W. H. Calvin, "The Emergence of Intelligence," Scientific American (1994)
Webbed Reprint Collection
William H. Calvin
University of Washington
Seattle WA 98195-1800 USA
Email || Home Page || publication list
William H. Calvin "The Emergence of
"The Emergence of Intelligence"
copyright ©1994 by William
H. Calvin and Scientific American
There is a 1998 revision of this article at
Copyright 1994 by the author. You may download this for personal use but may not redistribute or archive.
CAUTION: Some editorial improvements were made in this manuscript; see the printed version (or Scientific American On Line at aol.com) before quoting. There is one illustration (triangles within hexagons, bottom of p.106) that is seriously in error (it was a last minute invention that they didn't show to me). Just ignore it. If you really want to see the correct one, it's available here.
Language, foresight, musical skills and other hallmarks of intelligence are connected through an underlying facility that enhances rapid movements. Creativity may result from a Darwinian contest within the brain.To most observers, the essence of intelligence is cleverness, a versatility in solving novel problems. Bertrand Russell once wryly noted that "Animals studied by Americans rush about frantically, with an incredible display of hustle and pep, and at last achieve the desired result by chance. Animals observed by Germans sit still and think, and at last evolve the solution out of their inner consciousness." Besides being a commentary on the scientific fashions of 1927, Russell illustrates the false dichotomy usually made between random trial and error (which intuitively seems unrelated to intelligent behavior) and insight.
Foresight is also said to be an essential aspect of intelligence--particularly after an encounter with one of those terminally clever people who are all tactics and no strategy. Jean Piaget emphasized that intelligence was the sophisticated groping that we use when not knowing what to do. Personally, I like the way neurobiologist Horace Barlow of University of Cambridge frames the issue. He says intelligence is all about making a guess that discovers some new 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 a witty reply or guessing what's likely to happen next. Indeed, we all routinely predict what comes next, even when passively listening to a narrative or a melody. That's why a joke's punch line or a P.D.Q. Bach musical parody brings you up short--you were subconsciously predicting something else and were surprised by the mismatch.
We will never agree on a universal definition of intelligence because it is an open-ended word, like consciousness. Both intelligence and consciousness concern the high end of our mental life, but they are frequently confused with more elementary mental processes, such as ones we would use to recognize a friend or tie a shoelace. Of course, such simple neural mechanisms are probably the foundations from which our abilities to handle logic and metaphor evolved.
But how did that occur? That's both an evolutionary question and a neurophysiological one. Both kinds of answers are needed if we are to understand our own intelligence. They might even help us appreciate how an artificial or an exotic intelligence could evolve.
Did our intelligence arise from having more of what other animals have? The two-millimeter-thick cerebral cortex is the part of the brain most involved with making novel associations. Ours is extensively wrinkled but, were it flattened out, it would occupy four sheets of typing paper. A chimpanzee's cortex would fit on one sheet, a monkey's on a postcard, a rat's on a stamp.
Yet a purely quantitative explanation seems incomplete. I will argue that our intelligence arose primarily through the refinement of some brain specialization, such as that for language. The specialization would allow a quantum leap in cleverness and foresight during the evolution of humans from apes. If, as I suspect, that specialization involves a core facility common to language, the planning of hand movements, music and dance, it has even greater explanatory power.
Versatility is another characteristic of intelligence. Most animals are narrow specialists, especially in matters of diet: the mountain gorilla consumes 50 pounds of greenery each and every day. In comparison, a chimpanzee switches around a lot--it will eat fruit, termites, leaves, and even a small monkey or piglet if it is lucky enough to catch one. Omnivores have more basic moves in their general behavior because their ancestors had to switch between many different food sources. They need more sensory templates, too--mental search images of things such as foods and predators for which they are "on the lookout." Their behavior emerges through the matching of these sensory templates to responsive movements.
Sometimes animals try out a new combination of search image and movement during play, and find a use for it later. Many animals are only playful juveniles; being an adult is a serious business (they have all those young mouths to feed). Having a long juvenile period, as apes and humans do, surely aids intelligence. A long life further promotes versatility by affording more opportunities to discover new behaviors.
A social life also gives individuals the chance to mimic the useful discoveries of others. Researchers have seen a troop of monkeys in Japan copy one inventive female's techniques for washing sand off food. Moreover, a social life is full of interpersonal problems to solve, such as those created by pecking orders, that go well beyond the usual environmental challenges to survival and reproduction.
When the chimpanzees of Uganda arrive at a grove of fruit trees, they often discover that the efficient local monkeys are already speedily stripping the trees of edible fruit. The chimps can turn to termite fishing, or perhaps catch a monkey and eat it, but in practice their population is severely limited by that competition, despite a brain twice the size of their specialist rivals. Versatility is not always a virtue, and more of it is not always better. As frequent airline travelers know, passengers who only have carry-on bags can get all the available taxicabs while those burdened by three suitcases await their luggage. On the other hand, if the weather is so unpredictable that everyone has to travel with clothing ranging from swim suits to Arctic parkas, the "jack of all trades" has an advantage over the "master" of one. And so it is with behavioral versatility and brain size.
Whether versatility is advantageous depends on the time scales: for both the modern traveler and the evolving ape, it's how fast the weather changes and how long the trip lasts. Paleoclimatologists have discovered that many parts of the earth suffer sudden climate change, as abrupt in onset as a decade-long drought but lasting for centuries. A climate flip that eliminated fruit trees would be disastrous for many monkey species. It would hurt the more omnivorous animals, too, but they could make do with other foods, and eventually they would enjoy a population boom when the food crunch ended and few of their competitors remained.
Although Africa was cooling and drying as upright posture was becoming established 4 million years ago, brain size didn't change much. The fourfold expansion of the hominid brain did not start until the ice ages began, 2.5 million years ago. Ice cores from Greenland show frequent abrupt cooling episodes superimposed on the more stately rhythms of ice advance and retreat. Whole forests disappeared within several decades because of drastic drops in temperature and rainfall. The warm rains returned with equal suddenness several centuries later.
The evolution of anatomical adaptations in the hominids could not have kept pace with these abrupt climate changes, which would have occurred within the lifetime of single individuals. But these environmental fluctuations could have promoted the incremental accumulation of new mental abilities that conferred greater behavioral flexibility.
Language is the most defining feature of human intelligence: without syntax--the orderly arrangement of verbal ideas--we would be little more clever than a chimpanzee. For a glimpse of life without syntax, we can look to the case of Joseph, an 11-year-old deaf boy. Because he could not hear spoken language and had never been exposed to fluent sign language, Joseph did not have the opportunity to learn syntax during the critical years of early childhood.
As neurologist Oliver Sacks described him in Seeing Voices: "Joseph saw, distinguished, categorized, used; he had no problems with perceptual categorization or generalization, but he could not, it seemed, go much beyond this, hold abstract ideas in mind, reflect, play, plan. He seemed completely literal--unable to juggle images or hypotheses or possibilities, unable to enter an imaginative or figurative realm.... He seemed, like an animal, or an infant, to be stuck in the present, to be confined to literal and immediate perception, though made aware of this by a consciousness that no infant could have."
To understand why humans are so intelligent, we need to understand how our ancestors remodeled the ape symbolic repertoire and enhanced it by inventing syntax. Wild chimpanzees use about three dozen different vocalizations to convey about three dozen different meanings. They may repeat a sound to intensify its meaning, but they don't string together three sounds to add a new word to their vocabulary.
We humans also use about three dozen vocalizations, called phonemes. Yet only their combinations have content: we string together meaningless sounds to make meaningful words. No one has yet explained how our ancestors got over the hump of replacing "one sound/one meaning" with a sequential combinatorial system of meaningless phonemes, but it's probably one of the most important advances that happened during ape-to-human evolution.
Furthermore, human language uses strings of strings, such as the word phrases that make up this sentence. The simplest ways of generating word collections, such as pidgin dialects (or my tourist German), are known as protolanguage. In a protolanguage, the association of the words carries the message, with perhaps some assistance from customary word order (such as the subject-verb-object order in English declarative sentences).
Our closest animal cousins, the common chimpanzee and the bonobo (pygmy chimpanzee), can achieve surprising levels of language comprehension when motivated by skilled teachers. Kanzi, the most accomplished bonobo, can interpret sentences he has never heard before, such as "Go to the office and bring back the red ball," about as well as a 2.5-year-old child. Neither Kanzi nor the child constructs such sentences independently, but they can demonstrate by their actions that they understand them.
With a year's experience in comprehension, the child starts constructing sentences that nest one word phrase inside another. The rhyme about the house that Jack built ("This is the farmer sowing the corn/ That kept the cock that crowed in the morn/ ...That lay in the house that Jack built") is an extreme case, yet even preschoolers can understand how "that" keeps changing its meaning.
Syntax has treelike rules of reference that enable us to communicate quickly --sometimes with fewer than a hundred sounds strung together--who did what to whom, where, when, why and how. Generating and speaking a unique sentence quickly demonstrates whether you know the rules of syntax well enough to avoid ambiguities. Even children of low intelligence seem to acquire syntax effortlessly by listening, although intelligent deaf children like Joseph may miss out.
Short-term planning does occur to an extent, and it seems to allow an important increment in social intelligence. Deception is seen in apes, but seldom in monkeys. A chimp may give a call signaling that she has found food at one location, then quietly circle back through the dense forest to where she actually found the food. While the other chimps beat the bushes at the site of the food cry, she gets to eat without sharing.
The most difficult responses to plan are those to unique situations. They require imagining multiple scenarios, as when a hunter plots various approaches to a deer or a futurist spins three scenarios bracketing what an industry will look like in another decade. Compared to apes, humans do a lot of that: we are capable of heeding 18th-century admonition attributed to Edmund Burke, "The public interest requires doing today those things that men of intelligence and goodwill would wish, five or ten years hence, had been done."
Human planning abilities may stem from our talent for building syntactical, string-based conceptual structures larger than sentences. As the writer Kathryn Morton observes about narrative:
"The first sign that a baby is going to be a human being and not a noisy pet comes when he begins naming the world and demanding the stories that connect its parts. Once he knows the first of these he will instruct his teddy bear, enforce his world view on victims in the sandlot, tell himself stories of what he is doing as he plays and forecast stories of what he will do when he grows up. He will keep track of the actions of others and relate deviations to the person in charge. He will want a story at bedtime."Our abilities to plan ahead gradually develop from childhood narratives and are a major foundation for ethical choices, as we imagine a course of action, imagine its effects on others and decide whether or not to do it.
In this way, syntax raises intelligence to a new level. By borrowing the mental structures for syntax to judge other combinations of possible actions, we can extend our planning abilities and our intelligence. To some extent, we do this by talking silently to ourselves, making narratives out of what might happen next and then applying syntax-like rules of combination to rate a scenario as dangerous nonsense, mere nonsense, possible, likely or logical. But our thinking is not limited to languagelike constructs. Indeed, we may shout "Eureka!" when feeling a set of mental relationships click into place, yet have trouble expressing them verbally.
Language and intelligence are so powerful that we might think evolution would naturally favor their increase. As evolutionary theorists are fond of demonstrating, however, the fossil record is full of dead ends. Evolution often follows indirect routes rather than "progressing" via adaptations. To account for the breadth of our abilities, we need to look at improvements in common-core facilities. Though environments that give the musically gifted an evolutionary advantage over the tone deaf are difficult to imagine, there are multifunctional brain mechanisms whose improvement for one critical function might incidentally aid other functions.
We humans certainly have a passion for stringing things together: words into sentences, notes into melodies, steps into dances, narratives into games with rules of procedure. Might stringing things together be a core facility of the brain, one commonly useful to language, storytelling, planning ahead, games and ethics? If so, natural selection for any of these abilities might augment their shared neural machinery, so that an improved knack for syntactical sentences would automatically expand advance planning abilities, too. Such carryover is what Charles Darwin called functional change in anatomical continuity, distinguishing it from gradual adaptation. To some extent, music and dance are surely secondary uses of neural machinery shaped by sequential behaviors more exposed to natural selection, such as language.
As improbable as the idea initially seems, the brain's planning of ballistic movements may have once promoted language, music, and intelligence. Ballistic movements are extremely rapid actions of the limbs, that, once initiated, cannot be modified. Striking a nail with a hammer is an example. Apes have only elementary forms of the ballistic arm movements at which humans are expert--hammering, clubbing and throwing. Perhaps it is no coincidence that these movements are important to the manufacture and use of tools and hunting weapons: in some settings, hunting and toolmaking were probably important additions to hominids' basic survival strategies.
Compared to most movements, ballistic ones require a surprising amount of planning. Slow movements leave time for improvisation: when raising a cup to your lips, if the cup is lighter than you remembered, you can correct its trajectory before it hits your nose. Thus, a complete advance plan isn't needed. You start in the right general direction and then correct your path, just as a moon rocket does.
For sudden limb movements lasting less than one fifth of a second, feedback corrections are largely ineffective because reaction times are too long. The brain has to plan every detail of the movement in advance, as though it were silently punching a roll of music for a player piano.
Hammering, for example, requires planning the exact sequence of activation for dozens of muscles. The problem of throwing is further compounded by the launch window--the range of times in which a projectile can be released to hit a target. When the distance to a target doubles, the launch window becomes eight times narrower; statistical arguments indicate that programming a reliable throw would then require the activity of 64 times as many neurons.
If mouth movements rely on the same core facility for sequencing that ballistic hand movements do, then improvements in language might improve dexterity, and vice versa. Accurate throwing abilities open up the possibility of eating meat regularly, of being able to survive winter in a temperate zone. The gift of speech would be an incidental benefit --a free lunch, as it were, because of the linkage.
Doreen Kimura of the University of Western Ontario ("Sex Differences in the Brain," Scientific American, September 1992) has found that stroke patients with language problems (aphasia) resulting from damage to left lateral brain areas also have considerable difficulty executing novel sequences of hand and arm movements (apraxia). By electrically stimulating the brains of patients being operated on for epilepsy, George A. Ojemann of the University of Washington has also shown that at the center of the left lateral areas specialized for language lies a region involved in listening to sound sequences. This perisylvian region seems equally involved in producing oral-facial movement sequences--even nonlanguage ones.
These discoveries reveal that parts of the "language cortex," as people sometimes think of it, serves a far more generalized function than had been suspected. It is concerned with novel sequences of various kinds: both sensations and movements, for both the hands and the mouth.
By 1874, just 15 years after Darwin published The Origin of Species, the American psychologist William James was talking about mental processes operating in a Darwinian manner. In effect, he suggested, ideas might somehow "compete" with one another in the brain, leaving only the best or "fittest." Just as Darwinian evolution shaped a better brain in two million years, a similar Darwinian process operating within the brain might shape intelligent solutions to problems on the time scale of thought and action.
Researchers have demonstrated that a Darwinian process, operating on an intermediate time scale of days governs the immune response following a vaccination. Through a series of cellular generations spanning several weeks, the immune system produces defensive antibody molecules that are better and better "fits" against invaders. By abstracting the essential features of a Darwinian process from what is known about species evolution and immune responses, we can see that any "Darwin machine" must have six properties.
First, it must operate on patterns of some type; in genetics, they are strings of DNA bases, but the patterns of brain activity associated with a thought might qualify. Second, copies are somehow made of these patterns, just as in Richard Dawkins' memes. (Indeed, that which is reliably copied defines a unit pattern.) Third, patterns must occasionally vary, whether through mutations, copying errors, or a reshuffling of their parts.
Fourth, variant patterns must compete to occupy some limited space (as when bluegrass and crabgrass compete for my backyard). Fifth, the relative reproductive success of the variants is influenced by their environment; this result is what Charles Darwin called natural selection. And finally, the make-up of the next generation of patterns depends on which variants survive to be copied. The patterns of the next generation will be variations spread around the currently successful ones. Many of these new variants will be less successful than their parents, but some may be more so.
Sex and climate change may not be numbered among the six essentials but they add spice and speed to a darwinian process, whether it operates in milliseconds or millennia. Note that an "essential" isn't darwinian by itself, e.g., selective survival per se can be seen when flowing water carries away the sand and leaves the pebbles behind.
Evoking a memory is simply a matter of reconstituting such an activity pattern, according to the cell-assembly hypothesis of psychologist Donald O. Hebb (see "The Mind and Donald O. Hebb," by Peter M. Milner, Scientific American, January 1993). Long-term memories are frozen patterns waiting for signals of near resonance to reawaken them, like ruts in a washboarded road waiting for a passing car to recreate a bouncing spatiotemporal pattern.
Some "cerebral ruts" are permanent, while others are short-lived. Short-term memories are just temporary alterations in the strengths of synaptic connections between neurons, left behind by the last spatiotemporal pattern to occupy a patch of cortex; this "long-term potentiation" may fade in a matter of minutes. The transition from short- to long-term patterning is not well understood, but structural alterations may sometimes follow potentiation such that the synaptic connections between neurons are made strong and permanent, hardwiring the pattern of neural activity into the brain.
A Darwinian model of mind suggests that an activated memory can compete with others for "workspace" in the cortex. Both perceptions of the thinker's current environment and the memories of past environments may bias that competition and shape an emerging thought.
An active cerebral code moves from one part of the brain to another by making a copy of itself, much as a fax machine recreates a copy of a pattern on a distant sheet of paper. The cerebral cortex also has circuitry for copying spatiotemporal patterns in an immediately adjacent region less than a millimeter away, though our present imaging techniques lack enough resolution to see it in progress. Repeated copying of the minimal pattern could colonize a region, rather the way that a crystal grows or wallpaper repeats an elementary pattern.
The picture that emerges from these theoretical considerations is one of a quilt, some patches of which enlarge at the expense of their neighbors as one code copies more successfully than another. As you try to decide whether to pick an apple or a banana from the fruit bowl, so my theory goes, the cerebral code for "apple" may be having a cloning competition with the one for "banana." When one code has enough active copies to trip the action circuits, you might reach for the apple.
But the banana codes need not vanish: they could linger in the background as subconscious thoughts and undergo variations. When you try to remember someone's name, initially without success, the candidate codes might continue copying for the next half hour until, suddenly, Jane Smith's name seems to "pop into your mind" because your variations on the spatiotemporal theme finally hit a resonance and create a critical mass of identical copies. Our conscious thought may be only the currently dominant pattern in the copying competition, with many other variants competing for dominance, one of which will win a moment later when your thoughts seem to shift focus.
It may be that Darwinian processes are only the frosting on the cognitive cake, that much of our thinking is routine or rule-bound. But we often deal with novel situations in creative ways, as when you decide what to fix for dinner tonight. You survey what's already in the refrigerator nd on the kitchen shelves. You think about a few alternatives, keeping track of what else you might have to fetch from the grocery store. All of this can flash though your mind within seconds--and that's probably a Darwinian process at work.
It is difficult to estimate how often high intelligence might emerge, given how little we know about the demands of long-term species survival and the courses evolution can follow. We can, however, compare the prospects of a species by asking how many elements of intelligence each has amassed.
Does the species have a wide repertoire of movements, concepts or other tools? Does it have tolerance for creative confusion that allows individuals to create new categories occasionally? (Primatologist Duane Rumbaugh of Georgia State University has noted that small monkeys and prosimians, such as lemurs, often get trapped into repeating the first set of discrimination rules they are taught, unlike the more advanced rhesus monkeys and apes.)
Does each individual have more than a half dozen mental "workspaces" for concurrently holding different concepts? Does it have so many that it loses our human tendency to "chunk" certain concepts, as when we create the word "ambivalence" to stand for a whole sentence worth of description? Can individuals establish new relations between the concepts in their workspaces? These relations should be fancier than "is a" and "is larger than," which many animals can grasp. Treelike relations seem particularly important for linguistic structures; our ability to compare two relationships (analogy) enables operations in a metaphorical space.
Can individuals mold and refine their ideas off-line, before acting in the real world? Does that process involve all six of the essential Darwinian features, as well as some accelerating factors? Shortcuts that allow the process to start from something more than a primitive level? Can individuals make guesses about both long-term strategies and short-term tactics, so that they can make moves that will advantageously set the stage for future feats?
Chimps and bonobos may be missing a few of these elements but they're doing better than the present generation of artificial intelligence programs. Even in entities with all the elements, we would expect considerable variation in intelligence because of individual differences in implementing shortcuts, in finding the appropriate level of abstraction when using analogies, in processing speed and in perseverance.
William H. Calvin is a theoretical neurophysiologist at the University of Washington School of Medicine with a long association with neurosurgeons, zoologists, and psychiatrists. He studied physics at Northwestern University, made the transition to neuroscience at M.I.T. and Harvard Medical School, and received his Ph.D. in physiology and biophysics from the University of Washington in 1966. His literary efforts include The Throwing Madonna, The River that Flows Uphill, The Cerebral Symphony, The Ascent of Mind, How the Shaman Stole the Moon, and (with George Ojemann) Inside the Brain and Conversations with Neil's Brain.And, since the article, How Brains Think and The Cerebral Code.
The Ascent of Mind: Ice Age Climates and the Evolution of Intelligence, William H. Calvin. Bantam Books, 1990.
Tools, Language, and Cognition in Human Evolution, edited by Kathleen R. Gibson and Tim Ingold. Cambridge University Press, 1993.
Language Comprehension in Ape and Child, E. Sue Savage-Rumbaugh, Jeannine Murphy, Rose A. Sevcik, Karen E. Brakke, Shelley L. Williams, and Duane Rumbaugh. University of Chicago Press, 1993.
The Language Instinct, Steven Pinker. William Morrow and Company, 1994.
"Language" and Intelligence in Monkeys and Apes: Comparative Developmental Perspectives, edited by Sue T. Parker and Kathleen R. Gibson. Cambridge University Press, 1990.
Conversations with Neil's Brain: The Neural Nature of Thought and Language, William H. Calvin and George A. Ojemann. Addison-Wesley, 1994.
Forecasting and planning in an incoherent context, Donald N. Michael. Technological Forecasting and Social Change, vol. 36, pp. 79-87, 1989.
... I believe the brain plays a game -- some parts providing the stimuli, the others the reactions, and so on.... One is only consciously aware of something in the brain which acts as a summarizer or totalizer of the process going on and that probably consists of many parts acting simultaneously on each other. Clearly only the one-dimensional chain of syllogisms which constitutes thinking can be communicated verbally or written down.... If, on the other hand, I want to do something new or original, then it is no longer a question of syllogism chains. When I was a boy I felt that the role of rhyme in poetry was to compel one to find the unobvious because of the necessity of finding a word which rhymes. This forces novel associations and almost guarantees deviations from routine chains or trains of thought. It becomes paradoxically a sort of automatic mechanism of originality.... And what we call talent or perhaps genius itself depends to a large extent on the ability to use one's memory properly to find the analogies... [which] are essential to the development of new ideas.
Stan Ulam, Adventures of a Mathematician, 1976
[Current references to the throwing theory]