I.

A few years ago I had lunch with another psychiatrist-in-training and realized we had totally different experiences with psychotherapy.

We both got the same types of cases. We were both practicing the same kinds of therapy. We were both in the same training program, studying under the same teachers. But our experiences were totally different. In particular, all her patients had dramatic emotional meltdowns, and all my patients gave calm and considered analyses of their problems, as if they were lecturing on a particularly boring episode from 19th-century Norwegian history.

I’m not bragging here. I wish I could get my patients to have dramatic emotional meltdowns. As per the textbooks, there should be a climactic moment where the patient identifies me with their father, then screams at me that I ruined their childhood, then breaks down crying and realizes that she loved her father all along, then ???, and then their depression is cured. I never got that. I tried, I even dropped some hints, like “Maybe this reminds you of your father?” or “Maybe you feel like screaming at me right now?”, but they never took the bait. So I figured the textbooks were misleading, or that this was some kind of super-advanced technique, or that this was among the approximately 100% of things that Freud just pulled out of his ass.

And then I had lunch with my friend, and she was like “It’s so stressful when all of your patients identify you with their parents and break down crying, isn’t it? Don’t you wish you could just go one day without that happening?”

And later, my supervisor was reviewing one of my therapy sessions, and I was surprised to hear him comment that I “seemed uncomfortable with dramatic expressions of emotion”. I mean, I am uncomfortable with dramatic expressions of emotion. I was just surprised he noticed it. As a therapist, I’m supposed to be quiet and encouraging and not show discomfort at anything, and I was trying to do that, and I’d thought I was succeeding. But apparently I was unconsciously projecting some kind of “I don’t like strong emotions, you’d better avoid those” field, and my patients were unconsciously complying.

I wish I could say my supervisor’s guidance fixed the problem and I learned to encourage emotional openness just as well as my colleague. But any improvement I made was incremental at best. My colleague is a bubbly extravert who gets very excited about everything; I worry that to match her results, I would have to somehow copy her entire personality.

But all was not lost. I found myself doing well with overly emotional patients, the sort who had too many dramatic meltdowns to do therapy with anybody else. With me, they tended to give calm and considered analyses of their problems, as if they were lecturing on a particularly boring episode from 19th-century Norwegian history. Everyone assumed that meant I was good at dealing with difficult cases, and must have read a bunch of books about how to defuse crises. I did nothing to disabuse them of this.

Then a few days ago I stumbled across the Reddit thread Has Anyone Here Ever Been To An LW/SSC Meetup Or Otherwise Met A Rationalist IRL? User dgerard wrote about meeting me in 2011, saying:

His superpower is that he projects a Niceness Field, where people talking to him face to face want to be more polite and civil. The only person I’ve met with a similar Niceness Field is Jimmy Wales from Wikipedia…when people are around [Jimmy] talking to him they feel a sort of urge to be civil and polite in discourse 🙂 I’ve seen people visibly trying to be very precise and polite talking to him about stuff even when they’re quite upset about whatever it is. Scott has this too. It’s an interesting superpower to observe.

I should admit nobody else has mentioned anything like this, and that narcissism biases me toward believing anyone who says I have a superpower. Still, it would explain a lot. And not necessarily in a good way. I’ve always believed psychodynamic therapies are mostly ineffective, and cognitive-behavioral therapies very effective, because all my patients seem to defy the psychodynamic mode of having having weird but emotionally dramatic reactions to things in their past, but conform effortlessly to the cognitive-behavioral mode of being able to understand and rationally discuss their problems. And the more I examine this, the more I realize that my results are pretty atypical for psychiatrists. There’s something I’m doing – totally by accident – to produce those results. This is worrying not just as a psychiatrist, but as someone who wants to know anything about other people at all.

II.

New topic: paranoia and Williams Syndrome.

Paranoia is a common symptom of various psychiatric disorders – most famously schizophrenia, but also paranoid personality disorder, delusional disorder, sometimes bipolar disorder. You can also get it from abusing certain drugs – marijuana, LSD, cocaine, and even prescription drugs like Adderall and Ritalin. The fun thing about paranoia is how gradual it is. Sure, if you abuse every single drug at once you’ll think the CIA is after you with their mind-lasers. But if you just take a little more Adderall than you were supposed to, you’ll be 1% paranoid. You’ll have a very mild tendency to interpret ambiguous social signals just a little bit more negatively than usual. If a friend leaves without saying goodbye, and you would normally think “Oh, I guess she had a train to catch”, instead you think “Hm, I wonder what she meant by that”. There are a bunch of good stimulant abuse cases in the literature that present as “patient’s boss said she was unusually standoffish and wanted her to get psychiatric evaluation”, show up in the office as “well of course I’m standoffish, everyone in my office excludes me from everything and is rude in a thousand little ways throughout the day”, and end up as “cut your Adderall dosage in half, please”.

(“Why is that psychiatrist telling me to cut my Adderall in half? Does he think I’m lying about having ADHD? Is he calling me a liar? These doctors have always treated me like garbage. I HAVE RIGHTS, YOU KNOW!”)

Williams Syndrome is much rarer – only about 1/10,000 people, and most of them die before reaching adulthood. It’s marked by a sort of anti-paranoia; Williams patients are incapable of distrusting anyone. NPR has a good article, A Life Without Fear, describing some of what they go through:

Kids and adults with Williams love people, and they are literally pathologically trusting. They have no social fear. Researchers theorize that this is probably because of a problem in their limbic system, the part of the brain that regulates emotion. There appears to be a disregulation in one of the chemicals (oxytocin) that signals when to trust and when to distrust. This means that it is essentially biologically impossible for [them] to distrust.

The results are less than heartwarming:

As Isabelle got older, the negative side of her trusting nature began to play a larger role. A typical example happened a couple of years ago, when Jessica and her family were spending the day at the beach. Isabelle had been begging Jessica to go to Dairy Queen, and Jessica had been putting her off. Then Isabelle overheard a lady just down the beach.

“She was telling her kids, ‘OK, let’s go to the Dairy Queen,’ ” Jessica says. “And so Isabelle went over and got into the lady’s van, got in the back seat, buckled up and was waiting to be taken to Dairy Queen with that family.”

Jessica had no idea what had happened to Isabelle and was frantically searching for her when the driver of the van approached her and explained that she had been starting her car when she looked up and saw Isabelle’s face in the rearview mirror.

The woman, Jessica says, was incredibly angry.

“She said, ‘I am a stranger, you know!’ ” Jessica says. Essentially, the woman blamed Jessica for not keeping closer watch on her daughter — for neglecting to teach her the importance of not getting into a car with someone she didn’t know. But the reality could not be more different. “It’s like, ‘My friend, you have no idea,’ ” Jessica says.

In fact, because of Isabelle, Jessica has had to rethink even the most basic elements of her day-to-day life. She can not take Isabelle to the dog park. She tries not to take Isabelle to the store. And when the doorbell rings, Jessica will leap over a coffee table to intercept her.

It’s not just Jessica and her family who must be vigilant. Every teacher at Isabelle’s public school has been warned. Isabelle is not allowed to tell them that she loves them. Isabelle is not supposed to tell other schoolchildren that she loves them. And there are other restrictions.

“She’s not allowed to go to the bathroom alone at her school, because there have been numerous instances of girls with Williams syndrome being molested at school when they were alone in the hallway,” Jessica says. “And these are like middle class type schools. So it’s a very real problem. And, you know, I’d rather her be overly safe than be on CNN.”

Some of the research on these kids is fascinating – I’m not sure I believe the study finding that they’re incapable of racism, but the one finding a deficit detecting anger in faces seems pretty plausible.

Williams Syndrome usually involves mental retardation, but not always. Some of these people have normal IQ. It doesn’t really help. Threat-detection seems to be an automated process not totally susceptible to System II control. Maybe it’s like face-blindness. Intelligence can help a face-blind person come up with some systems to reduce the impact of their condition, but in the end it’s just not going to help that much.

Psychiatric disorders are often at the extremes of natural variation in human traits. For every intellectually disabled person, there are a dozen who are just kind of dumb. For every autistic person, there are a dozen who are just sort of nerdy. And so on. We naturally think of some people as more trusting than others, but maybe that isn’t the best frame. “Trusting” implies that we all receive the same information, and just choose how much risk we’re willing to tolerate. I don’t know if that’s true at all.

A recent theme here has been the ways that our sense-data is underdetermined. Each datum permits multiple possible explanations: this is true of visual and auditory perception, but also of the social world. A pretty girl laughs a little too long at a man’s joke; is she trying to flirt with him, or just friendly? A boss calls her subordinate’s work “okay” – did she mean to compliment him, or imply it was mediocre? A friend breaks off two appointments in a row, each time saying that something has come up – did something come up, or is he getting tired of the friendship? These are the sorts of questions everyone navigates all the time, usually with enough success that when autistic people screw them up, the rest of society nods sagely and says they need to learn to understand how to read context.

But “context” means “priors”, and priors can differ from person to person. There’s a lot of room for variation here before we get to the point where somebody will be so off-base that they end up excluded from society. Just as there’s a spectrum from smart to dumb, or from introverted to extraverted, so there’s a spectrum in people’s tendencies to interpret ambiguous situations in a positive or negative way. There are people walking around who are just short of clinically paranoid, or just shy of Williams Syndrome levels of trust. And this isn’t a value difference, it’s a perceptual one. These people aren’t bitter or risk-averse – or at least they don’t start off that way. They just notice how everyone’s hostile to them, all the time.

III.

Another change in topic: bubbles.

I’ve written before about how 46% of Americans are young-earth creationists, and how strongly that fails to square with my personal experience. I’ve met young-earth creationists once or twice. But of my hundred closest friends/co-workers/acquaintances, I think zero percent of them fall in that category. I’m not intentionally selecting friends on the basis of politics, religion, or anything else. It just seems to have happened. Something about my personality, location, social class, et cetera has completely isolated me from one particular half of the US population; I’m living in a non-creationist bubble in the midst of a half-creationist country.

What other bubbles do I live in? A quick look over my Facebook and some SSC survey results finds that my friends are about twenty times more likely to be transgender than the general population. There are about twice as many Asians but less than half as many African-Americans. Rates of depression, OCD, and autism are sky-high; rates of drug addiction and alcoholism are very low. Programmers are overrepresented at about ten times the Bay Area average.

I didn’t intend any of these bubbles. For example, I’ve never done any programming myself, I’m not interested in it, and I try my best to avoid programmer-heavy places where I know all the conversations are going to be programming-related. Hasn’t helped. And I’m about as cisgender as can be, I have several Problematic opinions, and I still can’t keep track of which gender all of my various friends are on a month-to-month basis. Part of it is probably class-, race-, and location-based. And I have some speculative theories about the rest – I think I have a pretty thing-oriented/systematizing thinking style, and so probably I get along better with other groups disproportionately made up of people whose thoughts work the same way – but I didn’t understand any of this until a few years ago and there are still some parts that don’t make sense. For now I just have to accept it as a given.

There are other bubbles I understand much better. Most of my friends are pretty chill and conflict-averse. This is because I used to have scarier conflict-prone friends, and as soon as I got into conflicts with them, I broke off the friendship. I’m not super-proud of this and it’s probably one of those maladaptive coping styles you always hear about, and a lot of people have told me I’m really extreme on this axis and need to be better at tolerating aggressive people – but whenever I try, I find it unpleasant and stop. I know some other people who seem to actively seek out abrasive types so they can get in fun fights with them. I don’t understand these people at all – but whatever their thought processes, we have different bubbles.

All of this goes double or triple for people I’ve dated. I don’t think of myself as clearly having a “type”, but people I date tend to turn out similar in dimensions I didn’t expect when I first met them. I’m going to be ambiguous here because it’s a small enough sample that I don’t want to give away people’s private information, but it’s true.

I think about this a lot when I meet serial abuse victims.

These people are a heartbreaking psychiatric cliche. Abused by their parents, abused by their high school boyfriend, abused by their first husband, abused by their second husband, abused by the guy they cheated on their first husband with, abused by the friend they tried to go to for help dealing with all the abuse. The classic (though super offensive) explanation is that some people seek out abusers for some reason – maybe because they were abused as children and they’ve internalized that as the “correct” model of a relationship.

And maybe this is true for some people. I have a friend who admits it’s true of her – her current strategy is to try to find someone in the sweet spot between “jerkish/narcissistic enough to be interesting” and “jerkish/narcissistic enough to actually abuse her”, and she’s said so in so many words to people trying to matchmake. I guess all I can do is wish her luck.

But for a lot of people, this sort of claim is just as offensively wrong as it sounds. I know people who have tried really hard to avoid abusers, who have gone to therapy and asked their therapist for independent verification that their new partner doesn’t seem like the abusive type, who have pulled out all the stops – and who still end up with abusive new partners. These people are cursed through no fault of their own. All I can say is that whatever mysterious forces connect me to transgender pro-evolution programmers are connecting them to abusers. Something completely unintentional that they try their best to resist gives them a bubble of terrible people.

I want to emphasize as hard as I can that I’m not blaming them or saying there’s anything they can do about their situation, and I have no doubt that despite my emphasis people are still going to accuse me of saying this, and I apologize if any of this sounds at all like anything in this direction. But something has to be happening here.

IV.

Sometimes I write about discrimination, and people send me emails about their own experiences. Many sound like this real one (quoted here with permission) from a woman who studied computer science at MIT and now works in the tech industry:

In my life, I have never been catcalled, inappropriately hit on, body-shamed, unwantedly touched in a sexual way, discouraged from a male-dominated field, told I couldn’t do something because it was a boy thing, or suffered from many other experiences that have traditionally served as examples as ways that women are less privileged. I have also never been shamed for not following gender norms (e.g. doing a bunch of math/science/CS stuff); instead I get encouraged and told that I’m a role model. I’ve never had problems going around wearing no make-up, a t-shirt, and cargo pants; but on the rare occasion that I do wear make-up / wear a dress, that’s completely socially acceptable…Hopefully my thoughts/experiences are helpful for your future social justice based discussions.

Other times they sound like the opposite. I don’t have anyone in this category who’s given me permission to quote their email verbatim (consider ways this might not be a coincidence), but they’re pretty much what you’d expect – a litany of constantly being put down, discriminated against, harassed, et cetera, across multiple jobs, at multiple companies, to the point where they complain it’s “endemic” (I guess I can quote one word) and that we need to reject a narrative of “a few bad apples” because really it’s a problem with all men to one degree or another.

These dueling categories of emails have always confused me. At the risk of being exactly the sort of creepy person the second set of writers complain about, I hunted down some of these people’s Facebook profiles to see if one group was consistently more attractive than the other. They weren’t. Nor is there any clear pattern in what industries or companies they work at, what position they’re in, or anything else like that. There isn’t even a consistent pattern in their politics. The woman I quote above mentions that she’s a feminist who believes discrimination is a major problem – which has only made it extra confusing to her that she never experiences any of it personally.

These people don’t just show up in my inbox. Some of them write articles on Slate, Medium, even The New Yorker, discussing not just how they’ve never experienced discrimination, but how much anger and backlash they’ve received when they try to explain this to everyone else. And all of them acknowledge that they know other people whose experiences seem to be the direct opposite.

I used to think this was pretty much just luck of the draw – some people will end up with nice people at great companies, other people will end up with bigots at terrible companies. I no longer think this explains everybody. Take that New Yorker article, by a black person who grew up in the South and says she was never discriminated against even once. I assume in her childhood she met thousands of different white Southerners; that’s a pretty big lucky streak for none of them at all to be racists, especially when you consider all the people who report daily or near-daily harassment. Likewise, when you study computer science in college and then work in half a dozen tech companies over the space of decades and never encounter one sexist, that’s quite the record. Surely something else must be going on here.

V.

And I think this has to come back to the sorts of things discussed in Parts I, II, and III.

People self-select into bubbles along all sorts of axes. Some of these bubbles are obvious and easy to explain, like rich people mostly meeting other rich people at the country club. Others are more mysterious, like how some non-programmer ends up with mostly programmer friends. Still others are horrible and completely outside comprehension, like someone who tries very hard to avoid abusers but ends up in multiple abusive relationships anyway. Even for two people living in the same country, city, and neighborhood, they can have a “society” made up of very different types of people.

People vary widely on the way they perceive social interaction. A paranoid schizophrenic will view every interaction as hostile; a Williams Syndrome kid will view every interaction as friendly. In between, there will be a whole range of healthy people without any psychiatric disorder who tend toward one side or the other. Only the most blatant data can be interpreted absent the priors that these dispositions provide; everything else will only get processed through preexisting assumptions about how people tend to act. Since things like racism rarely take the form of someone going up to you and saying “Hello, I am a racist and because of your skin color I plan to discriminate against you in the following ways…”, they’ll end up as ambiguous stimuli that everyone will interpret differently.

Finally, some people have personalities or styles of social interaction that unconsciously compel a certain response from their listeners. Call these “niceness fields” or “meanness fields” or whatever: some people are the sort who – if they became psychotherapists – would have patients who constantly suffered dramatic emotional meltdowns, and others’ patients would calmly discuss their problems.

The old question goes: are people basically good or basically evil? Different philosophers give different answers. But so do different random people I know who aren’t thinking philosophically at all. Some people describe a world of backstabbing Machiavellians, where everybody’s a shallow social climber who will kick down anyone it takes to get to the top. Other people describe a world where everyone is basically on the same page, trying to be nice to everyone else but getting stuck in communication difficulties and honest disagreements over values.

I think both groups are right. Some people experience worlds of basically-good people who treat them nicely. Other people experience worlds of awful hypocritical backstabbers. This can be true even if they live in the same area as each other, work the same job as each other, et cetera.

And it’s not just a basic good-evil axis. It can be about whether people are emotional/dramatic or calm/rational. It can be about whether people almost always discriminate or almost never do. It can be about whether they’re honest or liars, shun outsiders or accept them, welcome criticism or reject it. Some people think elites are incompetent parasites; others that they’re shockingly competent people who mean well and have interesting personalities. Some people think Silicon Valley is full of overpriced juicers, other people that it’s full of structured-light engines. And the people who say all these things are usually accurately reporting their own experiences.

Some people are vaguely aware of this in the form of “privilege”, which acknowledges different experiences at the cost of saying they have to line up exactly along special identity categories like race and gender. These certainly don’t help, but it’s not that simple – as proven by the article by that black Southerner who says she never once encountered discrimination. I’ve seen completely incomprehensible claims about human nature by people of precisely the same race, sex, class, orientation, etc as myself, and I have no doubt they’re trying to be truthful. The things that divide us are harder to see than we naively expect. Sometimes they’re completely invisible.

To return to a common theme: nothing makes sense except in light of inter-individual variation. Variation in people’s internal experience. Variation in people’s basic beliefs and assumptions. Variation in level of abstract thought. And to all of this I would add a variation in our experience of other people. Some of us are convinced, with reason, that humankind is basically good. Others start the day the same way Marcus Aurelius did:

When you wake up in the morning, tell yourself: the people I deal with today will be meddling, ungrateful, arrogant, dishonest, jealous and surly. They are like this because they cannot tell good from evil.

Notice this distinction, this way in which geographic neighbors can live in different worlds, and other people’s thoughts and behaviors get a little more comprehensible.

New Comment
16 comments, sorted by Click to highlight new comments since:

It's interesting what this means for psychiatry as a science. If two psychiatrist who do the same technique have radically different effects on their clients the assumptions behind evidence-based psychiatrist are shacky and it's basically what Feymann called cargo-cult science.

The evidence-based medicine paradigm won't bring the goods that are promised when different therapists who do the same standardized technique get radically different results.

We need to stop treating the evidence-based ideology as an ideal and move on to prediction-based medicine that can actually fulfill the promise of telling us whether treatment A or treatment B is more likely to cure us.


One of the best predictors of successful outcomes for a therapy patient, is that the patient trusts the therapist.

https://en.wikipedia.org/wiki/Therapeutic_relationship

This position is roughly 80 years old. Personally I think the best heuristic for telling wether a therapist is any good is whether they believe the connection between the two of you is more important than their personal preferred theoretic approach.

Relevant scholarship:
https://en.wikipedia.org/wiki/Dodo_bird_verdict
https://en.wikipedia.org/wiki/Common_factors_theory

I model therapy as an art and craft, a special subset of social skills, just like empathetic listening or building rapport, rather than a deterministic applied science like EEng or MechEng. Then, getting different results from the same therapist doing the same technique is exactly what I would expect because they probably aren't doing all the non-verbal communication which makes the largest impact in social situations. Just like Scott has the passive ability of making people around him devolve into really civil calm discussions, a desired social effect is not something easily reproduced by others, nor learned in many cases less rare than Scott's reality distortion field of civility.

I feel like I'm being a little mean and condescending here. My apologies, I've been up too late. But I think it's poor practice to throw stones at glass houses of the academy, without doing some good scholarship first.

I'm aware of the studies that suggest that empathy and alliance are more important than the theoretic approach of the therapist and made that point previously on LW. It's just that there are so many possible stones to throw that I'm not throwing them all at once.

Why do you expect prediction markets to be more useful for this than evidence-based methods which take into account interactions between the practitioner's characteristics and whatever method they are using?

I'm not advocating what Hanson calls prediction-markets. I'm advocating a different setup that's described in the linked article.

The core problem is that even if it would be possible in a perfect world to run evidence-based studies to gather the knowledge in the present system there are no economic incentives for anybody to run the required studies in a way that's likely to lead to effective clinical predictions. There's no accountability that pushes clinical trial design in a way that leads to clear clinical benefits. The incentives are mostly about overstating the effect of the intervention that's studied.

Even if there would be a sincere attempt at running the required studies it would be much more expensive than the way we currently study interventions and that means we are likely to study less interventions and thereby slowing down innovation by making the invention of new interventions more costly.

(Crossposted from the SSC comment thread)

Many years ago, I worked a security job that involved selecting some people for “additional screening”. We were issued a mechanical pseudorandom number generator. Essentially, you would click the button each time someone would come through the checkpoint. After some (non-constant) number of clicks, it would have a different readout that is supposed to prompt you to select the next person in line for additional screening. We were also instructed to never tell the public how they are selected.

The majority of the time, the person selected was a member of the majority demographic category in the region in which we were operating (as would be expected). They would be selected, shrug, and say something along the lines of “I guess it’s my lucky day”. Much more rarely, someone would get selected from a minority group. A fraction of these people would say something along the lines of “Oh, what a surprise, you picked the #{minority class/race/etc…} person” and be visibly agitated. At the time, I wondered how many of these incidents got recounted to others as clear instances of discrimination, not knowing that the result was determined by a dumb clicker.

No one seems to be mentioning here that subcommunications are a real, studied thing that encompass things like body language, facial expression, voice tonality, and eye contact. The facial action coding system is one example of how scientists have begun to codify these types of auras. Less studied, but in my experience still very valid, is the fact that one can change their subcommunications either through deliberate practice, or through changing internal beliefs and feelings.

I have periods where I have a "aura" that lets people on the street speak to me and other where that doesn't happen.

I remember once that in such a period a person was approaching me from behind while I was walking and reading on my Kindle. There seems to be a lot of information that goes beyond simply facial expression/voice tonality/naive conceptions of body-language.

Are you claiming that people were picking up on something other than the signals your body was giving off, (gait and posture would be the most obvious studied signals here) or simply that people don't often think about things like gait and posture when they talk about body language?

Some people claim to be able to see auras. At some point I got the vague impression that this was a ~real thing~, not in a supernatural way but in a "you can learn more than you realize about someone from looking at them, and in some people some of the things you can learn are represented as auras" way. (I may be completely wrong about that. I haven't seen anyone talk about them recently.)

So now I'm wondering whether that's the same thing we're talking about here. Would people who see auras be able to detect a niceness/meanness field? Would they see different auras in women who get frequently/rarely catcalled?

The claim that someone can see an aura is usually not only a claim about perceiving some information but also a claim of having a visual representation of it. To the extend that's there are people who really experience it, it's a form of synesthesia.

That seems true, but... I confess, I'm not sure why you're saying it. It feels like you're trying to tell me a relevant fact that I can use to answer my original question, or to approach it from another angle, or something. But if so, I'm failing to see how I can use it. Am I missing something?

My point is that the people who claim to see auras make a claim about having a visual perception of something.

Perceiving something as a "field of niceness" would be a more intuitive judgement. Whether or not the intuitive perception of a "field of niceness" is a real thing is a distinct question of whether someone actually has a visual perception of an aura.

Right, thanks. To be clear, I'm not suggesting that aura-viewers might perceive a niceness field *as* a niceness field; just that there might be some visual quality of an aura that suggests a niceness field. (I guess similar to how someone can look happy, even though I can't perceive happiness directly.)

(cross-posted from the SSC comment thread)

People who have close friends with a wide range of “fields” (i.e. behaviors that they unconsciously evoke in others): do you observe differences in the behavior? Is there anything you notice that could be replicated to achieve a desired effect?

I think part of this reply touches on your question: http://jadagul.tumblr.com/post/166000094778/how-social-bubbles-are-made (section "acting with intention", but I recommend the whole thing).