Cornell Archeologists Run Rings Around Others’ Dating Methods
In a basement laboratory at Cornell University, scientists are using chunks of wood to revolutionize the study of ancient history in the eastern Mediterranean.
They’re gathering tree rings for every year back to 7500 B.C. Every single year.
The payoff? A way to determine the age of wood or charcoal from archeological sites, right to the year that the tree toppled.
It won’t always be that precise. But for archeologists who are now delighted if they can date artifacts at give-or-take 50 years, the prospect is astounding.
Some accepted dates for historical events might shift by as much as 100 years. Movements like that--and the very precision of the dating technique--could overturn ideas about trade, migrations, wars and whether particular trends were evolutions or revolutions.
“It’s going to completely change our understanding of the chronological sequences of the Aegean and Near East,” said one observer, archeologist Gil Stein of Northwestern University.
“We will really know when these sites were occupied, when these styles of pottery were used, when these cities arose, when these areas were abandoned--very important things that up until now have been very much subject to debate.”
And all this for a time when pharaohs ruled, the world’s first major cities arose and collapsed, civilizations took root in Egypt and Mesopotamia, and classical Greece and Rome appeared and disintegrated.
At the center of the tree-ring effort is Peter Ian Kuniholm, 60, a husky classical archeologist who minces no words. He can dismiss a misguided analysis as “sheer howling nonsense” and a clutch of Byzantine samples as “major site, cruddy wood.”
Kuniholm began the work 25 years ago, and he knows his stuff. During a recent chat in his lab he spoke in encyclopedic detail. He repeatedly bolted from his chair in mid-conversation to fetch a phrase from the middle of an inch-thick file, a chunk of poplar from the Parthenon or purchasing records--in drachmas--of an ancient temple treasurer.
He’s not the first person to use tree rings for establishing dates. Such “dendrochronology” has been applied to thousands of years of history in Europe and the American Southwest, for example. But some colleagues initially scoffed at Kuniholm’s attempt to apply it to the eastern Mediterranean, which is not exactly covered with forests.
“I kept saying to him in the early days, ‘Well, Peter, where are you going to find wood?’ [It seemed] like counting blades of grass in the Sahara,” recalled archeologist Curtis Runnels of Boston University. “He had faith where the rest of us had doubts.”
Site by site, chunk by chunk, some 30,000 wood samples have been collected from about 600 sites in 18 countries, including Italy, Greece, Yugoslavia, Bulgaria, Turkey, Cyprus, northern Syria and Lebanon. The samples have included a piece of an ancient Turkish coffin lid, boxwood from a first-century-B.C. Roman boat and scorched timbers from a carpenter’s workshop--smothered by the eruption of Mt. Vesuvius in 79 A.D.
Kuniholm’s lab has made more than 9 million microscope measurements of tree rings, some finer than a human hair. Even a finger-sized piece of wood can hold enough rings to be useful, and a sample the size of a pizza slice can present a bounty of 750 rings.
Here’s how the analysis works:
Trees add a ring each year, sometimes thin, sometimes thick, depending on growing conditions. “When I have cool wet summers my trees grow like gangbusters,” Kuniholm said.
If a series of years provides a sequence of growing conditions like poor, poorer, average, great, poor, average, those years should produce a sequence of rings something like thin, thinner, average, thick, thin, average. And if two trees share the same climate, that imprint should be left on both of them, even if one is young and the other quite old.
So when Kuniholm and his assistants analyze a piece of wood, they see if its tree ring sequence matches one already in the database. If there’s a close enough match between two long sequences --generally at least 100 rings long--the two trees were alive during the same years. And if the new wood piece has bark, Kuniholm knows it was cut in the year represented by the last ring.
But what year was that? Kuniholm can tell, so long as the tree was felled after 360 A.D. His database reaches back to that date, starting with ring patterns from contemporary trees--which can be dated with certainty--and extended by sequences from older wood samples. When a sample matches the oldest rings in the database and provides still older rings, the older rings can be dated and the database extended.
It’s this process that Kuniholm hopes to push all the way back to 7500 B.C.
Even now, he says, he can date wood from 2660 B.C to 627 B.C., if he’s right in asserting that a volcano erupted on the Aegean island of Thera in 1627 B.C. and that a particularly thick ring in his samples corresponds to that event.
Tree rings can’t date every piece of wood precisely. If there’s no bark or other unambiguous sign that the last ring marked the death of the tree, Kuniholm can say only that the tree was living during certain years, but not when it was chopped down. And even if analysis does reveal a precise date on a beam from a building, that doesn’t always say how old the building is. The wood might have been reused from some previous building, or newly cut to repair an old building.
All told, the lab database includes tree ring segments covering about 6,500 years. But most segments are floating in time, not precisely datable until the database reaches back far enough to include them.
So, to get more wood, Kuniholm and assistants typically drive 10,000 miles a summer, visiting digs and museums, making their faces known, handing out scientific articles and brochures in eight languages--anything to locate samples and spread the word of their quest.
“Everybody is rushing to bring him wood,” said anthropologist Mary Voigt of the College of William & Mary. “It’s clear he has a long enough sequence that we have some confidence in the dates.”
She hopes Kuniholm can help her analyze ruins at Gordion, the Turkish home of King Midas. That’s where Kuniholm began his project, out of frustration with the imprecision of other methods.
Decades later, it’s still fun.
He once visited a stone castle at Pythion, on the border of Greece and Turkey. The archeologist there could say only that it was built in the first half of the 14th century, based on pottery, coins and written histories.
Kuniholm rappelled down its face and fished out some rotted timber, which the ancient builders had used as scaffolding.
“The wood was all cut in 1331,” Kuniholm said the other day with obvious satisfaction. “He can give me within 50 years. I can tell him the year.”
