The Emergence of Intelligence

Language, foresight, musical skills and other hallmarks of intelligence may all be linked to the human ability to create rapid movements. by William H. Calvin

To most observers, the essence of intelligence is cleverness, a versatility in solving

novel problems. Foresight is also said to be an essential aspect of intelligence -particularly after an encounter ith one of those terminally clever people ho are all tactics and no strategy. !ther observers ill add creativity to the list. "ersonally, I li#e the ay neurobiologist Horace $arlo of the %niversity of Cambridge frames the issue. He says intelligence is all about ma#ing a guess that discovers some ne underlying order. This idea neatly covers a lot of ground& finding the solution to a problem or the logic of an argument, happening on an appropriate analogy, creating a pleasing harmony or guessing hat's li#ely to happen ne(t. Indeed, e all routinely predict hat comes ne(t, even hen passively listening to a narrative or a melody. That's hy a )o#e's punch line or a ". *. +. $ach musical parody brings you up short -- you ere subconsciously predicting something else and ere surprised by the mismatch. We ill never agree on a universal definition of intelligence because it is an openended ord, li#e consciousness. $oth intelligence and consciousness concern the high end of our mental life, but they are fre,uently confused ith more elementary mental processes, such as ones e use to recogni-e a friend or tie a shoelace. !f course, such simple neural mechanisms are probably the foundations from hich our abilities to handle logic and metaphor evolved. $ut ho did that occur. That's both an evolutionary ,uestion and a neurophysiological one. $oth #inds of ans ers are needed to understand our o n intelligence. They might even help e(plain ho an artificial or an e(otic intelligence could evolve. *id our intelligence arise from having more of hat other animals have. The t omillimeter-thic# cerebral corte( is the part of the brain most involved ith ma#ing novel associations. !urs is e(tensively rin#led but, ere it flattened out, it ould occupy four sheets of typing paper. / chimpan-ee's corte( ould fit on one sheet, a mon#ey's on a postcard, a rat's on a stamp. $ut a purely ,uantitative e(planation seems incomplete. I ill argue that our intelligence arose primarily through the refinement of some brain speciali-ation, such as that for language. This speciali-ation allo ed a ,uantum leap in cleverness and foresight during the evolution of humans from apes. If, as I suspect, the speciali-ation involved a core facility common to language, the planning of hand movements, music and dance, it has even greater e(planatory po er.

/ particularly intelligent person often seems 0,uic#0 and capable of )uggling many ideas at once. Indeed, the t o strongest influences on your I+ score are ho many novel ,uestions you can ans er in a fi(ed length of time, and ho good you are at simultaneously manipulating a half do-en mental images -- as in those analogy ,uestions& / is to $ as C is to 1*, E or F2. 3ersatility is another characteristic of intelligence. 4ost animals are narro specialists, especially in matters of diet& the mountain gorilla consumes 56 pounds of green leaves each and every day. In comparison, a chimpan-ee s itches around a lot -- it ill eat fruit, termites, leaves and even a small mon#ey or piglet if it is luc#y enough to catch one. !mnivores have more basic moves in their general behavior because their ancestors had to s itch bet een many different food sources. They need more sensory templates, too -- mental search images of things such as foods and predators for hich they are 0on the loo#out.0 Their behavior emerges through the matching of these sensory templates to responsive movements. 7ometimes animals try out a ne combination of search image and movement during play, and find a use for it later. 4any animals are only playful as )uveniles8 being an adult is a serious business 1they have all those young mouths to feed2. Having a long )uvenile period, as apes and humans do, surely aids intelligence. / long life further promotes versatility by affording more opportunities to discover ne behaviors. / social life also gives individuals the chance to mimic the useful discoveries of others. 9esearchers have seen a troop of mon#eys in :apan copy one inventive female's techni,ues for ashing sand off food. 4oreover, a social life is full of interpersonal problems to solve, such as those created by pec#ing orders, that go ell beyond the usual environmental challenges to survival and reproduction. ;et versatility is not al ays a virtue, and more of it is not al ays better. When the chimpan-ees of %ganda arrive at a grove of fruit trees, they often discover that the efficient local mon#eys are already speedily stripping the trees of edible fruit. The chimps can turn to termite fishing, or perhaps catch a mon#ey and eat it, but in practice their population is severely limited by that competition, despite a brain t ice the si-e of their specialist rivals.

The Impact of Abrupt Climate Change

3ersatility becomes advantageous, ho ever, hen the eather changes abruptly. The fourfold e(pansion of the hominid brain started <.5 million years ago, hen the ice ages began. Ice cores from =reenland sho that arming and cooling episodes occurred every several thousand years, superimposed on the slo er

advances and retreats of the northern ice sheets. The vast rearrangements in ocean currents lasted for centuries, ith sudden transitions that too# less than a decade.

Ice core data of *ansgaard et al Nature >??@. Younger Dryas sho n in red. Aote the t o episodes during the arm period >@6,666 years ago.

The abrupt coolings li#ely devastated the ecosystems on hich our ancestors depended. $ecause of lo er temperatures and less rainfall, the forests in /frica dried up and animal populations began to crash. Bightning stri#es ignited giant forest fires, denuding large areas even in the tropics. There as very little food after the fires. !nce the grasses got started on the burnt landscape, ho ever, the surviving gra-ing animals had a boom time. Within several centuries, a succession of forests came bac# in many places, featuring species more appropriate to the cooler climate. Cool, crash and burn. !ur ancestors lived through hundreds of such episodes -- but each as a population bottlenec# that eliminated most of their relatives. Had the cooling ta#en a fe centuries to happen, so that the forests could have gradually shifted, our ancestors ould not have been treated so badly. The higher-elevation plant species ould have slo ly marched do n the hillsides to occupy the valley floors. Each hominid generation could have made their living in the ay their parents taught them, culturally adapting to the ne milieu. $ut hen the cooling and drought as abrupt, it as one unluc#y generation that suddenly had to improvise amidst crashing populations and burning ecosystems.We are the improbable descendants of those ho survived -- probably because they had ays of coping ith these episodes that the other great apes did not e(ploit. Improvising meant learning to eat grass -- or managing to regularly eat animals that eat grass. The trouble is, such animals are fast and ary, hether rabbit or antelope. 7mall or big, they're best tac#led by cooperative groups. $ut sharing a rabbit leaves everyone hungry, so the hunters ould have tried for the bigger animals that cluster in herds. /nd that had an interesting conse,uence. If a single hunter #illed a big animal, it as too much to eat8 best to give most of the meat a ay and count on reciprocity hen someone else succeeded. 7haring food also

meant fe er fights and more time available to see# out scarce food. Each population bottlenec# temporarily e(aggerated the importance of such traits as cooperation, altruism and hunting abilities. Even if each episode changed the inborn predilections of the hominids by only a small amount, the hundreds of repetitions of this scenario may e(plain some of the differences bet een human abilities and those of our closest relatives among the great apes. It is tempting to say that the abrupt coolings pumped up brain si-e, but hat ma#es for better survival is something much more specific& hunting abilities and perhaps altruism. What might they have to do ith intelligence.

Syntax and Structured Thought

!ne of the improvements that occurred during the ice ages as the capacity for human language. In most of us, the brain area critical to language is located )ust above our left ear. 4on#eys lac# this left lateral language area& their vocali-ations 1and simple emotional utterances in humans2 employ a more primitive language area near the corpus callosum, the band of fibers connecting the cerebral hemispheres. Banguage is the most defining feature of human intelligence& ithout synta( -- the orderly arrangement of verbal ideas -- e ould be little more clever than a chimpan-ee. For a glimpse of life ithout synta(, e can loo# to the case of :oseph, an >>-year-old deaf boy. $ecause he could not hear spo#en language and had never been e(posed to fluent sign language, :oseph did not have the opportunity to learn synta( during the critical years of early childhood. /s neurologist !liver 7ac#s described him& 0:oseph sa , distinguished, categori-ed, used8 he had no problems ith perceptual categori-ation or generali-ation, but he could not, it seemed, go much beyond this, hold abstract ideas in mind, reflect, play, plan. He seemed completely literal -- unable to )uggle images or hypotheses or possibilities, unable to enter an imaginative or figurative realm....He seemed, li#e an animal, or an infant, to be stuc# in the present, to be confined to literal and immediate perception, though made a are of this by a consciousness that no infant could have.0

To understand hy humans are so intelligent, e need to understand ho our ancestors remodeled the apes' symbolic repertoire and enhanced it by inventing synta(. Wild chimpan-ees use about three do-en different vocali-ations to convey about three do-en different meanings. They may repeat a sound to intensify its meaning, but they don't string together three sounds to add a ne ord to their vocabulary. We humans also use about three do-en vocali-ations, called phonemes.

;et only their combinations have content& e string together meaningless sounds to ma#e meaningful ords. Furthermore, human language uses strings of strings, such as the ord phrases that ma#e up this sentence. !ur closest animal cousins, the common chimpan-ee and the bonobo 1pygmy chimpan-ee2, can achieve surprising levels of language comprehension hen motivated by s#illed teachers. Can-i, the most accomplished bonobo, can interpret sentences he has never heard before, such as 0=o to the office and bring bac# the red ball,0 about as ell as a <.5-year-old child. Aeither Can-i nor the child constructs such sentences independently, but they can demonstrate by their actions that they understand them. With a year's e(perience in comprehension, the child starts constructing sentences that nest one ord phrase inside another. The rhyme about the house that :ac# built 10This is the farmer so ing the cornDThat #ept the coc# that cro ed in the mornD...That lay in the house that :ac# built02 is an e(ample of such a sentence. 7ynta( has treeli#e rules of reference that enable us to communicate ,uic#ly -sometimes ith fe er than a hundred sounds strung together -- ho did hat to hom, here, hen, hy and ho . Even children of lo intelligence seem to ac,uire synta( effortlessly, although intelligent deaf children li#e :oseph may miss out. 7omething very close to synta( also seems to contribute to another outstanding feature of human intelligence, the ability to plan ahead. /side from hormonally triggered preparations for inter, animals e(hibit surprisingly little evidence of advance planning. For instance, some chimpan-ees use long t igs to pull termites from their nests. ;et as :acob $rono s#i observed, none of the termite-fishing chimps 0spends the evening going round and tearing off a nice tidy supply of a do-en probes for tomorro .0 Human planning abilities may stem from our talent for building narratives. We can borro the mental structures for synta( to )udge combinations of possible actions. To some e(tent, e do this by tal#ing silently to ourselves, ma#ing narratives out of hat might happen ne(t and then applying synta(-li#e rules of combination to rate a scenario as unli#ely, possible or li#ely. Aarratives are also a ma)or foundation for ethical choices& e imagine a course of action and its effects on others, then decide hether or not to do it. $ut our thin#ing is not limited to languageli#e constructs. Indeed, e may shout 0Eure#aE0 hen feeling a set of mental relationships clic# into place, yet have trouble e(pressing them verbally.

Ballistic Movements and their Relatives

Banguage and intelligence are so po erful that e might thin# evolution ould

naturally favor their increase. $ut as the Harvard evolutionary biologist Ernst 4ayr once said, most species are not intelligent, hich suggests 0that high intelligence is not at all favored by natural selection0 -- or that it's very hard to achieve. 7o e must consider indirect ays of achieving it, rather than general principles. Evolution often follo s indirect routes rather than 0progressing0 via adaptations. To account for the breadth of our higher intellectual functions 1synta(, planning, logic, games ith rules, music2, e need to loo# at improvements in common-core facilities. We humans certainly have a passion for stringing things together& ords into sentences, notes into melodies, steps into dances, narratives into games ith rules of procedure. 4ight stringing things together be a core facility of the brain.

/s improbable as the idea initially seems, the brain's planning of ballistic movements may have once promoted language, music and intelligence. $allistic movements are e(tremely rapid actions of the limbs, that, once initiated, cannot be modified. 7tri#ing a nail ith a hammer is an e(ample. /pes have only elementary forms of the ballistic arm movements at hich humans are e(pert -- hammering, clubbing and thro ing. "erhaps it is no coincidence that these movements are important to the manufacture and use of tools and hunting eapons& in some settings such as cool-crash-and-burn, hunting and toolma#ing ere important additions to hominids' basic survival strategies. Compared to most movements, ballistic ones re,uire a surprising amount of planning. 7lo movements leave time for improvisation& hen raising a cup to your lips, if the cup is lighter than you remembered, you can correct its tra)ectory before it hits your nose. Thus, a complete advance plan isn't needed. ;ou start in the right general direction and then correct your path. For sudden limb movements lasting less than one fifth of a second, feedbac# corrections are largely ineffective because reaction times are too long. The brain has to plan every detail of the movement. Hammering, for e(ample, re,uires planning the e(act se,uence of activation for do-ens of muscles.

The problem of thro ing is compounded by the briefness of the launch indo -the range of time in hich a pro)ectile can be released to hit a target. $ecause the human sense of timing is inevitably )ittery, hen the distance to a target doubles, the launch indo becomes eight times narro er. To shrin# the timing )itter enough re,uires a chorus of independent timing mechanisms, about FG times as many neurons 0singing0 the same 0plainchant0 in unison. If mouth movements rely on the same core facility for se,uencing that ballistic hand movements do, then improvements in de(terity might improve language, and vice versa. /ccurate thro ing abilities, re arded by surviving the cool-crash-andburn episodes in the tropics, also open up some options, such as the possibility of

eating meat regularly, or of being able to survive inter in a temperate -one. The gift of speech ould be an incidental benefit -- a free lunch, as it ere, because of the lin#age. There certainly seems to be a se,uencer common to both hands and language. 4uch of the brain's coordination of movement occurs at a subcortical level in the basal ganglia or the cerebellum, but novel movements tend to depend on the premotor and prefrontal corte(. T o ma)or lines of evidence point to cortical speciali-ation for se,uencing, and both of them suggest that the lateral language area has a lot to do ith it. *oreen Cimura of the %niversity of Western !ntario has found that stro#e patients ith language problems 1aphasia2 resulting from damage to left lateral brain areas also have considerable difficulty e(ecuting novel se,uences of hand and arm movements 1apra(ia2. $y electrically stimulating the brains of patients being operated on for epilepsy, =eorge /. !)emann of the %niversity of Washington has also sho n that at the center of the left lateral areas speciali-ed for language lies a region involved in listening to sound se,uences. This perisylvian region seems e,ually involved in producing oral-facial movement se,uences -- even nonlanguage ones.

These discoveries reveal that the 0language corte(,0 as people sometimes thin# of it, serves a far more generali-ed function than had been suspected. It is concerned ith novel se,uences of various #inds& both sensations and movements, for both the hands and the mouth. The big problem ith creating ne se,uences and producing original behaviors is safety. Even simple reversals in order can be dangerous, as in 0Boo# after you leap.0 !ur capacity to ma#e analogies and mental models gives us a measure of protection, ho ever. We humans can simulate future courses of action and eed out the nonsense off-line8 as philosopher Carl "opper said, this 0permits our hypotheses to die in our stead.0 Creativity -- indeed, the hole high end of intelligence and consciousness -- involves playing mental games that shape up ,uality before acting. What sort of mental machinery might it ta#e to do something li#e that.

Natural Selection in the Brain

$y >HIG, )ust >5 years after *ar in published The !rigin of 7pecies, /merican psychologist William :ames as tal#ing about mental processes operating in a *ar inian manner. In effect, he suggested, ideas might someho 0compete0 ith one another in the brain, leaving only the best or 0fittest.0 :ust as *ar inian evolution shaped a better brain in t o million years, a similar *ar inian process operating ithin the brain might shape intelligent solutions to problems on the time

scale of thought and action. 9esearchers have demonstrated that a *ar inian process operating on a time scale of days governs the immune system. Through a series of cellular generations spanning several ee#s, the immune system produces defensive antibody molecules that are better and better 0fits0 against invaders. $y abstracting the essential features of a *ar inian process from hat is #no n about species evolution and immune responses, e can see that any 0*ar in machine0 must have si( properties. First, it must operate on patterns of some type8 in genetics, they are strings of *A/ bases, but the patterns of brain activity associated ith a thought might ,ualify. 7econd, copies are someho made of these patterns. Third, patterns must occasionally vary, either through mutations, copying errors, or superimposed patterns. Fourth, variant patterns must compete to occupy some limited space 1as hen bluegrass and crabgrass compete for my bac#yard2. Fifth, the relative reproductive success of the variants is influenced by their environment8 this result is hat *ar in called natural selection. /nd finally, the ma#e-up of the ne(t generation of patterns depends on hich variants survive to be copied. The patterns of the ne(t generation ill be variations based on the more successful patterns of the current generation. 4any of the ne variants ill be less successful than their parents, but some may be more so. Bet us consider ho these principles might apply to the evolution of an intelligent guess inside the brain. Thoughts are combinations of sensations and memories -- in a ay, they are movements that haven't happened yet 1and maybe never ill2. They ta#e the form of cerebral codes, hich are spatiotemporal activity patterns in the brain that each represent an ob)ect, an action or an abstraction. I estimate that a single code minimally involves a fe hundred cortical neurons ithin a millimeter of one another, either #eeping ,uiet or firing in a musical pattern. Evo#ing a memory is simply a matter of reconstituting such an activity pattern, according to the cell-assembly hypothesis of psychologist *onald !. Hebb Jsee 0The 4ind and *onald !. Hebb,0 by "eter 4. 4ilner8 7cientific /merican, :anuary >??@K. Bong-term memories are fro-en patterns aiting for signals of near resonance to rea a#en them, li#e ruts in a ashboarded road aiting for a passing car to recreate a bouncing spatiotemporal pattern. 7ome 0cerebral ruts0 are permanent, hile others are short-lived. 7hort-term memories are )ust temporary alterations in the strengths of synaptic connections bet een neurons, left behind by the last spatiotemporal pattern to occupy a patch of corte(8 they fade in a matter of minutes. The transition from short-term to longterm memory is not ell understood, but it appears to involve structural alterations in hich the synaptic connections bet een neurons are made strong and permanent, hard iring the pattern of neural activity into the brain. / *ar inian model of mind suggests that an activated memory can compete ith

others for 0 or#space0 in the corte(. $oth the perceptions of the thin#er's current environment and the memories of past environments may bias that competition and shape an emerging thought. /n active cerebral code moves from one part of the brain to another by ma#ing a copy of itself, much as a fa( machine re-creates a pattern on a distant sheet of paper. The cerebral corte( also has circuitry for copying spatiotemporal patterns in an ad)acent region less than a millimeter a ay, though our present imaging techni,ues lac# enough resolution to see it in progress. 9epeated copying of the minimal pattern could coloni-e a region, rather the ay that a crystal gro s or allpaper repeats an elementary pattern. The picture that emerges from these theoretical considerations is one of a ,uilt, some patches of hich enlarge at the e(pense of their neighbors as one code copies more successfully than another. /s you try to decide hether to pic# an apple or a banana from the fruit bo l, so my theory goes, the cerebral code for 0apple0 may be having a cloning competition ith the one for 0banana.0 When one code has enough active copies to trip the action circuits, you might reach for the apple. $ut the banana codes need not vanish& they could linger in the bac#ground as subconscious thoughts. !ur conscious thought may be only the currently dominant pattern in the copying competition, ith many other variants competing for dominance, one of hich ill in a moment later hen your thoughts seem to shift focus. It may be that *ar inian processes are only the frosting on the cognitive ca#e, that much of our thin#ing is routine or rule-bound. $ut e often deal ith novel situations in creative ays, as hen you decide hat to fi( for dinner tonight. ;ou survey hat's already in the refrigerator and on the #itchen shelves. ;ou thin# about a fe alternatives, #eeping trac# of hat else you might have to fetch from the grocery store. /ll of this can flash though your mind ithin seconds -- and that's probably a *ar inian process at or#.

Bootstrapping Intelligence
In both its phylogeny and ontogeny, human intelligence first solves movement problems and only later graduates to ponder more abstract ones. /n artificial or e(traterrestrial intelligence freed of the necessity of finding food and avoiding predators might not need to move -- and so might lac# the hat-happens-ne(t orientation of human intelligence. It is difficult to estimate ho often high intelligence might emerge, given ho little e #no about the demands of longterm species survival and the courses evolution can follo . We can, ho ever, evaluate the prospects of a species by as#ing ho many elements of intelligence each has amassed. Chimps and bonobos may be missing a fe of the elements -the ability to construct nested sentences, for e(ample -- but they're doing better

than the present generation of artificial intelligence programs. Why aren't there more species ith such comple( mental states. There might be a hump to get over& a little intelligence can be a dangerous thing. / beyond-the-apes intelligence must constantly navigate bet een the t in ha-ards of dangerous innovation and a conservatism that ignores hat the 9ed +ueen e(plained to /lice in Through the Boo#ing =lass& 0...it ta#es all the running you can do, to #eep in the same place.0 Foresight is our special form of running, essential for the intelligent ste ardship that 7tephen :ay =ould of Harvard %niversity arns is needed for longer-term survival& 0We have become, by the po er of a glorious evolutionary accident called intelligence, the ste ards of life's continuity on earth. We did not as# for this role, but e cannot ab)ure it. We may not be suited to it, but here e are.0

About the author

WIBBI/4 H. C/B3IA's career has ta#en a *ar inian course& his scientific interests have evolved significantly over the past four decades. He studied physics as an undergraduate at Aorth estern %niversity, but devoted his spare time to a research pro)ect e(ploring ho the brain processes color vision. This pro)ect led to graduate or# in neuroscience at 4.I.T. and Harvard 4edical 7chool, then to a "h.*. in physiology and biophysics from the %niversity of Washington in >?FF. His early research focused on neuron firing mechanisms. 0I iretapped neurons, trying to figure out ho they transformed information,0 he says. $ut in the >?H6s he too# on a bigger ,uestion -- ho the human brain evolved -- and his interests broadened to include anthropology, -oology and psychology. He has ritten several acclaimed boo#s, including The Cerebral Code, Ho $rains Thin# and 1 ith =eorge !)emann2 Conversations ith Aeil's $rain. 0The pu--les I'm trying to solve re,uire information from many different fields,0 he says. He is currently a theoretical neurophysiologist on the faculty of the %niversity of Washington 7chool of 4edicine.

urther Reading
Derek Bickerton, Language and Species 1%niversity of Chicago "ress,

>??62.More....

amazon.com

Derek Bickerton, Language and Human Behavior 1%niversity of Washington "ress, >??52. More.... amazon.com William H. Calvin, Derek Bickerton, Lingua ex machina !econciling "ar#in and $homsky #ith the Human Brain 14IT "ress, forthcoming in >???2. ;ou can currently read a draft of the manuscript on the eb. William H. Calvin, %he $erebral $ode %hinking a %hought in the Mosaics of the Mind 14IT "ress, >??F2. More.... amazon.com William H. Calvin, Ho# Brains %hink &volving 'ntelligence, %hen and No#17cience 4asters, $asic$oo#s, >??F2. More.... amazon.com William H. Calvin and George A. Ojemann, $onversations #ith Neil(s Brain %he Neural Nature of %hought and Language 1/ddison-Wesley, >??G2. More.... amazon.com Kat leen !. Gi"son and #im $ngol% 1Editors2, %ools, Language and $ognition in Human &volution 1Cambridge %niversity "ress, >??@2. More.... amazon.com &. 'ue 'avage(!um"aug , 'tuart ' anker, #al"ot ). #aylor, )pes, Language, and the Human Mind 1!(ford %niversity "ress, 4ay >??H2. More.... amazon.com #errence Deacon, %he Symbolic Species %he $o*&volution of Language and the Brain 1W. W. Aorton, /ugust >??I2. More.... amazon.com Daniel C. Dennett, +inds of Minds %o#ard an ,nderstanding of $onsciousness17cience 4asters, $asic$oo#s, >??F2. More.... amazon.com 'teven *inker, %he Language 'nstinct 14orro , >??G2. More.... amazon.com Oliver 'acks, Seeing -oices 1%niversity of California "ress, >?H?2. More.... amazon.com 'ue 'avage(!um"aug and !oger +e,in, +an.i 1Wiley, >??G2. More.... amazon.com