In order to have a norm like “debate or update” or something along those lines, you would have a rule like:
If you’re going to publicly claim X, then you have to defend it in public debate; this is especially so if X is important, if people are deferring to you about X, if actions are being taken based on deferential beliefs in X from your belief, if you are prestigious or powerful, etc.
(There are lots of exceptions, such as exfohazards.)
To have a rule like that, we’d have to balance this norm against people being busy. But people use that (and other such excuses) to spinelessly avoid public debate. On the flip side, people invoke the norm selectively as part of other bad behavior, such as clout chasing and spinelessly avoiding debates with lower status but scary debate opponents. How to balance these things?
For each person P who has some stated belief X subject to this norm, there’s the question of who and how often they have to debate to pass the norm. How to decide this?
A first draft is that you could essentially have an election, where anyone who believes not-X gets a vote. (Almost everyone won’t vote in any given election, but that is fine.)
This has various problems I assume. For example, how do you tell who believes not-X? Or if you allow anyone to vote, couldn’t the people who want to cheer for X vote for a poor debate opponent or an opponent who actually really agrees with X but just wants to play Devil’s Advocate or wants to just remind us of certain virtues of certain aspects of considering not-X sometimes?
A more reasonable, practicable sort of protocol is tracking properties like “sane” and “good-faith” and “knowledgeable” and “interesting”, which add up to a Fukuyama-like “Factor X” (no relation) of being a “worthy opponent” for debating about X, and indeed a worthy peer in some community of those who participate in this norm. But this leads to much huffiness. More precisely, this leads to lots of plausible deniability about who improperly declined a debate, or who doesn’t have the right to demand a debate on pain of the ”...or update”.
I’d worry about isolated demands for rigor, and I’d expect that unpopular people/ideas got challenged to a debate much more often than popular ones. This is a tax on contrarianism.
People should be free to state whatever they want. And you and I are free to ignore them if they don’t justify it. They don’t have to update on my disagreement any more than I have to update on their broken model (I would like to update if they have new evidence, though).
Um ok, but like, you know how people already do the thing I’m trying to gesture at, for good reasons, but really poorly, with all the costs and not that much of the benefit? What if we did it a little bit better, with some more of the benefit and not more costs?
Externally forced belief is the more menacing version of filtering what is permitted to be said in public. Debate as the escape valve doesn’t fix this, friction already shapes intent of the would-be debaters to initially make claims in public.
I don’t think it makes much sense to analyze “friction” as a blackbox category, some friction is just the wrong kind of friction. Debating in public shouldn’t be taken just like arbitrary friction. E.g. you don’t have to update, as long as you debate! Others can form their own opinions based on how the debate went (and the debate moderators can better learn how to elicit constructive dialogue etc.).
But yes, I think some friction is absolutely worth it! If Bob goes around being like “Yyyeah, ohhh yyeahhhh, X is so fucking true babyyy” and a lot of people are allocating resources on the back of that confidence, he should debate!
If Bob goes … and a lot of people are allocating resources on the back of that confidence, he should debate!
Or those people should reconsider listening to Bob, maybe there should be an Alice that informs them about issues with Bob’s claims. In your proposal, the menacing possibility of an external (public) belief-rewriting is still there, and even the lesser inconvenience of debating is distortionary. Some claims won’t be made as much, depending of what the proponents expect to happen as a consequence.
That’s the sense in which friction has a general use, the filtering effects. What Alices say is also friction that affects what Bobs will say based on the expected consequences, but at least it doesn’t make it directly more difficult for Bobs to say things. Only reputation and the effectiveness of the claims is on the line, depending on what the claims are.
What Alices say is also friction that affects what Bobs will say based on the expected consequences, but at least it doesn’t make it directly more difficult for Bobs to say things. Only reputation and the effectiveness of the claims is on the line, depending on what the claims are.
I don’t understand, I’m proposing a particular variety of this—a variety of “Bob might lose some reputation and effectiveness if he doesn’t debate.”. What’s on the line with the debate norm that isn’t on the line with Alices going around saying stuff? (TBC, I don’t endorse “debate or update”; I might come to endorse something in the genre of my proposed norm.)
Then it’s “debate or don’t”, not “debate or update”, which is completely different. You didn’t operationalize the penalty for refusing to debate even when you “have to” under your norm, so I was defaulting to “update” from the maxim you opened with. Both the menacing penalty of “update” and the baseline inconvenience of “debate” can be worse than taking the reputational damage (provided the Alices are allowed to speak, which isn’t truly a given), depending on the claim.
Ok. (From my POV that’s a rather uninteresting way to engage with what I wrote, as it is a butterfly idea and the stag you’re hunting seems to be different from the stag I’m hunting here.)
In order to have a norm like “debate or update” or something along those lines, you would have a rule like:
(There are lots of exceptions, such as exfohazards.)
To have a rule like that, we’d have to balance this norm against people being busy. But people use that (and other such excuses) to spinelessly avoid public debate. On the flip side, people invoke the norm selectively as part of other bad behavior, such as clout chasing and spinelessly avoiding debates with lower status but scary debate opponents. How to balance these things?
For each person P who has some stated belief X subject to this norm, there’s the question of who and how often they have to debate to pass the norm. How to decide this?
A first draft is that you could essentially have an election, where anyone who believes not-X gets a vote. (Almost everyone won’t vote in any given election, but that is fine.)
This has various problems I assume. For example, how do you tell who believes not-X? Or if you allow anyone to vote, couldn’t the people who want to cheer for X vote for a poor debate opponent or an opponent who actually really agrees with X but just wants to play Devil’s Advocate or wants to just remind us of certain virtues of certain aspects of considering not-X sometimes?
A more reasonable, practicable sort of protocol is tracking properties like “sane” and “good-faith” and “knowledgeable” and “interesting”, which add up to a Fukuyama-like “Factor X” (no relation) of being a “worthy opponent” for debating about X, and indeed a worthy peer in some community of those who participate in this norm. But this leads to much huffiness. More precisely, this leads to lots of plausible deniability about who improperly declined a debate, or who doesn’t have the right to demand a debate on pain of the ”...or update”.
What is to be done?
I’d worry about isolated demands for rigor, and I’d expect that unpopular people/ideas got challenged to a debate much more often than popular ones. This is a tax on contrarianism.
People should be free to state whatever they want. And you and I are free to ignore them if they don’t justify it. They don’t have to update on my disagreement any more than I have to update on their broken model (I would like to update if they have new evidence, though).
Um ok, but like, you know how people already do the thing I’m trying to gesture at, for good reasons, but really poorly, with all the costs and not that much of the benefit? What if we did it a little bit better, with some more of the benefit and not more costs?
Externally forced belief is the more menacing version of filtering what is permitted to be said in public. Debate as the escape valve doesn’t fix this, friction already shapes intent of the would-be debaters to initially make claims in public.
I don’t think it makes much sense to analyze “friction” as a blackbox category, some friction is just the wrong kind of friction. Debating in public shouldn’t be taken just like arbitrary friction. E.g. you don’t have to update, as long as you debate! Others can form their own opinions based on how the debate went (and the debate moderators can better learn how to elicit constructive dialogue etc.).
But yes, I think some friction is absolutely worth it! If Bob goes around being like “Yyyeah, ohhh yyeahhhh, X is so fucking true babyyy” and a lot of people are allocating resources on the back of that confidence, he should debate!
Or those people should reconsider listening to Bob, maybe there should be an Alice that informs them about issues with Bob’s claims. In your proposal, the menacing possibility of an external (public) belief-rewriting is still there, and even the lesser inconvenience of debating is distortionary. Some claims won’t be made as much, depending of what the proponents expect to happen as a consequence.
That’s the sense in which friction has a general use, the filtering effects. What Alices say is also friction that affects what Bobs will say based on the expected consequences, but at least it doesn’t make it directly more difficult for Bobs to say things. Only reputation and the effectiveness of the claims is on the line, depending on what the claims are.
I don’t understand, I’m proposing a particular variety of this—a variety of “Bob might lose some reputation and effectiveness if he doesn’t debate.”. What’s on the line with the debate norm that isn’t on the line with Alices going around saying stuff? (TBC, I don’t endorse “debate or update”; I might come to endorse something in the genre of my proposed norm.)
Then it’s “debate or don’t”, not “debate or update”, which is completely different. You didn’t operationalize the penalty for refusing to debate even when you “have to” under your norm, so I was defaulting to “update” from the maxim you opened with. Both the menacing penalty of “update” and the baseline inconvenience of “debate” can be worse than taking the reputational damage (provided the Alices are allowed to speak, which isn’t truly a given), depending on the claim.
Ok. (From my POV that’s a rather uninteresting way to engage with what I wrote, as it is a butterfly idea and the stag you’re hunting seems to be different from the stag I’m hunting here.)