Save

A Commentary on Two Scientific Studies of the ruma (רומא) Jar from Qumran

In: Dead Sea Discoveries
Author: Joseph Yellin1
View More View Less
Full Access

Two chemical studies of the ruma jar from Qumran Cave 7 are examined. It is shown that these researches, seemingly at odds, have more in common than not. It is argued that the evidence from both studies point to Jerusalem as the origin of the ruma jar although one of the studies concludes that the ruma jar was locally made in Qumran from local clay. The data from both studies are consistent but the interpretations differ.

Abstract

Two chemical studies of the ruma jar from Qumran Cave 7 are examined. It is shown that these researches, seemingly at odds, have more in common than not. It is argued that the evidence from both studies point to Jerusalem as the origin of the ruma jar although one of the studies concludes that the ruma jar was locally made in Qumran from local clay. The data from both studies are consistent but the interpretations differ.

Introduction

Several studies on the elemental composition of Qumran pottery were published in the last two decades.1 The object of these studies was to determine the origin of pottery found in Qumran, in neighboring caves and in other settlements in the proximity of Qumran. Two of these studies focused attention on an unusual jar from Qumran cave 7.2 The jar has pierced handles and bears the Hebrew inscription “רומא”, or “ruma”. The ruma jar (also translated “roma”) has drawn much attention and speculation concerning the significance of its inscription. Does the word refer to Rome? And if so, was the jar made in Rome for shipment of goods to Judaea? Perhaps it was made in Judaea and intended for shipment of goods to Rome. Or perhaps it was made by a Roman? Alternatively, does the word refer to ruma, an ancient village roughly in Kiryat Ata half way between modern day Tiberias and Haifa, in northern Israel? Was the potter from the village ruma? It is not the purpose of this commentary to speculate as to the inscription on the ruma jar but to examine the chemical evidence as it bears on the origin of the jar. It is vitally important to know whether differing conclusions concerning the provenance of Qumran pottery are due to faulty data or to flawed methodology in the interpretation of good data. In what follows I will argue that the latter is the case in the matter of the ruma jar.

As stated above, two different studies report on the analysis of the ruma jar. Both of these studies were based on instrumental neutron activation analysis (inaa). The first study, reported by yb2001 concluded that this jar was made in Jerusalem or vicinity. This conclusion was later disputed by gb2003 who propose that the jar was made in Qumran from local clay that is unrelated to Jerusalem pottery clay. Here I argue that the evidence for a local Qumran provenance of the ruma jar, as put forth by gb2003, is weak and that their data supports rather than refutes the conclusion of ybe2001.

At the heart of these works is the scientific data (measurements) and their interpretation in terms of provenience. The analytical work for ybe2001 was performed in Jerusalem while the analytical work for gb2003 was performed in Budapest.3 The disagreement between the conclusions of gb2003 and ybe2001 appear to be in the interpretation of the data rather than in the data itself.4

Background

Crucial to both of these studies is the chemical composition of clay from the Motza formation, and of Roman pottery made in Jerusalem and reported by Mommsen et al.5 This Jerusalem pottery composition was related to the Motza Formation clay. The same system of analysis was used for both the analysis of the ruma jar by ybe2001 and the Jerusalem reference group of Mommsen et al. Although instrumental and procedural changes occurred over time, tests were conducted to ensure that the measurements were correct and not affected by changes in laboratory procedures and instrumentation. Thus, although many years elapsed between the measurements associated with the Jerusalem reference group and the ruma jar and despite changes in instrumentation, the old and new measurements are equally precise and accurate. (As an example, see Yellin and Cahill.6 In this study of the Rosette Stamped handles a lmlk jar previously reported by Mommsen, Perlman and Yellin was reanalyzed for quality control.)

Mommsen et al. compared their Jerusalem Reference Group to samples of clay from the Motza Formation.7 Of the clays they sampled, the closest matches were to samples of clay from the towns of Beit ‘Ummar and Motza. Of these two matches, the clays from the town of Motza provided the best correlation.

Results from ybe2001

ybe2001 reported on the analysis by inaa at the Archaeometry Laboratory, Institute of Archaeology, The Hebrew University of Jerusalem (hu) of a sample of pottery from Qumran and from Ein Ghuweir, a settlement contemporaneous with Qumran and located ca. 15 km to the south, on the shore of the Dead Sea. A preliminary comparison of the Qumran and Ein Ghuweir specimens had been reported by Broshi;8 however, ybe2001 was a more thorough treatment and discussion of these and included some additional samples of Qumran pottery.

ybe2001 distinguished two compositional groups (with some outliers) within their dataset. One of these groups (khq-qrn) they presumed to be reflective of local ceramic production at Qumran.9 The other group (khq-jer) was observed to be most similar to the Jerusalem Reference Group reported by Mommsen et al.10 The Jerusalem Reference Group comprises 20 specimens of Roman-period pottery from Jerusalem, and is likely made from clay from the Motza Formation.

The ruma jar was one of the specimens analyzed by ybe2001, and was designated khq24.11 Statistical analysis of the compositional data for khq24 led ybe2001 to conclude that the jar has a composition most similar to the Jerusalem Reference Group. These similarities were based on concentrations of individual elements, patterning of Rare Earth Elements (rees), as well as naturally occurring elemental correlations (e.g., Sc and Fe). They noted in all respects excepting sodium (Na) the ruma jar and the khq-jer compositional group were statistically indistinguishable from the Jerusalem Reference Group. The significant difference in Na content was suggested to reflect the ubiquity of salt (NaCl) in the environment surrounding the Dead Sea.

Results from gb2003

gb2003 report the analysis of numerous pottery specimens from Qumran and the surrounding vicinity, including the ruma jar. In their narrative of the results, they designate at least five compositional groups within their sample. Two of these five groups are pertinent to the discussion here. What gb2003 designate as Group i is asserted to represent on-site pottery production at Qumran using local clays unrelated to the Motza Formation. Their Group ii is suggested to be of non-local origin, and they report finding that the closest-matching specimens in their database are Motza Formation clays from Beit ‘Ummar. It is unknown whether these particular clays are related to those discussed by Mommsen et al. as being strong statistical matches for the Jerusalem Reference Group.12

In their analysis, gb2003 designate the ruma jar as specimen qum134. They find that qum134 is a member of their Group i (i.e., it is a product of local ceramic production at Qumran and is not related to Motza Formation clay). In discussing their Group ii, they state:

[Group ii] has high potassium values and low sodium unique to the Motza Clay Formation, as exhaustively described by Perlman et al. in 1986. One of the reasons why we refuted a Jerusalem make of the roma jar in Jerusalem, is exactly the lack of the high potassium and the low sodium in the roma jar that should be there if it were from Jerusalem.13

No specific tests are reported to demonstrate that qum134 belongs to their Group i or that it does not belong to Group ii, and indeed no summary statistics for any of their groups are presented. In support of the assignment of qum134 to Group i, they reference Fig. 2 of a supplemental report by Balazs;14 however, Balazs’s Fig. 2 is a schematic of the addition of two components in a mixture, representing how two components (e.g., clay and temper) combine to produce a single chemical composition. They also reference Balazs’s Fig. 4, a point to which I will return below.

The composition of Group i is not given. However, on page 13 gb2003 compare the composition of the ruma jar to their Group ii and exclude the ruma jar from Group ii on the basis of the sodium, potassium concentration. Group ii composition is not given but this group of pottery is said to be chemically different from local Group i and it is suggested that the pottery (of Group ii) was made in Qumran from clay that may have originated in the Beit Ummar-Hebron region (cf. gb2003, pages 24, 26) and transported by flood waters from the high grounds of Hebron to the Dead Sea region. The composition of Group ii is said to reflect the composition of the Motza Formation clay from the vicinity of Hebron. However, the authors do not present a Beit Ummar or Hebron composition and, therefore, fail to substantiate their conclusion that the composition of the ruma jar does not match that of Beit Ummar or Hebron.

Additional Studies of Qumran Pottery Provenience

In the same volume as gb2003, Michniewicz and Krzyśko report on the analysis of ca. 50 ceramic specimens from Qumran and a large number of geological clay specimens from Qumran, Jerusalem, Jericho, Hebron, and elsewhere.15 Specimens in their study were analyzed by neutron activation and by thin-section petrography. inaa of their sample was performed at a commercial laboratory in Canada, and no attempt was made to compare their data with data from other laboratories. The ruma jar was not analyzed in their study, but results of their combined inaa and petrographic analysis led them to suggest that “[t]he Qumran jars under investigation were made of a raw material which is not present in the vicinity of the Qumran site [and that] most of the jars were probably made from the Moza [sic] Formation clays, known and widely utilized in Judaea.”16

Congruency

Three studies of pottery from Qumran have been briefly summarized above. Both ybe2001 and gb2003 distinguished a compositional fingerprint presumed to be local, and one that is most-likely associated with clays of the Motza Formation. However, whereas gb2003 maintained that the local pottery was made of local clay the authors of ybe2001 were more cautious. Lacking chemical evidence on local clays and citing evidence that pottery was produced in Qumran, the authors suggested that the non- Jerusalem pottery may represent local material. They noted, however, a chemical kinship between the Jerusalem and non- Jerusalem pottery as it is reflected in the rare earth pattern of the two groups. They suggested that the two groups may be from geologically close regions. Michniewicz and Krzyśko also concluded that the Motza Formation was the most-likely source, although they did not observe any evidence of local production.17 Samples sizes for both ybe2001 and for Michniewicz and Krzyśko were much smaller (n = 31 and n = 50, respectively) than that reported by gb2003 (n ≈ 200).18 Thus, it is unsurprising they (gb2003) report a larger number of compositional groups.

The consensus of all three studies then, is that the majority of pottery recovered from Qumran is made of clay compositionally identical (both geochemically and petrographically) to the Motza Formation. This is an important finding, and it provides an empirical warrant for continued provenance studies of Qumran pottery.19

Incongruency

A central incongruency between ybe2001 and gb2003 concerns the ruma jar. ybe2001 concluded that it is chemically most-similar—excepting the element Na—to the Jerusalem Reference Group, which Mommsen et al. reported as chemically matching Motza Formation clays from the towns of Motza and Beit ‘Ummar.20 Conversely, gb2003 claim to refute this assignment, and assign the ruma jar to their compositional group believed to represent local production (Group i).

Thus, two independent research teams arrived at radically different conclusions regarding the provenance of a single specimen. How is that possible? I suggest that these differences result from the analysis and explanation of the chemical data rather than in the data itself. That is, the chemical data generated by both research teams are accurate and precise characterizations of the composition of the ruma jar, but the manner in which gb2003 interpret their data is methodologically flawed. Indeed, as I demonstrate below, examination of their data for the ruma jar supports, rather than refutes, the conclusions of ybe2001.

Comments on gb2003

The composition of the ruma jar (number qum134) is listed on p. 48 of gb2003.21 However, these data and others appear to have been shifted downward such that the concentration listed for for Ce is actually the concentration of the element above (Ca). If the rows are shifted upward by one row, the values become reasonable and agree with those listed in Balla’s dissertation.22 These corrected data are reproduced in my Table 1.

table 1

Composition of the Jar ruma/roma as measured in Jerusalem and Budapest

table 1

* Corrected from Gunneweg and Balla 2003 (p. 48) and verified against Balla (2005).

† From Yellin, Broshi and Eshel 2001 (p. 76). Values in parentheses not originally reported.

‡ Jerusalem Reference Group from Perlman et al. 1986 (p. 79).

§ From Balla 2005 (p. 88).

** From Perlman et al. (1986).

As mentioned above, summary statistics for the compositional groups created by gb2003 are not listed. I find this particularly problematic given the criticism of the statistical validity of group-summary statistics provided by ybe2001.23 Regardless, the absence of any summary statistics means both that the statistical validity of the compositional groups reported by gb2003 and the statistical validity of assigning the ruma jar to their Group i cannot be directly evaluated by the reader except through reference to the figures and narrative of the text.

Examination of the Figures

Gunneweg and Balla do not themselves provide any detailed plots of their data; instead, they refer the reader to graphical data displays in the supplemental report by Balazs.24 Specifically they state that Balazs’s Figs. 2 and 4 demonstrate the chemical differences between Group i and Group ii.25 As mentioned above, Balazs’s Fig. 2 does not appear to be a plot of the actual Qumran ceramic data. His Fig. 4 is, however, and this is a bivariate plot of two elements, cobalt (Co) and lanthanum (La).

Group i and Group ii of gb2003 clearly separate along the Co axis: The confidence ellipse (presumably at 90% or higher) for Group i extends from just above 15 ppm to just above 21 ppm, and the Co abundance for Group ii is lower, ranging from ca. 10 ppm to 14 ppm. Curiously, comparison of this plot alone to the data reported for qum134 indicates that ruma jar has a Co concentration (11.1 ppm) entirely consistent with gb2003’s Group ii not their Group i. Similarly, the La concentration (22.3 ppm) reported for qum134 places it within the confidence ellipse of Group ii, and well outside the confidence ellipse of Group i. Remarkably there are two specimens at this approximate location on Balazs’s Fig. 4, and both are listed as outliers of Group ii. No specimen belonging to Group i is shown at this location.

Examination of the Narrative

As detailed in the block quotation given above from gb2003 (p. 13), they assigned the ruma jar to their Group i because it did not exhibit a “high potassium and low sodium” chemical profile they believe to be unique to clay from the Motza Formation. No other reasons for refuting a Jerusalem origin are given by gb2003. What is meant by “high” and “low” in this context, and how significantly does the ruma jar differ from the Jerusalem Reference Group and the Motza clay in these elements?

The corrected chemical data reported by gb2003 (reproduced here in Table 1) indicate that the abundances of potassium (K) and sodium (Na) reported by gb2003 for the ruma jar are 2.09% +/− 0.30% and 0.41% +/− 0.01% respectively. Though not stated, it is assumed that, following statistical convention, these errors are given at one standard deviation. These numbers compare quite favorably with the analysis of the ruma jar at the hu laboratory (ybe2001, p. 76). Though ybe2001 did not report the K value, it was measured at the time and the determination is provided in Table 1: K = 2.3% +/− 0.1%, and Na = 0.36% +/− 0.01%.

It must be noted that K is an element that is not measured as well as many of the trace elements by the inaa analytical system used in hu.26 The same applies to the measurements of K that led to a value for the Jerusalem reference group. And, it applies as well to the measurement reported by gb2003. Sodium is, however, an element that is measured very well by inaa. Yet, its value as a diagnostic signature is limited, as it is a highly mobile element, it is common in the environment, and there is a priori reason to believe that potters may have added Na-rich minerals to clay during the production of pottery in this region (I return to this last point below).

In spite of these cautionary statements, the determination of K between both laboratories is excellent, overlapping at one standard deviation. Agreement between the laboratories on Na is not as strong; however, I note that Balla observed the same poor agreement in Na determinations between the Budapest and hu laboratories.27 Despite the poor agreement in Na between the two laboratories, I consider the two determinations to be within the bounds of acceptance (ca. two standard-deviations) for a mobile element such as Na. The agreement between ybe2001 and gb2003 on the composition of the ruma jar is even better if we look at the element chromium (Cr). As stated in ybe2001,28 Cr was not reported because of concern of contamination in some samples. This suspected contamination was due to chrome-plated shafts of diamond-impregnated drills that were used to sample some of the pottery.29 There was no evidence that the particular sample from the ruma jar was affected by contamination, and I have included in Table 1 the Cr value for the ruma jar. As with K and Na, it is consistent with the findings of gb2003 and consistent with Cr previously observed in Roman period pottery.30 It is also consistant with Cr in Motza clay reported by Balla31 and Perlman et al.32 as can be seen in Table 1.

Now that it is established that both the hu and the Budapest analyses of the ruma jar are comparable, it is possible to compare these data to the Jerusalem Reference Group (produced at hu) of Mommsen et al.33 and of Perlman et al.34 As can be seen in Table 1, both assays of the ruma jar provided elemental determinations that are within two standard deviations (95% confidence) of the Jerusalem Reference Group. The data in Table 1 are also displayed graphically in Figure 1. The only two elements that do not agree well are K and Na. Does this in fact support the conclusion of gb2003?

The Jerusalem Reference Group and Motza Clay

The Jerusalem Reference Group is based on the chemical analysis of a variety of Roman-period specimens of different styles found in Jerusalem. These pots have a composition very similar to each other and to Motza Formation clay from the villages of Motza and Beit ‘Ummar—with the exception that the Na value for the clay is about one fourth of the value in the Jerusalem pottery. When initially reporting these data, Mommsen et al.35 and Perlman et al.36 proposed that one possible explanation for this discrepancy could be that ancient potters in Jerusalem added salts to the Motza clay. This is a common practice among many traditional pottery-making communities.37 Moreover, Landgraf’s work among traditional potters of the Jerusalem area documented this practice.38 Thus, according to gb2003, Mommsen et al.39 and Perlman et al.40 did not “exhaustively describe” the chemical profile of the Motza Formation clay.41 Rather, they advanced a hypothesis that would explain why ancient pottery might be chemically identical to Motza Formation clay, yet exhibit significantly higher Na than the raw clay. Might it not be that the potter who made the ruma jar added salt to the clay, just as modern-day potters do?

figure 1
figure 1

Comparing ruma jar measurements from Budapest, Jerusalem and the Jerusalem reference group (Jer). The error bars are ± .

Citation: Dead Sea Discoveries 22, 2 (2015) ; 10.1163/15685179-12341348

The composition of clay from the Motza Formation varies from place to place and also with depth from the surface. However, substantial similarity in composition is expected because of similar geochemical history underlying the clay processes. This is based also on limited analysis of clay from Beit ‘Ummar, el-Jib and Motza village which Perlman and Yellin conducted in the 1970s (unpublished) as well as by investigations by Michniewicz and Krzysko42 and Michniewicz43 who concluded that clay from Hebron is not distinguishable from Jerusalem clay. This does not necessarily mean that pottery from Jerusalem cannot be distinguished from Hebron pottery.44 Pottery may contain chemical signals related to the particular spot and depth from which the clay was drawn or related to pottery manufacture techniques which could differ from one production shop to another.

It is thus difficult to fathom why gb2003 rejected a possible Jerusalem origin for the ruma jar solely on the basis of Na and K. Indeed, by the same argument concerning the ruma jar, the Jerusalem Reference Group itself should be rejected as it does not have the low Na value of the Motza Formation that gb2003 demand! More puzzling is that in discussing their Group ii, which is believed to represent pottery made of Motza clay, gb2003 (p. 14) state that “of the 18 vessels that bear a similar chemical composition as the scroll jars [and to the Motza clay], five . . . have the same composition with the uniquely high K and Na values, whereas the other thirteen pots have a similar chemical composition but lack the high K and Na abundances.” Thus, on one hand, gb2003 claim that ybe2001 are “mistaken” by not considering K and Na concentrations of the ruma jar; while on the other hand they argue that 13 of their own specimens should be considered matches to the Motza clay despite poor agreement of these two elements.

Discussion

gb2003 argue that compositional data for the ruma jar indicates that it was produced using clays locally obtained at Qumran. This finding contrasts with that of ybe2001 who concluded that the jar was manufactured of Motza clay obtained from the vicinity of Jerusalem.

The argument of gb2003 is predicated in large part on the abundances of Na and K, and they maintain that if the ruma jar came from Jerusalem it should be enriched in K and depleted in Na—characteristics of Motza Formation clay.45 The Na value for the ruma jar, as measured by both ybe2001 and gb2003, is indeed high relative to Motza Formation clay, as is the average Na concentration for the Jerusalem Reference Group. Perlman et al. advanced a hypothesis to explain why Roman-period pottery from Jerusalem (i.e., the Jerusalem Reference Group) exhibited Na abundances nearly four times higher than Motza Formation clay but was otherwise chemically identical to the clay.46

gb2003 accept this argument when it concerns 13 other specimens that they believe are made of Motza clay.47 Yet, they decry the same argument as it concerns the ruma jar. It is difficult to imagine why gb2003 reject a Jerusalem origin for the ruma jar on the basis of a Na content that is much higher than what is found in the Motza clay, particularly when this anomaly has a reasonable explanation48 and that gb2003 find that explanation acceptable when applied to other ceramic vessels.

If, as gb2003 claim, the ruma jar matches better the composition of their Group i (local to Qumran), then this must be demonstrated quantitatively. They do not provide statistics of their group compositions, and there is ambiguity in the data tables of their appendix. Quantitative evaluation is therefore not provided by them nor is it possible with the information they have provided. The graphs provided by Balazs (2003) cited by gb2003 as depicting separation of their compositional groups indicate that their data do not support their conclusions concerning the ruma jar. That is, despite being assigned to Group i in the text, the concentrations of Co and La reported for the ruma jar are well outside the confidence ellipse of Group i, and within the confidence ellipse of Group ii (associated with Motza clay).

Conclusion

Despite statements to the contrary, the evidence offered by gb2003 regarding the chemical composition of the ruma jar supports the conclusion of ybe2001: The jar was likely not produced from clay available at Qumran, and it was most likely produced from Motza formation clay and brought to the site. This conclusion is fully in line with the findings of at least two other studies of the provenance of Qumran pottery.49

Although the specifics of my discussion have been limited to the results of a single ceramic vessel, the larger implication here involves the interpretation of the compositional groups reported by Gunneweg and Balla in gb2003. Specifically, the distinction between their Group i (local Qumran) and their Group ii (Motza clay) compositions is, so far as is presented by them, attributable entirely to variation in Na and K. It is well documented that the composition of Motza Formation clay varies internally in these elements. Moreover, the Dead Sea region is one of the world’s richest sources of evaporate salt minerals—present in as halite (NaCl), sylvite/potash (KCl), and sylvinite ([Na,K]Cl). A reliance on Na and K to discriminate these groups is therefore tenuous at best. It raises the prospect that Group i and Group ii might be variants of Motza clay pottery with varying degrees of Na and K either added intentionally or adsorbed diagenetically.

The variation in composition exhibited by the Jerusalem reference group is unusually narrow so that we should not expect a perfect match for the ruma jar. Inhomogenieties in the clay, temper, and in pottery-making practices could lead the values for some elements to shift from the average. So too could post-depositional processes. Both the composition reported by gb2003 and ybe2001 are in reasonable agreement with the Jerusalem reference group.

A possible future conclusion might be that it is not possible to distinguish between pottery made in Jerusalem (from Motza clay) and pottery made elsewhere in the Motza Clay Formation (e.g. Hebron, Beit ‘Ummar) but there is no data at present to support this hypothesis. The evidence points to the Motza Formation as the raw material from which the ruma jar was made. This is consistent with the research of Michniewicz and Krzysko and of Michniewicz.50

1 J. Yellin, M. Broshi, H. Eshel, “Pottery from Qumran and Ein Ghuweir: The First Chemical Exploration of Provenience,” Bulletin of the American Schools of Oriental Research 321 (2001): 65–78 hereafter ybe2001. J. Gunneweg and M. Balla, “Neutron Activation Analysis: Scroll Jars and Common Ware,” in Khirbet Qumran and ʿAin Feshkha ii: Studies in Anthropology, Physics and Chemistry (ed. J. Humbert and J. Gunneweg; Fribourg: Academic Press 2003), 3–55 hereafter gb2003. J. Michniewicz and M. Krzysko, “The Provenance of Scroll Jars in the Light of Archaeometric Investigations,” in Khirbet Qumran and ʿAin Feshkha ii: Studies in Anthropology, Physics and Chemistry (ed. J.-C. Humbert and J. Gunneweg; Fribourg: Academic Press 2003), 59–99.

2 Yellin, Broshi and Eshel, “Pottery from Qumran and Ein Ghuweir,” 65–78; Gunneweg and Balla, “Neutron Activation Analysis,” 3–55.

3 Archaeometry Laboratory, Institute of Archaeology, The Hebrew University of Jerusalem for ybe2001 and the Technical University of Budapest for gb2003.

4 However, it should be pointed out that the two laboratories were not inter-calibrated so that there is some uncertainty in comparing measurements between the two laboratories. gb2003 (page 7) maintain that measurements made in Budapest are compatible with those of Jerusalem and I will take that as a given.

5 H. Mommsen, I. Perlman and J. Yellin, “The Provenience of the lmlk Jars,” Israel Exploration Journal 34 (1984): 89–113.

6 J. Yellin and J. M. Cahill, “Rosette Stamped Handles: Instrumental Neutron Activation Analysis,” Israel Exploration Journal 54, 2 (2004): 191–213.

7 Mommsen et al., “Provenience of the lmlk Jars;” 107–9.

8 M. Broshi, “The Archaeology of Qumran—A Reconsideration,” in The Dead Sea Scrolls: Forty Years of Research (ed. D. Dimant and U. Rappaport; Leiden: Brill, 1992), 103–15, here 114–15.

9 This does not necessarily mean that the clay itself was local to Qumran.

10 Mommsen et al., “Provenience of the lmlk Jars;” see also I. Perlman, J. Gunneweg, and J. Yellin, “Pseudo-Nabataean Ware from Jerusalem,” Bulletin of the American Schools of Oriental Research 262 (1986): 77–82.

11 Note that the khq laboratory numbers used by ybe2001 should not be confused with the KhQ numbers assigned to Khirbet Qumran artifacts by gb2003.

12 Mommsen et al., “Provenience of the lmlk Jars.”

13 gb2003, 13.

14 L. Balazs, “Archaeological Data Analysis (Data Mining),” in Khirbet Qumran and ʿAin Feshkha ii: Studies in Anthropology, Physics and Chemistry (ed. J. Humbert and J. Gunneweg; Fribourg: Academic Press, 2003), 57–59.

15 Michniewicz and Krzysko, “The Provenance of Scroll Jars.”

16 Michniewicz and Krzysko, “The Provenance of Scroll Jars,” 76.

17 “The Provenance of Scroll Jars.”

18 The actual number of specimens analyzed by gb2003 is unclear. On p. 7 they state that “In May 1998, about 30 pieces of pottery” were analyzed, and that as the project has continued 170 additional specimens were analyzed (n = 200). Later they say that 28 specimens were taken from reference materials (geological) at Qumran, as well as 166 samples from pottery (n = 194). However, the Appendix of their chapter shows chemical data for only 98 specimens.

19 Gunneweg and Balla, “Neutron Activation Analysis,” 16, 18, 22–27, also report two pottery groups that they attribute to Jericho.

20 Mommsen et al., “Provenience of the lmlk Jars.”

21 It is assumed that inaa measurements done at the Technical University of Budapest are compatible with measurement done at The Hebrew University of Jerusalem although different methods were employed.

22 M. Balla, “Provenance Study of Qumran Pottery by Neutron Activation Analysis” (Ph.D. diss., University of Budapest, 2005), 101.

23 gb2003, 5.

24 Balazs, “Archaeological Data Analysis.”

25 gb2003, 13.

26 This is due to various factors among which are the short half-life of the potassium 42 isotope (12.4 hours), the low natural abundance of the potassium 41 isotope (6.7%) and other factors.

27 Balla, “Provenance Study of Qumran Pottery.”

28 See ybe2001, 66.

29 See also M.T. Boulanger, S.S. Fehrenbach, and M.D. Glascock “Experimental Evaluation of Sample-Extraction Methods and the Potential for Contamination in Ceramic Specimens,” Archaeometry 55, 5 (2013): 880–92.

30 Cf. Mommsen et al., “Provenience of the lmlk Jars” and Perlman et al., “Pseudo-Nabataean Ware.”

31 “Provenance Study of Qumran Pottery.”

32 “Pseudo-Nabataean Ware.”

33 “Provenience of the lmlk Jars.”

34 “Pseudo-Nabataean Ware.”

35 “Provenience of the lmlk Jars,” 109.

36 “Pseudo-Nabataean Ware,” 81.

37 See, e.g., D.E. Arnold, “Ethnomineralogy of Ticul, Yucatan potters: Etics and Emics,” American Antiquity, 36(1) (1971): 20–40. M. Bonifay, Études sur la Céramique Romaine Tardive d’Afrique (Oxford: bar-is, 2004); F.P. Matson, “A Study of Temperatures Used in Firing Ancient Mesopotamian Pottery,” in Science and Archaeology (ed. R.H. Brill; Cambridge: mit 1971), 65–79; O.S. Rye & C. Evans, Traditional Pottery Techniques of Pakistan: Field and Laboratory Studies (Washington, dc: Smithsonian Institution, 1976); and W.D. Stoner et al., “Taken with a Grain of Salt: Expermintation and the Chemistry of Archaeological Ceramics from Xaltocan, Mexico,” Journal of Archaeological Method and Theory doi (2013): 10.1007/s10816-013-9179-2.

38 J. Landgraf, “La Ceramique Byzantine,” in Tell Keisan (1971–1976): Une Cite Phenicienne en Galilee (ed. J. Briend and J.-B. Humbert; Fribourg: Editions Universitaires Fribourg, 1980), 51–99.

39 “Provenience of the lmlk Jars.”

40 “Pseudo-Nabataean Ware.”

41 gb2003, 13.

42 “The Provenance of Scroll Jars,” 76.

43 J. Michniewicz, “Qumran and Jericho Pottery: A Petrographic and Chemical Provenance Study” (Uniwersytet Im. Adama Mickiewicza W. Poznaniu Seria Geologia Nr 20, 2009), 27.

44 Cf. Michniewicz, “Qumran and Jericho Pottery,” 27.

45 Although Gunneweg and Balla, “Neutron Activation Analysis,” 13, state that Na, K values is one reason they rejected a Jerusalem origin for the ruma jar, they do not give any other reason.

46 “Pseudo-Nabataean Ware.”

47 gb2003, 14, 26.

48 See, e.g., Mommsen et al., “Provenience of the lmlk Jars” and Perlman et al., “Pseudo-Nabataean Ware.”

49 Michniewicz, “Qumran and Jericho Pottery” and Michniewicz and Krzysko, “The Provenance of Scroll,” 59–99.

50 Michniewicz and Krzysko, “The Provenance of Scroll” and Michniewicz, “Qumran and Jericho Pottery.”

  • 1

    J. Yellin, M. Broshi, H. Eshel, “Pottery from Qumran and Ein Ghuweir: The First Chemical Exploration of Provenience,” Bulletin of the American Schools of Oriental Research 321 (2001): 65–78 hereafter ybe2001. J. Gunneweg and M. Balla, “Neutron Activation Analysis: Scroll Jars and Common Ware,” in Khirbet Qumran and ʿAin Feshkha ii: Studies in Anthropology, Physics and Chemistry (ed. J. Humbert and J. Gunneweg; Fribourg: Academic Press 2003), 3–55 hereafter gb2003. J. Michniewicz and M. Krzysko, “The Provenance of Scroll Jars in the Light of Archaeometric Investigations,” in Khirbet Qumran and ʿAin Feshkha ii: Studies in Anthropology, Physics and Chemistry (ed. J.-C. Humbert and J. Gunneweg; Fribourg: Academic Press 2003), 59–99.

    • Search Google Scholar
    • Export Citation
  • 2

    Yellin, Broshi and Eshel, “Pottery from Qumran and Ein Ghuweir,” 65–78; Gunneweg and Balla, “Neutron Activation Analysis,” 3–55.

    • Search Google Scholar
    • Export Citation
  • 6

    J. Yellin and J. M. Cahill, “Rosette Stamped Handles: Instrumental Neutron Activation Analysis,” Israel Exploration Journal 54, 2 (2004): 191–213.

    • Search Google Scholar
    • Export Citation
  • 8

    M. Broshi, “The Archaeology of Qumran—A Reconsideration,” in The Dead Sea Scrolls: Forty Years of Research (ed. D. Dimant and U. Rappaport; Leiden: Brill, 1992), 103–15, here 114–15.

    • Search Google Scholar
    • Export Citation
  • 16

    Michniewicz and Krzysko, “The Provenance of Scroll Jars,” 76.

  • 37

    See, e.g., D.E. Arnold, “Ethnomineralogy of Ticul, Yucatan potters: Etics and Emics,” American Antiquity, 36 (1) (1971): 20–40. M. Bonifay, Études sur la Céramique Romaine Tardive d’Afrique (Oxford: bar-is, 2004); F.P. Matson, “A Study of Temperatures Used in Firing Ancient Mesopotamian Pottery,” in Science and Archaeology (ed. R.H. Brill; Cambridge: mit 1971), 65–79; O.S. Rye & C. Evans, Traditional Pottery Techniques of Pakistan: Field and Laboratory Studies (Washington, dc: Smithsonian Institution, 1976); and W.D. Stoner et al., “Taken with a Grain of Salt: Expermintation and the Chemistry of Archaeological Ceramics from Xaltocan, Mexico,” Journal of Archaeological Method and Theory doi (2013): 10.1007/s10816-013-9179-2.

    • Search Google Scholar
    • Export Citation
  • 38

    J. Landgraf, “La Ceramique Byzantine,” in Tell Keisan (1971–1976): Une Cite Phenicienne en Galilee (ed. J. Briend and J.-B. Humbert; Fribourg: Editions Universitaires Fribourg, 1980), 51–99.

    • Search Google Scholar
    • Export Citation
  • 44

    Cf. Michniewicz, “Qumran and Jericho Pottery,” 27.

  • 49

    Michniewicz, “Qumran and Jericho Pottery” and Michniewicz and Krzysko, “The Provenance of Scroll,” 59–99.

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 125 14 0
Full Text Views 186 1 0
PDF Views & Downloads 23 3 0