Rethinking the Fall of Easter Island
New evidence points to an alternative explanation for a civilization's collapse
Timing Is Everything
I did not expect when I first visited Rapa Nui, in May 2000, that I would end up questioning what I thought I knew about the island's past. Indeed, my initial trip to the island was primarily as a tourist, not as an archaeologist. But while I was there, I ran into Sergio Rapu, the first native Rapanui governor of the island and a former student of mine—Rapu had studied archaeology as a graduate student at the University of Hawaii. He invited me to do research on Rapa Nui.
Before that point, the thought of doing work there had not really crossed my mind. I had been planning to lead a continuing archaeological field school in Fiji later that summer, but when a violent coup d'état began there that May, my enthusiasm waned. Rapa Nui now seemed an increasingly appealing place to hold the field school sessions.
I anticipated that the work of my students and I, which we began in August 2000, would help put the finishing touches on a well-established story. But as I began to review archaeological survey data, studies of the moai and evidence for environmental changes, I realized that there were a number of gaps in what was known about Rapa Nui, and I grew increasingly skeptical of everything said about the island's prehistory.
Over the next few years, my students and I conducted field work for one or two months each year. My colleague Carl P. Lipo, an archaeologist at California State University, Long Beach, joined the effort and introduced me to the potential of satellite imagery, which we used to explore features such as the ancient roads on which the Rapanui transported the moai from the quarry at Rano Raraku to every corner of the island. Following road alignments also led to the documentation of a number of moai that were previously unrecorded.
In 2004, we began new excavations at a locale called Anakena. This white sand beach would have been the most inviting spot for the first colonists to land their boats (the shore in other places is for the most part made up of cliffs or rocky crags). Hence most anthropologists suspect the areas around Anakena to be the site of the earliest settlements. We intended to study subsistence and environmental change, not basic chronology, which we assumed was already settled.
We dug through sand whose beautifully undisturbed stratification proved to be an archaeologist's dream. The integrity of the layers would be helpful in determining when things happened, both in an absolute sense and relative to other events. But the excavations were not easy. The sand at Anakena is soft and unconsolidated. As we dug down a few meters, the pits became increasingly dangerous. Horses trotting by on the beach would cause nerve-wracking vibrations in the layers of sand; we worried someone would be buried alive in the pit.
Finally, we reached the bottom of the sand. In the top 3 to 5 centimeters of the underlying clay we unearthed abundant charcoal fragments (indicating the use of fire), bones (including those of the Polynesian rat, a species that arrived with the colonists) and flaked obsidian shards (a clear sign of human handiwork). Below, we found nothing suggesting human activity. Instead, the ancient clay was riddled with irregular voids—places where the soil had once molded itself around the roots of the long-gone Jubaea palm tree.
We had clearly found the layer with the earliest human-related materials at Anakena, and assuming that Anakena was likely the location of the first settlements on the island, we were in an excellent position to ascertain the timing of the initial colonization. So I was disappointed when I received an e-mail from the lab that did the radiocarbon dating on these samples. There seemed to be a mistake. The oldest dates were only about 800 years old, implying that occupation began around 1200 A.D. The dates from layers closer to the surface were progressively younger, which is inconsistent with the possibility that somehow our samples were contaminated with modern carbon. There was really no way to explain these numbers, at least not within the conventional model of Rapa Nui's development. I put aside the discouraging message for the moment and decided to try to figure out later what had gone wrong.
When the hardcopy of the report arrived a couple of weeks later, I examined the data again. The closer I looked, the more it seemed that our results were not the problem. I spoke with my friend and colleague Atholl Anderson of the Australian National University. He had done a careful screening of radiocarbon dates from New Zealand and concluded that the first settlers arrived there around 1200 A.D., several hundred years later than archaeologists had previously believed. The reaction to his ideas was initially quite cool, but time and additional evidence have proved him correct. Having had this experience, Anderson advised me to keep an open mind and to trust my data more than any preconceptions.
In 2005, Lipo and I returned with our students to Anakena and located another part of the dune where the deepest layers containing vestiges of occupation would be easier to expose. We uncovered a large area of the clay beneath the sand and took samples for more radiocarbon dating. The two additional dates from the basal layer were completely consistent with our earlier results.
Was the conventional chronology just plain wrong? Lipo and I took a closer look at the evidence for earlier human settlement. We evaluated 45 previously published radiocarbon dates indicating human presence more than 750 years ago using a "chronometric hygiene" protocol. We rejected dates measured from unreliable sources, such as marine organisms, which require corrections for the older carbon from the marine environment. We also rejected single dates that were not confirmed by a second date from the same archaeological context. Using only paired dates helps ensure the reliability of the data. Our standards were more inclusive than those used by Anderson in his study of New Zealand, but we still were left with only nine acceptable dates. With this culling, the evidence for first occupation around 800 A.D. simply fell apart.
Although our results did not fit with the accepted settlement date for Rapa Nui, they did match the chronology for deforestation that Orliac, Mann, and Mieth and Bork had worked out. You merely have to abandon the idea that a small or transient population endured for centuries. Instead, we posit that from the start the environmental impact was widespread.
The notion that humans did not arrive on Rapa Nui until about 1200 A.D. was not the only thing causing me to rethink my assumptions about the island. Research on other Pacific islands provides a compelling parallel and a possible explanation for the damage done to Rapa Nui's environment.