Dialogue form might actually work - Polya as a model might be good
- Honesty is telling the truth
- No, because you can lie through omission.
- Ok, it’s telling the whole truth
- No, because you can deceive without saying anything, just showing. (e.g. packing to go away on holidays to trick a spy)
- Ok, it’s about not causing other people to have false beliefs
- No, bc. you could be deluded
- Ok, it’s about not knowingly causing another person to have false beliefs
- No, because of it’s not dishonest to not-lie to a racist.
- Ok, what about intentionally causing them to have false beliefs?
- No, because of poker, and deceptive punchlines of jokes.
- Other points
- Speaking in systematically-false environments? How could rules about this fall to the Unilateralists curse
- Positive duties to either truth telling (deontic), or correcting false beliefs (consequentialist), which might require we lie, or come into conflict with privacy.
- Does honesty forbid bald-faced lies?
- How does the officialness of common grounds connect to the sub-ness of sub-agents.
- How does AI deception matter, e.g. the robot arm placing itself between the ball and camera? There’s no intention there. (though future systems could have intentions, even if they’re not phenomenally conscious) This would line up with functionalism about types of conscious state (Two types whose members always play the same functions are the same type) but not tokens of conscious states (two tokens may play the same function role without sharing all defining properties or something (this is kinda bs tbh))
Scope and degree stuff seems the most interesting. This connects the stuff about norms and delineating the common ground/bracketing agency in poker and stuff.
Not much disagreement in this one, more just Agnes presenting a puzzle, which is mostly a puzzle from an individualist normative lense, and Robin viewing it in a collective way, allowing for specialisation and pure description, and not seeming very puzzled.
Paradox is just that honesty seems like a terminal norm of one kind of action, communication, but all action only has one terminal norm, doing the good. Callard thinks this is perplexing, we’re pursuing the good right, so how can there be distinctive goods for more or less broad descriptions of action, that cna come apart? Maybe just like beauty being the goal of artmaking or something. I’m not sure how this would fit in with preference framings of valuing honesty, which would be quite sensible, and norm-approaches which would seem much harder to do, Paradoxes of deontology, virtue, and belief. Sure there can be a trade-off with preferences, but what’s the trade-off with norms and their enforcement?
- could there be a solution with the weird mixed goals of honesty (the extra-terminal goals problem) and the fact we kind of have different sub-agents, with their own terminal goals, who have some degree of existence/officialness corresponding to the official-ness of the common-ground to which the propose additions.
Hanson just views everything in terms of norms and coordination and enforcement and stuff. He’s more collective and descriptive, Callard is more individualist and normative.
Can you flag you’re being dishonest? Hanson gives the example of actors doing it but they’re really not.
How do you relax norms? e.g. we’re not in court, so we don;t have the 100% honesty norm, we’re at 95%, but what does that mean? Just weaker sanctions and investing less in detecting norm violation? What about intimate relationships wherein dishonesty is seen as a massive betrayal.
Bartering is an example of pretense. Dishonesty in these cases is totally fine, but only a certain degree of it, and a certain scope. Poker any degree is fine, but only a certain scope. Actors on stage have any degree and they can take almost any scope, aside from yelling fire and stuff. What might be certain scope/degree cases where we have the positive duty to tell the truth/inform? Mandatory reporting is kind of like this, but not quite.
Robin: So among intellectuals if you sort of date the first draft of your paper August and it was really September, we might not get very upset about that. But as other important parts of the paper—maybe the data set that was all made up—then that would be a much bigger deal. One communicative act can have different stakes.
When you want a relative ranking, you can correct for systemic exageration easily? Or something? It forces you to use context clues to adjust for the exageration, especially if people exagerate to different degrees. This is costly for those who can afford it, because it’s effortful and uncertain, but it’s even worse for people who fail to adjust in context, because they get the consequences of being wrong.