September 1999


The obsidian clock

ORNL method chips away barriers to dating prehistoric human artifacts

An obsidian arrowhead: How old really?
Obsidian—volcanic glass—was a material of choice for prehistoric tool and weapon makers. Determining the age of the objects they left behind has always been a challenge for archaeologists. Several ORNL and University of Tennessee researchers are developing an analytical method for obsidian that has the potential of ensuring a precise means of dating archaeological sites.

Prehistoric arrowheads, knives and spear points in many parts of the world, including Mexico and Central America, were made of obsidian flaked into very sharp cutting tools. Spanish conquistadors learned, to their annoyance, that a single blow from a good Aztec obsidian-edged sword could behead a horse. In modern times, obsidian has even been used for surgical scalpel blades.

Unfortunately, developing a reliable method of interpreting the “obsidian clock” has defied all previous efforts. Now the team of scientists—using secondary ionization mass spectrometry (SIMS), experimental calibration and sophisticated modeling methods—is closing in on a technique to accurately tell scientists when an obsidian tool was crafted.

The group, consisting of Lee Riciputi, Larry Anovitz and Dave Cole of the Chemical and Analytical Sciences Division and the UT Anthropology Department’s Mike Elam, report encouraging results with their technique, called ODDSIMS, for obsidian diffusion dating by SIMS. They are also learning surprising new things about how water diffuses into glass, which could be important to emerging waste management technologies.

Obsidian tools offer archaeologists a wealth of information on ancient technologies, economies and even religious practices. In the Americas, obsidian tools found at the site of Santa Isabel Ixtapan, Mexico, have been associated with a mammoth skeleton. Radiocarbon dating suggests that the tools may be as old as 10,000 years. Obsidian tools found in Ethiopia might be as old as 180,000 years.

Might be, because these dates are based on materials found near the artifacts and not the artifacts themselves, a fact that leaves the dates open to question. Years ago a U.S. Geological Survey scientist, Irving Friedman, in a method he called obsidian hydration dating, attempted to use optical methods to determine how far water had been absorbed into the surfaces of an obsidian artifact. By gauging the depth of hydration, coupled with simple diffusion modeling, he estimated when the artifact was made.

“A whole cottage industry developed around hydration dating, but with very marginal success,” says Cole. “Dates often conflicted with other findings or varied by thousands of years. Sometimes it caused significant controversy.”

The research team has discovered the reasons for the discrepancies. “The obsidian hydration dating theory was that you could cut thin sections of the tools and measure the width of the hydration rim accurately with an optical microscope,” says Anovitz. “We suspected there were problems both with the method of measuring rim width, which assumes that the interface between the rim and unhydrated core is sharply defined, and with the simple diffusion equations that were used to model how the rim grows with time. Diffusion alters the glass, making things complicated.

“With an optical microscope”, Anovitz continues, “you see a little white line, which was assumed to be the boundary between hydration and nonhydration. We found that the white line was a Becke line, typical of a boundary between two different materials. Becke lines appear to move as function of microscope focus, and this leads to large errors.”

Riciputi, Anovitz, Elam and Cole, backed by DOE Basic Energy Sciences and National Science Foundation funding, applied SIMS to the problem. “Our big advantage is in instrumentation,” says Riciputi. “With SIMS, you slowly sputter into the obsidian sample with an ion beam and measure how the composition changes as a function of depth. It offers a precise way of measuring the distribution of the hydrogen and the depth of penetration.”

Taking the new technique into the field, however, presents a number of problems. Environmental variables such as temperature and humidity affect the rate of water absorption into the glass, and controls present in laboratory experiments aren’t available at most archaeological sites. Not knowing local climatic and environmental conditions, and how they changed over hundreds or thousands of years, was one of the major limiting problems of the traditional hydration dating method.

The ORNL-UT team’s approach to the problem is to make two types of measurements: one on samples hydrated under precisely controlled experimental conditions and the other on obsidian artifacts found near and probably associated with materials, typically charcoal, that provide carbon-14 dates.

Collaborating with the Mexican National Institute of Anthropology and History, the group collected obsidian artifacts from a number of sites and research facilities. They focused on artifacts made from pachuca, a distinctive green obsidian that was mined from an extinct volcano called Sierra de las Navajas (“Mountain of the Knives”). Pachuca was highly prized and widely traded and, therefore, is abundant in much of Mexico and Central America.

The ORNL group has initially found that ODDSIMS estimates line up well with obsidian artifacts that have been dated with some confidence. “Preliminary findings have been very predictable and regular on samples from the Mexican Basin that have good C-14 dates; extremely good agreement with SIMS,” says Riciputi. “It means back to square one on a lot of dating interpretations.”

UT’s Elam adds, “Some established chronological systems may go out the window if the ODDSIMS method continues to provide consistent and reproducible results.”

An obsidian spearpoint
Riciputi believes interest in their research will go beyond archaeologists who want to know whether their artifacts are 5,000 or 8,000 years old. “We’re learning that the mechanisms of hydration are very different than previously thought. Hydrogen appears to be moving into the glass without oxygen uptake, which is a surprise. We’re getting a better insight into how glass hydrates, and that’s potentially very important to materials science and geology as well as archaeology. “DOE is interested in the effects of hydration on glasses used for encasing waste. These glasses have compositions that are relatively similar to those of obsidian, and we can look at the processes by which industrial glasses deteriorate as well.”

That’s a long technical leap from the prehistoric American hunched over a block of obsidian, flaking off spear points some 500 to 10,000 years ago. With more progress on the ODDSIMS dating technique, we may finally know exactly how long ago these ancient civilizations and cultures flourished.—B.C.