There’s nothing like the view from the edge of a watering hole.  Our tents are on a slight rise, and the view consists of several elephants wading knee-deep, hosing themselves down and flapping their ears in the afternoon heat.  Two wart hogs wander among our tents in an accustomed manner, showing no interest in anything except grass – which they eat while kneeling on their front legs.  This shows a proper reverence for grass.

     The baboons are not reverent, not about anything.  Should they prove obnoxious, the Maasai askari is armed with two short bare sticks looking like thick drumsticks, which he bangs together to threaten them.  To the baboons, the guard’s sticks must sound like giant canine teeth, gnashing in threat.  Perhaps after a second beer, I might speculate about how rattles and drumsticks came to be invented by our ancestors, but not so early in the day.

     Rivers tend to meander if they aren’t hurrying rapidly downhill.  Since meanders are often eliminated when a flood erodes through a bank to create a more direct route, there are lots of dead-end channels (“abandoned meanders”) which make excellent watering spots.  And they tend to fill up with delicious vegetation, something our upright ancestors might have noticed as well.  I can even imagine them as refuges from predators.  If you splash water toward an animal’s face, it stirs up the blink reflex and tends to make them cautious about proceeding.  Offshore islands with trees are a good refuge.

     Gallery woods may form along the river, trees extending as far from the water as roots will reach, and our ancestors might have found them comforting.  Could we somehow test australopithecines on Gordon Orians’ landscape esthetics, they might well prefer the combination of shoreline features and forest features – not isolated trees in a grassland with a pond, what we seem to prefer.

     The Maasai Mara is the northern end of the Serengeti Plain, grasslands interrupted by the occasional acacia tree and the even more occasional waterhole or stream.  With the aid of some DNA-coded recipes, the grass is readily turned into herds of elephant, zebra, bovids of various types, and antelope of even more types.  The various seeds feed the baboons.  The trees and bushes feed the giraffes and other browsers.  The grazers in particular feed the lions, cheetah, and leopards.  The leftovers feed the jackals, hyenas, and vultures (though they too will hunt on occasions).

     As Phillip Tobias said, “From 1925 to 1995 almost everyone grew up on the ‘received wisdom’ that the Hominidae (the family of mankind) was born on the savannah, believed to have been the ideal crucible in which the strange form of locomotion known as bipedalism came into being.” But since Serengeti-scale savanna scenes are only one or two million years old, our earliest after-the-apes ancestors didn’t move into this scene so much as they evolved with it, as the slower climate changes and uplift produced more grass and less forest.  Even worse for the savanna bipedalism hypothesis, the earliest bipedal hominid fossils are found with flora and fauna that are associated with forest.  So before Homo erectus days (they are found in much more arid areas), hominid habitat may have been forests – but perhaps ones near a river or water hole.

     For all three of the major steps up to becoming human – upright posture, bigger brains, and the modern mind (a separate problem, more in a minute) – we are back to square one, having eliminated the favorite ideas of prior decades through a lot of hard work by the primatologists, paleoanthropologists, and archaeologists.  Not one of those steps now has a fleshed out, widely agreed upon theory for why it happened, though there are some interesting candidates.  This is progress, but we badly need fresh ideas.

Forests are particularly vulnerable to fire, because they can spread it from tree to tree so easily.  And it can take centuries for a forest to come back, usually through a succession of grass to brush to small trees and eventually to climax-forest trees.  When brush burns, you may get grass for years afterward. 

     (No one knows how far back the practice goes of setting fires to clear the underbrush, improving the forage for antelope the next year, and favoring the various plants that produce berries, but such fires are a common practice of preagricultural peoples around the world.  When flying over agricultural parts of Kansas, one can still see mile-long fires as fields are burned on a windless day, with a vertical curtain of smoke hanging in the still air.  There is a great deal of forest and field burning in India and southeast Asia today.)

     When a drought hits, the lake margins become refugia for various species .  Many forests then burn down, once lightning discovers dry trees.  A year later, the former forests become temporary savannas.  This creates a boom time for the grazing animals, with their short generation times (no “childhood“ at all – many females become pregnant shortly after they themselves are weaned), allowing population doubling and redoubling in just a few years.

     So the aftermath of a drought, one severe enough to burn off a lot of forest, is potentially a boom time for experienced predators on grazing animals.  Which our ancestors likely were, and probably well before two million years ago.

Morning at the waterhole features birds that sound like oboes.  Now there is a family of four wart hogs mowing the grass, not the same ones as yesterday.  I see a fish eagle perched, eaglelike, in a crooked treetop with a good view.  There is a special quality of low-angle morning sunlight, casting features in high relief compared to the flat lighting of midday.  Perhaps the eagle has also noticed.

     What I need is some equivalent of early morning sunlight, for throwing into sharp relief the mental traits that evolved alongside, or ahead of, the stones and bones we can actually recover.  To appreciate the importance of our abilities to reflect and speculate, you have to have a foundation in the mindless stuff like feelings and emotions – among which are landscape esthetics.

     To appreciate the occasional emergent properties, and the new uses for old things, you need to get behind the names we give things.  (Is there really “language cortex” or have useful-for-language areas kept their original functions for the most part?  A “gene for symbolic behavior,” or just another application of a more general behavioral trait?)

     The physiologist Albert Szent-Gyorgyi liked to say that, for every complex problem, there is a simple, easy to understand, incorrect answer.  We humans are very much into storytelling, and it is likely that there are some gut-level mindless criteria which make us like one story better than another (though our rational facilities may sometimes allow us to supercede the gut-level guides).  We can be too readily satisfied when selected parts all seem to hang together nicely.  We may, for example, argue in a circle without realizing it, making no progress at all except to satisfy our esthetic sense about what constitutes a good story.  Scientists, in the course of our education and researches, encounter a lot of stories that no longer satisfy because we have learned that we (or our predecessors) were earlier fooled.

     With this caution, let me talk about acquisitiveness.  Birds are acquisitive of nesting materials.  We tend to connect that with rearing offspring and give it a purposeful label.  Packrats collect all sorts of things for building “houses.”  They cut branches, cactus pads, and leaves.  Then they pick up anything loose to add to the structure – bottles, cans, mule droppings, bones, papers, or even mouse-traps – just to fill space.  We can still ascribe purpose to this, but, because the class of objects is becoming more general, we have to be careful of our labels.

     From this, we move to the acquisitiveness of the local baboons, the ones that conduct raids on our tents near the waterhole.  They are reported to have a particular fondness for toothpaste, but their tastes are nonspecific enough that they are just as likely to make off with your camera bag, without checking first for anything edible.  Back home in Seattle, my neighborhood is plagued by someone who is so acquisitive of battered old cars that he has left dozens of them parked on our streets, coming around occasionally to push them a little farther down the block.

     Acquisitiveness is also the concept which I have applied (it’s somewhere in Lingua ex Machina) to how the human infant builds a language machine by listening for patterns.  As in nest building, there may be some general principles behind it – yet ones so loose that they don’t specify too many of the particulars.  First, infants begin forming up categories for the common speech sounds they hear, not whole words so much as the little units we call phonemes, less than a tenth of a second in duration.  Categories allow them to generalize across speakers, so that the mother’s /ba/ sound and the father’s somewhat deeper /ba/ sound are treated the same despite their differences.  By about a year of age, babies stop hearing many of these differences, having standardized them: a /ba/ is a /ba/ is a /ba/.

     The bird that sounds like a musical telephone bell is back, causing reflex jerks of my shoulders now and then.  I am trying to persuade myself that, because the call doesn’t repeat at an exact interval like a true telephone bell, it isn’t civilization intruding.  But categorical perception keeps winning, conforming the sound to my telephone-ring standard and grabbing my attention away from other things.

     By a year of age, babies are discovering patterns in the strings of phonemes and acquiring a few new words every day, just from the examples they hear (long before they begin speaking them).  You can say kids are like sponges soaking up words but that’s too passive a notion, one of the reasons I prefer the more active “acquisitive” as the characterization.  I suspect they’re searching for patterns in their environment.  Eventually the kids incorporate the more reliable guesses about underlying pattern as new mental categories.

     So kids have pyramided words atop the phonemes, and now have compound structures made from building blocks.  But then they do it again, discovering patterns in the strings of words they hear and inferring the grammar of that particular language:  ways of making plurals and past tenses and nested phrases.  This happens between the ages of 18 and 36 months.  Then they’re off detecting patterns on even longer time scales, that of the collection of sentences we call a story.  They infer that a proper story has a beginning, middle, and a wrap-up ending.

     Now you can talk about this as separate instincts for phonemes, words, syntax, and narrative – or you can try out an overarching principle.  Mine is that it is just repeated instances of pattern-finding acquisitiveness, on longer and longer time scales.  It may, of course, turn out to be something of both the general and the specific, but what’s interesting is how much you can build by just pyramiding ‘acquisitiveness’ concepts four times over.  It’s a pyramid, about like compounding atoms into molecules, molecules into crystals, and then making sharp tools out of the crystals.

Acquisitiveness by kids is why there might be more than just cultural spread involved when converting to widespread syntax, sometime before 50,000 years ago (more later).  Yes, I can see acquisitiveness for words per se having been around for a million years or more.  But pyramiding to syntax and then narratives in the preschool years, well before much plan-ahead or accurate throwing develops, happens so reliably in most modern kids that it makes me wonder if that higher-order acquisitiveness is an additional adaptation, backstopped by a speciation-like event.

     Of course, it need not be true speciation but simply assortative mating.  Though exceptions abound, tall people tend to select other tall people for mates, blacks tend to marry blacks, and so on.  It takes little imagination to suppose that articulate women, operating in a system that allows female mate choice, would prefer equally articulate men.  And so the rich get richer and the poor get poorer, once again.

     Unlike true speciation, assortative mating isn’t backstopped, with a ratchet to prevent backsliding.  Like dog breeds backsliding to mongrels, things can still become average.  Could syntax be lost if something like prolonged suspension of female mate selection (say, via a return to harems) removed this tendency towards divergent selection?  Perhaps not, because kids now have such strong multistage acquisitiveness for language that they can reinvent syntax, if no longer exposed to structured speech.  Those who acquire words early enough and fast enough (nine new words every day) might, like the deaf playmates of hearing parents, invent their own structuring.

The early-morning light has gone, the sun is hidden behind a cloud, and I can smell a proper English breakfast, not one of your out-of-the-box breakfasts.  “Out of the box” thinking is another matter, deriving from “boxed in” and the escape there from.  Escaping isn’t easy, and for some very fundamental reasons such as the simple categories we fall into, ones where their efficiencies for some uses serve to blind us in other ways.

     After a year or so of exposure to the Japanese phoneme that lies between the English L and R sounds, for example, an infant will no longer hear the difference between L and R if not early exposed to English speakers.  The infant has created an efficient box that, by standardizing variants, enables rapid processing of speech sounds.  But it does mean the English words, rice and lice, now sound the same.

     Well, we do exactly the same thing at higher levels, too – you see the problem in medicine all the time, where putting things in boxes serves to promote speed in decision making but also blinds the practitioners to the subtleties.  One expects that it even applies to those who learn some genetics, but not enough to be able to “think out of the box” on the occasions when mutations and chromosomal rearrangements just aren’t the name of the game.

     Certainly most of us who contemplate climate and human evolution have gotten trapped, at one time or another, in the Ice Age Box.  That name, “ice age,” traps us into thinking that the last 2.5 million years were all about giant ice mountains coming and going.  It traps us into thinking that such periods were icy cold when, here in the tropics, they were merely 3°^C to 5°^C cooler, not exactly fur coat weather.  It distracts us from characterizing periods as “dry” by focusing our attention on temperature per se.

    And the “glacial” pace of the ice ages, where it takes major ice sheets 100,000 years to come and go, may well focus us on the wrong time scale.  It’s the things that happen in a few years, like droughts and fires, that might be the relevant aspect of the “Ice Ages” for big brains and toolmaking.  Indeed it may be the decade-long transitions between states, not the steady states (warm-wet and cool-dry) themselves, that create the most relevant challenges and the most relevant new niche opportunities.

Most of the explanations offered for prehuman evolution, as I mentioned earlier, suffer from the “Why us?” problem.  An explanation needs to explain why all the changes happened to our ancestors and not to other omnivores such as the chimpanzees and the bears.  My explanation, whatever its faults, at least handles that one.

     The opportunities associated with bust-then-boom episodes were for a savanna-adapted ape.  As apes rather than big cats, they had the brains and hands suitable for using tools.  As waterhole predators they would survive the drought aspect better, with the larger lakes serving as refugia.  As savanna specialists, they would have a population boom in the next generation as the grazers expanded their range into the former forests.  Then dozens of generations later, when the sudden switch to warmer-and-wetter allowed a quick grasslands expansion into former deserts, there would have been a second boom time.  So – at least if you eat grass or grazing animals – the cycle is two ups for each downer:  Crash-Boom-Boom.

     A boom time gets rare variants surviving long enough to reproduce.  Nothing like this one-two opportunity happened to the chimps and bonobos, whose populations probably stayed small during the cool-and-dry phase, only slowly expanding when warm-and-wet finally came back.  Note that the contract-and-boom scenario repeats for hundreds of major episodes and thousands of minor ones, thanks to the climate instabilities and their rapid transitions.  Gradual changes would accumulate from repeated catastrophes.  “Catastrophic gradualism,” pumped by climate instability, affords us one candidate for how we ascended.

     Alas, boom-and-bust also illustrates how we could lose it.  I’ll save that for the long flight home.

Sunset at the waterhole seems promising, with its long shadows and renewed breezes, but we have been sobered by the prospect of leaving this Serengeti paradise tomorrow.  I learned a lot yesterday about the differences between lakes and the more ephemeral waterholes which dot the landscape here.  And it provides one way of summarizing what happened in the last five million years since hominids were still close to the apes.

     The big lakes that we saw up north in the Rift Valley, like Baringo and Nakuru and Naivasha, are clearly great magnets for wildlife.  They vary in size over the decades but seldom actually turn salty and then dry up.  One can imagine our early post-chimp ancestors hanging out around a lake, profiting from the way it focuses resources for predators.  Lakes bordered by forest, with a zone of former mud flat for grass, would have made for a nice transition, the comforting landscape esthetics of the chimplike forest still close at hand.

     Rivers too are always there, though some seasonally dry up enough so that you have to dig a little, to find the water flowing under the surface.  Rivers often have gallery woods bordering them, again a comfort and a refuge.  Australopithecines likely nested in trees at night, just to get away from the big cats.  Clearly rivers are prime expansion territory in good times, when lots of kids manage to grow up at the lake edge and need territory of their own.  When a drought hits later, the hominid population might have shrunken to what the lake margins could support.

     Rivers have waterholes, too, in those abandoned meanders that begin to silt up.  But the Serengeti also has a different type of waterhole, far from any obvious stream.  It is as if there were underground streams that had had a section of their roof cave in, probably with a little help from the elephants who like to dig for mud, perhaps because their low-frequency hearing enables them to hear flowing water beneath their feet.  (The Maasai cattle herders outside the park like to have elephants around, because they dig waterholes that the cattle can share, but their wildlife enthusiasm doesn’t extend to competitors for grass like zebras or to predators like lions.)

     The Serengeti has a foundation of crystalline rock that the ground water doesn’t penetrate very well, topped by volcanics that make good soils.  Indeed the elephant behavior provides many species with access to water at sites far from lakes and rivers.  The savanna is dotted with lone acacia trees and lone termite mounds and, less often, with unexpected waterholes.  Looking out across the savanna, you may not realize that a waterhole is there (many have no bushes or trees nearby) until several zebras climb out of it – the way horses climb up out of a folly, dug by an English landscape architect to hide a fence.  The waterhole really is a hole.

     Landscape esthetics suggest that our ancestors made the transition to being comfortable in such settings, not merely tolerating them.  And this is in spite of the fact that open savanna is prime predator territory, where there are few places to hide.  Savanna away from permanent water was the great expansion opportunity for hominids, tenfold greater numbers of grazing animals clustering around those isolated waterholes.  The waterholes tend to be ephemeral:  sometimes watery, sometimes muddy, and sometimes really dry.  This change can be seasonal but even your better waterholes can dry up for years.

     Just as the lakes-to-rivers expansion would reverse in droughts to center on lakes, so the further rivers-to-treeless-waterhole expansion would have been even more vulnerable in droughts.  Surely there was a lot of friction between groups and, when the drought crunch came, the peripheral savanna specialists probably tried to displace the ones that had comfortably stayed nearer the permanent water.  Expand-and-contract has a bloody side to it.

     Decades-long droughts, as I mentioned at Lake Naivasha, are rather common events, even in the good-times centuries of the historical record.  And the ice cores now show that worse droughts have happened worldwide every few thousand years when warm-and-wet flips into the cool-and-dry mode (and the Amazon flow is cut in half).  These droughts are far more severe and far longer, and sometimes they come in a series, a century when the climate goes mad with chattering whiplashes.

     The modern human population may not be limited by predators, but we’re still subject to population crashes from various causes.  The Four Horsemen of the Apocalypse supplemented death with warfare, pestilence, and famine.  The plague years in Europe were bad enough, where more than a third of the population died, but the natives of the Americas suffered far worse from the smallpox and tuberculosis brought by the somewhat resistant European settlers.  Most of the native population was lost simply from infectious disease, running ahead of the settlers.

     Famine from abrupt climate shifts has been equally dramatic and even faster, with some isolated populations wiped out entirely on many occasions.  What the population biologists call “die backs” were surely a common occurrence in human evolution.  The whiplash climate changes would have been far more stressful than the usual droughts, simply because of their widespread alterations in the availability of plants and prey animals.  Everywhere that humans lived, climate would likely have changed simultaneously – even the tropical regions would have had changes in rainfall patterns, disrupting the food supply.  The tropics may not have cooled more than 5°^C, but the rainfall patterns there would have abruptly shifted within the time scan of a single generation, creating patchy resource distributions, and raising questions of refugia.  Many subpopulations would not have been able to relocate before starving.

The modern human population expanded from millions to billions after agriculture revolutionized things.  In the United States and Canada, fewer than two percent of the population can now feed the rest, given modern technology, credit, and transportation (only a century or two ago, 80 percent were engaged in just raising food).

     But the population expansion does make us all very vulnerable.  Technology hasn’t done away with the more severe widespread droughts for which stockpiling will prove inadequate.  In good times like today, populations expand to fill the space available within only a few generations.  And it isn’t just physical space:  as farmers manage to feed more people from each acre of land, population growth fills in that “excess capacity.”  Let the productivity slip for any reason, such as a worldwide drought, and people starve if their society hasn’t devised good backup food sources.  Birth control, by keeping the population levels to what can still be supported in the least productive years, would be a humane alternative to the boom-and-bust cycle.

     Can we think our way out of this, produce yet another technofix?  Birth control might do the job if there were a long transition into a drier climate, slowly downsizing the human population to track the resources.  Though I doubt it, we might actually bypass the usual downsizing via the Four Horsemen of the Apocalypse, might even bypass what African history teaches us about the last thousand years of half-century-long droughts in otherwise good times.  Hothouse agriculture might, were it widespread enough, even feed a larger population than today’s.  Given enough time to make a gradual transition, all sorts of technofix might be possible, much as in the story that some thinkers advance – the notion that the Earth might actually be able to support twice as many people as now.

     Alas, they assume some form of stability that doesn’t exist, as they would know if they bothered to learn the paleoclimate story.  The transition-time factor cannot be left out of the equation.  We are very vulnerable to sudden decade-scale change.  When it happens, the struggle is not pretty.

The sun sets rather quickly in the tropics.  It’s dark already, with the distractions of the southern skies beginning to appear.  This isn’t at all the slow transition that I’m accustomed to, living at a latitude where twilight is prolonged.  Sunset in the Serengeti reminds me of the Eurostar train popping into Paris, without the slow transition from farmland to city via protracted suburbia.

     I hope that’s not a metaphor for our future climate, that it will arrive more suddenly than we expect.  But the record of past climates certainly makes it likely.  Until a dozen years ago, everyone lived in blissful ignorance about the severe climate flips that happen quickly, with just a few short years before hard times set in with a vengeance.