The Algernon Argument

Why most supplements fail: IQ improvement skepticism, Yudkowsky & Bostrom’s heuristics, nootropics
biology, psychology, nootropics, transhumanism, IQ, insight-porn
2010-03-232018-06-03 finished certainty: highly likely importance: 10

That intel­li­gence () in healthy peo­ple is nearly impos­si­ble to improve is clear from the fail­ure of psy­chol­ogy to pro­vide any such method. But why intel­li­gence would be so con­stant is not as clear: many other cog­ni­tive abil­i­ties are improv­able (like work­ing mem­o­ry), so why not intel­li­gence?

Arthur Jensen the fail­ure of inter­ven­tions in the 1960s, and the fail­ure remains com­plete now, half a cen­tury lat­er: if you are a bright healthy young man or woman gifted with an IQ in the 130s, there is noth­ing you can do to increase your under­ly­ing intel­li­gence by a stan­dard devi­a­tion. New meth­ods like or are trum­peted in the media, and years later are dis­cov­ered to increase moti­va­tion & not intel­li­gence, or to have been over­stat­ed, or work only in dam­aged or , or to be statistical/, or to be tan­ta­mount to train­ing on IQ tests them­selves which destroys their mean­ing (like mem­o­riz­ing vocab­u­lary), or to be so anom­alous as to verge on fraud­u­lent (like the Pyg­malion effect). The only ques­tion worth ask­ing is which of these expla­na­tions is the real expla­na­tion this time.

For IQ in par­tic­u­lar, peo­ple dis­cussing human-en­hance­ment (espe­cially ) have pro­posed a pes­simistic obser­va­tion & evo­lu­tion­ary expla­na­tion, dubbed the “Alger­non prin­ci­ple” or “Alger­non’s law” or my pref­er­ence, the “Alger­non Argu­ment”.

Algernon Argument

The famous SF story “” pos­tu­lates surgery which triples the IQ score of the retarded pro­tag­o­nist - but which comes with the dev­as­tat­ing side-­ef­fects of the gain being both tem­po­rary and some­times fatal; fic­tional evi­dence aside, it is curi­ous that despite the incred­i­ble progress mankind has made in count­less areas like build­ing cars or going to the moon or fight­ing can­cer or extinct­ing small­pox or invent­ing com­put­ers or arti­fi­cial intel­li­gence, we lack any mean­ing­ful way to pos­i­tively affect peo­ple’s intel­li­gence beyond cur­ing dis­eases & defi­cien­cies. If we com­pare the smartest peo­ple in the world now like to the smartest peo­ple of more than half a cen­tury ago like , there seems to be lit­tle dif­fer­ence. expands the thought out in his essay “Alger­non’s Law”, stat­ing it as:

Any sim­ple major enhance­ment to human intel­li­gence is a net evo­lu­tion­ary dis­ad­van­tage.

The les­son is that Mother Nature know best. Or alter­nate­ly, : “there ain’t ”.

Trade-offs are endemic in biol­o­gy. Any­thing which isn’t car­ry­ing its own weight will be elim­i­nated - organs which are no longer used will be and within a life­time, unused mus­cles & bones will start weak­en­ing or being scav­enged for resources, as ath­letes1 and the hard way2 and body­builders per­pet­u­ally fight3, while shrews cycli­cally shrink their brains & skulls by 15% to con­serve resources in win­ter (Lázaro et al 2017). Often, if you use a drug or surgery to opti­mize some­thing, you will dis­cover penal­ties else­where. If you delay aging & length lifes­pan as is pos­si­ble in many species, you might find that you have encour­aged can­cer or - still worse - decreased repro­duc­tion4 as evi­denced by the of or brown antech­i­nus5; if your immune sys­tem goes all-out against dis­ease, you either deplete your ener­getic and chem­i­cal reserves6 or risk autoim­mune dis­or­ders; sim­i­lar­ly, we heal much slower than seems pos­si­ble despite the clear advan­tage7; if you try to enhance atten­tion with an amphet­a­mine, you destroy cre­ativ­i­ty, or if the amphet­a­mines reduce sleep, you dam­age mem­ory con­sol­i­da­tion or periph­eral aware­ness8; or improv­ing mem­ory (which requires active effort to main­tain9) also increases sen­si­tiv­ity to pain10 and inter­feres with other men­tal tasks1112 (as increased WM does, slightly13); if a mouse invests in anti-ag­ing cel­lu­lar repairs, it may freeze to death14, and so on. (What are we to make of induc­ing savan­t-­like abil­i­ties by brute-­force sup­pres­sion of brain regions15, or improv­ing learn­ing?) From this per­spec­tive, it’s not too sur­pris­ing that human med­i­cine may be largely wasted effort or harm­ful16 (although most - espe­cially doc­tors - would stren­u­ously deny this). “Hardly any man is clever enough to know all the evil he does.”17

An anal­ogy to com­plex sys­tems is a super­fi­cial analy­sis at best. Many com­plex sys­tems are rou­tinely opti­miz­able on some para­me­ter of inter­est by orders of mag­ni­tudes, or at least fac­tors. Economies grow expo­nen­tial­ly, on the back of all sorts of improv­ing per­for­mance curves which make us richer than emper­ors com­pared with our ances­tors; the mir­a­cle of eco­nomic growth, built on thou­sands of dis­tinct com­plex sys­tems being opti­mized by humans, seems to go unno­ticed and be so nor­mal and taken for grant­ed. If we were com­put­ers, an ordi­nary nerd with access to some liq­uid nitro­gen could dou­ble our clock speed.

With intel­li­gence, on the other hand, not only do we have no inter­ven­tions to make one an order of mag­ni­tude smarter on some hypo­thet­i­cal mea­sure of absolute intel­li­gence (per­haps such a man would be to us as we are to dogs?) but we have no inter­ven­tions which make one a few fac­tors smarter (the smartest man to ever live?) nor do we even have any inter­ven­tions which can move one more than a few per­cent­age points up in the gen­eral pop­u­la­tion! We remain the same. It is as if sci­en­tists and doc­tors, after study­ing cars for cen­turies, shame­facedly had to admit that their thou­sands of exper­i­men­tal cars all still had their speed throt­tles stuck on 25-30kph - but the good news was that this new oil addi­tive might make a few of the cars run 0.1kph faster!

This is not the usual state of affairs for even extremely com­plex sys­tems. This raises the ques­tion of why all these cars are so uni­formly stuck at a cer­tain top speed and how they got to be so opti­mized; why are we like these fan­tas­ti­cal cars, and not com­puter proces­sors?


Intel­li­gence is an almost unal­loyed good when we look at cor­re­la­tions in the real-­world for income, longevi­ty, hap­pi­ness, con­tri­bu­tions to sci­ence or med­i­cine, crim­i­nal­i­ty, favor­ing of free speech etc18. Why is it, then, that we can find quotes like “the rule that human beings seem to fol­low is to engage the brain only when all else fails - and usu­ally not even then”19 or “In effect, all ani­mals are under strin­gent selec­tion pres­sure to be as stu­pid as they can get away with”20? Why does so much psy­cho­log­i­cal research, espe­cially & , seem to boil down to a dichotomy of a slow accu­rate way of think­ing and a fast less-ac­cu­rate way of think­ing (“sys­tem I” vs “sys­tem II” being just one of the innu­mer­able pairs coined by researchers21)?

Because think­ing is expen­sive and slow, and while it may be an unal­loyed good, it is sub­ject to dimin­ish­ing returns like any­thing else (if it is prof­itable at all in a par­tic­u­lar niche: no bac­te­ria needs sophis­ti­cated cog­ni­tive skills, nor most mam­mals) and other things become more valu­able. Just another trade­off.


In “The Wis­dom of Nature: An Evo­lu­tion­ary Heuris­tic for Human Enhance­ment”22 (Human Enhance­ment 2008), and Anders Sand­berg put this prin­ci­ple as a ques­tion or chal­lenge, “evo­lu­tion­ary opti­mal­ity chal­lenge” (EOC):

If the pro­posed inter­ven­tion would result in an enhance­ment, why have we not already evolved to be that way?

We could take this as a bio­log­i­cal ver­sion of Chester­ton’s fence. Evo­lu­tion is a mas­sively par­al­lel search process which has been run­ning on humans (and pre­de­ces­sor organ­isms) for bil­lions of years, ruth­lessly opti­miz­ing for repro­duc­tive fit­ness. It is an immensely stu­pid and blind idiot god which will accom­plish its goal by any avail­able means, and if that means evo­lu­tion­ary mech­a­nisms will cause indi­vid­u­als to drive their own species extinct because this was the fittest thing for each indi­vid­ual to do or if the highly arti­fi­cial & unlikely con­di­tions for are enforced by an exper­i­menter and evo­lu­tion causes group norms of mass infan­ti­ci­dal can­ni­bal­iza­tion to devel­op, so be it!

It is of course pos­si­ble for a new muta­tion to be fit­ter or for the envi­ron­ment to change and ren­der some alter­na­tive more fit. This is some­times true, but it is over­whelm­ingly usu­ally false. If you do not believe me, feel free to go try to beat just , or if you’re up for a chal­lenge, be more repro­duc­tively fit in a tuna fish’s niche than the tuna fish. Every so often you hear of a hedge fund which found , or of an that is beat­ing the native flo­ra: but remem­ber that they are fund or species out of thou­sands and thou­sands try­ing. :

It was a good answer that was made by one who when they showed him hang­ing in a tem­ple a pic­ture of those who had paid their vows as hav­ing escaped ship­wreck, and would have him say whether he did not now acknowl­edge the power of the gods, - “Aye”, asked he again, “but where are they painted that were drowned after their vows?” And such is the way of all super­sti­tion, whether in astrol­o­gy, dreams, omens, divine judg­ments, or the like; wherein men, hav­ing a delight in such van­i­ties, mark the events where they are ful­filled, but where they fail, though this hap­pens much often­er, neglect and pass them by.

They are excep­tions which prove the rule; they are, in fact, the excep­tions which cause the rule to be true, by exploit­ing the niche or oppor­tu­ni­ty. Sup­pose we turned out to be harm­fully miserly with calo­ries and there is some recep­tor (such as those acted upon by stim­u­lants like caf­feine or ) which trig­gers a cas­cade of changes lead­ing to behav­ior which is a supe­rior trade­off in caloric con­sump­tion vs activ­i­ty. Evo­lu­tion would slowly increase the mar­ket-share of alle­les which affect this recep­tor, and after a while, the new level of activ­ity would become opti­mal and now use of a stim­u­lant affect­ing the recep­tor would cease to be fit because it cranks it too high. There may be some such oppor­tu­ni­ties avail­able to humans today, since we know of past oppor­tu­ni­ties like adult lac­tose tol­er­ance which have been sweep­ing through gene pools over the past thou­sand years, but can we really claim that all the inter­ven­tions in which we dif­fer from our dis­tant ances­tors can be traced to such repro­duc­tive fit­ness jus­ti­fi­ca­tions? (And peo­ple think over­reaches and spec­u­lates with­out evi­dence!) The­o­ret­i­cal cal­cu­la­tions appar­ently indi­cate that in a chang­ing envi­ron­ment, the “repro­duc­tive fit­ness gap” between the cur­rent allele and its alter­na­tives will be small and large gaps expo­nen­tially rare23; this seems intu­itive - to con­tinue the mar­ket anal­o­gy, the big­ger the arbi­trage, the faster it will be exploit­ed.

Obvi­ously we humans do inter­vene all the time, and many of those inter­ven­tions are worth­while. Wom­en, for exam­ple, are big fans of , and if the female repro­duc­tive sys­tem isn’t con­trolled by evo­lu­tion, noth­ing is. How are we to rec­on­cile the the­o­ret­i­cal expec­ta­tion that we should find it nigh-im­pos­si­ble to beat evo­lu­tion at its own game with the observed fact that we seem to inter­vene suc­cess­fully all the time?


“What a book a dev­il’s chap­lain might write on the clum­sy, waste­ful blun­der­ing, low and hor­ri­bly cruel works of nature!”

, 1856-07-13 let­ter to , More Let­ters of Charles Dar­win, Vol­ume 1

“It is a pro­found truth—re­al­ized in the nine­teenth cen­tury by only a hand­ful of astute biol­o­gists and by philoso­phers hardly at all (in­deed, most of those who held and views on the mat­ter held a con­trary opin­ion)—a pro­found truth that Nature does not know best; that genet­i­cal evo­lu­tion, if we choose to look at it liv­er­ishly instead of with fatu­ous good humor, is a story of waste, makeshift, com­pro­mise and blun­der.”

Sir , The Future of Man

There may be no free lunch­es, but might there be some cheap lunch­es? Yud­kowsky’s for­mu­la­tion points out sev­eral ways to escape the argu­ment:

  1. inter­ven­tions may not be sim­ple

    So one might find major enhance­ments through some very com­plex surgery or pros­thet­ic; per­haps brain implants which expand mem­ory or enable (con­trolled) . Evo­lu­tion is a search pro­ce­dure for find­ing local opti­ma, which are not nec­es­sar­ily global opti­ma. Exam­ples like the demon­strates such traps, but how would evo­lu­tion fix them? Even if a muta­tion sud­den made the nerve go the shorter direc­tion, it’s not clear what other changes would have to be made to deal with this improve­ment, and this com­bi­na­tion of mul­ti­ple rare muta­tions may not hap­pen enough times for the small repro­duc­tive fit­ness improve­ment (less resources used on nerves) to make it to .

  2. the sim­ple inter­ven­tions may not lead to a major enhance­ment

    Nutri­tional sup­ple­ments are exam­ples; it makes per­fect sense that fix­ing a chem­i­cal defi­ciency could be a sim­ple mat­ter and enhance repro­duc­tive fit­ness - but one would expect only minor men­tal enhance­ments and this effect would not gen­er­al­ize to very many peo­ple. (Sim­i­lar­ly, most nootrop­ics do not do very much.)

  3. the inter­ven­tion may be sim­ple, give major enhance­ments, but result in a net loss of repro­duc­tive fit­ness

    The famous comes to mind. Accord­ing to this the­o­ry, the Ashke­nazi were forced into occu­pa­tions demand­ing intel­li­gence, and micro-s­e­lected for high intel­li­gence. Except the high IQ genes were not pre­vi­ously preva­lent among either Jews or gen­tiles because - like - when they became too preva­lent, they result in hor­ri­ble dis­eases like . In 2007, was found to increase ver­bal IQ in afflicted fam­ily mem­bers vs non by some­thing like 25 points; this would be great for them - except for how that muta­tion starts caus­ing blind­ness in one’s 20s or lat­er. (In gen­er­al, it’s much eas­ier to find muta­tions or other genetic changes break­ing intel­li­gence than help­ing in cases of retar­da­tion24 and autism25.)

Bostrom also offers 3 cat­e­gories of ways in which inter­ven­tions can escape his ‘EOC’:

  1. Changed Trade­offs. Evo­lu­tion ‘designed’ the sys­tem for oper­a­tion in one type of envi­ron­ment, but now we wish to deploy it in a very dif­fer­ent type of envi­ron­ment. It is not sur­pris­ing, then, that we might be able to mod­ify the sys­tem bet­ter to meet the demands imposed on it by the new envi­ron­ment.26
  2. Value Dis­cor­dance. There is a dis­crep­ancy between the stan­dards by which evo­lu­tion mea­sured the qual­ity of her work, and the stan­dards that we wish to apply. Even if evo­lu­tion had man­aged to build the finest repro­duc­tion-and-­sur­vival machine imag­in­able, we may still have rea­son to change it because what we value is not pri­mar­ily to be max­i­mally effec­tive inclu­sive-re­pro­duc­tive-­fit­ness opti­miz­ers.
  3. Evo­lu­tion­ary Restric­tions. We have access to var­i­ous tools, mate­ri­als, and tech­niques that were unavail­able to evo­lu­tion. Even if our engi­neer­ing tal­ent is far infe­rior to evo­lu­tion’s, we may nev­er­the­less be able to achieve cer­tain things that stumped evo­lu­tion, thanks to these novel aids.

An exam­ple of how not to escape the EOC, I believe, is offered in (Lynch et al 2012), when the authors attempt to argue that pow­er­ful nootrop­ics are pos­si­ble:

But per­haps the ‘room for improve­ment’ issue can be recast in terms of brain evo­lu­tion by ask­ing whether com­par­a­tive anatom­i­cal evi­dence points to strong adap­tive pres­sures for designs that are log­i­cally related to improved cog­ni­tive per­for­mance. Anatomists often resort to allom­e­try when deal­ing with ques­tions of selec­tive pres­sures on brain regions. Applied to brain pro­por­tions, this involves col­lect­ing mea­sure­ments for the region of inter­est - e.g., frontal cor­tex – for a series of ani­mals within a given tax­o­nomic group and then relat­ing it to the vol­ume or weight of the brains of those ani­mals. This can estab­lish with a rel­a­tively small degree of error whether a brain com­po­nent in a par­tic­u­lar species is larger than would be pre­dicted from that species’ brain size. While there is not a great deal of evi­dence, stud­ies of this type point to the con­clu­sion that cor­ti­cal sub­di­vi­sions in humans, includ­ing asso­ci­a­tion regions, are about as large as expected for an anthro­poid pri­mate with a 1350cc brain. The vol­ume of area 10 of human frontal cor­tex, for exam­ple, fits on the regres­sion line (area 10 vs. whole brain) cal­cu­lated from pub­lished data (Se­mende­feri et al., 2001) for a series com­posed of gib­bons, apes and humans (Lynch and Granger, 2008). Given that this region is widely assumed to play a cen­tral role in exec­u­tive func­tions and work­ing mem­o­ry, these obser­va­tions do not encour­age the idea that selec­tive pres­sures for cog­ni­tion have dif­fer­en­tially shaped the pro­por­tions of human cor­tex. Impor­tant­ly, this does not mean that those pro­por­tions are in any sense typ­i­cal. The allo­met­ric equa­tions involve dif­fer­ent expo­nents for dif­fer­ent regions, mean­ing that absolute pro­por­tions (e.g., pri­mary sen­sory cor­tex vs. asso­ci­a­tion cor­tex) change as brains grow larg­er. The bal­ance of parts in the cor­tex of the enor­mous human brain is dra­mat­i­cally dif­fer­ent than found in the much smaller mon­key brain: area 10, for instance, occu­pies a much greater per­cent­age of the cor­tex in man. But these effects seem to reflect expan­sion accord­ing to rules embed­ded in a con­served brain plan rather than selec­tion for the spe­cific pat­tern found in humans (Fin­lay et al., 2001).

…But our argu­ment here is that these expanded cor­ti­cal areas are likely to use generic net­work designs shared by most pri­mates; if so, then it appears unlikely that the designs are in any sense ‘opti­mized’ for cog­ni­tion. We take this as a start­ing posi­tion for the assump­tion that the designs are far from being max­i­mally effec­tive for spe­cial­ized human func­tions, and there­fore that it is real­is­tic to expect that cog­ni­tion-re­lated oper­a­tions can be sig­nif­i­cantly enhanced.

I would agree that the human brain’s archi­tec­ture does not seem to be opti­mal in any uni­ver­sal sense; and that this would con­sti­tute an inter­est­ing argu­ment if one were argu­ing that arti­fi­cial intel­li­gences will not inher­ently be lim­ited to a level of intel­li­gence com­pa­ra­ble to the great­est human genius­es.

How­ev­er, this does not offer hope for nootrop­ics because the human brain can eas­ily be sub­op­ti­mal in its gross anatom­i­cal archi­tec­ture but close to opti­mal in any fac­tor eas­ily tweaked by chem­i­cals! (A sug­ges­tion that brain region size is sub­op­ti­mal is a sug­ges­tion only that a large change in brain region size might lead to large gains - but large changes are nei­ther easy, sim­ple, nor pos­si­ble cur­rent­ly.)


Tele­ol­ogy is like a mis­tress to a biol­o­gist: he can­not live with­out her but he’s unwill­ing to be seen with her in pub­lic.27

Bostrom’s cri­te­ria are more gen­er­al, so we’ll use them.

Birth con­trol is a clear exam­ple of sat­is­fy­ing loop­hole #2, ‘value dis­cor­dance’. Ovu­la­tion is under the body’s con­trol and is linked in evo­lu­tion­ary psy­chol­ogy to many changes in behav­ior; unpro­cre­ative sex is com­mon through­out the ani­mal king­dom where it serves other pur­poses like form­ing social con­nec­tions in troupes. Hunter-­gath­erer women prac­tice spaced births by let­ting their child suckle them for years; mater­nal can­ni­bal­ism has been observed when moth­ers are under par­tic­u­lar stress (and per­haps also in human­s). So, it’s clear that there is birth con­trol capa­bil­ity already avail­able to hominids, and not too sur­pris­ing that it’s pos­si­ble to ren­der a healthy woman entirely infer­tile with­out major health con­se­quences. Many women would pre­fer evo­lu­tion have done just this! They do not value hav­ing a dozen chil­dren while young; they would rather have just 2 at a time of their choos­ing - if any at all. Why is evo­lu­tion not so oblig­ing? Well, it obvi­ously would not be very repro­duc­tively fit…

Pace­mak­ers are an exam­ple of #3: evo­lu­tion could­n’t afford to engi­neer more reli­able hearts, in part for lack of elec­tronic microchips and pos­si­bly because humans are already at the lim­its of the per­for­mance enve­lope28.

Many traits related to nutri­tion fall into the cat­e­gory of #1.29

How about sup­ple­ments? Most sup­ple­ments are just tweak­ing bio­chem­i­cal process­es, and don’t obvi­ously fall under 1, 2, or 3; and the few which seem to enhance healthy humans are finicky crea­tures (see my intro­duc­tion to ). , for exam­ple, may seem par­tic­u­larly ques­tion­able as one’s body secretes con­sid­er­able quan­ti­ties in an intri­cate cycle (but see later).

Flynn effect

The is a pos­si­ble coun­ter-ex­am­ple: it oper­ates broadly over many coun­tries, improves aver­age IQ by per­haps 10 or more points over the last cen­tury30, pre­sum­ably is envi­ron­men­tal, and oper­ates with­out any explicit expen­sive eugen­ics pro­grams or any­thing like that.

How­ev­er, there are sev­eral ways in which the Flynn effect respects Alger­non’s argu­ment and passes the loop­holes:

  1. the Flynn effect is lim­ited in its gains and so will result in not major gains

    the Flynn effect has already and reversed to some degree. The sit­u­a­tion in the US is unclear, but given the out­right losses in ver­bal & sci­ence skills seen 1981-2010 in the most intel­li­gent of Mid­west­ern stu­dents31, this is con­sis­tent with a Flynn effect oper­at­ing through elim­i­nat­ing defi­cien­cies & improv­ing the lower end or with a Flynn effect that has ceased to exist

  2. the Flynn effect is appar­ently envi­ron­men­tal, and one of the most plau­si­ble expla­na­tions is that it is due to either nutri­tional deficits or pub­lic health inter­ven­tions against infec­tious dis­eases.

    In nei­ther case are inter­ven­tions ‘easy’ in any sense, nor are the inter­ven­tions avail­able to evo­lu­tion - if one’s diet is lack­ing in an essen­tial ele­ment like , evo­lu­tion can­not sim­ply con­jure it away; nor can it invent any bet­ter immune sys­tems than it already has as part of the usual with infec­tious agents. As we already not­ed, we could expect nutri­tional inter­ven­tions to pro­duce small ben­e­fits, and we might expect that imple­ment­ing a whole bat­tery of pos­si­ble improve­ments (io­dine defi­cien­cy, iron defi­ciency, a dozen child­hood infec­tions etc) to pro­duce much what we see with the Flynn effect. But we would expect the gains to be spe­cific and quickly exhausted once the low-hang­ing fruit is exhaust­ed. (There can­not be indef­i­nitely many defi­cien­cies and infec­tion­s!) This too is what we observe with the halt­ing of the Flynn effect.

  3. The intel­li­gence gains from the Flynn effect may not be repro­duc­tive-­fit­ness-in­creas­ing; IQ cor­re­lates strongly with many desir­able things like income, hap­pi­ness, knowl­edge, edu­ca­tion, etc. - but not hav­ing more than aver­age chil­dren. The cor­re­la­tions are found both and . (It is of course pos­si­ble that the Flynn effect causes IQ gains and repro­duc­tive fit­ness increases on the lower end of the spec­trum and high IQ is intrin­si­cally repro­duc­tive-­fit­ness-re­duc­ing in the mod­ern envi­ron­ment, but the obser­va­tion is sug­ges­tive.)

  4. the Flynn effect does not actu­ally reflect intel­li­gence gains but dam­age to the valid­ity of the sub­test in which the gains appear, and is irrel­e­vant


Or in “Grow­ing up is hard”, remarks that Bostrom’s EOC is:

…one rea­son to be wary of, say, mem­ory enhancers [such as ]: if they have no down­sid­es, why does­n’t the brain pro­duce more already? Maybe you’re using up a lim­ited mem­ory capac­i­ty, or for­get­ting some­thing else…

Let’s con­sider the spe­cific case of pirac­etam. Pirac­etam is so old and has so many stud­ies on its effi­cacy (real if not sub­stan­tial) and safety (ut­ter­ly) that it screens off a lot of sec­ondary con­sid­er­a­tions.

  • Might pirac­etam escape the EOC with #3?

    No. What­ever recep­tors or but­tons pirac­etam pushes could already be pushed by the brain the usual way. There is noth­ing novel about pirac­etam in that sense.

  • Might pirac­etam escape the EOC with #2?

    Per­haps. Hard to see how pirac­etam trades off repro­duc­tive fit­ness for some­thing else, though. Since its syn­the­sis in 1964, or other safety issues have been not­ed, unlike other drugs such as caf­feine or aspirin.

  • Might pirac­etam escape the EOC with #1?

    Prob­a­bly. Many trade­offs are dif­fer­ent in con­tem­po­rary First World coun­tries than in the prover­bial Stone Age veldt. We should look more closely at what pirac­etam does and what trade­offs it may be chang­ing.

A ‘cholin­er­gic’ oper­ates by encour­ag­ing higher lev­els of the acetyl­choline neu­ro­trans­mit­ter; acetyl­choline is one of the most com­mon neu­ro­trans­mit­ters. If sero­tonin is loosely asso­ci­ated with mood, we might say that acetyl­choline is loosely asso­ci­ated with the ‘veloc­ity’ of thoughts in the brain. If one is using more acetyl­choline, one needs to cre­ate more acetyl­choline (the brain can­not bor­row indef­i­nitely like the US fed­eral gov­ern­men­t). Acetyl­choline is made out of the .

An inter­est­ing thing about pirac­etam use is that it does­n’t do very much by itself32. It is char­i­ta­bly described as ‘sub­tle’. The stan­dard advice is to take a choline sup­ple­ment with the pirac­etam: a gram of , choline bitar­trate, or choline cit­rate.

Isn’t this inter­est­ing? Pre­sum­ably we are not Irish peas­ants con­sum­ing wretched diets of pota­to, pota­to, and more pota­to, with some mut­ton on the hol­i­days. We are cog­nizant of how a good diet & exer­cise are pre­req­ui­sites to brain pow­er. Yet, a gram of straight choline still boosts pirac­etam’s effects from sub­tle or place­bo, to notice­able & mea­sur­able.

This sug­gests that per­haps a nor­mal First World diet is choline-d­e­fi­cient. If even well-fed humans must econ­o­mize on choline & acetyl­choline, then surely our ances­tors, who were worse off nutri­tion­al­ly, had to econ­o­mize even more severe­ly. Evo­lu­tion would frown on squan­der­ing acetyl­choline on idle thoughts like ‘what was that witty say­ing by Ugh the other day?’ That choline might be needed in the next famine! This sug­ges­tion is but­tressed by one small mouse exper­i­ment:

Admin­is­ter­ing choline sup­ple­men­ta­tion to preg­nant rats improved the per­for­mance of their pups, appar­ently as a result of changes in neural devel­op­ment in turn due to changes in gene expres­sion (Meck et al. 1988; Meck & Williams 2003; Mel­lott et al. 2004). Given the ready avail­abil­ity of choline sup­ple­ments, such pre­na­tal enhance­ment, may already (inad­ver­tent­ly) be tak­ing place in human pop­u­la­tions. Sup­ple­men­ta­tion of a moth­er’s diet dur­ing late and 3 months post­par­tum with long-chained fatty acids has also been demon­strated to improve cog­ni­tive per­for­mance in human chil­dren (Hel­land et al. 200333).34

Past our embry­o-­hood, we can’t tell our bod­ies that we have avail­able as much choline as it could pos­si­bly need, that we value our synapses blaz­ing at every moment more than a bet­ter chance of sur­viv­ing a famine (which effec­tively no longer exist). So we have to over­ride it, for our own good.

(It’s worth not­ing here that there is con­sid­er­able over­lap between #1 and #2. Whether you see pirac­etam as a con­flict in val­ues between evo­lu­tion’s worst-­case plan­ning and our desire for greater aver­age or peak per­for­mance, or as a shift in opti­mal expen­di­ture based on a his­tor­i­cal drop in the cost of bulk quan­ti­ties of choline, is a mat­ter of pref­er­ence.)


How about ? It is a clear-­cut exam­ple of fail­ing #3, but per­haps it passes under #1 like pirac­etam?

A is an obvi­ous case of value dis­cor­dance: humans are meant to work mostly dur­ing the day, with min­i­mal dan­ger­ous night-­time activ­i­ty. Shift work­ers per­versely insist on doing the exact oppo­site, even strug­gling against the cir­ca­dian rhythms (to the detri­ment of ). Evo­lu­tion wots not of your ‘employ­ment con­tract’, piti­ful human!

Reg­u­lar peo­ple have a less extreme ver­sion of the shift work­er’s dilem­ma. The mod­ern pop­u­la­tion does­n’t rise and set with the sun, for impon­der­able rea­sons. (My per­sonal the­ory is wide­spread : dark­ness over­comes and forced the ancients to bed, but we have elec­tric light­ing and can stay up indef­i­nite­ly.) This leads to a val­ues mis­match, and a sim­i­lar solu­tion.


is another drug that seems sus­pi­ciously like a free lunch. The side-­ef­fects are min­i­mal and rare, and the ben­e­fit quite unusual and strik­ing: not need­ing to sleep for a night. The research on gen­eral cog­ni­tive ben­e­fits is mixed but real35. (My own expe­ri­ence with armodafinil was that after 41 hours of sleep­-de­pri­va­tion, my and focus were actu­ally bet­ter than nor­mal as judged by scores! An anom­aly, but still curi­ous.) Yes, modafinil costs mon­ey, but that’s not really rel­e­vant to our health or to Evo­lu­tion. Yes, there is, anec­do­tal­ly, a risk of com­ing to tol­er­ate modafinil (although no addic­tion), but again that does­n’t mat­ter to Evo­lu­tion - there would still be ben­e­fits before the tol­er­ance kicked in.

What heuris­tic might we use?

  • Chem­i­cal­ly, modafinil does not seem to be so bizarre that evo­lu­tion could not stum­ble across it or an equiv­a­lent mech­a­nism, so prob­a­bly we can­not appeal to #3, “evo­lu­tion­ary restric­tions”. Its mech­a­nism is not very clear, but mostly seems to manip­u­late things like the his­t­a­mine sys­tem (and to a much lesser extent, dopamine), all things Evo­lu­tion could eas­ily do.

  • Nor is it clear what value dis­cor­dance might be involved. We could come up with one, though.

    If one the­o­rized that modafinil came with a mem­ory penal­ty, inas­much as mem­ory con­sol­i­da­tion and the hip­pocam­pus seem to inti­mately involve sleep, then we might have a dis­cor­dance where we value being able to pro­duce and act more than being able to remem­ber things. This might even be a sen­si­ble trade­off for a mod­ern man: why not sac­ri­fice some abil­ity to learn or remem­ber long-term, since you can imme­di­ately gain back that capac­ity and more by suit­able use of effi­cient mem­ory tech­niques like ?

  • #1 seems promis­ing. Like pirac­etam, there is some­thing in short sup­ply that modafinil would use more of: calo­ries! While you are awake, you are burn­ing more calo­ries than while asleep. Dur­ing the day, synapses , which get wiped out by sleep; is this because synapses and mem­o­ries are expen­sive36 and can­not be allowed to con­sume ever more resources with­out some sort of ‘’, synap­tic ? & stud­ies bear out some of the pre­dic­tions of the model and may lead to inter­est­ing new find­ings37 (see also Bom & Feld 2012 dis­cussing Chau­vette et al 2012).

    Pre­vi­ously noted was the meta­bolic cost of defend­ing against infec­tions; one ani­mal study found the prox­i­mate cause of death in sleep depri­va­tion to be bac­te­r­ial infec­tions38. You are also - in the ancient evo­lu­tion­ary envi­ron­ment - per­haps expos­ing your­self to addi­tional risks in the dark night. (This would be the .)

    Resource usage is a real con­cern for the human brain, along with scal­ing issues39: it uses <20% of ener­gy; 87% in infants. One blog­ger says:

    The human brain is also extremely “expen­sive tis­sue” (). Although it only accounts for 2% of an adult’s body weight, it accounts for 20–25% of an adult’s rest­ing oxy­gen and energy intake (Attwell & Laugh­lin 2001: 1143). In early life, the brain even makes up for up 60–70% of the body’s total energy require­ments. A chim­panzee’s brain, in com­par­ison, only con­sumes about 8–9% of its rest­ing metab­o­lism (: 330). The human brain’s energy demands are about 8 to 10 times higher than those of skele­tal mus­cles (Dun­bar & Shultz 2007: 1344), and, in terms of energy con­sump­tion, it is equal to the rate of energy con­sumed by leg mus­cles of a marathon run­ner when run­ning (At­twell & Laugh­lin 2001: 1143). All in all, its con­sump­tion rate is only topped by the energy intake of the heart (Dun­bar & Shultz 2007: 1344).

    There are addi­tional dis­ad­van­tages to increased intel­li­gence - larger heads would drive mater­nal & infant mor­tal­ity rates even higher than they are40. And it’s worth not­ing that while the human brain is , yet the human is not any big­ger than one would pre­dict be extrap­o­lat­ing from gib­bon or ape cor­tex vol­umes, despite the human lin­eage split­ting off mil­lions of years ago.41 The human brain seems to be spe­cial only in being a scaled-up pri­mate brain42, with close to the meta­bolic limit in its num­ber of neu­rons43 (which sug­gests a res­o­lu­tion to the ques­tion why despite con­ver­gent evo­lu­tion of rel­a­tively high intel­li­gence44, only pri­mates “took off”). There are other ways in which humans seem to have hit intel­li­gence lim­its - why did our ances­tors’ brains grow in vol­ume for mil­lions of years45, only to come to a halt with the Nean­derthals46 & Cro-­Magnons and actu­ally start shrink­ing47 to the mod­ern vol­ume, and why did old age only start increas­ing 50,000 years ago or later48, well after humans began devel­op­ing tech­nol­ogy like con­trolled fire (>=400,000 years ago49); or why are pri­mate guts (also resource-­ex­pen­sive) with brain size & in one fish breed­ing exper­i­ment, or mus­cles starved of sug­ars and brains favored50; or why do the seem to pay for their intel­li­gence with endemic genetic dis­or­ders51; or why does evo­lu­tion per­mit human brains to shrink dra­mat­i­cally with age, as much as 15% of vol­ume, besides the huge per­for­mance losses, while the brains of our clos­est rel­a­tive-species (the chim­panzees), do not shrink at all?52 For that mat­ter, why are heads, cen­tral ner­vous sys­tems, and pri­mate-level intel­li­gence so extremely rare on the tree of life, with no exam­ples of of intel­li­gence (as opposed to like basic eye­-spots, which are such a fan­tas­ti­cally adap­tive tool that they have inde­pen­dently evolved )?53

    The obvi­ous answer is that have kicked in for intel­li­gence in pri­mates and humans in par­tic­u­lar54. (In­deed, it’s appar­ently been argued that not only are humans not much smarter than pri­mates55, but there is lit­tle over­all intel­li­gence dif­fer­ences in ver­te­brates56. Humans lose embar­rass­ingly on even pure tests of sta­tis­ti­cal rea­son­ing; we are out­per­formed on the by pigeons and to a lesser extent mon­keys!) The last few mil­len­nia aside, humans have not done well and has appar­ently before, and the 57 and anthro­pogenic s sug­gest that our cur­rent suc­cess may be short­-lived (not that agri­cul­ture & civ­i­liza­tion were great in the first place). Some psy­chol­o­gists have even tried to make the case that increases in intel­li­gence do not lead to bet­ter infer­ences or choices (Her­twig & Todd 2003).

    Modafinil or modafinil-­like traits might be selected against due to increased calo­rie expen­di­ture, , or risks of night-­time activ­i­ty. Either expla­na­tion fails in a mod­ern envi­ron­ment; mod­ern soci­eties have mur­der and assault rates orders of mag­ni­tude lower than that seen among abo­rig­ines58, and calo­ries are so abun­dant that they have begun reduc­ing repro­duc­tive fit­ness (we call this poi­son­ing-by-­too-­many-calo­ries the ).

Is that last a con­vinc­ing defense of modafinil against the EOC or Alger­non’s prin­ci­ple? It seems rea­son­able to me, if not as strong a defense as I would like.


How about opi­ates? Mor­phine and other painkillers can eas­ily be jus­ti­fied as evo­lu­tion not know­ing when a knife cut is by a mur­der­ous enemy and when it’s by a kindly sur­geon (which did­n’t exist way back when), and choos­ing to make us err on the side of always feel­ing pain. But recre­ational drug abuse?

  • #1 does­n’t seem too plau­si­ble - what about mod­ern soci­ety would favor opi­ate con­sump­tion out­side of med­i­c­i­nal use? If one wishes to deaden the despair and ennui of liv­ing in a degen­er­ate athe­is­tic mate­r­ial cul­ture, we have beer for that.59
  • #3 does­n’t work either; opi­oids have been around for ages and work via the stan­dard brain machin­ery.
  • #2 might work here as well, but this dumps us straight into the debate about the and what harm drug use does to the user & soci­ety.

But even this analy­sis is help­ful: we now know on what basis to oppose drug use, and most impor­tant­ly, what kind of evi­dence which we should look for to sup­port or fal­sify our belief about hero­in.


is another pop­u­lar illicit drug. Read­ing accounts of early MDMA use or stud­ies on its ben­e­fi­cial psy­cho­log­i­cal prop­er­ties (a bit like those claimed for pre­vi­ous psy­che­delics like LSD or psilo­cy­bin), one is struck by how fear seems to be a com­mon trait - or rather, the lack of fear:

With Ecsta­sy, I had sim­ply stepped out­side the worn paths in my brain and, in the process, gained some per­spec­tive on my life. It was an amaz­ing feel­ing. Small incon­sis­ten­cies became obvi­ous. “I need mon­ey, I have a $500 motor­cy­cle that I’m too scared to ride, so why not sell it?” So did big psy­cho­log­i­cal ones: “The more angry I am at myself, the more crit­i­cal I am of my girl­friend. Why should I care how Carol chews her gum?” Ecstasy nudges you to think, very deeply, about one thing at a time. (It was­n’t that harsh LSD feel­ing, where every thought seems like an absurd para­dox - like the fact that we’re all, deep down, just a bunch of mon­keys.)..A gov­ern­men­t-ap­proved study in Spain has just begun in which Ecstasy is being offered to treat rape vic­tims for whom no treat­ment has worked, based on the premise that MDMA “reduces the fear response to a per­ceived emo­tional threat” in ther­apy ses­sions. A Swiss study in 1993 yielded pos­i­tive anec­do­tal evi­dence on its effect on peo­ple suf­fer­ing from post-­trau­matic stress dis­or­der. And a study in Cal­i­for­nia may soon begin in which Ecstasy is admin­is­tered to end-stage can­cer patients suf­fer­ing from depres­sion, exis­ten­tial crises and chronic pain. The F.D.A. will be review­ing the pro­to­col for Stage 2 of the tri­al; results are expected in 2002.

Read­ing, I can’t help but be reminded of the pop­u­lar self­-help prac­tice “” (an ), ele­ments of which reap­pear among businessmen/entrepreneurs, , shy­ness ther­a­pists60, nerds, and oth­ers: one goes out in pub­lic and makes small harm­less requests of var­i­ous strangers until one is no longer uncom­fort­able or afraid. Even­tu­ally one real­izes that it is harm­less to ask - the worst that will hap­pen is they will say no - and one will pre­sum­ably be more con­fi­dent, less fear­ful, hap­py, and effec­tive a per­son. What is the jus­ti­fi­ca­tion for this? After all, one does­n’t regard being afraid of, say, snake venom as a prob­lem and a good rea­son to under­take a long reg­i­men of ! Snake venom is dan­ger­ous and should be feared, and delib­er­at­ing destroy­ing one’s use­ful fear would be like a mouse doing ‘cat ther­apy’.

Rejec­tion ther­apy fans argue that there is a mis­match between fear and real­i­ty: our fears and social anx­i­ety are cal­i­brated for the world of a few cen­turies ago where >90% of the world lived on farms and vil­lagers where a poor rep­u­ta­tion & social rejec­tion could mean death; while in the mod­ern world, social rejec­tion is a mere incon­ve­nience because even if one is rejected by one’s extended cir­cle of there are 100x more peo­ple in a small town, and even more thou­sands of times more peo­ple in a city (to say noth­ing of a mega­lopo­lis like New York City where the num­bers get vague into the mil­lion­s). Risk-­tak­ing behav­ior which is opti­mal in the vil­lage will be ludi­crously con­ser­v­a­tive and inef­fi­cient in the big city.

If this the­ory were cor­rect (it is pos­si­ble but far from proven), and if MDMA worked the same way (un­like­ly), then we have a clear exam­ple of #1, “changed trade­offs”: we are too and fear­ful of social sanc­tion for a mod­ern envi­ron­ment. (Cu­ri­ous­ly, this is also a pro­posed expla­na­tion for the appar­ent increase in in mod­ern soci­eties: psy­chopaths are “” or “” who would nor­mally be sup­pressed or less fit in a tight­ly-net­worked tribe or vil­lage, but can thrive in the rep­u­ta­tion-poor mod­ern world as they move from place to place and social cir­cle to social cir­cle, leav­ing behind their vic­tims.61)

  1. , Eric­s­son et al 1993, gives a few exam­ples:

    The best evi­dence link­ing inten­sive train­ing directly to observed changes in heart size comes from lon­gi­tu­di­nal stud­ies ofy­oung ath­letes attain­ing expert per­for­mance and of older ath­letes ter­mi­nat­ing their careers and prac­tice reg­i­mens. Elo­vian­ioand Sund­berg (1983) found that elite long-dis­tance run­ners acquired greater aer­o­bic power and larger heart vol­umes dur­ing a 5-year period of train­ing but showed no ini­tial supe­ri­or­ity at age14. Rost (1987) found dur­ing a lon­gi­tu­di­nal study of chil­dren from age 8 to 11 that heart vol­umes increased much more iny­oung swim­mers than in non­trained chil­dren (con­trol). It appears that at least 1 year of intense train­ing is required before the size of a human heart begins to change. Sim­i­lar­ly, once ath­letes ter­mi­nate their train­ing the increased heart sizes remain, but in the absence of exer­cise the heart vol­ume regresses to within nor­mal range over a 10-year peri­od; Rost (1987) reports a vol­ume reduc­tion of 42% in one case. Howald (1982) reports case stud­ies of top ath­letes who were forced to stop or reduce train­ing because of injuries. Dras­tic decre­ments in the per­cent­age of their slow-twitch fibers occurred within 6 months to 1 year.

    Because most sports involve only some of the mus­cles in the body, it is pos­si­ble to con­trast these inten­sively trained mus­cles with other mus­cles in the same ath­letes. Tesch and Karls­son (1985) exam­ined the size and fre­quency of fast and slow-twitch fibers in the mus­cles of dif­fer­ent types of elite ath­letes as well as of stu­dents serv­ing as con­trol sub­jects. They found that dif­fer­ences in the per­cent­age of slow-twitch fibers in elite ath­letes’ mus­cles occur only for mus­cles specif­i­cally trained for a sport (legs in run­ners and back mus­cles in kayak­er­s), with no dif­fer­ences for untrained mus­cles.

    Some phys­i­o­log­i­cal changes, such as heart enlarge­ments, require years of increas­ingly intense prac­tice to emerge and take years to regress once train­ing is stopped. For exam­ple, Eriksson, Engstrom, Karl­berg, Salt­in, and Thoren (1971) found that swim­mers’ aer­o­bic abil­ity decreased by 29% five years after train­ing had stopped. The increased lungs and hearts of these swim­mers had not changed yet. Other changes are gained and lost more rapid­ly. For exam­ple, aer­o­bic power in bicy­clists (Burke et al., 1990) increases over 50% dur­ing the com­pet­i­tive sea­son every year. Female gym­nasts reduce the pro­por­tion of their body fat from aver­age lev­els by 50% dur­ing the com­pet­i­tive sea­son (Reilly & Secher, 1990). Within a week of no train­ing, swim­mers lose on aver­age 50% of the res­pi­ra­tory capac­ity of their mus­cles (Reil­ly, 1990b), but regain­ing this capac­ity takes con­sid­er­ably longer dur­ing retrain­ing.

  2. Pack­ing For Mars: The Curi­ous Sci­ence of Life in the Void, Mary Roach 2010; from the chap­ter “The Hor­i­zon­tal Stuff: What If You Never Got Out of Bed?”:

    The human body is a fru­gal con­trac­tor. It keeps the mus­cles and skele­ton as strong as they need to be, no more and no less. “Use it or lose it” is a basic mantra of the human body. If you take up jog­ging or gain thirty pounds, your body will strengthen your bones and mus­cles as need­ed. Quit jog­ging or lose the thirty pounds, and your frame will be appro­pri­ately down­sized. Mus­cle is regained in a mat­ter of weeks once astro­nauts return to earth (and bed-resters get out of bed), but bone takes three to six months to recov­er. Some stud­ies sug­gest that the skele­tons of astro­nauts on long-­du­ra­tion mis­sions never quite recov­er, and for this rea­son it’s bone that gets the most study at places like FARU.

    The body’s fore­man on call is a cell called the , embed­ded all through the matrix of the bone. Every time you go for a run or lift a heavy box, you cause minute amounts of dam­age to your bone. The osteo­cytes sense this and send in a repair team: osteo­clasts to remove the dam­aged cells, and osteoblasts to patch the holes with fresh ones. The repaving strength­ens the bone. This is why bone-­jar­ring exer­cise like jog­ging is rec­om­mended to beef up the bal­sa-­wood bones of thin, smal­l­-boned women of north­ern Euro­pean ances­try, whose genet­ics, post­menopause, will land them on the short list for hip replace­ment.

    Like­wise, if you stop jar­ring and stress­ing your bones - by going into space, or into a wheel­chair or a bed-rest study - this cues the strain-sens­ing osteo­clasts to have bone taken away. The human organ­ism seems to have a pen­chant for stream­lin­ing. Whether it’s mus­cle or bone, the body tries not to spend its resources on func­tions that aren’t serv­ing any pur­pose.

    Tom Lang, a bone expert at the Uni­ver­sity of Cal­i­for­nia, San Fran­cis­co, who has stud­ied astro­nauts, explained all this to me. He told me that a Ger­man doc­tor named fig­ured it out in the 1800s by study­ing X-rays of infants’ hips as they tran­si­tioned from crawl­ing to walk­ing. “A whole new evo­lu­tion of bone struc­ture takes place to sup­port the mechan­i­cal loads asso­ci­ated with walk­ing,” said Lang. “Wolff had the great insight that form fol­lows func­tion.”…

    How bad can it get? If you stay off your feet indef­i­nite­ly, will your body com­pletely dis­man­tle your skele­ton? Can humans become jel­ly­fish by never get­ting up? They can­not. Para­plegics even­tu­ally lose from 1/3 to 1/2 of their bone mass in the lower body. Com­puter mod­el­ing done by Den­nis Carter and his stu­dents at Stan­ford Uni­ver­sity sug­gests that a two-year mis­sion to Mars would have about the same effect on one’s skele­ton. Would an astro­naut return­ing from Mars run the risk of step­ping out of the cap­sule into Earth grav­ity and snap­ping a bone? Carter thinks so. It makes sense, given that extremely osteo­porotic women have been known to break a hip (ac­tu­al­ly, the top of the thigh­bone where it enters the pelvis) by doing noth­ing but shift­ing their weight while stand­ing. They don’t fall and break a bone; they break a bone and fall. And these women have typ­i­cally lost a good deal less than 50% of their bone mass.

  3. Body­builders & weightlifters lift weights and use drugs so heav­ily because they have to break through plateaus and over­come bod­ily home­osta­sis which cre­ates esca­lat­ing resis­tance to any fur­ther gains. The point of weightlift­ing is not to ‘train’ mus­cles in any sense (the nerves learn effi­ciency rel­a­tively quick­ly) but to inflict so much dam­age on mus­cles as to trick the body into over­re­pair­ing them. In “The Power and the Gory”, for­mer Mr Amer­ica Steve Micha­lik exem­pli­fies this end­less Sis­phyean strug­gle when, after being hos­pi­tal­ized for liver cysts due to his steroid abuse and unable to con­tinue exercise/protein loading/drug use, he lost ~110 pounds of mus­cle in 3 weeks (the diges­tion of which con­tributed to kid­ney fail­ure).↩︎

  4. “Excep­tional longevity is asso­ci­ated with decreased repro­duc­tion”; abstract:

    A num­ber of lead­ing the­o­ries of aging, namely The (Williams, 1957), The (Kirk­wood, 1977) and most recently The (Bowen and Atwood, 2004, 2010) sug­gest a trade­off between longevity and repro­duc­tion. While there has been an abun­dance of data link­ing longevity with reduced fer­til­ity in lower life forms, human data have been con­flict­ing. We assessed this trade­off in a cohort of genet­i­cally and socially homo­ge­neous Jew­ish cen­te­nar­i­ans (av­er­age age ~100 years). As com­pared with an Ashke­nazi cohort with­out excep­tional longevi­ty, our cen­te­nar­i­ans had fewer chil­dren (2.01 vs 2.53, p<0.0001), were older at first child­birth (28.0 vs 25.6, p<0.0001), and at last child­birth (32.4 vs 30.3, p<0.0001). The smaller num­ber of chil­dren was observed for male and female cen­te­nar­i­ans alike. The lower num­ber of chil­dren in both gen­ders together with the pat­tern of delayed repro­duc­tive matu­rity is sug­ges­tive of con­sti­tu­tional fac­tors that might enhance human life span at the expense of reduced repro­duc­tive abil­i­ty.

  5. “When do evo­lu­tion­ary expla­na­tions of belief debunk belief?”, Grif­fiths & Wilkins 2010:

    Resources allo­cated to form­ing true beliefs are resources unavail­able for mak­ing sperm or eggs, or fight­ing off the effects of aging by repair­ing dam­aged tis­sues. Mod­ern humans in first-­world coun­tries lead a shel­tered life and it is hard for us to appre­ci­ate just how direct these trade-offs can be. A dra­matic exam­ple comes from a small Aus­tralian mam­mal, the (Antech­i­nus Stu­ar­tii). In this and sev­eral related species a short, fren­zied mat­ing sea­son is fol­lowed by a period dur­ing which the male’s sex­ual organs regress and their immune sys­tem col­laps­es. Then all the males in the pop­u­la­tion die. The Antech­i­nus has lit­tle chance of sur­viv­ing to the next breed­ing sea­son and so it allo­cates all of its resources to the repro­duc­tive effort and none to tis­sue main­te­nance. There can be lit­tle doubt that if, like us, the Antech­i­nus had a mas­sively hyper­tro­phied cor­tex and engaged in a lot of costly think­ing, it would allow that neural tis­sue to decay in the mat­ing sea­son so as to allo­cate more resources to sperm pro­duc­tion and sex­ual com­pe­ti­tion.

  6. “The Evolved Self­-­Man­age­ment Sys­tem”, :

    Some years ago I drew atten­tion to the “para­dox of ”, a para­dox that must strike any evo­lu­tion­ary biol­o­gist who thinks about it. It’s this. When a per­son’s health improves under the influ­ence of placebo med­ica­tion, then, as we’ve noted already, this has to be a case of “self­-cure”. But if peo­ple have the capac­ity to heal them­selves by their own efforts, why not get on with it as soon as need­ed? Why wait for per­mis­sion - from a sugar pill, a witch doc­tor - that it’s time to get bet­ter?

    Pre­sum­ably the expla­na­tion must be that self­-cure has costs as well as ben­e­fits. What kind of costs are the­se? Well, actu­ally they’re fairly obvi­ous. Many of the ill­nesses we expe­ri­ence, like pain, fever and so on, are actu­ally defenses which are designed to stop us from get­ting into more trou­ble than we’re already in. So “cur­ing” our­selves of these defenses can indeed cost us down the line. Pain reduces our mobil­i­ty, for exam­ple, and stops us from harm­ing our­selves fur­ther; so, reliev­ing our­selves of pain is actu­ally quite risky. Fever helps kill bac­te­r­ial par­a­sites by rais­ing body tem­per­a­ture to a level they can’t tol­er­ate; so again, cur­ing our­selves of fever is risky. Vom­it­ing gets rid of tox­ins; so sup­press­ing vom­it­ing is risky. The same goes for the deploy­ment of the immune sys­tem. Mount­ing an immune response is ener­get­i­cally expen­sive. Our meta­bolic rate rises 15% or so, even if we’re just respond­ing to a com­mon cold. What’s more, when we make anti­bod­ies we use up rare nutri­ents that will later have to be replaced. Given these costs, it becomes clear that imme­di­ate self­-cure from an occur­rent ill­ness is not always a wise thing to do. In fact there will be cir­cum­stances when it would be best to hold back from deploy­ing par­tic­u­lar heal­ing mea­sures because the antic­i­pated ben­e­fits are not likely to exceed the antic­i­pated costs. In gen­eral it will be wise to err on side of cau­tion, to play safe, not to let down our defenses such as pain or fever until we see signs that the dan­ger has passed, not to use up our stock of ammu­ni­tion against par­a­sites until we know we’re in rel­a­tively good shape and there’s not still worse to come. Heal­ing our­selves involves - or ought to involve - a judg­ment cal­l…There’s plenty of evi­dence that we have just such a sys­tem at work over­see­ing our health. For exam­ple, in win­ter, we are cau­tious about deploy­ing our immune resources. That’s why a cold lasts much longer in win­ter than it does in sum­mer. It’s not because we’re cold, it’s because our bod­ies, based on deep evo­lu­tion­ary his­tory reckon that it’s not so safe to use our immune resources in win­ter, as it would be in sum­mer. There’s exper­i­men­tal con­fir­ma­tion of this in ani­mals. Sup­pose a ham­ster is injected with bac­te­ria which makes it sick - but in one case the ham­ster is on an arti­fi­cial day/night cycle that sug­gests it’s sum­mer; in the other case it’s on a cycle that sug­gests it’s win­ter. If the ham­ster is tricked into think­ing it’s sum­mer, it throws every­thing it has got against the infec­tion and recov­ers com­plete­ly. If it thinks it’s win­ter then it just mounts a hold­ing oper­a­tion, as if it’s wait­ing until it knows it’s safe to mount a ful­l-s­cale response. The ham­ster “thinks” this or that?? No, of course it does­n’t think it con­sciously - the light cycle acts as a sub­con­scious prime to the ham­ster’s health man­age­ment sys­tem.

    Humphrey also goes on to point out that exploit­ing the placebo effect sat­is­fies one of Bostrom’s EOC cri­te­ria (which we haven’t dis­cussed yet):

    But, as I said, the world has changed - or at least is chang­ing for most of us. We no longer live in such an oppres­sive envi­ron­ment. We no longer need to play by the old rules, and rein in our pecu­liar strengths and idio­syn­crasies. We can afford to take risks now we could­n’t before. So, yes, I’m hope­ful. I think it really ought to be pos­si­ble to devise placebo treat­ments for the self, which do indeed induce them to come out from their pro­tec­tive shells - and so to emerge as hap­pier, nicer, clev­er­er, more cre­ative peo­ple than they would ever oth­er­wise have dared to be.

  7. A scratch or worse injuries can take weeks to heal ful­ly, yet human cells can repli­cate far faster and fill the equiv­a­lent vol­ume in hours. Such fast repair has obvi­ous sur­vival val­ue, so why don’t we? seems to pre­dict heal­ing fairly accu­rately by a rough cal­cu­la­tion of the meta­bolic expen­di­ture of such all-out heal­ing and assum­ing that the body only has some meta­bolic energy to spare at any time.↩︎

  8. & “Con­verg­ing Cog­ni­tive Enhance­ments” (2006):

    Keep­ing awake using stim­u­lants pre­vents the mem­ory con­sol­i­da­tion that would have taken place dur­ing sleep, and enhanced con­cen­tra­tion abil­ity may impair the abil­ity to notice things in periph­eral aware­ness.

  9. One of the sur­pris­ing facts about mem­ory is that it seems that every time , the mem­ory is effec­tively destroyed and must be recre­at­ed. This con­stant cycle of cre­ation and destruc­tion seems to be key in how works, and also explains the well-­doc­u­mented , the ease of induc­ing and ‘’ mem­o­ries, and the dra­matic effect of drugs on recall. An arti­cle on those drugs describes the active process:

    The dis­ap­pear­ance of the fear mem­ory sug­gested that every time we think about the past we are del­i­cately trans­form­ing its cel­lu­lar rep­re­sen­ta­tion in the brain, chang­ing its under­ly­ing neural cir­cuit­ry. It was a stun­ning dis­cov­ery: Mem­o­ries are not formed and then pristinely main­tained, as neu­ro­sci­en­tists thought; they are formed and then rebuilt every time they’re accessed. “The brain isn’t inter­ested in hav­ing a per­fect set of mem­o­ries about the past,” LeDoux says. “Instead, mem­ory comes with a nat­ural updat­ing mech­a­nism, which is how we make sure that the infor­ma­tion tak­ing up valu­able space inside our head is still use­ful. That might make our mem­o­ries less accu­rate, but it prob­a­bly also makes them more rel­e­vant to the future.”

    …What does do? The mol­e­cule’s cru­cial trick is that it increases the den­sity of a par­tic­u­lar type of sen­sor called an on the out­side of a neu­ron. It’s an ion chan­nel, a gate­way to the inte­rior of a cell that, when opened, makes it eas­ier for adja­cent cells to excite one anoth­er. (While neu­rons are nor­mally shy strangers, strug­gling to inter­act, PKMzeta turns them into inti­mate friends, happy to exchange all sorts of inci­den­tal infor­ma­tion.) This process requires con­stant upkeep - every long-term mem­ory is always on the verge of van­ish­ing. As a result, even a brief inter­rup­tion of PKMzeta activ­ity can dis­man­tle the func­tion of a stead­fast cir­cuit. If the genetic expres­sion of PKMzeta is amped up - by, say, genet­i­cally engi­neer­ing rats to over­pro­duce the stuff - they become mnemonic freaks, able to con­vert even the most mun­dane events into long-term mem­o­ry. (Their per­for­mance on a stan­dard test of recall is nearly dou­ble that of nor­mal ani­mal­s.) Fur­ther­more, once neu­rons begin pro­duc­ing PKMzeta, the pro­tein tends to linger, mark­ing the neural con­nec­tion as a mem­o­ry. “The mol­e­cules them­selves are always chang­ing, but the high level of PKMzeta stays con­stant,” Sack­tor says. “That’s what makes the endurance of the mem­ory pos­si­ble.” For exam­ple, in a recent exper­i­ment, Sack­tor and sci­en­tists at the Weiz­mann Insti­tute of Sci­ence trained rats to asso­ciate the taste of sac­cha­rin with nau­sea (thanks to an injec­tion of lithi­um). After just a few tri­als, the rats began stu­diously avoid­ing the arti­fi­cial sweet­en­er. All it took was a sin­gle injec­tion of a PKMzeta inhibitor called zeta-in­ter­act­ing pro­tein, or ZIP, before the rats for­got all about their aver­sion. The rats went back to guz­zling down the stuff.

  10. Bostrom/Sandberg 2006:

    Genetic mem­ory enhance­ment has been demon­strated in rats and mice. In nor­mal ani­mals, dur­ing mat­u­ra­tion expres­sion of the NR2B sub­unit of the (NMDA) recep­tor is grad­u­ally replaced with expres­sion of the NR2A sub­unit, some­thing that may be linked to the lower brain plas­tic­ity in adult ani­mals. Tsien’s group (Tang et al. 1999) mod­i­fied mice to over­ex­press the NR2B. The NR2B mice demon­strated improved mem­ory per­for­mance, both in terms of acqui­si­tion and reten­tion. This included unlearn­ing of fear con­di­tion­ing, which is believed to be due to the learn­ing of a sec­ondary mem­ory (Falls et al. 1992). The mod­i­fi­ca­tion also made the mice more sen­si­tive to cer­tain forms of pain, sug­gest­ing a non­triv­ial trade-off between two poten­tial enhance­ment goals (Wei et al. 2001).

    , Hills & Her­twig 2011:

    If bet­ter mem­o­ry, for exam­ple, is unequiv­o­cally ben­e­fi­cial, why do seem­ingly triv­ial neu­ro­mol­e­c­u­lar changes that would enhance mem­o­ry, such as the over-­ex­pres­sion of NMDA recep­tors in the hip­pocam­pus (Tang et al., 1999), not (to our knowl­edge) exist in nat­ural pop­u­la­tions? If it is so easy to evolve supe­rior cog­ni­tive capac­i­ties, why aren’t we smarter already?

  11. “For­get­ting is Key to a Healthy Mind: Let­ting go of mem­o­ries sup­ports a sound state of mind, a sharp intel­lect - and supe­rior recall”, Sci­en­tific Amer­i­can (con­so­nant with some of the sleep research on for­get­ting and some the­o­ries explain­ing ).↩︎

  12. Taxi dri­vers for­feit some things in exchange for their nav­i­ga­tional skills; see the BBC’s “Brain changes seen in cab­bies who take ‘The Knowl­edge’”; from (Wool­lett & Maguire 2011):

    The mem­ory pro­file dis­played by the now qual­i­fied trainees mir­rors exactly the pat­tern dis­played in sev­eral pre­vi­ous cross-­sec­tional stud­ies of licensed Lon­don taxi dri­vers [3, 4, 20] (and that which nor­mal­ized in the retired taxi dri­vers [21]). In those stud­ies also, the taxi dri­vers dis­played more knowl­edge of the spa­tial rela­tion­ships between land­marks in Lon­don, unsur­pris­ing­ly, given their increased expo­sure to the city com­pared to con­trol par­tic­i­pants. By con­trast, this enhanced spa­tial rep­re­sen­ta­tion of the city was accom­pa­nied by poorer per­for­mance on a com­plex fig­ure test, a visu­ospa­tial task designed to assess the free recall of visual mate­r­ial after 30 min. Our find­ings there­fore not only repli­cate those of pre­vi­ous cross-­sec­tional stud­ies but extend them by show­ing the change in mem­ory pro­file within the same par­tic­i­pants. That the only major dif­fer­ence between T1 and T2 was acquir­ing ‘’ strongly sug­gests that this is what induced the mem­ory change.

  13. Hills & Her­twig 2011:

    The ben­e­fits of lim­ited mem­ory have also been pro­posed to explain the curi­ous con­straints on work­ing-mem­ory span to a lim­ited num­ber of infor­ma­tion chunks (for sev­eral related exam­ples, see Her­twig & Todd, 2003)…As an exam­ple, work­ing mem­ory is cor­re­lated with per­for­mance on many cog­ni­tive tasks, such as the Scholas­tic Apti­tude Test. How­ev­er, indi­vid­u­als with high work­ing-mem­ory capac­ity often fail to hear their own name in a cock­tail-­party task and recall fewer items from a list after expe­ri­enc­ing a con­text change (see Unsworth & Engle, 2007). These results demon­strate that the effects of enhance­ments should be viewed as we view adap­ta­tions: Enhance­ment is only mean­ing­ful with respect to spe­cific indi­vid­u­als in spe­cific envi­ron­ments.

  14. Dis­cussing the the­ory of aging, “Under­stand­ing the Odd Sci­ence of Aging” (Kirk­wood 2005):

    Somatic main­te­nance needs only to be good enough to keep the organ­ism in sound phys­i­o­log­i­cal con­di­tion for as long as it has a rea­son­able chance of sur­vival in the wild. For exam­ple, since more than 90% of wild mice die in their first year (Phe­lan and Aus­tad, 1989), any invest­ment of energy in mech­a­nisms for sur­vival beyond this age ben­e­fits at most 10% of the pop­u­la­tion. Nearly all of the mech­a­nisms required for somatic main­te­nance and repair (DNA repair, antiox­i­dant sys­tems, etc.) require [sub­stan­tial] amounts of energy (ATP). Energy is scarce, as shown by the fact that the major cause of mor­tal­ity for wild mice is cold, due to fail­ure to main­tain (Berry and Bron­son, 1992). The mouse will there­fore ben­e­fit by invest­ing any spare energy into ther­mo­ge­n­e­sis or repro­duc­tion, rather than into bet­ter capac­ity for somatic main­te­nance and repair, even though this means that dam­age will even­tu­ally accu­mu­late to cause aging. The three­-year lifes­pan poten­tial of the mouse is suf­fi­cient for its actual needs in the wild, and yet it is not exces­sive, given that some mice will sur­vive into their sec­ond year and that age-re­lated dete­ri­o­ra­tion will become appar­ent before max­i­mum life span poten­tial is reached. Thus, it makes sense to sup­pose that the intrin­sic life span of the mouse has been opti­mized to suit its ecol­o­gy. The idea that intrin­sic longevity is tuned to the pre­vail­ing level of extrin­sic mor­tal­ity is sup­ported by exten­sive obser­va­tions on nat­ural pop­u­la­tions (Rick­lefs, 1998). Evo­lu­tion­ary adap­ta­tions such as flight, pro­tec­tive shells, and large brains, all of which tend to reduce extrin­sic mor­tal­i­ty, are asso­ci­ated with increased longevi­ty.

  15. Hills & Her­twig 2011:

    Per­haps the clear­est nat­ural evi­dence for between-­do­main trade-offs in per­for­mance across tasks comes from savants, whose spec­tac­u­lar skills in one domain are asso­ci­ated with poor per­for­mance in other domains. Those asso­ci­a­tions are not coin­ci­den­tal. -like skills can be induced in healthy par­tic­i­pants by turn­ing off par­tic­u­lar func­tional areas of the brain - for exam­ple, via repet­i­tive (Sny­der, 2009).

  16. The fads for anti-ox­i­dants and vit­a­mins are good exam­ples (what, the body can’t clean up oxi­dants already?) and so we some­times find evi­dence of harm; med­i­cine and psy­chol­ogy are per­haps because it’s so hard to find any­thing which works (“bad money dri­ves out good”). Med­ical econ­o­mist has blogged for years on var­i­ous ways in which med­i­cine is inef­fec­tive, expen­sive, or (as the evo­lu­tion­ary per­spec­tive would pre­dict would be the case in all but excep­tional cases like bro­ken bones, the low-hang­ing fruit which may have been dis­cov­ered as much as mil­len­nia ago); here is a selec­tion of his med­ical posts in chrono­log­i­cal order:

    1. RAND Health Insur­ance Exper­i­ment”
    2. “Dis­agree­ment Case Study: Han­son and Cut­ler”
    3. “Cut Med­i­cine in Half”; from ‘Is More Med­i­cine Bet­ter?’; see also Overtreated (Brown­lee 2008)
    4. “Hos­pice Beats Hos­pi­tal”
    5. “Eter­nal Med­i­cine”
    6. “Beware Trans­fu­sions”
    7. “Beware High Stan­dards”
    8. “Free Docs Not Help Poor Kids”
    9. “Avoid Vena Cava Fil­ters”
    10. “Ques­tion Med­ical Find­ings” (Stu­art Buck)
    11. “Med­ical Ide­ol­ogy”
    12. “Meds to Cut”
    13. “Our Nutri­tion Igno­rance”
    14. “Wasted Can­cer Hope”
    15. “Africa HIV: Per­verts or Bad Med?”
    16. “Megan on Med”
    17. Hard Facts: Med
    18. “In Favor of Fever”
    19. “Death Pan­els Add Life”
    20. “How Med Harms”
    21. “Beware Knives”
    22. “The Ore­gon Health Insur­ance Exper­i­ment”
    23. “Skip Can­cer Screens”
    24. “Beware Can­cer Med”
    25. “For­get Salt”
    26. “All In Their Heads”
    27. “Don’t Tor­ture Mom & Dad”
    28. “Dog vs. Cat Med­i­cine”
    29. “Farm vs Pet Med­i­cine”
    30. “1/6 of US Deaths From Hos­pi­tal Errors”
  17. , Maximes 269↩︎

  18. Not all psy­cho­log­i­cal traits seem sim­ply good; often seem like too extreme a score could be a pretty bad thing (most obvi­ously for Neu­roti­cism and Open­ness). This may be vin­di­cated by look­ing at the influ­ence of genes on the two dif­fer­ent cat­e­gories. From Stanovich 2010, Ratio­nal­ity & The Reflec­tive Mind:

    For many years, evo­lu­tion­ary psy­chol­ogy had lit­tle to say about indi­vid­ual dif­fer­ences because the field had as a foun­da­tional assump­tion that nat­ural selec­tion would elim­i­nate her­i­ta­ble dif­fer­ences because her­i­ta­ble traits would be dri­ven to fix­a­tion (Buss, 2009 [The Hand­book of Evo­lu­tion­ary Psy­chol­ogy]). recently how­ev­er, evo­lu­tion­ary psy­chol­o­gists have attempted to explain the con­trary evi­dence that vir­tu­ally all cog­ni­tive and per­son­al­ity traits that have been mea­sured have her­i­tabil­i­ties hov­er­ing around 50%. have pro­posed a the­ory that explains these indi­vid­ual dif­fer­ences. Inter­est­ing­ly, the the­ory accounts for her­i­ta­ble cog­ni­tive abil­ity dif­fer­ences in a dif­fer­ent way than it accounts for her­i­ta­ble cog­ni­tive think­ing dis­po­si­tions and per­son­al­ity vari­ables. the basis of their the­ory is a dis­tinc­tion that I will be stress­ing through­out this book - that between typ­i­cal per­for­mance indi­ca­tors and opti­mal per­for­mance indi­ca­tors.

    Penke et al (2007) argue that “the clas­si­cal dis­tinc­tion between cog­ni­tive abil­i­ties and per­son­al­ity traits is much more than just a his­tor­i­cal con­ven­tion or a method­olog­i­cal mat­ter of dif­fer­ent mea­sure­ment approaches (Cron­bach, 1949 [Essen­tials of Psy­cho­log­i­cal Test­ing]), and instead reflects dif­fer­ent kinds of selec­tion pres­sures that have shaped dis­tinc­tive genetic archi­tec­tures for these two classes” (p.550) of indi­vid­ual dif­fer­ences. On their view, per­son­al­ity traits and think­ing dis­po­si­tions (re­flec­tive-level indi­vid­ual dif­fer­ences) rep­re­sent pre­served, her­i­ta­ble vari­abil­ity that is main­tained by dif­fer­ent bio­log­i­cal processes than intel­li­gence (al­go­rith­mic-level indi­vid­ual dif­fer­ences). Think­ing dis­po­si­tions and per­son­al­ity traits are main­tained by bal­anced selec­tion, most prob­a­bly fre­quen­cy-de­pen­dent selec­tion (Buss, 2009). The most famous exam­ple of the lat­ter is cheater-based per­son­al­ity traits that flour­ish when they are rare but become less adap­tive as the pro­por­tion of cheaters in the pop­u­la­tion rises (as cheaters begin to heat each oth­er), finally reach­ing an equi­lib­ri­um.

    In con­trast, vari­abil­ity in intel­li­gence is thought to be main­tained by con­stant changes in muta­tion load (Buss, 2009; Penke et al, 2007). As Pinker (2009 pg46) notes:

    …new muta­tions creep into the genome faster than nat­ural selec­tion can weed them out. At any given moment, the pop­u­la­tion is laden with a port­fo­lio of recent muta­tions, each of whose days are num­bered. This Sisyphean strug­gle between selec­tion and muta­tion is com­mon with traits that depend on many genes, because there are so many things that can go wrong…Un­likely per­son­al­i­ty, where it takes all kinds to make a world, with intel­li­gence, smarter is sim­ply bet­ter, so bal­anc­ing selec­tion is unlike­ly. But intel­li­gence depends on a large net­work of brain areas, and it thrives in a body that is prop­erly nour­ished and free of dis­eases and defect­s…­Mu­ta­tions in gen­eral are far more likely to be harm­ful than help­ful, and the large, help­ful ones were low-hang­ing fruit that were picked long ago in our evo­lu­tion­ary his­tory and entrenched in the species…But as the bar­rel gets closer to the tar­get, smaller and smaller tweaks are needed to bring any fur­ther improve­men­t…Though we know that genes for intel­li­gence must exist, each is likely to be small in effect, found in only a few peo­ple, or both [In a recent study of 6,000 chil­dren, the gene with the biggest effect accounted for less than one-quar­ter of an I.Q. point.]…The hunt for per­son­al­ity genes, though not yet Nobel-­wor­thy, has had bet­ter for­tunes. Sev­eral asso­ci­a­tions have been found between per­son­al­ity traits and genes that gov­ern the break­down, recy­cling or detec­tion of neu­ro­trans­mit­ters

    See also .↩︎

  19. Hull 2001, pg37; Sci­ence and selec­tion: Essays on bio­log­i­cal evo­lu­tion and the phi­los­o­phy of sci­ence.↩︎

  20. Rich­er­son & Boyd 2005, pg135; Not by genes alone: How cul­ture trans­formed human evo­lu­tion.↩︎

  21. pg 18, Table 1.1, “Some Alter­na­tive Terms for Type I and Type 2 Pro­cess­ing Used by Var­i­ous The­o­rists”, Ratio­nal­ity & The Reflec­tive Mind col­lates the fol­low­ing syn­onyms:

    auto­matic pro­cess­ing vs con­scious pro­cess­ing; want self vs should self; online think­ing vs offline think­ing; gist pro­cess­ing vs ana­lytic pro­cess­ing; heuris­tic pro­cess­ing vs sys­tem­atic pro­cess­ing; heuris­tic pro­cess­ing vs ana­lytic pro­cess­ing; tacit thought processes vs explicit thought process­es; type 1 processes vs type 2 process­es; mod­u­lar processes vs cen­tral process­es; asso­cia­tive processes vs propo­si­tional process­es; intu­itive sys­tem vs rea­son­ing sys­tem; implicit infer­ences vs explicit infer­ences; intu­ition vs rea­son­ing; reflex­ive sys­tem vs reflec­tive sys­tem; vis­ceral fac­tors vs tastes; hot sys­tem vs cool sys­tem; con­tention sched­ul­ing vs super­vi­sory atten­tional sys­tem; quick & inflex­i­ble mod­ules vs intel­lec­tion; auto­matic acti­va­tion vs con­scious pro­cess­ing; implicit cog­ni­tion vs explicit learn­ing; auto­matic pro­cess­ing vs con­trolled pro­cess­ing; asso­cia­tive sys­tem vs rule-based sys­tem; asso­ciate pro­cess­ing vs rule-based pro­cess­ing; impul­sive sys­tem vs reflec­tive sys­tem; doer vs plan­ner; stim­u­lus-bound vs higher order; adap­tive uncon­scious vs con­scious.↩︎

  22. The abstract:

    Human beings are a mar­vel of evolved com­plex­i­ty. Such sys­tems can be dif­fi­cult to enhance. When we manip­u­late com­plex evolved sys­tems, which are poorly under­stood, our inter­ven­tions often fail or back­fire. It can appear as if there is a “wis­dom of nature” which we ignore at our per­il. Some­times the belief in nature’s wis­dom - and cor­re­spond­ing doubts about the pru­dence of tam­per­ing with nature, espe­cially human nature - man­i­fest as dif­fusely moral objec­tions against enhance­ment. Such objec­tions may be expressed as intu­itions about the supe­ri­or­ity of the nat­ural or the trou­ble­some­ness of hubris, or as an eval­u­a­tive bias in favor of the sta­tus quo. This chap­ter explores the extent to which such pru­dence-derived anti-en­hance­ment sen­ti­ments are jus­ti­fied. We develop a heuris­tic, inspired by the field of evo­lu­tion­ary med­i­cine, for iden­ti­fy­ing promis­ing human enhance­ment inter­ven­tions. The heuris­tic incor­po­rates the grains of truth con­tained in “nature knows best” atti­tudes while pro­vid­ing cri­te­ria for the spe­cial cases where we have rea­son to believe that it is fea­si­ble for us to improve on nature.

  23. From the review “The pop­u­la­tion genet­ics of ben­e­fi­cial muta­tions”, Orr 2010:

    Under these so-­called strong-s­e­lec­tion weak-­mu­ta­tion con­di­tions, the pop­u­la­tion is essen­tially made up of a sin­gle wild-­type DNA sequence….Each of these [pos­si­ble] sequences, includ­ing the wild-­type, is assigned a [re­pro­duc­tive] fit­ness from some dis­tri­b­u­tion. The key point, how­ev­er, is that this over­all dis­tri­b­u­tion of fit­ness is unknown. Despite this, we do know two things. First, the wild-­type allele is highly fit; indeed it is fit­ter than all of its m mutant “neigh­bour” sequences (this is why it is wild-­type). Sec­ond, any ben­e­fi­cial muta­tions would be even fit­ter and so would fall even far­ther out in the tail of the fit­ness dis­tri­b­u­tion. (We assume for now that this tail falls off in some “ordi­nary” smooth way; see below.) At some point in time, the envi­ron­ment changes and the wild-­type allele slips slightly in fit­ness and one or more of the m muta­tions becomes ben­e­fi­cial. The ques­tion is: what is the size of the fit­ness gap between the wild-­type and a ben­e­fi­cial sequence?

    To answer this, Gille­spie assumed that only one ben­e­fi­cial muta­tion is avail­able. Tak­ing advan­tage of an obscure part of EVT con­cerned with “extreme spac­ings”, he showed that, more or less inde­pen­dently of the shape of the unknown over­all dis­tri­b­u­tion of fit­ness, this fit­ness gap - the fit­ness effects of new ben­e­fi­cial muta­tions - is expo­nen­tially dis­trib­uted. This result was later gen­er­al­ized by Orr (2003) to any mod­est num­ber of ben­e­fi­cial muta­tions (i.e. the wild-­type sequence might mutate to 5 or 10 or so dif­fer­ent ben­e­fi­cial muta­tion­s). Muta­tion should thus often yield ben­e­fi­cial alle­les of small effect and rarely yield those of large effect. In ret­ro­spect, it is clear that this expo­nen­tial dis­tri­b­u­tion of ben­e­fi­cial effects is a sim­ple con­se­quence of a well-­known result from so-­called peak­s-over-thresh­old mod­els in EVT (Lead­bet­ter et al. 1983). A large set of results, con­cern­ing both the first sub­sti­tu­tion dur­ing adap­ta­tion and the prop­er­ties of entire adap­tive walks to local optima rest on this result (Orr 2002, 2004, 2005; Rokyta et al. 2005)…Un­for­tu­nate­ly, the data avail­able thus far from the rel­e­vant exper­i­ments are mixed.

  24. For exam­ple, Baker et al 2002, Knight et al 1999 & Rauch et al 2012 & Deci­pher­ing Devel­op­men­tal Dis­or­ders Study 2015 & Gilis­sen et al 2014 do some genet­ics work with retarded chil­dren and turn up all sorts of muta­tions & change, while Zech­ner et al 2001 observes that lit­er­ally hun­dreds of muta­tions (many on the X chro­mo­some) have been linked with retar­da­tion.↩︎

  25. Autism like­wise impairs intel­li­gence mas­sive­ly, and seems to be caused by de novo muta­tions; see Ios­si­fov et al 2014, De Rubeis et al 2014, & Yuen et al 2015.

    In con­trast, the search for genet­ics lead­ing to greater intel­li­gence is still in its infancy with ten­ta­tive find­ings of small effect.↩︎

  26. “Indi­vid­ual organ­isms are best thought of as adap­ta­tion-ex­e­cuters rather than as fit­ness-­max­i­miz­ers.”↩︎

  27. ; quoted in Misha Gro­mow’s Struc­tures, Learn­ing and Ergosys­tems↩︎

  28. One old obser­va­tion in is that ani­mals and mam­mals in par­tic­u­lar have par­tic­u­lar math­e­mat­i­cal rela­tion between heart rate, mass, and longevity - except humans are an out­lier, liv­ing almost twice as long as they ‘should’, even com­pared to chim­panzees. See “Human Longevity Com­pared to Mam­mals”, “Ani­mal Longevity and Scale”; cf. the .↩︎

  29. “Adap­tive no more: A poten­tial ben­e­fit in pre­his­toric lean times, genetic vari­ant may increase risk of ges­ta­tional dia­betes today”, or see the .↩︎

  30. IQ scores of any reli­a­bil­ity are unavail­able from before the early 1990s; broad esti­mates from dys­genic con­sid­er­a­tions and com­puter mod­els (of uncer­tain reli­a­bil­i­ty) sug­gest geno­typic IQs (ceil­ings given good envi­ron­ment, nutri­tion etc) for west­ern Europe in the mid­dle 90s.↩︎

  31. I should note the authors claim their data shows that they found a Flynn effect and more­over, “The effect was found in the top 5% at a rate sim­i­lar to the gen­eral dis­tri­b­u­tion, pro­vid­ing evi­dence for the first time that the entire curve is likely increas­ing at a con­stant rate.” I dis­agree with this inter­pre­ta­tion; the scores increases come solely from the math­e­mat­i­cal sub­tests. As they acknowl­edge on page 5:

    In con­trast to the math­e­mat­i­cal abil­ity results, the ACT-S, ACT-E, and SAT-V all indi­cated a slight decrease (− 0.05 for the ACT-S and SAT-V and − 0.06 for the ACT-E). For 7th-­grade stu­dents the only ver­bal test that demon­strated a slight gain was the ACT-R (0.09). Appen­dixes A and B show increas­ing vari­ances for the SAT-V and ACT-R, but fairly sta­ble or slightly decreas­ing vari­ances for the ACT-S and ACT-E. There­fore, the small com­pos­ite gains on the SAT and ACT were gen­er­ally com­posed of large gains on the math sub­tests and slight losses on the sci­ence and ver­bal sub­tests.

    This is more than a lit­tle strange if the Flynn effect is gen­uinely oper­at­ing, as an increase in Gf ought to increase scores on all sub­tests; it is more con­sis­tent with pro­saic expla­na­tions like the increased empha­sis on math edu­ca­tion slightly increas­ing scores.↩︎

  32. See or look at the lit­er­a­ture, eg. “Pro­found effects of com­bin­ing choline and pirac­etam on mem­ory enhance­ment and cholin­er­gic func­tion in aged rats”↩︎

  33. But note that found the IQ increase was gone by the 7 year fol­lowup, the lat­est in a long line of infancy or early child­hood inter­ven­tions to dis­cover that promis­ing early IQ gains “faded out”.↩︎

  34. Empha­sis added; Bostrom/Sandberg 2006.↩︎

  35. Lynch et al 2011:

    There is a large and often con­flict­ing lit­er­a­ture on the effects of modafinil on com­po­nents of cog­ni­tion. Some stud­ies obtained a clear improve­ment in sus­tained atten­tion in healthy human sub­jects (Ran­dall et al., 2005) but oth­ers failed to find such effects (Turner et al., 2003). Sim­i­lar dis­crep­an­cies occur in the lit­er­a­ture on ani­mals (Waters et al., 2005). A recent, mul­ti­-­fac­to­r­ial analy­sis pro­vided con­vinc­ing evi­dence that mod­er­ate doses of modafinil improve atten­tion in healthy mid­dle-aged rats with­out affect­ing moti­va­tion or loco­mo­tor activ­ity (Mor­gan et al., 2007). Impor­tant­ly, these effects became evi­dent only as atten­tional demands were increased. In all, it seems rea­son­able at this point to con­clude that modafinil’s effects on basic psy­cho­log­i­cal state vari­ables - wake­ful­ness - can trans­late into selec­tive improve­ments in atten­tion.

    There is also a siz­able lit­er­a­ture sug­gest­ing that the above con­clu­sion can be extended to mem­ory encod­ing. An intrigu­ing aspect of these stud­ies in rodents (Bera­cochea et al., 2002) and humans (Turner et al., 2003; Baran­ski et al., 2004; Muller et al., 2004; Ran­dall et al., 2005) is that they gen­er­ally point to a drug influ­ence on work­ing mem­ory as opposed to the encod­ing of long-term mem­ory for spe­cific infor­ma­tion (Minzen­berg and Carter, 2008). (A sim­i­lar argu­ment was made ear­lier for Rital­in.) For exam­ple, the above noted work on mid­dle-aged rats found no evi­dence for accel­er­ated acqui­si­tion of a visual dis­crim­i­na­tion prob­lem, with min­i­mal demands on work­ing mem­o­ry, despite clear improve­ments in atten­tion. There are, how­ev­er, stud­ies show­ing that modafinil accel­er­ates the acqui­si­tion of sim­ple rules (‘win-s­tay’) (Bera­cochea et al., 2003), a spa­tial learn­ing pro­to­col (), and a non-­match to posi­tion prob­lem (Ward et al., 2004) in rodents. It is tempt­ing to spec­u­late that we are here see­ing hier­ar­chi­cal effects of modafinil such that enhanced wake­ful­ness pro­duces greater atten­tion that in turn improves both work­ing mem­ory and sim­ple rule learn­ing.

    But does the above sequence in fact improve the inte­gra­tive psy­cho­log­i­cal processes that con­sti­tute cog­ni­tion? By far the greater part of the human stud­ies with modafinil involves sub­jects with impair­ments to per­for­mance (sleep depri­va­tion) or psy­chi­atric dis­or­ders. None of the ani­mal stud­ies used recently devel­oped tests (see below) that are explic­itly intended to assess vari­ables such as recall vs. recog­ni­tion or ‘top-­down’ forc­ing of atten­tion. This leaves a small set of exper­i­ments involv­ing per­for­mance by healthy human sub­jects on rel­a­tively sim­ple learning/perceptual prob­lems. A ret­ro­spec­tive analy­sis of sev­eral stud­ies led to the con­clu­sion that modafinil does not pro­duce a ‘global’ enhance­ment of cog­ni­tion (Ran­dall et al., 2005).

  36. Tononi & Cirelli 2006:

    About 40% of the energy require­ments of the cere­bral cor­tex - by far the most meta­bol­i­cally expen­sive tis­sue in the body - are due to neu­ronal repo­lar­iza­tion fol­low­ing post­sy­nap­tic poten­tials.67 The higher the synap­tic weight imping­ing on a neu­ron, the higher this por­tion of the energy bud­get. More­over, increased synap­tic weight is thought to lead to increased aver­age fir­ing rates,68 and spikes in turn are respon­si­ble for another 40% of the gray mat­ter energy bud­get.67 There­fore, it would seem ener­get­i­cally pro­hib­i­tive for the brain to let synap­tic weight grow with­out checks as a result of wak­ing plas­tic­i­ty. Indeed, if PET data11 offer any indi­ca­tion, after just one wak­ing day energy expen­di­ture may grow by as much as 18%.

    …An­other ben­e­fit of synap­tic downscaling/downselection dur­ing sleep would be in terms of space require­ments. Synap­tic strength­en­ing is thought to be accom­pa­nied by mor­pho­log­i­cal changes, includ­ing increased size of ter­mi­nal bou­tons and spines, and synapses may even grow in num­ber (e.g.3,4,63). But space is a pre­cious com­mod­ity in the brain, and even minus­cule increases in vol­ume are extremely dan­ger­ous. For exam­ple, neo­cor­ti­cal gray mat­ter is tightly packed, with wiring (ax­ons and den­drites) tak­ing up ~60% of the space, synap­tic con­tacts (bou­tons and spines) ~20%, and the rest (cell bod­ies, ves­sels, extra­cel­lu­lar space) the remain­ing 20%.69 Thus, sleep would be impor­tant not just to keep in check the meta­bolic cost of strength­ened synaps­es, but also to curb their demands on brain real estate.

  37. “Acetyl­choline and synap­tic home­osta­sis”, Fink et al 2012, sug­gests the mech­a­nism of synap­tic upscal­ing & down­scal­ing to be related to acetyl­choline:

    We pro­pose that the influ­ence of acetyl­choline (ACh) may pro­vide a mech­a­nism for both upscal­ing and down­scal­ing of cor­ti­cal synaps­es. Specif­i­cal­ly, exper­i­men­tal stud­ies have shown that ACh mod­u­la­tion switches the phase response curves of cor­ti­cal pyra­mi­dal cells from Type II to Type I. Our com­pu­ta­tional stud­ies of cor­ti­cal net­works show that the pres­ence of ACh induces cel­lu­lar and net­work dynam­ics which lead to net synap­tic poten­ti­a­tion under a stan­dard STDP rule, while the absence of ACh alters dynam­ics in such a way that the same STDP rule leads to net depo­ten­ti­a­tion (see Fig. 1). Thus the well-estab­lished preva­lence of ACh in cor­ti­cal cir­cuits dur­ing wak­ing may lead to global synap­tic poten­ti­a­tion, while the absence of ACh dur­ing NREM sleep may lead to global depo­ten­ta­tion.

    ACh is bro­ken down by acetyl­cholinesterase, and ACh recep­tors can also be clas­si­fied as s; anti­cholin­er­gics like cause seda­tion, “brain fog”, and have been used to treat insom­nia. If the absence of ACh enables the down­scal­ing, could the process be sped up by inter­ven­ing with such drugs dur­ing sleep? Alter­nate­ly, the the­ory could be tested by inter­ven­ing with the oppo­site drugs and test­ing how high brain caloric con­sump­tion is upon wak­ing. (These would have to be ani­mal stud­ies; drugs like atropine or scopo­lamine are dan­ger­ous to use in human­s.)

    If this the­ory is borne out, it may sug­gest a more pre­cise excep­tion for pirac­etam or cholin­er­gics in gen­er­al: increas­ing poten­ti­a­tion may make the later depo­ten­ti­a­tion “too expen­sive” either ener­gy- or time-­wise. Addi­tional pre­dic­tions may be that: peo­ple who are more intel­li­gent (thanks to hav­ing ACh upreg­u­lated for what­ever rea­son) will need more sleep or ener­gy; use of cholin­er­gics will increase sleep or energy needs in nor­mal peo­ple. Retro­d­ic­tions include babies sleep­ing a great deal and the elderly sleep­ing less. (Since mem­ory for­ma­tion is already strongly linked to sleep and may increase sleep need itself, this is a con­found that needs to be taken into account along with oth­ers like the major sleep dis­tur­bances of the elderly & lack of mela­tonin secre­tion.)↩︎

  38. See Ever­son 1993, Ever­son 1995, Ever­son & Toth 2000; but also Rechtschaf­fen & Bergmann 1995 & Bergmann et al 1996, Rechtschaf­fen & Bergmann 2001, & Rechtschaf­fen & Bergmann 2002.↩︎

  39. , Brad­bury 2005:

    A big­ger, more com­plex brain may have advan­tages over a small brain in terms of com­put­ing pow­er, but brain expan­sion has costs. For one thing, a big brain is a meta­bolic drain on our bod­ies. Indeed, some peo­ple argue that, because the brain is one of the most meta­bol­i­cally expen­sive tis­sues in our body, our brains could only have expanded in response to an improved diet. Another cost that goes along with a big brain is the need to reor­gan­ise its wiring. “As brain size increas­es, sev­eral prob­lems are cre­ated”, explains sys­tems neu­ro­bi­ol­o­gist Jon Kaas (Van­der­bilt Uni­ver­si­ty, Nashville, Ten­nessee, United States). “The most seri­ous is the increased time it takes to get infor­ma­tion from one place to anoth­er.” One solu­tion is to make the axons of the neu­rons big­ger but this increases brain size again and the prob­lem esca­lates. Another solu­tion is to do things local­ly: only con­nect those parts of the brain that have to be con­nect­ed, and avoid the need for com­mu­ni­ca­tion between hemi­spheres by mak­ing dif­fer­ent sides of the brain do dif­fer­ent things. A big brain can also be made more effi­cient by organ­is­ing it into more sub­di­vi­sions, “rather like split­ting a com­pany into depart­ments”, says Kaas. Over­all, he con­cludes, because a big­ger brain per se would not work, brain reor­gan­i­sa­tion and size increase prob­a­bly occurred in par­al­lel dur­ing human brain evo­lu­tion. The end result is that the human brain is not just a scaled-up ver­sion of a mam­mal brain or even of an ape brain.

  40. Hills & Her­twig 2011:

    Con­sider the human female pelvis. Because its dimen­sions are small rel­a­tive to a baby’s head, obstet­ric com­pli­ca­tions dur­ing labor are com­mon. Why has­n’t evo­lu­tion improved the sur­vival chances of both mother and baby by select­ing for a larger female pelvis? The widely accepted expla­na­tion is that the opti­mal pelvis for bipedal loco­mo­tion and the opti­mal pelvis for encephal­iza­tion (the pro­gres­sive increase in the baby’s brain size) place com­pet­ing demands on the human pelvis. Bipedal loco­mo­tion requires sub­stan­tial skele­tal changes, includ­ing alter­ations in the pelvic archi­tec­ture (Wittman & Wall, 2007), and such changes must com­pete (in an evo­lu­tion­ary sense) with the obstet­ric demands of human babies’ rel­a­tively large brains.

  41. pg 2, “The like­li­hood of cog­ni­tive enhance­ment”, Lynch et al 2011:

    Anatomists often resort to allom­e­try when deal­ing with ques­tions of selec­tive pres­sures on brain regions. Applied to brain pro­por­tions, this involves col­lect­ing mea­sure­ments for the region of inter­est - e.g., frontal cor­tex - for a series of ani­mals within a given tax­o­nomic group and then relat­ing it to the vol­ume or weight of the brains of those ani­mals. This can estab­lish with a rel­a­tively small degree of error whether a brain com­po­nent in a par­tic­u­lar species is larger than would be pre­dicted from that species’ brain size. While there is not a great deal of evi­dence, stud­ies of this type point to the con­clu­sion that cor­ti­cal sub­di­vi­sions in humans, includ­ing asso­ci­a­tion regions, are about as large as expected for an anthro­poid pri­mate with a 1350 cm3 brain. The vol­ume of area 10 of human frontal cor­tex, for exam­ple, fits on the regres­sion line (area 10 vs. whole brain) cal­cu­lated from pub­lished data (Semende­feri et al., 2001) for a series com­posed of gib­bons, apes and humans (Lynch and Granger, 2008 [Big brain: the ori­gins and future of human intel­li­gence]). Given that this region is widely assumed to play a cen­tral role in exec­u­tive func­tions and work­ing mem­o­ry, these obser­va­tions do not encour­age the idea that selec­tive pres­sures for cog­ni­tion have dif­fer­en­tially shaped the pro­por­tions of human cor­tex. Impor­tant­ly, this does not mean that those pro­por­tions are in any sense typ­i­cal. The allo­met­ric equa­tions involve dif­fer­ent expo­nents for dif­fer­ent regions, mean­ing that absolute pro­por­tions (e.g., pri­mary sen­sory cor­tex vs. asso­ci­a­tion cor­tex) change as brains grow larg­er. The bal­ance of parts in the cor­tex of the enor­mous human brain is dra­mat­i­cally dif­fer­ent than found in the much smaller mon­key brain: area 10, for instance, occu­pies a much greater per­cent­age of the cor­tex in man. But these effects seem to reflect expan­sion accord­ing to rules embed­ded in a con­served brain plan rather than selec­tion for the spe­cific pat­tern found in humans ().

  42. , Her­cu­lano-Houzel 2012:

    Humans also do not rank first, or even close to first, in rel­a­tive brain size (ex­pressed as a per­cent­age of body mass), in absolute size of the cere­bral cor­tex, or in gyri­fi­ca­tion (3). At best, we rank first in the rel­a­tive size of the cere­bral cor­tex expressed as a per­cent­age of brain mass, but not by far. Although the human cere­bral cor­tex is the largest among mam­mals in its rel­a­tive size, at 75.5% (4), 75.7% (5), or even 84.0% (6) of the entire brain mass or vol­ume, other ani­mals, pri­mate and non­pri­mate, are not far behind: The cere­bral cor­tex rep­re­sents 73.0% of the entire brain mass in the chim­panzee (7), 74.5% in the horse, and 73.4% in the short­-finned whale (3).

    …If encephal­iza­tion were the main deter­mi­nant of cog­ni­tive abil­i­ties, smal­l­-brained ani­mals with very large encephal­iza­tion quo­tients, such as capuchin mon­keys, should be more cog­ni­tively able than large-brained but less encephal­ized ani­mals, such as the gorilla (2). How­ev­er, the for­mer ani­mals with a smaller brain are out­ranked by the lat­ter in cog­ni­tive per­for­mance (13).

    …How­ev­er, this notion is in dis­agree­ment with the obser­va­tion that ani­mals of sim­i­lar brain size but belong­ing to dif­fer­ent mam­malian orders, such as the cow and the chim­panzee (both at about 400 g of brain mass), or the rhe­sus mon­key and the capy­bara (at 70-80 g of brain mass), may have strik­ingly dif­fer­ent cog­ni­tive abil­i­ties and behav­ioral reper­toires.

    …De­spite com­mon remarks in the lit­er­a­ture that the human brain con­tains 100 bil­lion neu­rons and 10- to 50-­fold more glial cells (e.g., 57-59), no ref­er­ences are given to sup­port these state­ments; to the best of my knowl­edge, they are none other than ball­park esti­mates (60). Com­par­ing the human brain with other mam­malian brains thus required first esti­mat­ing the total num­bers of neu­ronal and non­neu­ronal cells that com­pose these brains, which we did a few years ago (25). Remark­ably, at an aver­age of 86 bil­lion neu­rons and 85 bil­lion non­neu­ronal cells (25), the human brain has just as many neu­rons as would be expected of a generic pri­mate brain of its size and the same over­all 1:1 nonneuronal/ neu­ronal ratio as other pri­mates (26). Bro­ken down into the cere­bral cor­tex, cere­bel­lum, and rest of the brain, the neu­ronal scal­ing rules that apply to pri­mate brains also apply to the human brain (25) (Fig. 3 A and C, arrows). Neu­ronal den­si­ties in the cere­bral cor­tex and cere­bel­lum also fit the expected val­ues in humans as in other pri­mate species (Fig. 3B), and the ratio between non­neu­ronal and neu­ronal cells in the whole human brain of 1:1 (not 10:1, as com­monly report­ed) is sim­i­lar to that of other pri­mates (25). The num­ber of neu­rons in the gray mat­ter alone of the human cere­bral cor­tex, as well as the size of the sub­cor­ti­cal white mat­ter and the num­ber of non­neu­ronal cells that it con­tains, also con­forms to the rules that apply to other pri­mates ana­lyzed (47). Most impor­tant­ly, even though the rel­a­tive expan­sion of the human cor­tex is fre­quently equated with brain evo­lu­tion, which would have reached its crown­ing achieve­ment in us (61), the human brain has the ratio of cere­bel­lar to cere­bral cor­ti­cal neu­rons pre­dicted from other mam­mals, pri­mate and non­pri­mate alike (Fig. 4A).

  43. Her­cu­lano-Houzel 2012:

    Con­trary to expec­ta­tions, divid­ing total glu­cose use per minute in the cere­bral cor­tex or whole brain (69) by the num­ber of brain neu­rons revealed a remark­ably con­stant aver­age glu­cose use per neu­ron across the mouse, rat, squir­rel, mon­key, baboon, and human, with no sig­nif­i­cant rela­tion­ship to neu­ronal den­sity and, there­fore, to aver­age neu­ronal size (70). This is in con­trast to the decreas­ing aver­age meta­bolic cost of other cell types in mam­malian bod­ies with increas­ing cell size (71-73), with the sin­gle pos­si­ble excep­tion of mus­cle fibers (74). The higher lev­els of expres­sion of genes related to metab­o­lism in human brains com­pared with chim­panzee and mon­key brains (75, 76) might there­fore be related not to an actual increase in metab­o­lism per cell but to the main­te­nance of aver­age neu­ronal metab­o­lism in the face of decreas­ing metab­o­lism in other cell types in the body. That the aver­age ener­getic cost per neu­ron does not scale with aver­age neu­ronal cell size has impor­tant phys­i­o­log­i­cal impli­ca­tions. First, con­sid­er­ing the oblig­a­tory increased cost related to a larger sur­face area (68), the evo­lu­tion of neu­rons with a con­stant aver­age ener­getic cost regard­less of their total cell size implies that the rela­tion­ship between larger neu­ronal size and a larger G/N ratio must not be related to increased meta­bolic needs, as usu­ally assumed.

    …Sec­ond, the con­stant aver­age ener­getic cost per neu­ron across species implies that larger neu­rons must com­pen­sate for the oblig­a­tory increased meta­bolic cost related to repo­lar­iz­ing the increased sur­face area of the cell mem­brane. This com­pen­sa­tion could be imple­mented by a decreased num­ber of synapses and/or decreased rates of exci­ta­tory synap­tic trans­mis­sion (69). Synap­tic home­osta­sis and elim­i­na­tion of excess synapses [e.g., dur­ing sleep (77)], the bases of synap­tic plas­tic­i­ty, might thus be nec­es­sary con­se­quences of a trade­off imposed by the need to con­strain neu­ronal ener­getic expen­di­ture (70). Another con­se­quence of a seem­ingly con­stant meta­bolic cost per neu­ron across species is that the total meta­bolic cost of rodent and pri­mate brains, and of the human brain, is a sim­ple, lin­ear func­tion of their total num­ber of neu­rons (70) (Fig. 6), regard­less of aver­age neu­ronal size, absolute brain size, or rel­a­tive brain size com­pared with the body. At an aver­age rate of 6 kcal/d per bil­lion neu­rons (70), the aver­age human brain, with 86 bil­lion neu­rons, costs about 516 kcal/d. That this rep­re­sents an enor­mous 25% of the total body ener­getic cost is sim­ply a result of the “eco­nom­i­cal” neu­ronal scal­ing rules that apply to pri­mates in com­par­i­son to rodents, and prob­a­bly to other mam­mals in gen­eral

    …Grow­ing a large body comes at a cost. Although large ani­mals require less energy per unit of body weight, they have con­sid­er­ably larger total meta­bolic require­ments that, on aver­age, scale with body mass raised to an expo­nent of ∼3/4 (84-87). Thus, large mam­mals need to eat more, and they can­not con­cen­trate on rare, hard-­to-find, or catch foods (88). Adding neu­rons to the brain, how­ev­er, also comes at a siz­able cost, as reviewed above: 6 kcal/d per bil­lion neu­rons (70). In pri­mates, whose brain mass scales lin­early with its num­ber of neu­rons, this implies that total brain metab­o­lism scales lin­early with brain vol­ume or mass, that is, with an expo­nent of 1, which is much greater than the much cited 3/4 expo­nent of Kleiber (84) that relates body metab­o­lism to body mass. The dis­crep­ancy sug­gests that, per gram, the cost of pri­mate brain tis­sue scales faster than the cost of non­neu­ronal bod­ily tis­sues, which calls for a mod­i­fi­ca­tion of the “expen­sive tis­sue hypoth­e­sis” of brain evo­lu­tion (89), accord­ing to which brain size is a lim­it­ing fac­tor. Given the steep, lin­ear increase in brain meta­bolic cost with increas­ing num­bers of neu­rons, we con­clude that meta­bolic cost is a more lim­it­ing fac­tor to brain expan­sion than pre­vi­ously sus­pect­ed. In our view, it is not brain size but, instead, absolute num­ber of neu­rons that imposes a meta­bolic con­straint on brain scal­ing in evo­lu­tion, because indi­vid­u­als with larger num­bers of neu­rons must be able to sus­tain their pro­por­tion­ately larger meta­bolic require­ments to keep their brain func­tion­al. The larger the num­ber of neu­rons, the higher is the total caloric cost of the brain, and there­fore the more time required to be spent feed­ing to sup­port the brain alone, and feed­ing can be very time-­con­sum­ing (90). Based on their brain mass [es­ti­mated from cra­nial capac­ity (91)], we pre­dicted that total num­bers of neu­rons in the brain increased from 27 to 35 bil­lion neu­rons in Aus­tralo­p­ithe­cus and Paran­thro­pus species to close to 50-60 bil­lion neu­rons in Homo species from Homo rudolfen­sis to Homo ante­ces­sor, to 62 bil­lion neu­rons in Homo erec­tus, and to 76-90 bil­lion neu­rons in Homo hei­del­ber­gen­sis and Homo nean­derthalen­sis (62), which is within the range of vari­a­tion found in mod­ern Homo sapi­ens (25). It can thus be seen how any increase in total num­bers of neu­rons in the evo­lu­tion of hominins and great apes would have taxed sur­vival in a lim­it­ing, if not pro­hib­i­tive, way, given that it prob­a­bly would have to occur in a con­text of already lim­it­ing feed­ing hours: The added 60 bil­lion brain neu­rons from an orang­utan-­sized hominin ances­tor to mod­ern Homo require an addi­tional 360 kcal/d, which is prob­a­bly not read­ily avail­able to great apes on their diet.

    It has been pro­posed that the advent of the abil­ity to con­trol fire to cook foods, which increases enor­mously the energy yield of foods and the speed with which they are con­sumed (92, 93), may have been a cru­cial step in allow­ing the near dou­bling of num­bers of brain neu­rons that is esti­mated to have occurred between H. erec­tus and H. sapi­ens (94).

  44. “How Hard Is Arti­fi­cial Intel­li­gence? Evo­lu­tion­ary Argu­ments and Selec­tion Effects”, Shul­man & Bostrom 2012:

    …con­ver­gent evo­lu­tion-the inde­pen­dent devel­op­ment of an inno­va­tion in mul­ti­ple tax­a-­can help us to under­stand the evolv­abil­ity of human intel­li­gence and its pre­cur­sors, and to eval­u­ate the evo­lu­tion­ary argu­ments for AI.

    The Last Com­mon Ances­tor (LCA) shared between humans and octo­pus­es, esti­mated to have lived at least 560 mil­lion years in the past, was a tiny worm­like crea­ture with an extremely prim­i­tive ner­vous sys­tem; it was also an ances­tor to nema­todes and earth­worms (Er­win and David­son 2002). Nonethe­less, octo­puses went on to evolve exten­sive cen­tral ner­vous sys­tems, with more ner­vous sys­tem mass (ad­justed for body size) than fish or rep­tiles, and a sophis­ti­cated behav­ioral reper­toire includ­ing mem­o­ry, visual com­mu­ni­ca­tion, and tool use. [See e.g. Mather (1994, 2008), Finn, Tre­gen­za, and Nor­man (2009) and Hochn­er, Shom­rat, and Fior­ito (2006) for a review of octo­pus intel­li­gence.] Impres­sively intel­li­gent ani­mals with more recent LCAs include, among oth­ers, corvids (crows and ravens, LCA about 300 mil­lion years ago),[­For exam­ple, a crow named Betty was able to bend a straight wire into a hook in order to retrieve a food bucket from a ver­ti­cal tube, with­out prior train­ing; crows in the wild make tools from sticks and leaves to aid their hunt­ing of insects, pass on pat­terns of tool use, and use social decep­tion to main­tain theft-re­sis­tant caches of food; see Emery and Clay­ton (2004). For LCA dat­ing, see Ben­ton and Ayala (2003).] ele­phants (LCA about 100 mil­lion years ago). [See Archibald (2003) for LCA dat­ing, and Byrne, Bates, and Moss (2009) for a review argu­ing that ele­phants’ tool use, num­ber sense, empa­thy, and abil­ity to pass the mir­ror test sug­gest that they are com­pa­ra­ble to non-hu­man great apes.] In other words, from the start­ing point of those worm­like com­mon ances­tors in the envi­ron­ment of Earth, the resources of evo­lu­tion inde­pen­dently pro­duced com­plex learn­ing, mem­o­ry, and tool use both within and with­out the line of human ances­try.

  45. See the chart on page 3 of “The pat­tern of evo­lu­tion in Pleis­tocene human brain size”; note also how high some of the recent skull vol­umes are - ~1800 cc - com­pared to mod­ern with an aver­age closer to 1500 cc (although appar­ently mod­ern extremes can still reach 1800-1900 cc).↩︎

  46. The Nean­derthals’ birth brain size was sim­i­lar to ours, and their adult brain size was notice­ably larger:

    Brain size reduc­tion in mod­ern humans over the past 40,000 years is well-­doc­u­ment­ed," the researchers said in their notes. "We hypoth­e­size that grow­ing smaller but sim­i­larly effi­cient brains might have rep­re­sented an ener­getic advan­tage, which paid off in faster repro­duc­tive rates in mod­ern [hu­mans] com­pared to Pleis­tocene peo­ple. Reduc­ing brain size thus might rep­re­sent an evo­lu­tion­ary advan­tage.

    Brain size, inci­den­tal­ly, cor­re­lates sur­pris­ingly well with intel­li­gence in both and espe­cial­ly.↩︎

  47. “Evo­lu­tion of the human brain: is big­ger bet­ter?”: “Since the Late Pleis­tocene (ap­prox­i­mately 30,000 years ago), human brain size decreased by approx­i­mately 10%” For pop­u­lar cov­er­age of expla­na­tions, see Dis­cover’s “If Mod­ern Humans Are So Smart, Why Are Our Brains Shrink­ing?”.↩︎

  48. The data is uncer­tain, but there seems to be a sub­stan­tial increase in ratio of old skele­tons found over time; Cas­pari & Lee 2004 (“Older age becomes com­mon late in human evo­lu­tion”), pg 2, find the ‘older to younger adults ratio’ for var­i­ous humanoid groups to be:

    1. : 0.12
    2. Early Homo: 0.25
    3. Nean­der­tals: 0.39
    4. Early : 2.08
    5. All: 0.28

    This could have many expla­na­tions (per­haps a slow accre­tion of technology/culture allowed older peo­ple to sur­vive with no con­nec­tion to human biol­ogy or evo­lu­tion), but in this con­text, I can’t help but won­der—­could old age be increas­ing because intel­li­gence is so expen­sive that old age is the only way for the genes to recoup their invest­ments? Or could it be that increases in human intel­li­gence used to pay off within a nor­mal lifes­pan because the humans learned faster, but now they have reached a limit on their intel­li­gence, how fast they learn, and so to be more effec­tive, the slow learn­ers have to learn for longer?↩︎

  49. See “Fire in the Earth Sys­tem”, Sci­ence↩︎

  50. See “Q & A: Brute Strength in Chimps” and “The Brain A Body Fit for a Freaky-Big Brain”, Carl Zim­mer:

    …Wray and his col­leagues com­pared SLC2A1 in humans and other ani­mals. They dis­cov­ered that our ances­tors acquired an unusu­ally high num­ber of muta­tions in the gene. The best expla­na­tion for that accu­mu­la­tion of muta­tions is that SLC2A1 expe­ri­enced nat­ural selec­tion in our own lin­eage, and the new muta­tions boosted our repro­duc­tive suc­cess. Intrigu­ing­ly, the Duke team dis­cov­ered that the muta­tions did­n’t alter the shape of the glu­cose trans­porters. Rather, they changed stretches of DNA that tog­gled the SLC2A1 gene on and off.

    Wray guessed that these muta­tions changed the total num­ber of glu­cose trans­porters built in the human brain. To test his the­o­ry, he looked at slices of human brain tis­sue. In order to make glu­cose trans­porters, the cells must first make copies of the SLC2A1 gene to serve as a tem­plate. Wray dis­cov­ered that in human brains there were 2.5 to 3 times as many copies of SLC2A1 as there were in chim­panzee brains, sug­gest­ing the pres­ence of more glu­cose trans­porters as well. Then he looked at glu­cose trans­porters that deliver the sugar to mus­cles. The gene for these mus­cle trans­porters, called SLC2A4, also under­went nat­ural selec­tion in humans, but in the oppo­site direc­tion. Our mus­cles con­tain fewer glu­cose trans­porters than in chimps’ mus­cles. Wray’s results sup­port the notion that our ances­tors evolved extra mol­e­c­u­lar pumps to fun­nel sugar into the brain, while starv­ing mus­cles by giv­ing them fewer trans­porters.

  51. I am not the only one to have noticed that the genetic dis­or­der the­ory of Ashke­nazi intel­li­gence seems like a beau­ti­ful exam­ple of trade-offs; Hills & Her­twig 2011:

    The Ashke­nazi Jew pop­u­la­tion pro­vides a less well-­known but more dra­matic exam­ple of between-­do­mains trade-offs (see Cochran, Hardy, & Harp­end­ing, 2006). Among the Ashke­nazi Jews, the aver­age IQ is approx­i­mately 0.7 to 1 stan­dard devi­a­tion above that of the gen­eral Euro­pean pop­u­la­tion. Recent evi­dence indi­cates that this rise in IQ was the con­se­quence of evo­lu­tion­ary selec­tion for greater intel­li­gence among Euro­pean Jews over approx­i­mately the last 2,000 years. How­ev­er, this greater capac­ity for learn­ing appears to have come with a spe­cific side effect: a rise in the preva­lence of dis­eases, such as , , , and . Cen­tral to our point, these dis­eases are cor­re­lated with the same neural causes that ren­dered pos­si­ble increased IQ, such as increased den­drite devel­op­ment.

  52. “Aging of the cere­bral cor­tex dif­fers between humans and chim­panzees”, PNAS 2011:

    …overt vol­u­met­ric decline of par­tic­u­lar brain struc­tures, such as the hip­pocam­pus and frontal lobe, has only been observed in human­s…In con­trast to humans, who showed a decrease in the vol­ume of all brain struc­tures over the lifes­pan [on fMRI], chim­panzees did not dis­play sig­nif­i­cant age-re­lated changes. Using an iter­a­tive age-range reduc­tion pro­ce­dure, we found that the sig­nif­i­cant aging effects in humans were because of the lever­age of indi­vid­u­als that were older than the max­i­mum longevity of chim­panzees. Thus, we con­clude that the increased mag­ni­tude of brain struc­ture shrink­age in human aging is evo­lu­tion­ar­ily novel and the result of an extended lifes­pan.

  53. , Lineweaver 2007 dis­cusses the rar­ity of intel­li­gence in the con­text of the :

    in “The Non­preva­lence of Humanoids” (1964) artic­u­lated the case that humans (or any given species) were a quirky prod­uct of ter­res­trial evo­lu­tion and there­fore we should not expect to find humanoids else­where. Thus stu­pid things do not, in gen­eral acquire human-­like intel­li­gence. The evi­dence we have tells us that once extinct, species do not re-e­volve. Evo­lu­tion is irre­versible. This is known as (Dollo 1893, Gould 1970). The re-evo­lu­tion of the same species is not some­thing that hap­pens only rarely. It never has hap­pened…

    “We are not requir­ing that they fol­low the par­tic­u­lar route that led to the evo­lu­tion of humans. There may be many dif­fer­ent evo­lu­tion­ary path­ways, each unlike­ly, but the sum of the num­ber of path­ways to intel­li­gence may nev­er­the­less be quite sub­stan­tial.” (Sagan 1995a)

    To which Mayr replied:

    “Sagan adopts the prin­ci­ple”it is bet­ter to be smart than to be stu­pid," but life on Earth refutes this claim. Among all the forms of life, nei­ther the prokary­otes nor pro­tists, fungi or plants has evolved smart­ness, as it should have if it were “bet­ter.” In the 28 plus phyla of ani­mals, intel­li­gence evolved in only one (chor­dates) and doubt­fully also in the cephalopods. And in the thou­sands of sub­di­vi­sions of the chor­dates, high intel­li­gence devel­oped in only one, the pri­mates, and even there only in one small sub­di­vi­sion. So much for the puta­tive inevitabil­ity of the devel­op­ment of high intel­li­gence because “it is bet­ter to be smart.”( Mayr 1995b)

    …These ances­tors and their lin­eages have con­tin­ued to exist and evolve and have not pro­duced intel­li­gence. All together that makes about 3 bil­lion years of prokary­otic evo­lu­tion that did not pro­duce high intel­li­gence and about 600 mil­lion years of pro­tist evo­lu­tion that did not pro­duce high intel­li­gence…What Drake, Sagan and Con­way-­Mor­ris have done is inter­pret cor­re­lated par­al­lel moves in evo­lu­tion as if they were uncon­strained by shared evo­lu­tion but highly con­strained by a uni­ver­sal selec­tion pres­sure towards intel­li­gence that could be extrap­o­lated to extrater­res­tri­als. I am argu­ing just the oppo­site – that the appar­ently inde­pen­dent evo­lu­tion toward higher E.Q. is largely con­strained by shared evo­lu­tion with no evi­dence for some uni­ver­sal selec­tion pres­sure towards intel­li­gence. If this view is cor­rect, we can­not extrap­o­late the trends toward higher E.Q. to the evo­lu­tion of extrater­res­tri­als. If the con­ver­gence of dol­phins and humans on high E.Q. has much to do with the 3.5 Gyr of shared his­tory (and I argue that it has every­thing to do with it) then we are not jus­ti­fied to extrap­o­late this con­ver­gence to other extrater­res­trial life forms that did not share this his­to­ry. Extrater­res­tri­als are related to us in the sense that they may be car­bon and water based - they may have poly­mer­ized the same monomers using amino acids to make pro­teins, nucleotides to make a genetic code, lipids to make fats and sug­ars to make poly­sac­cha­rides. How­ev­er, our “com­mon ances­tor” with extrater­res­tri­als was prob­a­bly pre-bi­otic and did not share a com­mon lim­ited set of genetic tog­gle switches that is respon­si­ble for the appar­ently inde­pen­dent con­ver­gences among ter­res­trial life forms.

    …If heads were a con­ver­gent fea­ture of evo­lu­tion one would expect inde­pen­dent lin­eages to evolve heads. Our short twig on the lower left labeled “Homo” has heads, but heads are found in no other branch. Our two clos­est rel­a­tives, plants and fungi, do not seem to have any ten­dency toward evolv­ing heads. The evo­lu­tion of heads (en­cephal­iza­tion) is there­fore not a con­ver­gent fea­ture of evo­lu­tion. Heads are mono­phyletic and were once the pos­ses­sions of only one quirky unique species that lived about six or seven hun­dred mil­lion years ago. Its ances­tors, no doubt pos­sessed some kind of pro­to-­head related to neural crests and pla­codes (Wada 2001, Man­zanares and Nieto 2003). Drake (2006) stated that “[in­tel­li­gence] is not a fluke that has occurred in some small sub­-set of ani­mal life.” How­ev­er, Fig. 4 shows that intel­li­gence, heads, even all ani­mal life or mul­ti­cel­lu­lar life, may well be a fluke that is a small sub­-set of ter­res­trial life. One poten­tial prob­lem with this con­clu­sion: It is pos­si­ble that exist­ing heads could have sup­pressed the emer­gence of sub­se­quent heads. Such sup­pres­sion would be dif­fi­cult to estab­lish….Life has been evolv­ing on this planet for ~4 bil­lion years. If the Planet of the Apes Hypoth­e­sis is cor­rect and there is an intel­li­gence niche that we have only recently occu­pied – Who occu­pied it 2 bil­lion years ago, or 1 bil­lion years ago or 500 mil­lion years ago? ? Algae? Jel­ly­fish?

    …To­day there are about a mil­lion species of pro­to­stomes and about 600,000 species of deuteros­tomes (of which we are one). We con­sider our­selves to be the smartest deuteros­tome. The most intel­li­gent pro­to­stome is prob­a­bly the octo­pus. After 600 mil­lion years of inde­pen­dent evo­lu­tion and despite their big brains, octopi do not seem to be on the verge of build­ing radio tele­scopes. The dol­phi­noidea evolved a large E.Q. between ~60 mil­lion years ago and ~20 mil­lion years ago (Marino et al 2004). Thus, dol­phins have had ~20 mil­lion years to build a radio tele­scope and have not done so. This strongly sug­gests that high E.Q. may be a nec­es­sary, but is not a suf­fi­cient con­di­tion for the con­struc­tion of radio tele­scopes. Thus, even if there were a uni­ver­sal trend toward high E.Q., the link between high E.Q. and the abil­ity to build a radio tele­scope is not clear. If you live under­wa­ter and have no hands, no mat­ter how high your E.Q., you may not be able to build, or be inter­ested in build­ing, a radio tele­scope.

    More on octo­puses and squids as pos­si­bly the only other exam­ple for intel­li­gence:

    That’s because other crea­tures that are believed intel­li­gent - such as dol­phins, chim­panzees, some birds, ele­phants - are rel­a­tively closely related to humans. They’re all on the ver­te­brate branch of the tree of life, so there’s a chance the intel­li­gence shares at least some char­ac­ter­is­tics. Octo­pus­es, how­ev­er, are inver­te­brates. Our last com­mon ances­tor reaches back to the dim depths of time, 500 mil­lion to 600 mil­lion years ago. That means octo­pus intel­li­gence likely evolved entirely sep­a­rately and could be very dif­fer­ent from that of ver­te­brates. “Octo­puses let us ask which fea­tures of our minds can we expect to be uni­ver­sal when­ever intel­li­gence arises in the uni­verse, and which are unique to us,” God­frey-­Smith said. "They really are an iso­lated out­post among inver­te­brates. … From the point of view of the phi­los­o­phy of the mind, they are a big deal.

    One of the major expla­na­tions for why pri­mates and humans evolved intel­li­gence is their close social rela­tions & pack struc­ture. Cephalopods are soli­tary, so why are they intel­li­gent? Cam­ou­flage seems like a pos­si­bil­i­ty… as does the endemic cephalo­pod can­ni­bal­ism; from “Can­ni­bal­ism in Cephalopods”:

    Can­ni­bal­ism is so com­mon in adult squids that it was assumed that they are unable to main­tain their daily con­sump­tion with­out a can­ni­bal­is­tic part in their diet, due to their high meta­bolic rates…­Cephalopods have the capac­ity to prey on both rel­a­tively small and large prey due to the skil­ful­ness of their arms and ten­ta­cles as well as the pos­si­bil­ity to shred their food with their beaks…Recog­ni­tion of famil­iar­ity in cephalopods is pos­si­ble, but not cer­tain…and the pos­si­ble lack of recog­ni­tion could pro­mote non het­ero-­can­ni­bal­ism in cephalopods.

  54. To exten­sively quote the June 2011 Sci­en­tific Amer­i­can cover story:

    “I think it is very likely that there is a law of dimin­ish­ing returns” to increas­ing intel­li­gence indef­i­nitely by adding new brain cell­s…­Size car­ries bur­dens with it, the most obvi­ous one being added energy con­sump­tion. In humans, the brain is already the hun­gri­est part of our body: at 2% of our body weight, this greedy lit­tle tape­worm of an organ wolfs down 20% of the calo­ries that we expend at rest. In new­borns, it’s an astound­ing 65%…

    For decades this divid­ing of the brain into more work cubi­cles was viewed as a hall­mark of intel­li­gence. But it may also reflect a more mun­dane truth…: spe­cial­iza­tion com­pen­sates for the con­nec­tiv­ity prob­lem that arises as brains get big­ger. As you go from a mouse brain to a cow brain with 100 times as many neu­rons, it is impos­si­ble for neu­rons to expand quickly enough to stay just as well con­nect­ed. Brains solve this prob­lem by seg­re­gat­ing like-­func­tioned neu­rons into highly inter­con­nected mod­ules, with far fewer long-dis­tance con­nec­tions between mod­ules. The spe­cial­iza­tion between right and left hemi­spheres solves a sim­i­lar prob­lem; it reduces the amount of infor­ma­tion that must flow between the hemi­spheres, which min­i­mizes the num­ber of long, inter­hemi­spheric axons that the brain needs to main­tain. “All of these seem­ingly com­plex things about big­ger brains are just the back­bends that the brain has to do to sat­isfy the con­nec­tiv­ity prob­lem” as it gets larg­er…“It does­n’t tell us that the brain is smarter.”…Neu­rons do get larger as brain size increas­es, but not quite quickly enough to stay equally well con­nect­ed. And axons do get thicker as brains expand, but not quickly enough to make up for the longer con­duc­tion delays…In fact, neu­ro­sci­en­tists have recently seen a sim­i­lar pat­tern in vari­a­tions within humans: peo­ple with the quick­est lines of com­mu­ni­ca­tion between their brain areas also seem to be the bright­est. One study…used func­tional mag­netic res­o­nance imag­ing to mea­sure how directly dif­fer­ent brain areas talk to one another - that is, whether they talk via a large or a small num­ber of inter­me­di­ary areas…Shorter paths between brain areas cor­re­lated with higher IQ…[Others] com­pared work­ing mem­ory (the abil­ity to hold sev­eral num­bers in one’s mem­ory at once) among 29 healthy peo­ple…Peo­ple with the most direct com­mu­ni­ca­tion and the fastest neural chat­ter had the best work­ing mem­o­ry.

    It is a momen­tous insight. We know that as brains get larg­er, they save space and energy by lim­it­ing the num­ber of direct con­nec­tions between regions. The large human brain has rel­a­tively few of these long-dis­tance con­nec­tions. But…these rare, non­stop con­nec­tions have a dis­pro­por­tion­ate influ­ence on smarts: brains that scrimp on resources by cut­ting just a few of them do notice­ably worse…There is another rea­son to doubt that a major evo­lu­tion­ary leap could lead to smarter brains. Biol­ogy may have had a wide range of options when neu­rons first evolved, but 600 mil­lion years later a pecu­liar thing has hap­pened. The brains of the hon­ey­bee, the octo­pus, the crow and intel­li­gent mam­mals, Roth points out, look noth­ing alike at first glance. But if you look at the cir­cuits that under­lie tasks such as vision, smell, nav­i­ga­tion and episodic mem­ory of event sequences, “very aston­ish­ingly they all have absolutely the same basic arrange­ment.” Such evo­lu­tion­ary con­ver­gence usu­ally sug­gests that a cer­tain anatom­i­cal or phys­i­o­log­i­cal solu­tion has reached matu­rity so that there may be lit­tle room left for improve­men­t…So have humans reached the phys­i­cal lim­its of how com­plex our brain can be, given the build­ing blocks that are avail­able to us? Laugh­lin doubts that there is any hard limit on brain func­tion the way there is one on the speed of light. “It’s more likely you just have a law of dimin­ish­ing returns,” he says. “It becomes less and less worth­while the more you invest in it.” Our brain can pack in only so many neu­rons; our neu­rons can estab­lish only so many con­nec­tions among them­selves; and those con­nec­tions can carry only so many elec­tri­cal impulses per sec­ond. More­over, if our body and brain got much big­ger, there would be costs in terms of energy con­sump­tion, dis­si­pa­tion of heat and the sheer time it takes for neural impulses to travel from one part of the brain to anoth­er.

    Coun­ter-ev­i­dence would be obser­va­tions that indi­cate evo­lu­tion try­ing to com­pen­sate for lim­its in one sys­tem by invest­ing even more into another sys­tem; for exam­ple, it has been observed that child­birth in humans is extremely risky and dan­ger­ous com­pared to other pri­mates because the infant head is so enor­mous com­pared to the birth canal. If intel­li­gence weren’t valu­able, one would expect the head size to remain con­stant or decrease, and one cer­tainly would not expect the over-­sized human brain to grow even faster after child birth; yet the human pre­frontal cor­tex grows much faster in infancy than the chim­panzee pre­frontal cor­tex does.↩︎

  55. eg. chim­panzees out­per­form humans on the sim­ple work­ing mem­ory task Mon­key Lad­der (but Sil­ber­berg & Kearns 2009 and Cook & Wil­son 2010 claim humans are equal or bet­ter with train­ing; see also “Super Smart Ani­mals”). Another fun sta­tis­tic is that besides obvi­ously being stronger, faster, and more dan­ger­ous than humans, chim­panzees have bet­ter immune sys­tems inas­much as they - over-re­act­ing being a major cause of com­mon issues like arthri­tis or asth­ma.↩︎

  56. , Stephen Budi­an­sky:

    Giv­ing a blind per­son a writ­ten IQ test is obvi­ously not a very mean mean­ing­ful eval­u­a­tion of his men­tal abil­i­ties. Yet that is exactly what many cross-species intel­li­gence tests have done. Mon­keys, for exam­ple, were found not only to learn visual dis­crim­i­na­tion tasks but to improve over a series of such tasks—they formed a learn­ing set, a gen­eral con­cept of the prob­lem that beto­kened a higher cog­ni­tive process than a sim­ple asso­ci­a­tion. Rats given the same tasks showed dif­fi­culty in mas­ter­ing the prob­lems and no abil­ity to form a learn­ing set. The obvi­ous con­clu­sion was that mon­keys are smarter than rats, a con­clu­sion that was com­fort­ably accept­ed, as it fit well with our pre­ex­ist­ing prej­u­dices about the dis­tri­b­u­tion of gen­eral intel­li­gence in nature. But when the rat exper­i­ments were repeat­ed, only this time the rats were given the task of dis­crim­i­nat­ing dif­fer­ent smells, they learned quickly and showed rapid improve­ment on sub­se­quent prob­lems, just as the mon­keys did.

    The prob­lem of moti­va­tion is another major con­found­ing vari­able. Some­times we may think we are test­ing an ani­mal’s brain when we are only test­ing its stom­ach. For exam­ple, in a series of stud­ies gold­fish never learned to improve their per­for­mance when chal­lenged with “rever­sal” tasks. These are exper­i­ments in which an ani­mal is trained to pick one of two alter­na­tive stim­uli (a black panel ver­sus a white pan­el, say) in order to obtain a food reward; the cor­rect answer is then switched and the sub­ject has to relearn which one to pick. Rats quickly learned to switch their response when the pre­vi­ously rewarded answer no longer worked. Fish did­n’t. This cer­tainly fit com­fort­ably with every­one’s sense that fish are dumber than rats. But when the exper­i­ment was repeated with a dif­fer­ent food reward (a paste squirted into the tank right where the fish made its cor­rect choice, as opposed to pel­lets dropped into the back of the tank), lo and behold the gold­fish sud­denly did start improv­ing on rever­sal tasks. Other seem­ingly fun­da­men­tal learn­ing dif­fer­ences between fish and rodents like­wise van­ished when the exper­i­ments were redesigned to take into account dif­fer­ences in moti­va­tion.

    Equal­iz­ing moti­va­tion is an almost insol­u­ble prob­lem for design­ers of exper­i­ments. Are three gold­fish pel­lets the equiv­a­lent of one banana or fif­teen bird seeds? How could we even know? We would some­how have to enter into the inter­nal being of dif­fer­ent ani­mals to know for sure, and if we could do that we would not need to be devis­ing round­about exper­i­ments to probe their men­tal processes in the first place. When we do con­trol for all of the con­found­ing vari­ables that we pos­si­bly can, the strik­ing thing about the “pure” cog­ni­tive dif­fer­ences that remain is how the sim­i­lar­i­ties in per­for­mance between dif­fer­ent ani­mals given sim­i­lar prob­lems vastly out­weigh the dif­fer­ences. To be sure, there seems to be lit­tle doubt that chim­panzees can learn new asso­ci­a­tions with a sin­gle rein­forced tri­al, and that that is gen­uinely faster than other mam­mals or pigeons do it. Mon­keys and apes also learn lists faster than pigeons do. Apes and mon­keys seem to have a faster and more accu­rate grasp of numeros­ity judg­ments than birds do. The abil­ity to manip­u­late spa­tial infor­ma­tion appears to be greater in apes than in mon­keys.

    But again and again exper­i­ments have shown that many abil­i­ties thought the sole province of “higher” pri­mates can be taught, with patience, to pigeons or other ani­mals. Sup­pos­edly supe­rior rhe­sus mon­keys did bet­ter than the less advanced cebus mon­keys in a visual learn­ing-set prob­lem using col­ored objects. Then it turned out that the cebus mon­keys did bet­ter than the rhe­sus mon­keys when gray objects were used. Rats were believed to have supe­rior abil­i­ties to pigeons in remem­ber­ing loca­tions in a radial maze. But after rel­a­tively small changes in the pro­ce­dure and the appa­ra­tus, pigeons did just as well.

    If such exper­i­ments had shown, say, that mon­keys can learn lists of forty-­five items but pigeons can only learn two, we would prob­a­bly be con­vinced that there are some absolute dif­fer­ences in men­tal machin­ery between the two species. But the absolute dif­fer­ences are far nar­row­er. Pigeons appear to dif­fer from baboons and peo­ple in the way they go about solv­ing prob­lems that involve match­ing up two images that have been rotated one from the oth­er, but they still get the right answers. They essen­tially do just as well as mon­keys in cat­e­go­riz­ing slides of birds or fish or other things. Euan Macphail’s review of the lit­er­a­ture led him to con­clude that when it comes to the things that can be hon­estly called gen­eral intel­li­gence, no con­vinc­ing dif­fer­ences, either qual­i­ta­tive or quan­ti­ta­tive, have yet been demon­strated between ver­te­brate species. While few cog­ni­tive researchers would go quite so far—and in deed we will encounter a num­ber of exam­ples of dif­fer­ences in men­tal abil­i­ties between species that are hard to explain as any­thing but a fun­da­men­tal dif­fer­ence in cog­ni­tive func­tion – it is strik­ing how small those dif­fer­ences are, far smaller than “com­mon sense” gen­er­ally has it. Macphail has sug­gested that the “no-d­if­fer­ence” stance should be taken as a “null hypoth­e­sis” in all stud­ies of com­par­a­tive intel­li­gence; that is, it is an alter­na­tive that always has to be con­sid­ered and ought to be assumed to be the case unless proven oth­er­wise.

    A recent exam­ple of teach­ing pigeons some­thing pre­vi­ously only rhe­sus mon­keys had been shown to learn is a 2011 paper demon­strat­ing that pigeons can learn the gen­eral con­cept of ‘ascend­ing’ or ‘larger’ groups - being taught to peck on groups of 3 rather than 2, or 4 rather than 3, and gen­er­al­iz­ing to peck­ing groups of 8 rather than 6.↩︎

  57. “The pur­suit of hap­pi­ness (with or with­out kids)”, BBC News Online Mag­a­zine 2003:

    Again, the fig­ures do not bear it out. While the birth rate in the UK is the low­est since records began in 1924, our level of con­tent­ment has remained fairly steady. Two of the fore­most thinkers on well-be­ing, Richard Layard and Andrew Oswald, agree that chil­dren have a sta­tis­ti­cally insignif­i­cant impact on our hap­pi­ness…In 2001, almost 90% of British peo­ple reported they were very or fairly sat­is­fied with life. Accord­ing to this new study, those with­out chil­dren are, by and large, every bit as con­tent as those with­…­For moth­ers in par­tic­u­lar, par­ent­hood brings a new sort of plea­sure, the result of spend­ing time with their chil­dren, see­ing them develop and pro­vid­ing a dif­fer­ent take on life. Yet this comes at a cost, both finan­cial and emo­tion­al, accord­ing to the report, which spoke to 1,500 adults, par­ents and non-­par­ents, between the ages of 20 and 40. “Ful­l-­time work­ing moth­ers are lower paid rel­a­tive to women with­out chil­dren,” says Kate Stan­ley, who car­ried out the sur­vey for the Insti­tute for Pub­lic Pol­icy Research. Most women also tend to take on the lion’s share of domes­tic and child-­care duties, accord­ing to the sur­vey. And since income and inde­pen­dence have a bear­ing on hap­pi­ness, what moth­er­hood giveth with one hand, it taketh away with the oth­er. The trade-off is less acute for men, but accord­ing to the sur­vey, they are less ecsta­tic about chil­dren any­way. While two-thirds of moth­ers say their chil­dren make them most hap­py, just over 40% of fathers agree…On the other side, those with­out chil­dren recog­nise they are freer to pur­sue their own inter­ests and enjoy­ment than their tied-up, fam­i­ly-­fo­cused friends.

    This is also true in the United States, accord­ing to Abma & Mar­tinez (2006); see also “Does Hav­ing Chil­dren Cre­ate Hap­pi­ness?” (an­swer: no).↩︎

  58. See “A His­tory of Vio­lence” and the bet­ter-ref­er­enced “A His­tory of Vio­lence”: Edge Mas­ter Class 2011, cul­mi­nat­ing in his 2011 book, The Bet­ter Angels of Our Nature.↩︎

  59. “Ale, man, ale’s the stuff to drink / For fel­lows whom it hurts to think.” –: LXII, “Ter­ence, this is stu­pid stuff”, by ↩︎

  60. The Lit­tle Book of Tal­ent (Coyle 2012), pg 80:

    The solu­tion is to ignore the bad habit and put your energy toward build­ing a new habit that will over­ride the old one. A good exam­ple of this tech­nique is found in the work of the Shy­ness Clin­ic, a pro­gram based in Los Altos, Cal­i­for­nia, that helps chron­i­cally shy peo­ple improve their social skills. The clin­ic’s ther­a­pists don’t delve into a clien­t’s per­sonal his­to­ry; they don’t try to “fix” any­thing. Instead, they focus on build­ing new skills through what they call a social fit­ness mod­el: a series of sim­ple, intense, grad­u­ally esca­lat­ing work­outs that develop new social mus­cles. One of the first work­outs for a Shy­ness Clinic client is to walk up to a stranger and ask for the time. Each day the work­out grows more stren­u­ous-­soon clients are ask­ing five strangers for the time, mak­ing phone calls to acquain­tances, or chat­ting with a stranger in an ele­va­tor. After a few months, some clients are “socially fit” enough to per­form the ulti­mate work­out: They walk into a crowded gro­cery store, lift a water­melon above their head, and pur­posely drop it on the floor, tri­umphantly endur­ing the stares of dozens of strangers. (The gro­cery store cleanup crew does­n’t enjoy this quite as much as the clients do.)

  61. Hand­book of Psy­chopa­thy, ed. Christo­pher Patrick 2005; “Psy­cho­pathic Per­son­al­i­ty: The Scope of the Prob­lem”, Lykken:

    For exam­ple, in her impor­tant study of men­tal ill­ness in prim­i­tive soci­eties, Mur­phy (1976) found that the Yupic-s­peak­ing Eski­mos in north­west Alaska have a name, kun­langeta, for the

    man who, for exam­ple, repeat­edly lies and cheats and steals things and does not go hunt­ing and, when the other men are out of the vil­lage, takes sex­ual advan­tage of many wom­en-­some­one who does not pay atten­tion to rep­ri­mands and who is always being brought to the elders for pun­ish­ment. One Eskimo among the 499 on their island was called kun­langeta. When asked what would have hap­pened to such a per­son tra­di­tion­al­ly, an Eskimo said that prob­a­bly some­body would have pushed him off the ice when nobody else was look­ing. (p. 1026)

    This is inter­est­ing since out of 500, the usual Amer­i­can base rates would pre­dict not 1 but >10 psy­chopaths. Is this all due to the tribal and closely knit nature of more abo­rig­i­nal soci­eties, or could Eskimo soci­ety really have been select­ing against psy­chopaths while big mod­ern soci­eties give scope for their tal­ents & ren­der them more evo­lu­tion­ar­ily fit? This may be unan­swer­able until the rel­e­vant genes are iden­ti­fied and sam­ples of gene pools exam­ined for the fre­quen­cies.

    “Psy­chopa­thy in spe­cific sub­pop­u­la­tions”, Sul­li­van & Kos­son (Hand­book):

    Ras­mussen and col­leagues (1999) hypoth­e­sized that in a nation such as Nor­way, where impris­on­ment is less fre­quent, more severe offend­ers who would be incar­cer­ated any­where are likely to com­prise a higher pro­por­tion of the inmate pop­u­la­tion. How­ev­er, this expla­na­tion does not likely apply to Scot­land: The incar­cer­a­tion rate for the United States is five to eight times that of Scot­land, but the base-rates of psy­chopa­thy in Scot­tish pris­ons are extremely low when com­pared to North Amer­i­can sam­ples (i.e., 3% in Scot­land vs. 28.4% in North Amer­i­ca, apply­ing the tra­di­tional cut­off of ≥30; Cooke, 1995; Cooke & Michie, 1999; Hare, 1991).

    “Treat­ment of Psy­chopa­thy: A Review of Empir­i­cal Find­ings”, Har­ris & Rice 2006 (Hand­book):

    We believe there is no evi­dence that any treat­ments yet applied to psy­chopaths have been shown to be effec­tive in reduc­ing vio­lence or crime. In fact, some treat­ments that are effec­tive for other offend­ers are actu­ally harm­ful for psy­chopaths in that they appear to pro­mote recidi­vism. We believe that the rea­son for these find­ings is that psy­chopaths are fun­da­men­tally dif­fer­ent from other offend­ers and that there is noth­ing “wrong” with them in the man­ner of a deficit or impair­ment that ther­apy can “fix.” Instead, they exhibit an evo­lu­tion­ar­ily viable life strat­egy that involves lying, cheat­ing, and manip­u­lat­ing oth­ers.

    The evo­lu­tion­ary hypoth­e­sis of psy­chopa­thy is strik­ing (eg. it’s par­tially her­i­ta­ble; or, sex offend­ers who tar­get post-pu­ber­tal women have the high­est PCL-R scores com­pared to any other sub­di­vi­sion of sex offend­er­s), but highly spec­u­la­tive. It’s dis­cussed a lit­tle skep­ti­cally in the chap­ter “The­o­ret­i­cal and Empir­i­cal Foun­da­tions” in the Hand­book.↩︎