Authoritarian Empiricism

post by Benquo · 2019-05-03T19:34:18.549Z · LW · GW · 10 comments

This is a link post for http://benjaminrosshoffman.com/authoritarian-empiricism/

Contents

10 comments

(Excerpts from a conversation with my friend Mack, very slightly edited for clarity and flow, including getting rid of most of the metaconversation.)

Ben: Just spent 2 full days offline for the holiday - feeling good about it, I needed it.

Mack: Good!

Ben: Also figured out some stuff about acculturation I got and had to unlearn, that was helpful

Mack: I'm interested if you feel like elaborating

Ben: OK, so, here's the deal.

I noticed over the first couple days of Passover that the men in the pseudo-community I grew up in seem to think there's a personal moral obligation to honor contracts, pretty much regardless of the coercion involved. The women seem to get that this increases the amount of violence in the world by quite a lot relative to optimal play, but they don't really tell the men. This seems related somehow to a thing where the men feel anxious about the prospect of modeling people as autonomous subjects - political creatures - instead of just objectifying them, but when they slap down attempts to do that, they pretend they're insisting on rigor and empiricism.

Which I'd wrongly internalized, as a kid, as good-faith critiques of my epistemics.

Story 1:

I was talking with my father about Adorno, the Enlightenment, and anti-Semitism, and the conversation was doing a reasonable-seeming thing, UNTIL he brought up the issue of high-fertility ethnic minorities with distinct political loyalties in democracies. So, naturally, first I explored the specific thing he brought up, which was that this strategy exploits a real security flaw in the democratic setup, and (since this came up in the context of Israel) that hypocritical ethnic majorities willing to occasionally violate their "standards" do a lot better patching the security flaw, than do ethnic majorities who insist on ACTUALLY having structurally neutral liberalism that takes care of and empowers everyone.

But, then, since we'd been talking about anti-Semitism, I had to point out that there's a structurally similar thing going on with Jews and credit-allocation systems in early financialized states like pre- and interwar Germany. If there had been actual coordination and an actual agenda, it would have been trivial to take over the state. (There wasn't and there wasn't, it's a trope in pre-WWII-era Jewish humor that the anti-Semitic newspapers kind of read like escapist fantasy). But, like, a double-digit percentage of elites is obviously enough, in a modern state where info-processing is abstract and mostly automated, to control quite a lot, given perfect coordination.

And he basically said, "you can't say that, because you don't have hard data."

Which, like, where am I gonna find hard data on the incidence of coups via groups with unreasonably high levels of coordination seizing control of the state's information-processing apparatus (thus causing the records to misreport reality as a side effect)?

When I poked him on this, he ended up retreating to the motte of "it's possible that what you're saying isn't true". Which, yes, obviously - it's speculation. But also obviously that isn't what he was originally saying. He was saying something like: It's wrong to reason about concrete situations based on hypotheticals about human potential; legitimate discourse is the sort of thing that could get into an academic journal (which is necessarily at least performing being apolitical in some sense, even in the journal's explicitly about political theory).

This helped a bunch of past stuff click for me, where e.g. he knows a lot about what the Rabbis of the Talmud said, and what later medieval commentators have said, and historical scholarship about how the text developed, and that's fine to talk about, but if I read them as though they were arguing about some specific real thing, try to understand and then talk about it, and use it to contextualize individual statements, that seems like "irresponsible" speculation to him.

Digression to an example I think is cool:

At the Passover Seder, we traditionally read a story about five Rabbis, in Roman times, staying up all night to study the Exodus from Egypt (the Passover story). (These are guys who were also associated both with the rebellions against Rome, and the successful transition to a permanently exilic Judaism.) And then in the morning their students come in and say "it's time to recite the morning Shema" (central affirmation that traditional Jews recite communally twice daily).

Turns out there's ANOTHER story in the Talmud about a rabbi staying up studying until his students come in to tell him it's time for the morning Shema, but this one is very different. It's Bar Yochai, a figure associated with mysticism / proto-Kabbalah. He's just generically studying Torah, not specifically the Exodus story. He's alone, not with peers. And when his students come in, he says that studying Torah takes precedence over anything else, so he's not going to come say the Shema with them, even though it's an obligatory commandment.

This is part of a broader disagreement between Bar Yochai and the other rabbis.

Another instance of the same disagreement:
Most of the Rabbis think that the commandment to attend to Torah (the teachings of Moses) all day means that if e.g. you're planting your crops, figure out how to do that in a Torah-ish way. Bar Yochai says you should literally just sit studying Torah, and if you do that well enough, gentiles will show up and plant your crops for you as a reward. So, Bar Yochai and his students tried it his way, and the other rabbis and their students tried it their way. And, empirically, Bar Yochai turned out to be mistaken. He got magic powers (the Talmud is very clear on this point), but his crops failed because he ... didn't plant or harvest them.

Basically he prioritized inner work over everything else, assuming that it's high enough leverage that other stuff would take care of itself, and the other rabbis thought that this stuff doesn't work outside the context of a community operating with some sort of synchronization, or outside the context of the mundane activities of life.

It's not hard to see why (a) the Talmud says that if there's any other school of thought available, never go with Bar Yochai's opinion on a legal matter, but also (b) the kabbalists saw him as an intellectual precursor.

So, linking this back to the underlying problem - describing the stories is OK, making inferences about them sort of registers as a kind of storytelling that can be fun/interesting, but my dad just can't engage with the idea that there's a fact of the matter about what these people were talking *about*, separate from what they explicitly said, and talk about kabbalah as political theory of change with concrete mundane implications.

Story 2:
I'd just talked with my mom a bunch about her adult ESL students - some of them are "unmotivated" and she'd recently realized it's in part because some are coerced to show up lest they lose their visas. I pointed out that she could just negotiate directly with them to work out a solution that allows the ones who want to learn to not be distracted, and that she's not morally obliged to force the ones who aren't interested in the class to pretend they are.

Then at 2nd seder a friend's father was talking with her about this, and as soon as he heard about the symptoms, he declared that she should set a firm boundary so that students that e.g. after n minutes the door of the classroom is locked and students who are too late are absent, that her first obligation is to her contract as a teacher, etc. And he basically just couldn't hear or wasn't interested in the fact that some of the students were under coercion, didn't seem to think that fact was morally relevant at all.

(None of these examples is hugely persuasive on their own, but each of them caused a long pattern of similar things to click).

When I pointed out that my mom wasn't morally obliged to collaborate with ICE he just denied that this had anything to do with what he was saying, without offering an argument.

Story 2b:
Same night, different incident.

My friend (the son of the guy from story 2) asked me how Pittsburgh was.

I responded with the following analogy:

While in Berkeley, it's like I was living on the first-class deck of the Titanic. In the distance, I can see the ship heading towards an iceberg. Meanwhile, all the first-class passengers are obsessed with scheming about how to become the captain, or otherwise take over the ship and get the nice staterooms and privileges.

I'm concerned with steering the ship to safety, but when I find people rallying around the stated intent to steer the ship to safety, they're mostly just another faction trying to take over the ship. I try to persuade individuals that ACTUALLY navigating is object-level important even though it doesn't affect anything in our immediate concrete environment, but this just seems to people like a weird bank-shot attempt to gain status by dominating the "steer the ship to safety" faction.

So, depressed and scared and emotionally scarred by this, I go to a place I've heard there are a bunch of sane competent engineers: the engine room!

It turns out, they ARE locally sane here. They're collaborating to do means-ends reasoning to keep the engine running, which keeps the lights on and keeps the ship moving forwards. Given the crazy situation we're in, keeping the ship moving forwards is not helping. But at least it's literally not their job to know about that, and they're doing what literally is their job. When I describe what's going on on the upper deck they don't seem particularly inclined to drop everything and come help, but they do seem sincerely concerned and interested in finding out whether they have any relevant resources they can direct to me. They understand in principle why steering the ship matters, and that hitting an iceberg would be bad in a way totally unrelated to factional politics.

Pittsburgh is the engine room.

So, I'm in the part of this analogy that's about the Bay, and my friend's dad jumps into the conversation to tell me that my analogy is too convoluted. So, I pause and ask him what part's hard to follow (he wasn't part of the conversation at first, but if someone wants to understand what I'm saying at a social event, it seems correct to try to include them), and he just keeps repeating that it's too convoluted, until eventually he changes his story and says "it's too crazy, I don't want to hear about it."

So, he was pretending to be critiquing my analogy, actually feels too much anxiety about the situation I'm describing to be OK letting someone else talk about it where he can hear, but felt the need to put himself above me by framing it as me making some sort of technical error in conversation.

Do you see how this seems like the same kind of thing my actual dad did?

Meanwhile, (back to the contracts thing), his wife works as a lawyer to advocate for kids whose needs aren't met by the family law & school system. She can't possibly do that job and think that the letter of the law even has an objective meaning, since it's literally her job to make it mean the thing that gets an okay outcome for the child.

The men of this category often end up in a position where they are the only one in their area who are technically adept at the thing people with their job description are supposedly certified to know about, or who care to do the object-level technical work.

I think this specific gendered dynamic might be particular to secular American Jews.

Mack: Okay that's interesting. Definitely seen similar things play out but not in such a gendered way. Thinking about my parents in particular, they end up on the "male" side of your stories occasionally. Not consistently at all. Hm maybe the examples coming to mind are only superficially similar.

Ben: Want to work through the details of one? Might be good to precisely formulate the distinction if there is one.

Mack: Re: not treating people like political entities, I can think of examples of that. But I suspect the reasons are different.

Ben: I suspect there's a shared sense to think of people of the other political party as defective parts of a machine, rather than as adversaries who might be negotiated with or fought but with whom there's not currently a shared paradigm. But, not a shared tendency to specifically dismiss attempts to model people as agents, as unscientific.

Mack: Things that come to mind: a knee jerk reaction among the older members on one side of the family to treat this kind of reasoning as...vulgar?

Ben: What does an example of the sort of thing they've reacted to this way look like? Actual or fictional examples both fine. Actual are better, but whatever prediction/generation function you learned is also valuable intel. (Just like fictional stories by competent poets are valuable intel)

Mack: I'm thinking of a cousin who is very similar to me. There are running jokes about us being in the same room and driving people crazy because we "start controversies." I think there was a conversation about Boise's homelessness policies, and she and I were talking about things like: the reasons the city might have taken recent aggressive action against the homeless population, essentially the different incentives at play.

We disagreed but it was sane disagreement, and her mother and grandmother were just visibly distressed. And they tried talking about ministry attempts, harsher drug laws, etc. Retreating to party lines on homelessness (red tribe). The conversation ended with her mother saying "Well then why bother!" as we poked at the policies they'd brought up.

It isn't the same retreat to what could be published in an academic journal, or to the obligation of a contract. But it is kind of like your Titanic analogy. Laying out the specific reasons a problem is hard, the normal party lines or grumbling not being sufficient or satisfying, and finding it rude to point out why a problem is hard, especially if it isn't about the outgroup being wrong or misled by satanic forces.

Ben: OK, so it sounds like your family is nondissociatedly anxious about politics, while mine (at least the men) retreats to dissociatedly identifying with an authority narrative that insists that only "apolitical" knowledge is speakable; your family more overtly identifies as members of a faction, while mine identifies with abstract shared authority.

Mack: That sounds right. Ah, so this side of the family is also pretty bound to contracts of a sort, though they aren't quite as aligned with the law.

Ben: All "legally binding" contracts, or just uncoerced personal agreements?

Mack: All legally binding contracts to an extent though that's more about avoiding punishment and being Good. Seeing the local social mores *as* binding contracts, I think.

Ben: That last thing seems noncrazy to me - like, an attitude I'd see in some fully functional societies.

Mack: It isn't crazy.

Ben: Whereas I think the thing I was pointing to is crazy, and the other things are somewhere in between.

Mack: I think I see the distinction. Feels like there's something familiar in my experience that's closer to the crazy side and I'm trying to figure out where that comes from.

Initial recognition was about the discomfort and retreat - I have a lot of examples of the role you took in those anecdotes being seen as extremely rude, uncomfortable, vulgar. I don't think it comes from the same place as the specific dynamic, though.

Recognition also of the realization that the people arguing around me were not arguing to try to understand something or solve the problem.

Ben: OK, I think the discomfort-and-retreat pattern is a specific kind of defensiveness, on behalf of the ruling regime by people identifying with it (where the ruling regime can be a local community's norms, or the state, or an ideology, etc etc.)

That's an important piece of model to have, it's one of the gears here. It connects to more than one possible type of defense or sense of threat.

Mack: I am curious about whether I've observed something closer.

Maybe this: at work some very expensive material was mixed. The timeline for new material was too long to meet even the revised deadlines for the product, there was no good mechanical solution, etc. So the bosses had been rotating employees through the tedious task of unmixing it by hand.

HR lady and I helped with this during some plant wide mandatory overtime.

She was insistent that the right thing to do would be to force the person responsible for the mess to devote all of their work hours plus overtime to fixing it.

I argued a little - not too hard because office norms. But her retreat was to a supposed alignment with company interests (even though, IMO, the solution was an okay compromise with multiple goals for the plant).

And it has come out over time that, as far as I can tell, she believes very strongly that when you begin employment you must suspend a large chunk of your personal interests and align them with the firm, or you're a subpar employee. And while this probably helps her a lot in some of her HR functions, she is resistant to discussing the individual incentives that prevent people from being "good employees" once they've come on to her radar as "bad employees."

Ben: The HR thing sounds like it might be an exact match with a big part of this. I do want to distinguish loyalty to a specific local institution, from loyalty to one's profession/contract. They're different kinds of implied coordination strategies.

Mack: Which loyalty is the one present in your stories?

Ben: The latter. So, the HR lady identifies her interests with the interests of the company she's attached to, that's her gang. But the guys I'm talking about identify with each other as members of a mercenary class with a perceived shared interest in upholding professional standards, so that they can be interchangeable pieces and charge for this.

Mack: Ahhh
Okay
Something clicked

Ben: Like, a doctor will identify with Doctors as a profession, not with the hospital and nurses. In-house counsel will often favor the class interests of lawyers over the interests of their company.

The Guild.

*****

Related: Blind Empiricism [LW · GW]

10 comments

Comments sorted by top scores.

comment by ryan_b · 2019-05-03T20:47:01.459Z · LW(p) · GW(p)
Which, like, where am I gonna find hard data on the incidence of coups via unreasonably high levels of coordination seizing control of the state's info-processing apparatus (thus causing the records to misreport reality as a side effect)?

You may be interested in the National Salvation Front. With only a little simplification, when Ceausescu was overthrown six people showed up to the state-controlled media center, announced the creation of the NSF and that they were now in charge. Four days later they were the government.

Replies from: Douglas_Knight
comment by Douglas_Knight · 2019-05-09T18:05:41.924Z · LW(p) · GW(p)

That sounds like a totally normal coup, openly seizing power and thus creating a Schelling point for everyone to fall in line behind them. Maybe they lied about having written earlier manifestos that also acted as a Schelling point, but that would just be a single lie; and just a bald assertion, not backed up by the authority of some "info-processing apparatus."

comment by Douglas_Knight · 2019-05-09T17:30:18.403Z · LW(p) · GW(p)
This seems related somehow to a thing where they

I think it would be clearer if you said who "they" is, men or women. Having read the post and found that it's all about men, it's pretty clear that it is men, but when I first read it I found it pretty ambiguous and in fact assumed that it was women. On a slightly different note, my failure to understand the relation between the two statements seems like a good example of the writing advice that you should repeat your thesis at the end, because it may be incomprehensible on the first pass. You do have a kind of thesis at the end, but it seems backwards, the more easily comprehended version at the end, rather than the beginning.

Replies from: Benquo
comment by Benquo · 2019-05-09T17:33:25.490Z · LW(p) · GW(p)

Thanks, the pronouns were unclear, hopefully my edit fixed that.

comment by habryka (habryka4) · 2019-05-04T02:06:34.134Z · LW(p) · GW(p)

Promoted to frontpage, with some caveats:

For context, the two core principles for the frontpage guidelines are:

1. The content should be understandable without being deeply intertwined in the social fabric of the community

The goal of this is to allow LessWrong to grow as an intellectual community, in a way that isn't constrained by people making social commitments to the more in-person and localized parts of the community. It seems very bad if a requirement for participating in LessWrong discussion necessarily becomes that you know what all the different organizational acronyms are, and that you know the social history of all the ideas, and that you have pledged allegiance in some kind of social contract, etc.

2. The content should be written with the intention of explaining, not persuading

The goal of this rule is to prevent the weaponization of posts on LessWrong and to maintain a certain form of epistemic trust between the participants on LessWrong. Obviously the separation between explaining something and persuading someone of something is quite nuanced and hard to clearly define, but I think it's a valuable guidelines for maintaining a high quality of discussion on the site.

(There are also some other guidelines, like not being about topics that reliably attract discussions that violate the above two principles, or just attract people to LessWrong that will cause the site culture to get destroyed, but for this case these are the core ones)

Now to this post:

I think this post is a bit at the edge of the first principle, with a variety of references to the Berkeley local community that I don't really want to be a prerequisite to participation on LessWrong, but it overall only makes up a small part of the post and so I think is overall fine.

I think it does pretty well on the second part, because it's just a transcript between the author and another person, which prevents the content from being optimized to persuade this specific audience (though there are complicated incentives and selection effects here that make this not a fully clear call).

Overall I think the post is a decent fit for the frontpage, but it seemed like a good time to clarify the frontpage guidelines and illustrate precisely where this post fits, since I think it's a bit of an edge-case and deserves further explanation.

Replies from: Benquo, Benquo
comment by Benquo · 2019-05-05T05:51:56.956Z · LW(p) · GW(p)

On 2, for the specific dimension you mentioned, this seems massively better (as a dialogue between two actual people) than e.g. the dialogue in Moloch's Toolbox, in which all sides were written by the same person, with a clear prior persuasive agenda, so if this isn't fully a clear call, what should we think of that?

What are the "complicated incentives and selection effects" here?

comment by Benquo · 2019-05-05T06:46:14.705Z · LW(p) · GW(p)

Overall I'm pretty confused by this comment, and don't really understand what the "edge case" reasoning is. The reference to Berkeley - which I take to be what you're saying makes this an edge case - doesn't reference any specifics that require insider knowledge to understand. It's literally an account of how I explained it to a few people who are not part of the Berkeley local community (most of whom seemed to find it perfectly intelligible). A friend on Twitter commented that he might as well have swapped the city names and titled it "why I left DC and what St Louis was like." Pretty much the only thing I can imagine requiring local knowledge is the subsequent unexplained reference to "the Bay," which is unnecessary so I'll fix it.

Replies from: habryka4
comment by habryka (habryka4) · 2019-05-06T01:09:16.609Z · LW(p) · GW(p)

I think it would be a bit confusing to an outsider, but I agree that it wouldn't be any significant obstacle to understanding, which was my whole point.

The edge-case-ness does not come from it being not a good fit, but more from the dimension of "there are common misunderstandings of our rules that might cause someone to think this post should not be on the frontpage". You should think of the comment above being written to help people understand what our guidelines are and why this post is a good fit for the frontpage, not as something that tries to indicate the post is only barely a good fit.

Replies from: Benquo
comment by Benquo · 2019-05-06T02:51:34.467Z · LW(p) · GW(p)

Thanks for explaining, makes sense to me now.

comment by Dagon · 2019-05-03T23:00:05.381Z · LW(p) · GW(p)

At least some of these examples seem more like role identification more than contractualism or group loyalty. It feels like just the traditional male role of enforcing expectations (whether agreed in contract or imposed) via authority seems to run throughout.