The confusion

I recently came upon a comment which I believe reflects a persistent confusion among rationalist/EA types. I was reading this post which contains ideas that the other has but doesn't have time to write posts about. One of those relates to the concept of "good faith", labelled "most arguments are not in good faith, of course":

Look, I love good faith discourse (here meaning "conversation in which the primary goal of all participants is to help other participants and onlookers arrive at true beliefs").

The definition given for "good faith discourse" seems incorrect to me, and it's not a close call in my opinion. The level of incorrectness in my view is something like saying "I like people who obey the law (here meaning never committing a social faux pas)". This isn't the first time I have seen someone in this community advance a similar view on the meaning of good/bad faith. For example, this post.

I thought it might be useful to bring this apparent disagreement to the foreground, so I will lay out my belief about what this concept means. I suspect this disagreement may also involve an underlying about what good norms are/should be related topics, but here I will attempt to put the prescriptive aspects aside and focus simply on the question of what the phrase means in "general" usage (i.e. beyond the rationalist/EA community).

My definition

"Good faith" has a general usage and also a more specific one as in "good faith discourse", both of which are highly related and consistent with each other. I go over my understanding of both here:

General meaning

I propose that "good faith" in its general meaning simply means acting genuinely or sincerely, acting without deceit.

Good faith discourse

As it relates to "good faith discourse" or "arguing in good faith", we can refine this to be more specific. In this context the phrase means genuinely representing one's own beliefs or intentions in the discussion. The opposite "bad faith" means to intentionally misrepresent one's beliefs or intentions. Relevantly, this means that "bad faith" must be intentional.

Evidence

I think two lines of evidence suggest my understanding of this phrase is correct. Common sources like dictionaries and Wikipedia, and prominent concepts that utilize the phrases or closely related phrases.

Sources

The second post that I link above quotes from the Wikipedia article on "bad faith":

What does "bad faith" mean, though? It doesn't mean "with ill intent." Following Wikipedia, bad faith is "a sustained form of deception which consists of entertaining or pretending to entertain one set of feelings while acting as if influenced by another." The great encyclopedia goes on to provide examples: the solider who waves a flag of surrender but then fires when the enemy comes out of their trenches, the attorney who prosecutes a case she knows to be false, the representative of a company facing a labor dispute who comes to the negotiating table with no intent of compromising.

What does that post conclude from this about the meaning of "bad faith"?

That is, bad faith is when someone's apparent reasons for doing something aren't the same as the real reasons. This is distinct from malign intent. The uniformed solider who shoots you without pretending to surrender is acting in good faith, because what you see is what you get: the man whose clothes indicate that his job is to try to kill you is, in fact, trying to kill you.

Now, we need to be precise with how we parse these terms. It is indeed true that bad faith is not the same as general bad intent. However, bad faith requires a specific bad intent. I think this is extremely clear from the examples listed in the Wikipedia article. All clearly involve intentional misrepresentation, not some type of accidental misrepresentation due to cognitive bias or some other mechanism. I think this is pretty clear if you read that section of the Wikipedia article:

Some examples of bad faith include: soldiers waving a white flag and then firing when their enemy approaches to take prisoners (cf. perfidy); a company representative who negotiates with union workers while having no intent of compromising; a prosecutor who argues a legal position that he knows to be false; and an insurer who uses language and reasoning which are deliberately misleading in order to deny a claim.

The later three all use pretty unequivocal phrases indicating the relevance of intent ("no intent of compromise", "knows to be false", "deliberately misleading"), and the first example involves an action that unambiguously indicates a specific intention (raising a white flag). It seems pretty clear to me that bad faith requires intentional misrepresentation.

Merriam-Webster has good faith as follows:

honesty or lawfulness of purpose

with the example usage:

bargained in good faith

the legal definition as:

honesty, fairness, and lawfulness of purpose : absence of any intent to defraud, act maliciously, or take unfair advantage

with example usages:

filed the suit in good faith negotiating in good faith

"Bad faith" from Merriam-Webster [is](lack of honesty in dealing with other people):

lack of honesty in dealing with other people

with the example usage:

She accused her landlord of bad faith because he had promised to paint the apartment but never did it.

Legal specific:

intentional deception, dishonesty, or failure to meet an obligation or duty no evidence of bad faith

American Heritage "good faith":

The sincere intention to be honest and law-abiding, as when negotiating a contract: bargained in good faith.

And "bad faith":

The malicious intention to be dishonest or to violate the law, as in negotiations over a contract.

The only one of these that seems debatable to me is the "she accused her landlord of bad faith..." one. That doesn't explicitly involve intent to deceive and one could infer that therefore intent isn't required. But you don't need to know for a fact someone did a thing to accuse them of it, and I definitely agree inconsistency alone can be a reason to suspect bad faith even if the intent is uncertain, so overall I think the balance of the evidence here is that "bad faith" does require intentional, rather than accidental or subconcious, misrepresentation.

Classic usage

I think some classic applications of the good faith concept also strongly suggest this interpretation:

Good faith errors or mistakes

A common application of the idea of good faith is that someone can still be operating in good faith even when they are mistaken. Although it is possible for this to happen even if bad faith doesn't require intent (the mistaken person straight-up lacks information), I think part of the reasoning behind the alternative meaning of "good faith" is that lot of "good faith mistakes" aren't actually in good faith. Take this for example:

I claim that this supposedly ordinary state of affairs is deeply weird at best, and probably just fake. Actual "good faith" disagreements—those where both parties are just trying to get the right answer and there are no other hidden motives, no "something else" going on—tend not to persist.

This is a common thread I see between the posts that I have linked to expressing a contrary view of what constitutes "good faith". People think that you should "assume good faith" and think good faith discourse is common, but actually people engage in dirty conversational moves and rhetorical game playing all the time! So the argument goes. But in my view, this gets the logic backwards. Whether an agreement is "actual good faith" depends on the definition of good faith, not on whether the disagreement is epistemically or morally good. A person's conduct can be bad without being "bad faith", as the solider example from this post above makes clear. The fact that the definition of good faith which does not require intent substantially narrows the category of "actual good faith mistakes" and in fact that this seems to be a major purpose of adopting this definition, suggests to me that there is a genuine change in meaning going on here. My interpretation is that the more common/standard understanding of bad faith does require intent and doesn't include mistakes that come about through things like cognitive biases, and that the argument being put forward in this post is effectively arguing that people should adopt a new definition/standard.

Good faith negotiation

The idea of "good faith" is commonly applied in the context of negotiations or bargains. I think this effectively puts the "good faith is when you are trying to help everyone find the truth" interpretation in its grave. In the context of a negotiation, the idea that the participants would primarily be focused on discovering truth doesn't make much sense. One would naturally assume that they have their own goals in the negotiation (e.g. getting a good deal for their business) that makes up their primary purpose. Instead of speaking to a persons goals or primary intent, "good faith" is actually a constraint on this intent. A person may be pursuing their primary goal, but not the point of bad faith.

I hypothesize that this may explain why this confusion might come up among rationalists/EAs. Many in these communities I think would describe themselves as concerned with "truth-seeking", and believe that norms and epistemics should prioritize encouraging this. In contrast, what I view as the common meaning of "good faith" focuses on truth as a constraint rather than as a goal. There is nothing wrong with either perspective in my view (although I have my own opinions on how they can be productively discussed and leveraged), but I think it is worth having clarity about terms. One can agree that "good faith" as I have defined it isn't sufficient as an epistemic norm or isn't that high of a standard, while still agreeing that it is how the term is commonly used.

Conclusion

"Bad faith" requires intentional misrepresentation, as the phrase is typically used outside rationalist/EA circles.