Correlating Clays, EB-Persian Pottery Fabrics, and Hellenistic-Roman Fabrics of the Southern Valley and : an Interim Report

Joseph Weinstein AIAR April 2018 Goals and Methods

• Collect all available • Identify potentially local published & archival data pottery by comparison to about soils, clays, and clay sources and a multi- period ”principal of relative ancient pottery of the abundance” Southern Jordan Valley and Dead Sea • Establish potential geographic spread of • Correlate petrographic and manufacturing centers by chemical classifications comparison to chemical (LBNL, MURR, etc.) & • Correlate pottery fabrics of petrographic (LCP, etc) various periods with each databases & geological other and with clay sources maps

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 2 Composition Sites

Chalcolithic – Persian Hellenistic-Roman

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 3 Composition Major Studies

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 4 Composition Labs & Data Sources: Chemistry Laboratory Method Sites Periods Samp Source Size BNL INAA Jericho, Safi, Feifa, MB 27 McGovern 2000 Bab edh-Dhra BNL INAA Jericho MB 29 Kaplan 2000 + MURR archive LBNL INAA Jericho LB 21 TDAR archive Manchester INAA Tell Iktanu, Jericho EB IV 172 Newton 1995 + MURR website MURR INAA Jericho LB 16 MURR archive MURR INAA Tell Nimrin MB – 26 McGovern 1988 Persian + MURR archive MURR INAA Iraq al-Amir Iron – 109 MURR archive Hellenistic SUNY Buffalo INAA Bab edh-Dhra EB IV unus Kipler-Koch able 1989 J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 5 Composition Labs & Data Sources: Optical Petrography

Laborator Method Sites Periods Samp Source y Size Optical Bab edh-Dhra’, EB IV ? Beynon, Petrogra Numeira Donahue 1986 phy Optical Qumran Iron 7 Master (LCP) Petrogra phy HU Optical Ein Gedi Chalcolithic ? Goren Petrogra phy Leiden Optical Jericho (Tell es- Neolithic, 150? Franken 1974 Petrogr Sultan) Iron II-III Braekman this aphy session

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 6 Composition Archival INAA Data Issues

• Sampling Issues • MURR and LBNL • Small, non-random samples measurements • Special goals, not necessarily representative appear to be excellent, but • Methodology varies much data missing • Measured elements differ • BNL • Precision unknown measurements • Intercalibration uncertain cover limited set • Information often missing of elements and • Find context precision is • Style/ware family uncertain • Manchester • Data Corruption measurements • Position of decimal point are badly • Elements and/or isotopes interchanged corrupted but • Unidentified items intermixed with pottery largely restorable • Original Classifications Questionable • SUNY Buffalo data is uninterpretable

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 7 Composition New Analytic 60 400 350 50 300 40 Methodology 250 30 200 150 • Log-Ratio (Aitchison) Transform/ 20 Chromium (ppm) Chromium

Lanthanum(ppm) 100 10 Relative Atomic Variation 50 • Convert N measurements [X ] and 1 residual 0 0 i 0 Scandium10 (ppm)20 30 0 Scandium10 (ppm)20 30 term to N log-ratios log([Yi]/ [Xj]) • Choose the N binary log-ratios to minimize Log-Ratio Transform of Lanthanum, intrinsic statistical dependencies Chromium, & Scandium. Data: Michniewicz. • Measurement uncertainty

• Grain Size Dependence (“Dilution effect”) 40 Group I (Silty, Foraminiferous) • Gain or Loss of Volatiles Group II (Motza + Calcareous) Group III (Motza + Quartz) • Mechanisms of incorporating trace elements Group IV (Terra Rossa) • Define inner product & metric Motza Clays Terra Rossa Clays (Hebron) • Unit variance across typical chemical group 20 Terra Rossa Clays () • Minimal co-variances Qumran Clays • Identify chemical groups by multivariate Other Clays statistical techniques

• Principal components analysis (PCA) Chromium/Scandium 10 • Agglomerative Hierarchical Clustering (AHC) • Identify and interpret chemical groups graphically • RAV scattergrams comparing two ratios 5 • Ternary scattergrams comparing any three 1 2 4 components Lanthanum/Scandium

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 8 Composition • Mathematically preferred technique for Advantages analyzing compositional data 20 • Matches data to the assumptions = Group I (Silty, Foraminiferous) underlying most multivariate statistical = Group II+IIIa (Motza) techniques = Group II+IIIb (Motza)15 = Group IV (Terra Rossa) • Avoids artifacts arising from the = Motza Clays constraint that components must add to 100% 10 • Ratios can be chosen so that they are unaffected by the grain size distribution 5 and by gain or loss of volatiles (“dilution effect”) F2 • Minimizes covariances within each 0 chemical group -15 -10 -5 0 5 10 15 20 25 • Simplifies identification, description, -5 and interpretation of chemical groups • Chemical groups can be adequately

-10 described by means and variances alone • Chemical groups can often be recognized graphically on plots of just two well- -15 chosen ratios F1 • Radically improves the performance of most multivariate classification Revised AHC analysis of Michniewicz’ techniques data, projected onto revised PCA • Pottery and clays can be compared directly analysis of the same data

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 9 Composition Local Clay Sources Rendzina, Terra Rossa soils

Rendzina, Terra Rossa soils AeolianLoessLoess Dust

Motza TuronianCenomanian Shueib Fm ShueibFm Fm Alluvial Fans Cenomanian & Naur Fms Lisan Fm Turonian Nezer Fm Lower CenomanianCenomanian Cretaceous Clays HazeraHazera FmFm DeadDead SeaSea BlackBlack MudMud

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 10 Composition Local clay sources – Chemistry

Alluvium, Wadi Qumran, downstream from Ghareb Fm Very high [Cr]/[Sc]

Lower Cretaceous Clays, Ghor al- Khabid (data from Mahis) [La]/[Sc] > 3.5

Rendzinal Soil: Jericho, Iktanu, & Iraq al-Amir Resembles Negev Loess: [La]/[Sc] ≈ 2.5-3.0 Cenomanian Naur, Fuheis, or Shueib Fm, Iraq al-Amir Cenomanian Hazera Fm, Ein Gedi Resembles Moza Fm: [La]/[Sc] ≈ 1.5

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 11 Composition EB-Persian Pottery - Chemistry

Rendzinal Soil or alluvium, note variability in [Cr]/[Sc]. Cooking pots, probably Terra Rossa with or without calcareous temper No known local source Cooking pots, probably Lower Some could be local loess Cretaceous clays with or without Others are very low calcium calcareous temper, but much lower [La]/[Sc] than Mahis clays

Pottery from Cenomanian Clays Large Group, All Sites, All Periods, All Types Local, Cisjordanian, & Transjordanian Clay Sources

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 12 Composition Hellenistic+Roman Qumran & Jericho – Michniewicz

Wadi Qumran Alluvium & Reference Items Not used for pottery ABSENT: Pottery from Lower Cretaceous Clays Rendzinal Soil Pottery (Michniewicz Group 1) Resembles EB-Persian Rendzinal Soil Pottery

Cooking Pots (Michniewicz Group 4) Resembles EB-Persian Terra Rossa Cooking Pots without calcareous temper Cenomanian Clay Pottery (Michniewicz Groups 2 & 3) Includes almost all “Scroll Jars” Resembles EB-Persian Cenomanian Clay Pottery

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 13 Composition Hellenistic+Roman Qumran – Balla

Wadi Qumran Alluvium & Reference Items Not used for pottery ABSENT: Pottery from Lower Cretaceous Clays Rendzinal Soil Pottery (Balla Group 4) Resembles EB-Persian Rendzinal Soil Pottery

Cooking Pots (probable Terra Rossa) Resembles EB-Persian Terra Rossa Cooking Pots without calcareous temper Cenomanian Clay Pottery (Balla Groups 1, 2, & 4) Includes almost all “Scroll Jars” Resembles EB-Persian Cenomanian Clay Pottery

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 14 Composition Petrographic Correlation - Rendzinal Soil

Jericho Neolithic

Michniewicz Group 1

LCP Fabric 7 (Qumran Iron 18)

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 15 Composition Petrographic Correlation – Dolomitic Cenomanian

Jericho Iron II-II

Michniewicz Group 2

Qumran Iron 1a Also: Nahal Mishmar, Ein Gedi

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 16 Composition Petrographic Correlation – Cenomanian w/Quartz Sand

Michniewicz Group 3

Bab edh-Dhra’ Nubian Sand Temper Chemistry may differ

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 17 Composition Petrographic Correlation – Terra Rossa

Michniewicz Group 4

Qumran Iron 2

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 18 Composition Petrographic Correlation – Wadi Qumran

Wadi Qumran Alluvium No Hellenistic/Roman Pottery

No EB – Persian Pottery

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 19 Composition Geographic Spread: ?Hamra

Recent Aeolian Clays Some Terra Rossa,

vs 20.00 Rendzina

18.00 Hamra Desert

16.00 Southern Rendzina Coastal 14.00 Loess

12.00 Negev

10.00 Loess Chromium/Scandium

8.00

6.00

4.00

2.00

0.00 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 LBNL Database Lanthanum/Scandium

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 20 Composition Geographic Spread: Cenomanian-Turonian Megiddo Clays Shechem

vs 20.00 Bethel Tell Nimrin 18.00 Gibeon Iraq el-Amir

16.00 Jerusalem Tall Umayri 14.00 Jericho Hisban

12.00 Qumran Bab ed-Dhra 10.00 Hebron Chromium/Scandium Safi, Feifa 8.00 Tell Masos Tawilan 6.00 Avdat

4.00 Petra Timna 2.00

0.00 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 LBNL Database Lanthanum/Scandium

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 21 Composition Geographic Spread:

Taqiye Fm & Ghareb Megiddo

vs 20.00

18.00

16.00 Wadi Ramle 14.00 Qumran

12.00

10.00 Chromium/Scandium

8.00 Arad

6.00

4.00

2.00

0.00 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 LBNL Database Lanthanum/Scandium

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 22 Composition Implications and Conclusions

• Chemical groupings reflect • Pottery clays used at Qumran resemble those used in Southern Jordan Valley geological clay sources more since time immemorial, except that precisely than specific Lower Cretaceous clays were not used manufacturing centers. • There are known local sources for all these pottery clays, except for the Terra • Within each geological clay Rossa used for many cooking pots source there are often smaller • Use of a Cenomanian (“Motza”) clay is differences that may represent not sufficient to demonstrate Jerusalem provenance different manufacturing • Potential local sources (Hazera Fm; centers Fuheis, Shueib, or Naur Fm) • Employed locally since time immemorial • The method of “Relative • We currently know very little about internal chemical or petrological Atomic Variation”/log-ratio variability within these Cenomanian clays transform provides a simple, useful graphical way for comparing pottery

J. Weinstein, Southern Jordan Valley Pottery AIAR April 29, 2018 23 Composition