Top banner


The Cutting Edge

Can stone-tool marks on fossils be distinguished from tooth marks?

Pat Shipman

The cutting edge of science is sharp and potentially dangerous. New techniques, new interpretations and new ideas are sometimes responsible for great advances in understanding—or for great mistakes.

2010-11MargShipmanFA.jpgClick to Enlarge ImageWhat would you do, as a responsible scientist, when a new discovery doesn’t jibe with the previous evidence?

First, you’d sit and ponder. What will be the broader implications, if the new evidence is correct? What ideas would be overthrown or contradicted? How different would the new interpretation be?

Then you’d probably do everything you could think of to double-check all of the evidence. Were you careful? Did your ideas go beyond what was supported by the finding? Could some form of confirmation—or refutation—be gleaned from another line of inquiry?

And finally, depending on your own convictions and how bold you are, you would make one of two choices. You could tuck the new evidence into a larger paper, discussing its possible import if confirmed. Or you could simply publish your data and its implications as fully and prominently as possible. In either case, you’d be hoping that later findings would show you were right, because science is essentially self-correcting.

Choosing to Publish

An international team found themselves in this sort of situation, with new evidence contradicting established theories, when surveying for fossils in the Lower Awash Valley of the Dikika region of Ethiopia—and they opted for publication. The Dikika Research Project team is led by Zeresenay Alemseged, curator of anthropology at the California Academy of Science. The team includes geologists, experts in radiometric dating, archaeologists and paleontologists, as well as experts in an obscure specialty called taphonomy, which is one of my own areas of expertise. Taphonomists strive to reconstruct what has happened to a fossil, from the time it was a living organism until it was found, and to deduce what that history tells us about the ancient environment and adaptive niche of the animal, including how it died.

The team has been working in the Lower Awash Valley for 11 years and has made impressive finds. The most spectacular has been the partial skeleton of a young Australopithecus afarensis—the same species as the Lucy skeleton—that has been nicknamed “Selam” or “the Dikika baby.” Selam is 3.3 million years old by geological standards and was about 3 years old at death by human standards.

2010-11MargShipmanFB.jpgClick to Enlarge ImageIn 2009 paleontologist Denis Geraads of the French Centre National de la Recherche Scientifique was collecting bones from the surface of one area and noticed, on one calcaneum (a foot bone) from an antelope, a set of obvious, unusual marks. Based on field inspection, he thought these might be cut marks made by a stone tool, so he brought the bone back to camp and showed it to the rest of the team. They decided to return to that area and several others to inspect every fragment of fossilized bone for similar marks. Eventually they collected four specimens with intriguing marks on their surfaces for further study. There were no identifiable stone tools or sharp stone fragments found in those areas.

Back in the United States, taphonomist Curtis Marean of Arizona State University examined the marks microscopically and compared them with his large collection of bones with known taphonomic histories. He became convinced that the marks on two of the specimens (a rib and a femur, both probably from antelopes) were unambiguous cut marks, some of which graded into percussion marks made by striking the bone with a hammerstone rather than slicing. Two other bones had marks that were too ambiguous to identify. His judgments were confirmed by an advanced graduate student and a former student in taphonomy, who made independent assessments of the marks.

Unfortunately, the two marked specimens were found on the surface and had very little rock adhering to them, so they could not be linked definitely to any particular stratigraphic layer. However, the surface on which they were found is bracketed by two well-dated deposits of volcanic tuff, the higher one at 3.24 million years and the lower one at 3.42 million years. Thus the surface is estimated to be 3.39 million years old.

On August 12 of this year, the team published the find as the cover story in the journal Nature, with the title “Evidence for stone-tool-assisted consumption of animal tissues before 3.39 million years ago at Dikika, Ethiopia.”

The applause from those who accepted their interpretation, and the criticism from those who didn’t, began almost immediately.

Why? The bones the team identified as having cut marks were more than 800,000 years older than the oldest-known type of stone tools, called the Oldowan industry, which come from nearby Gona, Ethiopia, and are 2.6 million years old. There is not a single confirmed stone tool anywhere that is older than the Oldowan tools at the Gona site—and there is not even one possible stone tool from this region of Dikika.

If the team’s interpretation is correct, their discovery “dramatically shifts the known timeframe of a game-changing behavior for our ancestors,” says Alemseged. “We are putting stone tools in the hands of Lucy and her kind.”

Cuts Versus Chews

I have great personal interest in this report on the Dikika fossils. Almost 30 years ago, Rick Potts, now of the Natural History Museum at the Smithsonian Institution, and I were the first to demonstrate (as we published in Nature) that the microscopic appearance of stone-tool cut marks is distinctly different from marks produced by animal teeth.

We had collected large samples of bones with marks of known origin, then “visually inspected” (eyeballed) and documented the marks under a scanning electron microscope. Our known samples included carnivore tooth marks, rodent gnawing, weathering, sedimentary abrasion, digestion, trampling, excavators’ marks, and root marks; the known samples expanded in number and diversity as years passed. By comparing the known pieces with each other and with the unknown ones, we were able to identify diagnostic characteristics that consistently showed up in true cut marks that were lacking in other kinds of marks.

Our paper on identifying surface marks on fossils kicked off a flurry of controversy and excitement. Some colleagues challenged our ability to make these identifications correctly. Others undertook parallel studies examining many different kinds of marks and surface damage, including percussion (hitting a bone with a stone tool to break it open). What once seemed simple is now complex. We now know that several types of marks can closely mimic cut marks, including trampling by animals and the action of crocodile teeth on their prey. Great care must be taken to distinguish such marks from cut marks.

In light of these findings, it’s no wonder that the startling interpretation of these two Dikika bones caused a major controversy to erupt over the identification of these marks.

Tools of Opportunity

So, how can there be cut marks prior to the existence of stone tools for cutting? There are a few possibilities, some more likely than others.

2010-11MargShipmanFC.jpgClick to Enlarge ImageFirst, maybe there are stone tools going back to 3.4 million years ago, but they have been missed. This is incredibly unlikely. This explanation would mean that all of the researchers on all of the expeditions that have combed the badlands of Ethiopia, Kenya and Tanzania for the past 40 years looking for fossils and early stone tools have been too incompetent to recognize stone tools in layers older than 2.6 million years. The timing of the invention of stone tools is a critical issue in paleoanthropology, and the oldest-known type of tools has been extensively studied. Furthermore, stone tools are sturdy objects not easily destroyed by most natural processes (unlike bones), so they don’t just vanish. It is completely unbelievable that Oldowan tools are scattered over the 3.4-million-year-old landscape and have been overlooked.

Second, the tools might be present but unrecognized because they are not deliberately flaked like the Oldowan tools, but naturally broken sharp stone fragments, or “expediency tools.” Marean favors this hypothesis.

“These will not look like Oldowan stone tools,” he says, “so the search image brought to the field will be inappropriate for finding this material.”

Alemseged also leans toward this possibility, but he cautions that if naturally occurring sharp objects were used as tools, without further modification, they will be very nearly invisible archaeologically. The only hope of identifying such an expediency tool would be if it had been excavated from fine-grained sediments and if it was of a rock type that does not naturally occur in that area (a so-called manuport). Manuports have not yet been recorded in any of the fossil sites older than 2.6 million years.

Third, the marks might be misidentified. This is the view of Tim White of the University of California at Berkeley—the leader of another team that works in an adjacent area of Ethiopia in this time period—who says that too much is being made of too little. “Extraordinary claims require extraordinary evidence,” he observes. “This is not perfectly good evidence that is being dismissed because its implications are startling. It is perfectly ambiguous evidence. This claim is based on about a dozen gouges and scratches on two small bone fragments with no associated context, from the surface of a site that is loaded with crocodiles.“

Because crocodiles chomp on their prey, leaving lots of marks with varied characteristics on the animals’ bones, White thinks crocodiles are probably responsible for the marked Dikika bones. The marks from Dikika are described as being unusually heavy and deep compared to most butchery marks, indicaing that whatever made them used a lot of force, as crocodiles do.

2010-11MargShipmanFD.jpgClick to Enlarge ImageJackson Njau, now a postdoctoral fellow at the University of California at Berkeley, did the most extensive study of crocodile marks on bones as part of his Ph.D. thesis. He is not only a lab analyst but is also the coleader of the Tanzanian International Paleoanthropological Research Project. Njau hasn’t yet seen the original Dikika specimens. But judging from the published photographs of the marked bones, he says:

The random patterning of the marks on those two specimens match those produced by crocodile action (under experimental conditions) or other unknown processes. Crocodile feeding techniques include twirling, twisting and hitting bones on the ground or rocks. Combined with the animal’s dental morphology, these techniques inflict marks of many different types, shapes and varying intensities, some of which look like cut or percussion marks, or like scoring marks that are curved similarly to ones documented on one of the Dikika specimens. One of these curved scoring marks is labeled J {see image at right}, which the authors consider to have been made by an unidentifiable carnivore.

In response to these challenges, Marean stands firm. He has examined the crocodile toothmarks from Njau’s controlled experiments as well as many other types of marks, and is still confident that his identification is correct.

“Most of the marks have features that indicate without doubt that they were inflicted by stone tools,” Marean asserts. “The range of actions that created the marks includes cutting and scraping for the removal of flesh, and percussion on the femur for breaking it to access marrow.” Crocodiles are not the only explanation for deep and heavy marks; hominins clumsily wielding heavy stone tools might also produce such damage.

Sufficient Data?

Is there enough evidence to support the stunning claim by the Dikika team? Obviously the team thinks so, and so do the scientists who reviewed their paper for Nature.

Sileshi Semaw of the Stone Age Institute in Indiana doesn’t think so. He is the leader of the ongoing research team that found the oldest known Oldowan tools and bones with cut marks at Gona. When his team published their discovery, they had excavated thousands of stone artifacts and many fossilized bone fragments from well-dated, fine-grained sediments, as well as collected surface samples.

Semaw asks, “What is the Dikika evidence? Two modified bones with possible cut marks from the surface, and not a single stone flake nor a single hominin-modified stone! Such a major leap in ancestral hominin tool-use behavior should have been supported by strong evidence, but the Dikika researchers are making a huge claim based on very meager data.”

Others challenge the interpretation of the Dikika bones because of the lack of context. The bones could not be linked by an adhering matrix (in other words, stone) to the particular rock layer from which they must have originated. Because it is the rocks that are dated, not the bones, the antiquity of the bones depends on their being matched to either a rock layer or, through excavation, to specimens still in place in the ground.

Speaking for the team, Alemseged responds to this objection: “We do not have any doubt that these bones are older than 3.24 million years and very likely 3.39 million years old. The juvenile skeleton Selam was discovered only 222 meters away, so intensive attention was paid to documenting the geologic history of that particular area. Almost all fossils from the Lucy site, Hadar and Dikika were dated using the same methods.” Of course, Alemseged adds, excavating additional pieces that confirm these finds would be great, and he hopes to find such material in his next field season.

But he doesn’t have the evidence yet, his critics retort.

The lines are clearly drawn.

The Dikika team has conducted all of the studies they can to support the identification and the antiquity of the marks. They chose to make a prominent announcement of their findings because they are confident that their work has been done correctly and because the implications of finding pre-Oldowan cut marks are huge.

If their interpretation is correct, it reshapes our understanding of the life of Australopithecus afarensis. This species may well have had more meat and fat in their diet than we supposed, but the addition of these important foods did not cause a rapid physiological change in Australopithecus afarensis that can be seen in the fossil record. Were the first tools used too rarely to have a dramatic impact on the species right away? Possibly.

Also, the odd pairing of bone with cut marks and an utter lack of recognizable tools suggests how tool use might have begun: opportunistically and infrequently with expediency tools. Maybe simply picking up handy objects showed australopithecines that what they really needed was a cutting edge. If so, then it took these early hominins 800,000 years to figure out how to make and use tools regularly.

In contrast, the skeptics feel the announcement was premature. They argue that the team should have waited for more, better and less ambiguous evidence before making such an enormous claim.

If stone tools were invented 2.6 million years ago and not much earlier, their first appearance coincides with the presence of Australopithecus garhi and is close in time to the oldest known occurrence of A. garhi’s possible descendent, an early species of the genus Homo. In this reading of the record, the invention of stone tools provoked a dietary change—an increase in carnivory—that led to a rapid increase in body and brain size seen in Homo.

Is the identification of marks on bone surfaces so reliable that a few cut marks can be taken as proof of stone tool use, even in the absence of recognizable tools?

Do we know enough about taphonomic agents that cause marks that closely mimic stone-tool cut marks and how to distinguish among them?

How long did it take to invent stone tools? Did using naturally occurring sharp stones give our ancestors the idea that they could make sharp objects out of stones?

And how long would a dietary and ecological shift—from herbivore to carnivore, from prey to predator—take to produce dramatic changes in the biology of our ancestors?

In this intriguing case, the truth is balanced precariously on a knife’s edge.



Bottom Banner