Book Review: The Secret Of Our Success

Link post

[Pre­vi­ously in se­quence: Epistemic Learned Hel­pless­ness]


“Cul­ture is the se­cret of hu­man­ity’s suc­cess” sounds like the most va­pid pos­si­ble the­sis. The Se­cret Of Our Suc­cess by an­thro­pol­o­gist Joseph Hein­rich man­ages to be an amaz­ing book any­way.

Hein­rich wants to de­bunk (or at least clar­ify) a pop­u­lar view where hu­mans suc­ceeded be­cause of our raw in­tel­li­gence. In this view, we are smart enough to in­vent neat tools that help us sur­vive and adapt to un­fa­mil­iar en­vi­ron­ments.

Against such the­o­ries: we can­not ac­tu­ally do this. Hein­rich walks the reader through many sto­ries about Euro­pean ex­plor­ers ma­rooned in un­fa­mil­iar en­vi­ron­ments. Th­ese ex­plor­ers usu­ally starved to death. They starved to death in the mid­dle of end­less plenty. Some of them were in Arc­tic lands that the Inuit con­sid­ered among their rich­est hunt­ing grounds. Others were in jun­gles, sur­rounded by ed­ible plants and an­i­mals. One par­tic­u­larly un­for­tu­nate group was in Alabama, and would have per­ished en­tirely if they hadn’t been cap­tured and en­slaved by lo­cal In­di­ans first.

Th­ese ex­plor­ers had many ad­van­tages over our ho­minid an­ces­tors. For one thing, their ex­plo­ra­tion par­ties were made up en­tirely of strong young men in their prime, with no need to sup­port women, chil­dren, or the el­derly. They were of­ten se­lected for their ed­u­ca­tion and in­tel­li­gence. Many of them were from Vic­to­rian Bri­tain, one of the most suc­cess­ful civ­i­liza­tions in his­tory, full of ge­niuses like Dar­win and Gal­ton. Most of them had some past ex­pe­rience with wilder­ness craft and sur­vival. But de­spite their big brains, when faced with the task our big brains sup­pos­edly evolved for – figur­ing out how to do hunt­ing and gath­er­ing in a wilder­ness en­vi­ron­ment – they failed pa­thet­i­cally.

Nor is it sur­pris­ing that they failed. Hunt­ing and gath­er­ing is ac­tu­ally re­ally hard. Here’s Hein­rich’s de­scrip­tion of how the Inuit hunt seals:

You first have to find their breath­ing holes in the ice. It’s im­por­tant that the area around the hole be snow-cov­ered—oth­er­wise the seals will hear you and van­ish. You then open the hole, smell it to ver­ify it’s still in use (what do seals smell like?), and then as­sess the shape of the hole us­ing a spe­cial curved piece of caribou antler. The hole is then cov­ered with snow, save for a small gap at the top that is capped with a down in­di­ca­tor. If the seal en­ters the hole, the in­di­ca­tor moves, and you must blindly plunge your har­poon into the hole us­ing all your weight. Your har­poon should be about 1.5 me­ters (5ft) long, with a de­tach­able tip that is teth­ered with a heavy braid of sinew line. You can get the antler from the pre­vi­ously noted caribou, which you brought down with your drift­wood bow.

The rear spike of the har­poon is made of ex­tra-hard po­lar bear bone (yes, you also need to know how to
kill po­lar bears; best to catch them nap­ping in their dens). Once you’ve plunged your har­poon’s head into the seal, you’re then in a wrestling match as you reel him in, onto the ice, where you can finish him off with the afore­men­tioned bear-bone spike.

Now you have a seal, but you have to cook it. How­ever, there are no trees at this lat­i­tude for wood, and
drift­wood is too sparse and valuable to use rou­tinely for fires. To have a re­li­able fire, you’ll need to carve a lamp from soap­stone (you know what soap­stone looks like, right?), ren­der some oil for the lamp from blub­ber, and make a wick out of a par­tic­u­lar species of moss. You will also need wa­ter. The pack ice is frozen salt wa­ter, so us­ing it for drink­ing will just make you de­hy­drate faster. How­ever, old sea ice has lost most of its salt, so it can be melted to make potable wa­ter. Of course, you need to be able to lo­cate and iden­tify old sea ice by color and tex­ture. To melt it, make sure you have enough oil for your soap­stone lamp.

No sur­prise that stranded ex­plor­ers couldn’t figure all this out. It’s more sur­pris­ing that the Inuit did. And al­though the Arc­tic is an un­usu­ally hos­tile place for hu­mans, Hein­rich makes it clear that hunt­ing-gath­er­ing tech­niques of this level of com­plex­ity are stan­dard ev­ery­where. Here’s how the In­di­ans of Tierra del Fuego make ar­rows:

Among the Fue­gians, mak­ing an ar­row re­quires a 14-step pro­ce­dure that in­volves us­ing seven differ­ent tools to work six differ­ent ma­te­ri­als. Here are some of the steps:

– The pro­cess be­gins by se­lect­ing the wood for the shaft, which prefer­ably comes from chaura, a
bushy, ev­er­green shrub. Though strong and light, this wood is a non-in­tu­itive choice since the
gnarled branches re­quire ex­ten­sive straight­en­ing (why not start with straighter branches?).

– The wood is heated, straight­ened with the crafts­man’s teeth, and even­tu­ally finished with a
scraper. Then, us­ing a pre-heated and grooved stone, the shaft is pressed into the grooves and
rubbed back and forth, press­ing it down with a piece of fox skin. The fox skin be­comes
im­preg­nated with the dust, which pre­pares it for the pol­ish­ing stage (Does it have to be fox skin?).

– Bits of pitch, gath­ered from the beach, are chewed and mixed with ash (What if you don’t in­clude the ash?).

– The mix­ture is then ap­plied to both ends of a heated shaft, which must then be coated with white clay (what about red clay? Do you have to heat it?). This pre­pares the ends for the fletch­ing and ar­row­head.

– Two feathers are used for the fletch­ing, prefer­ably from up­land geese (why not chicken feathers?).

– Right-handed bow­man must use feathers from the left wing of the bird, and vice versa for lefties (Does this re­ally mat­ter?).

– The feathers are lashed to the shaft us­ing sinews from the back of the gua­naco, af­ter they are smoothed and thinned with wa­ter and sal­iva (why not sinews from the fox that I had to kill for the afore­men­tioned skin?).

Next is the ar­row­head, which must be crafted and then at­tached to the shaft, and of course there is also the bow, quiver and archery skills. But, I’ll leave it there, since I think you get the idea.

How do hunter-gath­er­ers know how to do all this? We usu­ally sum­ma­rize it as “cul­ture”. How did it form? Not through some smart Inuit or Fue­gian per­son rea­son­ing it out; if that had been it, smart Euro­pean ex­plor­ers should have been able to rea­son it out too.

The ob­vi­ous an­swer is “cul­tural evolu­tion”, but Hein­rich isn’t much bet­ter than any­one else at tak­ing the mys­tery out of this phrase. Trial and er­ror must have been in­volved, and less suc­cess­ful groups/​peo­ple imi­tat­ing the tech­niques of more suc­cess­ful ones. But is that re­ally a satis­fy­ing ex­pla­na­tion?

I found the chap­ter on lan­guage a helpful re­minder that we already ba­si­cally ac­cept some­thing like this is true. How did lan­guage get in­vented? I’m es­pe­cially in­ter­ested in this ques­tion be­cause of my brief in­ter­ac­tions with con­lang­ing com­mu­ni­ties – peo­ple who try to con­struct their own lan­guages as a hobby or as part of a fan­tasy uni­verse, like Tolk­ien did with Elvish. Most peo­ple are ter­rible at this; their lan­guages are ei­ther un­us­able, or ex­act clones of English. Only peo­ple who (like Tolk­ien) already have years of for­mal train­ing in lin­guis­tics can do a re­motely pass­able job. And you’re tel­ling me the origi­nal lan­guages were in­vented by cave­men? Surely there was no com­mit­tee of Proto-Indo-Euro­pean no­mads that voted on whether to have an in­flect­ing or ag­glu­ti­nat­ing tongue? Surely no­body ran out of their cave shout­ing “Eureka!” af­ter hav­ing dis­cov­ered the in­ter­jec­tion? We just kind of ac­cept that af­ter cave­men work­ing re­ally hard to com­mu­ni­cate with each other, even­tu­ally lan­guage – still one of the most com­pli­cated and im­pres­sive pro­duc­tions of the hu­man race – just sort of hap­pened.

Tak­ing the gen­er­a­tion of cul­ture as sec­ondary to this kind of mys­te­ri­ous pro­cess, Hein­rich turns to its trans­mis­sion. If cul­tural gen­er­a­tion hap­pens at a cer­tain rate, then the fidelity of trans­mis­sion de­ter­mines whether a given so­ciety ad­vances, stag­nates, or de­clines.

For Hein­rich, hu­mans started be­com­ing more than just an­other species of mon­key when we started trans­mit­ting cul­ture with high fidelity. Some an­thro­poligsts talk about the Machi­avel­lian In­tel­li­gence Hy­poth­e­sis – the the­ory that hu­mans evolved big brains in or­der to suc­ceed at so­cial manuev­er­ing and climb­ing dom­i­nance hi­er­ar­chies. Hein­rich coun­ters with his own Cul­tural In­tel­li­gence Hy­poth­e­sis – hu­mans evolved big brains in or­der to be able to main­tain things like Inuit seal hunt­ing tech­niques. Every­thing that sep­a­rates us from the apes is part of an evolu­tion­ary pack­age de­signed to help us main­tain this kind of cul­ture, ex­ploit this kind of cul­ture, or ad­just to the new abil­ities that this kind of cul­ture gave us.


Se­cret gives many ex­am­ples of many cul­ture-re­lated adap­ta­tions, and not all are in the brain.

One of the most im­por­tant differ­ences be­tween man and ape is our puny di­ges­tive tracts:

Our mouths are the size of the squir­rel mon­key’s, a species that weighs less than three pounds. Chim­panzees can open their mouths twice as ide as we can and hold sub­stan­tial amounts of food com­pressed be­tween their lips and large teeth. We also have puny jaw mus­cles that reach up only to just be­low our ears. Other pri­mates’ jaw mus­cles stretch to the top of their heads, where they some­times even latch onto a cen­tral bony ridge. Our stom­achs are small, hav­ing only a third of the sur­face area that we’d ex­pect for a pri­mate of our size, and our colons are too short, be­ing only 60% of their ex­pected mass.

Com­pared to other an­i­mals, we have such at­ro­phied di­ges­tive tracts that we shouldn’t be able to live. What saves us? All of our food pro­cess­ing tech­niques, es­pe­cially cook­ing, but also chop­ping, rins­ing, boiling, and soak­ing. We’ve done much of the work of di­ges­tion be­fore food even en­ters our mouths. Our cul­ture teaches us how to do this, both in broad terms like “hold things over fire to cook them” and in spe­cific terms like “this plant needs to be soaked in wa­ter for 24 hours to leach out the tox­ins”. Each cul­ture has its own cook­ing knowl­edge re­lated to the lo­cal plants and an­i­mals; a fre­quent cause of death among Euro­pean ex­plor­ers was cook­ing things in ways that didn’t un­lock any of the nu­tri­ents, and so starv­ing while ap­par­ently well-fed.

All of this is cul­tural. Hein­rich is kind of cruel in his in­sis­tence on this. He recom­mends read­ers go out­side and try to start a fire. He even gives some helpful hints – flint is in­volved, rub­bing two sticks to­gether works for some peo­ple, etc. He pre­dicts – and sto­ries I’ve heard from un­for­tu­nate cam­pers con­firm – that you will not be able to do this, de­spite an IQ far be­yond that of most of our ho­minid an­ces­tors. In fact, some groups (most no­tably the abo­rigi­nal Tas­ma­ni­ans) seem to have lost the abil­ity to make fire, and never re­dis­cov­ered it. Fire-mak­ing was dis­cov­ered a small num­ber of times, maybe once, and has been cul­turally trans­mit­ted since then.

And food pro­cess­ing tech­niques are even more com­pli­cated. Nix­ta­mal­iza­tion of corn, nec­es­sary to pre­vent vi­tamin defi­cien­cies, in­volves soak­ing the corn in a solu­tion con­tain­ing ground-up burnt seashells. The an­cient Mex­i­cans dis­cov­ered this and lived off corn just fine for mil­len­nia. When the con­quis­ta­dors took over, they ig­nored it and ate corn straight. For four hun­dred years, Euro­peans and Amer­i­cans ate un­nix­ta­mal­ized corn. By offi­cial statis­tics, three mil­lion Amer­i­cans came down with corn-re­lated vi­tamin defi­cien­cies dur­ing this time, and up to a hun­dred thou­sand died. It wasn’t un­til 1937 that Western sci­en­tists dis­cov­ered which vi­tam­ins were in­volved and de­vel­oped an in­dus­trial ver­sion of nix­ta­mal­iza­tion that made corn safe. Early 1900s Amer­i­cans were very smart and had lots of ad­van­tages over an­cient Mex­i­cans. But the an­cient Mex­i­cans’ cul­ture got this one right in a way it took Western­ers cen­turies to match.

Hu­mans are per­sis­tence hunters: they can­not run as fast as gazel­les, but they can keep run­ning for longer than gazel­les (or al­most any­thing else). Why did we evolve into that niche? The se­cret is our abil­ity to carry wa­ter. Every hunter-gath­erer cul­ture has in­vented its own wa­ter-car­ry­ing tech­niques, usu­ally some kind of wa­ter­skin. This al­lowed hu­mans to switch to per­spira­tion-based cool­ing sys­tems, which al­lowed them to run as long as they want.

And hu­mans are con­su­mate tool users. In some cases, we evolved in or­der to use tools bet­ter; our hands out­class those of any other ape in terms of fi­nesse. In other cases, we de­volved sys­tems that were no longer nec­es­sary once tools took over. We are vastly weaker than any other ape. Hein­rich de­scribes a cir­cus act of the 1940s where the ring­mas­ter would challenge strong men in the au­di­ence to wres­tle a ju­ve­nile chim­panzee. The chim­panzee was tied up, dressed in a mask that pre­vented it from bit­ing, and wear­ing soft gloves that pre­vented it from scratch­ing. No hu­man ever lasted more than five sec­onds. Our com­mon an­ces­tor with other apes grew weaker and weaker as we be­came more and more re­li­ant on ar­tifi­cial weapons to give us an ad­van­tage.


But most of our differ­ences from other apes are in­deed in the brain. They’re just not nec­es­sar­ily where you would ex­pect.

To­masello et al tested hu­man tod­dlers vs. apes on a se­ries of tra­di­tional IQ type ques­tions. The match-up was sur­pris­ingly fair; in ar­eas like mem­ory, logic, and spa­tial rea­son­ing, the three species did about the same. But in abil­ity to learn from an­other per­son, hu­mans wiped the floor with the other two ape species:

Re­mem­ber, Hein­rich thinks cul­ture ac­cu­mu­lates through ran­dom mu­ta­tion. Hu­mans don’t have con­trol over how cul­ture gets gen­er­ated. They have more con­trol over how much of it gets trans­mit­ted to the next gen­er­a­tion. If 100% gets trans­mit­ted, then as more and more mu­ta­tions ac­cu­mu­late, the cul­ture be­comes bet­ter and bet­ter. If less than 100% gets trans­mit­ted, then at some point new cul­ture gained and old cul­ture lost fall into equil­ibrium, and your so­ciety sta­bi­lizes at some higher or lower tech­nolog­i­cal level. This means that trans­mit­ting cul­ture to the next gen­er­a­tion is maybe the core hu­man skill. The hu­man brain is op­ti­mized to make this work as well as pos­si­ble.

Hu­man chil­dren are ob­sessed with learn­ing things. And they don’t learn things ran­domly. There seem to be “bi­ases in cul­tural learn­ing”, ie slots in an in­fant’s mind that they know need to be filled with knowl­edge, and which they prefer­en­tially seek out the knowl­edge nec­es­sary to fill.

One slot is for lan­guage. Hu­man chil­dren nat­u­rally listen to speech (as early as in the womb). They nat­u­rally prune the phonemes they are able to pro­duce and dis­t­in­guish to the ones in the lo­cal lan­guage. And they nat­u­rally figure out how to speak and un­der­stand what peo­ple are say­ing, even though learn­ing a lan­guage is hard even for smart adults.

Another slot is for an­i­mals. In a world where megafauna has been rel­e­gated to zoos, we still teach chil­dren their ABCs with “L is for lion” and “B is for bear”, and chil­dren still read pic­ture books about Mr. Frog and Mrs. Snake hold­ing tea par­ties. Hein­rich sug­gests that just as the young brain is hard-coded to want to learn lan­guage, so it is hard-coded to want to learn the lo­cal an­i­mal life (lit­tle boys’ ve­hi­cle ob­ses­sion may be a weird out­growth of this; buses and trains are the clos­est thing to lo­cal megafauna that most of them will en­counter).

Another slot is for plants:

To see this sys­tem in op­er­a­tion, let’s con­sider how in­fants re­spond to un­fa­mil­iar plants. Plants are loaded with prickly thorns, nox­ious oils, sting­ing net­tles and dan­ger­ous tox­ins, all ge­net­i­cally evolved to pre­vent an­i­mals like us from mess­ing with them. Given our species wide ge­o­graphic range and di­verse use of plants as foods, medicines and con­struc­tion ma­te­ri­als, we ought to be primed to both learn about plants and avoid their dan­gers. To ex­plore this idea in the lab, the psy­chol­o­gists An­nie Wertz and Karen Wynn first gave in­fants, who ranged in age from eight to eigh­teen months, an op­por­tu­nity to touch novel plants (basil and parsley) and ar­ti­facts, in­clud­ing both novel ob­jects and com­mon ones, like wooden spoons and small lamps.

The re­sults were strik­ing. Re­gard­less of age, many in­fants flatly re­fused to touch the plants at all. When they did touch them, they waited sub­stan­tially longer than they did with the ar­ti­facts. By con­trast, even with the novel ob­jects, in­fants showed none of this re­luc­tance. This sug­gests that well be­fore one year of age in­fants can read­ily dis­t­in­guish plants from other things, and are primed for cau­tion with plants. But, how do they get past this con­ser­va­tive pre­dis­po­si­tion?

The an­swer is that in­fants keenly watch what other peo­ple do with plants, and are only in­clined to touch or eat the plants that other peo­ple have touched or eaten. In fact, once they get the ‘go ahead’ via cul­tural learn­ing, they are sud­denly in­ter­ested in eat­ing plants. To ex­plore this, An­nie and Karen ex­posed in­fants to mod­els who both picked fruit from plants and also picked fruit-like things from an ar­ti­fact of similar size and shape to the plant. The mod­els put both the fruit and the fruit-like things in their mouths. Next, the in­fants were given a choice to go for the fruit (picked from the plant) or the fruit-like things picked from the ob­ject. Over 75% of the time the in­fants went for the fruit, not the fruit-like things, since they’d got­ten the ‘go ahead’ via cul­tural learn­ing.

As a check, the in­fants were also ex­posed to mod­els putting the fruit or fruit-like things be­hind their ears(not in their mouths). In this case, the in­fants went for the fruit or fruit-like things in equal mea­sure. It seems that plants are most in­ter­est­ing if you can eat them, but only if you have some cul­tural learn­ing cues that they aren’t toxic.

After An­nie first told me about her work while I was vis­it­ing Yale in 2013, I went home to test it on my 6-month-old son, Josh. Josh seemed very likely to over­turn An­nie’s hard em­piri­cal work, since he
im­me­di­ately grasped any­thing you gave him and put it rapidly in his mouth. Com­fortable in his mom’s
arms, I first offered Josh a novel plas­tic cube. He delighted in grap­ping it and shov­ing it di­rectly into his mouth, with­out any hes­i­ta­tion. Then, I offered him a sprig of arugula. He quickly grabbed it, but then paused, looked with cu­ri­ous un­cer­tainty at it, and then slowly let it fall from his hand while turn­ing to hug his mom.

It’s worth point­ing out how rich the psy­chol­ogy is here. Not only do in­fants have to rec­og­nize that plants are differ­ent from ob­jects of similar size, shape and color, but they need to cre­ate cat­e­gories for types of plants, like basil and parsley, and dis­t­in­guish ‘eat­ing’ from just ‘touch­ing’. It does them lit­tle good to code their ob­ser­va­tion of some­one eat­ing basil as ‘plants are good to eat’ since that might cause them to eat poi­sonous plants as well as basil. But, it also does them lit­tle good to nar­rowly code the ob­ser­va­tion as ‘that par­tic­u­lar sprig of basil is good to eat’ since that par­tic­u­lar sprig has just been eaten by the per­son they are watch­ing. This an­other con­tent bias in cul­tural learn­ing.

This ties into the more gen­eral phe­nomenon of figur­ing out what’s ed­ible. Most Western­ers learn in­sects aren’t ed­ible; some Asi­ans learn that they are. This feels deeper than just some­one tel­ling you in­sects aren’t ed­ible and you be­liev­ing them. When I was in Thailand, my guide offered me a gi­ant cricket, tel­ling me it was deli­cious. I be­lieved him when he said it was safe to eat, I even be­lieved him when he said it tasted good to him, but my con­di­tion­ing won out – I didn’t eat the cricket. There seems to be some pro­cess where a child’s brain learns what is and isn’t lo­cally ed­ible, then hard-codes it against fu­ture change.

(Or so they say; I’ve never been able to eat shrimp ei­ther.)

Another slot is for gen­der roles. By now we’ve all heard the sto­ries of pro­gres­sives who try to raise their chil­dren with­out any ex­po­sure to gen­der. Their failure has some­times been taken as ev­i­dence that gen­der is hard-coded. But it can’t be quite that sim­ple: some mod­ern gen­der roles, like girls = pink, are far from ob­vi­ous or uni­ver­sal. In­stead, it looks like chil­dren have a hard-coded slot that gen­der roles go into, work hard to figure out what the lo­cal gen­der roles are (even if their par­ents are try­ing to con­fuse them), then latch onto them and don’t let go.

In the Cul­tural In­tel­li­gence Hy­poth­e­sis, hu­mans live in obli­gate sym­bio­sis with a cul­ture. A brain with­out an as­so­ci­ated cul­ture is in­com­plete and not very use­ful. So the in­fant brain is adapted to seek out the im­por­tant as­pects of its lo­cal cul­ture al­most from birth and fill them into the ap­pro­pri­ate slots in or­der to be­come whole.


The next part of the book dis­cusses post-child­hood learn­ing. This plays an im­por­tant role in hunter-gath­erer tribes:

While hunters reach their peak strength and speed in their twen­ties, in­di­vi­d­ual hunt­ing suc­cess does not peak un­til around age 30, be­cause suc­cess de­pends more on know-how and re­fined skills than on phys­i­cal prowess.

This part of the book made most sense in the con­text of ex­am­ples like the Inuit seal-hunt­ing strat­egy which drove home just how com­pli­cated and difficult hunt­ing-gath­er­ing was. Think less “Boy Scouts” and more “PhD”; a prim­i­tive tribe­sper­son’s life re­quires mas­tery of var­i­ous com­pli­cated tech­nolo­gies and skills. And the differ­ence be­tween “mediocre hunter” and “great hunter” can be the differ­ence be­tween high sta­tus (and good mat­ing op­por­tu­ni­ties) and low sta­tus, or even be­tween life and death. Hunter-gath­er­ers re­ally want to learn the es­sen­tials of their hunter-gath­erer lifestyle, and learn­ing it is re­ally hard. Their heuris­tics are:

Learn from peo­ple who are good at things and/​or widely-re­spected. If you haven’t already read about the differ­ence be­tween dom­i­nance and pres­tige hi­er­ar­chies, check out Kevin Sim­ler’s blog post on the topic. Peo­ple will fear and obey au­thor­ity figures like kings and chief­tains, but they give a differ­ent kind of re­spect (“pres­tige”) to peo­ple who seem good at things. And since it’s hard to figure out who’s good at things (can a non-mu­si­cian who wants to start learn­ing mu­sic tell the differ­ence be­tween a merely good performer and one of the world’s best?) most peo­ple use the heuris­tic of re­spect­ing the peo­ple who other peo­ple re­spect. Once you iden­tify some­one as re­spect-wor­thy, you strongly con­sider copy­ing them in, well, ev­ery­thing:

To un­der­stand pres­tige as a so­cial phe­nomenon, it’s cru­cial to re­al­ize that it’s of­ten difficult to figure out what pre­cisely makes some­one suc­cess­ful. In mod­ern so­cieties, the suc­cess of a star NBA bas­ket­ball player might arise from his:

(1) in­ten­sive prac­tice in the offsea­son
(2) sneaker prefer­ence
(3) sleep sched­ule
(4) pre-game prayer
(5) spe­cial vi­tam­ins
(6) taste for carrots

Any or all of these might in­crease his suc­cess. A naïve learner can’t tell all the causal links be­tween an in­di­vi­d­ual’s prac­tices and his suc­cess. As a con­se­quence, learn­ers of­ten copy their cho­sen mod­els broadly across many do­mains. Of course, learn­ers may place more weight on do­mains that for one rea­son or other seem more causally rele­vant to the model’s suc­cess. This copy­ing of­ten in­cludes the model’s per­sonal habits or styles as well as their goals and mo­ti­va­tions, since these may be linked to their suc­cess. This “if in doubt, copy it” heuris­tic is one of the rea­sons why suc­cess in one do­main con­verts to in­fluence across a broad range of do­mains.

The im­mense range of celebrity en­dorse­ments in mod­ern so­cieties shows the power of pres­tige. For
ex­am­ple, NBA star Le­bron James, who went di­rectly from High School to the pros, gets paid mil­lions to
en­dorse State Farm In­surance. Though a stun­ning bas­ket­ball tal­ent, it’s un­clear why Mr. James is qual­ified to recom­mend in­surance com­pa­nies. Similarly, Michael Jor­dan fa­mously wore Hanes un­der­wear and ap­par­ently Tiger Woods drove Buicks. Bey­once’ drinks Pepsi (at least in com­mer­cials). What’s the con­nec­tion be­tween mu­si­cal tal­ent and sug­ary cola bev­er­ages?

Fi­nally, while new med­i­cal find­ings and pub­lic ed­u­ca­tional cam­paigns only grad­u­ally in­fluence women’s ap­proach to pre­ven­tive medicine, An­gelina Jolie’s sin­gle OP-ED in the New York Times, de­scribing her de­ci­sion to get a pre­ven­tive dou­ble mas­tec­tomy af­ter learn­ing she had the ‘faulty’ BRCA1 gene, flooded clinics from the U.K. to New Zealand with women seek­ing ge­netic screen­ings for breast can­cer. Thus, an un­wanted evolu­tion­ary side effect, pres­tige turns out to be worth mil­lions, and rep­re­sents a pow­er­ful and un­der­uti­lized pub­lic health tool.

Of course, this cre­ates the risk of pres­tige cas­cades, where some ir­rele­vant fac­tor (Hein­rich men­tions be­ing a re­al­ity show star) cat­a­pults some­one to fame, ev­ery­one talks about them, and you end up with Mug­geridge’s defi­ni­tion of a celebrity: some­one fa­mous for be­ing fa­mous.

Some of this makes more sense if you go back to the evolu­tion­ary roots, and imag­ine watch­ing the best hunter in your tribe to see what his se­cret is, or be­ing nice to him in the hopes that he’ll take you un­der his wing and teach you stuff.

(but if all this is true, shouldn’t pub­lic aware­ness cam­paigns that hire celebrity spokes­peo­ple be wild suc­cesses? Don’t they just as of­ten fail, re­gard­less of how fa­mous a bas­ket­ball player they can con­vince to lec­ture schoolchil­dren about how Win­ners Don’t Do Drugs?)

Learn from peo­ple who are like you. If you are a man, it is prob­a­bly a bad idea to learn fash­ion by ob­serv­ing women. If you are a ser­vant, it is prob­a­bly a bad idea to learn the rules of eti­quette by ob­serv­ing how the king be­haves. Peo­ple are nat­u­rally in­clined to learn from peo­ple more similar to them­selves.

Hein­rich ties this in to var­i­ous stud­ies show­ing that black stu­dents learn best from a black teacher, fe­male stu­dents from a fe­male teacher, et cetera.

Learn from old peo­ple. Hu­mans are al­most unique in hav­ing menopause; most an­i­mals keep re­pro­duc­ing un­til they die in late mid­dle-age. Why does evolu­tion want hu­mans to stick around with­out re­pro­duc­ing?

Be­cause old peo­ple have already learned the lo­cal cul­ture and can teach it to oth­ers. Hein­rich asks us to throw out any per­sonal ex­pe­rience we have of el­ders; we live in a rapidly-chang­ing world where an old per­son is prob­a­bly “be­hind the times”. But for most of his­tory, change hap­pened glacially slowly, and old peo­ple would have spent their en­tire lives ac­cu­mu­lat­ing rele­vant knowl­edge. Imag­ine a world where when a Sili­con Valley pro­gram­mer can’t figure out how to make his code run, he calls up his grand­father, who spent fifty years cod­ing apps for Google and knows ev­ery pro­gram­ming lan­guage in­side and out.

Some­times im­por­tant events only hap­pen once in a gen­er­a­tion. Hein­rich tells the story of an Aus­tralian abo­rigi­nal tribe fac­ing a mas­sive drought. No­body knew what to do ex­cept Par­alji, the tribe’s old­est man, who had lived through the last mas­sive drought and re­mem­bered where his own el­ders had told him to find the last-re­sort wa­ter­holes.

This same dy­namic seems to play out even in other species:

In 1993, a se­vere drought hit Tan­za­nia, re­sult­ing in the death of 20% of the Afri­can elephant calves in a pop­u­la­tion of about 200. This pop­u­la­tion con­tained 21 differ­ent fam­i­lies, each of which was led by a sin­gle ma­tri­arch. The 21 elephant fam­i­lies were di­vided into 3 clans, and each clan shared the same ter­ri­tory dur­ing the wet sea­son (so, they knew each other). Re­searchers study­ing these elephants have an­a­lyzed the sur­vival of the calves and found that fam­i­lies led by older ma­tri­archs suffered fewer deaths of their calves dur­ing this drought.

More­over, two of the three elephant clans un­ex­pect­edly left the park dur­ing the drought, pre­sum­ably in search of wa­ter, and both had much higher sur­vival rates than the one clan that stayed be­hind. It hap­pens that these se­vere droughts only hit about once ev­ery four to five decades, and the last one hit about 1960. After that, sadly, elephant poach­ing in the 1970’s kil­led off many of the elephants who would have been old enough in 1993 to re­call the 1960 drought. How­ever, it turns out that ex­actly one mem­ber of each of the two clans who left the park, and sur­vived more effec­tively, were old enough to re­call life in 1960. This sug­gests, that like Par­alji in the Aus­tralian desert, they may have re­mem­bered what to do dur­ing a se­vere drought, and led their groups to the last wa­ter re­fuges. In the clan who stayed be­hind, the old­est mem­ber was born in 1960, and so was too young to have re­called the last ma­jor drought.

More gen­er­ally, ag­ing elephant ma­tri­archs have a big im­pact on their fam­i­lies, as those led by older ma­tri­archs do bet­ter at iden­ti­fy­ing and avoid­ing preda­tors (li­ons and hu­mans), avoid­ing in­ter­nal con­flicts and iden­ti­fy­ing the calls of their fel­low elephants. For ex­am­ple, in one set of field ex­per­i­ments, re­searchers played lion roars from both male and fe­male li­ons, and from ei­ther a sin­gle lion or a trio of li­ons. For elephants, male li­ons are much more dan­ger­ous than fe­males, and of course, three li­ons are always worse than only one lion. All the elephants gen­er­ally re­sponded with more defen­sive prepa­ra­tions when they heard three li­ons vs. one. How­ever, only the older ma­tri­archs keenly rec­og­nized the in­creased dan­gers of male li­ons over fe­male li­ons, and re­sponded to the in­creased threat with elephant defen­sive ma­neu­vers.


I was in­spired to read Se­cret by this re­view on Scholar’s Stage. I hate to be un­o­rigi­nal, but af­ter read­ing the whole book, I agree that the three sec­tions Tan­ner cites – on div­ina­tion, on man­ioc, and on shark taboos – are by far the best and most fas­ci­nat­ing.

On div­ina­tion:

When hunt­ing caribou, Naskapi for­agers in Labrador, Canada, had to de­cide where to go. Com­mon sense might lead one to go where one had suc­cess be­fore or to where friends or neigh­bors re­cently spot­ted caribou.

How­ever, this situ­a­tion is like [the Match­ing Pen­nies game]. The caribou are mis­match­ers and the hunters are match­ers. That is, hunters want to match the lo­ca­tions of caribou while caribou want to mis­match the hunters, to avoid be­ing shot and eaten. If a hunter shows any bias to re­turn to pre­vi­ous spots, where he or oth­ers have seen caribou, then the caribou can benefit (sur­vive bet­ter) by avoid­ing those lo­ca­tions (where they have pre­vi­ously seen hu­mans). Thus, the best hunt­ing strat­egy re­quires ran­dom­iz­ing.

Can cul­tural evolu­tion com­pen­sate for our cog­ni­tive in­ad­e­qua­cies? Tra­di­tion­ally, Naskapi hunters de­cided where to go to hunt us­ing div­ina­tion and be­lieved that the shoulder bones of caribou could point the way to suc­cess. To start the rit­ual, the shoulder blade was heated over hot coals in a way that caused pat­terns of cracks and burnt spots to form. This pat­tern­ing was then read as a kind of map, which was held in a pre-speci­fied ori­en­ta­tion. The crack­ing pat­terns were (prob­a­bly) es­sen­tially ran­dom from the point of view of hunt­ing lo­ca­tions, since the out­comes de­pended on myr­iad de­tails about the bone, fire, am­bi­ent tem­per­a­ture, and heat­ing pro­cess. Thus, these div­ina­tion rit­u­als may have pro­vided a crude ran­dom­iz­ing de­vice that helped hunters avoid their own de­ci­sion-mak­ing bi­ases.

This is not some ob­scure, iso­lated prac­tice, and other cases of div­ina­tion provide more ev­i­dence. In In­done­sia, the Kan­tus of Kal­i­man­tan use bird au­gury to se­lect lo­ca­tions for their agri­cul­tural plots. Geog­ra­pher Michael Dove ar­gues that two fac­tors will cause farm­ers to make plot place­ments that are too risky. First, Kantu ecolog­i­cal mod­els con­tain the Gam­bler’s Fal­lacy, and lead them to ex­pect floods to be less likely to oc­cur in a spe­cific lo­ca­tion af­ter a big flood in that lo­ca­tion (which is not true). Se­cond…Kan­tus pay at­ten­tion to oth­ers’ suc­cess and copy the choices of suc­cess­ful house­holds, mean­ing that if one of their neigh­bors has a good yield in an area one year, many other peo­ple will want to plant there in the next year. To re­duce the risks posed by these cog­ni­tive and de­ci­sion-mak­ing bi­ases, Kantu rely on a sys­tem of bird au­gury that effec­tively ran­dom­izes their choices for lo­cat­ing gar­den plots, which helps them avoid catas­trophic crop failures. Div­ina­tion re­sults de­pend not only on see­ing a par­tic­u­lar bird species in a par­tic­u­lar lo­ca­tion, but also on what type of call the bird makes (one type of call may be fa­vor­able, and an­other un­fa­vor­able).

The pat­tern­ing of bird au­gury sup­ports the view that this is a cul­tural adap­ta­tion. The sys­tem seems to have evolved and spread through­out this re­gion since the 17th cen­tury when rice cul­ti­va­tion was in­tro­duced. This makes sense, since it is rice cul­ti­va­tion that is most pos­i­tively in­fluenced by ran­dom­iz­ing gar­den lo­ca­tions. It’s pos­si­ble that, with the in­tro­duc­tion of rice, a few farm­ers be­gan to use bird sight­ings as an in­di­ca­tion of fa­vor­able gar­den sites. On-av­er­age, over a life­time, these farm­ers would do bet­ter – be more suc­cess­ful – than farm­ers who re­lied on the Gam­bler’s Fal­lacy or on copy­ing oth­ers’ im­me­di­ate be­hav­ior. What­ever the pro­cess, within 400 years, the bird au­gury sys­tem spread through­out the agri­cul­tural pop­u­la­tions of this Bor­neo re­gion. Yet, it re­mains con­spicu­ously miss­ing or un­der­de­vel­oped among lo­cal for­ag­ing groups and re­cent adopters of rice agri­cul­ture, as well as among pop­u­la­tions in north­ern Bor­neo who rely on ir­ri­ga­tion. So, bird au­gury has been sys­tem­at­i­cally spread­ing in those re­gions where it’s most adap­tive.

Scott Aaron­son has writ­ten about how easy it is to pre­dict peo­ple try­ing to “be ran­dom”:

In a class I taught at Berkeley, I did an ex­per­i­ment where I wrote a sim­ple lit­tle pro­gram that would let peo­ple type ei­ther “f” or “d” and would pre­dict which key they were go­ing to push next. It’s ac­tu­ally very easy to write a pro­gram that will make the right pre­dic­tion about 70% of the time. Most peo­ple don’t re­ally know how to type ran­domly. They’ll have too many al­ter­na­tions and so on. There will be all sorts of pat­terns, so you just have to build some sort of prob­a­bil­is­tic model. Even a very crude one will do well. I couldn’t even beat my own pro­gram, know­ing ex­actly how it worked. I challenged peo­ple to try this and the pro­gram was get­ting be­tween 70% and 80% pre­dic­tion rates. Then, we found one stu­dent that the pro­gram pre­dicted ex­actly 50% of the time. We asked him what his se­cret was and he re­sponded that he “just used his free will.”

But be­ing gen­uinely ran­dom is im­por­tant in pur­su­ing mixed game the­o­retic strate­gies. Hein­rich’s view is that div­ina­tion solved this prob­lem effec­tively.

I’m re­minded of the Ro­mans us­ing au­gury to de­cide when and where to at­tack. This always struck me as crazy; gen­er­als are go­ing to risk the lives of thou­sands of sol­diers be­cause they saw a weird bird ear­lier that morn­ing? But war is a clas­sic ex­am­ple of when a ran­dom strat­egy can be use­ful. If you’re de­cid­ing whether to at­tack the en­emy’s right vs. left flank, it’s im­por­tant that the en­emy can’t pre­dict your de­ci­sion and send his best defen­ders there. If you’re gen­er­ally pre­dictable – and Scott Aaron­son says you are – then out­sourc­ing your de­ci­sion to weird birds might be the best way to go.

And then there’s man­ioc. This is a tu­ber na­tive to the Amer­i­cas. It con­tains cyanide, and if you eat too much of it, you get cyanide poi­son­ing. From Hein­rich:

In the Amer­i­cas, where man­ioc was first do­mes­ti­cated, so­cieties who have re­lied on bit­ter va­ri­eties for thou­sands of years show no ev­i­dence of chronic cyanide poi­son­ing. In the Colom­bian Ama­zon, for ex­am­ple, in­dige­nous Tukanoans use a mul­ti­step, mul­ti­day pro­cess­ing tech­nique that in­volves scrap­ing, grat­ing, and fi­nally wash­ing the roots in or­der to sep­a­rate the fiber, starch, and liquid. Once sep­a­rated, the liquid is boiled into a bev­er­age, but the fiber and starch must then sit for two more days, when they can then be baked and eaten. Figure 7.1 shows the per­centage of cyanogenic con­tent in the liquid, fiber, and starch re­main­ing through each ma­jor step in this pro­cess­ing.

Such pro­cess­ing tech­niques are cru­cial for liv­ing in many parts of Ama­zo­nia, where other crops are difficult to cul­ti­vate and of­ten un­pro­duc­tive. How­ever, de­spite their util­ity, one per­son would have a difficult time figur­ing out the detox­ifi­ca­tion tech­nique. Con­sider the situ­a­tion from the point of view of the chil­dren and ado­les­cents who are learn­ing the tech­niques. They would have rarely, if ever, seen any­one get cyanide poi­son­ing, be­cause the tech­niques work. And even if the pro­cess­ing was in­effec­tive, such that cases of goi­ter (swol­len necks) or neu­rolog­i­cal prob­lems were com­mon, it would still be hard to rec­og­nize the link be­tween these chronic health is­sues and eat­ing man­ioc. Most peo­ple would have eaten man­ioc for years with no ap­par­ent effects. Low cyanogenic va­ri­eties are typ­i­cally boiled, but boiling alone is in­suffi­cient to pre­vent the chronic con­di­tions for bit­ter va­ri­eties. Boiling does, how­ever, re­move or re­duce the bit­ter taste and pre­vent the acute symp­toms (e.g., di­ar­rhea, stom­ach trou­bles, and vom­it­ing).

So, if one did the com­mon-sense thing and just boiled the high-cyanogenic man­ioc, ev­ery­thing would seem fine. Since the mul­ti­step task of pro­cess­ing man­ioc is long, ar­du­ous, and bor­ing, stick­ing with it is cer­tainly non-in­tu­itive. Tukanoan women spend about a quar­ter of their day detox­ify­ing man­ioc, so this is a costly tech­nique in the short term. Now con­sider what might re­sult if a self-re­li­ant Tukanoan mother de­cided to drop any seem­ingly un­nec­es­sary steps from the pro­cess­ing of her bit­ter man­ioc. She might crit­i­cally ex­am­ine the pro­ce­dure handed down to her from ear­lier gen­er­a­tions and con­clude that the goal of the pro­ce­dure is to re­move the bit­ter taste. She might then ex­per­i­ment with al­ter­na­tive pro­ce­dures by drop­ping some of the more la­bor-in­ten­sive or time-con­sum­ing steps. She’d find that with a shorter and much less la­bor-in­ten­sive pro­cess, she could re­move the bit­ter taste. Adopt­ing this eas­ier pro­to­col, she would have more time for other ac­tivi­ties, like car­ing for her chil­dren. Of course, years or decades later her fam­ily would be­gin to de­velop the symp­toms of chronic cyanide poi­son­ing.

Thus, the un­will­ing­ness of this mother to take on faith the prac­tices handed down to her from ear­lier gen­er­a­tions would re­sult in sick­ness and early death for mem­bers of her fam­ily. In­di­vi­d­ual learn­ing does not pay here, and in­tu­itions are mis­lead­ing. The prob­lem is that the steps in this pro­ce­dure are causally opaque—an in­di­vi­d­ual can­not read­ily in­fer their func­tions, in­ter­re­la­tion­ships, or im­por­tance. The causal opac­ity of many cul­tural adap­ta­tions had a big im­pact on our psy­chol­ogy.

Wait. Maybe I’m wrong about man­ioc pro­cess­ing. Per­haps it’s ac­tu­ally rather easy to in­di­vi­d­u­ally figure out the detox­ifi­ca­tion steps for man­ioc? For­tu­nately, his­tory has pro­vided a test case. At the be­gin­ning of the sev­en­teenth cen­tury, the Por­tuguese trans­ported man­ioc from South Amer­ica to West Africa for the first time. They did not, how­ever, trans­port the age-old in­dige­nous pro­cess­ing pro­to­cols or the un­der­ly­ing com­mit­ment to us­ing those tech­niques. Be­cause it is easy to plant and pro­vides high yields in in­fer­tile or drought-prone ar­eas, man­ioc spread rapidly across Africa and be­came a sta­ple food for many pop­u­la­tions. The pro­cess­ing tech­niques, how­ever, were not read­ily or con­sis­tently re­gen­er­ated. Even af­ter hun­dreds of years, chronic cyanide poi­son­ing re­mains a se­ri­ous health prob­lem in Africa. De­tailed stud­ies of lo­cal prepa­ra­tion tech­niques show that high lev­els of cyanide of­ten re­main and that many in­di­vi­d­u­als carry low lev­els of cyanide in their blood or urine, which haven’t yet man­i­fested in symp­toms. In some places, there’s no pro­cess­ing at all, or some­times the pro­cess­ing ac­tu­ally in­creases the cyanogenic con­tent. On the pos­i­tive side, some Afri­can groups have in fact cul­turally evolved effec­tive pro­cess­ing tech­niques, but these tech­niques are spread­ing only slowly.

Ra­tion­al­ists always won­der: how come peo­ple aren’t more ra­tio­nal? How come you can prove a thou­sand times, us­ing Facts and Logic, that some­thing is stupid, and yet peo­ple will still keep do­ing it?

Hein­rich hints at an an­swer: for ba­si­cally all of his­tory, us­ing rea­son would get you kil­led.

A rea­son­able per­son would have figured out there was no way for or­a­cle-bones to ac­cu­rately pre­dict the fu­ture. They would have aban­doned div­ina­tion, failed at hunt­ing, and maybe died of star­va­tion.

A rea­son­able per­son would have asked why ev­ery­one was wast­ing so much time prepar­ing man­ioc. When told “Be­cause that’s how we’ve always done it”, they would have been un­satis­fied with that an­swer. They would have done some ex­per­i­ments, and found that a sim­pler pro­cess of boiling it worked just as well. They would have saved lots of time, maybe con­verted all their friends to the new and eas­ier method. Twenty years later, they would have got­ten sick and died, in a way so causally dis­tant from their de­ci­sion to change man­ioc pro­cess­ing meth­ods that no­body would ever have been able to link the two to­gether.

Hein­rich dis­cusses preg­nancy taboos in Fiji; preg­nant women are banned from eat­ing sharks. Sure enough, these sharks con­tain chem­i­cals that can cause birth defects. The women didn’t re­ally know why they weren’t eat­ing the sharks, but when an­thro­pol­o­gists de­manded a rea­son, they even­tu­ally de­cided it was be­cause their ba­bies would be born with shark skin rather than hu­man skin. As ex­pla­na­tions go, this leaves a lot to be de­sired. How come you can still eat other fish? Aren’t you wor­ried your kids will have scales? Doesn’t the slight­est fa­mil­iar­ity with biol­ogy prove this mechanism is garbage? But if some smart in­de­pen­dent-minded icon­o­clas­tic Fijian girl figured any of this out, she would break the taboo and her child would have birth defects.

In giv­ing hu­mans rea­son at all, evolu­tion took a huge risk. Surely it must have wished there was some other way, some path that made us big-brained enough to un­der­stand tra­di­tion, but not big-brained enough to ques­tion it. Maybe it searched for a mind de­sign like that and couldn’t find one. So it was left with this tick­ing time-bomb, this ape that was con­stantly go­ing to be able to con­vince it­self of hare-brained and prob­a­bly-fatal ideas.

Here, too, cul­ture came to the res­cue. One of the most im­por­tant parts of any cul­ture – more im­por­tant than the tech­niques for hunt­ing seals, more im­por­tant than the tech­niques for pro­cess­ing tu­bers – is tech­niques for mak­ing sure no­body ever ques­tions tra­di­tion. Like the be­lief that any­one who doesn’t con­form is prob­a­bly a witch who should be cast out lest they bring de­struc­tion upon ev­ery­body. Or the be­lief in a God who has com­manded cer­tain spe­cific weird dietary re­stric­tions, and will tor­ture you for­ever if you dis­agree. Or the fairy tales where the prince asks a wiz­ard for help, and the wiz­ard says “You may have ev­ery­thing you wish for­ever, but you must never nod your head at a bad­ger”, and then one day the prince nods his head at a bad­ger, and his whole em­pire col­lapses into dust, and the moral of the story is that you should always obey weird ad­vice you don’t un­der­stand.

There’s a mon­ster at the end of this book. Hu­mans evolved to trans­mit cul­ture with high fidelity. And one of the biggest threats to trans­mit­ting cul­ture with high fidelity was Rea­son. Our an­ces­tors lived in Epistemic Hell, where they had to con­stantly rely on causally opaque pro­cesses with jus­tifi­ca­tions that couldn’t pos­si­bly be true, and if they ever ques­tioned them then they might die. His­tor­i­cally, Rea­son has been the villain of the hu­man nar­ra­tive, a cor­ro­sive force that tempts peo­ple away from adap­tive be­hav­ior to­wards choices that “sounded good at the time”.

Why are peo­ple so bad at rea­son­ing? For the same rea­son they’re so bad at let­ting poi­sonous spi­ders walk all over their face with­out freak­ing out. Both “skills” are re­ally bad ideas, most of the peo­ple who tried them died in the pro­cess, so evolu­tion re­moved those genes from the pop­u­la­tion, and suc­cess­ful cul­tures stig­ma­tized them enough to give peo­ple an in­ter­nal­ized fear of even try­ing.


This book be­longs alongside See­ing Like A State and the works of G.K. Ch­ester­ton as at­tempts to jus­tify tra­di­tion, and to ar­gue for or­gan­i­cally-evolved in­sti­tu­tions over top-down plan­ning. What unique con­tri­bu­tion does it make to this canon?

First, a lot more speci­fi­cally an­thro­polog­i­cal /​ pa­le­oan­thro­polog­i­cal rigor than the other two.

Se­cond, a much crisper fo­cus: Ch­ester­ton had only the fuzziest idea that he was writ­ing about cul­tural evolu­tion, and Scott was only a lit­tle clearer. I think Hein­rich is the only one of the three to use the term, and once you hear it, it’s ob­vi­ously the right fram­ing.

Third, a sense of how tra­di­tions con­tain the meta-tra­di­tion of defend­ing them­selves against Rea­son, and a sense for why this is nec­es­sary.

And fourth, maybe we’re not at the point where we re­ally want unique con­tri­bu­tions yet. Maybe we’re still at the point where we have to have this ham­mered in by more and more ex­am­ples. The temp­ta­tion is always to say “Ah, yes, a few sim­ple things like taboos against eat­ing poi­sonous plants may be re­lics of cul­tural evolu­tion, but ob­vi­ously by now we’re at the point where we know which tra­di­tions are im­por­tant vs. ran­dom looni­ness, and we can ra­tio­nally stick to the im­por­tant ones while throw­ing out the garbage.” And then some­body points out to you that ac­tu­ally div­ina­tion us­ing or­a­cle bones was one of the im­por­tant tra­di­tions, and if you thought you knew bet­ter than that and tried to throw it out, your civ­i­liza­tion would falter.

Maybe we just need to keep read­ing more similarly-themed books un­til this point re­ally sinks in, and we get prop­erly wor­ried.