The First Americans

by James Adovasio

  Junius Bird.

  Today we know that at the time Clovis points appeared in North America, people in South America were already practicing many lifeways. Some chiefly hunted camelids (guanacos and llamas) in the Andean highlands, while others chased horses and other herbivores on the pampas and savannas of Argentina. Others still lived chiefly on shellfish and other marinefoods along the coasts, as shell middens commonly found there indicate. Most people typically lived in small bands, which hunted small game and occasionally larger animals and depended as well on what plants could be gathered in their neighborhood. Some people evidently did actively hunt the Pleistocene megafauna, but it is highly unlikely that they could have survived only on those efforts without bagging smaller game and spending much time gathering fruit, nuts, and other plant foods.

  Fish-tail projectile points from South America, ca. 11,000 to 10,000 B.P.

  The chances are that most such people were merely opportunistic when it came to the large Pleistocene fauna, killing a mastodon only once it got stuck or disabled, for example. The spectacularly rapid environmental changes that were under way from about 14,000 years ago until about 10,000 years ago were sufficient to doom most of the Pleistocene fauna—an extinction event that took a bit longer in the south than in North America for reasons that remain to be explained fully. Horses, for example, plied the pampas after they were extinct in North America. The difference in timing, however, is essentially unimportant.

  The peopling of South America has emerged as far more complex, perhaps the result of several if not many separate entradas through Central America. By 11,000 to 10,000 years ago, the tradition of fluted points called fish-tail points became widespread throughout the continent. These points bear a considerable resemblance to one another wherever they are found. They typically encouraged the Clovis First model until it became clear that they were preceded in most places by a host of other, localized lithic (and nonlithic) traditions. It is quite possible that by the time these points came into fashion, people who had made their lonely way into the far reaches of the continent in relatively small bands were now living in larger and more numerous bands and were therefore more in touch with one another. In such circumstances, a tradition such as fluting could have passed from group to group very quickly.

  The first main question here—essentially forced on archaeologists in South America—is whether or not Clovis people came bursting through the isthmus at Panama and descended on South America with the same speed and lethality with which they supposedly overran North America. In an insight that is also relevant to North America, current work suggests that it was the technology—the idea of the fluted point—that traveled quickly, not the people. It now seems reasonable to suppose that Clovis points struck the fancy of people for some reason, perhaps religious, and the technique of making them spread across the country to people who were already there. The other critical question is whether or not people were indeed there before Clovis. We will review the many attempts to find pre-Clovis evidence before we get to Tom Dillehay, who was the one able to say “Bingo!”

  THEORY FIRST?

  Sherlock Holmes could look at a perfect stranger and from one insignificant feature or two spin a convincing description of the stranger's profession, his past, and even his character. Detectives in fiction can do that, but it is rare that scientists can develop a viable theory on the basis of a single datum. More often than not, successful theories arise as a result of many facts coming to light that need a new explanation. But in the history of archaeology, many practitioners develop a theory and then set out to find the data to support it. Ameghino, of Tertiary Man notoriety, was one such person; impelled by what was more of a dream than the desire for an explanation, he managed to find supporting evidence anywhere, even in the bones of monkeys. It seems silly to us now, but dreams still do inspire archaeological theories, and some dreamers can then persuade themselves that the evidence is supportive.

  And sometimes, as well, logic can lead one astray. In the early days of archaeology in Europe, as the finds delineating the Stone Age occurred, it was perfectly logical to assume that the different styles of toolmaking were a natural evolutionary progression from crude ones to more sophisticated ones—representing the move from savagery to increasing degrees of cultural refinement. It would be taken for granted for a long time that any crude tool was older than a more sophisticated one. This is in a sense much the same as saying that, given the availability of electric tools, no one in the twenty-first century will ever have made or used a manual screwdriver.

  In this way, some archaeologists, impelled by the dream of an extremely ancient peopling of the New World, can be easily persuaded that a broken cobble was clearly a unifacial tool and the work of humans that predated the sophisticates known as Clovis. Even such eminent scholars as Louis Leakey could persuade themselves that rocks broken by natural forces (ecofacts) were unifacial tools (i.e., artifacts). The Tom Lynches and Vance Hayneses have done the field a service in pointing this out when it actually is the case.

  In South American archaeology in the latter part of the twentieth century, many such mistakes were the work, ironically, of North American archaeologists. One, the maverick Scotty MacNeish, we have already met. As long ago as 1976, Scotty devised his own scenario: at some point—maybe 50,000 years ago, maybe even 70,000—humans came across the Bering land bridge and began a slow advance southward, reaching South America by 20,000 years ago. Their tool kit was simple: crudely worked cobbles and very simple unifacial and bifacial implements derived from the ancientchopper/chopping tool tradition of Asia. During what Scotty called Stage 2, they developed better, more sophisticated tools of stone and bone points. Thereafter, some 15,000 years ago in South America and earlier in the north, they came up with spiffy bifacial points and all the rest of the big-game hunter's tool kit. That there are practically no physical remains of any of the earlier artifacts is no surprise, Scotty said. It simply reflects the very small size of the populations in question and the ample time elapsed that has permitted geological forces to cover up what few artifacts were left behind.

  However plausible or implausible that might seem, even Scotty knew that it is difficult to prove the existence of something by the fact of its absence. Over the decades, he assiduously looked for signs that his scenario had in fact occurred. His theory was based on at least one excavation that showed to his satisfaction two or more pre-Clovis stages in the peopling of South America. This was at Pikimachay Cave high in the foothills of the Andes in southern Peru, where he worked in the late sixties into the early seventies. He came across layers that dated between 6,000 and 9,000 years B.P. and, below those, several deeper layers that had been sealed off by a huge rockfall from the cave's ceiling, perhaps the result of an earthquake. Just below the rockfall he found strata that included fossil bones of sloth, horse, and camelids, along with bifaced tools and scrapers. A bone from these layers yielded one date of 14,200 B.P. Below them he found four more layers with the remains of big sloths, a big carnivore, horses, and camelids along with hammerstones, choppers, and other fairly crude implements. There was no charcoal in these layers, but MacNeish obtained two radiocarbon dates from the sloth bones in the neighborhood of 20,000 years B.P., with a large potential variation. From this, MacNeish assumed that small bands of hunters had used the cave from 25,000 to 15,000 years ago when out looking for prey.

  Unfortunately, while virtually all archaeologists agree that the tools in the 14,000-year-old layers are indeed tools, most of them question the validity of the single date, and most believe that the “tools” from the earlier levels are simply not artifacts at all. They are generally made from the same volcanic tuff of which the cave itself consists, a very poor material for shaping into a cutting edge. More than likely they fell from the walls on their own. Some, such as Danièle Lavallée, hold out a wan hope that the14,000 B.P. level might be legitimate. If so, it is the earliest occupation known in South America. I myself find the lithics unconvinc
ing in this regard and the stratigraphy not well enough delineated.

  MORE THEORY FIRST

  Two other North Americans who have long held out for a pre-Clovis habitation of South America are a married couple, Ruth Gruhn and Alan Bryan, both on the faculty at the University of Calgary. Far less audacious but certainly as persistent as MacNeish, this couple has long stood among the most optimistic that pre-Clovis—and certainly non-Clovis—people reached South America long ago, perhaps as long ago as 20,000 or more years. They almost certainly will be vindicated in spite of the long and unbroken string of failed pre-Clovis sites they have dug. Bryan suggests that the early migrants to South America needed to adapt to a variety of diverse environments—forest, open land, and so forth—and therefore most likely used a simple tool kit of unifacial flakes and simple core tools that could beused to make other tools from wood, bone, fiber, and skins. Projectile points used in hunting big game would rarely be found in this scenario, and this has indeed been the case. At some point, some of the migrants would have taken up hunting large animals in the arid savanna regions, using stone projectile points such as those called El Jobo points, which are fairly common along the northern coast. Some time later yet, the notion of fluting arose, probably diffused throughout the south from North America.

  Alan Bryan and Ruth Gruhn.

  This scenario is plausible enough, but regrettably, in decades of excavation in both North and South America, the couple has never found much by way of solid evidence that anything prior to the El Jobo points really existed. Many archaeologists have found what appeared to be the early uni-facial tools that are part of Bryan's scenario, but they have almost entirely been found in open ground, without an archaeological context, stratigraphy, dating, or even a certainty that they are artifacts rather than ecofacts. Perhaps Bryan's most important find came when he reexcavated what had once been an ancient, spring-fed water hole in Venezuela called Taima Taima.

  The site had been dug earlier—in 1962—by the Venezuelan archaeolo-gist José Cruxent, who had produced carbon-14 dates of 14,000 to 12,000 years ago for some fossil animals. The dates had immediately been attacked on technical grounds, and other results at the site seemed confused, so in 1976 Bryan and Gruhn reinvestigated the site to clarify matters. In a layer of gray sand that lay sealed off below a sterile layer, they found the bones of horse, glyptodont, and a young mastodon lying on its left side, some of its parts still articulated, others missing. A couple of ribs showed markings that could have resulted from butchering. Various stone tools were found in the same layer, along with the remains of branches thought to have come from the animal's stomach. But most important, there was a fragment of a lanceolate bifacial point identified as El Jobo lying in the mastodon's pelvic cavity.

  In all, Bryan obtained nineteen carbon-14 dates from the bones, the branches, and charcoal from the gray sand. With one exception, they all fell between 14,800 and 12,800 years B.P., making this what Bryan called “one of the best dated kill-sites in America.” He hypothesized that the hunter had thrust his spear into the anus of the mastodon, which had expired from internal bleeding. One has to wonder if a wounded mastodonwould allow a hunter anywhere near his rectum, let alone in it! Tom Lynch and others were quick to hypothesize that the association of the bones and the stone tools was merely accidental, the result of sinking through water-saturated soils. This was, after all, a spring-fed water hole. Something similar had been noted in a nearby site dug by José Cruxent in the sixties. Critics asked where the tip of the projectile point was, thus insinuating that the broken point had to have sunk through the mud into the pelvis. But the point and the pelvis, along with the other remains and tools, were sealed in by a layer of clay above them that dates to about 10,000 B.P. Even the ever-suspicious Dena Dincauze conceded that both the point and the bones had to have arrived at their final resting place before 10,000 B.P. Supporters of the site have gone on to point out that had this been found in North America and dated to sometime before 11,000 B.P., no one would have questioned the context; many kill sites of mammoths and mastodons were at water holes. If the context is valid—and several Latin American archaeologists, including the Colombian Gerardo Ardila, find no problem with it— the dates still might not be. These could be resolved by revisiting the bones and artifacts and subjecting them to AMS dating, but for now the site remains controversial at best.

  THE PEDRA FURADA CONTROVERSY

  Another highly touted site of pre-Clovis habitation is a massive and dramatic sandstone rockshelter in the semiarid thorn forest in the state of Piaui in northeast Brazil, an area where some two hundred rockshelters have been found. Known as Pedra Furada, the rockshelter in question is 30 feet across and 60 feet deep, with the roof looming about 260 feet above the floor, which itself is more than 60 feet above the valley floor. First discovered in 1973, it was excavated beginning in 1978 by Niède Guidon of the School for the Advanced Study of Social Sciences in Paris. She and her French, Brazilian, and Italian colleagues found fragments of rock allegedly stained with paint—that is, cave paintings—and what appeared to be hearths and stone tools made of quartz and quartzite in a deep level that yielded some forty-six radiocarbon dates ranging from 46,000 years B.P. to 14,300 B.P. In all, this old layer produced around 560 “tools,” some chopping tools andthe likes of burins, but mostly what Guidon and her lithic analyst, Fabia Parenti, said were pieces with “blunt points obtained with two, three or four convergent flakings.” All this underlay a layer full of indisputable artifacts that yielded dates from 10,400 years B.P. to about 6000 B.P.

  The 1993 inspection team visiting Pedra Furada, Brazil.

  The finds were announced with great fanfare, first at a major scientific conference in Orono, Maine, in 1989, then in the scientific press (for example, the British journal Nature), and then in the popular press, with an article in The New York Times giving readers the impression that Brazil might well have been colonized before North America! No sooner did this pre-Clovis, even pre–North America, balloon go up than the naysayers descended. Curiously to some, traitorously to others, Tom Dillehay and I were among them.

  The quartzite tools, we and Dave Meltzer suggested, were almost surely broken rocks that had fallen into the rockshelter from a deposit of quartzite gravels some three hundred feet above. These could have enteredthe rockshelter in a number of ways, and moving water could have scattered them across the entire floor. The ancient “fireplaces” appeared to be nothing more than material blown in from nearby forest fires. Guidon replied that the region had been rain forest at the time, fires are hard to start in rain forests, and anyway, how come the charcoal flecks fetched up only in little areas the size of hearths? To which we naysayers said that charcoal might well have covered the entire floor only to be washed away by rains and floods, except for those flecks that were protected by rocks. Besides, we argued, pollen studies of the surround suggest that the area was drier than a rain forest at the time in question, 30,000 years B.P. During our inspection of the site, I observed no convincing evidence of fireplaces, despite protestations to the contrary. And so the arguments went, leaving the ancient level very much in doubt, chiefly because there really is no way to tell if any of the quartzite “tools” really are man-made and not simply rocks broken by falling.

  General view of the excavation area at Pedra Furada.

  Additional bad news for the pre-Clovis evidence is that because there are no animal bones—which should have been preserved and would be expected if the rockshelter had been a campsite—there are therefore no tell-tale cut marks on bones that would strongly indicate a human presence. Perhaps more telling, there are no tools made of any nonlocal rock, despite the presence of high-grade raw materials in the general study area.

  On the other hand, little doubt exists that humans used the rockshelterat least 11,000 years ago (the upper level) and perhaps somewhat earlier; for example, some valid human artifacts made of chert and other materials that had to have been imported to the shelter were found in
the more recent layer. It remains possible, at least to some investigators, that the rock-shelter could have been used far earlier, if not as a campsite, then perhaps as a quarry where nature had helped the toolmaking process along. In any event, all the pre-Clovis evidence at Pedra Furada remains moot, and I seriously doubt that the central issue of the origin of the alleged quartzite artifacts will ever be successfully settled in favor of an ancient human presence.

  (Meanwhile, it would be a mistake not to point out, by way of a sidelight, that in the Andean highlands as early as 10,000 years ago, the first steps to controlling the growth of certain plants were being taken: wild potatoes native to high altitudes were being moved to lower elevations. Here, then, were some of the first glimmerings of agriculture in this hemisphere—not so much later than such glimmerings as took place in the Near East and China, where modern humans had been for 20,000 and more years.)

  MONTE VERDE

  All this brings us back to the discovery made by my friend Tom Dillehay, the preeminent flak catcher of pre-Clovis archaeology. In 1976, he was teaching at the Southern University of Chile when a student returned from a site with a large mastodon tooth and some other bones, which he handed over to the university museum. The site itself had been discovered in the mid-1970s when some locals, clearing a path for their oxcarts, had cut back the creek's bank, revealing wood and stone artifacts along with the bones of a mastodon. Markings on the bones caught Dillehay's eye—either they were scratches made by animals trampling on the bones, or they were cut marks made during butchering. To find out which, Dillehay went to the site in 1977 and, in the course of a limited excavation, found more bones with what appeared to be butchering marks, along with clay-lined hearths and stone tools unmistakably made by humans. Assuming that the site might date as far back as 10,000 years, into the late Ice Age, he was startled and—like me a couple of years earlier—a bit appalled when radiocarbon dates on the charcoal from the hearths, as well as from bone and wooden implements, came back at more than 12,000 years B.P. Of course Tom was well aware of the Clovis Bar and had seen no reason to question it.