Book Review: The Elephant in the Brain

Link post

We don’t only con­stantly de­ceive oth­ers. In or­der to bet­ter de­ceive oth­ers, we also de­ceive our­selves. You’d pay to know what you re­ally think.

Robin Han­son has worked tire­lessly to fill this un­met need. To­gether with Kevin Sim­ler, he now brings us The Elephant in the Brain.

I highly recom­mend the book, es­pe­cially to those not fa­mil­iar with Over­com­ing Bias and claims of the type “X is not about Y.” The book feels like a great way to cre­ate com­mon knowl­edge around the claims in ques­tion, a sort of Han­so­nian se­quence. For those already fa­mil­iar with such con­cepts, it will be fun and quick read, and still likely to con­tain some new in­sights for you.

Two meta notes. In some places, I re­fer to Robin, in oth­ers to ‘the book’. This is some­what ran­dom but also some­what about which claims I have pre­vi­ously seen on Over­com­ing Bias. I nowhere mean any dis­re­spect to Kevin Sim­ler. Also, this is a long re­view, so my apolo­gies for not hav­ing the time to write a shorter one, lest this linger too long past the book’s pub­li­ca­tion.

I

The book di­vides into two halves. In the first half, it is re­vealed (I’m shocked, shocked to find gam­bling in this es­tab­lish­ment) that we are poli­ti­cal an­i­mals and con­stant schemers that are con­stantly look­ing out for num­ber one, but we have the de­cency to pre­tend oth­er­wise lest oth­ers dis­cover we are con­stantly schem­ing poli­ti­cal an­i­mals. The eas­iest way to pre­tend oth­er­wise is of­ten to fool your­self first. That’s ex­actly what we do.

What are our real mo­tives? Ex­actly what you’d think they would be. We want the loyal and strong poli­ti­cal al­lies, true friends, the best sex with the most fit mates, food on the table, en­force­ment of our preferred norms, high sta­tus, re­spect and other neat stuff like that. The whole re­pro­duc­tive fit­ness pack­age. To get these things, we must send the right sig­nals to oth­ers, and de­tect the right ones in oth­ers, and so forth.

This in­sight is then used to shine a light on some of our most im­por­tant in­sti­tu­tions:

This book at­tempts to shine light on just those dark, un­ex­am­ined facets of pub­lic life: ven­er­ated so­cial in­sti­tu­tions in which al­most all par­ti­ci­pants are strate­gi­cally self-de­ceived, mar­kets in which both buy­ers and sel­l­ers pre­tend to trans­act one thing while covertly trans­act­ing an­other. The art scene, for ex­am­ple, isn’t just about “ap­pre­ci­at­ing beauty”; it also func­tions as an ex­cuse to af­fili­ate with im­pres­sive peo­ple and as a sex­ual dis­play (a way to hob­nob and get laid). Ed­u­ca­tion isn’t just about learn­ing; it’s largely about get­ting graded, ranked, and cre­den­tialed, stamped for the ap­proval of em­ploy­ers. Reli­gion isn’t just about pri­vate be­lief in God or the af­ter­life, but about con­spicu­ous pub­lic pro­fes­sions of be­lief that help bind groups to­gether. In each of these ar­eas, our hid­den agen­das ex­plain a sur­pris­ing amount of our be­hav­ior—of­ten a ma­jor­ity. When push comes to shove, we of­ten make choices that pri­ori­tize our hid­den agen­das over the offi­cial ones.

I know it is as a shock to you, but Robin doesn’t think X is about the Y it claims to be about. In­stead it’s about sig­nal­ing and sta­tus. I know, it’s a lot to take in. Stop. Catch your breath.

Robin and Kevin re­al­ize this is good news:

This may sound like pes­simism, but it’s ac­tu­ally great news. How­ever flawed our in­sti­tu­tions may be, we’re already liv­ing with them—and life, for most of us, is pretty good.

So if we can ac­cu­rately di­ag­nose what’s hold­ing back our in­sti­tu­tions, we may fi­nally suc­ceed in re­form­ing them, thereby mak­ing our lives even bet­ter.

There are many ex­am­ples through­out the book where par­tic­u­lar things are held up as about sta­tus and sig­nal­ing. In some cases, it is ex­plained why these ac­tions are valuable or so­cially benefi­cial, but in most cases the clear im­pli­ca­tion is that it is all pointless zero-sum games and a ter­rible waste.

The perfect en­cap­su­la­tion of this at­ti­tude is some­thing I re­mem­ber from Over­com­ing Bias, rather than from the book, in a post called Har­ness­ing Po­lariza­tion:

Hu­man sta­tus com­pe­ti­tion can be waste­ful. For ex­am­ple, of­ten many ath­letes all work hard to win a con­test, yet if they had all worked only half as hard, the best one could still have won.

As a sports fan and a former pro­fes­sional game player, I re­spond that yes, the same ath­lete wins. But it wouldn’t count. The win would be tainted, the spec­ta­cle ru­ined. We live to com­pete with all we have, and to watch oth­ers com­pete with all they have. We shake each oth­ers’ hands, say ‘good game’ and mean it. After we’re done, we an­a­lyze what hap­pened to gain un­der­stand­ing. We cel­e­brate striv­ing, ex­cel­lence and achieve­ment.

Com­pe­ti­tion is not about win­ning.

What looks like the defi­ni­tion of a zero-sum game is quite the op­po­site.

Thus the bat­tles be­tween our elephants to gain ad­van­tage, in their in­evitably in­creas­ing com­plex­ity, cre­ate the arms race of big­ger and bet­ter brains, more and more com­plex be­hav­iors, cul­tures and tools, and all the nice things. Each of us has an in­cen­tive to ad­just the in­cen­tives of those around us back to that which is hard, that which is use­ful, that which makes us stronger, up and down the chains of meta lev­els.

Every­thing is Bayesian ev­i­dence of oth­ers’ true mo­tives and loy­alties and ca­pa­bil­ities. We need elephant-sized brains to have any hope of pro­cess­ing all of it.

Bask in the won­drous deeds of un­con­scious Bayesian anal­y­sis and leaky self-de­cep­tion. The Elephant in the Brain is a won­der­ful defense of hypocrisy.

I, for one, take great joy in the merely real, the great­est show on Earth.

II

The next few sec­tions will be a quote that stood out, and my as­so­ci­ated thoughts.

Who would you rather team up with: some­one who stands by while rules are flouted, or some­one who stands up for what’s right?

Good ques­tion!

The book as­sumes the an­swer is the one who stands up for what’s right. But are you sure about that?

Are you sure you don’t want a win­ner? Who will be smarter than to waste re­sources re­in­forc­ing norms?

A cen­tral con­cept in the book is that norms are en­forced be­cause the meta-norm of en­forc­ing norms is also en­forced by the meta-meta-norm of en­forc­ing the en­forc­ing of rules, and so on. The meta-norm au­to­mat­i­cally acts on it­self, so the cy­cle never ends.

For that to work, the an­swer to the team up ques­tion must be the one who stands up for what is right. For the elephant that is what be­ing right means. In turn, for that to work, ev­ery­one must have a com­mon knowl­edge ex­pec­ta­tion that the meta-norm will be en­forced.

The Dar­win Game illus­trates (among other things, also spoiler alert) what hap­pens when there can be no en­force­ment of the en­force­ment of norms.

In a world… of ven­ture cap­i­tal talks about what hap­pens when be­ing a win­ner be­comes the norm. When the meta-norm en­forces ‘be a win­ner’ it also pun­ishes all other con­sid­er­a­tions. You de­stroy all pro-so­cial norms.

One could also point to cer­tain poli­ti­cal cam­paigns and at­ti­tudes. Nor­mally one must pre­tend to be in fa­vor of al­lies be­ing in fa­vor of norm en­force­ment, and tear­ing that mask off is re­ally, re­ally bad.

That gen­er­al­izes. Any ac­tions not re­in­forced by norms are seen by the meta-norm as ri­val ac­tions to norm en­force­ment, and are pun­ished.

Neu­tral­ity is not a thing. What is a thing is a war of norm against norm, re­ward against re­ward, en­force­ment against en­force­ment, level upon level of com­plex­ity.

III

What’s not ac­cept­able is syco­phancy: brown-nos­ing, bootlick­ing, grov­el­ing, toad­y­ing, and suck­ing up. Nor is it ac­cept­able to “buy” high-sta­tus as­so­ci­ates via cash, flat­tery, or sex­ual fa­vors. Th­ese tac­tics are frowned on or oth­er­wise con­sid­ered ille­gi­t­i­mate, in part be­cause they ruin the as­so­ci­a­tion sig­nal for ev­ery­one else. We pre­fer celebri­ties to en­dorse prod­ucts be­cause they ac­tu­ally like those prod­ucts, not be­cause they just want cash. We think bosses should pro­mote work­ers who do a good job, not work­ers who just sleep with the boss.

The first rule of sig­nal­ing is cheat.

The sec­ond rule of sig­nal­ing is catch cheaters.

The third rule, there­fore, is don’t get caught.

When too easy a strat­egy earns too big an as­so­ci­a­tive pay­off, that’s a great deal for you. But it ru­ins the sig­nal for oth­ers. They must bring the sig­nal back in line with the difficulty of the ac­tion, by mak­ing the strat­egy more difficult and/​or as­sess­ing a penalty against the sig­nal.

So the game moves up a level, to sig­nal­ing of difficulty, where the rules still ap­ply. The game gets more com­plex and the brains get big­ger.

Like the best games, the sig­nal­ing game is self bal­anc­ing, be­cause the play­ers change the rules to make it so.

How fun!

IV

This line stood out as quite odd:

You botched a big pre­sen­ta­tion at work. Feel the pang of shame? That’s your brain tel­ling you not to dwell on that par­tic­u­lar in­for­ma­tion.

Things that make me feel shame bother me for years. Decades, even. We pun­ish and dis­ci­pline and shape be­hav­ior by as­so­ci­at­ing it with shame. Th­ese days most of us con­sider shame toxic.

If shame is the brain’s way of tel­ling you not to dwell on that par­tic­u­lar in­for­ma­tion, it’s do­ing quite the epic fail of a job.

It seems more like the op­po­site. Shame is a tool for mak­ing damn sure you do dwell on that in­for­ma­tion, and ev­ery­one else knows it. It re­stores bal­ance. It is ex­plic­itly a tax.

Shame roy­ally sucks. That’s the whole point. That’s how you pay the tax. Much bet­ter to do with­out, but we must then re­store bal­ance an­other way.

V

Of course, we re­al­ize that a few ex­pert opinions don’t nec­es­sar­ily re­flect a con­sen­sus among all ex­perts—nor, it should be noted, is con­sen­sus opinion nec­es­sar­ily the truth.

Well said. Con­sider this your pe­ri­odic re­minder.

In part two, the book breaks down some of the mo­tives be­hind ma­jor ar­eas of life: Body lan­guage, laugh­ter, con­ver­sa­tion, con­sump­tion, art, char­ity, ed­u­ca­tion, medicine, re­li­gion and poli­tics.

The sec­ond half did an ex­cel­lent job of point­ing out the sig­nal­ing, sta­tus and strate­gic mo­tives be­hind these ar­eas of life. In this re­gard, I bought most of the book’s claims. Such mo­tives are all around us and cen­tral to al­most ev­ery­thing in­volv­ing mul­ti­ple peo­ple. We see over and over the struc­tures from the first half of the book. As the chap­ter ti­tles put it: Norms, cheat­ing, self-de­cep­tion, coun­terfeit rea­sons.

The pat­tern is to pre­sent the stan­dard ex­pla­na­tions for what hap­pens in such ar­eas of life, and then point out that this ex­pla­na­tion doesn’t make sense – much of our be­hav­ior re­mains un­ex­plained. So in­stead, such ar­eas must be about other rea­sons, and such other rea­sons are found and ex­plored.

Thus Robin’s claim that, con­trary to pop­u­lar be­lief, X is not about Y:

Food isn’t about Nutri­tion
Clothes aren’t about Com­fort
Be­d­rooms aren’t about Sleep
Mar­riage isn’t about Ro­mance
Talk isn’t about Info
Laugh­ter isn’t about Jokes
Char­ity isn’t about Helping
Church isn’t about God
Art isn’t about In­sight
Medicine isn’t about Health
Con­sult­ing isn’t about Ad­vice
School isn’t about Learn­ing
Re­search isn’t about Progress
Poli­tics isn’t about Policy

For most peo­ple, such state­ments provide a use­ful nudge in the cor­rect di­rec­tion. In an im­por­tant sense, all the above state­ments are true. In an­other im­por­tant sense, they’re all false.

Every­thing is, in an im­por­tant sense, about these games of sig­nal­ing and sta­tus and al­li­ances and norms and cheat­ing. If you don’t have that per­spec­tive, you need it.

But let’s not take that too far. That’s not all such things are about. Y still mat­ters: you need a McGuffin. From that McGuffin can arise all these com­plex be­hav­iors. If the McGuffin wasn’t im­por­tant, the fighters would leave the arena and play their games some­where else. To play these games, one must make a plau­si­ble case one cares about the McGuffin, and is helping with the McGuffin.

Other­wise, the other play­ers of the broad game no­tice that you’re not do­ing that. Which means you’ve been caught cheat­ing.

Robin’s stan­dard rea­son­ing is to say, sup­pose X was about Y. But if all we cared about was Y, we’d sim­ply do Z, which is way bet­ter at Y. Since we don’t do Z, we must care about some­thing else in­stead. But in there’s no in­stead; there’s only in ad­di­tion to.

A fine move in the broad game is to ac­tu­ally move to­wards ac­com­plish­ing the McGuffin, or point out oth­ers not do­ing so. It’s far from the only fine move, but it’s usu­ally enough to get some amount of McGuffin pro­duced.

There’s also the ques­tion of what we do when we re­al­ize peo­ple don’t care much about the McGuffin (here­after sim­ply Y). There are three paths one can take.

Path one is to say we could ac­com­plish Y bet­ter by do­ing Z, so let’s do Z. Here’s how to make medicine about health and poli­tics about policy! That can be a very good re­sult, but it would also kill the sig­nal­ing com­po­nent of the en­ter­prise. That likely means pay­ing for it in other ways, so Z do­ing Y bet­ter is not a suffi­cient case that it’s a bet­ter sys­tem.

Path two is to dis­par­age and dis­card the McGuffin, and play the broad game more openly. Medicine isn’t about health, so here’s the treat­ment that shows you care. Poli­tics isn’t about policy, so I’m not even go­ing to pre­tend to care about that. This is very bad. This is a huge vi­o­la­tion of the broader game’s rules. It’s cheat­ing. It’s too easy, it doesn’t show you’re clever or have spare re­sources, and the norms it im­plies are dis­as­trous.

The broader game is very clear that when you re­al­ize that X is mostly not about Y, you’re sup­posed to keep say­ing that X is about Y, be­cause that’s both how we keep the game’s difficulty high enough to send worth­while sig­nals, and how we get a lit­tle bit of Y out of the whole thing. A lit­tle is of­ten all you need; thanks to the civ­i­liza­tion this game has built, there re­ally is room for most of us to waste most of our time. The whole sys­tem is built on this hypocrisy.

I’ll now go through the chap­ters in part II, and give my take on the par­tic­u­lar claims.

VI – Body Language

I think this chap­ter is mostly spot on. Body lan­guage is about send­ing leaky and in­stan­ta­neous, and there­fore more hon­est, com­mu­ni­ca­tion, of­ten about sta­tus, while also hav­ing rich band­width and be­ing de­ni­able if re­ported to oth­ers. We pre­fer nat­u­ral body lan­guage be­cause it is a sig­nal we can trust, whereas un­nat­u­ral body lan­guage is ba­si­cally ly­ing. By us­ing such com­mu­ni­ca­tion meth­ods, we test each other on many lev­els, in­clud­ing our skill in in­ter­pret­ing and giv­ing such com­mu­ni­ca­tion, and get to see in a de­ni­able fash­ion what oth­ers are think­ing is go­ing on. Every move we make is based on ev­ery­thing it im­plies; just say­ing what you mean is im­pos­si­ble.

Mak­ing more things more ex­plicit, on the mar­gin, is of­ten a good idea. As are fur­ther ex­pla­na­tions of un­clear sig­nals, es­pe­cially to those who strug­gle with them. Ask Cul­ture gives peo­ple the valuable tool of di­rect com­mu­ni­ca­tion with­out the need to vi­o­late norms, but try­ing to screen off rich im­plicit com­mu­ni­ca­tion chan­nels, or treat them as some­thing we aren’t re­spon­si­ble for pars­ing, is a mis­take. Life is full of Bayesian ev­i­dence and we have a brilli­ant pro­ces­sor for rich data sets. We’d be fools not to use it.

Body lan­guage is about com­mu­ni­ca­tion, es­pe­cially com­mu­ni­ca­tion about sta­tus, but then there isn’t re­ally an­other thing for body lan­guage to not be about.

VII – Laughter

This chap­ter also seems mostly spot on, and seems much bet­ter than my pre­vi­ous mod­els of hu­mor. Alter­nate ex­pla­na­tions of laugh­ter don’t ex­plain ob­served be­hav­ior, whereas laugh­ter as a play sig­nal fits the data quite well. Laugh­ing at some­thing means it is playful, which has far reach­ing im­pli­ca­tions. It ex­plains why good-na­tured teas­ing and laugh­ter strength­ens re­la­tion­ships, while mean-spir­ited teas­ing and laugh­ter weak­ens them: the state­ment that some­thing is playful changes its im­pli­ca­tions. Cal­ling the right thing play shines a pos­i­tive light on your mo­tives, whereas over­reach­ing does the op­po­site. It means you don’t care. And of course, laugh­ter lets you com­mu­ni­cate things you can’t oth­er­wise say, as in the chap­ter’s clos­ing quote from Os­car Wilde: “If you want to tell peo­ple the truth, make them laugh; oth­er­wise they’ll kill you.”

The book doesn’t pay much at­ten­tion to the sim­ple fact that laugh­ing is fun, and thus in­cen­tivized. Laugh­ter be­comes a tool that iden­ti­fies and pos­i­tively re­in­forces play in the ex­plo­ra­tory, clever and ex­per­i­men­tal senses, which is great be­cause such play is highly use­ful. The bet­ter and more clever and un­pre­dictable and origi­nal the play, the bet­ter the laugh, en­courag­ing more use­ful tasks and big­ger brains. And of course, the bet­ter one’s sense of hu­mor – e.g. know­ing which things are clever and un­pre­dictable and origi­nal – the bet­ter one looks, and thus the vir­tu­ous arms race con­tinues.

Laugh­ter isn’t about jokes, but that was always clear. Jokes are an at­tempt to hack the hu­mor func­tion and get laugh­ter. Use­ful, but not the point.

VI – Consumption

Clearly this is in­tended to be a pos­i­tive as­so­ci­a­tion for many view­ers, but in Kevin’s case, the ad ac­tu­ally back­fired. There’s noth­ing wrong with the product it­self; it smells great and masks body odor effec­tively. But the cul­tural as­so­ci­a­tions were enough to dis­suade Kevin from us­ing the product. This shows how ar­bi­trary images can turn cus­tomers away, but by similar prin­ci­ples, other lifestyle ads must be hav­ing an op­po­site, pos­i­tive effect.

Kevin (cor­rectly) not want­ing to con­sume [product] due to its as­so­ci­a­tions is not a bug. It’s a fea­ture. The point of ad­ver­tis­ing [product] in this way is to as­so­ci­ate it with peo­ple who are very much Not Kevin. If Kevin were to use the product, it would be do­ing quite a bad job as­so­ci­at­ing with and sig­nal­ing Not Kevin! You wouldn’t want peo­ple smelling way too much [product] on peo­ple, and think­ing it was Kevin.

I’m sure [product] likes ex­tra sales, but there is a big­ger plan in mo­tion here.

You could also look at it as good Bayesian in­for­ma­tion for Kevin. If [product] is pay­ing to cre­ate these as­so­ci­a­tions, that is ev­i­dence [product] is a rel­a­tively bet­ter fit for peo­ple who would re­spond pos­i­tively. By re­spond­ing nega­tively, Kevin gets the in­for­ma­tion that the product is a poor fit for him. Since [product] ex­ists in a com­pet­i­tive mar­ket, that prob­a­bly means it’s a bad choice. Even if Kevin re­acted neu­trally to the mes­sage, that is a worse than ex­pected re­ac­tion to ad­ver­tis­ing, so Kevin should up­date nega­tively on the product. Smart elephant.

At a min­i­mum, he’d be pay­ing for the sig­nal, so it’s likely over­priced.

This is also quite healthy from an in­cen­tive per­spec­tive. If cre­at­ing an as­so­ci­a­tion with your product helped with some cus­tomers with­out hurt­ing with other cus­tomers, it would be an overly effec­tive, free ac­tion. It would be too easy, un­earned sta­tus. In other words, it would be cheat­ing. We catch such cheaters who are so low qual­ity as to be caught, and pun­ish their ac­tions with equal and op­po­site re­ac­tions.

In an­other ex­am­ple of good ad­ver­tis­ing:

The U.S. Marine Corps, for ex­am­ple, ad­ver­tises it­self as a place to build strength and char­ac­ter. In do­ing so, it’s not ad­ver­tis­ing only to po­ten­tial re­cruits; it’s also re­mind­ing civili­ans that the peo­ple who serve in the Marines have strength and char­ac­ter. This helps to en­sure that when sol­diers come home, they’ll be re­spected.

And what would we do with­out ad­ver­tis­ing?

To get a bet­ter sense for just how much of our con­sump­tion is driven by sig­nal­ing mo­tives (i.e., con­spicu­ous con­sump­tion), let’s try to imag­ine a world where con­sump­tion is en­tirely in­con­spicu­ous.

The book pre­sents the thought ex­per­i­ment of the Oblivi­ated world, where we are no longer able to form mean­ingful im­pres­sions of other peo­ple’s things. What hap­pens then? How much of our con­sump­tion would then be pointless?

The book’s an­swer is quite a lot:

To­day it’s con­sid­ered in­ap­pro­pri­ate to wear sweat­pants to a din­ner party or around the office. But in an Oblivi­ated world, where no one is even ca­pa­ble of notic­ing, why not?

Liv­ing rooms—which are of­ten dec­o­rated lav­ishly with guests in mind, then used only spar­ingly—will even­tu­ally dis­ap­pear or get re­pur­posed.

Who would care?

I would. I would care.

I spend a large por­tion of my life in my liv­ing room. Even with no guests. I want it to be a nice place. For me. The idea that they’d dis­ap­pear is down­right bizarre to me. They’re for liv­ing! It’s a thing!

I buy most of my lux­ury goods where they are not easy to no­tice.

In the medicine chap­ter it is noted:

When peo­ple buy choco­lates for their sweet­hearts on Valen­tine’s Day, for ex­am­ple, they usu­ally buy spe­cial fancy choco­lates in elab­o­rate pack­ag­ing, not the stan­dard gro­cery-store Her­shey’s bar.

Yes, part of that is be­cause you want to sig­nal that Valen­tine’s Day is spe­cial. But part of that is that you pre­sum­ably like your sweet­heart, and wouldn’t want him or her eat­ing a Her­shey’s bar.

I do not think, as I be­lieve Robin does, that most product va­ri­ety and cus­tomiza­tion is worth­less aside from sig­nal­ing. In­stead, I thank the sig­nal­ing mar­ket for en­courag­ing such great va­ri­ety, and curse it for of­ten mak­ing the good stuff so much more ex­pen­sive.

Robin has always dis­missed the value of in­con­spicu­ous con­sump­tion. I think it is vastly un­der­rated. I of­ten end up buy­ing the pack­age of con­spicu­ous and in­con­spicu­ous con­sump­tion to­gether, but there’s no other way to get the half I want. Per­haps that also means I am the ex­cep­tion that proves the rule. I can be­lieve most peo­ple in­stead throw away the other half.

If your restau­rant con­sump­tion is, as the book claims, ‘more for show­ing off’ than for per­sonal use, you are do­ing it wrong. Same with (from the same chart) mo­bile phones and liv­ing room fur­ni­ture.

In the long run, if con­sump­tion were some­how to­tally in­con­spicu­ous, would I slowly lose the as­so­ci­a­tions I have and de­value my in­con­spicu­ous con­sump­tion as well? Would the next gen­er­a­tion get no joy from nice things? Per­haps the next gen­er­a­tion. I think I’d stick to my guns. But the good stuff of­ten is ac­tu­ally good. Con­sump­tion of­ten re­ally is about con­sum­ing, even if much of it (and much of the more ex­pen­sive ver­sions of it) mostly aren’t. The sys­tem works.

The sys­tem also works in the sense that get­ting cap­i­tal­ists to cre­ate more and bet­ter con­spicu­ous con­sump­tion is both an en­g­ine of in­no­va­tion and tech­nolog­i­cal progress, and also an effi­cient method of con­spicu­ous sig­nal­ing. It’s cer­tainly not a perfect sig­nal of con­spicu­ous sig­nal­ing, we’d pre­fer we were all com­pet­ing to donate to effi­cient char­i­ties (hold that thought for four sec­tions), but mov­ing the com­pe­ti­tion into in­no­va­tion and de­sign and pro­duc­tion, and the bid­ding up of scarce re­sources, is quite the eco­nomic en­g­ine.

Con­sider the al­ter­na­tive. In an Oblivi­ated world, where would all that sig­nal­ing go? We’d have to turn to other sources. They would likely have to be be­hav­iors, which seems quite nasty and nega­tive sum with­out the benefi­cial eco­nomic side effects. The beauty of con­spicu­ous con­sump­tion is it eats your money but that money ends up el­se­where. Mostly harm­less. If it ate your time that time is gone. The al­ter­na­tive of us­ing mea­sure­ment and hav­ing a mer­i­toc­racy is great up to a point and then turns into a dystopia, es­pe­cially when it starts in­volv­ing ma­chine learn­ing on more and more data. Then it eats your time and your money and your re­la­tion­ships and ev­ery­thing else.

There is spec­u­la­tion that this is hap­pen­ing in many jobs. It used to be that one could “suit up” and oth­er­wise con­sume rote for­mal­ity, and use that as a sig­nal of se­ri­ous­ness and pro­fes­sion­al­ism. That has in­creas­ingly stopped work­ing, so we rely on other sig­nals, which turns over even worse ar­eas of our be­hav­ior to the sig­nal­ing mar­ket, forc­ing peo­ple to turn to cul­tural and class sig­nals, and other Bayesian ev­i­dence, eat­ing up much more of peo­ple’s re­sources and mak­ing it that much harder to move up in the world. Every lit­tle piece of body lan­guage and ev­ery tone of voice gets ob­sessed over, even more than by de­fault.

Which is ter­rible, but then again, I re­ally hate suits.

VII – Conversation

Con­ver­sa­tion isn’t about in­for­ma­tion is a strange claim. Of course con­ver­sa­tion is about in­for­ma­tion. That doesn’t mean there isn’t any­thing else go­ing on; the same ‘any­thing else’ is always go­ing on when peo­ple in­ter­act. In this case, I think ‘con­ver­sa­tion is about info’ is more im­por­tantly true than ‘con­ver­sa­tion is not (en­tirely) about info.’ Info is quite the im­por­tant McGuffin.

We can as­sume that ev­ery­one is calcu­lat­ing when it is to their benefit to share in­for­ma­tion, ver­sus when it is bet­ter to hoard it. When they should talk ver­sus when they should listen. So when I see state­ments like:

A full ac­count­ing will in­clude two other, much larger costs:

  1. The op­por­tu­nity cost of mo­nop­o­liz­ing in­for­ma­tion.

  2. The costs of ac­quiring the in­for­ma­tion in the first place.

In light of these costs, it seems a win­ning strat­egy would be to re­lax and play it safe, let­ting oth­ers do all the work to gather new in­for­ma­tion. If they’re just go­ing to share it with you any­way, as an act of al­tru­ism, why bother?

But that’s not the in­stinct we find in the hu­man an­i­mal. We aren’t lazy, greedy listen­ers. In­stead we’re both in­tensely cu­ri­ous and happy to share the fruits of our cu­ri­os­ity with oth­ers. … If speak­ers are giv­ing away lit­tle in­for­ma­tional “gifts” in ev­ery con­ver­sa­tion, what are they get­ting in re­turn?

Hello, en­tire the­sis of the book? Once again, we see the naive view of ‘I thought this was a nar­row, iso­lated ac­tion, I’d be shocked, shocked to see hu­mans try­ing to use it to max­i­mize their re­pro­duc­tive fit­ness in this es­tab­lish­ment.’ Then when a sim­ple one-ac­tive-path­way model is sug­gested, in this case a quid pro quo ar­range­ment of “I’ll share some­thing with you if you re­turn the fa­vor,” we are ‘puz­zled’ to find that it doesn’t fully ex­plain the rich­ness of hu­man be­hav­ior.

Hav­ing read Robin for many years, it feels like he’s al­most play­ing dumb. Or per­haps he’s con­tin­u­ously ex­pect­ing read­ers to for­get or dis­miss the cen­tral the­sis of both blog and book, and think hu­mans do things in iso­lated ways that don’t in­ter­act with the wider world and wider hid­den (and open) mo­tives.

What ex­cites me most about The Elephant in the Brain is that, hav­ing put the cen­tral point in book form, we can hope to take it as a given so we can get on with work­ing out the im­pli­ca­tions and ex­ten­sions, rather than con­stantly be­ing shocked at what we find in this es­tab­lish­ment.

So we have the fol­low­ing puz­zles:

Puz­zle 1: Peo­ple Don’t Keep Track of Con­ver­sa­tional Debts

Puz­zle 2: Peo­ple Are More Eager to Talk Than Listen

Puz­zle 3: The Cri­te­rion of Rele­vance (in gen­eral, what­ever we say needs to re­late to the topic or task at hand)

Puz­zle 4: Subop­ti­mal Ex­changes (when two peo­ple meet for the first time, they rarely talk about the most im­por­tant top­ics they know).

The book’s re­s­olu­tion to this, which it draws from Ge­offrey Miller’s The Mat­ing Mind and Jean-Louis Des­salles’ Why We Talk is ti­tled “Sex and Poli­tics” where it is sug­gested we should “stop look­ing at con­ver­sa­tion as an ex­change of in­for­ma­tion, and in­stead try to see the benefits of speak­ing as some­thing other than re­ceiv­ing more in­for­ma­tion later down the road.”

There’s that word again. In­stead. I would sug­gest ‘in ad­di­tion to’. If I share in­for­ma­tion with you it might benefit me in ways other than you giv­ing me back in­for­ma­tion of value in this con­ver­sa­tion. Well, sure. Like any other time I do any­thing that benefits some­one else, I’m go­ing to ac­cu­mu­late good­will and a sort of debt, and it’s likely that benefit­ing you benefits me, or that how you will act given this in­for­ma­tion will benefit me, and also there’s all the sig­nal­ing and other sec­ondary im­pli­ca­tions.

Miller and Des­salles fo­cus on the sig­nal­ing as­pect, in par­tic­u­lar show­ing off. Miller speaks of im­press­ing po­ten­tial mates, Des­salles of po­ten­tial al­lies. The book sug­gests the metaphor that we each carry around a back­pack full of tools, and when we pull one out oth­ers get a du­pli­cate for free if they don’t have one already. By pul­ling out tools that are valuable and rele­vant, I provide ev­i­dence of an ex­ten­sive and valuable tool set:

You want to know whether the ap­pli­cant is sharp or dull, plugged-in or out of the loop. You want to know the size and util­ity of the ap­pli­cant’s back­pack.

Every con­ver­sa­tion is like a (mu­tual) job in­ter­view, where each of us is “ap­ply­ing” for the role of friend, lover or leader.

One of the most im­por­tant tools is the re­spect and sup­port of oth­ers, and our pres­tige.

Our ob­ses­sion with news, it is noted, makes lit­tle sense based on news’ di­rect use­ful­ness, but makes perfect sense if it is about main­tain­ing the abil­ity to con­verse im­pres­sively about hot top­ics that will count as rele­vant. We are in­ter­ested be­cause we ex­pect ev­ery­one else to be in­ter­ested, and need to be seen as be­ing in the loop.

Mo­tives that seem im­por­tant to con­ver­sa­tion, and that help ex­plain the puz­zles, but that don’t get much men­tion, in­clude: Per­sua­sion, De­cep­tion, Ne­go­ti­a­tion, Fram­ing, Credit (first per­son to pull out the tool gets the credit for it), Value (rele­vant things are more likely to be valuable, es­pe­cially in con­text, and seek­ing them is likely why we’re hav­ing the con­ver­sa­tion), Un­der­stand­ing (want­ing to be un­der­stood), Steer­ing (where the con­ver­sa­tion goes, in­clud­ing ask­ing for what would be valuable to you), At­ten­tion, Peo­ple plain old en­joy talk­ing and shar­ing in­for­ma­tion, Re­ac­tion (how you re­act to in­for­ma­tion is valuable to me), Play and Prac­tice, and gen­er­ally all the mo­ti­va­tions of the elephant even when they aren’t be­ing ex­plic­itly named.

Also, are the puz­zles even ac­cu­rate?

Peo­ple can and do keep track of con­ver­sa­tional debt. If you tell me some­thing valuable, I owe you, and vice versa. That debt is like any other debt. Lit­tle gifts that cost us noth­ing we don’t track care­fully, big gifts that do cost us, es­pe­cially ones that were re­quested, we track more care­fully. And some­times be­ing al­lowed to talk is the gift. A given con­ver­sa­tion need not bal­ance, but the debts do ac­cu­mu­late, and they do mat­ter. That there isn’t an ex­act ledger is only a puz­zle for ab­surdly eco­nomic man.

In turn, if debt is be­ing tracked, be­ing ea­ger to talk rather than listen makes sense. If you have a chance to gain cred­its, you’d be ex­cited to do that, so you can spend those vir­tual ap­prox­i­mate cred­its later, or pay back debt you’ve in­curred.

The crite­rion of rele­vance is nec­es­sary for con­ver­sa­tion to be a con­ver­sa­tion rather than an ex­change of facts. Rele­vant things build upon pre­vi­ously said things in valuable ways, and are likely to be of higher value. If I talk about what we should have for din­ner and you tell me the cap­i­tal of Brazil, chances are that’s both not some­thing I es­pe­cially care about right now and also not helping.

Similarly sub­op­ti­mal ex­change, since info has rel­a­tive value based on con­text and what differ­ent peo­ple care about at differ­ent times. I don’t have a ‘most valuable piece of knowl­edge’ to go re­peat­ing to ev­ery­one. But yes, peo­ple would benefit greatly if we paid more at­ten­tion to talk­ing about more valuable top­ics, and ex­chang­ing more valuable in­for­ma­tion. I am re­minded of the idea of go­ing around ask­ing ev­ery­one “what are the most im­por­tant prob­lems in your field?” and then “why aren’t you work­ing on them?”

The chap­ter ends with some claims about re­search prefer­ences of­ten be­ing about im­pres­sive­ness and pres­ti­gious as­so­ci­a­tion, which few read­ing ei­ther this or the book would ar­gue against, other than as a mat­ter of de­gree.

VIII – Art

Art isn’t about in­sight. Art is about sac­ri­fic­ing re­sources to show that you can af­ford to, and a gen­eral-pur­pose fit­ness dis­play (and courtship dis­play). Thus we care deeply about who made the art, in what con­text, and how difficult and ex­pen­sive it was to cre­ate. Artists in­ten­tion­ally choose difficult mediums and sub­ject mat­ter. For ex­am­ple, ex­act re­pro­duc­tion was prized as art when it was hard, then dis­carded when re­pro­duc­tion be­came easy. We like im­prov over the­ater, and the­ater over movies, hold­ing qual­ity con­stant, be­cause of difficulty and also as­so­ci­a­tion with pres­ti­gious folk. Lob­ster was peas­ant food when plen­tiful, a lux­ury now that it is rare.

In turn, we care deeply about our abil­ity to dis­cern good art from bad, lest we be mis­led, and oth­ers see us as there­fore un­fit.

I buy all that. I won­der how deeply it is linked to the fact that Choices are Bad. Rais­ing difficulty through re­stric­tion while hold­ing re­sults con­stant equals greater satis­fac­tion. Plus re­stric­tions breed cre­ativity.

It also ex­plains why so much of what we call ‘art’ is so ter­rible. When in­sid­ers com­pete to do things other in­sid­ers rec­og­nize as difficult, you get mu­se­ums full of things I would pay to not pos­sess. You get re­spect paid to ‘pres­tige’ pic­tures no one en­joys. There is lots of great art and fine work that re­quires taste to en­joy, but there’s also lots of ‘art’ where the en­joy­ing is about how ab­stractly difficult it is, and I can­not stand the stuff. Or lob­ster. We were right the first time on that one.

I can ap­pre­ci­ate when con­text makes things difficult, but there’s also difficulty in figur­ing out the right easy thing to do, and the difficult thing that’s ugly and ter­rible is still ugly and ter­rible in my eyes.

And yet, I like im­prov and ap­pre­ci­ate the­ater and live mu­sic and live sports (even if I don’t con­sider it usu­ally worth the time, money and trou­ble enough to ac­tu­ally go too of­ten, al­though I think I’m mak­ing a mis­take not go­ing more). Part of that is the lack of dis­trac­tion, and that real life is su­per high re­s­olu­tion, and the chance to in­ter­act with oth­ers, and even the sense of be­ing there with the high sta­tus folk. A lot of it is the abil­ity to see skil­led peo­ple tackle re­stric­tions and added difficulty, but that’s still far from all.

The im­por­tant miss­ing el­e­ment is un­pre­dictabil­ity and non-op­ti­miza­tion. I love watch­ing perform­ers think on their feet, and strate­giz­ing with them. When things are ed­ited and con­figured for the best perfor­mances and ex­actly what sells best, they be­come pre­dictable and bland. When you know some­thing is go­ing to be ex­cit­ing, that’s not ex­cit­ing; part of what makes a great game great is not know­ing if it’s go­ing to turn into a rout. Watch­ing great im­prov is about it some­times fal­ling on its face.

IX – Charity

Any­one read­ing this is al­most cer­tainly fa­mil­iar with Effec­tive Altru­ism, and the idea that peo­ple are char­i­ta­ble more for the warm glow of feel­ing helpful and show­ing off that they’re helping, rather than ac­tu­ally do­ing good. At this point all of that is un­con­tro­ver­sial around these parts.

As care­ful read­ers of mine have no doubt noted, I have con­cerns about Effec­tive Altru­ism. One of those con­cerns is the ten­dency of such folk to be­come smug and act su­pe­rior, treat­ing those who give oth­er­wise as ‘not re­ally car­ing.’ This chap­ter comes off as quite smug in a way I worry is off-putting. Yes, our giv­ing cares about visi­bil­ity, and peer pres­sure, and prox­im­ity, and re­lata­bil­ity and mat­ing mo­tives. We are not sim­ply try­ing to do the great­est good for the great­est num­ber slash score all the utilons. As op­posed to the Effec­tive Altru­ists, who are build­ing a cul­ture that at­tempts to bet­ter line up what serves those other mo­tives with what ac­tu­ally helps peo­ple.

Which is great! But that does not mean oth­ers do not care. Nor does failure to op­ti­mize X mean one does not care about X; hu­mans are not au­to­mat­i­cally strate­gic. They don’t go around op­ti­miz­ing things. If you’re go­ing to ac­cuse them of not car­ing, don’t ac­cuse them of not car­ing about peo­ple, ac­cuse them of not car­ing about op­ti­miza­tion! Cause on that, they’re guilty guilty guilty.

Definitely don’t act like car­ing about prox­im­ity or re­lata­bil­ity makes you a bad per­son, let alone effec­tively guilty of neg­li­gent homi­cide. You don’t get or keep civ­i­liza­tion if you don’t care about prox­im­ity or re­lata­bil­ity, or listen too care­fully to Peter Singer, and even if that wasn’t a con­cern, hu­mans flat out do care about those things. At this point I con­sider the drown­ing child ar­gu­ment a Basilisk, and wish it was treated ac­cord­ingly: as some­thing memet­i­cally haz­ardous that ev­ery­one needs to over­come and defeat as part of their com­ing-of-age rit­u­als. It doesn’t even point out hypocrisy be­cause peo­ple read­ily ad­mit that life does not work like that.

X – Education

In­di­vi­d­ual stu­dents can ex­pect their in­comes to rise roughly 8 to 12 per­cent for each ad­di­tional year of school they com­plete. Na­tions, how­ever, can ex­pect their in­comes to rise by only 1 to 3 per­cent for each ad­di­tional year of school com­pleted by their cit­i­zens on av­er­age.

School is about sig­nal­ing, and learn­ing to sub­mit. We knew that already. Film at eleven.

For the full ver­sion, see Bryan Ca­plan’s up­com­ing book, The Case Against Ed­u­ca­tion.

One note I would make is about the sheep­skin effect, where the last year of a col­lege de­gree is much more valuable than pre­vi­ous years. There’s been some de­bate about this on­line lately be­tween Bryan Ca­plan and Noah Smith. I agree that this is largely a sig­nal­ing effect, with ‘com­pleted all eight terms’ much more im­pres­sive than ‘com­pleted seven of eight terms’ since you don’t know how many more terms the first stu­dent could have finished if nec­es­sary.

What the dis­cus­sion misses, it seems to me, is that only af­ter grad­u­a­tion do you know that the first three years were real. It is easy to be­come ‘a se­nior’ through com­ple­tion of a num­ber of cred­its, sav­ing the stuff they find hard­est for last or even be­ing in ter­rible shape to match up with grad­u­a­tion re­quire­ments. I strongly sus­pect that a lot of peo­ple who drop out in year four are much farther from finished than they would have you be­lieve.

XI – Medicine

Medicine is about con­spicu­ous car­ing. The point is to show your care by do­ing all the things, on a per­sonal and a so­cietal level. We want to sym­bol­i­cally ap­ply care, and know that such sym­bolic care has been ap­plied. We ac­tively pre­fer treat­ments with painful side effects, be­cause they show how much we care. We want to con­struct a nar­ra­tive where we did all that we could, and no one could blame us. There are lots of ways to im­prove pa­tient out­comes that the med­i­cal sys­tem shrugs off and ig­nores for decades.

In this model, the fact that much of our med­i­cal care ac­tu­ally works is mostly a co­in­ci­dence. When older med­i­cal sys­tems used care that was ac­tively harm­ful, the same be­hav­iors were mostly ob­served. On the mar­gin, our med­i­cal care does not work, and likely does harm, which is why health in­surance doesn’t lead to more phys­i­cal health (al­though not wor­ry­ing about health care bills im­proves fi­nan­cial and men­tal health). We would do well to spend far less on health care, es­pe­cially end of life care, while keep­ing core ser­vices that ac­tu­ally work like trauma care and vac­ci­na­tion.

Hav­ing run a med­i­cal start-up, I can re­port that medicine is even less about health than Robin makes it out to be. Even for them­selves, with no one in their lives pre­sent to sig­nal, peo­ple would rather con­sume the sym­bolic ver­sion of the thing. They care about how your re­search re­port looks, not about what it con­tains. They want to feel helped, not be helped. We fool­ishly had the hy­poth­e­sis that peo­ple cared about their health enough to think for them­selves. We re­jected that hy­poth­e­sis.

XII – Religion

Reli­gion is a strange case where “X is not about Y” is of­ten con­sid­ered ac­tively good news. God is not of­ten con­sid­ered a wor­thy cause (or even to ex­ist) around these parts. Com­mu­nity and so­cial sys­tems, on the other hand, are rec­og­nized as vi­tal by all right think­ing folk. Com­mu­nity pro­vides a com­mit­ment mechanism for com­mu­nity mem­bers and a way to cre­ate com­mon knowl­edge of pro-so­cial norms that make those com­mu­ni­ties work as places to live, sup­port fel­low mem­bers and raise fam­i­lies. It also serves as a badge, al­low­ing oth­ers both in and out of your com­mu­nity to trust you will abide by your com­mu­ni­ties’ stan­dards.

In­stead of mock­ing re­li­gion as an es­pe­cially false be­lief, the book sug­gests, we should rec­og­nize its strate­gic value in build­ing and re­in­forc­ing so­cial sys­tems, norms and com­mu­ni­ties. The ab­sur­dity of re­li­gious be­lief is even an ad­van­tage, en­sur­ing that the sig­nals sent are gen­uine signs of com­mit­ment. Reli­gious folk have more chil­dren, stay mar­ried longer, have more so­cial con­nec­tions, get and stay mar­ried more, tend to be hap­pier, and have many things athe­ist com­mu­ni­ties strug­gle at great cost to recre­ate. Is there re­ally that big a differ­ence other than de­gree, the book asks, be­tween the Mus­lim Hajj and the pledge of alle­giance, or even trav­el­ing to watch your fa­vorite sports team?

I see the ap­peal. I have looked at and adopted many Jewish prac­tices and rit­u­als for their mun­dane util­ity, with great suc­cess. The book points out this is com­mon, and many Jews con­tinue to fol­low the com­mu­nity norms de­spite be­ing Athe­ists.

This at­ti­tude is summed up in this di­rect ap­peal:

Fi­nally, we’d like to make a plea for some char­ity and hu­mil­ity, es­pe­cially from our athe­ist read­ers. It’s easy for non­be­liev­ers to de­ride su­per­nat­u­ral be­liefs as “delu­sions” or “harm­ful su­per­sti­tions”

Nev­er­the­less, we think peo­ple can gen­er­ally in­tuit what’s good for them, even if they don’t have an an­a­lyt­i­cal un­der­stand­ing of why it’s good for them. In par­tic­u­lar, they have a keen sense for their con­crete self-in­ter­est, for when things are work­ing out in their fa­vor ver­sus when they’re get­ting a raw deal. So when­ever ad­her­ents feel trapped or op­pressed by their re­li­gion, as many do, they’re prob­a­bly right. But in most times and places, peo­ple feel pow­er­fully at­tracted to re­li­gion. They con­tinue to par­ti­ci­pate, week af­ter week and year af­ter year-not with re­luc­tance but with tremen­dous zeal. And we’d like to give them the benefit of the doubt that they know what’s good for them.

If we are mostly run by our elephants, and our elephants are ex­pert Bayesian an­a­lysts with a keen un­der­stand­ing of what will help our re­pro­duc­tive fit­ness, that is a strong ar­gu­ment in fa­vor of peo­ple know­ing what is good for them. Cer­tainly peo­ple are tak­ing much into ac­count when think­ing about such mat­ters, in ways they can’t ar­tic­u­late and that peo­ple of­ten aren’t given credit for. In the other chap­ters of the book, we see peo­ple do­ing things that don’t seem to make sense, but turn out to be rather strate­gic. Peo­ple are re­ally good at this in some con­texts.

It’s even true that much of the time, if we ask peo­ple what is ‘good for them,’ they will inuit re­mark­ably ac­cu­rate an­swers, even iden­ti­fy­ing when their own be­hav­iors are bad for them, or other be­hav­iors they are not do­ing would be good. If you think peo­ple go around hav­ing no idea what is good for them, and that peo­ple do things for no good rea­son, this is an im­por­tant mes­sage for you.

The coun­ter­ar­gu­ment is that we’ve all met hu­mans, so we know that they of­ten have no idea what is good for them, in any con­text. It is not a co­in­ci­dence that re­li­gion is sin­gled out as the place the book ex­plic­itly asks us to trust oth­ers to make such de­ci­sions. Reli­gion is a place we’ve con­structed a norm that we should mostly trust peo­ple to make their own choices, un­less the re­li­gion is young or too weird, and we call it a cult. Then we rec­og­nize that peo­ple can be very wrong about what is good for them.

That’s not to say that re­li­gion (set­ting aside the truth value of its be­liefs) is usu­ally bad for peo­ple. It seems es­pe­cially not bad for peo­ple in the con­text of its ex­ist­ing com­mu­ni­ties. Given ev­ery­one around you is do­ing this thing, not do­ing the thing might be quite the bad idea.

What I am push­ing back against is the more gen­eral claim that peo­ple can in­tuit what is good for them, to the ex­tent that peo­ple know ac­cu­rately sense when their re­li­gion is or isn’t work­ing out. Reli­gion is mostly an adap­ta­tion, and re­li­gious be­liefs are of­ten real what­ever the origi­nal mo­ti­va­tion for se­lect­ing those be­liefs. Peo­ple are Adap­ta­tion-Ex­e­cuters, not Fit­ness-Max­i­miz­ers.

Usu­ally peo­ple aren’t try­ing to in­tuit any­thing at all, they’re just do­ing what they are in the habit of do­ing, and what they see oth­ers in the habit of do­ing, fol­low­ing the an­cient prin­ci­ples of fit­ting into a tribe, be­cause that’s typ­i­cally a recipe for lo­cal suc­cess. Often these strate­gies are much smarter than they look. But when what is good for peo­ple doesn’t match those in­stincts and habits, peo­ple don’t see it. Try­ing to out-think oth­ers, or in­ten­tion­ally co­or­di­nate with them, is a rough road.

XIII – Politics

Once again, we have puz­zles: Peo­ple vote with­out (much) re­gard for vote de­ci­sive­ness, they vote when they know they are un­in­formed, they have en­trenched opinions and strong emo­tions on poli­ti­cal is­sues. If we are poli­ti­cal ‘do-rights’ try­ing to en­act the best poli­cies, these ac­tions do not make sense. Even more puz­zling, we don’t seem to vote ac­cord­ing to our nar­row self-in­ter­ests ei­ther.

But as we all know, choos­ing the best poli­cies is not what most poli­tics is mostly about. Poli­tics is mostly about be­ing in coal­i­tions and show­ing loy­alty to that coal­i­tion. In many times and places, mem­bers of the poli­ti­cal out­group are not taken kindly to, so one needs to show loy­alty to the in­group and its poli­ti­cal view­points.

That doesn’t mean poli­tics isn’t ul­ti­mately largely about policy. Th­ese coal­i­tions in­volve many ac­tors who do care deeply about cer­tain poli­cies, of­ten out of nar­row self-in­ter­est but also of­ten as gen­uine do-rights. The policy wonks and ideal­ists are real, and views on is­sues of­ten do shift for the right rea­sons, not only the wrong ones. We all un­der­stand that a poli­tics com­pletely about al­li­ances would re­sult in the rapid col­lapse of the re­pub­lic, with dev­as­tat­ing con­se­quences for al­most ev­ery­one. Many fear that is well un­der­way.

So we re­ally do re­ward those who work for the com­mon good and pun­ish those who do not, in ad­di­tion to wor­ry­ing about al­li­ances. Others re­ward those who re­ward the com­mon good, and so forth. The rel­a­tive power work­ing to­wards policy ebbs and flows with the times, the cul­ture, the norms. Oc­ca­sion­ally this even re­sults in good policy. It hasn’t been an es­pe­cially good time for that, re­cently. But I am old enough to re­mem­ber a time when this was far more of a thing.

XIV – Scorecard

Chap­ter % Claims Ac­cepted Y= Z= X about Y more than Z?

Body Lan­guage 100% N/​A Info N/​A

Laugh­ter 100% Jokes Playful­ness No

Con­ver­sa­tion 50% Info Show­ing Off Yes

Con­sump­tion 50% You Con­spicu­ous­ness Yes

Art 100% In­sight Fit­ness Dis­play No

Char­ity 90% Helping Fuzzies/​Sig­nal­ing No

Ed­u­ca­tion 100% Learn­ing Sig­nal­ing No

Medicine 110% Health Con­spicu­ous Car­ing No

Reli­gion 90% God Com­mu­nity No

Poli­tics 100% Policy Alli­ances No

On av­er­age, this adds up to buy­ing about 90% of claims. For a book mak­ing so many bold claims, that’s very good.

Ranked from most about Y to least about Y:

Food isn’t about Nutrition

Be­d­rooms aren’t about Sleep

Talk isn’t about Info

Con­sump­tion isn’t about You /​ Clothes aren’t about Comfort

Re­search isn’t about Progress

Poli­tics isn’t about Policy

Con­sult­ing isn’t about Advice

Mar­riage isn’t about Romance

School isn’t about Learning

Medicine isn’t about Health

Art isn’t about Insight

Church isn’t about God

XV – Conclusion

X is about a lot more than Y. Often X is about Z far more. Even if X is mostly about Y on some lev­els, it is mostly about Z on oth­ers. But we fear pun­ish­ment and so­cial col­lapse if we claim or even be­lieve this.

This is a sort of ar­gu­ment for mod­esty in the ‘be­lieve and act like if you know what’s good for you’ sense – you should pro­fess the same be­liefs you see around you be­cause they are cho­sen strate­gi­cally, even if they’re false. You should do the same ac­tions you see, be­cause they have hid­den so­cial mo­tives and pur­poses, and peo­ple will pun­ish you for act­ing differ­ently even if they don’t know why act­ing differ­ently might be bad here – hold­ing out for that ex­pla­na­tion is not a chance they are will­ing to take. Nice hu­man you have there. It would be a shame if some­one were to os­tra­cize it or lower its sta­tus.

It’s also an ar­gu­ment that­ev­ery­one is ly­ing to you, all the time, and you know it. Dis­in­for­ma­tion is ev­ery­where, re­quiring lo­cal mod­el­ing. Trust­ing the ‘ex­pert con­sen­sus’ means trust­ing peo­ple who are ly­ing their asses off. All the time. Be­ing more ac­cu­rate than peo­ple who are ly­ing their asses off all the time sounds doable. Every­one is already do­ing it! Our elephants in­stinc­tively ad­just for this ly­ing. It also means ev­ery­one’s effort is mostly go­ing to other things, so again you should have a rel­a­tively easy time do­ing bet­ter. The ques­tion is how.

All this hypocrisy and self-de­cep­tion is how we got here and have all the big brains and nice things. Without it we’d be able to ac­com­plish things more effi­ciently, but we’d also lose our pro-so­cial norms and things would fall apart. So how do we ex­plic­itly ex­plore these con­cepts and get the benefits, with­out bring­ing the whole tower down on our heads?

Every­one pun­ishes those who are not hyp­ocrites in such ways. They also pun­ish any­one who is caught be­ing a hyp­ocrite. When peo­ple figure out that no one does the thing they are sup­pos­edly do­ing, and ev­ery­one who knows this is ly­ing about it, they al­most always con­clude there’s a norm to keep ly­ing about it, and start ly­ing them­selves lest they be torn apart by un­known en­force­ment mechanisms. How do we com­mu­ni­cate about such things ex­plic­itly with­out bring­ing this wrath down on us?

You just do it. A lit­tle. Then you re­al­ize noth­ing hap­pened. No one is com­ing for you.

So you do it some more. Noth­ing hap­pens. No one comes.

Un­til you shout it from the roof­tops, for years. And still noth­ing hap­pens.

Our in­stincts say such talk is dan­ger­ous, but it isn’t. What is dan­ger­ous are to speak against key norms, or the in-vogue sym­bols of group loy­alty. There are cer­tain things you ab­solutely can­not say with­out risk of dis­aster. They vary from place to place and time to time. Some are true, some are false. All of them are on the list be­cause enough peo­ple ac­tu­ally say them, or they are ev­i­dence of things peo­ple ac­tu­ally be­lieve.

So… don’t say them in the wrong com­pany. That sim­ple. You don’t need to. When they’re true, they’ll be im­plied. When they’re false, they won’t be.

No one ex­pects you to go meta on the situ­a­tion. Han­so­nian claims are fine. Vas­sar­ian claims are fine. Yud­kowskian claims are fine.

You just can’t be a hyp­ocrite about it.

Or rather, you ab­solutely must be a hyp­ocrite about it, but be care­ful to not be caught.

There’s no need to be a hyp­ocrite about be­ing a hyp­ocrite. Peo­ple are watch­ing you to see what will hap­pen if they press a but­ton and see what stock re­sponse you send. Will you pull out the slip of pa­per con­tain­ing the ap­pro­pri­ate an­swer? That’s what they are check­ing. They don’t care what your un­der­ly­ing log­i­cal al­gorithm is; that’s not very cor­re­lated with the slips of pa­per that come out when you press but­tons, whereas your loy­alty is highly cor­re­lated with those pa­per slips. If your logic says that what you write on those slips is about loy­alty, but you pull out the right slips of pa­per, does that make you less loyal? Or more loyal? I’m not even sure.

The CliqueBots roam the land to strike down those who do not say the Shib­bo­leth. They’re not look­ing for your logic, un­less it bumps into one of the for­bid­den moves.

So now that you can talk about it, now what? How do we make things bet­ter?

On a per­sonal level, we have a bet­ter world-model. We now know that medicine isn’t about health and poli­tics usu­ally isn’t about policy, and so on. So we can bet­ter un­der­stand what oth­ers will do, and how to mo­ti­vate them – we need to give them mo­ti­va­tions about con­spicu­ous car­ing and al­li­ances, with­out say­ing ex­plic­itly that we are do­ing that. We in turn can no­tice when our ac­tions are con­spicu­ous car­ing or al­li­ance for­ma­tion, and de­cide whether that’s a wise thing to do in con­text.

On a so­cietal level, we now un­der­stand why we can’t ‘just’ do the medicine that im­proves health, or the poli­tics that im­proves policy. We know more about why peo­ple are do­ing what they are do­ing, how the sys­tem ticks, what keeps it work­ing, what might break it down. We can more use­fully ask where to push on it, or how to com­pete against it. If we want to sell peo­ple on pre­dic­tion mar­kets, we can ask how we might frame them such that they might get buy-in. If we want to cut health care spend­ing, we can ask how to frame such cuts as car­ing rather than not car­ing. Per­haps, for ex­am­ple, we should more share hor­ror sto­ries about over-treat­ment at end of life, and pres­sure to use it, and ask how we could care so lit­tle as to al­low that. Use the sys­tem against it­self.

We can also rec­og­nize that the Zs in ques­tion are not bad things. There’s noth­ing wrong with sig­nal­ing, con­spicu­ous car­ing, form­ing al­li­ances, build­ing com­mu­ni­ties and dis­play­ing one’s fit­ness. Th­ese are hu­man needs. If we squeeze them out of one place, they’ll show up in an­other. So is search­ing for hyp­ocrites. It gives us mo­ti­va­tion to stay con­sis­tent; ra­tio­nal­ists might not need such mo­ti­va­tion, but most oth­ers do. Fear of look­ing hyp­o­crit­i­cal keeps them on their toes. We shouldn’t be the hyp­ocrites who con­demn such mo­tives, any more than we should be the ex­plicit hyp­ocrites who say we’re do­ing med­i­cal care to show car­ing and play­ing poli­tics to build al­li­ances, ex­cept in­stru­men­tally or ad­di­tion­ally. Let us not dis­re­spect the Y that X is not about, for it is about that too!

And this al­lows us to listen to the elephant, and co­op­er­ate with it. It knows more than you do, and faster. It brings wis­dom. We can and must be aware of it, guide it, train it, but also trust it, and play the game with ev­ery­one else. After en­light­en­ment, chop wood, carry wa­ter. But the good wood, and the right wa­ter.