In the example provided, you get arrested by a non-corrupt cop for attempted bribery. In a dating scenario, you get rejected. This seems obvious, maybe you mean something else?
Hmm, my understanding of conflict theory is that the proponents consider life a zero-sum game where you divide a fixed-size utility pie, in the first approximation. An alternative (whether to call it mistake theory or not, I am not sure), is that cooperation increases the size of the pie. Assuming this understanding is correct, do you think the non-corrupt cop’s logic is accurately modeled as “it’s either me or them?”
My understand of conflict theory is that proponents consider the situation in question to be a conflict where there are different people who have opposing interests, whereas for mistake theory, proponents consider the situation in question to be a cooperation where everyone agrees that there is a problem in a shared goal and there’s just uncertainty about the appropriate solution.
In the usual case, “the situation in question” would be political debate, but in the corrupt-cop case would be an interaction where society has tried to ensure that there are non-corrupt cops who arrest people who do bad stuff and people who try to offer bribes to corrupt cops after having done bad stuff.
To be clear, the following would be my understanding of a mistake theoretic analysis of the cop issue:
It is unclear how quickly people should drive cars on the road. There are guidelines called “speed limits” which put an upper bound on it, but there are of course two sides to any issue. For instance, there are honest people who feel like since they kind of gotta pee, it’s worth driving extra fast so they don’t pee their pants. There are also other honest people who drive fast so that they can feel the excitement of the speed.
These people disagree with the speed limits, and they feel like that is mainly because the people who made the speed limits didn’t realize their needs, or because they happen to have some incorrect cognitive heuristics or similar.
In order to make the speed limits work, we have cops who try to identify who is speeding, so they can inform them about the need to avoid speeding. To better communicate the importance, they may fine the drivers so they can see how big of a deal it is. If the drivers are not smart enough to understand the cops’ concern about speed, they may have to take away their driver’s license.
Of course, the presence of the cops is not great for speeders who feel that the speed limits are set wrong. Fortunately from their perspective, there are some sympathetic cops who are willing to suspend the rules for a bit of extra pay.
If there are a lot of speeders, maybe that indicates that the speed limits are set wrong, and that we need to have more open debate so we can figure out what the appropriate speed limits would be.
Someone who is being mistake theorist about this might be unwilling to acknowledge that the indirect language serves the purpose to avoid arrest, or they might begrudgingly admit it but frame it as a way to minimize conflict or be polite or avoid biases (similar to how mistake theorists about politics will be unwilling to acknowledge wolf whistles, will prefer to avoid claims that someone is lying, etc.).
Meanwhile a conflict theorist would analyze it as a conflict between the driver who wants to drive fast vs other people who want safe roads, and analyze the cops as having the job of enforcing the safe roads side over the drive fast side of the conflict, and analyze corrupt cops as people who betray their side.
This is ingenious but feels like rather a stretch, compared with the simpler analysis that says that the cop-bribing situation just isn’t really the sort of thing the mistake/conflict dichotomy is meant for. The job of a police officer is to enforce the law, pretty much everyone agrees with this (including e.g. people who think the laws are bad and shouldn’t be enforced, and people who think cops are often bad and not doing their jobs right), and if you want to break the law then you and the police are in conflict, end of story. Favouring “mistake theory” in political arguments doesn’t require anyone to dream up ingenious ways to reframe your relationship with the law and its enforcers that allow them to see that in conflict-free terms.
Favouring “mistake theory” in political arguments doesn’t require anyone to dream up ingenious ways to reframe your relationship with the law and its enforcers that allow them to see that in conflict-free terms.
I don’t primarily think of conflict theory and mistake theory as properties of people. Rather, I primarily think of conflict and mistake as properties of situations, and then one might have different theories about what best describes those situations.
So for example, it is possible for someone to have a conflict theory about the relationship between criminals and society (as you correctly point out most people have, though I would say it seems to me conservatives have this theory more than progressives do, hence “tough on crime”) while having a mistake theory about how politics works.
I think it’s worth distinguishing the general phenomenon where people can have opposed interests to varying degrees (which happens everywhere) from the more specific question of what attitude a person should have or does have towards the people they’re arguing with.
Your interests and those of a police officer who’s just pulled you over for speeding (or, staying closer to Zack’s example, the interests of more-bribeable and less-bribeable police officers in such a situation) may be more or less in conflict with one another, but I don’t see how it helps anything to try to fit that into the same framework as we use for assessing different attitudes to political debate.
There’s definitely lots of distinctions that can be made. The aspect of conflict that I find most important to know about is the “epistemic” part of it. Basically, does Aumann’s agreement theorem apply?
In ordinary non-conflict conversations such as asking someone you are visiting where you can get a cup of water, you can simply copy other’s beliefs and do reasonably well, whereas in conflict conversations such as politics, copying your opponent’s beliefs about factual matters is a serious security vulnerability.
(It doesn’t seem uncommon for mistake theorists to come up with explanations of why Aumann’s agreement theorem wouldn’t apply to politics despite both sides being honest, but nobody has come up with compelling explanations, whereas the conflict theory analysis of it seems compelling and commonly for people to self-endorse.)
My understand of conflict theory is that proponents consider the situation in question to be a conflict where there are different people who have opposing interests, whereas for mistake theory, proponents consider the situation in question to be a cooperation where everyone agrees that there is a problem in a shared goal and there’s just uncertainty about the appropriate solution.
Right, I think it is another way to distinguish between believing in zero-sum and believing in positive sum games, no?
I suspect that this framework is not well suited for analyzing the traffic stop issue. At least I do not think the reasoning you describe is sufficiently common… A non-corrupt cop just follows the rules without thinking about the size of the pie, while a corrupt cop sees their position as an opportunity to grab some extra piece from someone else’s pie… which can be interpreted from either perspective, without providing any useful insight. Basically, this is not a nail, so using a hammer is not a great approach.
Right, I think it is another way to distinguish between believing in zero-sum and believing in positive sum games, no?
No. While zero-sum games is one way that people can be in conflict, they are not the only way:
Zero-sum games are just one example of a broader class of (game-theoretically equivalent) Pareto-frontier games, where one person’s benefit is another person’s loss, but where in the general case one person may benefit more than the other loses. (For instance, some might analyze the traffic case as a Pareto-frontier game, where the loss due to crashes is worse than whatever gain the driver might get from driving fast.)
Even if there globally is opportunity for Pareto improvements, locally one may be in a setting where there is no system to cooperate to achieve these improvements, and there is insufficient trust to negotiate to achieve them. In such a case there may be conflict and local opposing interests, even if there would be gain for deescalating the conflict and finding ways to help everyone.
In order for it to make sense to analyze others as being basically honest and merely making a mistake when they disagree, it would have to be that people aren’t intentionally trying to work against each other’s interests.
The police officer example is about safe escalation of shared knowledge, not mistake theory or conflict theory: https://scottaaronson.blog/?p=2410
Safe from what?
Safe from negative consequences of revealing your intentions unilaterally.
Where do those negative consequences come from?
In the example provided, you get arrested by a non-corrupt cop for attempted bribery. In a dating scenario, you get rejected. This seems obvious, maybe you mean something else?
How is getting arrested by a non-corrupt cop for attempted bribery not about conflict?
Hmm, my understanding of conflict theory is that the proponents consider life a zero-sum game where you divide a fixed-size utility pie, in the first approximation. An alternative (whether to call it mistake theory or not, I am not sure), is that cooperation increases the size of the pie. Assuming this understanding is correct, do you think the non-corrupt cop’s logic is accurately modeled as “it’s either me or them?”
My understand of conflict theory is that proponents consider the situation in question to be a conflict where there are different people who have opposing interests, whereas for mistake theory, proponents consider the situation in question to be a cooperation where everyone agrees that there is a problem in a shared goal and there’s just uncertainty about the appropriate solution.
In the usual case, “the situation in question” would be political debate, but in the corrupt-cop case would be an interaction where society has tried to ensure that there are non-corrupt cops who arrest people who do bad stuff and people who try to offer bribes to corrupt cops after having done bad stuff.
To be clear, the following would be my understanding of a mistake theoretic analysis of the cop issue:
Someone who is being mistake theorist about this might be unwilling to acknowledge that the indirect language serves the purpose to avoid arrest, or they might begrudgingly admit it but frame it as a way to minimize conflict or be polite or avoid biases (similar to how mistake theorists about politics will be unwilling to acknowledge wolf whistles, will prefer to avoid claims that someone is lying, etc.).
Meanwhile a conflict theorist would analyze it as a conflict between the driver who wants to drive fast vs other people who want safe roads, and analyze the cops as having the job of enforcing the safe roads side over the drive fast side of the conflict, and analyze corrupt cops as people who betray their side.
This is ingenious but feels like rather a stretch, compared with the simpler analysis that says that the cop-bribing situation just isn’t really the sort of thing the mistake/conflict dichotomy is meant for. The job of a police officer is to enforce the law, pretty much everyone agrees with this (including e.g. people who think the laws are bad and shouldn’t be enforced, and people who think cops are often bad and not doing their jobs right), and if you want to break the law then you and the police are in conflict, end of story. Favouring “mistake theory” in political arguments doesn’t require anyone to dream up ingenious ways to reframe your relationship with the law and its enforcers that allow them to see that in conflict-free terms.
I don’t primarily think of conflict theory and mistake theory as properties of people. Rather, I primarily think of conflict and mistake as properties of situations, and then one might have different theories about what best describes those situations.
So for example, it is possible for someone to have a conflict theory about the relationship between criminals and society (as you correctly point out most people have, though I would say it seems to me conservatives have this theory more than progressives do, hence “tough on crime”) while having a mistake theory about how politics works.
I think it’s worth distinguishing the general phenomenon where people can have opposed interests to varying degrees (which happens everywhere) from the more specific question of what attitude a person should have or does have towards the people they’re arguing with.
Your interests and those of a police officer who’s just pulled you over for speeding (or, staying closer to Zack’s example, the interests of more-bribeable and less-bribeable police officers in such a situation) may be more or less in conflict with one another, but I don’t see how it helps anything to try to fit that into the same framework as we use for assessing different attitudes to political debate.
There’s definitely lots of distinctions that can be made. The aspect of conflict that I find most important to know about is the “epistemic” part of it. Basically, does Aumann’s agreement theorem apply?
In ordinary non-conflict conversations such as asking someone you are visiting where you can get a cup of water, you can simply copy other’s beliefs and do reasonably well, whereas in conflict conversations such as politics, copying your opponent’s beliefs about factual matters is a serious security vulnerability.
(It doesn’t seem uncommon for mistake theorists to come up with explanations of why Aumann’s agreement theorem wouldn’t apply to politics despite both sides being honest, but nobody has come up with compelling explanations, whereas the conflict theory analysis of it seems compelling and commonly for people to self-endorse.)
Right, I think it is another way to distinguish between believing in zero-sum and believing in positive sum games, no?
I suspect that this framework is not well suited for analyzing the traffic stop issue. At least I do not think the reasoning you describe is sufficiently common… A non-corrupt cop just follows the rules without thinking about the size of the pie, while a corrupt cop sees their position as an opportunity to grab some extra piece from someone else’s pie… which can be interpreted from either perspective, without providing any useful insight. Basically, this is not a nail, so using a hammer is not a great approach.
No. While zero-sum games is one way that people can be in conflict, they are not the only way:
Zero-sum games are just one example of a broader class of (game-theoretically equivalent) Pareto-frontier games, where one person’s benefit is another person’s loss, but where in the general case one person may benefit more than the other loses. (For instance, some might analyze the traffic case as a Pareto-frontier game, where the loss due to crashes is worse than whatever gain the driver might get from driving fast.)
Even if there globally is opportunity for Pareto improvements, locally one may be in a setting where there is no system to cooperate to achieve these improvements, and there is insufficient trust to negotiate to achieve them. In such a case there may be conflict and local opposing interests, even if there would be gain for deescalating the conflict and finding ways to help everyone.
In order for it to make sense to analyze others as being basically honest and merely making a mistake when they disagree, it would have to be that people aren’t intentionally trying to work against each other’s interests.