Old Sherds, New Science

Hershel Shanks’s First Person in the May/June 2013 issue of BAR

The Biblical Archaeology Society (BAS) has recently published a must-read for dig directors with the imposing and somewhat intimidating title Cyber-Archaeology in the Holy Land—The Future of the Past (available online for free).a It is written by Tom Levy and his team from the University of California, San Diego,1 who have been excavating at the Edomite copper-mining site in Faynan, Jordan, since 1997.b According to the Bible, the site was visited by the Israelites on their Exodus journey from Egypt (Numbers 33:43).

We still dig the old-fashioned way, Levy tells us, “like our 19th-century predecessors with shovels, picks, trowels, dustpans, toothbrushes and so on. What is different is the use of digital tools to record data—and lots of it. In short, the way we collect and analyze the data that we recover is completely changing. Cyber-archaeology faces the challenges of quickly and accurately collecting masses of archaeological data, visualizing it and sharing it with colleagues and the public. This process can be visualized with a four-part model that focuses on acquisition, curation, analysis and dissemination of data.”

The ubiquitous pieces of broken pottery—sherds—found at every archaeological site in the Near East have been used for dating the various archaeological strata since the days of Flinders Petrie. With cyber-archaeology that will change:

“Instead of relying on the traditional archaeological assumptions that are used to date pottery to a chronological period based on style, we decided to use a large number of high-precision radiocarbon dates anchored into the stratigraphy of the site with the best contextual/cultural data that cyber-archaeology can offer.”

But just when we thought science had solved our dating problems, we are stopped short.

Download the FREE eBook “Cyber-Archaeology in the Holy Land — The Future of the Past,” featuring the latest research on GPS, Light Detection and Ranging Laser Scanning, unmanned aerial drones, 3D artifact scans, CAVE visualization environments and much more.

For more cutting-edge archaeological technology, visit the BAS Archaeological Technology page.

The plenary speaker at last year’s annual meeting of ASOR, the leading American professional association of Near Eastern archaeologists,2 was one of the world’s most eminent archaeologists: Manfred Bietak of the University of Vienna. Dates used in Near Eastern archaeology are ultimately dependent on ancient Egyptian dates, and Bietak digs at the leading site in Egypt for determining these dates: Tell el-Daba, a harbor town in the Nile delta. His work is universally admired.

His talk was extremely technical but shocking in its way: He questioned the precision of much radiocarbon (carbon-14) dating! To be sure I quoted Bietak accurately, I asked him if I could have a copy of his talk. He replied that he spoke without a manuscript, but he sent me an article covering much of the same material, titled “Antagonisms in Historical and Radiocarbon Chronology,” that is to be published in a forthcoming volume.3

“In some circles” Bietak observed, it has “become a crime to cast doubt on the reliable accuracy of radiocarbon dating.” Bietak calls it “the present deadlock,” referring to the differences between the two systems of dating, one by pottery sherds and the other by carbon-14.

In short, after years of extremely careful work at Tell el-Daba and by comparing it with other sites all over the Mediterranean, Bietak has determined a chronology based on what he calls historical data (mainly pottery seriation), which significantly differs from the chronology produced by carbon-14 tests. The difference in dates that Bietak refers to amounts to about 80 years, roughly the same as that between the “minimalists” and “maximalists” in Israelite chronology.c The carbon-14 tests focus especially on the Iron Age, the period ascribed to the Israelite monarchy, where the approximately 80-year difference can be crucial—and very controversial. A “critical attitude toward radiocarbon dating,” Bietak tells us, “becomes a casus belli instead of reflecting if and where there could be complications.”

BAS Library Members Read “Glossary: How to Date a Cooking Pot” by John C. H. Laughlin and “Archaeological Views: Carbon 14—The Solution to Dating David and Solomon?” by Lily Singer-Avitz as they appeared in Biblical Archaeology Review.

Not a BAS Library member yet? Sign up today.

Bietak outlines the scientific difficulties in attaining accuracy and precision in carbon-14 dating. The difference (or offset) in what Bietak calls “historical” dating as opposed to “scientific” (i.e., radiocarbon) dating can be accounted for in a number of ways. For example, the radiocarbon in the atmosphere fluctuates in different growing seasons, especially in winter and summer wheat. Samples from sites near the sea, even the Dead Sea, give carbon-14 dates different from what we would expect. In short, differences in region and time of year could play a role in explaining the deviation of carbon-14 dates.

Moreover, “dates in radiocarbon chronology are in most cases obtained by combining the results of samples, even if spread over a considerable time. Outliers are usually omitted, although the decision of what is an outlier or not is sometimes strangely biased.”

And then there is the need in all cases for a Bayesian calibration of raw carbon-14 dates. I wasn’t sure I could explain this for a lay audience, so I asked Tom Higham, deputy director of the Oxford Radiocarbon Accelerator Unit in England. This is what he told me: The calibration allows for a more refined chronometry (dating) by using Bayesian statistical approaches. It is a statistical adjustment that allows the inclusion of archaeological evidence such as stratigraphic layers, other dating evidence from coins, volcanic ash, artifacts and the like, along with the calibrated radiocarbon ages. The modeling of these combined sets of data results in probability age estimates for specific archaeological events that are more robust and usually much more precise than those produced with single calibrated date ranges.

That’s about as plain and simple as you can get.

In short, radiocarbon does not offer a quick fix. We can’t stop dating pottery sherds “historically.”

As between Bietak’s dates based on his pottery chronology and the dates based on radiocarbon tests, I’ll put my money on Bietak.

Download the FREE eBook “Cyber-Archaeology in the Holy Land — The Future of the Past,” featuring the latest research on GPS, Light Detection and Ranging Laser Scanning, unmanned aerial drones, 3D artifact scans, CAVE visualization environments and much more.

For more cutting-edge archaeological technology, visit the BAS Archaeological Technology page.



1. Actually the Center of Interdisciplinary Science for Art, Architecture and Archaeology (CISA3) at the California Institute of Telecommunications and Information Technology (Calit2) at the University of California, San Diego.

2. The American Schools of Oriental Research (ASOR).

3. “Antagonisms in Historical and Radiocarbon Chronology,” in A.J. Shortland and C. Bronk Ramsey, eds., Radiocarbon and the Chronologies of Ancient Egypt (Oxford: Oxbow, 2013), pp. 78–110.

a. Thomas E. Levy, Neil G. Smith, Mohammad Najjar, Thomas A. DeFanti, Falko Kuester and Albert Yu-Min Lin, Cyber-Archaeology in the Holy Land—The Future of the Past (Washington, DC: Biblical Archaeology Society, 2012). www.biblicalarchaeology.org/cyber.

b. See Thomas E. Levy and Mohammad Najjar, “Edom and Copper,” BAR 32:04; Thomas E. Levy and Mohammad Najjar, “Condemned to the Mines,” BAR 37:06.

c. See Yosef Garfinkel, “The Birth and Death of Biblical Minimalism,” BAR 37:03; Philip R. Davies, “What Separates a Minimalist from a Maximalist? Not Much,” BAR 26:02; Hershel Shanks, “Face to Face: Biblical Minimalists Meet Their Challengers,” BAR 23:04.

Posted in Archaeology Today.

Tagged with , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , .

Add Your Comments

6 Responses

Stay in touch with the conversation, subscribe to the RSS feed for comments on this post.

  1. JAllan says

    Another problem is that radiocarbon dating requires CARBON, that is organic material. And it does not date from when a structure was built, but from when an animal or plant DIES. As an example, BAR has a current article on reuse of cedar timbers previously in another building in a new one, pointing out that cedar (especially the Cedar of Lebanon) is very durable. The article has carbon dates on cedar beams that were reused in a mosque on the Temple Mount, and reported that that mosque was rebuilt in the 1800’s using timbers from the previous building, with some of the timbers left in a pile that could amount to storage. Some of these timbers have carbon dates back to the Second Temple or even the First Temple! If we did not KNOW that the mosque which was torn down and replaced had been built during the Islamic period, we would think it dated from Herod or even Solomon! And if a future archeologist were to uncover the ruins of this mosque (hypothetically; I don’t want to be accused of anything), and test one of the older reused beams, that archeologist might think the MODERN mosque was built thousands of years earlier!

    So one piece of caution is, if you carbon date wood from a building ruin, rather than AVERAGE the ages of samples, pick the NEWEST sample you can get. In our hypothetical example, this would include trees cut in the 19th century. But then, what if the building IS older, but this piece of wood was REPLACED? Then you could err in the opposite direction.

    And this does not only apply to Biblical sites. There is a small specialty market in the U.S. for “salvage” logs that had been lost and “drowned” underwater in 19th century logging operations while being floated downriver to be shipped. Logs that became waterlogged (pun intended) and sunk to the bottom of deep rivers and were covered with mud are very well preserved. People are building houses TODAY with wood that was harvested during the CIVIL WAR (1861-1865 for the international audience)!

    If it is possible, it would be good to date pottery by some similar process using different elements; possibly the oxygen-18 content of patina caused by weathering (as opposed to the original silicate minerals in the clay), or of rust/corrosion on ancient iron/bronze weapons and tools.

  2. Saul says

    With all due respect, I find Mr. Shank’s thesis hard to swallow. He quotes Bietak statement that “the radiocarbon in the atmosphere fluctuates in different growing seasons, especially in winter and summer wheat.”. Even if true, the question is whether there is a fluctuation is the RATIO of C-14 and C-12 in the atmosphere, not whether the ABSOLUTE amount of C-14 changes. Shanks also quotes Bietak as stating that “Samples from sites near the sea, even the Dead Sea, give carbon-14 dates different from what we would expect”. Expect from what? Pottery sherd dating? Why assume that is correct?

    Shanks quotes Bietak as stating “Outliers are usually omitted, although the decision of what is an outlier or not is sometimes strangely biased.” Wow. Omitting an outlier would be considered fraud in science-based disciplines (physics, chemistry, biology, evidence-based medicine, etc). Does Bietak claim that his colleagues routinely perpetrate scientific fraud? Do archaeolgists who use only “traditional methods” like pottery sherd dating also omit outliers?

    Shanks ends by stating “As between Bietak’s dates based on his pottery chronology and the dates based on radiocarbon tests, I’ll put my money on Bietak”. I’d take up that bet. In a conflict between a validated, quantitative scientific technique (radiocarbon dating) and a non-validated, subjective, qualitative technique (such as pottery sherd dating) I too have no doubt in my mind which to rely on.

    I find the last quote, from the director of the Oxford Radiocarbon Accelerator Unit, the most interesting. “The modeling of these combined sets of data results in probability age estimates for specific archaeological events that are more robust and usually much more precise than those produced with single calibrated date ranges.”. Note the terms Dr. Higham used – “more robust” and “more precise”. He did not say “more accurate”. “Robustness” and “precision” mean the results are more reproducible, but possibly wrong. Had Dr. Higham stated that the Bayesian calibration yielded more accurate results, Bietak and Shanks might have had a leg to stand on. Precision alone, without accuracy, is valueless in this case.

    Referring to end of JAllan’s letter, oxygen-18 is a stable isotope and cannot be used to date anything (though it has other uses in paleoclimatology and studies of plant physiology).

  3. Kurt says


    Definition: Dates mark the time at which events occur. The Bible expresses dates in relation to the lifetime of individuals, the period during which certain rulers were in office, or other notable events. It contains the only complete chronology reaching back to the time of Adam’s creation. Bible chronology also pinpointed in advance the time when certain important events in the fulfillment of God’s purpose would take place. The Gregorian calendar, which is now popular in much of the world, did not come into use until 1582. In secular sources there is disagreement on dates given for events in ancient history. However, certain key dates, such as 539 B.C.E. for the fall of Babylon, and hence 537 B.C.E. for the Jews’ return from captivity, are well established. (Ezra 1:1-3) Using such dates as starting points, it is possible to express in terms of current calendars the dates for ancient Biblical events.

    Have scientists proved that humans have been on earth for millions of years, not merely some thousands of years as the Bible indicates?
    The dating methods used by scientists are built on assumptions that can be useful but that often lead to very contradictory results. So, dates given by them are constantly being revised.
    A report in New Scientist of March 18, 1982, reads: “‘I am staggered to believe that as little as a year ago I made the statements that I made.’ So said Richard Leakey, before the elegant audience of a Royal Institution evening discourse last Friday. He had come to reveal that the conventional wisdom, which he had so recently espoused in his BBC television series The Making of Mankind, was ‘probably wrong in a number of crucial areas.’ In particular, he now sees man’s oldest ancestor as being considerably younger than the 15-20 million years he plumped for on television.”—P. 695.
    From time to time, new methods of dating are developed. How reliable are these? Regarding one known as thermoluminescence, The New Encyclopædia Britannica (1976, Macropædia, Vol. 5, p. 509) says: “Hope rather than accomplishment mainly characterizes the status of thermoluminescence dating at the present time.” Also, Science (August 28, 1981, p. 1003) reports that a skeleton showing an age of 70,000 years by amino acid racemization gave only 8,300 or 9,000 years by radioactive dating.
    Popular Science (November 1979, p. 81) reports that physicist Robert Gentry “believes that all of the dates determined by radioactive decay may be off—not only by a few years, but by orders of magnitude.” The article points out that his findings would lead to the conclusion that “man, instead of having walked the earth for 3.6 million years, may have been around for only a few thousand.”
    It should be noted, however, that scientists believe that the age of the earth itself is much greater than the age of man. The Bible does not disagree with that.
    Were the ages of people before the Flood, as stated in the Bible, measured according to the same kind of years that we use?
    If it was reasoned that the “years” must be equivalent to our months, then Enosh became a father when he was seven years of age, and Kenan was only five years old when he fathered a son. (Gen. 5:9, 12) That is, clearly, impossible.
    Detailed chronology provided in connection with the Flood indicates the length of months and years used at that time. Comparison of Genesis 7:11, 24 and Ge 8:3, 4 shows that five months (from the 17th of the 2nd month to the 17th of the 7th month) was equal to 150 days, or five 30-day months. Specific reference is made to a “tenth month” and to further periods beyond that before coming to the next year. (Gen. 8:5, 6, 8, 10, 12-14) Evidently, their years were made up of twelve 30-day months. At a very early time, the strictly lunar calendar was adjusted periodically to the length of the solar year, as indicated by Israel’s holding seasonal festivals of ingathering on specified dates. In that way the festivals continued to fall in the appropriate seasons.—Lev. 23:39.
    Keep in mind that God made humans to live forever. It was Adam’s sin that led to death. (Gen. 2:17; 3:17-19; Rom. 5:12) Those who lived before the Flood were closer to perfection than we are today, so they lived much longer. But each one died within a thousand years.

  4. Scott says

    Just want to note directly above, the WatchTower/Jehovah’s Witnesses, so called, have inflated their date for the flood by 20 tyears to protect their precious interpretation of the 70 years for Babylon to being 70 years for Judah being abandoned, and so they add 20 years further back to the flood date. It should be 2350 BC, not 2370. Other than that, they have bee trying to keep up with my material and direction (they will never admit it). Not bad for an apostate, so called, eh? they haven’t seen nothing yet!

  5. Scott says

    Much of this really hinges on floating chronologies. Though we can match pottery shards and other such material culture to link cultures and time periods in a broad general sense, they can not, of themselves, determine precise dating, since shards can not reflect accurate measures of time. Radiocarbon represents the general best method for more carefully determining age, since carbon is a relatively easy material to find in remains and culture. It is not perfect, but it is very helpful and a useful tool up to a certain point.

    What all have failed to admit is that there have been articles published in the RadionCarbon Journal that show that Carbon with carbon 14 can be contaminated by the environment, with environmental carbon 14 replacing in situ carbon remains, without any ability to distinguish contamination. I have the exact references up on my site in due time. But old carbon can infect and alter carbon in situ. “New” carbon can also infect and alter carbon 14 measurements and no way to detect or separate it.

    For example, if piece of wood had a real date of 1000 BC, and was near to a water source that had old calcium carbonate, which had very old carbon dioxide in the carbonate, which then is leeched from the lime into the water, which then flows through the ground over the piece of wood, individual atoms and molecules of carbon in the wood can be replaced by the lime carbon atoms and molecules so that the date of the wood now reflects an older age, than 1000 BC, say 2000 BC. If rain, with current levels of CO2 ,were to rain and filter down into the soil where the wood piece is, and leech and exchange carbon atoms and molecules, then the wood, would now appear to be of a younger than 1000 BC age, say 500 BC instead.

    Herein is the potential problem of carbon 14. RC14 journals have shown many such examples that doe not reflect known established historical volcanic events which cause things nearby to look much older, since the carbon dioxide gas released is “dead” carbon which will immediately increase the apparent age of things around it.

    That is why most dates of material that reflects the flood, show 10,000 BC, rather than 4350 BC, due to massive localized releases of volcanic and tectonic carbon dioxide gas, which look like dead carbon and mix with flood remains and look a little more than twice as old in many cases. The massive explosion of Mt. Toba, concurrent with the flood says I, reflect 70,000 years BC, rather than 10,000 BC, since Toba spewed far more gases and ash in that area. The megafauna die-off of Australia, not far from Toba, show dates of 40,000 in general, whereas in the Americas, artic remains show 10,000 BC. Places in Siberia will often show 30,000 or 40,000 BC. The further you venture away from Mt. Toba, the old age decreases.

    Carbon 14 remains in the Americas usually show near to 10,000 BC, with some occasionally showing near to 19,000 BC or so, though these are in the minority.

    So here, carbon 14 is still useful, for it shows the distribution pattern of volcanic and tectonic releases of gas that accompanied massive releases of waters beneath the crust, that led to temporary flooding. Continuing Upheavals cause mount chain rises and subduction of the released waters so that water levels came back down to within 150-300 feet of their former levels before flood. We have higher water levels to day than we did before the flood.

  6. Lawrence says

    From my understanding, Carbon 14 dating works on the premise that all living animals (Humans included) when they die have a certain amount of carbon in them that decays at a uniform rate over time. The problem is that this isn’t so and from what I’ve read on the subject within the first six months of death 80% of carbon is lost.

    In 2005 a samlpe of rock that came into being after Mount St Helens erupted was sent to some of the top radiocarbon dating labs in the world. the dates given for the age of the rock was anything from 3 to 20 million years old, that was despite the fact that it only became ‘rock’ in 1986 and before that it was molten magma! Carbon 14 is based on a faulty premise

Some HTML is OK

or, reply to this post via trackback.

Send this to friend

Hello! You friend thought you might be interested in reading this post from http://www.biblicalarchaeology.org:
Old Sherds, New Science!
Here is the link: http://www.biblicalarchaeology.org/daily/archaeology-today/old-sherds-new-science/
Enter Your Log In Credentials

Change Password