something they've read elsewhere, matters a lot for evaluating a factual statement independently.

[RussianCow]

The one thing with this is that you have to be consciously careful of who you view as an authority on different subjects, or whose opinion on something is relevant. For a simple, everyday example, it's easy to follow a friend's recommendation to eat at a particular restaurant, even if that friend has completely different tastes in food than you do (which, in all likelihood, will result in a poor dining experience for you). Likewise, I think even smart, educated people often make the mistake of treating the word of powerful individuals as fact even if those individuals have absolutely no experience with or authority on the matter.

[guelo]

> "NYT, which quotes that UN report, Iraq had WMDs"

You left out the last step,

!= Bush administration says Iraq had WMDs

[nradov]

That is a major part of the formal training program for professional intelligence analysts. Of course they still often get the probabilities wrong.

[jlg23]

Let's just talk in Quechua then, evidentiality is engraved in their grammar!(https://en.m.wikipedia.org/wiki/Quechuan_languages)