Home Page || Publication List || Public Bookmarks || The Calvin Bookshelf (August 97)
William H. Calvin
This page is at http://WilliamCalvin.com/bk3/bk3day10.htm
The River That Flows Uphill (Sierra Club Books 1987) is my river diary of a two-week whitewater trip through the bottom of the Grand Canyon, discussing everything from the Big Bang to the Big Brain. It became a bestseller in German translation in 1995. AVAILABILITY limited; the US edition is now out of print. There are German and Dutch translations in print.
The River That Flows Uphill
A Journey from the Big Bang
to the Big Brain

Copyright 1986 by William H. Calvin.

You may download this for personal reading but may not redistribute or archive without permission (exception: teachers should feel free to print out a chapter and photocopy it for students).

This is a Deluxe edition in an unusual sense: the photographs and sound files are from Leonard Thurman’s Grand Canyon River Running web pages. What you get on your web browser is assembled, before your very eyes, using text delivered from Seattle (Washington State USA, near the Canadian border), and pictures and sound being sent from Tucson (Arizona USA, near the Mexican border).

DAY 10

Mile 137
Overhang Camp

NO HANGOVERS THIS MORNING as Overhang Camp awakens. As I stand at the shore brushing my teeth, I notice that the river seems exceptionally swift. It's probably because the Canyon is so narrow here. I smell bacon cooking on our homely hearth. It's what woke me up -- the overhang "cave" tends to fill up with the cooking smells.

I suppose the Anasazi cooked bighorn sheep instead. And I've heard that there are some Anasazi granaries just downriver. They're in the Tapeats, just above the underlying schist.

I've been trying to see up above the overhang, at the canyon above. We still haven't figured out the geology, but I suspect that an old fault comes through here -- there is a lot of broken rock above the Tapeats cliff, but no creek along its path. It's something like the Eminence Break back at Mile 43, where it looked as if a meat clever had fallen out of the sky and shattered the rocks. The broken rock here extends up through the Muav and a bit into the Redwall. We're now on the same side of the river as where Jeremy spotted those bighorns yesterday morning, and we're only three miles downstream. But I cannot spot them.

We're all in the shade, though the sun almost peeked in when it first rose. But Overhang Camp faces north, and probably stays shaded most of the day. It fills with the sounds of a sociable river camp. And we're a group comfortable with one another by now. There's always a lively round-table discussion going. Without a table.

"So Australopithecus and Homo habilis never made it out of Africa, never got to see an ice age?", asked Jackie as she sipped her second cup of tea.

Barbara nodded. "Of course, the ice ages probably changed the climate in the tropics too, simply because the ocean currents and weather patterns altered. So they'd have had to move around some in Africa, to follow the game or find enough rainfall to keep them supplied in plant food. But that doesn't mean it was a more severe climate."

"But Homo erectus saw the ice ages?", asked Ben.

"Surely. Homo erectus made it out of Africa not long after the time that the two toolkit styles developed, about 1.5 million years ago. Taking his handy spinning projectiles along. Even if you stick to the classically-described four ice ages that left rockpiles rearranged by the glaciers, you've got dates of 0.8, 0.46, 0.2, and 0.1 million years ago. That's when those ice ages started, anyway. Homo erectus surely saw the first two or three, at least -- plus a lot of the others that didn't leave such a nice record."

"And Homo erectus was far enough north to see them?", Jackie inquired.

"Well, Peking Man was late Homo erectus, living up in that northern latitude by 0.5, dying out by 0.2 million years ago. There are sharp, cold winters in Beijing these days -- much colder than in other cities like Naples or St. Louis that are the same latitude -- and I doubt Beijing was any warmer in an ice-age winter. So Homo erectus sure must have been wearing fur coats and living in cozy shelters by then," explained Barbara.

"Didn't Peking Man live in caves, though?", Dan Richard asked. "Isn't that where they found them?"

"That's right. That cave system got used repeatedly for a quarter of a million years. And they presumably had fire. It's not just that the climate would seem to require a source of heat. They've found charcoal in the caves. And the cave floor has been repeatedly burned," Barbara said. "So they were probably cooking food, and huddling around campfires for warmth at night."

"Is that when fire-making was invented?", asked Jackie. "No one is very sure about controlled fire. There are some reports from East Africa of fire-damaged stones arranged in what looks like a hearth, but it's still hard to eliminate grass fires and other accidental causes. That's several million years back. I expect that sometime between then and Peking Man at a quarter-million years ago, fire was invented. And lost. And then reinvented."

Ben brought over the coffee pot and refilled cups for most of us. "But you say that the tools didn't change much over that period?", he asked. "Well, the Mousterian style came in late in that period, but that's mostly an improvement in manufacturing, handy in places where the raw material is scarce. True, you can get some nice long blades that are hard to produce simply by touching up a handy random rock fragment. I'd say that the improvement was one of economy rather than an advance in what they could do with the tools they made. It isn't like the jump from wearing skins to sewing clothing that keeps the drafts out, or the jump from eating raw food to preparing cooked food, which greatly expands the possible diet by inactivating plant toxins. But I could be wrong," Barbara added.

"And brain size increased all though that Homo erectus period?", asked Jackie again.

"At first glance, that's what you'd think. They start at about 800 cubic centimeters in Africa 1.7 million years ago, and Homo erectus finishes up with Peking Man at 0.23 million years ago. And some of those Peking skulls are as big as 1140 cubic centimeters, though they average about 1088. But there's a lot of scatter. Some experts, like Philip Rightmire, argue that there isn't much steady improvement."

"So what replaced Homo erectus?", asked Ben, returning from putting the coffeepot back on the breakfast table. "Was it Homo sap?"

"Well, I wish that I knew," said Barbara, brushing her hair back with annoyance. "There are some skulls 0.4 million years old in Europe which we class as 'archaic Homo sapiens' because they are more like Homo sapiens than Homo erectus. But modern-type Homo sapiens doesn't appear until about 100,000 years ago in South Africa. So there's 300,000 years of transitional types, and they're really not very well defined yet. And 100,000 years ago is also about when the Neanderthal type appears. How Homo erectus changed into the archaic form, and then into the two Homo sapiens types is a big question."

"Neanderthal is Homo sapiens too?", asked Ben.

"Well, it sure wasn't an ancestor -- they were heavily-built contemporaries, with bigger brains than ours. So we just call them Homo sapiens neanderthalensis. And so we got a subspecies name, too. Homo sapiens sapiens."

"Man the wise, but doubly wise? I don't think so, not the way we're heading," commented Jackie.

"Now let me get this straight," said Ben after thinking a few moments, turning to me. "You don't think that the big brain era is just a continuation of the savannah phase in Africa, right?"

"Maybe the beginning -- Homo habilis and early Homo erectus -- was in the African savannah or hills or the Danakil tidelands. But no, I doubt that the Rift Valley was a good place to repeatedly run the ratchet for a bigger brain," I replied. "I think that the interesting events were probably happening elsewhere after, say, 1.4 million years ago. That's when Homo erectus started spreading all over. And I think we were hunting in a big way by then. Loren Eiseley used to say that meat supplied the energy that carried man around the world."

"But couldn't brain enlargement have been due to toolmaking?", queried Jackie.

"But," interjected Ben, "tools just don't change all that much during the period of enlargement. That was just my point."

I shrugged. "When you get down to it, we're just not an order-of-magnitude better than chimps are, when it comes to hammering skills. I think it's hunting and, while hunting is handy anywhere there is game, it is essential on the frontiers. In Africa, it's just too easy to fall back on gathering if the hunting skills fail. It's where groups are over-extended, living beyond their normal niche, that selection is most do-or-die."

"But why is a bigger brain so great for hunting? Wouldn't a big brain be better for outsmarting animals, or remembering your territory, things like that?", persisted Jackie.

"Carnivores," I explained, "are the real experts at outsmarting other species, and they haven't been experiencing a bigger brain boom. Nor have squirrels and packrats, who remember where they hide things, or orangutans who find their way around big territories. I think we're dealing with something fundamentally new in this third phase of post-chimp evolution -- that big brains involve something that other apes just don't do very much."

"Like language?", suggested Rosalie, looking up from threading film in her camera. "That's sure fundamentally new, and it's what our cultural boom is built on."

"That, and general intelligence itself, are surely most people's favorite candidates for why big brains became so popular," I agreed. "But I think they're too slow to do the job -- I prefer to put my money on something like hunting. And particularly on the kinds of hunting that the apes don't do -- such as throwing rocks. That's really a new invention, not just an improvement in hammering or social communication. And throwing also has this great growth curve -- bigger and bigger brains continue to be better and better performers."

"But why is a big brain so much better at throwing stones?", asked Jackie. "That's not exactly obvious to me. I can understand why a bigger brain could store more information, or why it might have to enlarge if it were to add on a new language center. But throwing is an arm motion -- it's not unlike hammering. And those chimps seem to already have that. Why does it take a bigger brain to throw? Isn't it just a little improvement in hammering-type motions?"

"And why all the emphasis on natural selection?", asked Ben. "I thought that variations were the flip side of the selection coin. Why aren't we talking about the anatomical variations that lead to bigger brains?"

I started to reply, but Gary called everyone over to the shore to discuss what we were doing today. So we picked ourselves up and walked down to the boats. Head boatman's announcement time. The answer wouldn't have been brief, anyway.

"We're going down today to a real pretty place called Matkatamiba. It's down at Mile 148, and we'll hang out there all afternoon," Gary began. "We'll probably stop for lunch just before Matkat at a neat waterfall in Olo Canyon. But it'll be a quiet morning. We'll go through a couple of fault lines, Sinyala and Fishtail. And Fishtail Rapid's a good ride. Then we'll get to Kanab Creek, probably stop there and let you hike up to see the meanders."

Gary held up a cautionary hand. "Just don't drink the water from that creek, or just downriver of it -- Kanab Creek comes all the way down from Utah. The town of Kanab's up there at the head of the canyon. We'll be filling our water jugs before getting to Kanab Creek. Why don't you fill up your canteens this morning, top them off before we break camp?"

"Keep your eyes open for bighorn today," concluded Gary. "Oh, yes. There's a big rapid this afternoon, after we leave Matkat. It's called Upset Rapid. A real good ride. But that's the last big one until we get to Lava Falls, day after tomorrow. We'll camp tonight down around Mile 155 or 156, so we can do Havasu tomorrow, bright and early. Anybody got any questions?"

"This Oreo Canyon where we're having lunch," asked Marsha. "Is that where the cookie monster lives?"

Groan. It's going to be one of those days. First we rename Overhang Camp as Hangover Camp, and now poor Olo Canyon is about to be renamed too.

Mile 143
Kanab Creek

THIS SIDE CANYON IS LONG, with the characteristic tree-like branching of a river that isn't in a hurry. Just up the creek from the confluence, it twists and turns with the typical meanders that one sees when flying over the Mississippi River. Such long rivers in flat country can change their meanders every century or so, but Kanab Creek cannot. Its meanders have become "entrenched" -- that is to say, they've dug such a deep canyon into harder rock that they can no longer actively meander about the way they did in the softer rock layers above; they are locked into an old wiggle from ages ago.

One wonders what old patterns have become entrenched in human evolution. Abundant toolmaking and brain enlargement may have started up together more than 2 million years ago but to judge from the merely minor progression of the Acheulean toolkit from 1.5- until 0.3 million years ago, it wasn't the demands made by toolmaking that caused our ancestors' brain size to keep increasing. Yet, if not toolmaking, then what? What used natural selection to make bigger brain variants survive better, reproduce better?

The usual answer to this question would surely be "general intelligence": smarter is better when trying to catch food that can run away from you, better in coping with changing environments. If bigger brains are also smarter, then bigger brains are better. It fits in perfectly with our preconceptions: that intelligence is what humans are all about.

The first problem with this appealing explanation is that one doesn't see other examples of it. Why didn't some other primate double and triple its brain size too? Even if no one made it as far as we have, surely there would be some robust examples in evolution of bigger-is-smarter-is-better, demonstrating its efficacy for us to see. It doesn't take other examples of 200 percent. Even a 50 percent increase would help make the point. But none obliged. The monkey-to-ape enlargement is about all we've got with which to compare.

The second problem is that within the present population of the presumptuously-named Homo sapiens sapiens, bigger isn't particularly smarter or wiser -- though some correlation between brain size and some aspect of "intelligence" isn't ruled out, the geniuses certainly come with all sizes of brain, persons of low intelligence don't have a smaller than average brain size, and generally all bets are off. Knowing someone's hat size won't do you a bit of good.

And to look at dolphins and gorillas and orangs and quite a few other animals, there are certainly species that are more clever than they need to be. These animals could perhaps get along fine with much less learning ability, manipulative ability, and mimicry ability than they have -- in their present niches. Maybe they were once faced with a series of environmental challenges that selected for those qualities, but they've settled into lives which no longer demand them for survival -- they found a reliable way of making a living and stuck with it, like the taxi drivers with Ph.D.s. One may not get to see the world that way, but by sticking to fruit and leaves, one doesn't have to worry if the game animals or the grain harvest will be sufficient to see one's family through the winter. Smarter may be better, but only if one's environment continues to make demands, only if speciation opportunities continue to save those biased genomes from subsequent dilution.

BIGGER DOESN'T EVEN MEAN MORE CELLS, much less a smarter brain. A bigger adult body usually means a bigger brain as well, but it may not have any more nerve cells than before! Typically, when a mammalian species gets bigger like the horses did, the brain enlarges along with the rest of the body. And the nerve cells simply spread out -- the same number, just more space between them. But sometimes there are also more cells as a result, and the reason may lie in one of those Darwinian aspects of development: many more nerve cells are produced early in development than can ever survive in subsequent stages. As gardeners know, the sprouts come up too close together in flats of young seedlings, so not all survive -- but if the gardener spreads them out, more will make it. Our threefold larger human brain doesn't have a proportional increase in the number of nerve cells it contains; the current estimates suggest that there are perhaps 25 percent more nerve cells in the human cerebral cortex than in a chimpanzee's cerebral cortex. The other 175 percent growth is just because they're aren't as close together.

How does one relate this to fossils, when all there is to go on is their cranial capacity and some estimates of their body weight, derived from the thicknesses of their bones and the size of their muscle-attachment zones? One can compensate somewhat for the differences in body size by reducing things to the brain/body ratio. There are formulas for brain/body weight ratios, but the safest thing is to compare primates of the same body weight, such as a 50 kilogram bear, a 50 kilogram chimp, and a 50 kilogram human. On that basis, we still have a brain that is 3.6 times as large as most apes would need to run a body our size, and 8.6 times that of an average mammal of our body size.

One way of judging brain growth over successive species is to fall back on the analogies between ontogeny and phylogeny. Brain size certainly grows as the fetus-infant-child enlarges, but at very different rates during the various stages. In the first trimester of gestation, a human fetus may be half head, making the brain/body ratio quite high. The head continues growing quite rapidly, but eventually it slows and the body grows faster, the brain/body ratio dropping steadily through childhood until reaching its adult value. To get a bigger adult brain in a new species means modifying those rates somewhere along the line, and that means varying the brain/body ratios. Comparative anatomy shows that there is a very simple way to increase relative brain size. Just grow young. It is called juvenilization, and it results in a bigger adult brain/body ratio. It enables evolution to work around entrenched meanders in matters developmental.

Whenever you happen to take your children to the Zoo you may observe in the eyes of the apes, when they are not performing gymnastic feats or cracking nuts, a strange strained sadness. One can almost imagine that they feel they ought to become men, but cannot discover the secret of how to do it.

Mile 145
Olo Canyon

LUNCH IS NEXT TO AN OVERHANGING WATERFALL which is quite a beautiful sight. And not a cookie monster in sight, just a rope dangling from atop the waterfall, left by someone trying to get up-canyon. It's rotted by now, and Gary cautions us against trying that route.

We were just saying that sometimes in evolution, one gets something for free, in apparent contradiction to the adage "There is no such thing as a free lunch." Most random change is, of course, bad -- and natural selection may eliminate it. Natural selection may seemingly operate on individual features of the human body, selecting for hairlessness separately from the diving reflex, so that our body is a mosaic of the ancient and the recent. This adaptationist view of the body tends to suggest that every feature has been shaped for a purpose, that nothing came gratis. Mosaic evolution.

But it's misleading. That's because natural selection isn't perfect; the physiologist Lloyd Partridge emphasizes that the result is more like "good enough engineering" in which a sufficient solution to an environmental problem removes the feature from exposure to the natural selection that might lead to further improvements in it. A neurophysiologist can imagine ways to improve our sense of taste; important as it is in the evolution of omnivores, our present system hasn't been streamlined. Natural selection also has yet to eliminate nearsightedness, the inflammation-prone appendix, muscle cramps, flat feet, headaches, premenstrual tension, and other such evidence of imperfection in human evolution. And natural selection may exercise little influence on some change, neither rewarding nor penalizing it. Mosaic evolution can happen, but selection on a one-feature-at-a-time basis has been greatly oversold as the architect of change.

Thus, there are "free" ways to get change: because many traits are anatomically linked, functional selection for one will carry other anatomical features along for free. Hitch-hikers, no less. One family of linked traits, of crucial importance for human evolution, is commonly known as juvenilization.

You may remember a fad in architecture in which new buildings looked unfinished, even though the occupants had moved in and the building was open for business. The structural beams were still exposed, the heating ducts and plumbing visible as if the builders hadn't yet gotten around to installing the usual false ceiling. In short, the building looked unfinished by the usual standards of its ancestors, the buildings of earlier decades. I remember wondering how far they'd carry this trend, whether they'd leave off the windows as the next phase.

I don't know if the architectural critics knew enough biology to see the analogy, but a good name for the unfinished-look fad would have been "architectural juvenilization." Juvenilization (also known as neoteny and paedomorphosis, by those who can figure out how to pronounce them) is a biological fad, one important way in which new animal species evolve from old ones -- and there's nothing trivial about this fad: It's the fad that helped produce the apes from the monkeys, and the humans from the apes. And probably the vertebrates from the invertebrates, back much earlier in the Cambrian, during Tapeats time. That's not a bad track record.

Juvenilization doesn't really refer to juveniles -- it's a fad affecting adults. Jackie explained it by analogy to fashions in women's clothing. It's as if this year's style for "Women's" dresses were to look suspiciously like last year's fad in dresses for "Junior Misses." Our society seems to be full of people trying to look younger than they are. Well, we're talking about the analogous biological trend that produces adults who look and act younger than an adult of the previous model year. There are several mechanisms for inducing this fad, though we try not to confuse them with the fad itself.

In its simplest manifestations, juvenilization appears as a way of escaping overspecialization, of backing up in evolution. Sometimes the latest model of the body finds itself unsuited to the environment, as when an amphibian discovers that a prolonged heat wave has dried out its swamp. An adult cannot return to the safety of the water by regressing on the spot to its fetal form, retrieving its gills. But a youngster coming along can put the brakes on its development, so that it keeps its gills. This is best seen in the Mexican salamander, much valued in the restaurants of Mexico City, as it was by Julian Huxley for his developmental studies. This newt-like amphibian, Ambystoma, usually goes through a larval stage corresponding to the tadpole stage of the frog, then loses its gills and emerges from the water as an air-breathing, land-dwelling animal getting around on all fours. But when the weather has been bad for salamanders, this metamorphosis from tadpole to salamander doesn't happen. The immature salamander is better off staying in the water; by putting the brakes on its development, it retains its gills and swims happily ever after.

Now the usual trouble with remaining a child forever is that you don't reproduce your kind. Happily, however, the tadpole form of Ambystoma becomes sexually mature; in fact, this early sexual maturity is part of the reason that the tadpole's body development comes to a halt before it makes the water-to-land modifications in its body. A minor form of this phenomenon can be seen in humans: the main reason that men are taller than women is that girls mature earlier than boys, slowing growth. And earlier-than-average sexual maturity in girls is likely to lead to their being shorter adults than girls whose menstruation begins at age 16. Accelerated puberty is, however, just one mechanism for inducing the juvenilization fad.

A delay in puberty, but with even more delay in the general body-development timetable, is another mechanism used. And it seems to have been used repeatedly by the primates: the whole primate lineage has been postponing the later phases of its development in favor of retaining the more juvenile forms.

Reflecting this is a tendency for the primate lineage to become more and more juvenile in appearance -- not only do the faces of successive species often become flatter but the teeth become smaller and the brain becomes larger (relative to body size). That is, chimp adults are more like juvenile monkeys; adult humans are more like juvenile chimps. This long-term trend toward a more juvenile adult was originally termed neoteny in 1884 by its discoverer, Julius Kollmann, a zoologist at Basel. Havelock Ellis in 1894 applied the idea to human evolution. Louis Bolk, an anatomist in Amsterdam, called it "fetalization" in comparing primate fetal stages with human developmental stages back in 1926. Stephen Jay Gould calls it paedomorphosis ("child-shaped"). Julian Huxley in 1952 liked "juvenilization," and that's my favorite too, since my tongue trips on the others. But take your pick.

Ben asked if names lasted only a quarter of a century, or whether each new generation of biologists had to rediscover juvenilization for itself?

Since the perimeter of the [scallop's] shell grows at a faster rate than the center, the perimeter curls and wrinkles. No genes carry an image of how to place the wrinkles; no genes remember the shape of the shell; they only permit or encourage faster growth at the perimeter than at the center.
...... PETER S. STEVENS, Patterns in Nature, 1974.
THE LINKAGE BETWEEN ALL THOSE JUVENILE TRAITS comes about because they share a common mechanism: altered clocks. Rather than simply speeding up the sexual maturity clock, as in Ambystoma, the road toHomo sapiens is marked by a simple slowing of two rates of maturation. Sexual maturity comes later and later: a monkey may become sexually mature in 3 to 4 years, a chimp in 8 to 9 years, and humans much later. The puberty alarm clock is slowed down; its mechanism is thought to be related to the secretion of melatonin by the pineal gland and some other midline areas of the brain.

The second slowing is in somatic growth. A juvenile chimp and a juvenile monkey may become playmates one year, but the next year the monkey will have grown much more than the chimp -- and the "retarded" chimp usually has to find a younger playmate. The regulation of the somatic growth rate is traditionally attributed to the growth hormone secreted by the pituitary gland (though the issue is known to be more complicated than that). A hypopituitary dwarf can, as a child, be treated with a two year course of human growth hormone that will accelerate its body growth (originally, the hormone was gathered from 100-200 human pituitaries at autopsy, but it can now be produced by bacteria fooled with recombinant DNA technology).

The slowing is seen in all stages of life, with longer lifespans for the more juvenilized species (one reason why ageing isn't just a matter of wearing out). If both sexual and somatic clocks were slowed down equally, the eventual adult body form would probably remain the same -- it would just take much longer to reach it. But at each stage of primate juvenilization, body growth is slowed down more than the puberty alarm clock -- this achieves much the same end as speeding up sexual maturity a la Ambystoma but additionally provides the extended childhood for cultural learning. Thus, puberty arrives when the body form is still juvenile by the standards of the ancestral adults. This is inferred from the numerous "neotenous" traits, things which adults now have but which are merely retained from their juvenile selves, things which were no longer altered in adulthood.

Take the little matter of flat faces. Monkey, chimp, and human infants all share a flat face, with the nose and lips lined up vertically beneath the eyes -- and the brain coming far enough forward to sit atop the eyes. But as these three primates grow up, the lower face starts to grow forward until -- in the monkey and chimp -- the adult brain sits largely to the rear of the eyes, the eyes to the rear of the nostrils, with the lips protruding even further forward. The same thing happens to many animals, such as our pet cats and dogs. We find the young ones particularly appealing because, like human infants, they have flat faces with large eyes (relative to their face size). Child face shape is a trigger feature that attracts adults, a matter important if the child is to receive adult care and protection. Much like the cuckoo's scarlet throat attracts the vireo step-parents, the attractiveness of a baby face may also serve other ends besides the one for which evolution first imperfectly designed it, such as identifying the appropriate target for parental affection and care-giving.

Then there's playfulness. Behavior is one of the big differences between juveniles and adults. The young are more inclined to play, amusing themselves by exercising their newfound abilities. And they are far more flexible than adults: the classic story of the Japanese monkeys tells of how novel foods were gradually adopted by a troop of macaques on a small island. The young would learn to unwrap caramel candies and eat them. They would learn how to wash sand off of potatoes. They would throw grains intermixed with the sand on the beach into the shallow water nearby, scooping up the floating grain after the sand sank. Not only didn't the adults ever invent these adaptations to new food, but they were slow to adopt them even after observing the success of the juveniles. Maybe chimps and prehumans adapted themselves to new niches by learning to eat a wider variety of foods than their fruit-loving ancestors. It is easy to imagine that juvenile curiosity and experimentation, retained into the adult stage of life, would have made possible the exploitation of many new foods.

And then there are the teeth. Our ancestors had big molars for grinding down food. Indeed, for an animal that tries to eat everything, it is hard to cite advantages of small teeth. Yet the grinding surfaces of the cheek teeth shrink during prehuman evolution, decreasing by half in the time that the brain triples. Were the smaller teeth somehow better, to use a traditional adaptationist approach, or was the tooth-size reduction another case of a hitchhiker: retaining the half-size juvenile molars into adulthood because some other aspect of juvenilization was rewarded by natural selection?

A bigger brain is also part of juvenilization. But isn't the adult's brain larger than the juvenile's? True, but it is the ratio of brain size and body size, rather than the absolute size itself, that seems all-important. Early on, the embryo is half head. By birth, the head is only a quarter of the crown-to-rump distance. In adults, maybe 10 percent. So one simple way to get a bigger brain/body ratio in an adult is to retain the relatively larger juvenile head size.

Human adults exhibit many other juvenile traits: for example, the big toes face forward in juvenile monkeys and chimps, but come to point outward (are "rotated") in adults. The human forward-pointing big toe might again be most simply explained as "juvenilization strikes again," though adaptationists strive mightily to link upright walking to modifying that big toe.

Some primate offspring have more juvenilization, others less. Selection operates on this genetic variation, biasing the gene pool one way or the other. The adult of the pygmy chimpanzee, Pan paniscus, looks like a juvenile of the common chimp, Pan troglodytes (you only get one guess: why is the pygmy chimp shorter?). Pygmy chimps look even "more human" than the adult common chimp, thanks to juvenilization. They split off from the common chimps about 3 million years ago. They're rare, and given how important they could be for understanding prehuman behaviors, we really ought to be building breeding colonies to assure their survival.

Dog breeders have artificially selected for juvenilization in the various steps starting with a jackal-like dog and ending up with a pug, simply by breeding together the flatter-faced offspring. Most of our domestic animals are juvenilized versions of their wild ancestors: pigs, cows, sheep, dogs and, to a lesser extent, cats. It seems likely that the artificial selection that humans exercised on these animals operated on some aspect of the juvenilization collection of traits, such as behavioral plasticity. But since young animals also solicit care, the ones that continued this trait into adulthood were also more likely to be fed and sheltered by humans. Perhaps humans too were "domesticated" by a selection for juvenilization.

Juvenilization was important at another juncture in our evolution: in the passage between the invertebrates to the vertebrates. The embryonic form of the sea-squirt looks surprisingly like a primitive chordate, and it is thought that a truncation of adult development in the sea-squirt led to the first of the chordates. There's a French phrase which summarizes all this: reculer pour mieux sauter ("step back to leap better").

The problem which remains is in fact not "how have vertebrates been formed by sea-squirts", but how have vertebrates eliminated the [adult] sea-squirt stage from their life history? It is wholly reasonable to consider that this has been accomplished by paedomorphism [juvenilization].
......the pioneer neurobiologist J. Z. YOUNG, 1950

ADAPTATION OR "FOR FREE"? Because the traits that exhibit juvenilization all share one common mechanism -- the puberty alarm clock is slowed down less than the body-growth clock -- natural selection affecting any one of the these traits can haul the others along for free. For example, if our ancestors became more successful via expanding the acceptable items in their diet, the selection for behavioral plasticity would have favored those individuals who were more juvenilized over those who were less. But a bigger brain/body ratio, a flatter face, smaller teeth, and a nonrotated big toe might have been incidental changes. In short, not all of these traits need have been directly shaped by selection -- it is sufficient for only one to have been a big success to have indirectly carried along the others. However, while one gene controlling many seemingly unrelated features is a well-recognized fact in genetics (it's called pleiotropy), neoteny-paedomorphism-juvenilization is not the traditional teaching in human biology, though first recognized over a century ago.

The variations tend to be families of anatomical features; the natural selection tends to act on the function of one of the family. Anatomy varies but physiology succeeds.

Do we owe our big brains, then, to the success of the juvenile's playfulness, or perhaps only to the tendency of individuals to select mates whose flat-faced, big-eyed appearance mimics the child's? Sexual selection could, after all, do the job rather than natural selection. I once asked students on a final exam a comic relief question: construct a hypothesis that could connect bigger brains to the common practice of changing the appearance of a woman's face via eye-shadow and -liner. What happens with eye makeup, of course, is that it is used to enhance the apparent size of the eyes, making the face look more juvenile. I wanted the students to recognize that such makeup was used to mimic juvenile appearance, and that adults were "tuned up" to like children's faces and to protect and provide for them. And that relatively larger eye appearance in an adult woman might stimulate a man to provide her with the care he might otherwise reserve for a child. But since women are even more tuned up to respond to young faces, this ought to also work the other way: as one of the best students pointed out to my surprise, this means that eye shadow for men might attract women. And that maybe I should shave off my beard if....

Ah, but back to juvenilization. If all these traits are linked, thanks to their sharing the sluggish clock genes, cause and effect become harder to distinguish. The selection for juvenile facial appearance in mates might occur at the same time as selection for juvenile experimentation with novel foods, at the same time as direct selection for neuron numbers via general intelligence. There is no single "cause" for any evolutionary happening, but for linked traits like those of the juvenilization family, cause is particularly fuzzy, a child of many parents.

While juvenilization has been recognized for a hundred years under various names, it has not been widely adopted as an explanation for human phylogeny; you'll seldom hear an anthropologist mention it when spinning a scenario for hominid evolution. And Dan Hartline voiced what has probably been a widespread reaction to the juvenilization hypothesis even among biologists: How can you advance by retreating? We're not juvenile chimps, after all.

Juvenilization goes against the widespread notion of "progress" in evolution. After all, it suggests that an adult ape-hominid was overspecialized, that it had to back up to produce us. It suggests that the child may be better than the adult, contrary to our personal experience of becoming more competent as we grow older. Yet once you begin studying the subject and get beyond these surface reactions, you see that juvenilization has to be at least part of the answer to hominid evolutionary mechanisms -- there are just too many facts starring one in the face, like the doubled and redoubled childhood, the flatter faces and smaller teeth, those greater brain/body ratios.

The issue isn't juvenilization, but whether there is some other important mechanism involved in addition to juvenilization. For example, there might be a developmental tendency for the head end of the body to enlarge faster than the tail end. This may happen in the squirrel monkey, whose 1/31 adult brain/body ratio is even bigger than our 1/49. One can see the opposite in the cheetah, whose body is much larger relative to its head size than an average cat. If there is an additional mechanism for relatively larger head ends or tail ends, we don't know much about it yet: it's undoubtedly a matter of relative clock rates in development, but what are the linked traits? What selects for those traits? Which get hauled along for free? What are its disadvantages?

Tune in next year. But in the meantime, the advice we'd offer ambitious chimps is: Grow young.

I do not know what I may appear to the world; but to myself I seem to have been only like a boy playing on the seashore, and diverting myself now and then finding a smoother pebble or a prettier shell than ordinary, while the great ocean of truth lay all undiscovered before me.
......the natural philosopher ISAAC NEWTON (1642-1727)

In our innermost soul we are children and remain so for the rest of our lives.
.......the early neuroscientist SIGMUND FREUD (1856-1939)

Man's maturity: to have regained the seriousness that he had as a child at play.
......the philosopher FRIEDRICH NIETZSCHE (1844-1900)

[Children] are fantastically interested in making things and in asking "Why? Why? Why?" Then, at a certain age, the children just become adults and are no longer very deeply interested in anything, except in the process of making a living and in sex and power.... Profound curiosity happens when they are young. I think physicists are the Peter Pans of the human race. They never grow up, and they keep their curiosity. Once you are sophisticated, you know too much -- far too much.
......the nuclear physicist ISIDOR ISAAC RABI, 1975

In a sense, all science, all human thought, is a form of play. Abstract thought is the neoteny of the intellect, by which man is able to continue to carry out activities which have no immediate goal (other animals play while young) in order to prepare himself for long-term strategies and plans.
......the mathematician JACOB BRONOWSKI (1908-1974)

Mile 148
Matkatamiba Canyon

MATKATAMIBA CANYON is special. We pull into a narrow side canyon just before hitting the rapid; unless you knew about this secret place, you'd zip right past, spotting the canyon entrance too late to pull in. It's hidden around a corner. The boatmen come down hugging the left shore. Then, as we round the corner, a lot of hard rowing and we enter the quieter waters.

The entrance to Matkat is very narrow, with only enough room for a few boats to moor. The others tie up behind them and the passengers pick their way across the other boats. Carefully.

Matkat stays narrow: usually I can keep one hand on each wall. As at Silver Grotto, I have to wedge myself up in some places where there are waterfalls and some greenery. I can manage to walk along one side of the creek while pushing with both hands on the opposite wall, edging along at an angle. The creek looks easier to descend than ascend -- there are narrow spaces where one could slide downhill, as in a children's playground. Getting up them is the problem, as the stone is worn smooth and the microgreenery makes it slick near the flowing water. But the canyon layers of Muav limestone have handholds, and so I can wedge up in chimney fashion, always keeping at a 45°dg angle.

No sooner is this first obstacle surmounted than I reach a pleasant flat section that wets my ankles again as I slosh along happily. Then comes another chute which I must wedge myself up. But if you arrive at a rock-plugged crevasse that requires both a boost from below and maybe a hand extended down from above, you've gone too far -- go back and climb the Muav cliff to the right, up away from the creek.

And then the canyon opens out, soon reaching an amphitheater in the Redwall, right where it sits atop the Muav. The unconformity is very obvious here, those missing layers from between 535 and 360 million years ago. You may not be able to see them, but here the Muav looks to have been sanded flat and then another layer-cake glued down atop it.

The gray-green limestone-sandstone forms the floor of this natural theater, and the red walls are streaked with black desert varnish, punctuated by patches of greenery. The creekbed in the Muav has a series of bathtub-sized pools with slowly flowing water. The water is warm, and we tried them all, one after another. I kept returning to the middle one, and decided -- as I lay there with the warm waters flowing around me -- that if I ever made a fortune, I'd have a landscape architect copy this magical place, right down to the desert plants that surround it, and have it installed somehow in my back yard. Unfortunately, it would require a crew of gardeners to maintain. But Matkat manages nicely without help, because each of the ingredients has been selected by the environment over the centuries to work together with the others. It is a small ecosystem that cannot be readily transported elsewhere.

The rock slab where I am resting, I now notice, is another masterpiece of Cambrian wormworks, just as at Deer Creek. The fossil casts are in just one layer of the Muav; the lower layers lack any sign of the finger-sized grooves running every which way.

Matkat at Mile 148, from Leonard Thurman's Grand Canyon River Running web pages.

BUT BIGGER BRAINS MIGHT ALSO BE BETTER, not just freeloaders, to come back to the usual hypothesis. Bigger-is-smarter-is-better is representative of a large number of explanations for the Great Encephalization that come readily to mind. Some "explanations" are mere romantic nonsense, articles of faith such as Robert Ardrey's resounding dictum: "We do not think because our brain is big; our brain has grown big because we think." (This dramatic statement is at least from a dramatist; alas, it echoes the statements of many anthropologists and neurobiologists who should be more suspicious of such comfortable notions).

Other "explanations" are based on mechanistic analogies to systems that work very differently from the biological brain: users of digital computers, for example, immediately comment that a larger memory capacity might have been useful (but such computers use pigeonhole memory storage of the kind that fills up, not the overlapping committees of the brain's distributed memory storage system, which may not saturate but merely require progressively longer times to find something). My objections do not, of course, rule out the possibility that there is more than a grain of truth in such proposals. Indeed, almost everyone's favorite proposal might well be right, in that the proposed mechanism could have played some role in hominid evolution. Fortunately, it is possible to focus the discussion a little.

Of the many things which might have benefited from a larger and more versatile brain, which was in the best position to work the evolutionary ratchet? And to do it most often? Was there a fast track for brain enlargement, some one factor operating over and over again to increment brain size? The problem any candidate faces is not only to have evolved a bigger brain, but to have done it very rapidly, in the last two million years or so. From the lessons of evolutionary biology, one can create a recipe for rapid evolution:

First, take a variable population, some with smaller, some with bigger brains. Expose it to severe natural selection. That means severe limitations on food, severe culling by predators, severe selection by disease, severe actions of climate -- with which some variants in the population can cope better than others. Overextended subpopulations, barely making a go of it in a somewhat foreign habitat that strains their existing skills, is the sort of setup that comes to mind. Small batches are best for this recipe. Many small isolated bowls are better than one big bowl.

Second, do not stir after selection. Not only can natural selection bias a small genome faster than a large one, but boiling the population down to a residue, and then having a subsequent population explosion based on those few survivors, is a fast way to make the genome idiosyncratic enough so that speciation occurs. Just repeat this bust-and-boom cycle (in the business, this is known as the Founder or Bottleneck Effect followed by a Population Flush) a few times and (at least in insects) reproductive isolation will start becoming noticeable. Speciation means that when the severely selected folk do eventually find their way back into the parent central population, they won't interbreed very well. Since their successful copulations are only with each other, it saves their own new-fangled genome from being badly diluted. After all, that central population has a lot more individuals than any isolated subpopulation. Though the biased genome of the selected subpopulation will shift the gene frequencies of the whole species a little upon remixing with the central population, it will be nothing compared to the shift achieved in the subpopulation itself before mixing. It may work eventually, but here we're talking about speed. The way to make rapid changes is to build atop the subpopulation, rather than by the tiny biases to the whole population that Darwin first envisaged. Thus the recipe calls for the speciation ratchet now and then, to prevent dilution of the progress achieved so far by selection, just as the ratchet on an automobile jack prevents the car from slipping back to the ground, just as the ratchet in the clockworks prevents time from seeming to run backwards.

Third, repeat the cycle as often as possible. The unsettled geology and tides of Danakil? But even more frequent cycles can be achieved by a locale in temperate climates, where there are real winters unlike the tropics, so that the selection process can be nudged along by yearly episodes of hard times. And there is that 100,000-year cycle in climate, starting 3 million years ago, with the mini-ice ages taking rainfall out of circulation. In our classic examples of adaptive radiation, such as those arising after a mass extinction, there are many niches available and few rivals, and so there is a rapid diversification. But here we are concerned with repeated adaptation in a particular direction. That's a very different problem. Repeating the cycle only produces results as long as natural selection continues to have an effect. Suppose, for example, that bigger brains were a mere side effect of selecting for hairlessness via juvenilization. Infants do have finer fur than adults, and maybe an aquatic environment's selection for less hair would therefore sometimes use the juvenilization family of features, when selecting for the clock rates that give less hair in adults, dragging along a bigger brain for free. But continuing selection pressure won't have any effect after a while: one can, after all, become only so naked. That less-hair growth curve flattens out at a limit. But some things can grow over a wide range. So it would be best to base natural selection for bigger brains on some feature other than hair -- a feature that could forever encourage brain growth, preferably one with a steep growth curve.

Now my reliable recipe for rapid ratcheting is not obligatory; for example, the speciation step might be eliminated if one doesn't mind the risks of losing everything to a premature remixing with the central population. Cultural aspects of behavior can substitute for speciation in many cases: when remixing, the two groups might tend to maintain their own cultures, just as the Developed Oldowan toolmakers and those preferring the Acheulean tool kits (the hand ax throwers) managed to coexist for so long. And that same separation by culture might tend to minimize interbreeding. There are many other such "barriers" (as they are called in evolutionary biology) to interbreeding.

But because there is always the chance of tumultuous remixing, as in raids and rape, episodes of regular speciation are good insurance against the loss of a specially selected genome. Who knows what creatures developed, omitting the insurance of a ratchet, and then got diluted out of existence? And it is certainly not obligatory to start in temperate climates with the selective nudge provided by every winter; one doesn't have to speed up evolution in every possible way all at once to ensure rapid growth of the brain. In reality, the mix of rapidity-encouraging elements probably changed, with episodes of two steps forward, then one backward.

... the increase in brain size in such a short period from an average of 460 g to more than three times as much is almost unbelievably fast.
......the evolutionary theorist ERNST MAYR, 1973

THE FAST-TRACK ARGUMENTS for human brain enlargement thus center around ape-human differences that are climate-dependent, more important on the periphery or islands than in the central tropical population, directly related to survival skills rather than contemplative intelligence, and likely to involve some aspect of the juvenilization collection of traits open to natural selection.

Could diet have done it, since the smartest animals are often omnivores? Learning to eat more types of food is useful in expanding one's niche, enabling one to live in different kinds of environments. But however good a general principle that may be, I doubt that diet will explain the special case of the Great Encephalization: that's because chimps are already pretty accomplished omnivores; we may, or may not, have picked up a taste for dead meat along the way, enabling us to take advantage of dead animals, steaks that chimps would pass up. I don't want to minimize scavenging; it might have aided the development of throwing (driving away hyenas with thrown rocks, etc.) and hammering (well-protected brains and bone marrow are among the leftovers), and gotten hominids into the hunting-by-throwing business gradually. But while applicable to the savannah of equatorial Africa where there are lots of game animals, scavenging is intrinsically a way of life that spells a low prehuman population even there: we'd have been dependent on the top predators for our living -- even though not eating them -- and it takes a lot of game animals to support a single lion (and even more if we were stealing part of the lion's food). For standard food-chain reasons often forgotten by enthusiasts of the scavenging hypothesis, that means far fewer prehumans than lions. And scavenging would give a very low yield outside the densely-populated savannas. Scavenging would thus surely have been only one minor part of a more complex food economy but, because of its influence on hammering and throwing, perhaps of major importance.

Human gathering isn't all that different from what chimpanzees or baboons do. True, we may make use of digging sticks; true, we tend to carry things back to others, postponing consumption and sharing our food. But what is the connection of gathering to juvenilization or bigger brains? What kind of growth curve might it have?

How about predation of the usual snatch-or-chase type? One starts with the baboon and chimpanzee picture -- again already at a near-human level. Their predation involves group cooperation in stalking -- of a kind we used to think was uniquely human. One can say that improved spatial skills would be useful for hunting, but one comes up against the considerable baseline of chimp and orangutan skills -- they unerringly and repeatedly locate fruit trees in distant corners of their range during just that week of the year when the fruit will become ripe. So navigation doesn't look like such a dramatic improvement.

What about throwing-type hunting? It fits a number of the aspects of the recipe. It's more important in temperate and arctic climates, if only because gathering is so restricted at certain times of the year. Though they throw in order to threaten, apes don't seem to use aimed missiles for hunting to any extent. However, their skillful hammering and threat throwing are obvious stepping-stones toward displacing predators and other scavengers from a kill, toward developing crudely-aimed predatory throws at groups of animals visiting a waterhole (the hand ax story), and then to aimed throws of the familiar type. There is a long growth curve in hominid throwing, just judging from improvements in missile type along the way: selected rocks, hand-ax-type discuses, wooden spears, rock-tipped spears, throwing sticks, slings and slingshots, bow-and-arrow -- not to mention all the advanced projectiles.

There is another substantial growth curve: in throwing technique. You can see it by watching children of various ages. At first a child tosses things without much aim or control, reminiscent of how chimps and gorillas throw in their threat displays. Then smaller rocks are selected, with more of an emphasis on distance achieved, and on hitting a target. The child learns to "get set," concentrating on the task, throwing in a stereotyped way over and over, making minor modifications in speed or release point in order to change the impact point of the projectile.

Applied to hunting, you immediately see a third kind of growth curve: accuracy, or, as it manifests itself, "approach distance." It's easier to hit a nearby target than one farther away, and so the hunter tries to get as close to the prey as possible. While prey animals will often ignore a two-legged human when they would run from a four-legged carnivore, there is still a point at which they will move away from you (their "approach distance"), sometimes just moving back to maintain a certain distance, at other times simply picking up and running. The farther back you keep, the less likely they are to be spooked by the beginning of your throwing motion. So an ability to throw farther with the same accuracy means more successes.

But there is a big bonus for learning to hit a target from twice your former distance. To throw twice as far typically means launching the missile with double the former speed, as distance achieved with a nearly-flat trajectory is proportional to initial velocity. Ignoring air resistance for the moment, this means that the missile arrives at its target with twice the velocity, so that its kinetic energy -- being proportional to the square of the velocity -- may be four times greater than formerly. Thus the "stopping power" of the missile quadruples when one doubles the distance. This means that one can tackle bigger prey, graduating from birds to rabbits, rabbits to bushpigs, bushpigs to gazelle, and so on. Farther is also better for this second reason. This has a nice kind of growth curve too, following a square law rather than merely being linear.

So there are four kinds of growth curves for throwing, all seemingly unlimited. The only trouble is that throwing twice as far means making decisions more than twice as fast. Indeed, nearly eight times as fast.

TIME IS OF THE ESSENCE, once one starts looking carefully at throwing with a neurophysiologist's eye for detail. Traditionally, the first thing about which a neurophysiologist would remark is the impossibility of feedback corrections late in the throw. Muscles have sensors embedded in them, as do the tendons and joints; they tell the brain where the arm is located, via messages coded in nerve impulses.

But they don't tell the neighboring muscles, except via that long round trip into the spinal cord and back out again. The brain only knows where the arm was about 1/25 of a second earlier. It can take a message 1/50 of a second to travel from the arm into the spinal cord; it can take another 1/50 of a second (and usually longer) for the spinal cord to tell the brain about it too. Unlike wires through which electrical signals travel at almost the speed of light, nerves use a relay system not unlike a burning fuse or a row of falling dominos. It's not as slow as the mail service, but it isn't as instantaneous as the telephone. It takes even longer for a command message to travel back out, from the spinal cord to a muscle in the arm (the motor nerves aren't designed for speed in the same way as sensory nerves). And it also takes time to make decisions in the brain -- the reaction time is anywhere from 1/10 to 1/4 second, or even longer if you're indecisive.

One can make little corrections early during the throwing motion. Though not, of course, after you've let loose -- that's one of the disadvantages to unguided missiles. But one also can't make corrections in the last 1/10 second or so before letting go -- there just isn't time to gather new data, make the decision, and send the new commands out to a muscle. So there's a point of no more feedback. Once past it, one can't make corrections any more. Your brain is on its own.

The faster the throw, the shorter the throwing time. But the "period of no more feedback" doesn't change; it stays at about 1/10 sec. Thus, a larger and larger percentage of the throwing time is impossible to modify as one throws faster -- one has to predict what to do on the basis of early data fed back from the arm, telling how rapidly it is actually accelerating, then do the calculations of the trajectory of your rock based on that, and set the release time accordingly.

For some particularly rapid movements (eye flicks, for example), you just have to forget about muscle and visual feedback altogether and send out exactly the right command sequence to the muscle in the first place. You plan carefully for ballistic movements, so that you don't have to correct the command sequence in midcourse. Throwing farther may be better, and faster may be its critical mechanism, but it sure does make life hard for the poor brain.

Take a simple overhand throw, with the body considered rigid just to simplify things down to the arm's motion. You cock your elbow, hand atop your shoulder grasping a rock, and then contract both sets of muscles: the extensor muscle groups that uncock the elbow, and the flexor muscle groups that cock it. Contracting both opposing sets (known as co-contraction) serves to stretch the muscle tendons, storing energy in them just as if you'd stretched a spring. You keep the two tensions large, but exactly equal so that the arm doesn't move in either direction. Then the brain gives the command to stop contracting the flexor muscles. And the elbow starts uncocking, both the stored elastic energy and the active forces from the ever more active extensor muscles serving to accelerate your forearm so that it moves faster and faster in its upward arc. At some point, the rock you're grasping flies loose. The time at which this happens is under the control of your hand muscles. Essentially, you want to open your fingers at just the right moment and so let the rock slip free. Imagine a robot hand, flapping open upon command. Human hands are more complicated than that, but whatever the thumb and finger muscles do together has to be timed with a precision equal to when that command would be given to a robot hand.

And what is the "right time", the right moment for releasing the rock? That depends on how far away your target is. And how big it is. Release too soon, and the rock will lob too high, going too far and impacting behind the target. Release too late, and the rock will hit the ground in front of the target. The right time is whenever the resulting trajectory will cause the rock to come down somewhere on the target.

Suppose that the target is a rabbit -- the rabbit is facing you calmly eating a bit of greenery. Little does it know, evolution not having informed it yet, that you're a new-fangled action-at-a-distance predator. The standard rabbit, let us say, is 10 centimeters high and 20 centimeters from front to rear (never mind its width, since that turns out not to be as crucial). The range of correct times for release (in analogy with a moon rocket, we can call it the "launch window") are those between an early release where the rock travels a bit too far and hits the top rear of the rabbit, and a late release that would land the rock on the rabbit's front paws.

Now it is a simple matter for any freshman physics student to calculate the rock's trajectory. So just plug in the numbers for various different release times and see which ones hit the target. This is equivalent to what artillery gunners do in "walking" the impact point of a shell onto the target, except that they just adjust the angle of the barrel. But that's essentially what adjusting release time is doing: release early and your rock will head upwards at an initial angle; release late and it'll head out horizontally instead. Intermediate release times are just the intermediate angles.

To hit a standard rabbit from 4 meters away is pretty easy; most of us could do it with a minor amount of practice, since the distance is the length of a subcompact automobile. Just imagine standing alongside the front bumper and throwing at a stuffed rabbit sitting on the ground near the rear bumper. For 4-meter throws, the average launch window is 11 milliseconds. This is about as long as a camera shutter stays open when set at 1/100 sec. Release anywhere within that 11-millisecond window, and the rock will hit the standard rabbit somewhere on its front or top.

Now move the rabbit so that it is twice as far away: two small cars, parked bumper-to-bumper. Throwing will become much harder, though most of us nonexperts would succeed with some practice. You're throwing twice as fast, so you expect the launch window to halve just from that. But the rabbit also presents a smaller-looking target to you; at twice the distance, the visual angle between the top and bottom of the rabbit will more than halve. Thanks to the trigonometry of small angles, the angle drops to a quarter of its original value. So it comes as no surprise that the launch window at 8 meters has dropped to 1.4 millisecond, 1/8 of its value for the 4-meter throw. You have to get eight times better in your timing to throw twice as far with equal success rates. That's why it's so much harder.

Now real machines are not infinitely accurate. Neither are we. When we practice and manage to hit the target at double the distance, we've done something to improve the precision of our brain-and-muscle combination. We've improved our timing.

Accustomed as some of us are to clocks that can split a second into a billion equal parts, it may come as some surprise that cells aren't capable of doing this too. In fact, as clocks, they're pretty poor, jittery as can be. But brains still pull off some great feats of timing, as an expert thrower demonstrates. How? They use lots of jittery cells, all trying to do the same job.

Hearts use the same trick. Take a single embryonic heart cell, resting on the bottom of a glass dish. If one uses a microscope, it can be seen twitching a few times every second. It isn't a very regular beat, with some intervals between beats being twice as long as others. Played through a loudspeaker, a lone cell sounds like rain on the roof: highly irregular. Now take another such cell and push it over until it touches the first cell. Heart cells are not only sticky, but they exchange electrical currents with one another. Both cells were beating independently of one another when separate, but once stuck together, their beats synchronize -- they beat together. And a funny thing happens: the beat starts sounding more regular. There aren't as many long pauses or short, quick double beats.

Just keep sticking cells together (how to build a heart!) and one will soon have a mass of synchronously contracting cells. And the beat will get more and more rhythmic, ticking along with great regularity, each interval pretty much the same as the one before. It sounds like a rapidly-dripping faucet, not at all like the irregular beat of the single, isolated cell. The jitter -- the range of fluctuation -- narrows by half when you quadruple the number of cells in the cluster. When there are a hundred cells in the cluster, the fluctuation range is ten times narrower than for one cell by itself. More and more is better and better. To make a really rhythmic beat, use lots of cells. Our regular heartbeats come from thousands of heart cells (in a region called the sino-atrial node) all beating together like this, setting the pace for the rest of the heart to follow. If only a few dozen pacemaker cells were on the job in the sino-atrial node, our heartbeats might be rather erratic, bouncing around between too fast and too slow.

Nerve cells can use the same trick, solving the precision problem with large numbers of cells, even though they don't visibly twitch like muscle cells (to a neurophysiologist, a muscle cell is just a nerve cell that can also contract). A nerve cell can beat in an electrical rhythm (you can hear it by hooking the cell up to a hi-fi system; it sounds like a dripping water faucet going tap-tap-tap), and that's how cells count time. To make a large number of nerve cells "beat" together in synchrony, you don't have to literally stick them together; nerve cells are much more sophisticated than heart cells, and their wiring diagrams can accomplish the same end. And it doesn't even take special, fancy wiring patterns -- the simplest kind of parallel summing circuit will suffice to create really precise beats, timing with any degree of accuracy you need.

All it takes is lots of cells. To make a fluctuation range 8 times narrower, as you need to do to double your throwing distance to a standard target, you just use 64 times as many nerve cells as you used before. To triple the throwing distance, all you need is 729 times as many cells. This isn't just square law for a growth curve: the number of nerve cells needed rises as the sixth power of throwing distance. Accurate throwing has an insatiable appetite for more and more synchronized nerve cells.

MORE NEURONS, BUT FROM WHERE does one get so many additional timing neurons? This isn't just a doubling that we're talking about. In the short run, I suspect you borrow them from elsewhere in the brain, the region primarily concerned with timing getting the neighbors to come and lend a hand.

Where are we talking about? Rapid movement sequences like hammering and throwing are likely orchestrated from a region of the left brain in the frontal lobe, just in front of the motor strip for the hand and arm; besides this premotor region of cerebral cortex, the cerebellum (which also enlarges nearly three times from ape to human) probably plays a major role in coordinating things too. When the cells in those areas have done their collective best to reduce the jitter, you perhaps create an even bigger circuit by using other regions of the frontal lobe, or perhaps by borrowing from the language areas of the temporal and parietal lobes. These regions have lots of connections with one another; you just turn off the business-as-usual connections and turn on the connections that create the massive parallel circuit that synchronizes them all. That means that you might not be able to talk and throw accurately at the same time. Or listen carefully to speech while launching. This suggests that as you concentrate and get set to throw, you're creating a really big parallel brain circuit just for the throwing command sequence. After the big push, the circuits relax and go back to business as usual -- it says here, with more than a little guessing and wishful thinking.

And so evolution might be a matter of selecting for brains that are wired in a manner that allows such temporary synchronizations to occur. Do any developmental trends hold out the promise of facilitating that? Yes, indeed. Early in development, it sometimes appears as if everything is connected to everything else, with many connections eliminated later. As I mentioned the other day when we got to discussing cell death, neurons from nearly all brain areas send connections down to the spinal cord; later in development, most of these connections have been broken, with only those from the motor strip remaining. The other connections have been somehow pruned or withdrawn. Thus, juvenilization might also select for wider connections, if this ontogeny observation can be so extrapolated. This fits nicely with a new theory for mammalian brain development; the neuroanatomist Sven Ebbesson says that things are indeed wired up widely at first, then selectively "pruned" (my word) to more clearly define the subsystems of the adult brain. Freezing the brain development at an immature stage might thus maintain a more widespread system of connections useful for those occasions when large numbers of neurons are needed in parallel.

But evolution could also just select those variants with brains that have extra neurons in the particularly important places, such as premotor cortex. And the easy way to do this is also through juvenilization: rather than selectively increasing the size of the premotor area, it may be easier to simply make the entire brain bigger. The other regions of the brain would thus get increases in their cell population "for free," even though they hadn't contributed to the cause. This might improve hearing, or memory for faces, or any number of functions unrelated to the throwing skills.

Thus, those prehuman adults who had greater juvenilization might have made better hunters than those with less-than-average juvenilization. A bigger brain might, in its own right, have been useful for the throwing skills so exposed to natural selection via hunting success and failure.

The growth curve for throwing-style hunting being what it is, there is no danger that this selection effect would saturate in the manner of hairlessness. Faster and faster is always better and better for throwing. Bigger and bigger brains can be faster and faster with the required timing precision. Thus bigger is faster is better. At last, a suitable mechanistic scenario for bigger-is-better.

This mechanism is, of course, only one possible way in which a bigger brain might also be a better brain, or in which a bigger brain might have arisen incidentally as a consequence of the success of some other juvenilized traits. But it fits the recipe for rapid evolution especially well: it provides an immediate reward in hunting success for any bigger-brained variants that come along from genetic permutations, and the hunting success is particularly important during winters and ice ages, capable of repeatedly driving the selection cycle. Bigger brains for better throwing looks like a fast track for evolution. Whether other tracks are even faster remains to be seen.

The hand is the cutting edge of the mind.
......the polymath JACOB BRONOWSKI
MATKATAMIBA'S AMPHITHEATER has a Muav floor with a series of cracks running along parallel to one another. Indeed, they almost subdivide the floor into a great checkerboard. Aha, the map says that the Matkatamiba Syncline is near here, so the Muav was probably flexed and cracked during some ancient upheaval.

A throwing contest has developed, using my poor hat for a target. We are standing on one crack and trying to hit the target at the next crack, then moving the target back another crack to make things more difficult as the contestants get better and better with practice. Now if only chimps had learned to hold such throwing contests, they might have indeed bootstrapped themselves up.

PARENTAL PRIDE BEING WHAT IT IS, you need to discount some of my enthusiasm for the throwing theory. I stumbled on it while throwing rocks at the beach one day. Back in the dark ages, when I was doing my Ph.D. dissertation, I happened to study the sources of jittery beats in the nerve cells of the spinal cord that run the limb muscles (the primary source turned out to be "bumps" from the small, but quantized, inputs to the cells -- not unlike the random walk of the Brownian Motion). I was, though I didn't realize it at the time, taking a lesson from Charles Darwin: he recognized that the motive force in evolution was not the species type but the individual variations about that average type. If you like, the "jitter." I had investigated whether there was something important about the variations in cell timing, rather than the average time. I didn't strike gold at the time, but learned a lot and filed it away in my head.

Sitting on that beach, I knew that individual cells just couldn't time things very accurately, no closer than about 5 parts per hundred. And my instincts told me that throwing surely needed better timing than that -- but how good? You don't have to take slow-motion movies of baseball pitchers to answer that question -- just work backwards from the physics of trajectories. So when I got home from the beach, I finally cranked up the computer and did the physics calculations, though it turned out that the needed equations weren't quite the ones in the physics textbooks (they assumed initial conditions that were too simple). There was a little delay while I derived the equations from scratch, starting from Newton's Laws and using integral calculus. And then wrote a computer program in BASIC.

Sure enough, throwing sometimes required an accuracy more like 1 part per thousand. And so how did circuits of many nerve cells become more accurate than an individual cell? Then I remembered reading about those heart cells in the Biophysical Journal, recalled reading in Science about some computer simulations of common nerve-cell circuits involved with circadian rhythms in sand fleas. Lots and lots of cells could do the job, and thus there might be a way in which bigger brains were better. Later, I remembered what the mathematicians call the Law of Large Numbers, and discovered that the physiological examples were just manifestations of this fundamental law. It seems that nature discovered the Law of Large Numbers even before Bernoulli did in 1713, using it to make heartbeats regular and throwing more accurate.

IF BIG BRAINS DEPEND ON HUNTING SKILLS and the hunters were male, then does that mean we owe our human brain development to our male ancestors? That's too simplistic. A little knowledge is a dangerous thing, and all that.

Even if men did 95 percent of the hunting, the occasions when women hunted could have been of crucial importance. Hunting is a hazardous business, what with the dangers of being gored by an animal or freezing in a storm when caught out in the open. It was surely not uncommon for the man to fail to return from the hunt. Then it would have fallen to the mother to provide all of the food, and in wintertime her hunting skills would have been the only way to stave off starvation. Even if infrequently exercised, female throwing skills would have been used in life-or-death situations rather than everyday ones.

But of more fundamental importance is that the evolutionary base for the throwing skills is probably the same rapid muscle-sequencing abilities that chimpanzees use for hammering on nuts. They involve the same arm motion, the same precise control of a ballistic movement, only without the release and with the elbow held lower. Since female chimps do more than 92 percent of the most skillful hammering tasks, we may owe much of our throwing skills to female ancestors; the subsequent brain enlargement may only have duplicated these basic ballistic sequencing circuits many times over. The neurological foundation for throwing may be female, even if males make more regular use of the extra copies in tandem. Which is more important, the foundation or its overelaboration? That seems a useless question.

Mile 150
Upset Rapid

TODAY'S ONLY BIG RAPID is rated an 8, and it is our last rapid of any size until we get to Lava Falls the day after tomorrow, another 30 miles downriver. The rapid's name brings to mind an incident that makes the boatmen very careful about how they rig their boats, and keeps them wearing a belt knife while going through big rapids. In 1967, a boatman was drowned here when his motor rig flipped and his lifejacket became entangled with the boat ropes. The passengers all survived.

The small oar-powered boats are rigged so their load remains intact if they flip; if there are no loose ropes to start with, there shouldn't be any after a flip, if the boat has been rigged correctly. When three of the big motor rigs flipped in Crystal in 1983 during the Fool's Flood, their ice chests, black bags, and ammo cans were found scattered downstream over a hundred-mile stretch; many probably wound up buried in the mud flats of Lake Mead.

We all got very wet in Upset, but stayed upright. Our last real warm-up for Lava Falls. And its black hole. Well, its alleged black hole.

Sinyala Rapid, down at Mile 153 is only rated a 4, and the boatmen consider it permissible to let passengers swim some such rapids if they have an overwhelming urge to do so. The river water has had three days since its release from Lake Dominy to warm up, and it's almost tolerable now. The air's downright hot.

Rosalie had for days been contemplating swimming a rapid. She had, I thought, a certain morbid fascination with the subject (I myself have never been the least bit interested in swimming a rapid). She decided that she was going to do it, having her adrenaline still up from our trip through Upset Rapid. So Jimmy told her how to do it: she should keep her feet out in front of her, so as to push away from rocks. And should keep her mouth shut.

Rosalie tightened up her lifejacket. With a great whoop she jumped overboard and went bobbing down the river, Jimmy keeping the boat just behind her and off to one side.

All was fine, with only minor complaints about the water temperature, until we went into the first big wave of the rapid. Which, when one's face is so near the surface, must have looked a lot larger than we were used to seeing from up in the boat.

Quite audibly, we heard a quick, breathless "Hail Mary, full of..." and about that time, a large cold wave struck Rosalie full in the face. She came up sputtering "Holy Shit!".

We were all still laughing when we picked her up below the rapid and hauled her back aboard. She didn't understand what was so funny, and asked what we were laughing about. So we told her about the interesting sentence that she had constructed. And Rosalie turned a deep shade of purple. Proving that with sufficient cause superimposed upon the right education, one can successfully blush despite a sunburn.

I then commented that her supplementary motor area must have really been disinhibited, something of an in-group joke among the more clinical neurobiologists among us, and which, being a stroke and paralysis expert, Rosalie understood just fine. But we had to explain it to everyone else. Most people know about the language center being located in the left side of the brain, out just above the left ear. But there is a second language area in the middle of the brain, just above the corpus callosum, known as the supplementary motor area. The two areas are far away from each other. People who have strokes that leave them unable to speak or understand what others say (aphasia) can usually still swear like sailors, much to the distress of their families. The only strokes that leave someone entirely mute are those of the supplementary motor area, not those of the main language cortex. And so the swearing center of the brain (it's also Sir John Eccles' most recent candidate for the seat of the soul, but not for that reason) tells you something rather interesting about the origins of language.

In monkeys, the brain areas that have something to do with monkey's vocalizations -- cries and barks and chattering -- are not the areas that you would guess from a knowledge of human language. In the monkeys it is that supplementary motor area that is the main piece of cortex involved in vocalization; all of the cortex homologous to the main human language area seems to have little to do with monkey vocalizations. So it looks as if swearing in humans is analogous to the monkey's vocalizations -- and indeed they are both rather emotional kinds of utterances. People who say swearing is rather primitive and unsophisticated may be more correct than they know.

But this leaves you wondering from whence the main human language cortex arose in evolution, if human speech didn't build on top of the more common emotional vocalizations. If you don't know, one starting hypothesis would be to assume that an adjacent area expanded, and then later specialized in language. So to what is the main human language cortex a neighbor? Most obviously, its neighbors are the auditory cortex, where sounds are deciphered. And the motor cortex for running the throat, mouth, lips and face, which is in turn right next door to the motor strip for the hand and arm. So those become the logical candidates for the origins of human language specializations. Did we get especially good at hearing? Or mouth movements? Or maybe hand and arm movements? Did language then build atop one such improvement to achieve our elaborate language abilities?

While chimpanzee calls do serve to convey basic information about some situations and individuals, they cannot for the most part be compared to a spoken language. Man by means of words can communicate abstract ideas; he can benefit from the experiences of others without having to be present at the time; he can make intelligent cooperative plans. All the same, when humans come to an exchange of emotional feelings, most people fall back on the old chimpanzee-type of gestural communication -- the cheering pat, the embrace of exuberance, the clasp of hands. And on the occasions when we also use words, we often use them in rather the same way as a chimpanzee utters his calls -- simply to convey the emotion we feel at that moment.... This usage of words on the emotional level is as different from oratory, from literature, from intelligent conversation, as are the grunts and hoots of chimpanzees.
......JANE GOODALL, In the Shadow of Man, 1971.

Mile 155
The Ledges
Tenth Campsite

THIS PLACE LOOKS LIKE A CAMBRIAN RESORT HOTEL, its ledges forming suites on three levels stacked up above the river. There is a seep dripping quietly in its middle, building up travertine. There is greenery around the wet rock. The Canyon is narrow here, the open sky almost looking like a skylight in a roof. There are a series of scalloped outcrops into the river itself, looking like a series of small piers in a waterfront development. It's all enough to make one look around for the registration desk. And for the little bronze plaque that tells who the architect was.

The suites are in natural caves formed by overhangs. That means they're hot, and Dan and I have learned by now. If instead we pick part of the flat Muav platform next to the river's edge, so as to get the breeze, we run the risk of losing things into the rushing river. I have already littered the river, a full can of soda got away from me and rolled in. Everyone is carefully weighting down anything that can blow away. And as usual, none of the rooms has a private bath. The toilet has been located a two-story climb up a rock-filled gully. It was hard enough to find at dusk; I hate to imagine someone looking for it by flashlight during the night. We also have to watch out for the wall-to-wall carpeting. This is travertine-coated Muav we're walking on; Gary warns us to keep our shoes on.

We were lucky to get this camp; we can't camp at Havasu (or any other of the places that get heavy day use, such as Redwall Cavern, the Little Colorado, Elves Chasm, and Deer Creek), and there is only one more campsite before we get there -- a little spot downstream called Last Chance Camp, which has all the disadvantages of The Ledges plus being smaller and more precarious to navigate at night.

Neither place has any sand beach. The Muav ledge just drops off steeply into the river, which is therefore flowing past the shore much more rapidly than usual, without any sand to slow it down. Gary suggested that we wear lifejackets if making a nocturnal visit to the river's edge, just in case we were to slip.

PLAYING CHARADES reminds us that nonverbal communication is quite sufficient for many purposes. One can tell a lot just from a person's posture. Moving toward someone, or staring at them, carries a message, as does turning one's back. Gestures are an elaboration of this, where hand and arm movements convey additional information. Facial expression is particularly important in monkeys and apes; one reason why we find chimpanzees "so human" is that they hug and kiss, raise their eyebrows, pucker their lips, and sometimes look sad. And sometimes angry. Human brains have a region of the right brain that specializes in interpreting facial expressions; if the area is put temporarily out of commission, the patient will mistakenly label a happy face as sad, or a sad face as disgusted, enraged or neutral. The extensive communication between a human mother and her baby makes use of postures and facial expressions, plus some soothing words whose function may be prelinguistic, and since mothers tend to hold infants in their left visual field (which reports first to the right brain), we surmise that this emotional-facial-judgment region is being extensively used in this elementary form of human communication.

Verbal communication adds to all this, being particularly important when two animals are out of sight of one another, as when up in the trees -- or have a particularly urgent message to deliver, concerning the arrival of a leopard. But when we start counting up the different types of vocalizations, we may wind up with a dozen different messages in a monkey, several dozen in a chimp. It is a long way from human language.

Human language doesn't utilize that many more basic sounds than in a chimpanzee vocabulary (though our phonemes are different, mostly shorter). Instead, we have evolved the trick of stringing those sounds together, with the order of the sounds being especially important for conveying information. We interpret a string of sounds terminated by a silent pause as a word, fitting that phoneme string to an auditory schema in our brains and coming up with a set of associations from memory -- the connotations of the word.

Just as a string of phonemes can make a word, so can a string of words make a sentence. And the order of the words is very significant -- the interesting sentence that Rosalie constructed (and she blushed again when we told the story after dinner, then threw a cookie at me with considerable accuracy) would not have been embarrassing had the order of the phrases been reversed. We interpret the word string using mental rules for word order, so that we interpret "Bill called Rosalie" differently from "Rosalie called Bill." The subject-verb-object order of a simple direct English sentence is not universal; Japanese, for example, assumes a subject-object-verb order and classical Arabic assumes a verb-subject-object construction. One of the things that comes with learning a language is a set of expectations about word order, which enables us to interpret the sentence in the same way as other speakers of our language do. Those expectations regarding word order are called grammar, or syntax.

So the brain needs a greatly improved sequencing ability for language, and a trainable memory for sequential order, but those capabilities need not be special to language per se. Just as the right brain has an emotional reputation, the left brain (more exactly, the language-dominant hemisphere) has a general reputation in humans for being obsessed with temporal sequences. Language is only part of it: rapid movement sequences of hand and arm, whether on the right or left side, are controlled from left brain. Ditto for oral-facial movement sequences: it is left brain that controls both sides of the face in sequential facial expressions. It is the left auditory cortex that specializes in detecting rapid sound sequences, whether they are language, musical phrases, or nonsense noise sequences. And it is the left brain that puts together the string of motor commands for such rapid ballistic movements as hammering and throwing (they are the most strongly right-handed of skills). Perhaps it is sequencing that is the key function, and language is just one of its latter-day applications.

As far as we can tell, human language results from a certain type of mental organization, not simply from a high level of intelligence.
........the linguist NOAM CHOMSKY, Language and Mind, 1965

LINGUISTICS HAS NOT HAD any traditional relationship to biology, and it perhaps took some audacity for Noam Chomsky to propose that there is a "language bioprogram" (as it is currently called) in the brains of all humans, and that this "innate bias" accounts for the many puzzling similarities in diverse languages and the ways in which they are learned and in the characteristic mistakes made while learning them. The bioprogram does not supply word order -- as can be seen from the way languages differ around the world -- but it does supply case relations (agent of, goal of) and grammatical functions (subject of, direct object of).

In many ways, Chomsky's proposal for a uniquely human "language organ" in the brain is merely using Descartes' "language is uniquely human" dictum to explain such regularities in the comparative study of languages. It has been criticized as being an organum ex machina, in analogy to the way in which ancient Greek playwrights solved thorny plot problems by bringing in the gods, who lectured the players and audience and resolved the difficulty. There was literally a "god machine" in classical tragedies, an elevated lecture platform on wheels that was rolled onto the stage and from which the gods spoke. Thus the phrase deus ex machina ("god in a machine") has come to signify any particularly contrived resolution of a storyteller's difficulty; organum ex machina is a commentary on Chomsky's language organ made by people who think the explanation contrived, that language may instead be an emergent principle, arising from the coordinated use of other mental facilities such as cognition, memory, perception -- and, I would add to the head of the list, sequencing.

Someone asked about whether Neanderthals could talk, at what stage of hominid evolution the speech apparatus was sufficient? Barbara explained how the issue came up (on most campuses, if one has a question about some aspect of primate anatomy, the local expert may be in the anthropology department -- anatomists don't always do anatomy these days). The upper respiratory tract changes a lot in mammals and also in human infants during development: the larynx, or voice-box, moves down in the neck. When the larynx is up high in the neck, the animal can simultaneously breathe and swallow, thanks to an interesting criss-cross anatomical arrangement called the piriform sinus, which solves an earlier design blunder: the trachea being in front of the esophagus rather than behind it. Until about 18 to 24 months of age, a human infant has a high larynx, much as do most mammals. Newborns breathe, swallow, and vocalize much as do the chimps and monkeys, for example. Sometime during the baby's second year, however, the larynx starts to move down, and this dramatically alters the way in which the baby breathes, swallows, and vocalizes. In this lower position, the larynx does not allow for simultaneous swallowing and breathing; rather, the two must be carefully coordinated to avoid aspirating food and water. And suffocation.

It is not known why the larynx descends. It is surely not part of juvenilization, for example. It has all sorts of disadvantages, such as choking when the coordination fails. But there is one lovely advantage: the vibrations arising from the vocal cords can be modulated by the shape of the throat, tongue, and lips over a much wider range of sounds than is possible with a high larynx. Maybe this is why babies cannot talk sooner than they do (they certainly hammer and throw at an earlier age!). But that also suggests that perhaps our ancestors lacked the human range of sound production too.

How does one investigate prehuman sounds, given that the larynx does not fossilize? The comparative anatomists have discovered an interesting correlation between the larynx position and the shape of the bottom of the skull, which does fossilize. The base of the skull is rather flat in most mammals and in human infants before the larynx descends. In humans after the descent, the base of the skull becomes flexed. Thus the obvious question (in retrospect!) is: When in hominid evolution does the flexed base of the skull show up? Australopithecines have flat bases much like the chimps and monkeys. Homo erectus shows signs of incipient flexing, suggesting that its larynx was moving down and its vocal repertoire improving. Some people see this as meaning that human language had to wait for crucial anatomical developments.

But, as was pointed out rather forcefully in our discussion, chimps have dozens of different vocalizations, not too different from the number of phonemes that any one human language uses. They might not be our phonemes, but chimps can sure tell the difference between them all. What chimps lack is the practice of stringing them together in a meaningful special order. The deus ex machina of the people who try, and fail, to teach chimps and gorillas a spoken language is that the ape larynx is insufficient for producing our range of vocal expression. So what? They can produce -- and distinguish -- almost as many elementary vocalizations as we ordinarily use, and if they had the proper neural sequencing machinery, should therefore able to string them together in an orderly way to achieve the benefits that we accrue from this clever coding method for conveying information.

We might have to learn their "phonemes" just as we learn those of the dolphins and whales, and it might be a slower language in which to speak a long sentence than human languages, but they'd have a language if they could just master the sequencing problems: producing a sequence; listening to a sequence and holding it in their short-term memory long enough to match it up with sequence schemata (word-order rules) in long-term memory. And doing their planning in terms of such sequence schemata, just as we do while talking to ourselves, would convey to them more of what we call consciousness.

One can't tell whether apes have sequential language from the failures and half-successes of teaching sign language to apes, because such gestural languages do not usually rely on sequential ordering; in many sign languages, several different parts of the message may be expressed simultaneously, just as in our ordinary nonverbal communication methods.

It wasn't vocalizations that made the language revolution -- it was sequential ordering and its rules. Maybe the chimps sequence their facial expressions and body postures instead. If there were rules about the order of elements, that greatly multiplied the number of possible messages that were sent, we'd have to concede that the chimps had a real language with syntax, even if it didn't involve sound.

But, of course, emotional speech doesn't utilize the rules of sequential speech: it's mainly one-word or stock phrases which aren't varied. Therefore, Rosalie pointed out, we were applying the wrong rules when stringing together her interesting construction in the rapid: each phrase stood independently, it isn't fair to string them together for an additional meaning since they were emotional speech.

Quite right, we agreed. We'd take back our laughs, we said, if we could watch her take back her blush. Like a movie run backwards.

Language, like other cognitive structures, is useful for some tasks and worthless for others. I cannot tell you, because I do not know, what my language prevents me from knowing. Language is itself like a work of art; it selects, abstracts, exaggerates, and orders.
.......ANNIE DILLARD, Living by Fiction, 1982

SEQUENCE IS THE SINE QUA NON of language. So if you want to start looking at the neighbors of the lateral language cortex for cues as to how it developed, out there away from emotional language predecessors, you might want to ask about the extent to which sequence is important for each neighboring function.

Auditory cortex -- well, the most prominent sound sequences (as opposed to single sounds) in the ape's environment are those emotional vocalizations of other apes and monkeys and leopards, at least until you get to language itself. Escalating sequences of vocalizations signal increasing social tension. There might be an intermediate sound sequence skill that's important between the apes and us, but we couldn't think of an example.

Then there's the motor cortex and premotor regions of the frontal lobe; the premotor cortex in particular has a "planning sequential movements" reputation. Down at the bottom of the motor strip is the control area for the larynx and pharynx, mouth and lips. Now at some point, we had to learn a lot of breathing regulation for diving. And it probably took some new coordination of breathing and swallowing whenever the larynx started moving down, along about the time of Homo erectus. But things like breathing and swallowing sequences usually aren't handled in the cerebral cortex -- they're handled down in the brainstem, closer to the spinal cord. So again, short of language itself, our group had a hard time coming up with any suitable examples of sequential movements of those structures that would have been exposed to selection pressures during hominid evolution.

Next on the motor-strip map comes the face, then the thumb and other fingers, the hand, the wrist, the arm, and the shoulder. You can find language regions in the frontal lobe just in front of such regions of motor strip, so it isn't getting to be too far away. This premotor cortex, with its sequential reputation, has particularly extensive interconnections with the wrist area of motor cortex.

It is possible that language developed through serial improvements in hand and arm gestures, then in facial expression, and finally in spoken sequences. Or the sequencer abilities in the neural machinery might originally have had nothing to do with language or gestures, only being used later for sequencing sounds. And the most prominent rapid-movement sequences are for hand and arm motions such as clubbing, hammering, and throwing. They certainly are exposed to natural selection, though each has a different growth curve; some, such as hammering, may not have improved much over chimpanzee abilities. And one may have improved as a result of improvements in another; converting a throw into a hammering motion (or vice versa), for example, doesn't take much change in the motor sequencer. The fast-track hypothesis provides one way of dealing with many interrelated causes, and throwing seems to win that conceptual competition so far, though the results are only beginning to come in.

And so, as the evening talk wound down, we were left facing a hypothesis that assigns our rapid acquisition of language, as well as the rapid growth of our brain, to getting better and better at a non-language sensorimotor skill: throwing rocks and spears at prey animals. Starting from general philosophical principles about human qualities in Descartes' manner, you'd never have stumbled upon such a hypothesis. Nor, starting from the knowledge base of linguistics, would you ever arrive there.

Assigning a major role in language evolution to throwing will probably remain heresy for a long time, even if it turns out to be the least awkward solution to the difficulties. Given the usual fate of most scientific hypotheses, it may well turn out to be another deus ex machina when we are farther down the road. But maybe it is the fast track, maybe language is an emergent property of brain circuits facilitating fancy time sequences, itself selected by hunting success, out in the fringe subpopulations where selection was harder and speciation was easier.

The ancient question is still awaiting an answer: What features in our brain account for our humanity, our musical creativity, infinitely varied artifacts, subtlety of humor, sophisticated projection (in chess, politics, and business), our poetry, ecstasy, fervor, contorted morality, and elaborate rationalization?
.......the neurobiologist THEODORE H. BULLOCK, 1984.

I JUST WOKE UP from a wild dream, thanks to some noise nearby (so I'm now scribbling in my notebook by the light of a dim penlight). My wife, in my dream, was looking out the window and remarking upon a dead mouse or shrew that the cat had hauled home as a present. It was left just outside the cat's basement entrance to the house, as if the cat had second thoughts about its acceptability inside the house. Katherine went outside to pick it up, and I came over and leaned out the window to look. It was some strange animal we hadn't seen before. It had a tail (bushy and striped, however, most unratlike). But its head, as she described it, "has a high domed forehead, just what they need for eating cheese."

What? And then I recalled, in the dream, that some animals like pigs have special head adaptations for rooting around, butting their heads into the earth in search of roots and other such goodies. Those side-projecting teeth are also what make piglets so dangerous to other piglets competing for the same teat (farmers routinely clip them to prevent bloody sibling rivalry). Why did the rat-shrew-whatever have a high domed forehead? I had, in the dream, a good knowledge about the inside of the backward-sloping skull of such animals (it comes with being a neurophysiologist), and so knew that it was strange for any of them to have domed foreheads. Except humans. And the animal, whatever it was, certainly wasn't even vaguely human.

The domed forehead was, my zoologist-wife said in my dream, for butting into a block of cheese, pounding off a chunk so that it could be carried away to eat elsewhere. (I now recognize this take-the-money-and-run scenario as coming from the pictures I've recently seen of chimpanzees carrying off bananas into the forest to hide them from the other chimps, so they can be eaten in peace).

But pounding with the forehead on a block of cheese? Good grief! What animal does that? At that point I woke up.

And I lay there in the moonlight trying to puzzle out where I could have possibly gotten the notion of animals pounding on something cheese-like with their heads -- in which the shape of the head was somehow important for pounding. Had I really invented something entirely out of whole cloth, or did I really have those concepts already in my brain from some source? The striped tail was easy -- that was appropriated from the ringtail episode we discussed the other morning back at Tapeats Creek. But the head...?

It has finally come to me. It wasn't cheese, but wax. The dome-shaped head was the hemispherical shape of the bee's head (well, I do mix up my phyla sometimes). They pound their heads against the wax walls of the tunnels through their honeycomb. What's so interesting about that is that hexagons miraculously appear when a lot of bees are all pounding on the same piece of wax with their round heads. Suppose some miners were tunneling through soft clay, and in order to shore up the walls in the network of tunnels, they butted into the soft walls with their round miner's helmets. And that miners in the neighboring tunnels were doing the same thing, all without coordination, just butting at random. The tunnels would begin to take on a hexagonal shape, quite without anyone intending it to happen. It's an emergent property.

LATER BACK HOME: I located the source of my dream schemata in a scientific journal I'd read just before the river trip:

A casual observer noting the perfect hexagonal structure of honeycombs is tempted to conclude that the universality and perfection of the hive structure are ensured by "instinct" or, more specifically, by some kind of innate hexagonal principle responsible for the bee's construction behavior. [I hope no anatomist ever went looking for a hexagon in the bee's brain].

However, it is now well understood that the hexagonal structure is an inevitable outcome of the "packing principle," a mathematical law governing the behavior of spheres packed together at even or random pressure from all angles. The bees' "innate knowledge of hexagons" need consist of nothing more than a tendency to pack wax with their hemispheric heads....

By the same line of argument, grammars [this is from Elizabeth Bates' critique of Chomsky's innate bioprogram for language] may be taken to represent a set of possible solutions to a much more complex formal problem, with some solutions falling out more easily than others on purely formal grounds.

It's an organum ex machina criticism, suggesting that a lot of neurons pushing around their electrical signals might have produced some patterns just like the hexagons, and that language makes use of them. Emergent principles, in short, have struck again. And I had a dream about them (I doubt the Anasazi would have). I wonder if any of those Cambrian wormworks we saw had hexagonal cross-sections? Ah, well, back to sleep.

...The way we think in dreams is also the way we think when we are awake, all of these images occurring simultaneously, images opening up new images, charging and recharging, until we have a whole new field of image, an electric field pulsing and blazing and taking on the exact character of a migraine aura.... Usually we sedate ourselves to keep the clatter down.... I don't necessarily mean with drugs, not at all. Work is a sedative. The love of children can be a sedative.... Another way we keep the clatter down is by trying to make it coherent, trying to give it the same dramatic shape we give to our dreams; in other words by making up stories. All of us make up stories. Some of us, if we are writers, write these stories down, concentrate on them, worry them, revise them, throw them away and retrieve them and revise them again, focus on them all our attention, all of our emotion, render them into objects.
......JOAN DIDION, 1979.

The throwing theory, for those interested, is best explained here and in chapter 8 of The Ascent of Mind . The more academic version is
W. H. Calvin, "The unitary hypothesis: A common neural circuitry for novel manipulations, language, plan-ahead, and throwing?" In Tools, Language, and Cognition in Human Evolution, edited by Kathleen R. Gibson and Tim Ingold. Cambridge University Press, pp. 230-250 (1993).

Home Page || END-NOTES || Table of Contents || CONTINUE to Day 11