Sometime before 1650, small scattered tribes on the Hawaiian Islands apparently coalesced into a centralized society. Archaeologists believe this partly because many temples built at approximately that time on Maui are similar to one another, suggesting a single ruler. But exactly when those temples were built and just how rapidly the tribes assimilated has been an impenetrable mystery until very recently.
The problem stymied Patrick V. Kirch, an archaeologist at the University of California at Berkeley, for years because the standard technique of discovering how old objects are, radiocarbon dating, turns out to be rather imprecise for short time frames, in this case a 40- to 250-year range. So Kirch turned to geologists Warren D. Sharp and Kenneth R. Ludwig at the nearby Berkeley Geochronology Center for a better technique. Because the temples incorporate pristine examples of local corals, Kirch and Sharp tried a method often used by geologists called uranium-series dating.
"Corals are ideal materials for this kind of dating," says Sharp. "They take in uranium from the seawater, but not thorium." Therefore, the coral would have stopped taking in uranium at the moment it was pulled from the sea. The uranium-238 in the coral eventually decays into the more stable thorium-230, at a known rate. Sharp used a mass spectrometer to measure the ratio of uranium to thorium in the coral and then, with a single age equation, calculated when the coral first grew on a reef.
All the coral fell in a range between 1580 and 1640. Some fell within a tight time span of just 30 years. The dates indicate that a single ruler wrought a major social transformation in one generation.
U-series dating is hardly new and has been used on corals before, but usually for the purpose of dating sea-level changes, glaciation, and world climate shifts over periods of 100,000 years. The use by Kirch and Sharp opens up a new world of accuracy for archaeologists trying to understand human development in the deep past.
Nevertheless, all dating methods have some constraints. Dendrochronology, for example, the dating of sites with tree rings, requires the survival of ancient trees nearby. Paleomagnetism requires that specific minerals in the sample retain their magnetic polarity after Earth's magnetic field flip-flops, as it does about every 300,000 years. Dating by radioactive decay (over a long enough period of time, uranium decays into lead) requires the presence of radioactive elements in the sample. Carbon-14 dating requires organic materials like bone or wood. Argon-40/argon-39 dating requires potassium or argon in the sample.
"Argon-argon is the most powerful and widely applicable method for dating," says geologist Paul Renne, director of the Berkeley Geochronology Center. "It can work from about 2,000 years ago all the way back to the early solar system. It's our bread and butter." It is how scientists know that Earth is 4.5 billion years old and not 6,004 years, as some biblical literalists believe—or 5,765 years old, as some Orthodox rabbis believe. It is how we know that the earliest hominids walked the earth more than 6 million years ago and how we know the first modern humans lived more than 160,000 years ago.
The laboratory techniques for dating almost anything by radioactive decay, whether a hominid fossil or a coral horn, are very similar. Scrupulously cleaned and refined samples are placed in one end of a mass spectrometer, then heated either by means of a CO2 laser or an electrified rhenium filament until the sample turns into a gas. The isotopes in the sample are formed into an ion beam that is detected at the far end of the mass spectrometer, where the number of isotopes are counted. The ratio between quantities of different isotopes provides the data for an equation that calculates the age.
As techniques and methods have improved, scientists are practicing a lot of revisionism. Take, for instance, the great Permian-Triassic extinction, the largest obliteration of life in the history of our planet. A catastrophic event, possibly an asteroid impact, killed all but about 10 percent of living things. Until very recently, the date of the extinction could not be nailed down precisely enough to link the extinction to any known catastrophe. Then in 2004, geochronologists at Berkeley—Ludwig, Renne, and Roland Mundil, working with Australian geologist Ian Metcalfe—improved the pretreatment of volcanic zircon samples gathered in China. They were able to use pieces as small as a single crystal, greatly reducing the chance of error by contamination, and pinpointed the extinction to 252.6 million years ago, plus or minus 2 million years. That date coincides with the Siberian flood volcanism—a continent-size flow of lava pouring out on Earth's surface from below its crust. Linking the year to the event was impossible prior to the recent breakthrough in the processing of zircon samples.
Recent advances in uranium-series dating make the procedure so accurate "we could distinguish between materials dating from the first or second Roosevelt administrations," says Sharp. "Relatively few types of materials have been extensively used, but there are a lot of candidates, and one of the things we are trying to do is expand the range of materials that U-series can be applied to."
Among the candidate materials is pedogenic carbonate, carbonate material that forms in soils over time. Another is travertine, or spring-deposited carbonate. "If you can find those in a place where you want to understand how the land surface has changed with time," Sharp says, "they can be very useful, datable markers." Sharp, for example, is part of a team studying what he calls "the incision history of the Grand Canyon." Ancient river gravels embedded in travertine high on the canyon walls can be dated to reveal "the time when the river was at that old elevation."
At Olorgesailie in Kenya, the walls of a basin in the East African Rift, where early humanlike creatures evolved, span nearly a million years of time. Some of the strata contain abundant fossils and stone tools. They also contain deposits of opaline, a material that's very promising for uranium-series dating. "These are actually siliceous replacements of roots and stems of plants like petrified wood," Sharp says. "They look like thick white pencils." The samples come from an area that has already been dated by argon-argon methods. If Sharp gets corresponding dates via the uranium-series method, "then we can carry it elsewhere, where we don't have volcanic ashes, like in South Africa, where there are tons of important early human fossils."
In the meantime, Sharp and Kirch plan to continue their work on coral-bearing temples on islands in the Pacific. They will sample more areas on Maui and adjacent islands, and Sharp adds, "There are other areas around Tahiti where Polynesians used big fan corals to decorate the facades of the temples. We plan to go sample those and see if we can make the U-series dating work there too."