Guess culture screens for trying to cooperate

My friend Miri (quoted with permission) wrote this on Facebook a while back:

Midwesterners are intolerably passive aggressive. My family is sitting among some grass in the dunes because it's the only shady place and a park ranger drives by and says, "That grass you're sitting in--we try to protect that." I say the only thing that makes sense to say in response, which is, "Thanks for letting me know! We'll be careful with it." And I go back to my reading.

Then I look up and she's still there. I look at her for a few moments and she says, "You need to get out of there." I'm like, ok. Why can't you just say that the first time? Not everyone grew up in your damn convoluted culture. Say what you fucking mean.

In the comments, someone replied:

One of the best parts of NYC is that no one dances around what they mean to say here. On the contrary, once I heard a guy on the subway say, to confused-looking strangers, "Do you need some fucking help or what?”

This particular incident seems like obnoxious behavior on the part of the park ranger, but it got me curious about why this sort of norm seems to win out over more explicit communication in many places.

I think it's relevant that following local guess-culture and other implicit communication norms requires some active cognitive effort to infer people's preferences/boundaries. Our ability to detect shirkers and malicious actors is imperfect, and telling someone in every observed case exactly how their behavior seems wrong is just telling them how to slip in between the gaps in your ability to detect problems. For this reason, in high-trust situations, we want people to display some sort of active effort to infer what our preferences are, in order to accommodate them. People who don't do this thus appear not just slightly more costly to interact with, but rude and offensive - so it's easier to coordinate to exclude them.

Expending effort to infer others' preferences, needs, and boundaries, and how your actions affect these, is a classic "costly signal" since it's something a friendly person would want to do anyway, and something an unfriendly person would not want to do without norms that force it.
This would explain why explicit, overt communication is more prevalent in melting pot style situations where people from a lot of different subcultures are mixing in one-off interactions. New York City is a great example of this, as the commenter pointed out.

An important cost here is that making a scene harder for outsiders to exploit maliciously also makes it harder for harmless strangers to interface with, and imposes some cognitive overhead on all participants. Probably for people sufficiently acculturated (and not neurodivergent in certain ways) the training creates a passive process where they don't even see how they could do otherwise, which is fairly low-effort as long as they don't have to code-switch or do anything unusual.

This would also explain why explicit communication norms seem to coincide with liberal values like individual tolerance - people can harmlessly do unusual things only in environments where those don't send a lot of other unrelated and undesired signals.

Some evidence against this explanatory model is the extent to which places like New York seem to behave surprisingly like high-trust environments despite the lack of superficial social conformity. People seem happy to give directions, return lost wallets, etc. I don't know if there are smaller-town interactions that I don't know I'm missing because I'm used to more New York-like norms.

Explicit communication about preferences can be very valuable for bridging cultural gaps, or simply misaligned expectations, and often what we need is simply to tell each other what's wrong. But if someone crosses boundary after boundary with no apparent attempts to generalize, it doesn't matter whether they're malicious, or simply very much unlike you. If the other person isn't able to do some of the work towards being decent towards you by your standards, you can't explicitly tell them how. Not mustn't - can't. There's no litmus test for "try to infer my preferences, rather than directly optimizing on the metrics I can observe directly," unless you set up bullshit tests:

26 thoughts on “Guess culture screens for trying to cooperate

  1. lahwran

    I feel like this is getting close to the core of what I think is why the community I hang out with has always been so uncomfortable to me. I think that someone who already knows how to do the thing will recognize this blog post as important, but I don't think it's quite to the point that I feel like it'd have worked to explain it to me a year ago. I'm super glad you posted it.

    Reply
    1. Benquo Post author

      Fortunately, I'm going to be writing more on this topic, as it's the exact intersection of "how to do friendship with me" and "why integrity is important." Generally I think many small attempts at persuasion outperform one comprehensive attempt, since different people have different stuck points.

      Reply
  2. Nick T

    At least some of this is true for complying with (apparently-)arbitrary norms in general, and is why people showing a strong aversion to complying with any arbitrary norms has always bothered me.

    Reply
      1. Benquo Post author

        This opinion seems pretty surprising to me if I'm reading Nick correctly - I think he's basically saying that people who bulldoze other people's fences unless there's a sign explaining what they do are rude.

        Reply
        1. Michael

          Think about the things that seemed arbitrary or subtly or not so subtly 'wrong' when you were 8. How many of them *didn't* ultimately make the most sense as a mechanism for preserving and concealing power-relations.

          Reply
  3. Aceso Under Glass

    I think there's two separate tests going on: how well you actually guess someone's preferences, and how much energy you invest in doing so. Someone who's psychic and does things I like when they're low cost doesn't actually predict much about what they'll do when it's high cost. Someone who will put a lot of energy into figuring out what I want and then fails is a good ally for some projects and a terrible one for others. It's also easy to fake putting a lot of energy in, if you don't actually care about the answer. OTOH, it's easy to care a lot and put a lot of energy in and fail, if two people's frames of reference are different enough.

    Reply
    1. Benquo Post author

      This seems right, which is part of why it's important to be able to reject people as bad matches even if you're not sure they're malicious.

      Reply
  4. Howie

    This bit didn't ring true to me:
    "This would also explain why explicit communication norms seem to coincide with liberal values like individual tolerance - people can harmlessly do unusual things only in environments where those don't send a lot of other unrelated and undesired signals."

    1) I haven't actually noticed a coincidence between liberal values and explicit communication norms. I haven't spent much time in the midwest but I think it scores relatively high on individual tolerance but is the prototypical example of guess culture. On an individual level, a disproportionate number of the most tolerant people I've known find guess culture to be most natural to them. I think you could as easily make the opposite argument - guess culture only works in places with unusually high tolerance for people from other groups because it forces you to try to learn their norms.

    2) It's true that explicit communication norms decrease irrelevant signals somewhat when someone behaves unusually. But I think this effect is pretty narrow. Let's say I'm new to a culture and I know it has norms of "explicit communication" but I don't know much else about it. I think if I behaved unusually, I'd end up unintentionally and unknowingly sending all kinds of signals.

    In particular, you can't really have a norm of explicit communication in the abstract. Even in a culture of explicit communication, I can't possibly count on the fact that somebody will tell me if I unintentionally signal something to them (and I can't count on the fact that people will assume none of my signals have information value if they were not explicit). At most, a community might have norms of explicit communication about [some set of things.] If I know what that set of things is then I know I'm safe from accidentally sending signals about those things but not about anything else.

    I'm noticing that this also means that cultures emphasizing explicit communication can still vary on important dimensions and still require code switching and social intelligence to get things right. Even in "Ask" cultures, I've definitely experienced the phenomenon where somebody commits a faux pas by being explicit in ways that are not acceptable in that particular Ask culture. (When someone earnestly explains a faux pas with "I thought it would be ok bc we're in Ask culture" you've got an example of this.)

    Reply
    1. Benquo Post author

      On (2) I think I want to stand behind a more modest claim than the one you're arguing against - I think it's true on the margin that ask culture / explicit communication enables a wider range of nonaggressive behaviors, not that you can get cultural barriers down to near zero in some objective sense.

      On (1), I'm pretty uncertain what's true here, and more uncertain after reading your comment. Thanks for disabusing me of my imagined understanding!

      Reply
      1. Howie

        Cool. In case it's helpful/non-obvious, I'm going to point out the specific language that suggested a broader claim.

        1) The word "only"
        2) "don't send a lot of other unrelated and undesired signals" sounds pretty broad.

        If you'd said "people have more leeway to do unusual things in environments where those send fewer unrelated and undesired signals" I would've understood you to be making the more limited claim that I think you now endorse.

        Reply
  5. Howie

    "I think it's relevant that following local guess-culture and other implicit communication norms requires some active cognitive effort to infer people's preferences/boundaries."

    -For what it's worth, my experience is that Ask Culture takes me much more active cognitive effort than "guess culture." I generally can't turn off the part of my brain that's trying to infer other people's preferences so that's cognitive effort that's going to happen for me no matter what culture I'm in. But Ask Culture requires the additional cognitive effort of having to constantly monitor my own boundaries. In some sense, actually doing the work of speaking up about your own boundaries is a costly signal of cooperation in Ask Culture just like trying to infer someone else's boundaries is a costly signal of cooperation in Guess Culture. And, depending on which culture feels more native to you, signaling cooperation will probably be more costly to you in one culture or the other.

    Reply
    1. Benquo Post author

      I agree that there are cognitive burdens imposed by both cultures; I think the main difference is who bears the burden.

      In Ask Culture, the burden of causing boundaries to be respected is more on the people with the boundaries, and in Guess Culture it's more on the people at risk of crossing those boundaries.
      In Ask Culture, the burden of recognizing a need is on the person with the need; in Guess Culture, it's on the person with the resources to satisfy that need.

      Reply
      1. PDV

        Framed that way, Ask culture is obviously superior for the same reasons markets are powerful; it puts the decisions on the people with the most information about the value of their options.

        Do Ask/Guess cultures lines split up according to history of market-based society?

        Reply
  6. Howie

    Don't have time to fully formulate this comment so it may or may not make much sense. But something clicked for me where I think I realized the direction you're going with this and the point you're going to make a blog post or two from now.

    Let's assume I'm right about your main point but too lazy to write it out here. I wonder if you really need all the connotations of the "Guess Culture" framework to make the point you intend. It seems possible to me that Guess Culture related stuff is only adjacent to the main issue which is an issue of whether norms should be bright line based v. fuzzy. (I think Ask Cultures and Guess Cultures can both have rules that are bright lines or have fuzzy standards). For example, a culture could be guess culture in that the actor is supposed to anticipate another person's boundaries. But it could also be the case that if you guessed the boundary correctly it would be a bright line rules (e.g. stand at least 2 feet away from me) instead of a fuzzy one (stand at a distance that shows you're respectful of my personal space).

    Norms that are bright lines can be easier to game while evading their spirit. Thus, in communities with bright-line norms, following norms is less of a signal of cooperation.

    Is that close to what you mean?

    Reply
  7. John Salvatier

    Interesting!

    I really like the phrase 'trying to cooperate'.

    I am frequently trying to convince people of things. I'm usually trying to communicate an (often subtle) point I think is important, that I want the other person to get. It almost always feels to me like people are *not* 'trying to cooperate' in the conversation. They're not trying hard to guess what I mean. They're not checking that they understand the point I intended to communicate (as opposed to merely understanding my words). If they interpret me as meaning something wrong or trivial, they're not second guessing themselves. They're not guessing at my reasons for wanting to communicate that point. They don't make informative conversation noises and facial expressions ('oooooh!' -> insight, 'oh, I see' -> resolved confusion, *scrunched up face* -> 'that seems wrong to me', confused sounding 'huh...' and scrunched eyes -> something feels confusing, and many many more).

    When the rolls are reversed, when I guess someone else has a point they'd like me to get, I try really hard to guess the thing they'd like me to get. And work to make sure I haven't misunderstood it. I make sure to communicate where my current understanding is. If I imagine that someone was trying to communicate something that *meant* something to me, and I didn't try hard to actually get it, that feels *shameful*. They had a real point that meant something, and I was too cowardly to really check whether they were right. If I notice that I have been avoiding engaging with someone's point for the last 5 minutes, I feel humbled and feel ashamed to think of myself as a truth seeker. After all, they were right in front of me trying to give me truth.

    And I think this works really well for me. I am frequently able to get subtle points that actually affect me from people.

    At the same time, when I have a point, I try really hard to say what I mean directly and clearly. And to focus on the most important things. And I implicitly expect others to do that too.

    I really really really wish other people had this same attitude. We have a lot of people in the community who know some very excellent things, but that's not actually very useful if they don't actually get transferred to other people. And right now the diffusion of such things is very slow. If we want to do really excellent things, that probably has to change.

    The burden of making communication work mostly falls on the listener/student. Its that way because the student just has much better feedback about whether they've understood in a useful way. In a way that actually affects their actions. So if we want to do hard things, we're going to have to figure out how to make people better at 'trying to cooperate' in communication.

    Reply
    1. Aceso Under Glass

      Argh yes.

      This clicked something for me that's more relevant to Ben's main post than your comment, but it clicked here so I want to group them together.

      I have had issues with romantic partners before. I wasn't particularly mad that they were doing a suboptimal thing or failing to guess the optimal. I was mad that *all of the work* for improving was on me. What I really wanted was for them to figure out the correct questions to ask (perhaps after I brought up the general problem), so I didn't have to do the work of identifying the problem and figuring out what the root solution was on my own. This is closer to ask culture than guess, but it still means both parties have to do some emotional labor.

      Reply
  8. Nita

    I would guess the causality goes the other way: the need for deep inference / "mind-reading" skills is a side effect driven by the other side of the interaction -- runaway signaling of politeness and respect.

    In some cultures, "telling someone what to do" is something you do to your subordinates or "inferiors" (e.g., children), and showing proper respect to your peers or superiors involves using hints instead of direct requests. It signals that you see them as free persons who can choose whether or not to cooperate.

    You can observe some of this even in mainstream language. Obviously, when someone says, "Could you close the window?" they want the window to be closed, but the indirect form makes the request "softer" than "Close the window, please".

    Unfortunately, when everyone tries very hard to appear respectful and polite, the result can be something like a euphemism treadmill. As individuals invent new ways to appear even more polite than the norm, old signals of respect become insufficient.

    But for communication to work, people still need to read the actual request through the layers of conspicuous respect, which results in growing demands on their inference skills.

    In some other cultures, a sufficient level of respect is assumed by default, and redundant attempts to signal it via indirect communication are considered annoying or even rude.

    Reply
  9. Pingback: Honesty and perjury | Compass Rose

  10. Sophie Grouchy

    This is really similar in idea (but much better written) to a blog I wrote a bit ago. The problem I have with tell culture is that the people who practice it are often particularly unskilled at modeling other people's preferences, which leads them to trample all over boundaries constantly, which then has to be dealt with by me.

    As an example of boundary trampling, I let an out of town friend crash in my bed. They knew I had to be at work the next day at 8am, but stayed out until 1am without letting me know in advance that they even had any plans, and of course woke me up at 1am to go to bed. This sort of behavior, but CONSTANTLY.

    https://sophiegrouchy.wordpress.com/2016/01/26/morality/

    Reply
  11. Anonymous

    > Some evidence against this explanatory model is the extent to which places like New York seem to behave surprisingly like high-trust environments despite the lack of superficial social conformity. People seem happy to give directions, return lost wallets, etc. I don't know if there are smaller-town interactions that I don't know I'm missing because I'm used to more New York-like norms.

    I think this is the first time I've ever heard new york mentioned in the high-trust low-trust dichotomy as anything except an example of an extremely low trust environment 🙂

    Note that I'm not from a small town, so I can't tell you about that, but trust level isn't about whether you'll give directions. When someone starts speaking to you on the street, do you stop and wait for them to finish? Or do you keep walking?

    I give directions to people here pretty regularly, but if you start speaking to me I'll keep walking unless I hear the actual question - otherwise I'm likely to assume you're going to ask me for money (which *is* the right assumption, >90% of the time). That's a habit I had to learn.

    Reply
  12. Pingback: GiveWell and partial funding | Compass Rose

  13. Pingback: On the construction of beacons | Compass Rose

Leave a Reply to Benquo Cancel reply

Your email address will not be published. Required fields are marked *