UC San Diego Electronic Theses and Dissertations
Peer Reviewed Title: Technology and society : some insights on the development of metallurgy in the Southern Levant in the light of new dates of slag deposits Author: Ben-Yosef, Erez Acceptance Date: 01-01-2008 Series: UC San Diego Electronic Theses and Dissertations Degree: M.A., UC San Diego Permalink: http://escholarship.org/uc/item/03f2f3vx Local Identifier: b6636008 Abstract: An ongoing project for reconstructing the behavior of the geomagnetic field intensity during the last seven millennia has yielded several new dates for archaeometallurgical sites in the Southern Levant. These dates shed new light on the dawn of metallurgy in the region as well as on the quality of technological development and its relation to social and political structures. This paper introduces the methodology and concepts behind the archaeomagnetic project as well as the principles of the applied dating technique. In addition, the paper presents the archaeomagnetic results, discusses the alternative dating of several archaeometallurgical sites and explores the implication of these results on our understanding of the interaction between technology and society in the past. For the latter, the results particularly challenge the "Standard View of Technology" (Pfaffenberger, 1992), and suggest a complex, nonlinear evolution of copper industry in the Southern Levant

eScholarship provides open access, scholarly publishing services to the University of California and delivers a dynamic research platform to scholars worldwide.

UNIVERSITY OF CALIFORNIA, SAN DIEGO
Technology and Society: Some Insights on the Development of Metallurgy in the Southern Levant in the Light of New Dates of Slag Deposits

A Thesis submitted in partial satisfaction of the requirements for the degree Master of Arts

in

Anthropology

by

Erez Ben-Yosef

Committee in charge:

Professor Thomas E. Levy, Chair Professor Guillermo Algaze Professor Geoffrey E. Braswell

2008

The Thesis of Erez Ben-Yosef is approved, and it is acceptable in quality and form for publication of microfilm:

Chair

University of California, San Diego 2008

iii

TABLE OF CONTENTS Signature page ……………………………………………………………………… Table of Contents …………………………………………………………………... List of Figures ……………………………………………………………………… List of Tables ………………………………………………………………………. Acknowledgments ………………………………………………………………….. Abstract …………………………………………………………………………….. Introduction ………………………………………………………………………… Part I: Reconstructing the geomagnetic intensity in an archaeological context …… Atchaeointensity research ……………………………………………………... Thellier-Thellier experimental design and data interpretation ……………. Representation of geomagnetic intensity results – a comment about units … The contribution of archaeology to geomagnetic intensity research ……….. Archaeointensity in archaeometallurgical context …………………………….. Copper slag as an archaeointensity recorder ……………………………… Other artifacts from archaeometallurgical context ………………………... Seven millennia of geomagnetic intensity changes in the southern Levant …... Research methodology ……………………………………………………. Results …………………………………………………………………….. Part II: Archaeological and anthropological implications …………………………. Archaeometallurgy in the Southern Levant and the problem of dating ……….. Implications on dating of archaeometallurgical sites ………………………….. Timna 39b ………………………………………………………………… Archaeometallurgical sites from later periods ……………………………. Technological development in a wide perspective ……………………………. The Standard View of technology ………………………………………... The development of copper metallurgy in the southern Levant ………….. Conclusions ………………………………………………………………………… Archaeointensity in the Levant – New Horizons …………………………. Archaeointensity as a Dating Tool ………………………………………... Challenging the Standard View of Technology …………………………... Appendix I: Figures ………………………………………………………………... Appendix II: Tables ………………………………………………………………... References ………………………………………………………………………….. iii iv v vi vii viii 1 2 2 5 9 10 11 11 13 14 14 17 19 19 23 23 27 32 32 33 38 38 39 40 42 51 54

iv

LIST OF FIGURES

Figure 1: of archaeointensity data from the Near and Middle East for the last seven millennia (after Ben-Yosef et al., 2008) …………………………………... Figure 2: The Earth’s magnetic field and its elements ………………………....... Figure 3: A graphic representation of the “Thellier-Thellier” type experiment …. Figure 4: Embedded charcoal in a slag sample …………………………………... Figure 5: Slag deposits in the Southern Levant ………………………………….. Figure 6: Slag attached to the walls of in-situ furnaces ………………………….. Figure 7: Examples of slag samples …………………………………………........ Figure 8: Examples of baked clay artifacts from archaeometallurgical context …. Figure 9: The major copper production centers in the Southern Levant ………… Figure 10: Archaeometallurgical sites that were sampled in the current study ….. Figure 11: Summary of all acceptable sample intensities results in this study …... Figure 12: copper smelting installation in site Timna 39b and the excavated area surrounding it …………………………………………………………………...... Figure 13: Curve combining Syrian and southern Levantine archaeointensity results …………………………………………………………………………….. Figure 14: Schematic diagram showing two models for the development (evolution) of technology, from simple to complex for a given region (or society)

42 43 43 44 44 45 45 46 46 47 48

49

49

50

v

LIST OF TABLES

Table 1: Archaeometallurgical sites and samples in this study …………………….

52

Table 2: Reliable archaeointensity results from well-dated archaeometallurgical sites in the Southern Levant ………………………………………………………...

53

Table 3: Reliable archaeointensity results from poorly dated or controversial archaeometallurgical sites in the Southern Levant …………………………………

53

vi

ACKNOWLEDGEMENTS

First I would like to acknowledge the most important advisors of this work, Professor Hagai Ron (Hebrew University of Jerusalem [HUJI]), Professor Amotz Agnon (HUJI), Professor Lisa Tauxe (University of California, San Diego [UCSD]) and Professor Thomas E. Levy (UCSD). Without their invaluable contribution, this work could never see its successful completion. It is a pleasure to thank the following people as well: Jason Steindorf for many of the measurements, Angès Genevey for her contribution to the experimental part of this work, and Uzi Avner for his insights in the field. Thanks are also due to Zeev Herzog, Assaf Holtzer, Michael Levy, Ron Shaar, Sariel Shalev, Naama Yahalom and Yuval Yekutieli for help in various aspects of this research. We are grateful to Mohammad Najjar and the Jordanian Department of Antiquity for their help with the field work. This study was supported by the FIRST program of the Israel Science Foundation Grant No. 1334/05, US-Israel Binational Science Foundation Grant No. 2004/98, NSF grant EAR0636051 and the Academic Senate of UCSD.

vii

ABSTRACT OF THE THESIS
Technology and Society: Some insights on the development of metallurgy in the Southern Levant in the light of new dates from slag deposits

by

Erez Ben-Yosef Master of Arts in Anthropology University of California, San Diego, 2008 Professor Thomas E. Levy, Chair

An ongoing project for reconstructing the behavior of the geomagnetic field intensity during the last seven millennia has yielded several new dates for archaeometallurgical sites in the Southern Levant. These dates shed new light on the dawn of metallurgy in the region as well as on the quality of technological development and its relation to social and political structures. This paper introduces the methodology and concepts behind the archaeomagnetic project as well as the principles of the applied dating technique. In addition, the paper presents the archaeomagnetic results, discusses

viii

the alternative dating of several archaeometallurgical sites and explores the implication of these results on our understanding of the interaction between technology and society in the past. For the latter, the results particularly challenge the “Standard View of Technology” (Pfaffenberger, 1992), and suggest a complex, nonlinear evolution of copper industry in the Southern Levant.

ix

Introduction This work consists of two major parts. The first part presents an interdisciplinary enterprise aims to reconstruct the geomagnetic intensity using archaeometallurgical artifacts from dozens of sites in Israel and Jordan. This is a collaboration of archaeologists, geologists and geophysicists from Israel, Jordan and the USA in an attempt to better understand one of the most fascinating geophysical phenomena as it was recorded in the archaeological deposits. I conducted the field work in Israel and Jordan during the years 2004-2005. The laboratorial analysis was carried out in the University of California, San Diego, during the years 2006-2007. The second part of this work discusses archaeological and anthropological implications of the archaeomagnetic experiments and data. In this section I present the archaeometallurgical research in the Southern Levant and suggest new dates for several key archaeometallurgical sites in the light of the archaeomagnetic experiments results. These new dates contribute to the discussion about the development of metallurgy in the region as well as to our general understanding of social anthropology of technology.

1

PART I: RECONSTRUCTING THE GEOMAGNETIC INTENSITY IN AN ARCHAEOLOGICAL CONTEXT

Archaeointensity research, the study of the intensity of the geomagnetic field as recorded by archaeological artifacts, has produced a vast amount of data since the 1950s (Donadini et al., 2006; Thellier & Thellier, 1959). The data for the Near and Middle East obtained heretofore are inconsistent and highly scattered (Figure 1), principally as a result of different materials used, varying experimental methods, different standards for evaluating the reliability of the laboratorial results and poor time constraints. This inconsistency, together with low data resolution in certain periods, influences the accuracy and predictability of derived geomagnetic field models (notably CALSK7.2, see Korte & Constable, 2005a, 2005b; Korte et al., 2005, and dashed line in Figure 1). In search of increasing reliability of archaeointensity data we conducted a systematic investigation of a virtually unexploited recording medium, namely copper slag deposits. Together with a recently improved experimental design for the archaeointensity experiment, we demonstrated the suitability of this medium, as well as other materials from archaeometallurgical context, for archaeointensity studies (Ben-Yosef et al., 2008).

Archaeointensity research Although the Earth’s magnetic field has been recognized for nearly 2000 years (Kono, 2007), it is still one of the least understood geophysical phenomena. Its behavior provides insights on the inner workings of the Earth, including geodynamics of the early

2

3 planet and changes in boundary conditions through time. Its strength modulates the amount of cosmic radiation hitting the Earth, thus contributing to factors such as the production of cosmogenic isotopes in the atmosphere (including radiocarbon, see e.g. Frank, 2000; Kitagawa & Plicht, 1998; Peristykh & Damon, 2003) and potentially even climatic changes (e.g. Courtillot et al., 2007; Gallet et al., 2005). The geomagnetic field is dynamic and undergoes random changes. Small scale variations (known as “secular variations”) occur constantly, independent of the larger scale directional changes of reversals and excursions (e.g. Yamazaki & Oda, 2004). They show similar characteristics over an areal extent of order 103 km, and they consist of significant non-dipolar components whose magnitudes are debated (e.g. Constable et al., 2000; Courtilot et al., 1992). Reconstructing geomagnetic field behavior for the last several millennia focuses on studying its secular variations, and thus depends strongly on position. Improved prediction of geomagnetic field vectors awaits more sophisticated archeosecular variation models, based on reliable data from various regions of the world. Comprehensive investigation of the geomagnetic field requires full vector information for a known point in time (Figure 2). For the directional components there are instrumental records for the last 400 years, and for the intensity we have records since 1830s, all included in the GUFM model of Jackson et al. (2000). Reconstructing the geomagnetic field prior to the instrumental recording depends on geological and archaeological recorders. In most cases, these recorders are volcanic rocks and archaeological artifacts that acquired a thermal remanent magnetization (TRM) after cooling from Curie temperature (usually in the range of 300°-600°C). Materials like basaltic rocks, pottery sherds and fired clay bricks are examples of paleomagnetic and

4 archaeomagnetic recorders which preserve the properties of the geomagnetic field from the last moment of cooling. While reconstructing the directional properties of the geomagnetic field is a relatively easy procedure, extracting the ancient intensity is a complex and laborious process (Valet, 2003). In principle, it is possible to determine the intensity for ancient magnetic fields because the primary mechanisms by which rocks and artifacts become magnetized can be approximately linearly related to the ambient field for low fields such as the Earth’s. Thus, we have by assumption

M NRM ≅ α anc H anc and M lab ≅ α lab H lab
Where α lab and α anc are dimensionless coefficients; M NRM and M lab are natural (i.e. original) and laboratory remanent magnetizations, respectively; and H anc and H lab are the magnitudes of the ancient and laboratory fields, respectively. If α lab and α anc are equal, we can divide the two equations and rearrange terms to get
H anc = M NRM H lab M lab

In other words, if the laboratory remanence has the same proportionality constant with respect to the applied field as the ancient one, the remanences are linearly related to the applied field, and the natural remanence (NRM) is composed solely of a single component, all one needs to do to get the ancient field is measure M NRM , give the

5 specimen a laboratory proxy remanence M lab and multiply the ratio between them by H lab . In practice, paleointensity is not so simple. The remanence acquired in the laboratory may not have the same proportionality constant as the original remanence (e.g., the specimen has altered its capacity to acquire remanence or was acquired by a mechanism not reproduced in the laboratory). The assumption of linearity between the remanence and the applied field may not hold true. Or, the natural remanence may have multiple components acquired at different times with different constants of proportionality. A sophisticated experimental design is needed for validating the basic assumptions of the method, for tracking changes in the magnetic characteristic of a specimen throughout the experiment and for evaluating the reliability of the intensity results. For materials with thermal remanent magnetization, the most common

experimental design derives from the basic "Thellier-Thelier" experimental protocol (Thellier, 1938; Thellier & Thellier, 1959).

Thellier-Thellier experimental design and data interpretation

The theoretical basis for how ancient magnetic fields might be preserved was clarified with the Nobel Prize-winning work of Néel (1949; 1955). The theoretical basis for the experiments, including detailed description and comparison with other methods was recently reviewed by Tauxe and Yamazaki (2007). Here we present only the basic principles of the laboratory work and data interpretation as a background for the archaeological discussion.

6 The basic experiment involves heating specimens up in stages, progressively replacing the NRM with partial thermal remanences (pTRMs) in the hope of establishing the ratio M NRM / M lab prior to the onset of alteration. This step-wise approach relies on the assumptions that pTRMs acquired by cooling between any two temperature steps are independent of those acquired between any other two temperature steps, and that the total TRM is the sum of all independent pTRMs. There are several options for ordering the sequential steps. For simplicity, the one of Coe (1967) is presented here. In the first step the specimen is heated to some temperature and cooled in zero field. The measurement of the specimen will give us

M first = M NRM _ remaining
As an illustration, we plot MNRM_remaining for a series of temperature steps as the blue line in Figure 3a. In the second step the specimen is heated again to the same temperature and cooled in the laboratory field, H lab . The measurement of the combined remanence (what is left of the natural remanence plus the new laboratory pTRM) is

M sec ond = M NRM _ remaining + pTRM
Simple vector subtraction allows the determination of the pTRM for this temperature step. The pTRM of each temperature stage is plotted as the red line in Figure 3a. Then, we plot the pTRMs against the relevant NRMremaining, and the result is a useful diagram ("Arai plot", Nagata, 1961) for analyzing the behavior of the specimen throughout the experiment (Figure 3b). The proportion between the pTRM and the NRMremaining should be constant, and the slope of the line is the desired proportional constant

7
NRM remainging − NRM pTRM

slope =

⇒ H anc = H lab × slope

Additional steps in the same or lower temperatures provide tests for various reliability checks in the experiment. For example, repeating a lower temperature step and checking the pTRM acquired (“pTRM check”) indicates if the ability to acquire a pTRM has changed during the experiment. Demagnetizing the specimen after it acquired a pTRM in the same temperature (“pTRM tail check”) checks whether the blocking temperature is equal to the unblocking temperature, an important prerequisite for reliable intensity results. These tests can be represented on the Arai plot (Figure 3b). Interpretation of the results has to take into account numerous factors, and should be done for each specimen separately. First, the segment of the experiment which represents the ancient magnetic field should be identified, usually using standard demagnetization vector end-point diagrams for determining the original magnetic component and Arai plot for spotting alteration. Then the reliability of the relevant segment should be evaluated, using aspects such as the linearity of the line in the Arai plot, the results of the relevant pTRM and pTRM-tail checks, the number of data points in the relevant segment and others. Many of these aspects can be quantified, and different combinations are used as “selection criteria” for determining a reliable intensity results (see review in Tauxe, 2006). The criteria used and their acceptance values vary among different studies, and they depend on the experimental protocols, the materials used and the personal methodology of the researcher.

8 There are several other considerations regarding the reliability, precision and accuracy of the intensity results. For example, if the specimen is anisotropic with respect to the acquisition of thermal remanence, the anisotropy tensor must be determined and intensity corrected (e.g. Aitken et al., 1981; Selkin et al., 2000). Moreover, because the approach to equilibrium is a function of time, slower cooling results in a larger TRM; hence differences in cooling rate between the original remanence acquisition and that acquired in the laboratory will lead to erroneous results (e.g. Fox & Aitken, 1980). Compensating for differences in cooling rate is relatively straight forward if the original cooling rate is well known and the sample behaves according to single-domain theory. This theory derives its name from the distribution of atomic magnets within the macroscopic sample, where no domains of mutually contradicting magnetization might cancel each-other. Alternatively, one could take an empirical approach in which the specimen is allowed to acquire a pTRM under varying cooling rates, an approach useful for cooling periods of up to a day or two. For pottery fragments, originally cooled inside kilns, the over-estimation was shown experimentally to be by as much as 15-20% with an original cooling time of a day (from the Curie temperature) and an experimental cooling time of half hour (Genevey & Gallet, 2002). Finally, the intensity results should be evaluated in the sample level, according to the agreement between different specimens from the same original sample (i.e. the standard deviation cut-off). Usually a minimum number of “well-behaved” specimens per sample (N) is also determined as an additional cut-off value.

9
Representation of geomagnetic intensity results – a comment about units

The Système international (SI) basic unit for representing magnetic induction (B) is tesla (T). Induction is often used interchangeably with the term magnetic field (H), with units of A/m because in cgs units there is no difference between field and induction. While there is a significant difference in SI units (a factor of µo, or 4πx10-7henries/m), most researchers for simplicity, continue to refer to the induction as the magnetic field but quote values in tesla. For the Earth’s magnetic field, which is relatively weak, it is convenient to use µT. The field varies strongly as a function of latitude, as expected from an essentially dipolar field (which is twice as strong at the poles than at the equator). Therefore, when comparing data from different localities (i.e. different

longitudes/latitudes) in the same region, it is useful to ‘reduce’ them to a reference latitude, by simple manipulation (e.g. Odah et al., 1995)
⎛ 4 − 3 cos 2 θ reduced = Bsite ⎜ ⎜ 4 − 3 cos 2 θ site ⎝ ⎞2 ⎟ ⎟ ⎠
1

Breduced

A more common way to compare geomagnetic intensity data from different localities and regions is by presenting them as virtual axial dipole moment (VADM)
VADM = 4πr 3 Bancient 1 + 3 cos 2 θ

µ0

(

)

−

1 5

(Where r = Earth's radius [~6372000m]; µ0 = permeability of free space constant; θ = colatitude). Magnetic moments (as the VADM) are measured in Am2 so magnetic fields (A/m) can be thought of as volume normalized magnetic moments. Conversion to VADM eliminates the effect of the dipole on intensity and allows the possibility of

10 regional differences derived from sources of non-dipole moments to be assessed. Represented as VADM, the current geomagnetic intensity is 77.8 ZAm2 (Zeta=1021).

The contribution of archaeology to geomagnetic intensity research

Understanding the behavior of the geomagnetic field’s intensity over the last millennia is a key for studying various related phenomena, such as solar activity (e.g. Usoskin et al., 2006), the production of radiocarbon and other cosmogenic isotopes (e.g. Peristykh & Damon, 2003), the mechanisms of the geomagnetic field itself (e.g. Constable et al., 2000) and perhaps even climate changes (e.g. Courtillot et al., 2007). Moreover, the geomagnetic field has significantly reduced in strength over the last few decades, leading to speculation that it could collapse entirely as it undergoes a reversal of polarity (Constable & Korte, 2006; Hulot et al., 2002). The decay of the field has been observed since the beginning of instrumental recording over 160 years ago (Bloxham, 2003), yet a better understanding of the geomagnetic intensity throughout the last millennia is needed for assessing the nature of the recent change. For the last millennia, as for the entire Holocene, the best source for reconstructing the secular variations of the geomagnetic field derives from the archaeological context (Folgheraiter, 1899; Thellier, 1938). Since the innovation of pyrotechnological industries in the Neolithic, heated materials are abundant in the archaeological record. The most commonly used archaeomagnetic recorders are artifacts of baked clay, typically pottery sherds, fired mud bricks and kilns’ walls (see examples in Figure 1 and caption). The primary advantage of these recorders is the ability to determine their age by the archaeological context. For young (1100ºC) and in many cases have unique tempering and complex structure as making them resistant to the smelting and melting processes. Thus, clay samples from archaeometallurgical context are distinct from the commonly used baked clay artifacts such as pottery sherds (typically baked between 400-800ºC) and fired mud bricks. As part of the current study we also measured 28 specimens derived from five samples of refractory ceramics from archaeometallurgical sites in the Southern Levant (Ben-Yosef et al., 2008). The experiments yielded successful results for 25 specimens (~89% success rate), and for all of the samples (using rigorous selection criteria of N>2 and σ ≤ 10%). Although the number of clay samples was small, the results indicate that they are highly suitable for archaeointensity studies. We hope to test this observation with a much larger sample of refractory clay objects in the future.

Seven millennia of geomagnetic intensity changes in the southern Levant
Research methodology

As part of an investigation into slag material as an archaeointensity recorder and in an effort to improve the resolution and reliability of the geomagnetic intensity curve for the last seven millennia, we collected slag, furnace and crucible fragments from 27

15 archaeometallurgical sites in Israel and Jordan (Figure 10, table 1). Most of the samples were collected during a field survey from a variety of archaeological contexts, and others were taken from collections of previous archaeological excavations with the exact locations well known (e.g. the sites of Shiqmim and Khirbat Hamra Ifdan), providing the best reference for further analysis. The main criteria used for choosing the sites were: 1) dating quality, with priority given to sites that have well established archaeological dating or reliable results from radiocarbon measurements, 2) sites from periods that have distinct geomagnetic archaeointensity trends in previous studies, such as the conspicuous peak in the Iron Age (ca. 3,000 years ago) and the low in the Chalcolithic – Early Bronze Age (ca. 5,500 years ago), and 3) sites in which paleointensity data might help solve questions concerning the history of metallurgical technology, such as Tima 39b. All of the dates assigned to our samples are based on prior archaeological investigation of the sites. We have not measured radiocarbon samples in this stage of the research, although in many cases associated charcoal pieces are abundant and might be used in the future. The archaeological context constraining the age information of the sample collection (see Table 1) is of variable quality, depending on the collection method and the previous archaeological work. We have developed a scheme for characterizing the age uncertainty of a sample based on the complex reality of archaeological investigation in our research area. While the age assigned might be precise (i.e. having a small deviation from the mean), the archaeological context tying a given sample to a given age may be weak or controversial. In order to characterize the context itself, we make use of various objective categories that relate to the methods of the original dating

16 (e.g. radiocarbon measurements versus ceramic typology), the characteristic of the site (e.g. presenting multi-periods or single period) and our sample collection strategy (e.g. from confined excavated loci or surface survey). To summarize the relative reliability of our samples ages, we have assigned each age a number from 1 to 6 whereby 1 is considered excellent and 5 is poor. Controversial sites are assigned a number 6. For the purposes of geomagnetic field modeling, only the samples with quality determinations of 1 and 2 should be considered. The results from the rest of the samples are part of the discussions on the quality of slag as an archaeointensity recorder (Ben-Yosef et al., 2008) and on the dating of the sites from which they were collected (below). In this study, every coherent fragment (piece of slag or clay) that we collected is called “sample”, and every chip of a sample is called “specimen”. From each sample we isolated four to twelve specimens ranging from 2 to 7 mm in diameter. The full name of a specimen designates its location. JS stands for Jordan, IS stands for Israel and the next two digits represent the site. The sample piece is designated with a letter and the specimen number with the last two digits. For example, specimen JS01b03 is the third specimen from the b sample from the Wadi Fidan 4 site in Jordan (JS01). We catalogued and stored all of our samples in the paleomagnetic laboratory of the Institute of Earth Sciences in the Hebrew University of Jerusalem, and they constitute a large inventory for future research. The specimens were inserted into non-magnetic glass tubes (1 cm in diameter) and went through a Thellier-Thellier type experiment, using a sophisticated experimental protocol (the “IZZI” protocol, see Tauxe & Staudigel, 2004; Yu & Tauxe, 2005; Yu et

17 al., 2004). A detailed description of the experiments, the selection criteria used and our methodology in determining the cut-off values together with comprehensive results and statistical analyses are given in Ben-Yosef et al. (2008).

Results

Our archaeointensity curve (Figure 11a and Table 2) is based on well dated samples (age quality 1 and 2) with at least 3 successful specimens (N≥3) that are in good agreement with each other (σ cut off = 20% of the mean or within 5 µT). Figure 11b and Table 3 show the additional samples that passed the experimental and statistical requirements, but originated from a poorly dated or controversial context (age quality 35). For perspective, we plot the recently published data set from archaeointensity investigation of Syrian sites (Gallet et al., 2006; Gallet & Le Goff, 2006; Genevey et al., 2003) together with the predicted VADM for the region from the CALS7K.2 model of Korte and Constable (Korte & Constable, 2005a). In total, 30 samples out of 80 show reliable geomagnetic intensity results, therefore representing a success rate (on a sample basis) of 37.5%. At the specimen level, 236 out of 400 passed the experimental requirements, giving a general success rate of ~60%. Comparing between specimens of furnace and tapping slag in terms of success rate shows a slight preference towards furnace slag. The success rate of baked clay from archaeometallurgical context was extremely high (89% in the specimens level and 100% in the sample level), although the total number of specimens is only 28. Bronze production slag show similar success rate to furnace copper slag, but in this case the number of specimens is limited, making this inference tentative.

18 Our archaeointensity curve shows acceptable agreement with the data set from Syria (Gallet et al., 2006; Gallet & Le Goff, 2006; Genevey et al., 2003 see Figure 11a). As this region is close to the Southern Levant and as these researchers used samples from careful archaeological contexts and modern, strict, experimental procedures, we consider the comparison useful, and the different data sets as complementary. The intensity of the geomagnetic field fluctuated rapidly over the last 7000 years. Major trends observed in previous studies were confirmed with our new results. This includes the conspicuous peak in intensity around 3000 years ago, now shown to be even higher during the Iron Age I, and the relatively long period of low intensity prior to 5000 years ago (Chalcolithic – Early Bronze Age I). Two less prominent peaks are corroborated around 4500 years ago (Early Bronze Age II-III) and 1200 years ago (Early Islamic). Our data suggest a slightly lower trough 2000 years ago (Early Roman). Not surprisingly, the details of the archaeointensity curve do not agree precisely with the smoother depiction of the global model of Korte and Constable (Korte & Constable, 2005a) (see Figure 11a). Nevertheless, most of the major trends of the geomagnetic intensity are reflected in the model. It seems to us that the reasons for the discrepancy are the current low resolution of the global model and the use of some less rigorously obtained data as constraints. The published data include a variety of approaches, materials, and quality controls on paleointensity and dating, hence may contain a less than optimal recording of the geomagnetic field.

PART II: ARCHAEOLOGICAL AND ANTHROPOLOGICAL IMPLICATIONS

As part of applying and testing the new approach to archaeointensity investigation described above, we obtained highly reliable archaeomagnetic results from hundreds of specimens that originate from archaeometallurgical sites in the Southern Levant. These results shed new light on the dating of some sites in the Timna valley, including the controversial site of Timna 39b, situated in Israel’s southern Negev desert. The new dates contribute to the discussion about the role of technology in society and challenge a common perception of technological evolution. In particular, they question the “Standard View of Technology” (Pfaffenberger, 1992, discussed in detail below), which suggests a unilinear evolution of technologies from simple to complex.

Archaeometallurgy in the Southern Levant and the problem of dating The copper ore rich districts of southern Israel and Jordan are some of the richest ancient mining and metal production regions in the Old World, comprising widespread evidence of archaeometallurgical sites and slag deposits. Together they provide key areas for understanding the role of technology on social change and an exciting new sample set for archaeointensity research for the time span of the last seven millennia. The first evidence of copper production in the Southern Levant goes back as early as the 5th millennium BCE and corresponds with the period of metallurgical innovation throughout the ancient Near East (e.g. Görsdorf 2002; Hauptmann, 2000, 2007; Levy, 2007; Levy & Shalev, 1989; Rothenberg & Merkel, 1998). The archaeometallurgical sites

19

20 in the region span almost all of the archaeological periods from the beginning of metal production in the Chalcolithic period , although at different resolutions (e.g. Avner, 2002; Rothenberg, 1999b), thorugh the Mamluk period, in the 13th century CE (Hauptmann, 2007). The main centers of copper production in the Southern Levant are Faynan and Timna, located along either side of the Wadi Arabah (the Arava Valley) (Figure 9). They are situated in the vicinity of natural exposures of rich copper ore that are typically part of sandstone and dolomite host layers (Hauptmann, 2007). Except for few other copper smelting sites located near small exposures of copper ore along the Wadi Arabah and in the Sinai Peninsula, other sites of copper industry required transportation of the ore for a relatively long distance. The Chalcolithic site of Shiqmim (Shalev & Northover, 1987) and the Early Bronze Ia site of Ashqelon-Afridar (Segal et al., 2004) are examples of copper production industries that transported copper ore from Faynan, more than 150 km away. The region of Timna has been intensively investigated by Beno Rothenberg, the director of the Arava archaeological expedition, between the years 1959-1990 (e.g. Rothenberg, 1962, 1999a, 1999b, 1990b). As part of this work more than 300 copper mining and production sites were documented (Wilson, 1983), some of which were excavated. Intermittent archaeological research in Timna continues to the present by the Israeli Antiquity Authority and University College London. The archaeometallurgy of the Faynan district was systematically investigated by Andreas Hauptmann and a team from the Deutsches Bergbau-Museum Bochum (DBM) between the years 1983-1993 (e.g. Hauptmann, 2007). Their work included surveys,

21 small scale excavations and complementary laboratory analysis of the

archaeometallurgical finds. Since 1997 the area has been the focus of intensive investigation as part of the Edom Lowland Regional Archaeology Project of the University of California San Diego (UCSD) and the Department of Antiquity Jordan (DOAJ) under the direction of Thomas Levy and Mohammad Najjar (e.g. Levy, 2006). As the largest center of copper production in the eastern Mediterranean, the Faynan district is a prolific source for archaeometallurgical studies. Moreover, the current UCSD - DOAJ research in this area provides samples from well defined context(Levy & Smith, 2007), usually with dating constrained by radiocarbon measurements. In Timna, however, the situation with regard to the dates of many sites is much more complex – in part because the excavations mostly took place over 25 years ago. In spite of the intensive research and the abundance of surveyed and excavated sites, only scarce radiocarbon dates are available (Avner, 2002 see in particular table 2 which covers all the periods). The paucity of radiocarbon dates generates a significant challenge for dating sites in the desert areas of the Wadi Arabah. These ancient sites, being remote from the populated centers of the Mediterranean and semi-arid regions where agriculture is relatively easy to practice, show distinct regional characteristics in the material culture. The ceramic typology for this region is much less refined, especially in the early periods form the Chalcolithic to the Iron Age (Avner, 2002; Rothenberg & Glass, 1992), thus hampering the possibility for high resolution contextual dating. In some periods, such as the Chalcolithic and Early Bronze, there are very little stylistic changes in the ceramic assemblage. This results, inter alia, in difficulty for identifying desert sites to the

Chalcolithic period in many of the early sites in the Wadi Arabah, both in the Faynan

22 area (e.g. Adams, 1998; Genz, 1997) and in the outskirts of Aqaba (e.g. Görsdorf 2002; Khalil, 1987, 1992, 1995; Khalil & Eichmann, 1999). In the Jordanian sites the ambiguity in dating was eventually resolved using high precision radiocarbon measurements. In Timna, however, the dating of some of the sites is still highly controversial, such as the copper smelting furnace of site Timna 39b (e.g. Rothenberg, 1990a and see below; Rothenberg & Merkel, 1998). The difficulty of establishing high resolution dates based on the material culture in the region of Timna led Rothenberg and Glass to develop a different and more crude typological/chronological scheme for the desert sites divided into three assumed phases of the “Sinai-Arabah Copper Age” (Rothenberg & Glass, 1992). In addition to distinctive ceramic and lithic types, each phase was characterized also by an archaeometallurgical typology, including slag types (Rothenberg, 1990b). For example, slag features such as glassy textures, viscosity, amount of left-over copper, mineralogy and chemistry, were considered as chronological markers. The reliability of archaeometallurgical typology as a dating tool was questioned by members of the Arava archaeological expedition themselves and other scholars (e.g. Avner, 2002), and it became clear that the technological development was not unilinear. Moreover, the chemical composition of slag varies according to the original ore and flux mixture which depends primarily on the geographical location rather than on the advances in technologies. Nevertheless, the archaeometallurgical typology was used for dating many sites, such as N3 (Segal et al., 1998) and 250b (Rothenberg & Shaw, 1990a, 1990b). These were dated to the Chalcolithic according to a similar “technological horizon” as Site 39b, a contentious site in itself.

23 In many of the earliest archaeometallurgical sites it is difficult or impossible to retrieve radiocarbon samples. Slag samples, as archaeointensity recorders, might hold the key for solving some of the dating problems and clarify the archaeological picture of the dawn of metallurgy in the region. Since the archaeointensity curve for the Southern Levant is yet in low resolution, a comparison with results from well dated archaeometallurgical sites is in cases necessary. As part of the current study we investigated slag also from sites of the more populated areas of the Beersheva Valley (Shiqmim), the western Negev (Ashqelon-Afridar) and the central coastal plain of Israel (Tell Dor and Tell Gerisa). In the latter we investigated Iron Age I bronze production sites (Ilan, 1999). However, before focusing on the problem of the 5th millennium BCE, it is important to examine the archaeointensity results for the entire seven millennia trajectory.

Implications on dating of archaeometallurgical sites Samples with reliable archaeointensity readings from poorly dated or controversial sites can contribute for constraining the age of their context. The results of the current research provide some insights into the dating of certain archaeometallurgical sites in the Southern Levant, mainly in the region of Timna. This includes the controversial site of Timna 39b.

Timna 39b The site of Timna 39b is considered by its excavator, Beno Rothenberg, to be the most ancient copper smelting installation ever found anywhere (Rothenberg, 1990a and

24 many other publications). Since its discovery (1960) and excavation (1965), there has been a ceaseless debate regarding its age (e.g. Avner, 2002; Craddock, 2001), which has not reached a satisfactory resolution so far. The site is located in the southeastern part of Timna Valley, on top of a small hill facing the Wadi Arabah plain. It was excavated together with a domestic site situated ca. 130 m to the southeast, on the lower slopes of the hill (Timna 39a). The final report (Rothenberg, 1978) connects the two sites and concludes that both are dated to the early phase of the Chalcolithic. Site 39a, a household unit with scarce evidence of ore and metal processing, was first dated primarily by the lithic assemblage (Bercovici, 1978). The Chalcolithic age was confirmed later by radiocarbon measurement yielding the date of 5485±45 BP (4351±98 BCE 95.4% probability using OxCal 4.0) (Rothenberg & Merkel, 1998). Site 39b is a “pit in the ground” smelting furnace, surrounded by many fragments of small furnace slag with homogeneous visual characteristics (Figure 12). It is 30-40 cm in diameter and ca. 40 cm in depth, although its partially stone lining suggest an upper structure of additional 40 cm (Rothenberg, 1978). It was dated to the early phase of the Chalcolithic primarily by relying on the typology of the lithics uncovered in the small excavation around the furnace, the slag and furnace characteristics and the supposed connection to Site 39a (Rothenberg, 1978, 1990a; Rothenberg & Merkel, 1998). Critical reservations regarding the early date of the furnace in Site 39b were raised, even before the publication of the final report, by J.D. Muhly (1973; 1976). He extended his criticism later on (Muhly, 1984), and was followed by various of other scholars (e.g. Adams, 1998; Avner, 2002; Craddock, 2001; Hanbury-Tenison, 1986;

25 Weisgerber & Hauptmann, 1988). In general, these objections for the early date are based on two aspects of the archaeometallurgical research of the site. The first is related to a comprehensive understanding of the metal production in the Chalcolithic (e.g. Shalev, 1994), which claims that copper smelting was done in specialized communities far from the origin of ore. This is the case in Beersheva valley (e.g. Gilead et al., 1992; Levy & Shalev, 1989), and in recently discovered industries near Aqaba (Hauptmann et al., 2004). The second aspect is related to the quality of the archaeological evidence (see updated summary and discussion in Avner, 2002). The main arguments regarding the quality of the archaeological evidence include reassessment of the technology, reservations of the models employed by the investigators and a previously unpublished radiocarbon date from the furnace itself. The furnace structure and the characteristics of the slag were used by Rothenberg as evidence for a suggested technology that is even earlier than the Chalcolithic of Beersheva Valley (Rothenberg & Merkel, 1998). However, revisiting of the evidence suggests an advanced, presumably late industry (e.g. Avner, 2002). The supposed connection between Site 39a and the furnace is not decisive, and the original publication of the lithic assemblage did not distinguish between the two sites (Bercovici, 1978) creating ambiguity in the interpretation. Most surprising is the radiocarbon date from the furnace, yielding the result of 1945±309 BP (Burleigh & Hewson, 1979) (761BCE – 645CE, 95.4% probability, using OxCal 4.0). Rothenberg, who characterizes this date as “Late Bronze Age” (Rothenberg, 1990a), explains the date as being derived from refill of the excavation pit that was brought from a different location. Others suggest the possibility of

26 reusing the smelting location and/or installation in the course of more than one period (Avner, 2002). Revisiting the site in 2004-5, we collected 10 samples of furnace slag from the furnace itself and its close vicinity. Four samples (based on 16 specimens) passed all of our rigorous selection criteria and yielded reliable archaeointensity results. They clearly show three distinct groups of ancient geomagnetic intensity (Figure 13), implying at least three periods of copper production in the site of Timna 39b. The group showing the lowest intensity (66±7 ZAm2 VADM) might indeed represent copper smelting during the Chalcolithic. It is within a one standard deviation agreement with the archaeointensity results obtained for the Chalcolithic site of Shiqmim (58±4 ZAm2 VADM) and is consistent with the general low intensity throughout this period. Nevertheless, this group is compatible with copper smelting in other periods, mainly the Early Bronze Age I. The middle group, as well, might represent several different periods of copper production including Early Bronze Age II-III, Middle and Late Bronze Age, and Byzantine – Early Islamic periods. The latter corresponds to the radiocarbon measurement from the site. The group with the highest intensity (145±11 ZAm2 VADM) fits best to the Iron Age I period, the latest phase of the intensive copper production in Timna region under the Egyptian influence (Rothenberg, 1999b). The archaeointensity results from Site 39b provide additional support for Rothenberg’s early Chalcolithic dating, although they do not decisively prove it. Moreover, there might be a difference between the dating of copper production in the site and the dating of the installation found in situ today. While our results support the idea

27 that smelting activities occurred in more than one period, the installation itself might represent only the latest one. We do not find the evidence of copper production near the origin of the ore during the Chalcolithic to be unique. The evidence of metallurgical activities in the Chalcolithic site of Timna 39a (Rothenberg, 1978), together with other small sites in the Timna region such as N3 (Segal et al., 1998), F2 (Rothenberg, 1999a; Rothenberg & Merkel, 1995) and 250b (Rothenberg & Shaw, 1990a) might suggest small scale domestic copper production in periods as early as the Chalcolithic, although this evidence is problematic (e.g. Avner, 2002) and more research is needed. Moreover, in the light of other sites in the Wadi Arabah, the connection between sites 39a and 39b is a reasonable supposition. In many cases, the “cold industry” of crushing the ore and flux and processing slag was done at the foot of the hill, while the pyrotechnological industry, taking advantage of the wind, was done on the top of the hill (e.g. Avner, 2002; Site 189a: Avner & Naor, 1978; Site 201a: Rothenberg, 1999a; Rothenberg, 1990b). There is no doubt that the vast majority of data for Chalcolithic smelting in the southern Levant comes from the Beersheva and supports the ‘monopoly’ model of metallurgical control from this region. However, the new archaeointensity data points to more than one mode of production during the 5th millennium BCE.

Archaeometallurgical sites from later periods The site of Timna 149 (Rothenberg, 1999a; Rothenberg & Glass, 1992; Rothenberg & Shaw, 1990a, 1990b) is located in the northeastern part of the Timna Valley, and considered by its excavator to be a key site for understanding the

28 development of metallurgy in the Early Bronze Age IV (ca. 2200-2000 BCE). The site consists of two separate parts, one on top of a hill facing the Wadi Arabah and the other on a plain to the west of the hill. The latter was excavated (1984, 1990) and dated by indicative ceramics from well defined context to the Early Bronze Age IV. The excavated area contains two shallow lines of walls, ground stones, slag fragments and clay rods, and was interpreted as a preparation camp for the smelting process which took place on the top of the hill. In addition, the excavation suggests slag processing and probably a secondary melting for the production of ingots (Rothenberg & Shaw, 1990b). The date of the finds from the hilltop is much less secure and based primarily on the supposed connection to the excavated site of the hillside. They include slag fragments and stones that were interpreted as part of sophisticated furnaces that replaced the earlier “pit in the ground” type. According to the excavator, they represent a progress in copper production attributed to this period (e.g. Rothenberg & Shaw, 1990a). Our archaeointensity results (Figure 13) show clearly that there is no connection between the metallurgical activities of the hillside and the hilltop. While results from the former are indeed in agreement with data from previous studies and fit well in the Early Bronze Age IV, the results from the hilltop are distinct and represent a different period. This period is most probable the Late Bronze IIB (13th century BCE), when the copper production activity in the area reached a climax under the Egyptian influence. Several other periods are also compatible with our results, including Early Islamic (638 – 1099 CE) and Early Bronze Age II-III (ca. 3000 – 2200 BCE) (Figure 13). The alleged sophistication of the furnaces on the hilltop and the claims for industrial scale of copper production, with a breakthrough in technology (e.g. first

29 appearance of tapping slag) are contentious, still regardless of their date (e.g. Avner, 2002). The conclusion about metallurgical activities during the Early Bronze Age IV should be reassessed under the light of the recently discovered large scale industry from this period in Faynan district (Levy et al., 2002), as well as the interpretation of the finds from the excavated industry in the hillside. We suggest that the industry of the hillside included smelting in addition to preparation and processing activities. The clay rods, considered by the excavators to be components of crucible manufacturing (Rothenberg & Shaw, 1990b), might be part of the smelting installation, as suggested for the same type of finds from Faynan district (Hauptmann, 1989, 2000). In Faynan, however, the clay rods are part of wind-driven furnaces common in the Early Bronze II period. The samples from the site of Timna 30 were collected from layer I, considered by the excavator to represent the most advanced ancient copper smelting technology (Rothenberg, 1999b). The site was excavated (Bachmann, 1980; Rothenberg, 1980, 1999b, 1990b) and layer I was dated by Egyptian ceramic to the 22nd dynasty, in particular to the reign of Shishanq I. A radiocarbon date yield even later date from the 8th century BCE (Rothenberg, 1990b Footnote 71). The advanced technology represented in layer I and the uniqueness of the Iron Age II period raised some reservations concerning the date (e.g. Avner & Magness, 1998 Footnote 7). Our archaeointensity results fit well in the Iron Age II, both to the period of Shishanq I as well as to the 8th century BCE. Because of the high peak in the geomagnetic intensity in this period, it is difficult to assign this layer to any other period. The site of Givat Yocheved (Nahal Amram, Timna 33) is located 15 km south of Timna Valley, near an intensive mining district. It consists of several structures and

30 mounds of broken tapping slag. The Arava expedition dated the site to the New Kingdom (14th-12th centuries BCE) (Rothenberg, 1967; 1990b Footnote 23), a date that was confirmed with a radiocarbon measurement from the bottom of the slag mound (Rothenberg, 1990b Footnote 21). However, based on the advanced metallurgical technology evidenced at the site, other scholars date the site to the Early Islamic period (Avner & Magness, 1998) and point out another radiocarbon measurement from the same site, yielded a date from the 8th – 9th centuries CE (Burleigh & Hewson, 1979). Our archaeointensity results (Figure 13) fit neither of the suggestions above, and indicate most probably copper smelting in the Early Roman period. A date from the Middle Bronze Age or earlier (Figure 13) is inconsistent with the advanced tapping technology, and the Early Roman period is compatible with the intensive mining of copper ore from this period in the close vicinity (Avner & Magness, 1998; Willies, 1990). However, the site very likely represents more than one period, including the New Kingdom and Early Islamic as well. The site of Eilot Quarry was surveyed in the 1970s (Avner & Naor, 1978). Its original Early Islamic date was changed to Early Bronze Age according to new finds of lithic and ceramics (Avener, personal communication, 2006). Our archaeointensity results (Figure 13) support the early date and constrain it to the Early Bronze Age I / early phase of Early Bronze Age II. The sites of Hai-Bar and Yorvata-EB in the Timna region are considered to be early according to the slag type and arhcaeometallurgical typology. According to our archaeointensity results (Figure 13), both are dated to later periods. Hai-Bar can most probably be dated to the Late Bronze Age – Iron Age I, the climax of copper production

31 in the area under the Egyptian influence. Nevertheless, other periods are also possible for this site, such as the Early Islamic. The results from Yotvata-EB indicate Iron Age II smelting activities, a date which makes it the second known site from this period in the southern part of the Wadi Arabah. The revised dating of these sites demonstrates that slag and archaeometallurgical typology cannot be used as a chronological marker, and that the advancement in copper production technologies was accompanied by continuation of small scale production using less sophisticated techniques. The site of Ashqelon-Afridar (Gophna, 2004) is a large scale Early Bronze Age I settlement, located in the southern part of the coastal plain of Israel. The excavation encountered ample archaeometallurgical remains (Segal et al., 2004), representing melting and casting activities, as well as smelting of copper ores. Our samples originated in area 10, excavated by Y. Yekutieli in 1998. Although the finds from this area were dated to the Early Bronze Age Ia and show similar characteristic to the finds from nearby area E (Golani, 2004), the specific samples (IS20a,b) came from an insecure context of refill in pits. Our archaeointensity results suggest a later date for this phase of metallurgical activities associated with the pits, most probably Early Bronze Age II-III (Figure 13). Our archaeointensity results from Tell Gerisa (Figure 13) suggest a different date than Iron Age I. The excavations were not published yet and the exact context of the samples is not known, thus hampering further discussion.

32 Technological development in a wide perspective Revisiting a large number of archaeometallurgical sites from a broad range of periods as part of the current study provides a wide perspective for assessing theories regarding the social aspects of technology. The new dates obtained for several sites in the area of Timna have significant implications on our understanding of the technological development in the region, and might be a model for similar interpretations in different parts of the world. Moreover, revising the dates previously assigned incorrectly to many copper production sites in the study area questions some of the preconceptions of early scholars, as well as their basic paradigm in regard to evolution of technology.

The Standard View of technology As an analogue to the Standard View of science (Mulkay, 1978), Bryan

Pfaffenberger (1992) has termed “the Standard View of technology” to refer to what underlies scholarly and popular thinking about technology in the majority of publications. Although the extent of which this Standard View is present in the different publications varies, in general it comprises three major paradigms: ‘necessity is the mother of invention’, ‘form follows function and style, and meaning is a surface matter’, and ‘development is unilinear evolution, from simple to complex’ (Pfaffenberger, 1992). The latter can be demonstrated using Pfaffenberger’s description: This record shows a unilinear progression over time, because technology is cumulative. Each new level of penetration into Nature’s secrets builds on the previous one, producing ever more powerful inventions. The digging stick had to precede the plough. Those inventions that significantly increase Man’s reach bring about revolutionary changes in social organization and subsistence. Accordingly, the ages of Man can be expressed in terms of technological stages, such as the Stone Age, the Iron

33 Age, the Bronze Age, and so on (Pfaffenberger, 1992, the use of Man as representing of humankind is deliberately done in this source to emphasize the gender ideology accompanying the Standard View). For rebutting the Standard View, Pfaffenberger uses a concept from socio-historical study fields, namely the sociotechnical system (Pfaffenberger, 1992). Applying this concept in the field of anthropology puts technology as an inseparable part of society, a subsystem among other social subsystems. Technology cannot be studied alone. It has complex reciprocal interactions with other subsystems within society, such as social, political and economical institutions, ideology, value systems and meaning, and religion. In this light, the basic paradigms of the Standard View are challenged and a different approach to the study of technology is developed. Pfaffenberger avoids discussing in detail the relations between sociotechnical systems and the assumption of unilinearity in technological development. However, this concept is useful when discussing alternatives to this assumption (below).

The development of copper production technologies in the southern Levant The Standard View of technology as termed by Pfaffenberger (1992, see previous section) is a useful reference for assessing the early research in the copper production districts of the Southern Levant. The question of how much the Standard View virtually underlies previous publications in the field of anthropology in general, and if it is indeed so dominant as Pfaffenberger suggests, might still be open. It is beyond the scope of this work to evaluate the prevalence of the derived paradigms throughout the course of progress in anthropological research, although it seems that the claim for their predominance in this field is somewhat exaggerated (see e.g. Killick, 2004; Schiffer,

34 2001 and other references there). However, the early stages of archaeometallurgical research in the Wadi Arabah was not done by anthropologists (Hauptmann, 2007; Rothenberg, 1990b), and the basic paradigms of the Standard View dominant all of the early publications, especially in regard to the research in Timna. When discussing ancient copper production technologies Rothenberg states that The typology of furnaces, slag, tuyeres, etc. is also of considerable importance as a means of establishing stratigraphic groups of technological characteristics. For example, the fact that small, platy, broken tapped slag, small clay tuyeres, bowl-furnaces with clay-like mortar lining, are characteristic of Layers III-II of Site 30 and Layer II of Site 2, enables us to synchronize the two sites and identiry isolated elements of such groups found in excavations or surface surveys. (Rothenberg, 1990b) And in other place However, the Arabah slag typology, supported by analytical studies of the visually different types, demonstrates that the physical shape and texture of slag of a given area can indeed serve as a reliable chronological indicator. After all, the visual differences between the different slag types are simply the outside representation of characteristics related to techniques and efficiency of smelting installations. In this sense, differences do have chronological significance, often even of a fundamental nature. This has been widely confirmed by comparisons of the Arabah material with metallurgical remains of metal producing in other areas, even from wide-apart continents. (Rothenberg, 1990b emphasis is mine) Rothenberg’s statements above are deeply rooted in the third paradigm of the Standard View, namely ‘development is unilinear evolution, from simple to complex’. Many sites were dated by the Arabah expedition based on this paradigm, assuming a simplistic evolution of copper technology in the region. Rothenberg was not alone. Others archaeologists surveying in the area used the same assumptions to date unexcavated sites

35 such as Hai-Bar and Tell Hara-Hadid mentioned in this work (Uzi Avner and Assaf Holtzer, Pers. Comm. 2006).

Our revised dates for many archaeometallurgical sites in the region of Timna suggest a different model for technological development (Figure 14). The new dates show that simple technologies were in use in later periods than those assigned to them by early scholars, sometimes taking place simultaneously with very complex industry in the close vicinity. The terms “simple” and “complex” in this paper stand for the efficiency and scale of the technology, e.g. tapping technology enabled the production of much more copper metal than bowl-furnaces technology, for the same given unit of energy invested in the system. In the light of the new archaeointensity dates presented here, there is no longer a strong basis for the technological-typological-chronological scheme used by earlier researchers. At best, technological ‘horizons’ can serve as a terminus post quem, but even that should be made carefully, after establishing well dated scheme for advances in technology and exploring the process of technological innovations in anthropological, or sociotechnical context. The development of copper technologies in the Southern Levant is complex, and can be best explained using the sociotechnical concept as described by Pfaffenberger (Pfaffenberger, 1992). It is clear that simple technologies were still in use when advanced and more complex technologies were developed (Figure 14b, columns 1,2,4). It is also evident that technological knowledge possessed by certain societies in certain regions can get lost (Figure 14b, column 2) (see also Pfaffenberger, 1992). This observation raises interesting questions: Does the presence of simple and complex technologies in the same

36 region indicate the presence of different societies? Or does it indicate different classes or sectors within one society? Does regaining lost knowledge indicate the arrival of new society/culture to the region, or is a diffusion-model is more adequate to explain such process? What can we infer from the presence of different technologies in one region about the societal hegemony, structure and distribution of power in a given region? Even if we consider different societies for the region of the southern Wadi Arabah (e.g. sedentary society versus nomadic), in this relative small area interactions must have occurred. Technology, as mentioned above, is a one facet of complex subsystems of society, with direct connections to political, economical, religious and social structures. To untangle this complex network of interactions in a deep time perspective out of the silent material remains is a hard task (Levy, 2006). For the region of Timna, one should take into account the historical evidence as well as the archaeological and geographical layout. The Egyptian hegemony in copper production during the Late Bronze Age – Iron Age I periods (ca. 1300-1100 BCE) might have been challenged by marginal groups of local nomadic population that still practiced “ancient” simple copper production technology in the Arabah valley, close to the Egyptian center in Timna, but far enough and in isolated locations, to be relatively safe from the direct Egyptian control (The site of Hai-Bar, re-dated in this work, might be an example of such site). Was it always marginal groups that continued to practice the simple technologies? Was the presence of simple-technology workshops manifests some sort of a challenge on the dominant economical power, represented by the contemporary practice of complex technology? One should consider also a peaceful picture of legitimate household industries that were practiced side by side with the complex, industrial scale technologies. Although complex

37 technologies are much more efficient, they demand more communal effort and more centralized power to operate. As this was the domain of the elite, others could still practice the simple technology with their products aimed to be part of the collective good, paid as the family taxes, or distributed in a local trade. The latter might be practiced while the complex mass production was aimed to be shipped for long distances, to different regions and states, creating a special products class for “export only”. Another aspect of technology to consider when analyzing the complex quality of technological development is the treatment of technological knowledge in complex societies. Already in the first appearance of chiefdom society in the Chalcolithic (ca. 4500-3500 BCE) centers of specialized workshops were recognized in the archaeological record, including ones focused on copper metallurgy. These copper production workshops were found typically far from the ore deposits (see discussion above under Site 39b), and were interpreted as representing the practice of specific social group possessing exclusive technological knowledge (e.g. Levy & Shalev, 1989; Shalev, 1994). Knowledge and technological knowledge in particular, is an important source of power. Interpreting simultaneous variations in complexity of technology in a given region should be done also in the light of accessibility to knowledge sources, intertwined with the distribution of power. The possibilities for interpreting the lines of technological evolution in a given region and/or society are numerous, especially in the light of the sociotechnical concept. Each case in each period should be analyzing separately, considering the archaeological, historical and geographical evidence available. Only a complex model based on

38 ethnoarchaeological, anthropological and sociological observation can bring us close to a thorough understanding of the ancient society under consideration.

Conclusions Archaeointensity in the Levant – New Horizons The research presented here illustrates the importance of melding science-based research methods and anthropological theory to achieve new and testable insights concerning the relationship between technology and society through time. The focus of this study has been ancient copper production in the southern Levant. The results from the current study demonstrate the suitability of copper slag material in archaeointensity research (see also Ben-Yosef et al., 2008). Together with the application of a sophisticated experimental protocol (the “IZZI” protocol of Tauxe and Staudigel 2004), we introduced a new and promising tool for studying the behavior of the geomagnetic intensity during the last 7 millennia. The abundant archaeometallurgical sites in the Southern Levant provide an invaluable source of samples for archaeointensity research. Together with complementary sites in Cyprus (e.g. Balthazar, 1990) and Anatolia (e.g. Yener, 2000), slag deposits present a relatively high time resolution for the periods since the dawn of metallurgy. In this study, 15 reliable archaeointensity results from well dated contexts have been added to the archaeointensity curve of the Levant. They are in good agreement with previously published data from Syria (Gallet et al., 2006; Gallet & Le Goff, 2006; Genevey et al., 2003), and emphasize some of the heretofore observed trends in the geomagnetic intensity behavior. Further reliable archaeointensity data from well dated

39 archaeological context are needed for improving the resolution of the highly fluctuating curve. Such a high-resolution curve, in turn, might be used in the archaeological research.

Archaeointensity as a Dating Tool The resolution of the current available archaeointensity curve is poor and its application as a dating tool is limited. In most cases, other archaeological methods of dating, such as radiocarbon or material culture typologies, are more probable to yield accurate results. However, in certain sites, where radiocarbon samples are unavailable and the material culture typology is problematic or in low resolution, the archaeointensity curve might be used as a reference for dating. This is the case in many of the archaeometallurgical sites in the southern Wadi Arabah where the material culture cannot provide a decisive date. Our reliable archaeointensity results from such sites were compared to results from well dated samples and to the available archaeointensity curve, providing several insights regarding the archaeometallurgy of this region. The research presented here has local historical implications for the archaeological record of the southern Levant. Our results show that metal production activities in site Timna 39b occurred in more than one period, most probably including the Chalcolithic. These results help lay to rest a long-standing debate concerning the beginning of copper production in the southern Levant (cf. Avner, 2002). The site of Timna 149 had hosted copper smelting in the Early Bronze Age IV only in the excavated hillside part, while the remains on the hilltop are from a distinct period, probably related to the proliferation of copper industry during the New Kingdom.

40 Archaeointensity research focuses only on one component of the geomagnetic field. Combining data from high resolution curves of inclination and declination changes provide a strong dating tool for the archaeologist, based on a statistical matching of the three different components (Lanos, 2003). Applications of such a dating technique provides excellent results (e.g. Jordanova et al., 2004; Kovacheva et al., 2004) and demonstrate the need for further reliable archaeomagnetic data in the Southern Levant (see also Le Goff et al., 2002).

Challenging the Standard View of Technology A significant conclusion derived from the new dates for several

archaeometallurgical sites in the southern Wadi Arabah is the nonlinear development of copper smelting technologies. Our results show clearly that ancient technologies were still in use in later periods, along with the advanced large-scale production industry. Slag and archaeometallurgy typology cannot, therefore, be used as a chronological marker. They might, however, be related to social and political structures, implying differential accessibility to resources of knowledge and power. The nonlinear development of copper smelting technologies in the Southern Levant as demonstrated in this work challenges a common perception about the evolution of technology, namely the Standard View of Technology (Pfaffenberger, 1992). Early scholars of archaeometallurgy in the region practiced the basic paradigms of this view, and many sites were dated in the light of such approach. More anthropological approach that makes use of the sociotechnical concept is better to explain the complex evidence as presented in the current study.

APPENDIX I: FIGURES

Figure 1: Examples of archaeointensity data from the Near and Middle East for the last seven millennia, the period since the inception of copper smelting (after Ben-Yosef et al., 2008). The magnetic field strength is expressed as Virtual Axial Dipole Moment (Z=1021). Large green triangles are data from Syria of Genevey et al. (2003), blue squares are from Gallet & Le Goff (2006) and brown dots are the Syrian data from Gallet et al. (2006). Open red circles and squares are compilation of 11 other sources, mostly based on fired clay (see Ben-Yosef et al., 2008 for references). Predicted VADM values for Syria by CALS7K.2 of Korte and Constable (2005) are shown as dashed line. The recent dipole value is shown as a solid black line (~80 ZAm2).

42

43

Figure 2: The Earth’s magnetic field and its elements. a) Magnetic field lines as predicted by a simple model of geocentric axial dipole; b) magnetic field lines predicted from the international geomagnetic reference field from 1980 in the Earth’s mantle (green) (Courtesy of R.L. Parker). The core is the source of the field and is shown in yellow. The field is somewhat more complicated than that shown in a) owing to non-axial dipole contributions to the field. The enlarged circle represents the vector of the geomagnetic field (B) for the specific location on the Earth’s surface (I=Inclination angle); c) three elements of the geomagnetic field’s vector: Inclination angle (I), declination angle (D) and intensity, represented by the length of line B.

Figure 3: A graphic representation of the “Thellier-Thellier” type experiment. The figures represent one specimen (IS07a01) and the fractional pTRM it obtained in the laboratory throughout the experiment. Each point stands for a different temperature step. a) A plot showing the gradual destruction and replacement of NRM (the original [natural] remanent magnetization) by laboratory partial TRM for each temperature step. The fraction NRM remaining after cooling in zero field is the blue line with circles and the pTRM gained when the specimen is cooled in a laboratory field is the green line with the red squares. b) An “Arai plot”: the NRM remaining at each temperature step is plotted against the pTRM gained. The absolute value of the slope of the line connecting temperature steps (1.871) is reflects the ratio of NRM/TRM. When multiplied by the lab field (30 µT), the slope gives the absolute archaeointensity (56 µT). The linearity of this line and other parameters are used to determine the reliability of the archaeointensity result. In addition, the plot shows the pTRM checks (blue triangles, expected to be in the same location of the corresponding temperature step, e.g. at 530ºC) and the pTRM tail checks (blue square, expected to be close to zero). For further discussion see text and Ben-Yosef et al. (2008).

44

Figure 4: Embedded charcoal in a slag sample. The charcoal enables a direct dating of the sample, without relying on the archaeological context.

Figure 5: Slag deposits in the Southern Levant. a) A ~2 m profile of a partiallly excavated slag mound in the site of Timna 30, representing probably the Late Bronze Age II, Iron Age I and Iron Age II; b) Abundant slag mounds in Beer Ora Valley (Timna 28). The slag deposits represent intensive copper production in the Early Islamic period.

45

Figure 6: Slag attached to the walls of in-situ furnaces enables sampling for full geomagnetic vector analysis. a) The lower part of furnace “Z” in site Timna 2 is a clay-lining “pit in the ground” (Rothenberg, 1990b); b) The stone built furnace “E” in site Timna 2 with slag attached (Rothenberg, 1990b).

Figure 7: Examples of slag samples. a) Broken tapping slag with flow textures, looking at its top (Khirbat en-Nahas, Jordan). Flat areas indicate the horizontal position of the slag when cooling, enabling the reconstruction of the geomagnetic inclination angle. b) Broken tapping slag with “slag droplet” embedded (Khirbat al-Jariya, Jordan). The droplets indicate the horizontal position of the sample when cooling, enabling the reconstruction of the geomagnetic inclination angle. The glassy texture makes the droplet itself a good source for archaeointensity experiments. c) Intact tapping slag sample Khirbat Hamra Ifdan, Jordan). d) and e) glassy fragments of tapping slag (Khirbat en-Nahas, Jordan). f) “Slag cake” (Beer-Ora Valley, Israel). g) broken furnace slag from site Timna 39b.

46

Figure 8: Examples of baked clay artifacts from archaeometallurgical context. a) Clay rods (“lady fingers”) and furnace fragments from site Fenan 15, Jordan (Early Bronze Age II-III). b) Clay crucible with slag coating, Tell Gerisa, Israel (Iron Age I). c) Clay furnace fragment, Khirbat en-Nahas, Jordan (Iron Age II). d) Tuyère fragment with slag coating Khirbat en-Nahas, Jordan (Iron Age II). e) Tuyère fragment, back side, Khirbat en-Nahas, Jordan (Iron Age II). Note the composite structure of clay material. f) Bellow tube fragment, Khirbat en-Nahas, Jordan (Iron Age II). g) Clay mould for casting copper ax, Khirbat Hamra Ifdan, Jordan (Early Bronze IV) (after Levy et al., 2002).

Figure 9: The major copper production centers in the Southern Levant.

47

Figure 10: Archaeometallurgical sites that were sampled in the current study.

48

Figure 11: Summary of all acceptable sample intensities (with standard deviation cutoff values of 20% of the mean and N≥3). a) All samples have an age reliability index better than 3 (Table 2). b) Same as in a) but including samples with uncertain ages (triangles). Small blue squares are data from Syria (Gallet et al., 2006; Gallet & Le Goff, 2006; Genevey et al., 2003). Predicted VADM values for Syria by CALSK7K.2 of Korte and Constable (2005) are shown as dashed line.

49

Figure 12: The copper smelting installation in site Timna 39b and the excavated area surrounding it.

Figure 13: Curve combining Syrian (Gallet et al., 2006; Gallet & Le Goff, 2006; Genevey et al., 2003) and Southern Levantine results (this study). We averaged results to the site level, excluding Timna 39b where three distinct groups of data were obtained. Also shown are reliable archaeointensity results from poorly dated or controversial sites (green circles; 149t=hilltop, 149s=hillside) (see text for discussion).

50
1 2 simple 3 4

Complex

simple

complex complex

Time

Simple a

Simple b

Figure 14: Schematic diagram showing two models for the development (evolution) of technology, from simple (light color) to complex (dark color) for a given region (or society). a) The Standard View of technology: a unilinear development from simple to complex. b) A more complex evolution, as demonstrated by the results of this study. Simple technologies can still be in use in the same region and/or the same society simultaneously with complex technologies and technological knowledge can get lost. The diagram represents several possibilities out of myriad of possible combinations (see text).

APPENDIX II: TABLES

51

52
Table 1: Archaeometallurgical sites and samples in this study Lat/Long Age CalBP C1 Co2 Q3 S#4 SN5 Type6
29.763/34.994 29.830/35.020 29.890/35.058 29.885/35.046

Site name Timna 39b Hai-Bar
Yotvata Fortress

S7 * * * * * * * * * * * * * * * * * -

Ref.8
6,1113, 33

6150±250
5350±1100

* * * * * * * * * * * -

S S S S S S S S S S S S S S S S S E E S S S E

6 5 2 5 1 5 2 2 6 2 2 3 2 3 6 3 1 1 1 1 4 1 1 3

10 6 1 1 1 1 4 3 6 4 3 2 2 5 1

Yotvata Eilot Quarry Beer-Ora Hill Timna 149 Timna 2 Timna 3 Timna 30 Timna 28 Tell Hara-Hadid GivatYocheved Mitzpe Evrona Fidan 4 Fenan 15 Khirbat Hamra Ifdan
Khirbat al-Jariya Khirbat en-Nahas

29.589/34.952 29.717/34.985 29.792/35.001 29.784/34.948 29.779/34.952 29.771/34.947 29.716/34.984 29.589/34.965 29.648/34.939 29.695/34.987 30.673/35.385 30.629/35.497 30.663/35.393

5810±500 4600±350 1950±100 4600±350 3735±20 4100±150 3175±75 3175±75 2810±60 1100±150 1150±150 1150±150 1150±150 5200±250 4550±300 4400±150 4050±100 2950±200 2980±110 2800±50 2950±200

IS11 IS24 IS01 IS17 IS15 IS16 IS19 IS07 IS03 IS02 IS05 IS06 IS08 IS09 IS04 IS10 IS18 1IS25 JS01 JS04 JS09 JS08 JS06 JS02 JS03 JS07

FS FS FS FS TS FS FS FS FS TS TS TS TS TS TS TS FS/C FS/C FS FS TS TS/C TS FS

1
17,8,6

1 1 20,1 6 2,3,5 ,6 10 5 10
4,7,9

1 9,18, 10 9
27,28

30.707/35.452 30.681/35.437 30.674/35.391 30.630/35.495 30.626/35.495

Khirbat Glueck
(WF77a)

3 2 1 1 1 2 3 1

28 31 31 32 28
28,29

32

2850±100 S 5 1 JS10 TS 28 1790±145 * S 2 1 JS11 TS 28 30.675/35.447 700±50 S 1 1 JS05 TS * 28 31.195/34.639 14-16 6225±50 * E 1 3 IS14 FS/C * Ashqelon-Afridar 31.679/34.556 21-24 5425±125 * E 4 3 IS20C FS/C * 32.091/34.806 Tell Gerisa 3050±100 E 3 1 IS21 BS * 25 32.617/34.916 Tell Dor 3050±100 E 1 1 IS22 BS 26 1 14C dates available; 2 Collection method: S: surface collection during survey, E: Collection from excavation; 3 Age quality: 1: Excellent, 2: Moderate to excellent, 3: Moderate, 4: Moderate to Poor, 5: Poor, 6: Controversial; 4 Number of samples measured from the site; 5 Sample names; 6 Type (crude categories): FS: furnace slag, TS: tapping slag, C: clay, BS: bronze production slag; 7 Successful experiment results for the site (according to our criteria, see Ben-Yosef et al., 2008); 8 Most relevant references: [1] Avner, U., Pers. Comm.. 2006, [2] (Rothenberg & Shaw, 1990b), [3] (Rothenberg & Shaw, 1990a), [4] (Rothenberg, 1999), [5] (Rothenberg & Glass, 1992), [6] (Avner, 2002), [7] (Sharon et al., 1996), [8] (Segal & Carmi, 1996), [9] (Avner & Magness, 1998), [10] (Rothenberg, 1990b), [11] (Rothenberg, 1978), [12] (Rothenberg, 1990a), [13] (Muhly, 1984), [14] (Gilead, 1994), [15] (Shalev & Northover, 1987), [16] (Burton & Levy, 2001), [17] (Meshel, 1993), [18] (Willies, 1990), [20] (Avner & Naor, 1978), [21] (Segal & Carmi, 2004), [22] (Gophna, 2004), [23] (Golani, 2004), [24] Yekutieli, Y., Pers. Comm.. 2006, [25] Herzog, Z., Pers. Comm. 2006, [26] (Ilan, 1999), [27] (Adams, 1999), [28] (Hauptmann, 2000), [29] (Levy et al., 2004), [31] (Levy et al., 2002), [32] Levy, T., Pers. Comm. 2007, [33] (Burleigh & Hewson, 1979).

Fenan 7 Fenan 1 El-Furn Shiqmim

53
Table 2: Reliable archaeointensity results from well-dated archaeometallurgical sites in the Southern Levant (figure 11). For discussion on selection criteria applied see Ben-Yosef et al. (2008), and text. (Q=age quality index; N=number of successful specimens). Sample IS03b IS04b IS05a IS06a IS06b IS07a IS14c IS16a JS01b JS01c JS02b JS04a JS04b JS05a JS08a Site Timna 149 hillside Timna 28 Timna 2 Timna 2 Timna 2 Ora Hill Shiqmim Yotvata Nabataean Wadi Fidan 4 Wadi Fidan 4 Khirbat Jarieh Fenan 15 Fenan 15 El-Furn Khirbat Hamra Ifdan Age 4100 1100 3175 3175 3175 3735 6225 1950 5200 5200 2980 4550 4550 700 4050 +/150 150 75 75 75 20 50 100 250 250 110 300 300 50 100 Q 2 2 2 2 2 2 1 2 1 1 1 1 1 1 1 N 3 5 3 6 4 3 3 3 4 5 3 4 3 3 5 Bancient 44 52 62 61 55 56 30 40 23 29 82 53 50 43 36 %σ 8.6 12.0 2.7 13.4 9.3 2.2 7.3 20.0 6.4 5.4 6.8 17.9 1.5 4.1 14.8 VADM 85.5 102 121 120 108 110 57.8 78 43.6 55.6 159 102 95.7 83.9 68.6 1σ 7.39 12.2 3.32 16 10.1 2.45 4.23 15.6 2.8 3.02 10.8 18.2 1.41 3.44 10.2

Table 3: Reliable archaeointensity results from poorly dated or controversial archaeometallurgical sites in the Southern Levant (figure 11b). For discussion on selection criteria applied see Ben-Yosef et al. (2008), and text. (Q=age quality index; N=number of successful specimens). Sample IS01a IS01b IS02a IS02e IS02f IS09a IS11b IS11d IS11e IS11i IS15a IS18a IS19a IS20c IS21a Site Hai-bar Hai-bar Timna 149 hilltop Timna 149 hilltop Timna 149 hilltop Timna 30 Timna 39b Timna 39b Timna 39b Timna 39b Yotvata (EB) Givat Yocheved Eilot quarry Ashkelon-Afridar Tell Gerisa Age 5350 5350 4100 4100 4100 2810 6150 6150 6150 6150 4600 1150 4600 5425 3050 +/1100 1100 150 150 150 60 250 250 250 250 350 150 350 125 100 Q 5 5 6 6 6 3 6 6 6 6 5 6 5 4 3 N 3 4 6 5 7 3 3 4 4 5 3 7 3 11 3 Bancient 57.9 59.7 57.2 55 54 64 54.6 73.9 34 49.7 67.7 45 35 56.4 45.7 %σ 13.5 9.2 4.4 12.9 8.5 18.1 8.0 7.9 10.0 9.5 2.6 10.2 8.4 16.0 4.0 VADM 113 117 112 108 106 125 107 145 66.5 97.3 132 88.3 68.6 108 86.8 σ 15.3 10.8 4.95 13.9 8.97 22.7 8.51 11.4 6.68 9.25 3.37 8.96 5.73 17.2 3.47

REFERENCES

Adams, R. B. (1998). On Early Copper Metallurgy in the Levant: A Response to Claims of Neolithic Metallurgy. In H. G. K. Gebel, Z. Kafafi & G. O. Rollefson (Eds.), The Prehistory of Jordan II. Perspectives from 1997. Studies in Early Near Eastern Production, Subsistence, and Environment 4. (pp. 651-656). Berlin. Adams, R. B. (1999). The development of copper metallurgy during the Early Bronze Age of the Southern Levant: evidence from the Faynan Region, Southern Jordan., University of Sheffield. Aitken, M. J., Alcock, P. A., G.D., B., & Shaw, C. J. (1981). Archaeomagnetic determination of the past geomagnetic intensity using ancient ceramics: allowance for anisotropy. Archaeometry, 23, 53-64. Avner, U. (2002). Studies in the Material and Spiritual Culture of the Negev and Sinai Populations, During the 6th-3rd Millennia B.C., Hebrew University of Jerusalem, Jerusalem. Avner, U., & Magness, J. (1998). Early Islamic Settlement in the Southern Negev. Bulletin of the American Schools of Oriental Research(310), 39-57. Avner, U., & Naor, A. (1978). A Survey in the Eilat Area. Hadashot Arkheologiot(67/68), 66-68 (in Hebrew). Bachmann, H. G. (1980). Early Copper Smelting Techniques in Sinai and in the Negev as Deduced from Slag Investigations. In P. T. Craddock (Ed.), Scientific Studies in Early Mining and Extractive Metallurgy. (pp. 103-134). London. Balthazar, J. W. (1990). Copper and Bronze Working in Early through Middle Bronze Age Cyprus. Partille. Ben-Yosef, E., Ron, H., Tauxe, L., Agnon, A., Genevey, A., Levy, T. E., et al. (2008). Application of copper slag in geomagnetic archaeointensity research. Journal of Geophysical Research, (revised). Bercovici, A. (1978). Flint Implements from Timna Site 39. In B. Rothenberg, R. F. Tylecote & P. J. Boydell (Eds.), Chalcolithic Copper Smelting, ArchaeoMetallurgy 1. (pp. 16-20). London. Bloxham, J. (2003). Dipole decay, secular variation and reversals. Eos, Trans. AGU, Fall Meet. Suppl, 84, F34.

54

55

Burleigh, R., & Hewson, A. (1979). British Museum Natural Radiocarbon Measurements XI. Radiocarbon, 21(3), 339-352. Burton, M., & Levy, T. (2001). The chalcolithic radiocarbon record and its use in southern Levantine archaeology. In H. Bruins, I. Carmi & E. Boaretto (Eds.), Near East Chronology: Archeology and Environment (Vol. 43, pp. 1223-1246): Radiocarbon. Coe, R. S. (1967). Paleointensities of the Earth's magnetic field determined from Tertiary and Quaternary rocks. Jour. Geophys. Res., 72, 3247--5281. Constable, C., & Korte, M. (2006). Is Earth's magnetic field reversing? Earth and Planetary Science Letters, 246(1-2), 1-16. Constable, C. G., Johnson, C. L., & Lund, S. P. (2000). Global geomagnetic field models for the past 3000 years: transient or permanent flux lobes? Phil Trans Roy Soc London, Series A, 358(1768), 991-1008. Courtillot, V., Gallet, Y., Le-Mouel, J.-L., Fluteau, F., & Genevey, A. (2007). Are there connections between the Earth's mangetic field and climate? Earth Planet. Sci. Lett., 253, 328-339. Courtilot, V., Valet, J. P., Hulot, G., & Mouel, J. L. L. (1992). The Earth's Magnetic Field: Which Geometry? Eos Trans. AGU, 73, 337, 340-342. Craddock, P. T. (2001). From hearth to furnace: evidence for the earliest metal smelting technologies in the Eastern Mediterranean. Paléorient, 26(2), 151-165. Donadini, F., Korthonen, K., Riisager, P., & J., P. L. (2006). Database for Holocene Geomagnetic Intensity Information. Eos Trans. AGU, 87(14), 137. Folgheraiter, M. (1899). Sur les variations séculaires de l'inclinaison magnétique dans l'antiquité. Jour. de Phys., 5, 660-667. Fox, J. M. W., & Aitken, M. J. (1980). Cooling-rate dependence of thermoremanent magnetization. Nature, 283, 462--463. Frank, M. (2000). Comparison of cosmogenic radionuclide production and geomagnetic field intensity over the last 200,000 years. Phil Trans. R. Soc. Lond. A, 358, 10891107. Gallet, Y., Genevey, A., & Fluteau, F. (2005). Does Earth’s magnetic field secular variation control centennial climate change? Earth Planet. Sci. Lett., 236, 339– 347.

56

Gallet, Y., Genevey, A., Le Goff, M., Fluteau, F., & Eshraghi, S. A. (2006). Possible impact of the Earth's magnetic field on the history of ancient civilizations. Earth Planet. Sci. Lett., 246, 17-26. Gallet, Y., & Le Goff, M. (2006). High-temperature archeointensity measurements from Mesopotamia. Earth Planet. Sci. Lett., 241, 159-173. Genevey, A., & Gallet, Y. (2002). Intensity of the geomagnetic field in western Europe over the past 2000 years: New data from ancient French pottery. J. Geophys. Res., 107 (B11), 2285. Genevey, A., Gallet, Y., & Margueron, J. (2003). Eight thousand years of geomagnetic field intensity variations in the eastern Mediterranean. J. Geophys. Res, 108, doi:10.1029/2001JB001612. Genz, H. (1997). Problems in Defining a Chalcolithic for Southern Jordan. In H. G. K. Gebel, Z. Kafafi & G. O. Rollefson (Eds.), The Prehistory of Jordan II. Perspectives from 1997. Studies in Early Near Eastern Production, Subsistence, and Environment 4 (pp. 441-448). Berlin. Gilead, I. (1994). The History of the Chalcolithic Settlement in the Nahal Beer Sheva Area: The Radiocarbon Aspect. Bulletin of the American Schools of Oriental Research, 296, 1-13. Gilead, I., Rosen, S., Fabian, P., & Rothenberg, B. (1992). New Archaeological Evidence for the Beginning of Metallurgy in the Southern Levant. Excavation at Tell Abu Matar, Beersheba (Israel) 1990/1. Institute for Archaeo-Metallurgical Studies, 18, 11-14. Golani, A. (2004). Salvage Excavations at the Early Bronze Age Site of Ashqelon, Afridar – Area E. 'Atiqot, 45, 9-62. Gophna, R. (2004). Excavations at Ashqelon, Afridar – Introduction. 'Atiqot, 45, 1-8. Görsdorf , J. (2002). New 14C-Datings of Prehistoric Settlements in the South of Jordan. Orient-Archaologie, 5, 333-339. Gram-Jensen, M., Abrahamsen, N., & Chauvin, A. (2000). Archaeomagnetic intensity in Denmark. Phys. Chem Earth (A), 25, 525-531. Hanbury-Tenison, J. W. (1986). The Late Chalcolithic to Early Bronze I Transition in Palestine and Transjordan. Oxford.

57 Hauptmann, A. (1989). The earliest periods of copper metallurgy in Feinan, Jordan. In A. P. Hauptmann, E. Wager, G.A. (Ed.), Old World Archaeometallurgy (pp. 119135). Bochum: Deutsche Bergbau-Museum. Hauptmann, A. (2000). Zur frühen Metallurgie des Kupfers in Fenan/Jordanien. Bochum: Der Anschnitt. Hauptmann, A. (2007). The Archaeometallurgy of Copper - Evidence from Faynan, Jordan. Berlin: Springer. Hauptmann, A., Khalil, L., & Schmitt-Strecker, S. (2004). Evidence for Late Chalcolithic/Early Bronze Age I Copper Production from Timna Ores at Tall Magass, Aqaba. Levant. Hulot, G., Eymin, C., Langlais, B., Mandea, M., & Olsen, N. (2002). Small-scale structure of the geodynamo inferred from Oersted and Magsat satellite data. Nature, 416, 620-623. Ilan, D. (1999). Northeastern Israel in the Iron Age I: Cultural, Socioeconomic and Political Perspectives., Tel-Aviv University, Tel-Aviv. Jackson, A., Jonkers, A. R. T., & Walker, M. R. (2000). Four centuries of geomagnetic secular variation from historical records. Phil Trans Roy Soc London, Series A, 358(1768), 957-990. Jordanova, N., Kovacheva, M., & Kostadinove, M. (2004). Archaeomagnetic investigation and dating of Neolighic archaeological site (Kovachevo) from Bulgaria. Phys. Earth Planet. Inter., 147, 89-102. Khalil, L. (1987). Preliminary Report on the 1985 Season of Excavation at el-MagassAqaba. Annual of the Department of Antiquities of Jordan, 31, 481-483. Khalil, L. (1992). Some Technological Features from a Chalcolithic Site at MagassAqaba. Studies in History and Archaeology of Jordan, IV, 143-148. Khalil, L. (1995). The Second Season of Excavation at al-Magass-Aqaba, 1990. Annual of the Department of Antiquities of Jordan, 39, 65-79. Khalil, L., & Eichmann, R. (1999). Archaeological Survey and Excavation at Wadi AlYutum and Tall Al-Magass Area-‘Aqaba (ASEYM). A Preliminary Report on the First Season 1998. Annual of the Department of Antiquities of Jordan, 43, 501520. Killick, D. J. (2004). Social Constructionist Approaches to the Study of Technology. World Archaeology, 36(4), 571-578.

58

Kitagawa, H., & Plicht, J. (1998). Atmospheric Radiocarbon Calibration to 45,000 yr B.P.: Late Glacial Fluctuations and Cosmogenic Isotope Production. Science, 279, 1187-1190. Kono, M. (2007). Geomagnetism in perspective. In M. Kono (Ed.), Geomagnetism (Vol. 5, pp. 1-32). Amsterdam: Elsevier. Korte, M., & Constable, C. (2005a). Continuous geomagnetic field models for the past 7 millennia: 2. CALS7K. Geochem., Geophys., Geosyst., 6, Q02H16: DOI 10.1029/2004GC000801. Korte, M., & Constable, C. (2005b). The geomagnetic dipole moment over the last 7000 years - new results from a global model. Earth Planet. Sci. Lett., 236, 348-358. Korte, M., Genevey, A., Constable, C., Frank, U., & Schnepp, E. (2005). Continuous geomagnetic field models for the past 7 millennia: 1. A new global data compilation. Geochem., Geophys., Geosyst., 6(Q02H15), Q02H15; DOI 10.1029/2004GC000800. Kovacheva, M., Hedley, I., Jordanova, N., Kostadinove, M., & Gigov, v. (2004). Archaeomagnetic dating of archaeological sites from Switzerland and Bulgaria. Journal of Archaeological Science 31, 1463-1479. Lanos, P. (2003). Beyesian inference of calibration curve, application to Archaeomagnetism. In C. E. Buck & A. R. Millard (Eds.), Tools for Chronology, Crossing Disciplinary Boundaries (pp. 43-82). London: Springer-Verlag. Le Goff, M., Gallet, Y., Genevey, A., & Warmé, N. (2002). On archeomagnetic secular variation curves and archeomagnetic dating. . Phys. Earth Planet. Inter. , 134, 203-211. Levy, T. (2006). Grand Narratives, Technological Revolutions and the Past: Deep-Time Studies of Metallurgy and Social Evolution in the Eastern Mediterranean. In O. LaBianca & S. A. Scham (Eds.), Connectivity in Antiquity - Globalization as a Long-Term Historical Process (pp. 10-25). London: Equinox. Levy, T. (2007). Journey to the Copper Age - Archaeology in the Holy Land. San Diego: San Diego Museum of Man. Levy, T. E., Adams, R. B., Hauptmann, A., Prange, M., Schmitt-Strecker, S., & Najjar, M. (2002). Early Bronze Age Metallurgy: a newly discovered copper manufactory in southern Jordan. Antiquity, 76, 425-437.

59 Levy, T. E., Adams, R. B., Najjar, M., Robinson, M., & Higham, T. (2004). Reassessing the chronology of Biblical Edom: new excavations and 14C dates from Khirbat en Nahas (Jordan). Antiquity, 78, 863-876. Levy, T. E., & Shalev, S. (1989). Prehistoric metalworking in the southern Levant: archaeometallurgical and social perspectives. World Archaeology, 20, 352-272. Levy, T. E., & Smith, N. G. (2007). On-Site Digital Archaeology: GIS-Based Excavation Recording in Southern Jordan. In T. E. Levy, M. Daviau, R. Younker & M. Shaer (Eds.), Crossing Jordan – North American Contributions to the Archaeology of Jordan (pp. 47-58). London: Equinox. Merkel, J. F. (1990). Experimental Reconstruction of Bronze Age Copper Smelting Based on Archaeological Evidence from Timna. In B. Rothenberg (Ed.), Researches in the Arabah 1959-1984, Vol. 2, The Ancient Metallurgy of Copper. (pp. 78-122). London: Institute for Archaeo-Metallurgical Studies. Meshel, Z. (1993). Yotvata. In E. Stern (Ed.), The New Encyclopedia of Archaeological Excavations in the Holy Land (Vol. IV, pp. 1517-1520). Jerusalem. Muhly, J. D. (1973). Copper and Tin. New-Haven. Muhly, J. D. (1976). Supplement to Copper and Tin. Hamden. Muhly, J. D. (1984). Timna and King Solomon. Bibliotheca Orientalis, XLI(3-4), 276292. Mulkay, M. (1978). Science and the Sociology of Knowledge. London: George Allen and Unwin. Nagata, T. (1961). Rock Magnetism. Néel, L. (1949). Théorie du trainage magnétique des ferromagneétiques en grains fines avec applications aux terres cuites. . Annales de Geophysique, 5, 99-136. Néel, L. (1955). Some theoretical aspects of rock-magnetism. Advances in Physics 4, 191-243. Odah, H., Heider, F., Hussain, A. G., Hoffmann, V., Soffel, H., & ElGamili, M. (1995). Paleointensity of the geomagnetic field in Egypt from 4000 BC to 150 AD using the Thellier method. J. Geomag. Geoelectr., 47, 41-58. Peristykh, A. N., & Damon, P. E. (2003). Persistense of the glessberg 88-year solar cycle ofver thelast 12,000 years: Evidence from cosmogenic isotopes. J. Geophys. Res., 108, doi: 10.1029/2002JA009390.

60

Pfaffenberger, B. (1992). Social Anthropology of Technology. Annual Review of Anthropology, 21, 491-516. Rothenberg, B. (1962). Ancient Copper Industries in the Western Arabah. Palestine Exploration Quarterly, 5-71. Rothenberg, B. (1967). Negev: Archaeology in the Negev and the Arabah. Tel-Aviv: Masada. Rothenberg, B. (1978). Excavation at Timna Site 39. In B. Rothenberg (Ed.), Chalcotithic Copper Smelting. (pp. 1-26). London: Institute for Archaeometallurgical Studies. Rothenberg, B. (1980). Die Archaeologie des Verhüttungslagers Site 30. In H. G. Conrad & B. Rothenberg (Eds.), Antikes Kupfer im Timna-Tal. (pp. 187-214). Bochum. Rothenberg, B. (1990a). The Chalcolithic Copper Smelting Furnace in the Timna ValleyIts Discovery and the Strange Argument Surrounding Its Dating. Institute for Archaeo-Metallurgical Studies News letter, 15-16, 9-12. Rothenberg, B. (1999a). Archaeo-Metallurgical Researches in the Southern Arabah 19591990. Part 1: Late Pottery Neolothic to Early Bronze IV. Palestine Exploration Quarterly, 131, 68-89. Rothenberg, B. (1999b). Archaeo-Metallurgical Researches in the Southern Arabah 19591990. Part 2: Egyptian New Kingdom (Ramesside) to Early Islam. Palestine Exploration Quarterly, 131, 149 - 175. Rothenberg, B. (Ed.). (1990b). Researches in the Araba 1959-1984, Vol. 2, The Ancient Metallurgy of Copper. London: Institute for Archaeo-Metallurgical Studies. Rothenberg, B., & Glass, J. (1992). Beginnings and Development of Early Metallurgy and the Settlement and Chronology of the Western Arabah, from the Chalcolithic Period to Early Bronze Age IV. . Levant, 24, 141-157. Rothenberg, B., & Merkel, J. F. (1995). Late Neolithic Copper Smelting in the Arabah. Institute for Archaeo-Metallurgical Studies News letter, 19, 1-7. Rothenberg, B., & Merkel, J. F. (1998). Chalcolithic, 5th Millennium B.C., Copper Smelting at Timna. Institute for Archaeo-Metallurgical Studies News letter, 20, 13. Rothenberg, B., & Shaw, C. T. (1990a). Chalcolithic and Early Bronze Age IV Copper Mining and Smelting in Timna Valley (Israel) – Excavations 1984-1990. In P.

61 Pertovic & S. Durdekanovic (Eds.), Ancient Mining and Metallurgy in Southern Europe. (pp. 281-294). Belgrad. Rothenberg, B., & Shaw, C. T. (1990b). The Discovery of a Copper Mine and smelter from the End of the Early Bronze Age (EBIV) in the Timna Valley. Institute for Archaeo-Metallurgical Studies News letter, 15-16, 1-8. Schiffer, M. B. (2001). Toward an Anthropology of Technology. In M. B. Schiffer (Ed.), Anthropological Perspectives on Technology (pp. 1-15). Albuquerque: University of New Mexico Press. Segal, D., & Carmi, I. (1996). Rehovot Radiocarbon Date List V. 'Atiqot, XXIX, 79-106. Segal, D., & Carmi, I. (2004). Determination of age using the 14C method on archaeobotanical samples from Ashqelon, Afridar – Area E. 'Atiqot, 45, 119-120. Segal, I., Halicz, L., & Kamenski, A. (2004). The Metallurgical Remains from Ashqelon, Afridar – Areas E, G and H. 'Atiqot, 45, 311-330. Segal, I., Rothenberg, B., & Bar-Matthews, M. (1998). Smelting Slag from Prehistoric Sites F2 and N3 in Timna, SW Arabah, Israel. In T. Rehren, A. Hauptmann & J. D. Muhly (Eds.), Metallurgica Antiqua 8 (pp. 223-234). Bochum. Selkin, P. A., Gee, J. S., Tauxe, L., Meurer, W. P., & Newell, A. (2000). The effect of remanence anisotropy on paleointensity estimates: A case study from the Archean Stillwater complex. Earth Planet. Sci. Lett., 182, 403-416. Shalev, S. (1994). Change in metal production from the Chalcolithic Period to the Early Bronze Age in Israel and Jordan. Antiquity, 68, 630-637. Shalev, S., & Northover, P. (1987). The Chalcolithic Metal and Metalworking from Shiqmim. In T. E. Levy (Ed.), Shipmim I – Studies Concerning Chalcolithic Societies in the Northern Negev Desert, Israel (1982-1984). (pp. 357-371). Oxford: British Archaeological Reports, International Series 356. Sharon, M., Avner, U., & Nahlieli, D. (1996). An Early Islamic Mosque Near Be'er Ora in the Southern Negev: Possible Evidence for an Early Eastern Qiblah? 'Atiqot, XXX, 107-119. Tauxe, L. (2006). Long-term trends in paleointensity: The contribution of DSDP/ODP submarine basaltic glass collections. Physics of the Earth and Planetary Interiors, 156(3-4), 223-241. Tauxe, L., & Staudigel, H. (2004). Strength of the geomagnetic field in the Cretaceous Normal Superchron: New data from submarine basaltic glass of the Troodos

62 Ophiolite. Geochem. Geophys. Geosyst., 5(2), Q02H06, doi:10.1029/2003GC000635. Tauxe, L., & Yamazaki, T. (2007). Paleointensities. In M. Kono (Ed.), Geomagnetism (Vol. 5, pp. 509-564). Amsterdam: Elsevier. Thellier, E. (1938). Sur l'aimantation des terres cuites et ses applications géophysique. Ann. Inst. Phys. Globe Univ. Paris, 16, 157-302. Thellier, E., & Thellier, O. (1959). Sur l'intensité du champ magnétique terrestre dans le passé historique et géologique. Ann. Geophys., 15, 285--378. Usoskin, I. G., Solanki, S. K., & Korte, M. (2006). Solar activity reconstructed over the last 7000 years: the influence of geomagnetic field changes. Geophys. Res. Lett., 33, doi: 10.1029/2006GL25921. Valet, J. P. (2003). Time variations in geomagnetic intensity - art. no. 1004. Reviews of Geophysics, 41(1), 1004. Weisgerber, G., & Hauptmann, A. (1988). Early Copper Mining and Smelting in Palestine. In R. Maddin (Ed.), The Beginning of the Use of Metals and Alloys. Papers from the Second International Conference on the Beginning of the Use of Metals and Alloys, Zhengzhou, China, 21-26 October 1986. (pp. 52-62). Cambridge. Willies, L. (1990). Exploring the Ancient Copper Mines of the Wadi Amram (South Arabah). Institute for Archaeo-Metallurgical Studies News letter, 15-16, 12-15. Wilson, A. J. e. (1983). (Editor notes). Institute for Archaeo-Metallurgical Studies News letter, 5. Yamazaki, T., & Oda, H. (2004). Intensity-inclination correlation for long-term secular variation of the geomagnetic field and its relevance to persistent non-dipole components. In J. E. T. Channell, et al. (Ed.), Timescales of the Paleomagnetic Field (Vol. 145, pp. 287-298). Washington, D.C.: American Geophysical Union. Yener, S. A. (2000). The Domestication of Metals. The Rise of Complex Metal Industries in Anatolia. Brill. Yu, Y., & Tauxe, L. (2005). Testing the IZZI protocol of geomagnetic field intensity determination. Geochem., Geophys., Geosyst., in press. Yu, Y., Tauxe, L., & Genevey, A. (2004). Toward an optimal geomagnetic field intensity determination technique. Geochemistry, Geophysics, Geosystems, 5(2), Q02H07, doi:10.1029/2003GC000630.