Carbon dating the Dead Sea Scrolls


Carbon dating the Dead Sea Scrolls refers to a series of radiocarbon dating tests performed on the Dead Sea Scrolls, first by the AMS lab of the Zurich Institute of Technology in 1991 and then by the AMS Facility at the University of Arizona in Tucson in 1994–95. There was also a historical test of a piece of linen performed in 1946 by Willard Libby, the inventor of the dating method.

Testing

One of the earliest carbon dating tests was carried out on November 14, 1950. This was on a piece of linen from Qumran Cave 1, the resulting date range being 167 BCE233 CE. Libby had first started using the dating method in 1946 and the early testing required relatively large samples, so testing on scrolls themselves only became feasible when methods used in the dating process were improved upon. F.E. Zeuner carried out tests on date palm wood from the Qumran site yielding a date range of 70 BCE90 CE. In 1963 Libby tested a sample from the Isaiah Scroll, which provided a range of 200 BCE1 CE.
In 1991, Robert Eisenman and Philip R. Davies made a request to date a number of scrolls, which led to a series of tests carried out in Zurich on samples from fourteen scrolls. Among these were samples from other sites around the Dead Sea, which contained date indications within the text to supply a control for the carbon dating results. A similar battery of tests was carried out in 1994–95 in Tucson, this time with samples from twenty-two scrolls as well as another piece of linen.

14C Test results

The following table shows all the Qumran-related samples that were tested by Zurich, Tucson and Libby. The column headed "14C Age" provides a raw age before 1950 for each sample tested. This represents the ideal date for the amount of 14C measured for the sample. However, as the quantity of 14 absorbed by all life fluctuates from year to year, the figure must be calibrated based on known fluctuation. This calibrated range of dates is represented in the last column, given with a 2-sigma error rating, which means at 95% confidence. With the exception of the first text from Wadi-ed-Daliyeh, the texts in the table below are only those from the caves around Qumran. The table orders them chronologically, based on 14C age.
-LabDescription14C AgeCalibrated Age
1Z2289 +/- 55408–203 BCE
2ZTestament of Qahat2240 +/- 39395–181 BCE
3T1QIsaiaha2141 +/- 32351–295 or 230–53 BCE
4ZFrg. 3 2139 +/- 32351–296 or 230–53 BCE
5Z1QIsaiaha2128 +/- 38351–296 or 230–48 BCE
6Z4Q213 Levia ar2125 +/- 24344–324 or 203–53 BCE
7T4Q249 pap cryptA2097 +/- 50349–304 or 228 BCE–18 CE
8Z4Q53Samuelc2095 +/- 49349–318 or 228 BCE–18 CE
9L1QIsaiaha2050 +/- 100200 BCE – 1 CE
10T4Q208 2095 +/- 20172–48 BCE
11T4Q2672094 +/- 29198–3 BCE
12T4Q317 Phases of the Moon2084 +/- 30196–1 BCE
13T1QpHab Habakkuk Commentary2054 +/- 22160–148 or 111 BCE–2 CE
14T4Q22 paleoExodusm2044 +/- 65342–324 or 203 BCE–83 CE or 105–115 CE
15T1QS Community Rule2041 +/- 68344–323 or 203 BCE–122 CE
16Z11Q19 Temple Scroll2030 +/- 32166 BCE–67 CE
17T4Q22 paleoExodusm patch2024 +/- 39161–146 or 113 BCE–70 CE
18Z1QApGen Genesis Apocryphon2013 +/- 3289 BCE–118 CE
19T4Q521 Messianic Apocalypse1984 +/- 3349 BCE–116 CE
20Z1QH Thanksgiving Scroll1979 +/- 3247 BCE–118 CE
21T4Q258 Comm. Rule, 2nd sample1964 +/- 4550 BCE–130 CE
22T4Q266 Damascus Documenta1954 +/- 3844 BCE–129 CE
23T4Q171 Psalms Commentarya1944 +/- 233–126 CE
24T4Q258 Comm. Rule, 1st sample1823 +/- 24129–255 or 303–318 CE

Non-scroll material tested:
25TQumran 4Q Linen with leather thong2069 +/- 40197 BCE–46 CE
26LQumran 1Q linen1917 +/- 200167 BCE–233 CE

Many of the date ranges provided are actually two date ranges, for example the Habakkuk Commentary, which is given as 160–148 or 111–2 CE. The section of the calibration curve for the 14C age of the Habakkuk Commentary is complex, so that the 14C age of 2054 cuts through a few spikes on the curve, providing two date ranges.

Observations

The Great Isaiah Scroll 1QIsaa has been tested three times, once by Libby, once at Zurich and once at Tucson. The results from the latter two were almost identical, which is a good indicator of the basic accuracy of this dating method. 1QS, tested at Zurich, and 4QSamc, tested at Tucson, provide overlapping date ranges, which is expected when both texts are attributed to the same scribe. When 4Q258 was tested at Tucson its result was so anomalous that the laboratory was asked to retest another sample from the same document. The second test yielded a result that was deemed more satisfactory.