In conversations on LessWrong you may be surprised (in fact, dismayed) to find an apparent majority of the community agreeing with each other, and disagreeing with some view you hold dear. You may be tempted to call "groupthink". Whenever that happens, please hold yourself to at least as high an epistemic standard as the people who are participating in the community, and substantiate your accusation of groupthink with actual evidence and analysis.

"Groupthink" can be an instance of applause lights, terms or explanations used not so much for their semantic content as for the warm fuzzies they are intended to trigger in your audience. Or... since "groupthink" isn't so much intended to generate applause for you, but to generate disapproval of those who disagree with you, we might coin the phrase "boo lights".

At any rate, you may be cheaply establishing (in your own eyes and the eyes of people "on your side") your status as a skeptic, without actually doing any critical thinking or even basic due diligence. Are you sure you that's what you want?

(N.B. links in this post either point to examples, or to more complete definitions of the concepts referenced; they are intended as supplementary material and this post stands on its own, you can ignore the links on a first read-through.)

Apparent consensus is not sufficient grounds for suspecting groupthink, because the "groupthink" explanatory scheme leads to further predictions than the mere appearance of consensus. For instance, groupthink results in "selection bias in collecting information" (from the Wikipedia entry). If the community has shown diligence in seeking contrary information, and yet has not rallied to your favored point of view, your accusations of groupthink are unjustified.

Disapproval of your contributions (in the form of downvoting) is not sufficient grounds for suspecting groupthink. Communities establish mechanisms of defence against disruption, in a legitimate response to a context of discourse where disruption is an ever present threat, the flip side of open participation. The voting/karma system is the current mechanism, probably flawed and probably better than nothing. Downvotes signal "we would like to see fewer comments like this one". The appropriate thing to do if you receive downvotes and you're neither a troll nor a crackpot is to simply seek feedback: ask what's wrong. Complaining only makes things worse. Complaining that the community is exhibiting censorship or groupthink makes things much worse.

Disapproval of your accusations of groupthink is still not sufficient grounds for suspecting groupthink. This community is aware of information cascades and other effects leading to groupthink, discusses them openly, and strives to adopt countervailing norms. (Note that this post generalizes to further concepts, such as censorship. Downvotes are not censorship; they are a collaborative filtering mechanism, whereby readers are encouraged to skip over some content; that content is nevertheless preserved, visible to anyone who chooses to read it; censorship, i.e. banning, does occur but much more seldom than downvoting.)

Here is a good example of someone substantiating their accusations of groupthink by reference to the actual research on groupthink. Note how much more work this is.

If you're still thinking of calling "groupthink" without doing that work... or, perhaps, if you have already done so...

Please reconsider: your behaviour devalues the technical meaning of "groupthink", which this community does have a use for (as do other communities of sincere inquiry). We want the term groupthink to still be useful when we really need it - when we actually succumb to groupthink.

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

I've made the point that we tend to scrutinize posts/comments more carefully for flaws when they argue against beliefs that we hold, which results in subtly flawed arguments supporting a majority position being voted up despite being flawed, while a similarly flawed argument against a majority position will more likely be discovered as flawed and voted down. This results in the appearance that there are more valid arguments supporting the majority position than there really are, and drives away those who argue against the majority position because they think they are being treated unfairly.

Does anyone disagree with this, or think that Less Wrong voters have already adequately compensated for it? Or, if you think this is a real effect, but shouldn't be called groupthink, what is the right name for it?

7Morendil
Real effect but not full-blown groupthink. For instance, I wouldn't expect people on LW to be accused of "disloyalty". I've said once or twice that I din't think of myself primarily as rationalist, to no raised eyebrows. I have not always been aware of the technical meaning of groupthink. But even then I was careful to distinguish between group cohesion one one hand, and irrational commitment to failing courses of action (or poisoned beliefs) on the other. I suspect that we're not criticizing the CraigsList project as harshly as we should. I suggested as much, but there was little follow-up. This strikes me as counterproductive; such a project should be subjected to the harshest tests and launched only if it survives unscathed. The phrase "insider bias" comes to mind, but I'm not sure it captures what you mean either.
7Blueberry
I made one or two skeptical comments on that page, but I didn't criticize the project as much as it deserved. It seemed so obviously nugatory. I didn't want to spend the time and effort thinking about something that I wasn't at all interested in, especially when everyone else seemed so excited about it. I just figured the people who really care about it can do it if they want. There is a type of selection bias at work here, where people who aren't at all interested in something may not feel the desire to spend time fighting the tide, which then makes the tide appear stronger. There are probably also people who feel this way about cryonics. (Not me: I think it's awesome.)
0[anonymous]
There seems too much discussion on this site regarding definitions. I call them "word arguments" and tend to skip them when I sense their presence. It seems people like to categorize things, in which case a word argument is an attempt to write dictionaries so everything make sense. I much prefer to try and understand what the other person is attempting to say, perhaps even without some specific, technical word to describe it, and continue the discussion in profitable directions. In which case, the proper response to accusations of groupthink would be, "I disagree for reasons X and Y. Now to cover your othre points..." Or, in my estimation, even better, "Why does it seem like groupthink to you?" You don't even have to ask them for their definition of the word to understand what they want to tell you, so long as they spell it out. And you don't necessarily have to answer the accusation of groupthink, then, either, especially if their reasons are not contained in the definition of the word. Instead, you answer the points themselves: what the person was trying to convey in the first place.
5Paul Crowley
It's not I think that we don't suffer from confirmation bias - of course we do, and I doubt that we do properly compensate for it - it's that saying "groupthink" or "confirmation bias" without providing evidence doesn't advance the debate any, and seems to constitute no more than a way of sneering at people for not agreeing with you.
7Rain
There seems too much discussion on this site regarding definitions. I call them "word arguments" and tend to skip them when I sense their presence. It seems people like to categorize things, in which case a word argument is an attempt to write dictionaries so everything make sense. I much prefer to try and understand what the other person is attempting to say, perhaps even without some specific, technical word to describe it, and continue the discussion in profitable directions. In which case, the proper response to accusations of groupthink would be, "I disagree for reasons X and Y. Now to cover your other points..." Or, in my estimation, even better, "Why does it seem like groupthink to you?" You don't even have to ask them for their definition of the word to understand what they want to tell you, so long as they spell it out. And you don't necessarily have to answer the accusation of groupthink, then, either, especially if their stated reasons are not contained in the definition of the word. Instead, you answer the points themselves: what the person was trying to convey in the first place. And you might be able to suggest they use a better word or phrase, once you understand them.
0Wei Dai
Perhaps they are not trying to convince you that there is more bias than you think (which would require evidence), but just reminding you that it does exist and is probably not being adequately compensated for? If we agree that a problem exists, then we should welcome such reminders (at least until we find some other way to solve the problem).
5RobinZ
You can do that on a case-by-case basis, but the problem is that accusations of groupthink are too easy. Edit: Come to think of it, it wouldn't be a good boo-light if it wasn't easy to make the accusation.
4komponisto
If you're referring to what I think you are (can you perhaps point me to another instance where you've made this point?), I simply disagree that the respective arguments in question are "similarly flawed". One circumstance in which you can usually be safe in dismissing a minority argument is when the minority arguer has ignored the arguments already put forth for the majority position. If you want to be a contrarian, that's fine, but you can't expect to persuade the majority to your point of view without addressing the reasons they believe what they believe.
2Wei Dai
I came across the information that on the individual level, this is called belief bias: (ETA: A search of LW and OB shows that we've never discussed this particular bias before, at least not by name.)
2brazil84
I agree 100%, and it occurs to me that the problem somewhat feeds on itself since there is a natural temptation to downvote in retaliation; to upvote "allies"; and so forth.
1Nanani
It seems likely that, given limited time to read comments, any given member will stick to the more controversial (and therefore interesting) comment threads, leading to an apparent piling on when disagreement is expressed.

For the record, my views differ frequently from those on Less Wrong (I'm a frequentist, theist, platonist, etc.) but I think Less Wrong does not suffer from groupthink = censorship or repression of different views. In fact, it's amazing how little it does.

Less Wrong does have a lot of clustering of views, which is a very interesting phenomenon. I think there are reasons for why the views cluster together other than the simplest explanation that these views are all correct! (For example, considering the clustering of views that are neither correct nor incorrect, but represent preferences.)

7Aurini
"Less Wrong does have a lot of clustering of views, which is a very interesting phenomenon. I think there are reasons for why the views cluster together other than the simplest explanation that these views are all correct! " 90% of the time I agree with this assessment, but I suspect that there is a subtler selection-bias amongst the members determining the remaining 10%; there have been some suggestions in research literature that a biological basis for morality influences whether somebody becomes Liberal or Conservative (that awfully reported study from a year ago, for instance). LW and OB tend to attract those who are comfortable in an Academic environment, and (for instance) I've observed that University graduates tend to underestimate the utility of violence. There are probably a number of other traits that are positively correlated to Academics, but not necessarily to intelligence; unfortunately the nature of our discourse here usually bores people in, for instance, SpecOps*, the Arts, Engineering, and other fields known for requiring intelligence; this bias would be worth keeping in mind, but I don't see that LW could be 'fixed' to become more user friendly. *Average IQ of SpecOps soldiers is 120-140.
3Jack
About what exactly?? Do most people here identify as nominalists (about most abstract objects)?
2byrnema
I don't think that my 'platonian' beliefs are on-topic here on Less Wrong, but since they seem to form the foundation/substrate of my beliefs they leak through sometimes. (For whoever is down-voting me, I agree it was unwise of me to link to those old comments...)
3Jack
Really? Understanding and clarifying the concepts and categories we use to organize our beliefs seems pretty damn on topic compared cryonics, personal identity and creating a billion dollars out of nothing. Not to mention I'm pretty sure nominalism is wrong/bad so I'd really like to know if all the smart people here disagree with me. And if indeed we find overlapping views here we might find that a particular 'metaphysical' view (terrible term) is more conducive to thinking rationally, which would be interesting.
2arundelo
Data point: I think I'm what's called a conceptualist. I believe exactly this.

Everyone who has argued that the "Boo lights" post intends to make it more difficult to charge groupthink is in fact right.

I'm sticking my neck out and arguing that it is presently too easy (relative to what I believe is the standard most conducive to furthering this site's mission). Yes this could well be wrong.

Some of the evidence I base this judgement on is participation in other "communities of sincere inquiry", and in particular the one where my "secret identity" is established. There the words used differed ("dogma&... (read more)

Re: "Please reconsider: your behaviour devalues the technical meaning of "groupthink""

It sounds like an attempt to raise the barrier for groupthink accusations - if you have to do a lot of extra work documenting the evidence before expressing your opinion.

IMO, people's opinions are often of interest - even without detailed references - and realizing that doesn't devalue any technical meanings.

It is best to see groupthink accusations as a cheap warning mechanism, IMO. Like pain in other words. Some false alarms are OK - while a failure to notice where a problem is occurring can be more problematical.

6Morendil
I probably should have linked to this article, but I'm not done reading it yet. It makes somewhat similar points to yours. Yes these accusations can serve a useful purpose. The point of this post is that the next time a comment like this pops up, it can be dealt with in a shorter reply, "Downvoted for boo lights; you are calling groupthink on a discussion which is actually a sincere inquiry, either do your homework or move along."
3timtyler
Down-voting is an even less-informative kind of booing. At least a groupthink accusation explains what it thinks the problem is. Yes, maybe it would be nice to have a detailed analysis, but is it really desirable to encourage deliberate punishment of posters who are trying to point our biases for not going into more detail? It seems as though the probable effect of such a policy would be fewer groupthink accusations - and more unidentified groupthink.
4Morendil
I believe it's desirable, for the same reason we should discourage the boy who cries wolf. I also understand where you're coming from. The danger of escalation - increasing committment to a failing course of action - can be big. Maybe, if we knew the relevant parameters, we would want to tolerate more such accusations than we currently do, rather than fewer, given a big enough payoff in the rare cases where they'd turn out to be on the money.
1timtyler
Here we seem to be discussing the policy of down-voting the wolf-crier who doesn't come with recent video footage of the wolf. I figure that is reasonable if there is reasonable independent evidence of no wolf, or a history of false positives - but otherwise not.
5Morendil
Remember the context of the comment linked upthread: this is someone voicing accusations of groupthink in the "Survey of Anti-Cryonics Writing" post... the equivalent of crying wolf in the middle of a quiet evening in the sheepfold.
1timtyler
The post tries to make a general point, not confined to one particular comment. The "cryonics groupthink" accusation is not obviously wrong. IMO, it probably has some truth to it. Blogs naturally attract like-minded individuals. There are often major "yes men" effects at work - along with mindless copying of group ideas, eagerness to display badges of group membership - and so on. IMO, there is more to it than groupthink. Identification with your ego is ubiquitous in western society - and the expression of that in geeks tends to result in an interest in cryonics. Ultimately, it probably has something to do with traditional western religions being so farcical, while western secular institutions offer little guidance.

I am quite surprised that more Karma functions do not track both the Up Votes and the Down Votes and display a running tally of each.

On a relative basis I would be far more interested in reading a post that has had 1,000 Up Votes and 999 Down Votes, than one which has had a single Up Vote.

4Jack
Well I guess you could solve this problem by posting duplicate comments- one for positive karma and one for negative karma. But I feel like it would be distracting. ;-)

shit. :( i was hoping this article would tell me how i could be wrong in calling some of my friends susceptible to "groupthink", but all it's done is make it seem even moreso. not that the group is overtly conformist, but when you do something different, it feels like you stick out like a sore thumb.

really really good article. thank you for sharing it. i learned a lot. :)

2Kevin
I hope that we don't scare you off too soon, because there are lots of other good things to read on this site. http://lesswrong.com/lw/qk/that_alien_message/ is probably my favorite single post. I think that the best or at least easiest to read introduction to the values of this site is Eliezer's novella length work of fiction, Three Worlds Collide. http://lesswrong.com/lw/y4/three_worlds_collide_08/
2AngryParsley
I've got to warn you: this community holds comments to a higher standard than you might be used to. I suggest you take a look at some of the other comments to get an idea of what is expected. With that out of the way, welcome to Less Wrong. I hope you like it here. Since you're new, you probably want to check out the sequences and the about page. Edit: By "check out" I meant "click on titles you find interesting," not "read all of these."
9Jack
I'm not blaming you for this because there isn't another option but linking new users to the entirety of the sequences is, like, the most ridiculous habit we have here. Imagine you've never been here. You sign up because you read something clever someone wrote about groupthink. Then someone suggests you check out a list of some 600+ blog posts covering a tremendous variety of seemingly unrelated topics which the user may already know more about than Eliezer or lack even the basic background knowledge necessary for comprehension. I also doubt that the sequences are even the 600 whatever best/most important/most helpful introduction to what is going on here. Sure, some new users will be enthralled by the Sequences or parts of them but I have trouble imagining a worse gateway than that giant list with subsection upon subsection.
3Kevin
It's made worse because the first major topic in the sequences is Bayesian probability. It's important, sure, but most people find it really boring. I don't have a better alternative reading order though, I wouldn't recommend my mishmashed and nearly random flow through the sequences to others. I'm hoping that Eliezer isn't planning on editing his rationality book entirely by himself and that a professional editor can turn it into something with more intuitive readability and flow that could serve as a better introduction to this site than the well intentioned advice of "read the sequences." For now, at least for people that like sci-fi, I think "go read Three Worlds Collide" is better advice than "go read the sequences".
4Kevin
I thought panda's comment deserved positive karma. Nothing wrong with expressing sincere appreciation for an article. Also, since panda's comment gave nothing to indicate that we should actively discourage them from participating here, we shouldn't vote the comment into the negative because panda would probably interpret it as a sign for "go away".
1AngryParsley
I hate to be the resident spelling/grammar Nazi, but I've commented on this before.
3panda
Thanks for that. :) Definitely wasn't expecting it. I may have wandered in with the wrong idea. I found this site linked from someone's blog, so I thought it was a plain old blog too, not a discussion community. Sorry - my comment didn't add much to the discussion at all! But thanks for being helpful, everyone. After checking out your links, AngryParsley, this place is a bit too clever for me! Fascinating stuff, but I probably can't contribute. But I've recommended it to my more clever friends irl, who'll probably join in your sequences at some point. Sorry again, guys! As you were. :)
[-][anonymous]00

I am quite surprised that more Karma functions do not track both the Up Votes and the Down Votes and display a running tally of each.

On a relative basis I would be far more interested in reading a post that has had 1,000 Up Votes and 999 Down Votes, than one which has had a single Up Vote.

EDIT: This comment does not apply to the current version of the post.

Considering the post to which you linked (about labeling the cryonics opinion on LW as groupthink) has no attempts at rebuttal, it seems like it merits an answer instead of merely dismissing it as an "attempt" - which is just saying, "This isn't sufficient" without any statement about what would be sufficient.

3Morendil
Consider the full facts about that comment. It starts thus: "I'm only posting this to play devils advocate" - which is a good reason (see ciphergoth's forthcoming list of phrases to never use) to not even attempt a rebuttal. It suffices as an example of what I want an example of, which is someone who is at least trying, and acknowledging groupthink as a technical term. I don't have to set the bar at "trying and succeeding", not for cryonics: the debate on cryonics has enough evidence of being a debate, so we already know that groupthink isn't happening on that particular topic. Don't demand particular proof that groupthink isn't happening, such as someone saying "groupthink!" with strong evidence plus rebuttals of their points. (Hey, I'm agreeing with Eliezer, and linking to his post. Groupthink!)
0bgrah449
I didn't accuse anyone of groupthink or demand any particular proof opposing or supporting claims of groupthink. I said it warranted a rebuttal before being dismissed as an attempt.
3Morendil
Calling it an attempt is no dismissal. Successful attempts are a subset of all attempts.
1bgrah449
ADBOC - People refer to successful attempts as "successes," not as "attempts."
2Morendil
So conceded. Suggest an edit to the post?
0bgrah449
I don't want to disrespect the graciousness of conceding this minor point, but I also don't have a great suggestion. Maybe something as simple as becoming ? But up to you, I just wanted to point out that "attempt" was bringing in some probably-unintended judgments.
0Morendil
Edited, with strikethrough. I wish one could mark comments as applying to a past version of a post - just making the edit would make this exchange meaningless.
0bgrah449
I can edit my comment, if that helps - "This comment does not apply to the current version of the post."
2Unknowns
The fact that that post is presently at +5 karma is actually evidence against groupthink.
3bgrah449
It seems dirty to say, "It raises a bunch of good points," and then, without answering any of them, say, "See? I just said it has good points. No groupthink here, thus disproving his points."
0Unknowns
Maybe that's true, but still, given groupthink, you would have expected negative karma for the post.
1wedrifid
It merits an answer the first time. After the tenth time it does not. There is sufficient discussion already available on both cryonics and accusations of groupthink regarding cryonics. Just downvoting the cry of a groupthinking wolf is adequate.
0bgrah449
At the time I posted this comment (and at the time I post this one), the post to which I'm referring still has no rebuttals. I will be very disappointed if someone posts a very weak rebuttal which is subsequently held up as sufficient for no reason other than it exists.
0prase
In some sense, absence of rebuttals can be viewed as silent agreement, thus implying the presence of groupthink (both because it is an agreement with the accusation thereof, and more so because the agreement is silent). On the other hand, if there was a real groupthink, people would be trying hard to rebut all accusations, so the lack of the rebuttals may be interpreted as evidence against groupthink. Which leads me to think that playing devil's advocate may not be the best way to detect groupthink. Perhaps the groupthink hypothesis is not testable, in the same way conspiracy theories aren't. Is it correct? If not, are there some reliable tests, some questions whose anwers (or lack of answers) would tell us whether there is groupthink present inside the group or not?

Lots of different words and phrases "devalue" different technical terms, since they exist outside of their technical definition. From what I can see from the OED, group think has been used as a term since 1923 and similar phrases like group mind were used in the late 19th century. Because someone makes a definition in a field it does not strip the original word or phrase of its meaning. If that was the case I'm sure lawyers would have a field day with all of us and that I could pick out quite a few misuses of onto on this site.

The technical mean... (read more)

6Morendil
Re "less talking", yeah. One thing that particularly disappointed me was when I proposed starting up a Jaynes study group and got pretty much zero uptake. More disappointed even when people cited "too hard" as the reason. Good points about technical meaning. I guess it's not really people respecting the exact framework from Janis that I'd like, so much as saying things which are more interesting than "you guys are biased". It's just too easy to troll that way. I'm not saying "no one should be able to charge groupthink". I'm hoping that the discussion below this post will encourage and help newcomers (or even regulars) to make clearer and more pointed diagnoses of the pathologies of this community. Just saying "Robin and Eliezer have really biased the crowd on this one", and particularly in the context of a discussion which was devoted to close critical examination of "this one", is providing no value. I'm a relative newcomer, so not really motivated by the celebrity status from OB. I want to encourage discourse that provides value and discourage that which does not. I am so motivated not out of righteousness, but for purely selfish motives: I have learned a lot from this site already (Bayes, Jaynes, MWI, a bunch of smaller conceptual tools from Eliezer's OB writings) and I want more of that good stuff. Positive ROI is what this is about.
3jimrandomh
Downvoted for wasting my time with many paragraphs of empty status signals, making an unreasonable request for censorship, and demanding work (software) that you don't have the right to ask for.
1rortian
This was done in the past and I think it was a great request. It was a suggestion that I think people here would enjoy doing. I always have fun when I'm coding something experimental (of course these things are always fun until they're not). Interesting charge. Don't worry though, I caught your hostile signals.
6jimrandomh
Sorry, that came off more strongly than I intended. I think the main reason I reacted that way is because your post seemed to put status concerns above truth. In particular, this paragraph Alleges that Less Wrong members are committing groupthink without claiming that they're actually wrong about anything. That cleanly separates further discussion from the facts of any particular matter, and, while I acknowledge that this is an Eliezerish view to hold, I consider that a very bad thing. Technically, that was only a moratorium. But more to the point, topics come and go on Less Wrong, posts about particular topics tend to be temporally clustered, and there're always some complaints about the current topic du jour. I wouldn't worry about cryonics and FAI crowding out other topics; there've been enough posts about them recently that there isn't much left to say, so we probably won't hear much about them for awhile.
3Cyan
Voted you up for some of your points and suggestions, but I gotta say, if you think the OP suggested then you didn't read what it actually said.
2rortian
Fair point. It would have been more precise to say raise the difficulty in charging etc.
-1Eliezer Yudkowsky
Why was this voted up to +4? Y'all are way too scared of being labeled cultish if you're voting this stuff up. I really wish there was some way to teach arrogance. It seems to be such a large factor in whether people actually make progress as rationalists or not.
6Jack
I liked the suggestions. I wouldn't permanently ban any topics but if every couple of months we stopped talking about transhumanist topics I think the results would be really constructive and help grow the community.
4Morendil
Just keep modeling. ;)
4D_Alex
Why was this not downvoted to -10? Y'all are way too cultish if you are not voting this stuff down. I really wish there was some way to teach irony. It seems to be such a large factor in whether people actually make progress as rationalists or not.
3wedrifid
I don't usually find cause to say this in reply to downvoted comments, but that is worth a post. Particularly because the conception you have of arrogance (or at least, the conceptions that I infer you have about arrogance) crosses some significant inferential barriers so is lost somewhat in this context.
4Nick_Tarleton
I think The Proper Use of Humility, The Proper Use of Doubt, Science Doesn't Trust Your Rationality, and Einstein's Superpowers contain most of the essential ideas, which isn't to say that a post amplifying the point wouldn't be useful.
0wedrifid
I agree on both counts, and thanks for rounding up the links. Somewhere in my collection of half-baked drafts is a post specific to arrogance itself, how the definition is tied to status, when it is useful and when it isn't.
-5rortian
[-]taw-20

Downvoting is a good evidence for groupthink. This site uses the same mechanism for up/downvoting comments as Reddit, so:

  • Go to a random subreddit and figure out what their groupthink is. It's usually very blatant.
  • Make some comments agreeing or disagreeing with groupthink but otherwise being similar.
  • See which are upvoted and which are downvoted.
  • Extrapolate to lesswrong by reference class forecasting.

So for example, reddit believers largely believe in evolution. Make a comment in favour and a comment against, the former goes up while the latter goes down, and groupthink is proven. I think you need a more rigorous experimental method.

1bgrah449
A position can be well-supported by facts and still be well-supported by a group for reasons other than facts. EDIT: The statement above is a truism. I also don't think a group's support for or opposition against evolution is a sufficient indicator of all other group opinions being correct.