Rock Dates Thrown Into Doubt, Researcher Under Fire

See allHide authors and affiliations

Science  26 Jun 1998:
Vol. 280, Issue 5372, pp. 2041-2042
DOI: 10.1126/science.280.5372.2041

When amateur rock art enthusiast Ekkehart Malotki discovered ghoulish human and animal carvings in isolated sandstone cliffs in the Arizona desert in 1993, he suspected that they were among the oldest he had ever found. So he was elated in 1996 when Ronald Dorn, a prominent geoscientist at Arizona State University (ASU) in Tempe, agreed to date the carvings. Using a technique he developed, based on measuring tiny amounts of carbon trapped beneath rock varnish, Dorn concluded that the carvings were perhaps 4800 years old, making them some of Arizona's oldest scientifically dated petroglyphs. An archaeology journal accepted a paper by the two men announcing the discovery. “I was honored that this world-renowned rock art dater had done this for me,” recalls Malotki, a linguist at Northern Arizona University in Flagstaff.

Now, however, Malotki says his elation has given way to anguish, and the age of his find—and of hundreds of petroglyphs, stone tools, and rock surfaces around the world—is in doubt. On page 2132, Malotki and seven co-authors, including several prominent geologists, challenge the validity of Dorn's technique. They report that rock samples Dorn processed contain microscopic granules of coal and charcoal, which, they say, render the dating results meaningless. If so, then many of Dorn's results—with implications for such debates as the peopling of the Americas—may be in jeopardy.

Center of controversy.

Charges and countercharges swirl around the dating of these Arizona petroglyphs.


The findings also raise the specter of scientific misconduct by Dorn, who says the researchers are accusing him of manipulating samples. The Science Technical Comment, whose lead author is geoscientist Warren Beck of the University of Arizona, Tucson, stops short of accusing Dorn of fraud. “We do not intend to use the ‘f’ word with any reporters,” says Beck, who, like Dorn, has consulted with attorneys in this matter. But the paper reports that the carbon granules do not appear in samples processed by other researchers. Those close to the case say that the National Science Foundation (NSF) and ASU are reviewing the possibility of misconduct.

In a lengthy response, Dorn acknowledges that his technique is flawed and produces “ambiguous” results. But he writes that the suggestion of tampering is “utterly false” and that the carbon granules are naturally occurring; he offers detailed technical explanations of why Beck and colleagues could not reproduce his results. This “controversy not only threatens my reputation but sidetracks a very important body of scientific research,” he told Science.

The controversial technique relies on accelerator mass spectrometry (AMS) radiocarbon analysis, one of several rock-surface dating methods pioneered since the 1980s by Dorn, a tenured professor who earned his Ph.D. at the University of California, Los Angeles (UCLA), and was hired by ASU in 1988. The method assumes that microscopic quantities of carbon-rich organic material, such as the remains of plants or bacteria growing on the rock's surface, become trapped beneath a thin layer of natural varnish. Although some scientists doubt the trapped material can be accurately dated, Dorn has argued—beginning with a 1986 Science paper—that it holds measurable quantities of radioactive carbon-14, which decays at a known rate. To date a rock, Dorn scraped organic material from beneath the varnish and extracted it with acid, then sent samples to one of the dozen AMS facilities around the world.

In June 1996, Dorn sent samples from Malotki's petroglyphs to Beck's AMS lab. Noticing unusual color variations, Beck says he examined the samples under a microscope and discovered that in addition to mineral grains, they contained two kinds of black granules. One type was identified by an independent expert as coal and the other appears to be charcoal, as the Technical Comment explains. Using radiocarbon dating, Beck and colleagues found that the coal averaged about 28,000 years old and the charcoal 4000 years. The different ages, says Beck, mean that the sample “will not yield a reliable radiocarbon age.”

Mystified by how grains of such disparate ages were incorporated into the same thin slice of rock surface, Beck and Malotki collected and processed new samples from Dorn's original site. They report that none of the new samples contained the carbon granules. Concerned, Malotki says he immediately withdrew the paper on his treasured find.

Meanwhile, similar observations had been made by geochemist Wallace Broecker and his colleague Tanzhuo Liu at Columbia University's Lamont-Doherty Earth Observatory in Palisades, New York, who wanted to use rock varnish to date geologic events. According to Broecker, samples Dorn processed for them contained the granules, but samples they prepared themselves did not. Liu, a former graduate student of Dorn's, then returned to ASU in August 1996 especially to receive training in sample preparation from Dorn. Samples that Liu prepared in Dorn's lab contained the granules; those he processed at Columbia did not, says Broecker. Beck says he has examined more than 50 other samples prepared by Dorn and other investigators since the mid-1980s and found the same pattern: “The bottom line is that we found one or both of these materials in virtually every sample that was processed by Ron Dorn, and [we were] unable to find them in any sample that anyone else had processed,” he told Science.

Dorn counters, however, that faulty technique explains why Beck, Broecker, and Liu couldn't find the grains. He adds that the presence of two different carbon grains on some, but not all, rock surfaces is well documented. “I and other scientists have found both of these materials associated with rock varnish samples from a wide variety of locations, and the literature abounds with those findings,” he told Science. He cites two other researchers who have found the grains in independently processed samples: a former graduate student in his lab, and geologist Ramon Arrowsmith, an associate professor at ASU. Arrowsmith says his unpublished study, designed to look for the two kinds of grains, did reveal them in varnished rock samples from Arizona. He did not date the grains and says “it is a bit of a mystery” how they got into the rock.

As further argument against the tampering charge, Dorn claims that doctoring samples with different-aged contaminants would make it difficult to consistently bias results and “would be insane” if cheating were his intent. Finally, he cites a 1995 double-blind test in which he and geologist Alan Watchman of James Cook University in Townsville, Australia, independently dated Portuguese rock art using varnish-trapped organics—and got similar results. However, they concluded that the art was about 4000 years old—and most archaeologists put it at about 18,000 years old. Dorn says that the Portuguese results, and his realization that his samples held different-aged materials, made him recognize “fatal flaws” in his method.

But Dorn and his challengers differ on the timing of this admission. Dorn says he began to have concerns about the method as early as mid-1995; David Whiteley, a colleague from Dorn's UCLA days and now a consulting archaeologist with W & S Consultants in Simi Valley, California, confirms that Dorn discussed the problems with him then. “My impression was that he was attempting to broadcast his doubts widely,” says Whiteley. In mid-1996, Dorn says he sent a note about the method's flaws to Nature; after it was rejected, he published an article in the December 1996 issue of La Pintura, the newsletter of the American Rock Art Association. It details “critical mistakes on the radiocarbon dating of organic matter associated with rock varnish. Enough data has accumulated to put the brakes on the use of this technique.” He sounded a narrower cautionary note in a March 1997 Antiquity paper, which described “fundamental problems in the use of radiocarbon dating” at the Portuguese sites.

Beck and Broecker, however, say Dorn began publicly backing away from his technique only after he learned of Beck's concerns in August 1996. And Malotki says that Dorn didn't warn him of problems with the method even though Malotki was preparing their paper in the spring and summer of 1996. Dorn flatly contradicts this and says that he did indeed tell Malotki that the paper needed revision. Finally, in September 1996, Broecker says he forwarded the team's preliminary findings to the NSF Inspector General's office. NSF responded to him in late 1997, asking for a list of researchers qualified to review the issue. And in April, ASU moved ahead with its own inquiry, asking Beck to testify before a review committee.

Some researchers who know Dorn say that they don't believe the charges of tampering. Whiteley, for example, calls them “absurd.” Ken Hedges, editor of La Pintura and an archaeological curator at the San Diego Museum of Man in California, says that “most of us who know Ron don't think they hold any water.”

No matter how the NSF and ASU inquiries turn out, Dorn agrees that dates included in more than 20 of his publications over the last dozen years must now be viewed with skepticism. However, many archaeologists have long been skeptical of rock art dating anyway, notes archaeologist Benjamin Swartz Jr. of Ball State University in Muncie, Indiana. Thus, although Dorn estimated some southwestern rock art to be more than 15,000 years old—implying an early peopling of the Americas—“his dates weren't any better than any others,” Swartz says. “The general consensus remains that the earliest [North American] rock art is about 12,000 years old.”

The controversy is likely to cast a cloud over rock-varnish science, says geologist Fred Phillips of the New Mexico Institute of Mining and Technology in Socorro, who published a number of papers with Dorn. Says Phillips: “I would not submit any type of proposal involving rock varnish at this point.”

Navigate This Article