Edit - Barring a major surprise, this post should be regarded as a worthless artifact of my impulse to do things instead of talking about them.  I apologize for any time wasted on this, and would recommend ignoring it unless it is for historical purposes.  I'll just stick to things I'm less bad at from now on.


This article will be edited as people post and discuss.  

I believe that we need to have a clear, concise statement about the beliefs, practices, and taboos that it is rational to hold, and that we already hold as a group.  To be clear, this is not an attempt to make new norms, but an attempt to codify the ones that we already hold and to get a rough estimate of the popularity/importance of each.

Core Rational - skills, meta-beliefs, and habits that enhance personal rationality

Social Rational - norms that enhance working in groups rationally

LessWrong Norms - norms for dealing with Less Wrong specifically

Common Knowledge - basic, useful beliefs to build on

Please post one phrase at a time and then give your reasoning under it.  Once any idea has a common consensus, I'll add it to this article in the appropriate list.  

Edited - Removed the word 'should' as someone has suggested a better phrasing.  Edited again - category change, remove extra now-useless examples.

New Comment
71 comments, sorted by Click to highlight new comments since:
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

We should believe that our beliefs, taboos and practises are subject to change based on future evidence and circumstance, much of which is unknown, or even unknowable, at this time.

We also suppose that codifying them will add both a cognitive and administrative overhead to carrying out such changes.

It may be useful to codify beliefs in particular as "what we believed in (May of) 2011" rather than "what we believe". That seems to me like it will cut down on a lot of the overhead - there's much less implication that someone should believe any particular thing on the list, and it's more obvious that 'so what?' or 'yes, because...' are good responses to 'you don't believe [particular thing] any more!'. I'm not sure that beliefs are an appropriate thing to have in a list of LW norms in the first place, though. There are some obvious ones that seem to belong here, like the belief that believing true things is generally likely to be helpful, but other than that, well, changing beliefs is what we're about; our norms should be something a bit more meta than that.
That's an interesting point, actually. If it were a descriptive exercise rather than a normative one, it could have merit as a historical record.
Perhaps add a version number with a datestamp? I took the categories of the norms from this post you made earlier. The 'updating based on evidence' seemed more like a skill than a belief, and the other two categories you mention explicitly. If they end up not being a useful division, then they'll end up getting changed. I don't have any investment in them.
Eh? I mentioned updating beliefs based on evidence, and acting based on one's beliefs, but not having particular beliefs in general. The last bit, about the list of common beliefs, was meant to be along the lines of "If you believe A, you should do B. If you believe C, you should do D. If you believe E...", not a list of endorsed beliefs. Also, rationalist taboo is not about taboos in the usual sense.
Oh, so it isn't. Oops. Hm. Norms do usually contain taboos, but there isn't any particular reason that we have to. Should that category be deleted, or do we have things that we think should be taboo? One that springs to mind as a possibility would be the use of the Dark Arts. edit - I was thinking that our list of endorsed beliefs would be slightly more basic, things like 'rationality is worth pursuing' or something. http://lds.org/library/display/0,4945,106-1-2-1,FF.html is the small amount of norms that the LDS church has codified as an easy reference to point potential converts to, and while I don't like the idea of converting people, being able to point to a set of beliefs that sounds boringly sane might be a good thing.
Codifying taboos seems reasonable if we're codifying norms in general - taboos are basically norms of not doing certain things, after all. We seem to have a pretty strong norm of not lying to each other, and a weaker (and possibly not generally endorsed) norm of not talking about any religious or spiritual practices that we might have.
Not lying to each other sounds like a very good one.
Both of these seem true. Making our norms rigid will decrease our ability to adapt, and codifying them will make them harder to change. If the group consensus ends up determining that this isn't a worthwhile endeavor, then I'll delete the article and if anyone wishes to contact me I'll apologize for the wasted time involved.

Of course the norm would be "don't reply to words with violence", not "it is irrational to reply to words with violence".

By the same token, taboos are not actions that it is irrational to take, taboos are either actions that it's bad to take (prescriptively) or actions that it's frowned upon to take (descriptively).

Rationality is not generic virtuousness and irrationality generic sinfulness. "Rational" is not a label you slap on something when you want to endorse it. At least it shouldn't be ;)

Note that arguments are different to words. With respect to the more general norm of responding to words we already inherit that norm via the laws of most countries where we reside. In those countries where such a norm is not enforced following the norm is quite possibly a bad idea. There are (and even more so have been) places where violence is an acceptable and even necessary response to certain types of words. This perhaps even applies in certain subcultures within the more developed countries. I don't have a problem with that - it is their culture, not mine. Because we don't have the ability to punch each other in the face over the internet and only meet in person in the location where we physically exist it is the norms of that location that are relevant. Such a norm does not belong here.
Eh, I could see us having a norm of thinking poorly of people who are known to have punched third parties in the face, which is how I was parsing that one. I'm mostly nitpicking, though.
I personally would be thinking poorly of people who are known to advocate a norm of thinking poorly of those known to have punched third parties in the face. Primarily because this isn't church but also because it encourages a short-sighted and incomplete understanding of social dynamics and ethics.
Still nitpicking here, but this seems to me to be related to the most important difference between norms and rules: Norms apply to normal situations and leave room for personal judgement; rules apply to all situations and generally don't. I expect that it is normally not a good thing to punch someone in the face - that something more than 60% of a random sampling of non-consensual face-punches would generally be considered on reflection to have been a bad idea. If that's true (and I'll grant that I'm not terribly confident of it), then having a norm of reacting to the information that a given person has punched someone in the face as if that's indicative of a significant lapse in judgement seems reasonable. If there are extenuating circumstances, then that can be taken into account, and it's not 'breaking a rule' to do so and conclude that the normal response is inappropriate. It doesn't even disprove the usefulness of the norm, whereas finding a situation where a rule leads to poor results can be taken to prove that the rule needs changing. (I don't necessarily advocate this specific norm here, though. I haven't thought enough about it to weigh in one way or the other.)
There is a difference between thinking that on average punching people in the face is a dumbass move and codifying a norm to the effect.
Expand? I'm not sure why you think that, but it's entirely possible that I'm just failing to notice something.
I am confused (perhaps?). One is a codified norm - the sort of thing that people will link to or cite. Another is an individual (but common) expectation that a given punch was a suboptimal move and somewhat of an indicator of undesired traits. They just seem... entirely different to me. ie. One is highly unpleasant and the other is blatantly obvious.
I'm still lost. We write things down and link to them and cite them all the time; I don't see why you think that's bad in this case. It does add the implication that the cited norm is generally agreed on, but if we've talked the norms over and only listed ones where that's the case that doesn't seem likely to be a problem - and, even in that case, citing a listed norm only seems to me to carry about equivalent conversational weight to someone posting "[behavior] is normal here; you should have expected it" and getting upvoted in agreement a couple of times. Plus, it gives new people an opportunity to learn those norms more quickly, and also allows us to explicitly not count certain things as norms even if they're common habits, if they're things we want to allow anyway. There is some context that might be notable, here: Alicorn and I happen to have been working on a house norms list for the last several days, since we decided to offer CronoDAS the opportunity to come stay with us for a while. That's been going quite well, and appears to have several advantages over a house rule list - the flexibility mentioned above is one, and the fact that norms can semi-contradict each other in ways that rules mostly can't is another, for example. So I have some positive affect built up around the concept, where I suspect that you have some negative affect around it because of how the concept came up here.
That is something that I do when I live in shared accommodation as well.
Huh. You probably have even more experience with it than I do, then. What about that experience makes you think it won't work here, since it's apparently something you find useful enough at home (given that you 'do', rather than 'have done')?
Implemented, thank you for your input.

Let's try codify our norms.

Fuck that!

The above seems to be the most natural translation of my instinctive evaluation of the suggestion. I expect such a process of codification to have a negative effect on the community. Any reference to such a code would be an obnoxious interruption for reasons similar to those Eliezer recently mentioned.

Codify concepts not norms.

I'll be happy to change the title if you have a better suggestion.
I don't mind the title and don't even object to your post - I didn't downvote. I just disagree with you about the value of formal codification of the type you are advocating. My vehement rejection is to your idea, not your expression thereof.
I apologize for wasting your time.
Don't. There is an example. At my dojo we have a norm (courtesy of me) wherein anyone who apologizes unnecessarily gets hit (in the mild sparring sense).
Can you unpack what you think makes apology (un)necessary?
This particular apology was unnecessary because I clearly wasn't offended. It conveys a frame wherein I was objecting to my time being wasted, as opposed to actually engaging with the topic with a disagreement - distracting from my actual point. A net negative message.
I suppose I should be looking forward to getting hit, then. I thank you for the warning.
I believe you may be mis-parsing the statement. Apologising when you've done nothing wrong is a signal that you consider yourself inferior. This attitude makes ones' life worse, in general, by reinforcing bullying behaviour. Even people who won't bully will be less likely to want to hang out with you, because being around someone who does not feel comfortable is unlikely to be fun, and speaking personally, feeling the (metaphorical) urge to kick the puppy makes my skin crawl.
Yes, I'm aware of this. I didn't mis-parse his statement, I know what he meant, and what his example of a norm was supposed to provoke in me. It doesn't matter. The subject matter of the posts that led to me posting this article and my memories apparently affect me more than I had thought they would, which in its own way shows that I am indeed an inferior rationalist. Truth is truth, and denying it does no one any good. From now on I'll just concern myself with the local group here, or not, depending on what happens over the next few weeks. In any case, I can let this account lie fallow or delete it. Which would disrupt the chain of comments less? edit - I do apologize for how you have ended up feeling, as that was an unintended consequence.
The only way we become better at anything is through practice. I'm a pretty shitty instrumental rationalist, as a 27 year old with no great skills or qualifications, and very patchy work experience, but I aspire to better. If you enjoy participating in the Berkeley rationalist community, keep doing it. Consider continuing to participate here, your posts, Mitigating Social Awkwardness and Insufficiently Awesome mean you have definitively contributed more here than I have. FWIW you didn't provoke any unfortunate emotions in me; that's almost exclusively something that happens in person.
You do? Roughly it was supposed to provoke a short term frustration but a net reduction in sensitivity to the real or (more commonly) imagined effects you may have on the internal state of other agents. Also a somewhat reduced impulse to publicly distance yourself from your own (minor) actions and instinctively withdraw. For what it is worth people would prefer you to continue engaging with the community here. If you don't think you would personally benefit from participating here then by all means move along. But for crying out loud, you don't need to second guess what other people want you to do. Too much tiptoeing about like you're walking on eggshells is perhaps one of the worst things excessive exposure to religion results in.
Just as soon as you sign up to a suitable martial arts club. I encourage it, the training generalises well. :)

This is a terrible idea.

I would welcome alternatives. Do you have any suggestions?
Not prescribing the beliefs, taboos and practices of LW members.
Our norms exist already, they're just unwritten right now. There are things that we can do that will cause everyone to shun us. (Posting a discussion like this may end up being one of them, in which case I will have learned something valuable about Less Wrong...)
I wouldn't have a problem with describing the norms that currently happen to hold among LW users, but I do with prescribing them. EDIT: to elaborate, there's a difference between following a norm, and thinking that it's the best norm and that it deserves to be codified and officially explicitly endorsed.
I don't follow. What is your actual reasoning as to why this is a universally bad thing?
Agreed, and on closer reading the OP seems to need to be reworded to reflect this.
Edited to try to make this clearer. I may still need to alter the phrasing more to make it less offensive, and I welcome all suggestions.
*looks* The intentions of the various categories seem to have changed considerably in this version, which suggests that maybe we need to talk about what we want this list of norms to do before we figure out what subcategories it needs. To that end, I see three obvious goals. In rough order of importance: 1. Codify skills, habits, and meta-beliefs that will help people be rational, e.g. rationalist taboo. 2. Codify norms of this group, to make it easy for people to join up - rather like Silas' 'signs about how things are done here' idea. 3. List useful, basically-settled beliefs for people to build on, e.g. Ocham's Razor or Bayseanism.
Would it be better to categorize them by goal, then? That would suggest three levels of norms: core rational, social rational, and common knowledge.
The term "social rational" sounds like it would be used for core-type skills used for working in groups. "LW-specific norms" might work better for that one. Otherwise, yeah, sounds good to me.
Describing them is my goal. The only way that I can think of to get a complete list is to ask everyone to post the ones that they feel exist, and then to see what the consensus is. If you have a better method, please let me know so that I can use it instead.
Implemented, thank you for your input.

Pretty as a page full of detached levers might look, I'm not keen.

Things don't magically stop working just because they're written down somewhere.

I believe that we need to have a clear, concise statement about the beliefs, practices, and taboos that it is rational to hold, and that we already hold as a group.

We don't know what beliefs, practices and taboos it is rational to hold. They're probably not the same ones that we already hold as a group. Thus we won't discover them by just investigating the ones we currently have.

Decide whether you want to list what norms LW has, or to investigate what norms it would be "rational" to have (a loftier project to say the least).

I'd like to know what we have now. Really if we were going to make a new, 100% more rational version, we would still need to know what we're starting with. That said, this is probably just a sunk cost now, and not worth contributing to. I'll just concentrate on things I'm less bad at, and ignore my step-forward impulses from now on. Edit - please disregard this post
Step-forward impulses are an awesome thing to have, something we "should" all cultivate. Don't apologize for them. As far as your being bad at this, I am skeptical and suspect you are backing down too quickly. At worst this is an opportunity for people who are opposed to your approach to explain (and perhaps figure out more clearly for themselves) why it is a bad idea.
I have a policy of only doing fun things, or things that lead directly to fun. As soon as something becomes non-fun, I find something else to do. The potential benefits of this, and the use I would get out of them, aren't worth the cost to me or the time others have sunk into it. In a lot of ways I already live in the least convenient possible world, so I'm just going to assume that the result I would have gotten was the one that would have been the worst. I won't waste anyone else's time with this anymore, so the whole effort is moot. If someone wishes to try to do this in a more adept fashion, then I wish them well. Apologies for wasting your time with this. Edit - please disregard this post

Can you explain why this is a very high priority for you?

Any phrase beginning with "we should believe that" seems outright irrational. Rational beliefs are grow from evidence, not moral reasoning of any kind.

  • Instead of saying "we should believe that the Earth is round because it looks round when seen from the space", just say "the Earth is round because it looks round when seen from the space". The latter sentence is clearer and doesn't lack anything important which the former has.
  • If you rather had in mind something like
... (read more)
Norms really are a 'should' type of thinking. I don't like using 'should' in any capacity, because it sounds like I'm telling someone what to do, but in this case that's exactly what norms do. It's a high priority because of the recent posts suggesting that we adopt select practices from religions. I want to know now if I should walk away, and if any efforts I'm in the process of making are just sunk costs already. Before we start adopting things from other groups, we need to have something to compare them to so that we can make sure that there aren't hidden conflicts. In general, isn't writing things down a way to avoid or expose biases?
I am also worried about adopting religious practices. But to ascertain whether it happens or not, it seems easier to ask directly: "do we want to adopt this set of practices?" rather than to ask "what norms we should adopt". I would even be afraid that the simple fact of having explicit norms, especially norms on beliefs, would move the community closer to the realm of religions. I really don't want to be told that I should support cryonics, believe in many-worlds QM and be an atheist, or leave LW for good (and if there is no "or leave" or analogical punishment for norm violation, why call that norms?). I prefer when people spread beliefs by argument, not by social pressure.
With the reception that this article has gotten so far, I suspect that it won't result in a list of extant norms. I'll give it a day or two so that everyone that wants to can weigh in, and then I'll probably end up deleting it,.
I would prefer if you don't delete it. As for now, you aren't losing any karma from it, and there are already several comments. I am often frustrated when comments with replies get deleted and you can read the replies while the context is inaccessible. I don't know what happens when a post with comments is deleted. Moreover, even if the suggestion isn't agreed upon, its existence at least makes clear what it is not agreed upon, and saves effort of future readers who may propose the same.
I would suspect that the whole thing including comments all vanishes, but I haven't tested it yet. edit - I really don't care about karma, the only use it seems to have is voting people down and being able to post on the front page, and I doubt I'll ever do either of those things. I'll happily let other people be the top contributors.
The post disappears from the list of posts, and probably can't be found by searching, but it still exists and can be linked to and commented on, and any comments on it still appear in the new comments feed. I think it also can't be voted on after being deleted, but I wouldn't swear to it.
How refreshingly counterintuitive. (-_-)
The latter sentence parses as either malformed or false, to me. The earth appears to be round from space because it's round, not vice versa; the earth is round because of the forces that were involved in its creation. I do agree that the 'should' needs to go, but I think the formulation should look something like 'it is rational to believe [thing] because [evidence]' or 'I/we believe [thing] because [evidence]'.
Fair enough. "Because" itself isn't perfectly transparent word: X because Y may mean that * (etymologically) Y is a cause of X: "I have been arrested because I have robbed a bank." * Y is a purpose for X: "I robbed the bank because I wanted the money." * X is logically deducible from Y: The apple fell down because of the laws of gravity. * (and I believed that also) X is probabilistically deducible from Y. I have used the word "because" in this sense, as a shorthand for "and the evidence for the previous claim is that", which after all may be ungrammatical. In any case, should is redundand.
This is a good way to formulate it. I'll implement this now.

Be very careful what you assert, even if it's true. It's really easy to start with good ideas and end up with a doctrine.

If you make a statement like "we should believe cryonics is desirable" or even something as basic (for LW) as "we believe in reductionism", it's really easy to exclude or piss off all the people we actually want to reach.

A very minimal set of skepticism might work. Something like RAW's advice in Prometheus Rising: "Avoid coming to any strong conclusions prematurely. [...] Believe it possible that you do not know everything yet, and that you might have something still to learn."

I meant those as examples only. I would welcome suggestions on less-offensive alternatives.

We should encourage and support self-experimentation by our members.

Unless the experiment is obviously harming the experimenter, encouraging this will help us find more efficient ways of doing things. I think that respecting a fellow rationalist's decision is a way of respecting their rationality as well.

We should have a clear and concise list of our norms.

This will give us a clearer picture of what is expected of us, by us.
'Social norms are the behavioral expectations and cues within a society or group. This sociological term has been defined as "the rules that a group uses for appropriate and inappropriate values, beliefs, attitudes and behaviors. These rules may be explicit or implicit.' - http://en.wikipedia.org/wiki/Norm_(sociology)

Without explicit norms, we invite debates over exactly what a 'real' rationalist should believe, and whether or not ... (read more)

Doesn't actually happen all that much, to be honest.
No, it doesn't.
It doesn't have to happen often to cumulatively waste more time than this process will take. Another reason to do this is that it will give us something very easy to point people to when they ask us 'what is a rationalist, anyway?'