Tell Fara / Shuruppak

• OxA-33,185 (4178 ± 35 B.P.)

• Object type: bone tool (unknown species)

• Museum no.: Penn 33-13-432

• Context info.: DE 39; Grave, Level 2; Exc. no. 436

This small broken bone pin was the only one of five sampled Early Dynastic (hereafter ED) I/II tools from Fara to produce sufficient collagen yield, which is usually used for radiocarbon dating (Fig. 3). There seems to be a slight conflict of information regarding the context of this sample. On the one hand, according to the museum records, this object came from Level 2. On the other, its recorded depth is 3.4 to 3.5 m below the surface, which would place it in Level 4 (Martin Reference Martin1988: 24). This makes me believe that the object likely came from the fill of one of the graves sunk from Level 2. Though the number of the grave is not recorded, one may expect the pin to have originally come from the shaft of one of the following deep graves in D E 39: 11, 13, or 14. All three of these graves are believed to have corresponded to the late ED I/II, pre-ED IIIa period on the basis of stratigraphy and associated finds (Martin Reference Martin1988: 26–31). Given the Fara-style (ED IIIa) tablet found just below Level 1, the date provides a good terminus post quem for the ED IIIa period on the site.

Fig. 3

Bone pin, Penn 33-13-432 from Tell Fara/Shuruppak, sampled for radiocarbon dating

The pin was polished (possibly waxed) and glued, therefore the sample's surface was air-abraded prior to solvent wash. A section of the pin was drilled and the powder underwent standard solvent extraction and acid-alkali-acid pretreatment, followed by collagen extraction.

As visible in Figure 4, a plateau in the calibration curve c. 2850–2550 B.C. adversely affects the precision of the measurement. Nevertheless, the result suggests a range of c. 2880–2700 B.C. as the most likely date for the pin. The date finds its closest parallels in the samples from Levels IX and VIII at the Nippur Inanna sounding (Stuckenrath Reference Stuckenrath1963; Stuckenrath and Ralph Reference Stuckenrath and Ralph1965) plotted in Figure 5. Together with the ceramic evidence (Hansen Reference Hansen and Ehrich1965; Hansen and Dales Reference Hansen and Dales1962; Wilson Reference Wilson, Finkbeiner and Röllig1986) as well as the few cuneiform tablets recovered from these levels (Buccellati and Biggs Reference Buccellati and Biggs1969), the radiocarbon sample from Fara confirms the ED I/II date of the context. Moreover, it corresponds closely to the single radiocarbon measurement performed on a sample from pre-Ur-Nanshe levels at Tello/Girsu (Hritz et al. Reference Hritz, Pournelle, Smith, Albadran, Issa and al-Handal2012).

Fig. 4

Probability distribution showing the calibrated date OxA-33,185 from Fara

Fig. 5

Comparison of radiocarbon dates from ED I/II contexts from the Nippur Inanna Temple Sounding, Girsu, and Fara; P-801 is highlighted as an obvious outlier

Tell Muqayyar/Ur

• OxA-33,848 (4166 ± 32 B.P.)

• Object: wood (Cedrus libani)

• Museum no.: BM 1935,0116.87

• Context: coffin wood from PG1648

PG1648 was described by Woolley as the smallest of all the royal tombs - a single limestone chamber no larger than 2 x 1.5 metres (Woolley Reference Woolley, Burrows, Keith and Legrain1934: 133–134). It was dominated by a wooden coffin, which contained a poorly preserved skeleton of a young man. Beads of gold, silver and lapis lazuli placed around the individual's head suggested an elaborate headdress. The chamber contained three subsidiary burials: a male and two females, one young and one older. Molleson and Hodgson (Reference Molleson and Hodgson2003) analysed the skeletons and found that the bodies had been exposed to high temperatures and that the male body showed stress commensurate with head basket-carrying. Not much can be said about the stratigraphic position of the grave, except that it is overlaid by another small royal tomb, PG1618. The coffin fragment was identified by Dr C. Cartwright as cedar wood, at least 35 years old at the time of its felling. The sample is therefore likely affected by the old wood problem.

Wood samples from the Royal Cemetery at Ur proved extremely difficult to date due to poor preservation. Standard pretreatment methods involving the alkali phase were too destructive, and so an alternative approach was adopted. The sample was cleaned with a weak HCl solution, followed by seven rounds of microfiltering and ultrasound baths in ultrapure water. This method is likely to be less effective in removing humic acids from the sample; however, previous studies (Wild et al. Reference Wild, Steier, Fischer and Höflmayer2013) showed the potential effects of exogenous humic acids on Bronze Age samples from the Eastern Mediterranean to be minimal, and therefore it is unlikely that the choice of the pretreatment protocol affected the final readings.

• OxA-34,011 (4211 ± 32 B.P.)

• Object: wood (Cedrus libani)

• Museum no.: BM 1935,0116.87

• Context: coffin wood from PG1648

The sample was taken from the same piece of wood as OxA-33,848. The two measurements were combined into a single date estimate.

• OxA-X-2685-8 (3992 ± 36 B.P.):

• Object: ox tooth

• Museum no.: BM 1935,0116.17

• Context: PG789

A detached molar (no. 7 in Fig. 6) was found near an ox jaw in the richly furnished PG789. The tooth most likely belonged to the ox which pulled the ceremonial chariot. The animal itself was decorated with jewellery, as evidenced by a gold nose-ring (c. 9 cm in diameter). There is therefore little doubt that the ox tooth can be associated with the royal burial.

Fig. 6

Sketch of the ox jaw found in PG789; molar marked with 7 was sampled for dating. Courtesy of Dr. A. Fletcher, British Museum

Unfortunately, none of the teeth recovered from the Royal Cemetery yielded enough collagen to allow dating. As a result, the radiocarbon measurement was performed on the mineral fraction. Previous attempts to date the mineral fraction of bones and teeth (Cherkinsky Reference Cherkinsky2009; Zazzo Reference Zazzo2014; Zazzo and Saliège Reference Zazzo and Saliège2011) showed that there is a visible difference between dates performed on bioapatite and collagen from the same organism. The likely explanation for this difference is the fact that mineral fractions are known to exchange carbon with the environment post-deposition, unlike collagen which simply degrades and decomposes. Unfortunately, the difference is by no means systematic and can result in the apatite dates being either younger or older by anywhere between 0 and 300 years for samples from the last 10,000 years, depending on the type of environment, water percolation and pre-treatment method used. For these reasons, bioapatite dates are seen as less reliable than their collagen counterparts. ORAU protocols do not include a standard methodology for bioapatite samples. For the purpose of this study, the chemical pretreatment method developed by Hopkins et al. (Reference Hopkins, Snoeck and Higham2016) was adopted. The study tested several pretreatment methods and identified one which produced dates which, while still possibly younger (by 0-100 years), were the closest to those performed on the collagen fraction. Therefore, although not as reliable as a potential collagen date, this measurement provides the most likely age estimate for the tomb's construction.

• BM-64 (3920 ± 150 B.P.)

• Object: human bone

• Context: PG755

• Source publication: Barker and Mackey Reference Barker and Mackey1961

This bone sample came from the skeleton of a young man usually identified as king Meskalamdu of Ur. This sample, together with BM-70 and BM-76, was published in 1961 and represents the very earliest attempt to apply radiocarbon dates to archaeological contexts in the region, hence the enlarged measurement error.

• BM-70 (4030 ± 150 B.P.)

• Object: human burnt bone

• Context: PG1515

• Source publication: Barker and Mackey Reference Barker and Mackey1961

The identity of the individual buried in the grave is unknown.

• BM-76 (3990 ± 150 B.P.)

• Object: human bone

• Context: PG800

• Source publication: Barker and Mackey Reference Barker and Mackey1961

This bone sample came from the female skeleton usually identified as queen Pū-abim.

• P-724 (3959 ± 59 B.P.)

• Object: wood

• Context: unknown

• Source publication: Stuckenrath and Ralph Reference Stuckenrath and Ralph1965

This reading was performed on a bulk sample of wood collected from various axes found within the graves. Little can be said about the context of this sample, however the fact that measurement has been performed on material likely affected by the old wood effect gives sufficient grounds to treat it as a terminus post quem for the bone samples.

Figure 7 clearly shows the difference in age between the wood and bone samples. As expected, the wood samples are older, most likely indicating an offset due to inbuilt age. The date of P-724 seems closer to the bone apatite date, however the credibility of the sample is questionable due to the mixed source material. The early British Museum dates are too imprecise to allow for any meaningful inference. Therefore, OxA-X-2685-8 remains the most reliable age estimate for the construction of the early graves in the Royal Cemetery, placing it most likely in the 2500 ± 50 B.C. range.

Fig. 7

Radiocarbon dates from the Royal Cemetery of Ur; dates taken on wood samples are highlighted

Table 2

New radiocarbon results for the Royal Cemetery of Ur

The comparison of radiocarbon dates in Figure 8 shows that Nippur Inanna levels VII-V (Stuckenrath Reference Stuckenrath1963; Stuckenrath and Ralph Reference Stuckenrath and Ralph1965) appear to be slightly older than the age estimate for the Royal Cemetery. Considering, however, that the Nippur dating was performed on charcoal samples, this may be the result of inbuilt age. Though the Inanna levels VIII-V were rather poor in datable ceramic, some finds (especially stone bowls, inlays and seals) from Level VII suggest that the stratum was roughly contemporary with the Royal Tombs of Ur (Hansen & Dales Reference Hansen and Dales1962: 79; Hansen Reference Hansen and Ehrich1965: 209). Radiocarbon samples from Areas A and E in Abu Salabikh (Bowman et al. Reference Bowman, Ambers and Leese1990; Burleigh and Matthews Reference Burleigh and Matthews1982; Fishman et al. Reference Fishman, Forbes and Lawn1977) plotted in the graph (Fig. 8) all came from contexts closely associated with the Fara-style tablets, often seen as characteristic of the ED IIIa period (Biggs Reference Biggs1966; Biggs Reference Biggs1974; Biggs and Postgate Reference Biggs and Postgate1978; Krebernik and Postgate Reference Krebernik and Postgate2009). Except for two very imprecise readings (BM-2328R and P-2051), the dates seem younger than the Royal Cemetery Tombs, which may be surprising. The BM readings from this batch, however, are problematic for technical reasons (Bowman et al. Reference Bowman, Ambers and Leese1990; Tite et al. Reference Tite, Bowman, Ambers and Matthews1987), and a new absolute chronology for Abu Salabikh (Wencel in preparation) allows us to doubt the credibility of these readings. A single date from the Ywn sounding in Kish (Zaina Reference Zaina2015) appears younger, which is to be expected considering the fact that its context was identified as more or less contemporaneous with Cemetery A at Kish, i.e. late ED IIIb/Akkadian in date. In summary, the radiocarbon dating of some of the earliest graves in the Royal Cemetery at Ur places it on the verge of ED IIIa and ED IIIb, corroborating archaeological periodisation.

Fig. 8

Comparison of radiocarbon dates from ED IIIa/b contexts from Nippur, Kish, Abu Salabikh, and the Royal Cemetery of Ur; P-806 is highlighted as an obvious outlier

Concluding Remarks

Although the results presented in this report do not contradict the conventional absolute chronology, they do contribute to the general discussion about the transformations which occurred in Southern Mesopotamia during the early third millennium B.C. Firstly, they offer a useful benchmark for the construction of more refined chronological models. Secondly, they provide independent support for a chronological scheme which, until now, has been highly speculative. The dates discussed in this paper have not been modelled, which means that the numbers were presented as they are, without any supplementary assumptions. The ultimate goal of this dating project is to follow a holistic, interdisciplinary methodology combining historical, archaeological, and radiocarbon data, using modern statistical techniques, in order to arrive at reliable and precise date estimates. Nevertheless, even the most sophisticated statistical technique will not be able to produce reliable estimates if the model has to rely on flawed, incomplete, or skewed datasets. For this reason, the importance of an in-depth assessment of the quality of individual radiocarbon dates and their archaeological provenience cannot be overstated. Further work on materials from Abu Salabikh and Tell Brak will provide additional radiocarbon measurements to address the problem of the limited radiocarbon dataset, before the entire database of radiocarbon measurements can be combined into a regional chronological model. Radiocarbon dating as well as other scientific chronometric techniques provide the unique opportunity to weigh in on relevant archaeological debates such as the date and duration of controversial periods or cultures (e.g. Jemdet Nasr, ED II), correspondence between history and archaeology (e.g. the Royal Cemetery of Ur and the First Dynasty of Ur), or the philological and palaeographic development of the cuneiform script (e.g. Fara tablets and ED IIIb texts). Future radiocarbon work should target such problematic periods in order to fully exploit the strengths of the technique.