WEIRD may be weirder than you think. We Aren't The World writes of psychological experiments on non-Westerners that give vastly disparate results from results that have been assumed to be hardwired, and the implications of this:

Henrich used a “game”—along the lines of the famous prisoner’s dilemma—to see whether isolated cultures shared with the West the same basic instinct for fairness. In doing so, Henrich expected to confirm one of the foundational assumptions underlying such experiments, and indeed underpinning the entire fields of economics and psychology: that humans all share the same cognitive machinery—the same evolved rational and psychological hardwiring.  The test that Henrich introduced to the Machiguenga was called the ultimatum game.


To begin with, the offers from the first player were much lower. In addition, when on the receiving end of the game, the Machiguenga rarely refused even the lowest possible amount. “It just seemed ridiculous to the Machiguenga that you would reject an offer of free money,” says Henrich. “They just didn’t understand why anyone would sacrifice money to punish someone who had the good luck of getting to play the other role in the game.”


At the heart of most of that research was the implicit assumption that the results revealed evolved psychological traits common to all humans, never mind that the test subjects were nearly always from the industrialized West.

Edit: The actual papers this article writes about are covered in this post by Ciphergoth from a few years ago.

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

I realized I was a bad test subject when they had me play the Ultimatum Game and I thought, "Naively I'd want to offer $1 out of $10, but I read about a study that says most people won't accept less than $3. OK, I'll offer $3."

Psych students and habitual test subjects (I lived on psych studies for about a year) are even worse than Westerners in general.

Psychology and economy have this unfortunate effect that publishing how people react sometimes changes how people react, so the successful theories can be automatically killed by their own success.


The piece from that article that I found most disturbing was that, for the Asch Conformity Experiment, we Americans fail the least. Google suggests we fail pretty hard. (~70%) If the rest of the world fails worse....yeah, that's bad.

Note that "failure" means going along with the group even once. IIRC, most people answered most questions correctly.

Hrm. I somehow managed to miss that part; thank you. I guess the more relevant number is the percentage that incorrectly went along with the group, consistently.

(off-topic: You didn't happen to spend some time on a certain gaming-related Usenet group in the late-90s, did you? Your name is familiar.)

(off-topic: You didn't happen to spend some time on a certain gaming-related Usenet group in the late-90s, did you? Your name is familiar.)

As a matter of fact, I did indeed spend time on gaming related Usenet groups in the late 90s, using this very name. (My particular favorite was one where giant signature files were in style.) Small world, huh?

That is the one I was thinking of. Small world indeed. If you're interested, some of the regs moved to a private forum when Usenet began to die, and it's still in use. I'll PM you.

Eh. Conformity is a really, really good heuristic most of the time.

True. You do run into informational cascades once in a while (a few people start doing something, other people think they must have a good reason and follow them, and soon everyone is doing it merely because everyone else is) but most of the time going along with the group isn't going to screw you; even if it's objectively stupid, at least it won't change your relative position within the group after the feces hit the fan. Also, to use a really bad analogy, if everyone else in your hunter-gatherer tribe jumps off the cliff and you don't, you haven't gained much because lone survivors don't have much of a life expectancy.

A more modern example:

You work at a mutual fund. The other people in the fund are making a specific investment that has a 50% chance of tripling their money and a 50% of being worthless. You discover another investment that you believe has a 75% chance of tripling your money and a 25% chance of being worthless, but everyone else has already committed their share of the firm's funds and no longer cares what you have to say. Which of the two investments should you allocate your share to? These are the possible results:

1) The 50% and the 75% investment both pay off. Regardless of which investment you made, the result is the same: you made money for the firm and keep your job. 2) The 50% investment pays off, the 75% investment doesn't If you took the 50% investment, you're okay. If you took the 75% investment, you look stupid and get fired. 3) Neither investment pays off. The firm goes bankrupt and you lose your job, regardless of which investment you made. 4) The 50% investment doesn't pay off, but the 75% investment does If you picked the 75% investment, well, you made money, but the firm is still bankrupt. You lose your job anyway.

So if everyone else has already picked the 50% investment, there's no reason for you to invest in anything else, because if their bet fails, you lose, too. Might as well go along with the crowd!

One interesting thing about the Machiguenga is the explanation that they give for their allocations -- it's not that they don't care about fairness; it's that they think that since the players' roles were selected at random, the game is already fair. Schelling makes a related point: it matters what we call the roles in games, because that affects how people will play. If we called the game "lottery" instead of "ultimatum", then WEIRD subjects might well be inclined to share less.

My father was also confused when he read that people would reject low offers. After all, it was never their money in the first place, so why should someone think they have any right to get more of it?

Has the ultimatum game been repeated with westerners with the western equivalent of "several days wages" between people who know each other well rather than randoms?

It's been repeated somewhat. Joseph Henrich explains his methodology in his paper here. He used UCLA grad students because of their similar community closeness, offering $160 (2.3 days of wages) vs 20 soles, which is supposedly the Machiguenga equivalent. It was compared to tests done by another group in other non-western locations (Tokyo and Java), which show similar results to those of the UCLA students.


You should probably spell out the acronym "WEIRD" (Western, Educated, Industrialized, Rich, and Democratic) at the start. I thought it was just a capitalize-the-first-word-in-an-article formatting until I read gwern's link.

So, Japanese, Taiwanese and South Koreans are EEIRD?

Yeah, probably. And then India and China are another not-totally-huge step away.

I figured it was well known, sorry about that. Link added!

Anthropology is a mindkiller.

I suspect that what many people say about other cultures reflects their own political opinions. Someone should do a research on that -- ask people about their political opinions, and then ask them about whether they consider some descriptions of foreign cultures likely or unlikely.

Seems to me that these are the attractors of describing foreigners:

1) They are exactly like us. All differences are just superficial details, analogies of what we do or think. When the difference is undeniable, it is only because they are forced or brainwashed to be different, but deep inside they would really prefer to be like us. Their different state is so unstable that any change (like sending them missionaires, or overthrowing their government) will start the inevitable process of them becoming exactly like us.

2) They are worse than us. Every difference without exception shows their intellectual or moral deficiency. Any negative description of them is credible; even if it does not make much sense (it's because they are idiots). They would do best by trying to copy us as much as possible, but it is a goal they can never achieve alone. Our mastery over them would be a blessing for them.

3) They are noble savages, infinitely morally and spiritually superior to us. If something about them seems imperfect, it is just our blindness to the higher truth. They live in perfect harmony with nature and other people, and if you have an evidence against that, you are just brainwashed by our evil culture. (Maybe they sometimes murder and torture people horribly, but... well, you simply don't understand it in the proper context. Also, did I tell you that you are brainwashed by our evil culture?) We are probably not worthy enough to copy them, but the least we can do is admire them and use some of their words, proverbs, or traditional clothing.

4) They are as different as different species or someone from a different planet would be. There is no hope of ever understanding each other. Protect our borders. Trade with them, if convenient, but don't care about anything else they do. (If their part of the trade with you involves slave labor, torture or murder of their people, well, it's their people and maybe in their morality it is okay, and you are not responsible for any of that. Just treat it as a black box and focus on your business.)

These descriptions can probably be mapped onto our political or other opinions. Each description suggests that we should do something, so probably each of them will sound credible to people who already think that we should do that.

Meta: Do other cultures also have similar attractors? Seems to me that I should include at least two to the list:

5) They are evil spirits who cause us harm. (Of course we, the good people, must kill them!)

6) They are strong but stupid. (We, the clever people, are allowed by our gods to steal from them and otherwise exploit them, but we must be careful to never confront them openly.)

Note: The common assumption in all descriptions is that all people from the foreign country are the same. (Maybe except the first one, which assumes an evil powerful minority preventing the rest of the population from becoming like us.)


Anthropology is a mindkiller.

I don't understand - you've listed a number of failure modes, many of which are likely results of mindkiller / Keep your identity small failures.

And you conclude from this that trying to use anthropology-type analysis should be avoided to the extent we avoid political analysis? It seems like a lot of the questions we'd like to able to answer to define Friendly or solve problems in a community need that kind of analysis.

At this point, I'm fairly certain I've misunderstood you in some way.

When people speak about politics, they often get funny in the head, so the credibility of what they say is extremely low. Something similar seems to happen in anthropology -- the "foreign people" are used as a canvas to illustrate our political opinions. Therefore the credibility of anything related to anthropology should be also very low... unless accompanied with a lot of data. By which I mean: if you say that X is true for people everywhere, you better show me you did real experiments in different cultures. Also, if you tell me that X is different.

I liked your analysis of the "attractors", but the "anthropology is a mind-killer" seems under-justified. More talk about politics really does seem to do harm by causing people to break into factions. Are you arguing that anthropology talk creates factions? Or does it just manifest factions that already exist?

You don't seem to me to have picked out anything particularly dangerous about anthropology talk. It does often serve as a "canvas to illustrate our political opinions", but the same can be said of any subject where people mostly talk and think in a "far" mode.


For what it is worth, I think Villiam_Bur is completely right that personal identity (as Graham means the term) is the mindkiller. Thus, all personal identity discussion should be viewed with exactly the skepticism of political discussion.

The problem is that discussions of these types of topic seem to incorrectly assume that they are not discussing personal identity, and so don't need to be wary of the mindkiller.

There's a potential issue in keeping your identity small; I've noticed it within myself: I dislike people who don't. It annoys the hell out of me, and I regard them as part of the outgroup.

Which suggests to me that "keeping your identity small" ("individualism," loosely speaking) is part of my identity.

I'm not sure if this is an actual problem, so for now I simply keep it in the back of my mind as a potential biasing factor.


Which suggests to me that "keeping your identity small" ("individualism," loosely speaking) is part of my identity.

I think incorporating Graham's lesson is valuable to being rational.

That said, I don't think "individualism" is the only way to keep your identity small. For example, a collectivist is perfectly capable of having a small identity.

Now I ponder what "small" means, exactly.

On the one hand, having fewer identifiers seems a useful definition; it means fewer potential points of bias.

On the other, having fewer -shared- identifiers -also- seems a useful definition; it means fewer potential agents you are going to share a bias with. From the individualist perspective, you want the most accurate information for yourself, and you're less likely to regard ideas from the ingroup with sufficient criticism. But from the collectivist perspective, you want the most accurate information for your group, which suggests that the rationalist collectivist doesn't include the collective as part of their identity.

Unless I miss an alternative logic?


I don't think Graham expects one's identifiers not be shared with others. He's arguing the point more in line with your first definition, in that the more aspects you add to your personal identity, the more sources of bias one must avoid. As Graham says, "all other things being equal, the best plan is to let as few things into your identity as possible."

In this context, individualist and collectivist are labels that describe an agent's preferred social arrangements and social norms. There's nothing about keeping one's identity small that prohibits one from having a higher priority for group preferences than individual preferences.

Are you arguing that anthropology talk creates factions? Or does it just manifest factions that already exist?

It provides justifications for them. It allows existing factions to pretend they are fact-based, by creating a (possibly fictional) description of foreign people which confirms the map of the world used by the faction. This will make other people join the faction, even if they originally just wanted to see the territory as it is.

As an example, if I am part of a faction that worships unicorns, I can gain more followers by publishing a textbook describing a foreign society which worships unicorns and therefore has no wars and violence, and a lot of cookies. Now when you know how much happiness unicorn worship brings, you are more likely to join us. And you know it must be true, because it was taught at a university.


if you say that X is true for people everywhere, you better show me you did real experiments in different cultures. Also, if you tell me that X is different.

That's very reasonable standard, but I don't think people apply it consistently in this part of concept-space. I agree that people go funny in the head when dealing with personal identity, so one should be more critical of assertions about the universality (or distinctiveness) of a particular identity.

For example, this comment makes an assertion that certain types of effective self-improvement universally have particular properties that some people find undesirable. The universality of that property can (and is) challenged on an empirical basis, but such a possible change was not addressed - or even acknowledged as possible.

Further, sometimes people use your personal-identity-is-the-mindkiller point as a fully general counter-argument to policy positions they don't like.

I definitely agree that these attractors seem really common but I am not sure how much they make up all anthropology.

I suspect that what many people say about other cultures reflects their own political opinions.

Is that different in some way from what people say about their own culture? Shouldn't all statement about psychology suffer from the same issue?


In your meta point, you portray other cultures as antagonistic and employing different axes. This opinion would belong to the 2,4 quadrant (worse, different), right? Or is it just me sitting in 1,3 relative to you?

I meant that the positions 1,2,3,4 are typical for large and complex societies (not only Western ones),and 5,6 for small and primitive societies. The main difference is that in a primitive society, your tribe interacts with the strangers often. In a large society, unless you live on the border, the vast majority of your interactions is with your neighbors. Your opinions about strangers are mostly for signalling for your neighbors, because you interact with strangers very rarely.

My assumption is that people are usually antagonistic to their competitors. This is why in a small society I expect more overt hostility to strangers.

A large society usually has some rules to supress hostility, but more importantly, most of the remaining hostility is targeted towards your neighbors belonging to the opposing faction. Thus less hostility towards strangers. Actually, your faction can use friendly attitudes towards strangers as their attire. It does not actually have to be sincere. (For example, as an anti-racist you are not required to have friends among people of other races. And having friends among people of other races does not make you an anti-racist; it will usually not be even accepted as an evidence. The only requirement of an anti-racist is that they must hate racists and support other anti-racists. And symetrically, the only requirement of a racist is to hate anti-racists and support other racists. The actual hate towards people of other races in a private life is not required. It's about belonging to a faction in the internal conflict; the actual strangers are mostly irrelevant. A person worshiping noble savages is not expected to adopt their lifestyle. Only to worship them, and thus show superiority over the less enlightened neighbors who don't worship the same noble savages.) So the complex society can split along "anti-stranger" and "pro-stranger", or "strangers are the same" and "strangers are different" axes. Unless is has other, more important internal divisions, for example religious ones.

My position is that people are biologically very similar (1), but living in some culture often makes them accept the cultural norms, which can seem completely crazy to us (4 - with the exception that we should morally care about what they do to their own people; we just shouldn't naively expect that they will share our concerns). I do consider complex cultures mostly better than primitive ones (2), with one significant exception that a life in a primitive society, although shorter and more painful, is probably less frustrating (3 - in one specific aspect, which is rather low on my personal priority list).

I think that in Western societies, political correctness plays the role religion has in other societies. So I would expect other societies to be less concerned about strangers, and more concerned about whatever is the burning religious topic for them. So instead of using strangers as arguments for their political positions, they would use something else. But they may be mindkilled about strangers because of what their religion says about them.

I don't see what you mean by political correctness playing the role of religion.

People have been doing the ultimatum game in a wide variety of cultures for a while, and what this article calls the 'western' behavior isn't exactly restricted to westerners.

We were talking about WEIRD before it was cool.

And this is about the same Henrich that article discusses, too.


Westerners are only terrible subjects if you are primarily interested in results from these "isolated cultures." If you live in a western culture, you probably are interested in the implications of research for your own society.

Also, I suspect "isolated cultures" aren't representative of most cultures throughout history because most cultures weren't isolated.

Yeah. There's a difference between 'USA', 'North America plus Europe', 'North America, Europe, China, Japan, India, and Parts Of The Middle East' and 'Highly Isolated Small Cultures'.

Nothing of that is exactly news.

This reminds me a bit of Dan Ariely's Amazon gift card experiment. He said people were irrational to take the $10 card for free rather than the $20 card for $7. I thought this was completely wrong because he wasn't factoring in the overhead of keeping track of and actually using gift cards, which I am personally really bad at. (I emailed him about this and he actually replied but I don't think he understood what I was talking about.) It seems like these other cultures have other factors associated with giving or accepting gifts (or money) that researchers initially didn't take into account. I'm also pretty confident that if the researchers asked the people to explain their decision, they'd have a reasonable, articulate explanation. It seems like that article is leaning toward, "loook, a mysterious difference in braaain fuuunction.~"

I think the point isn't "loook, a mysterious difference in braaain fuuunction.~" but that nurture is more important than nature when it comes to human psychology.

But they act as if the impact of nurture on decision-making is this inaccessible thing. The people making the decisions could have been able to tell them why they decided the way they did.

But they act as if the impact of nurture on decision-making is this inaccessible thing.

Who do you mean with "they"?

The people making the decisions could have been able to tell them why they decided the way they did.

I would guess that the orginal papers do include some discussion of why they decided the way they did.

Oh sorry! I mean the people who wrote the article. (Which is similar to the way Dan Ariely wrote in his book that people have no good reason to pick the $10 card, even though we do sometimes. He just concluded that we irrationally value free things more than we should.)

I think I'm concerned that by not providing any explanations, this article jumps into @Villiam_Bur's post's category #4. Other cultures are so different! They make these totally alien decisions with this game! If the researchers themselves did have explanations, then that makes the article even more sketchy because its authors decided to overplay the cultural difference instead of presenting the results as "probably if you were raised [like this], you'd also play this game differently" -- which is what the research likely found.

EDIT: I caught a weird bug where I keep saying "they" when referring to the article authors. I fixed it myself this time!

I mean the people who wrote the article.

The article isn't written by multiple people but by one person called Ethan Watters who reports about the scientific work of Joe Henrich and his collegues to a nonscientific audience.

As a result there are multiple different groups you could mean with "they".

I think I'm concerned that by not providing any explanations

Explanations aren't what science is about. Science is about having a theory and using it to make predictions. Then you see whether those predictions are accurately describing reality.

this article jumps into @Villiam_Bur's post's category #4. Other cultures are so different!

I don't think there anything wrong with making that argument. The orginial WEIRD peer reviwed paper makes that point with data.

They make these totally alien decisions with this game!

It's not about whether the decision violates "common sense" and is alien. It's about whether it violates theoretic models that were developed in academia.

On the positive side, running tests on WEIRD subjects does at least tell you about the psychology of WEIRD people, and there are enough of them around to make the knowledge worthwhile even if it doesn't generalize to other populations.

The biggest problem is that it probably means these results won't hold going forward even for future WEIRD people.

Is there something in the article that's new and worth reading if you are already with the discussion from a few years ago?

Not really :-) I'd forgotten about the earlier post until Gwern mentioned it below. (I'd even commented on that thread.)