The Southern High Plains were dusted with snow, and from 30,000 feet they appeared to be a vast marble slab. Since the flight from San Francisco to Dallas wasn’t full, I slipped back and forth across the aisle, peering out the windows, hoping to spot a landmark. There aren’t many on this flat, featureless terrain. I was looking, in vain, for a sign that we were passing over Clovis, New Mexico, home of the archeological site that’s given its name to the people who first set foot in America. That is supposed to have been about 11,200 years ago.
I’d been thinking about Clovis since we took off from San Francisco that morning last December, mostly because I’d heard so much about it the day before, at the American Anthropological Association’s annual meeting. There molecular biologists who use human DNA as a clock to date population migrations gave a message to us hands-in-the-dirt archeologists: the Clovis-based arrival date is wrong. Very wrong.
The messenger was molecular anthropologist Tad Schurr, representing the Emory University laboratory of geneticist Douglas Wallace. Schurr and Wallace had examined the mitochondrial DNA of various
Native American peoples. This DNA exists in little organelles-- mitochondria--outside a cell’s nucleus, and you inherit it solely from your mother. Because it doesn’t get mixed up with DNA from your father, any changes in the molecule as it moves from generation to generation are a result only of random mutations. These, it turns out, happen at the conveniently steady rate of 2 to 4 percent per million years. That makes mitochondrial DNA a genetic clock.
Schurr and Wallace found that the Native Americans belong to four distinct lineages.
The Emory researchers then counted the mitochondrial DNA mutations in each lineage, figured out how much time was needed for them to occur, and deduced when these lineages were last together--that is, when they first diverged from a common maternal ancestor. That ancestor was someone who walked across the land bridge that once connected Siberia with Alaska, and the genetic clock started ticking when her descendants then spread across the New World. According to the biologists, that ancestor took her stroll long before Clovis time. Upwards of 21,000 to 42,000 years ago, Schurr told us, without batting an eye.
That’s hearsay testimony, of course: genes cannot be directly dated. Still, the news caused something of a buzz, and not just because modern genetics was failing to uphold a cherished tenet of old-fashioned archeology. What got people talking was that Schurr’s date provided support, from an unexpected quarter, for one of the most talked-about archeological finds of the last decade. From 1978 to 1985, researchers working in southern Chile excavated a site they claim was occupied earlier than Clovis times--some 2,000 years earlier, in fact. Ever since the news began spreading, that site, called Monte Verde, has received a lot of attention from archeologists seeking traces of a pre-Clovis human presence in the Americas. Schurr’s date fits beautifully with the Monte Verde evidence, and together they may revolutionize our views on the peopling of the Americas.
That won’t come as much of a surprise to those of us who were at another meeting, three years earlier, on the Orono campus of the University of Maine. A couple hundred archeologists had assembled to wrangle over the origin and antiquity of the first Americans. We were there for three long days and three very long nights, and by the last afternoon session I was-- all of us were, I suppose--tired, hungry for dinner, and ready to go home. The New York Times reporter had already left: no more news fit to print.
And then Tom Dillehay of the University of Kentucky, the final speaker, began to tell us about Monte Verde.
Ten minutes into the talk, the fellow sitting next to me whistled softly in astonishment, then asked aloud of no one in particular, What planet is this stuff from? I was wondering that myself. The evidence from Monte Verde was unlike anything most of us--who early in our careers learned to be thankful just for stone tools and scraps of bone--had ever encountered before in our archeological earth. When Dillehay finished 45 minutes later, the hall erupted in applause. I forgot about dinner.
That day in Maine many of us saw for the first time pictures of the extraordinary archeological treasures of Monte Verde: artifacts of stone and bone, of course, but also of wood and ivory; freshly preserved leftovers from meals of leaves, fruits, nuts, and seeds; the remains of crayfish and paleocamel; the torn flesh of an extinct mastodon; even the footprint of a child.
Those are hardly the usual contents of an archeological site. They are unheard of for one in the Americas dating to some 13,000 years before the present. The pros in the hall that afternoon knew exactly what that meant: the Clovis-first barrier to the peopling of the Americas, standing strong at 11,200 years, was threatened.
Established in the 1930s with the discovery of the Clovis site, and buttressed with radiocarbon dating in the 1950s, the idea that Clovis people were the first Americans once made perfect sense. As envisioned, they came out of northeast Asia across the Bering land bridge (Beringia) to Alaska, then headed south. Their migration was thought to be timed to the rhythm of glaciers. By 25,000 years ago those vast ice sheets had frozen 5 percent of the oceans’ water on land, plunging global sea levels and uncovering the land bridge. But 20,000 years ago the glaciers had grown so much that they blocked the routes south from Alaska. Only around 12,000 years ago, once the glaciers had melted back and the terrain dried and was reforested, did a passable southern route reopen, roughly along the present border of Alberta and British Columbia.
The first Americans must have headed south soon thereafter, and fast, for by 11,200 years ago groups were camping at a freshwater pond at Clovis, and by 11,000 years ago they had reached Tierra del Fuego, at the southern tip of South America. One can recognize traces of the Clovis people in the fluted stone spear points they left behind--a design first noted at the New Mexico pond. The explosive move to the south corresponded neatly with the extinction of over 150 million mammoths, mastodons, ground sloths, and other giant Ice Age mammals. To some that was no coincidence: these rapacious hunters, encountering big game that had never before peered down the shaft of a spear, would have had easy pickings and been spurred onward by visions of still more prey.
Not a bad model, all things considered, and for six decades it held up well while pretenders came and went. But the Clovis-first model has problems, and one of them is enormous: How, and why, did people race from Alaska--where archeologists have found Clovis-like traces in sites about 11,300 years old--down to Tierra del Fuego, nearly 10,000 miles away, in scarcely 300 years?
Granted, that’s only 33 miles a year, unbearably slow by today’s standards; many of us have longer daily commutes. Yet it’s a breakneck pace for hunter-gatherers, easily four times faster than the current world record for prehistoric colonization of an empty area, set by ancestral Thule Eskimos. In just a couple of centuries around A.D. 1000, the Thule flashed from Alaska to Greenland. But they had it easy, following a familiar corridor of animals they had lived with for millennia. The corridor had just stretched thousands of miles eastward following a long period of warmer-than-average temperatures.
The first Americans had no such advantage. They were pioneering an infinitely trackless, ever changing, and (to former Siberians) ecologically exotic realm, from high mountains to high plains, and near- polar deserts to tropical forests. They were slowed each time they entered a new habitat and had to find plants, animals, water, stone, and other resources vital to their survival. Sadly, the romantic vision of fast- moving, mammoth-chasing hunters has no archeological reality. They were slowed by obstacles along the way, such as rivers swollen by glacial meltwater, mountains shrouded in ice, and freshly deglaciated barren landscapes. They were also slowed by the demands of keeping contact with kin, finding mates, and raising families. (Ever try to go anywhere fast with kids?)
But if the first Americans didn’t race through the continent, why do Clovis sites suddenly spring like dragon’s teeth from the ground in the centuries around 11,000 years ago? Maybe Schurr and his colleagues have the answer. Assume for the moment that the Americas were peopled 21,000 to 42,000 years ago, and then around 11,200 years ago someone invented the Clovis point--a handy, versatile tool useful in dozens of applications. How fast might such a good invention travel among groups?
Of course, if Schurr’s group is right, then we ought to find lots of pre-Clovis archeological sites. Some pre-Clovis proponents say we already have, pointing to sites throughout the hemisphere that brandish dates of 13,000, 33,000, and even 200,000 years ago.
But these sites have few believers--and for good reason. Of the scores of pre-Clovis archeological finds made in the last 60 years, none so far has withstood the harsh glare of critical scrutiny. Either their ages were inflated, their artifacts proved of natural and not human origin, or they hid some other fatal flaw. Exposing these flaws usually takes less than a decade, and then the site is tossed on the archeological scrap heap. Archeologists have long memories--it’s part of the job, after all--and in the face of many false alarms over the years, they have grown deeply skeptical of any and all pre-Clovis claims. The first site to topple the Clovis barrier will have to have undeniable artifacts in an undisturbed setting accompanied by unimpeachable dates, and it will have to win over the severest pre-Clovis critics: the Jackie Robinson rule, my colleague Mott Davis calls it.
That’s why Monte Verde rivets our attention. It may well be archeology’s Jackie Robinson.
Monte Verde sits along a small tributary creek of the Río Maullín, some 30 miles inland from the Pacific Ocean. This is a region shrouded in mist and clouds and a thick, verdant cover of forest and marsh that softens and rounds the landscape. At the site itself the land opens into a grassy plain about the size of a football field, through which shallow Chinchihuapi Creek slowly meanders. The sharp, snow-capped spine of the Andes looms to the east, but only on a clear summer’s day can you see the steam and smoke rising from its active volcanoes. And those days are rare enough in this damp and chilly climate.
But it’s the damp climate that makes the site special. The archeological debris that Dillehay told us about in Maine rests on a sandy bank, which, soon after the residents departed, was blanketed by water- saturated, grass-matted, oxygen-deprived peat. Beneath this anaerobic quilt the normal decay of organic materials was checked, preserving the site. Monte Verde came to light only in the mid-1970s, when local woodsmen, cutting back the banks of the creek to widen trails for their oxcarts, dislodged some buried wood and mastodon bones from their resting place.
Dillehay’s subsequent excavation showed that the bank was littered with the roots, stems, fruits, and nuts of nearly 70 species of plants, and even 3 types of marine algae. That’s many more plant spe cies than might be expected in a comparable-size natural deposit. We know that because Dillehay’s excavators hiked about a mile upstream from Monte Verde to dig where there was no hint of any human presence, just to see what was deposited naturally on surfaces of the same age. Such preemptive shoveling helps muffle critics who argue that the plants could as easily have been left at the site by floodwaters, for instance, as by people.
If nature was responsible for depositing the plants on the site, then nature is awfully hardworking--and more than a little bit devious. More than a third of the plants were imports, brought from their native habitats on the Pacific coast, high in the Andes, or from grasslands and other settings 30 to 250 miles distant. Coincidence or not, 42 of the species found on the sandy bank are still used by contemporary native Mapuche for food, drink, or medicine. Only the usable parts of many of those plants made it to Monte Verde and were found burned in some food pits, on the floors of what once were huts, and in shallow hearths. Even more unusual was the discovery of several plugs of chewed boldo leaves-- prehistoric chaw, they’d call that here in Texas--mixed with what appears to be seaweed and a third, as yet unidentified, plant. Boldo leaves are still used today to cure stomach ills and relieve colds and congestion.
The chaw was lying on a wishbone-shaped foundation of sand and gravel, which appeared to be glued together by animal fat. Along the foundation’s edges were vertical wood stubs and scraps of animal skin, the remnants of a hide-draped frame hut that once stood there. Fronting the structure was a small cache of salt crystals, plant remains, mastodon bones, a chunk of animal meat (which, based on preliminary DNA analysis, most likely belonged to a mastodon), hearths, and stone tools.
Monte Verde’s stone artifact inventory, now 700 pieces strong, includes finely crafted spear points, a slender and polished basalt drill, and cores, choppers, and flakes, several dozen of which are made of rock quarried many miles from the site. One can imagine that with these tools the Monte Verdeans worked, ate, and took cures in the wishbone structure. Most of them, however, lived 40 yards away in a group of 12 rectangular huts, each some 45 feet square. Nine of these huts were arranged in two rows, like row houses, lined by log planks staked together, framed by poles, and draped with a common roof of mastodon hide. Their floors were of sediments high in nitrogen and phosphate--a sure chemical sign of human waste--and littered with ash and grit. The huts surrounded two large communal hearths, two dozen smaller hearths (the child’s footprint alongside one of them), and more tools: digging sticks, mastodon-tusk gouges, grinding slabs, knives, spear points, and bola stones (rounded, grooved stones; when sinew is wrapped around the groove, the stone can be whirled and flung). Some of the artifacts were still speckled with the tar that bound them to their wooden or bone handles. In the huts were still more traces of the Monte Verdeans’ meals: plant remains, animal remains large and small (mastodon and camel bones, as well as a bird’s charred feathers, eggshells, and bone), and even a few human coprolites (the politely scientific word for fossilized excrement).
The age of Monte Verde is anchored by a chain of radiocarbon dates run on artifacts and samples of different materials, including charcoal, wood, bone, and ivory. The dates range in age from 11,790 to 13,565 years ago. Dillehay thinks the oldest one, from charcoal that was sealed and preserved in a clay-lined hole, best represents the age of the site, making it a good 2,000 years older than Clovis times.
But there the skeptics pounce. University of Massachusetts archeologist Dena Dincauze, for example, accuses Dillehay of uncritical use of the radiocarbon ages. A better approach, she argues, is to discard the oldest date, then look at the time ranges for each radiocarbon date and use their overlap as the best age for the site. In response, Dillehay suggests dropping the younger dates on bone and ivory (which are more susceptible to contamination) and averaging the dates run on charcoal and wood. Doing so, he says, puts the occupation at around 12,250 years ago, which, as retired archeologist Tom Lynch happily observes, lies at the very margin of Paleo-Indian time.
Of course, the passage south from Alaska didn’t open up for another 250 years, making the trip to Monte Verde an unlikely matter of time travel. But let’s assume there’s some slop in both dates, and that both the passage and the site appear a little before 12,000 years ago, with the passage opening first. Monte Verde is still 6,000 miles farther south than Clovis. To reach the Chilean site on time, these first Americans had to have left Alaska immediately after the passage opened up--giving them maybe 200 or 300 years to make the trip (our record-setting Thule could never have kept up). Leaving Alaska before the glaciers advanced 20,000 years ago, or earlier, seems a little more plausible.
So is Monte Verde just another Paleo-Indian site, albeit slightly older than most? Lynch growls that we cannot be sure, since so few of the artifacts have been illustrated or described in print. Dillehay’s first published volume on Monte Verde said next to nothing about the site’s artifacts or architecture, saving them for the (yet unpublished) second volume. This was a novel gambit on Dillehay’s part to convince his readers, solely on the testimony of the nonarcheological remains, that Monte Verde was a genuine archeological site. It caused a fair amount of grumbling, but he nearly pulled it off. Lynch, for example, admits the plant remains didn’t get to the site by chance--he’s just not willing to have them be there as early as Dillehay is.
Ultimately, making the case for Monte Verde--as Dillehay knows-- will require more detailed descriptions and photographs of the artifacts. It will require precise maps of the layout of the huts and hearths, showing why these features are demonstrably of human origin. And it will require showing the distribution of the artifacts, organic remains, and dated samples within these features. Jackie Robinson didn’t have it easy his first few years in the majors, either.
But will the publication of Dillehay’s second volume be all that’s needed to reach consensus on this site, and perhaps break the Clovis barrier? I put the question to University of Arizona archeologist Vance Haynes, the dean of the skeptics:
But why not?
Have you ever visited a site, Haynes asked me, that looked just the way you thought it would, based on what you’d read of it beforehand? I had to admit I hadn’t (truth is, sometimes I have trouble recognizing sites I’ve dug from descriptions I’ve written of them). In archeology, what one reads and what one sees are often very different. Archeology is like that: unlike researchers in the experimental sciences, we cannot replicate a crucial study in our own labs nor fully re-create a site’s evidence in words and pictures. Having Haynes visit Monte Verde for a guided tour of the site and an explanation of its history would be the best way for Dillehay to dispose of any lingering doubts the hard-boiled skeptics might have.
The next book’s important, Haynes said, but one day on the site of Monte Verde would be worth all the words they could write.
Dillehay’s response: Fine. I’ve been inviting people since 1979 to come to the site. Let’s go look at it.
So the wrangling over Monte Verde will last several more years, while we wait for Dillehay and his team of specialists to wrap up volume two, and for Haynes and the other skeptics to hustle funds for the plane tickets to Chile. In the meantime, though, Emory’s molecule hunters are doing their part to hasten the end of the Clovis hegemony.
Wallace and his colleagues inaugurated their mitochondrial DNA studies among Native Americans in the mid-1980s, just about the time excavations at Monte Verde were winding down. Almost immediately their results were extraordinarily encouraging. In analyzing mitochondrial DNA from Arizona’s native Pima, Wallace’s group spotted a mutation that occurs in 1 to 2 percent of Asian people (HincII morph 6 at bp 13259, in the alphabet-soup idiom of genetics). Among the Pima, however, the mutation’s incidence was 20 times higher. That telltale clue does more than affirm the Pima’s shared ancestry with Asians. The mutation is so frequent among them that the Pima must have descended from a small number of Asian immigrants, nearly all of whom carried that mutation in their genetic baggage.
Half a dozen years and a couple of hundred more samples later (including mitochondrial DNA samples from Navajo and Apache in North America, Yucatán Maya in Central America, and Ticuna in South America), Wallace’s team was able to show that all native peoples throughout the Americas share four mitochondrial DNA founding lineages. That’s powerful evidence that the very first Americans were few in number. Genetically speaking, they left their Siberian sisters carrying but a fraction of the mitochondrial DNA gene pool.
If Wallace’s group is right, the genetic clock started ticking the moment those first Americans left Siberia, at least 21,000 years ago by mitochondrial DNA reckoning. As it happens, the first Siberians appear archeologically about then. If they left Siberia soon thereafter, they could have crossed Beringia, moved south before glacial ice became an obstacle, and been at Monte Verde exactly on time (and without having to break any hunter-gatherer speed records).
Of course, Wallace’s group may not be right. There are biologists who think the genetic clock was already ticking back in Siberia--long before the future first Americans departed--and some of those naysayers were at the San Francisco meeting to say so. Ryk Ward, of the University of Utah, and Svante Pääbo, of the University of Munich, have also analyzed mitochondrial DNA variation among Native Americans--in their case, the Nuu- Chah-Nulth of Vancouver Island. Of the 63 individuals they studied, they found 28 separate molecular variants. That’s an astonishingly high rate of molecular diversity in just one tribe. Measured by the mitochondrial DNA clock, the Nuu-Chah-Nulth’s ancestors had to have left Siberia up to 78,000 years ago.
That startling number implies one of two things: either that Americans are of an antiquity inconceivable to all but the most passionate pre-Clovis crusaders--a humbling prospect for archeologists, because it means people were here for tens of thousands of years before Clovis and we’ve utterly failed to detect them--or that the Emory group is dead wrong about how the Americas were colonized.
Ward thinks the latter. The genetic diversity evident in the Nuu- Chah-Nulth, he believes, cannot have evolved in America, at least not until archeologists find some evidence of a 70,000-year-old occupation. Therefore, it must have originated in Asia long before the first Americans left Siberia. In other words, the first emigrants were already genetically diverse at departure, and the mitochondrial DNA clock says nothing about when the populations departed for America. Even if we are able to use molecular data to define the divergence times of a set of lineages, Ward told his Bay Area audience, we cannot state at what time the representative populations themselves split.
But Wallace’s group isn’t giving an inch. We’re looking at the genetic tree very differently, Schurr says. For the Emory researchers, the separate branches on which Asians and Native Americans perch are so genetically distinct that there must have been a branching event somewhere near Siberia. But for Ward, the twigs on the American branch are too numerous for the branching to have happened only here.
Are we back where we started? Not entirely--the archeology is proceeding apace. Dillehay expects to complete the second volume on Monte Verde next year, and he, Haynes, and others are planning to visit the site together. If the skeptics leave converted, then the biologists will have a reason to pick one molecular-clock scenario over another. And the rest of us will have a much clearer picture of the hardy pioneers who long ago slipped across the unmarked border between Siberia and Alaska and found a truly new world.