tl;dr: Sometimes it seems like in order to accomplish something, you need to hold a particular belief. However, the effect of your beliefs on what you accomplish can be screened off from what you actually do.

Also, thank you to Benquo for reading over a rough draft of this and providing very helpful comments.

Foma: Beliefs that Cause Themselves to be True

Live by the foma [harmless untruths] that make you brave and kind and healthy and happy ~ Cat's Cradle

When I was younger, I had formed an idea that there were some beliefs that, when believed, caused themselves to be true. I even had a name picked out for them – foma. These are just a few examples of how I came to think that.

“This is awkward” very often makes things awkward.

Consider walking through a room with a group of people that you don't know very well all talking and laughing. One or two look at you, and you just sort of stare back. “Well,” you think, “this is awkward”.

You stare blankly before letting out an uneasy laugh, and you go on your way.  You can feel people watching you walk out the door.

If you just walk through the room without thinking about it at all, its not even emotionally salient enough for you to wonder how it feels.

When I got over my fear of public speaking, it was basically because of a fluke. I decided to do a presentation on the mistakes of Odysseus' crew in character as Odysseus. People then assumed that my shaky arms, legs, and voice were the result of me doing a good portrayal of a shaken Odysseus, rather than my being nervous.

After that, I thought public speaking wasn't so hard as long as I feel comfortable doing it. Taking a few steps to mitigate my physical signs of nervousness (like walking around, or standing behind a podium), I quickly became pretty comfortable doing it.

“I'm not a good public speaker” worsened my public speaking skills, and “I can do this” strengthened them. Areas like self-confidence seem to possibly be foamy.

However on closer reflection, that model is incomplete.

Anticipations Influence Action

Clearly, there are beliefs that don't cause themselves to be true. Foma that work in some instances don't work in others. If I thought I that I was such a great  speaker that I could go in front of a group, stare at the ground, and then stutter into some note cards while mumbling offensive things to the audience and have them like it, I'd be wrong. If I even just thought I was going to do a good presentation and then didn't do anything, I'd be wrong.

A belief alone isn't actually enough to do anything. There needs to be a causal reason for your holding a belief to influence the world. Your beliefs influence what you do, and what you do influences the rest of the world.

Some religious people argue that their belief in God allows them to be a good person. As we know, you can be a good person without believing in God. Controlling for what you do, and with tight enough definitions on "what you do", your beliefs are effectively screened off from the rest of the world unless you're being brain scanned or something.

A causal diagram can be drawn as such:

Alice believes in X → Alice's anticipations based on X make her choose to do Y → Y has effect Z

Consider Alice believes that she is funny → Alice's expectations of delivering a funny joke leads her to deliver a joke well  → Alice's good joke delivery makes Bob think that she's funny

In this diagram, if Alice just does Y her belief in X is screened off from effect Z.

Taking apart Foma

People's brains don't normally think particularly rigorously. When language combines two different things into one or frames something as an intrinsic property of an object ("Alice thinks Alice is funny" and "Bob thinks Alice is funny" often becomes "Alice is funny"), it can seem like foma occur. On top of that, we often act to fulfill or preserve our self image (remember Bruce?).

It's easy for Alice to think that her belief that she's funny is causing her to be funny. If she were more precise, she could get away with:

Alice believes that actions Y have effect Z → Alice does Y → Y has effect Z

With specifics,

Alice knows that people find a joke funny when its set up right and has good timing → Alice sets up the joke correctly and has good timing → Bob finds the joke funny

Every step of that chain is entirely true.

When Foma are Practical

Part of the reason that I took so long to take apart foma was probably that, in the areas that I experienced foma, I wasn't consciously processing my actions. The awkward things that I did while I thought I was being awkward were never part of an intentional strategy, they just sort of happened. Since changing my beliefs seemed to do something, it felt like they had a direct causal effect on the world.

With the unpacking of foma in mind however, it's easier to discover which behaviors are actually influencing the world. When you feel like you're experiencing foma, you have an opportunity to learn something about what you do or don't do successfully.

For instance, when I feel awkward I'll start speaking when I feel that someone made a pause in conversation, but then stop and let someone else to speak, then continually almost interrupt them.

When I feel like I'm funny, I'll extend pauses after particularly emphasized parts of a joke, and slightly vary my volume and speed of talking based on relevance to the punchline.

In some instances, it probably is easier to believe in foma than to act on the relevant beliefs. As of right now, my unconscious mind knows much more about how to be confident than my conscious mind does, and on top of that it has much better processing power with which to act on its knowledge, and keep track of other people's responses. It runs more automatically, and continues to deliver while I'm consciously distracted. When I need to act confident, I find it to be much more time efficient and effective to just "psych myself up" than to review everything I know about body language and whatnot.



New Comment
39 comments, sorted by Click to highlight new comments since: Today at 3:48 PM

I'm in the process of trying to unpack the process of making friends, which is kind of mysterious to me. My old implicit model was:

I like someone -> We (sometimes) become friends

Of course my feelings don't directly cause anything but my actions. The simple insight that I was attributing a consequence to something that could only cause it indirectly, gave me the opportunity to notice some of the mediating causes, and fine-tune them separately:

I like someone -> I behave pleasantly toward them -> Increase P(Friendship)

I like someone -> I invite them to things or accept their invitations -> Increase P(Friendship)

I like someone -> I ask them questions about things that interest them -> Increase P(Friendship)

Now if I want to be better friends with someone, instead of just uselessly liking them more, I can do things like strike up more conversations, or invite them to hang out. There's lots of stuff I'm missing, but finding mediating causes is a powerful tool.


Well, merely liking someone does increase P(Friendship), in that not liking someone presumably decreases it. Usually you don't become friends with people you don't like.

That's the whole point though. Merely liking someone doesn't do anything, but cause the behavior/actions that actually have an effect (pleasantries, invitations, questions). If I didn't like the person, but did everything else the same, we would be just as good of friends. I can like someone to death, but unless I show it in some way, it's useless for forming a friendship.

But only in the form that attitudes affect your actions, and these affect your friendship, as mentioned above.

It's like a wossname... markov blanket.

This is related to Schelling's theory of focal points only there it's necessary for a group of people to believe something (and more importantly to believe that everyone else believes it) for it to become true.

Interesting, I didn't make the connection to this before, but was going to talk about something similar in the next thing I write up, assuming I get around to it.

This seems closely related to performative sentences, which are sentences whose utterance causes them to become true, and which are usually marked with "hereby".

For example:

"I name this ship the Nausicaa"

"I now pronounce you husband and wife"

"Well, I say!"

Well, 'foma' as this article and Vonnegut use it, is about things where believing in them has some useful effect, even if they are not true in themselves (yet, or ever). I guess society is based on a lot of such things -- e.g. warrior-ethos in the olden days.

Self-help aside, the most basic foma is "I believe this sentence".

It's equally coherent for you to believe or disbelieve that sentence, and the uttering doesn't seem to be related to its truth-value.

I would suggest something more like, "This sentence has been written."

Another nice one is "Bongo cannot believe this sentence".

I think he can get round the problem by saying that he can, if he chooses, believe it but chooses not to.

This doesn't solve the problem that he's disbelieving an obviously true sentence, though.

No, its false. It asserts that he cannot believe it, whereas we are saying that he can. Just because he chooses not to doesn't mean he can't (I choose not the jump out the first story window next to me, but this doesn't mean I can't).

Its the same trick as resolving the "everything I say is false" paradox, some things I say are false, some are true, and that sentence was the first kind.

Counterfactuals are fun :-) How do you know that he can believe it? Just because he asserts so? It seems to me that he cannot.

Why not? Sure, it would be a false belief, but people are capable of holding those.

Hm, good point. I guess we can close that loophole by saying something like "Bongo cannot believe this sentence and stay consistent".

Or "Bongo doesn't believe this sentence".

We should really start a site dedicated to hacking social life and reducing awkwardness. It seems like very many people around here have problems at social interaction (for example, consider this thread).

Perhaps a subreddit once we get those. I for one would like to see less of these self-helpy posts on the main lw and even discussion, not least because they kind of make lw look bad.

they kind of make lw look bad.

Could you expand on this point? How/why?

I think I get this point. Suppose an intelligent extrovert steps by. He/she will not be able to usethe social advice in any way, as it's just obvious for him/her. Instead, he/she might think that we are a bunch of nerds who are claiming to attempt to be more rational, but just have problems with social interaction. He/she leaves.

Well, that might be quite improbable, but it is plausible that similar things may indeed "make lw look bad".

However, I'd claim that creating a subreddit or something similar is a good thing for (1) getting this stuff away from lw and (2) exploring the problem of bad social skills more thoroughly. Here, those posts are quickly overwhelmed by other matters such as meetups, cryonics and the other stuff.



Well, before saying what the lw community should do, one needs to figure out what lw is supposed to be. If it is about pushing the boundaries of knowledge, akin to a scientific journal, then we should not be held accountable to those who don't "get" it any more than a mathematical publication is.And if this is true then lw needs to decide what standards it enforces, that is, whether or not social interaction instruction is a worthwhile publication. However, if it is about the lesswrong community improving their ability to "think and decide more successfully", as it says on the About page, then one should consider how certain groups would respond to posts. And if it is the latter, one should be careful about replying "why pander to them" to critiques, as this response can be used generally against many disenfranchised group that may have a legitimate complaint. Misogynistic? Why pander to overly sensitive people who can't take it like it is. You don't buy into the singularity/transhumanism/cryonics/atheism? Well, I'm not going to pander to your inferior intellect.

Personally, I think TrE did an excellent job describing the impression this site can sometimes give. I really enjoy reading this site, and the posts on epistemology are some of the most influential things I have ever read. And I know that many people on this site lack the social graces that most people have. And this does not necessarily affect their ability to write on the types of topics I like to read about. I also know that there are very fundamental tasks, making friends, getting dates, that are obviously important for people to learn to do. But much of the population, as well as myself, may not glean a lot of insight from this type of instruction. One could even turn the logic around and say, why pander to the socially incompetent?

Since I know lw, I can skip the occasional article that doesn't strike my fancy, but as a newcomer I'm not sure if I would have stuck around. Some of the discussions about socializing, especially the whole PUA episode, really made me wonder, is this my tribe? Are these the people I should be learning from? People who cannot do, in my mind, very rudimentary tasks that illiterates in Appalachia can excel at? And I really do wonder that sometimes. I feel that lw is full of incredibly intelligent people who in their real lives aren't actually "winning." And I know that is what this site is supposed to improve, of course. But I don't feel that the current level of lw elitism is really justified when many people can only theoretically "win."



If someone on this site did not understand basic algebra, and wanted to talk about it, would you think that would be a valid lw discussion? You write that "Any smart extrovert that I would want to know would not be so easily turned off by social difficulty", but what would you think of a rationality community that had to teach its members basic algebra? And these individuals may be trying very hard to understand it, but they still really struggle. Would you be turned off by their mathematical difficulty?

But algebra is SO EASY, one might say. To many, social skills are easy and math is hard. And I'm not saying that "easy" things should not be focused on on lw, only that lw content is generally "hard", with the conspicuous example of "easy" social skills. (easy meaning much of the general population can do it) Currently the lesswrong community focuses less on algebra and more on social skills because that is the skill set the community needs, but that focus in turn influences what the community becomes. If we accept the current composition of the lw community, which I would warn against as I think it is too homogeneous, then sure, we can deal with the existing needs of the community, ie teaching social skils. But if we are trying to foster exclusively discussions of high level winning, and if we would reject a discussion on algebra as being too easy, we should similarly reject basic discussions on social skills.

Figuring out how to explain math to people who aren't naturally good at it might be a valuable topic for the site.



Maybe the more pressing problem is not making LW look bad, but that this may not be a particularly effective way to learn social skills.

The easiest way to learn pattern of behavior X is to spend lots of time around people who exhibit a high level of X.

(Social skills, or character traits like generosity or patience, recovery from addictive behaviors, etc.)

It's true that you probably won't learn social skills solely by reading a text in an empty room and then leave it, having fully assimilated the described social skills.

Beyond that -- and not to single you out, of course -- your comments present what I have found to be a widespread, counterproductive misconception. What I have found, instead, is that:

1) Believing that learning a skill you possess requires extensive experience, is the quickest way to "compartmentalize" and weaken your own understanding of the skill, and dull your ability to pass it on to others. If you start from the premise that it's all an inarticulable black block, you will completely miss out on the parts that can be verbally communicated. I have seen this all the time in instructors who lament that they can't just tell me how to do something, and at the end I find that "er, you could have just told me all along that ...".

2) Spending time around others who exhibit a high level of X will do very little for your skill at X. Everything that was a mystery will remain so, because you won't see the "model" that led the expert to act one way rather than another, and people woefully overestimate their ability to "infer" the "algorithm" that the expert is using. (This is very much related to the underdetermination problem.) Time and again, I've apprenticed with others and learned precisely nothing, while I've trained others up to my level in a fraction of the time I required to learn it, or of the time expected for the person to learn the skill.

3) Much experiential learning -- not all, of course -- can be obviated by a relatively small amount of verbal instruction, because it singles out the non-obvious, hard-to-experientially-infer part of the problemspace. Again, with the instructors I've been involved with (and contrasted with my role as an instructor), I could make no progress learning alongside them until I could verbalize the skill -- which typically reveals holes in the instructor's own understanding!

This has all led me to strongly suspect that people who fall back on an inarticulability defense, typically lack understanding in crucial ways themselves.

Recently, I wrote an introduction to asymmetric cryptography that was lauded as far more helpful than anything else available on the matter, even despite covering less material. My trick? Actually have explicit understanding of the matter, not just the "learn by watching" kind.

I should have been more specific.

I think you are right, pertaining to purely intellectual topics such as asymmetric cryptography.

But with social interaction where most of the stuff goes on under the conscious level and we have lots of built-in heuristics, I think being around people who are good at skill X is very useful, as long as you observe them and frequently ask, how did you do that?

Also trying it yourself and having them critique afterwards. A mix of theoretical instruction and actual practice seems to be what teaches my unconscious how to do things, from social skills to habits to sports.

Actually have explicit understanding of the matter, not just the "learn by watching" kind.

Oddly, I tend to get the exact same "that was so helpful!" response, despite generally having a knowledge base that is almost entirely "learn by watching" and "well, this works but I have no clue why". It seems to help that I'm very good at making this clear.

Obviously, this only works if I'm teaching someone who can get by on that level of understanding; teaching people to understand something better than I do is tricky ;)

I really like this comment, especially 1).

3) Much experiential learning -- not all, of course -- can be obviated by a relatively small amount of verbal instruction, because it singles out the non-obvious, hard-to-experientially-infer part of the problemspace. Again, with the instructors I've been involved with (and contrasted with my role as an instructor), I could make no progress learning alongside them until I could verbalize the skill -- which typically reveals holes in the instructor's own understanding!

That's very true. I have had trouble with being shown "how to do something" by someone, but not understanding a part of it and asking them to explain it in words. I hadn't realized, until now, why this should be more helpful. It was sometimes frustrating for both of us when they realized that couldn't verbalize the process (until after some thought, hopefully).

I believe that this is also why teaching something is the best way to learn it. I, at least, have found this to be true; I know simple chemistry and algebra extremely well from having to explain them often. When I first tried to do so (thanks sab!), I often encountered places where I either didn't know how to proceed, or did know but couldn't say why it should be so at first.

Isn't the purpose of this blog self-help, i.e. learning to become a better thinker and a more rational actor? Yes, there has been lots of domain-specific advice that maybe would be better placed elsewhere, but I didn't think this post in particular had that failing. It seems like this is a somewhat more general insight.

It is indeed a general insight, and I don't criticize this post in particular. In fact, I upvoted it. However, it seems that many posts on this site deal especially with social awkwardness / failures in social life, so creating a separate site may be a good idea. I could have posted this in the "socially awkward penguin" thread as well (see above for link), but this was the most recent post available for such a remark.