Qumran Revisited: A Reassessment of the Archaeology of the site and its texts. BAR International Series 2520 (Oxford 2013).

Table of Contents Page Forewords: David Stacey Gregory Doudna 3 5 A Reassessment of the Stratigraphy of Qumran by David Stacey 1. Early conclusions reached during the excavations in the 1950s. 7 2. When was the ‘main’ aqueduct built? 2.1 – The aqueduct and the dam north-west of the inhabited building 2.2 – The Southern Wall, ‘W1’, of the ‘Main Building’ 2.3 – Pool Locus 48/9 11 15 18 21 3. Qumran in the Hasmonean and early Herodian Period 100-31/20 BCE. (Plans 1, 2) 3.1 – The Tower 3.2 – A note on the man made caves in the terraces near to Qumran 34 36 37 4. The expansion of Qumran in the time of Herod, 20-15 BCE – 1 BCE. (Plans 3-7) 4.1 – The aqueduct and pools 4.2(i) – Two pools associated with L56/58 4.2(ii) – The implications 4.3 – Overflows within the system 4.4 – The main building 38 38 40 40 42 43 5. Herod and post Herod occupation. 5.1 – The south-west of the ‘main building’ 5.2 – Other Loci in the ‘Main Building’ 5.2(i) – Loci in the north-east of the ‘Main Building’ 5.3 – Some Loci ouside of the ‘Main’ Building 5.4 – A brief note on the archaeological methodology, and its possible implications 45 45 47 49 49 51 6. What was the character of the industrial activity at Qumran? 6.1 – Potential uses of the ‘industrial area’ and the water resources 6.1(i) – Industries associated with sheep and goats 6.1(ii) – Industries not associated with sheep and goats 6.1(iii) – Other seasonal activities 6.2 – Seasonal activities and the character of habitation at Qumran 52 53 53 58 61 61 7. The change of character of Qumran in the Herodian period. 66 8. Qumran after Herod 70 71 1 List of Figures 19 19 20 19 20 20 24 24 25 25 48 56 69 Fig. 1 Fig. 2 Fig. 3 Fig. 4 Fig. 5 Fig. 6 Fig. 7 Fig. 8 Fig. 9 Fig. 10 Fig. 11 Fig. 12 Fig. 13 Figs. 1- 6: Copyright Shimon Gibson Figs. 7-11: Copyright David Stacey Fig. 12: From now defunct journal, Old West Riding. Copyright holder unknown. Fig. 13: Copyright Nikos Kokkinos List of Plans 26 27 28 29 30 31 32 33 33 33 33 Plan 1 Plan 2 Plan 3 Plan 4 Plan 5 Plan 6 Plan 7 Plan 8 Plan 9 Plan 10 Plan 11 The Sect of the Qumran Texts and its Leading Role in the Temple in Jerusalem During Much of the First Century BCE: Toward a New Framework for Understanding by Gregory L. Doudna 75 Who Were Interred in the Qumran Cemetery? On Ethnic Identities and the Archaeology of Death and Burial by Gideon Avni 125 List of Figures Fig. 1 Fig. 2 Fig. 3 Fig. 4 Fig. 5 Fig. 6 Fig. 7 Fig. 8 133 133 134 134 135 135 136 136 Fig. 1: Copyright Yitzhak Magen Figs. 3, 5: Copyright Gideon Avni Fig. 7: Copyright York Archaeological Trust Fig. 8: Copyright D. Adar Bibliography 137 2 Forewords Introduction from an Archaeologist’s Point of View by David Stacey I became interested in the archaeology of Qumran after ten seasons working as a field archaeologist with Ehud Netzer on his comprehensive excavations of the Hasmonean and Herodian Palaces and Royal Estate in Jericho, and in Herodium and Masada. Our interpretations of how Hasmonean Jericho had developed differed: Ehud saw it as the result of an overall grand plan whereas I detected a more gradual, piecemeal development, with additions, often necessitating minor alterations in the way the aqueduct(s) distributed water, made when revenue from the gradually expanding agricultural estate made them affordable. The archaeology never allowed us to ‘prove’ either interpretation. Small probes conducted where we both agreed crucial ‘proof’ could be found invariably led to disappointment: the ‘proof’ having long since been eroded, collapsed into the moat, or been destroyed by later alterations. Following the publication of the Final Reports, which only gave Ehud’s interpretation, I posted on the internet the suggestion that it was impossible to view Qumran in isolation from the Jericho estate (Stacey 2004c) and followed it with a short article (Stacey 2006) discussing the way I saw the development of Hasmonean Jericho. Two articles on Qumran followed (Stacey 2007 & 2008) and I want to thank Koninklijke Brill NV for permission to reprint a greatly revised version of Some Archaeological Observations on the Aqueducts of Qumran which appeared in DSD Vol 14:2 (2007) ISSN 09290761, Pp. 222-243; and the Anglo-Israel Archaeological Society for permission to reprint, within the text of this volume, a revised version of Seasonal Industries at Qumran in BAIAS Vol 26 (2008) ISSN 0266-2442, Pp. 7-29. There had been a settlement in Jericho since Neolithic times which drew its waters from Ain es-Sultan, a spring on the Jericho plain, whose waters were exploited to the full by the late Hellenistic village. The Hasmonean dynasty, which was in urgent need of a steady source of income, realised that large quantities of two high value crops, dates and balsam, could be grown on previously uncultivated land if only they could find a new source of water with which to irrigate it. To do so they had to go to considerable lengths to bring water by aqueduct from Ain Qelt, some 8 km to the west in the Wadi Qelt. Their endeavours were rewarded and they were successful in developing irrigated agriculture on the relatively flat ground to the north of where the wadi debouches from its canyon. The flat land to the south, however, was irrigated via a pool, Birket Musa ‘the largest man-made water tank... in the country’ (Netzer and Garbrecht 2002, 377), which must have been fed by diverting some of the water that ran in the wadi. Only as agriculture flourished did palaces begin to be built on the north bank utilising the same water supply. To expand agriculture to new fields east of the first plantations, and to allow for larger palaces to be built, the Hasmoneans had to bring water from Ain Na’aran, to the north of Ain es-Sultan, by an aqueduct that ran for some 5 km along ‘tortuous contours’ (ibid: 373) on the scarp face. As it would have required frequent repairs following winter floods its water would have been ‘expensive’. It was this aqueduct that was, in the main, exploited for use in the gradually expanding palace complexes, whose growth over ever wider areas necessitated the occasional diversion of parts of the main channel northwards to by-pass the developments. There were four main diversions over time. To track the phases of the additions it was essential to keep in mind from which of these diversions that particular phase was drawing water. When Herod wanted to further increase the water supply for the estate he brought water to the head of the Na’aran aqueduct from Ain el Auja a further 8 km to the north. The total course to the fields would have been at least 14 km. He also exploited the two upper springs in Wadi Qelt by building a 20 km long aqueduct, which ‘incorporated ten bridges and five tunnels in order to overcome the many small ravines descending into the wadi’ (Garbrecht and Peleg 1994: 167). The Qelt water was used to develop fields south of those irrigated by Birkat Musa, fields which may have reached to within 8 km of Qumran. Thus ultimately all the water required for the success of the Royal Estate had to be brought to the site either by a 14 km long aqueduct from the north; by two from the west, one 8 km long, the other over 20 km; or was gathered in the largest open pool in Palestine. With the difficulty that was encountered bringing water to the Royal Estate it seemed certain to me that both the Hasmoneans and Herod would have exploited, and kept under their control, the locally falling rain water that could be gathered with comparative ease at Qumran only 14 km to the south. Moreover to relinquish a tight hold over a site at the foot of a track leading up to the Buqe’ah, and in proximity to a number of desert forts and palaces would be a sign of political weakness. When visiting Qumran my knowledge of the archaeology of Jericho and other nearby, contemporary sites, in particular the development of the water systems, was always in the back of my mind and it seemed to me that some of the interpretations appearing in publications on the archaeology of Qumran by, inter alia, de Vaux, Magness, Hirschfeld and Humbert, were mistaken. Very early on in the history of Qumran exploration the ‘Qumran-Essene’ hypothesis – in which it was claimed that the scrolls found in nearby caves were written at Qumran and were the library of a sectarian, ‘Essene’ community living there – was formulated as a ‘fact’. The more I studied the archaeology of the site the less 3 likely it seemed that any scroll was ever written there. Moreover I suspected that the understandable excitement with which modern scholars first delved into the scrolls may have led them to overinflate the scrolls’ importance in the general scheme of things. For those on the Dead Sea littoral the pragmatic realities of economically exploiting a marginal land were probably of greater concern. My conclusions are based on what can be seen in the field and on what is published on the archaeology. No final report of the excavations has ever been published and it is possible that such a final publication would demand changes to my conclusions. A better understanding of the development of certain loci, and more precision in the dating is certainly to be hoped for. Many people have written their interpretations of the archaeology, far too many for me to claim that I am familiar with them all. I refer to points raised by some but it was not my intention here to review them all. To move beyond the archaeology to consider the manner in which the site may have been exploited, and the part it may have played in the regional economy requires more speculation. Explicit evidence is often lacking but my suggestions are no more speculative than is the Qumran-Essene hypothesis. I am not a scholar of ancient languages or of the Dead Sea Scrolls but I am honoured to have been able to draw on the expertise of Greg Doudna who has contributed an important and insightful essay on textual matters. I thank him for his diligence and enthusiasm, particularly when glitches seemed to delay publication ever further into the future, and for his ability to clarify certain issues. I want to thank Gideon Avni, head of the Excavation and Survey Department of the Israel Antiquities Authority, for contributing his chapter on the cemeteries of Qumran. It is a revised version of an article that first appeared in Hebrew in Cathedra 131 in 2009. Many thanks to Prof. Oleson for sending me a pdf of his 2007 article, Christian Cloke for a pdf of his unpublished 2007 MA, and Dr. Kamash for a pdf of her 2010 book, none of which were immediately available in my local academic library. Special thanks to Leen Ritmeyer for his clear, illustrative drawings on p. 35 and p. 39. Over the years I have had many pleasant discussions on the archaeology of Qumran with a number of people, in particular, Benny Arubas, Dennis Mizzi, Rachel Bar-Nathan and Joan Taylor, and I hope that we may have more in the future. I want to thank Mizzi for permission to quote selectively from his Oxford PhD thesis (2009) The Archaeology of Khirbet Qumran: a comparative approach. The editors and reviewers of Archaeopress made some helpful suggestions. I must thank my daughter, Abi Stacey, without whom a computer ready copy of the text would never have seen the light of day. Any mistakes are, of course, my responsibility. An unfortunate hiatus between the initial writing of much of this work and its final publication has meant that some recent publications have not received the attention they deserve. David Stacey Saffron Walden, Essex. UK 4 Introduction from a Text Scholar’s Point of View by Greg Doudna I am honored to have been invited by David Stacey to participate in the publication of his important and fresh perspective on the archaeology of Qumran, where ancient Jewish texts were found in caves adjoining and radiating northward from the site. David Stacey writes from his experience as an archaeologist who worked for ten years on the Netzer excavations at Hasmonean/Herodian Jericho, the nearest major site to Qumran. Out of this experience Stacey does not interpret Qumran as if Qumran was sui generis, a unique site to be interpreted in light of the contents of the texts found in its caves, its material remains interpreted as reflections of a separatist sect inhabiting the site in an ideological world of its own. David Stacey analyzes Qumran from the perspective of regional economic development and industries affecting the site without regard to the contents of the texts. Rather than interpret the remains of Qumran in terms of reconstructed sectarian practices (activities presumed to have included collection and production of the texts themselves), Stacey examines Qumran independently of the texts beyond accounting for the texts’ physical presence. Stacey accounts for the presence of the scrolls in Qumran’s caves in terms of permanent disposals of texts brought to the site for that purpose. Of particular interest is Stacey’s argument that Qumran was an extension of the Hasmonean estate in Jericho. Although there are precursors to this idea—notably Pessah Bar-Adon’s influential 1981 argument that Qumran was part of a Hasmonean program of fortification and exploitation of economic resources near the Dead Sea, such that an interpretation of Qumran as beginning as a Hasmonean state initiative is now embraced by most current studies of Qumran’s archaeology (Humbert 1994, Hirschfeld 2004b, Magen and Peleg 2007, Cargill 2009, Mizzi 2009, Taylor 2012b)—yet Stacey differs in his analysis in comparison to most others who advocate a Hasmonean state initiative origin of Qumran. For Stacey sees Hasmonean Qumran being further developed by Herod with no fundamental rupture or break in its industries and functioning. This is in contrast to several of the analyses cited (Humbert, Cargill, Mizzi, Taylor) who suppose there was a radical change in occupation of the site in the time of Herod. In the line of thinking of these analyses, a new group—sectarians—came to Qumran late in the 1st century BCE bringing their scrolls with them and producing more scrolls at the site, which they continued to do until the First Revolt of the 1st century CE. The reason these scholars identify a new sect arriving at the site late in the 1st century BCE, and replacing the site’s former inhabitants, is because they wish to account for the presence of the scrolls at the site, and believe this is the way to do so: first a Hasmonean period at the site (no connection to the scrolls), then replaced by the sect connected to the scrolls. This was not de Vaux’s scheme (de Vaux had the sect at Qumran from the beginning), and abandoned in this scheme is the old idea that a 2nd century BCE Teacher of Righteousness of the texts came to Qumran with his little band of followers to build the site after he was exiled from Jerusalem. But it is a way to keep the traditional model of the sect of the texts and its relationship to the site of Qumran intact, in a modified form. But why not, a reader who has been following the discussion closely to this point might ask, simply identify the people of the scrolls as the people operating the site when Qumran was a Hasmonean outpost? Why the necessity to suppose two distinct sets of people? Such a question seemingly comes from a naive reader unfamiliar with Qumran scholarship. For if there is one point upon which scholars have been in agreement with almost complete unanimity for as long as anyone can remember, it is that the sect of the Qumran texts was opposed to the Hasmonean high priests. In prevailing Qumran scholarship it is considered simply inconceivable that the people of the texts of Qumran could have been favorable to, for example, Hyrcanus II, the high priest of the temple in Jerusalem 76-67 and 63-40 BCE. My essay in the present volume builds on a paper I was invited to deliver at a 2009 conference held in Copenhagen (Doudna 2011). My essay shows that the naive reader’s question referred to above is actually quite astute. My essay challenges the reasons claimed for supposing that the Qumran texts were opposed to the Hasmonean high priests. I show that nothing in the Qumran Community Rule (S) texts calls for reading those texts as opposed to the temple or to the priests who controlled the temple. I show that such interpretations of the Qumran S texts are unfounded and chimerical, no matter how deeply ingrained such interpretations have been in scholarly discourse. I argue that, in fact, nothing in the Qumran texts calls for supposing the sectarian texts were other than supportive of most Hasmonean high priests. Texts which have been interpreted as polemical against all Hasmonean rulers and their regimes instead become polemical against one Hasmonean ruler’s regime from supporters of the previous high priest. Condemnations in some texts of a Hasmonean regime become not a rejection of all Hasmonean high priests but rather arise out of the authors’ loyalty to, as they interpreted it, the legitimate Hasmonean high priest who had been cast into exile. 5 I show that traditional arguments for supposing an adversarial relationship between the sect of the Qumran texts and the Hasmonean high priests evaporate upon examination: there is no sign in the texts of calendar conflict between the sect and the Hasmonean high priests; no criticism for combining king and high priest; no notion of rival priestly ancestries; no opposition to Alexander Jannaeus, or to John Hyrcanus I before him. Contrary to common conceptions, none of these notions are in the Qumran texts in any way. Instead of the sect of the Qumran texts being opposed to the Hasmonean high priests, the sect of the texts was the sect of the Hasmonean high priests. In the past, archaeological analyses of Qumran which follow the method of David Stacey—of interpreting the material remains of the site independently of the contents of the texts—have been marginalized on the grounds that they ‘ignore the evidence of the scrolls’ (which are part of the archaeological realia of the site, it is pointed out). Although my essay stands independently, I place it in this volume for the purpose of shielding from this line of criticism the work of archaeologists who seek to evaluate the material remains of Qumran objectively. I intend my essay to open spaces to allow such archaeologists’ voices to be heard, liberated from the stifling constraints of a long-dominant interpretive filter promulgated in the name of the ancient Essenes. In fact Stacey’s analysis of Qumran’s archaeology done independently of the texts agrees very well with the texts. Stacey’s argument that Qumran was an extension of Hasmonean Jericho becomes very sensible, given the congruence of ruling priests at Hasmonean Jericho and the disposals of large numbers of religious texts reflective of those priests at nearby Qumran. David Stacey and I are each solely responsible for our respective pieces in the present volume: Stacey’s dealing with archaeological interpretation, and mine engaging the texts. Stacey and I have made no attempt to harmonize every detail, and that is as it should be. Stacey’s study and my article, as well as Gideon Avni’s analysis of the Qumran cemetery, should be read as distinct perspectives, distinct voices, offering a significant degree of overlap on relevant points, yet retaining autonomy. I am thankful to Marie-France Dion, Russell Gmirkin, Søren Holst, James Pasto, Kaare Lund Rasmussen, Annette Steudel, Joan Taylor, Thomas Thompson, Adam Ward, Ian Young, and my wife, Anne Caroline Doudna, as well as David Stacey, for deeply appreciated support in reading and commenting on sections as I prepared this contribution for this volume. Barbara Siegel was helpful in making available publications of her father, Manfred Lehmann. I thank Abi Stacey, and the editors and reviewers at Archaeopress, for the behind-the-scenes work involved in bringing this volume to press. I hope this volume will assist all of us, text scholars and archaeologists alike, as well as readers from other disciplines and the interested public, in approaching a better understanding of the ancient texts of Qumran and the site where these texts were found. Greg Doudna Bellingham, Washington USA 6 1. Early conclusions reached during the excavations in the 1950s prescribed manner) (Sukenik 1948-50), but de Vaux, in 1949, before excavations on the ruins had taken place, contended that they ‘are not some pieces thrown aside, but constitute archives, or a library, taken to safety at some particularly critical moment’ (de Vaux 1949: 236; 1973: 55-6, 103). Although there is no reason to suppose that geniza deposits were, as he implied, carelessly ‘thrown aside’, indeed they would be treated with respect, other scholars dismissed the geniza concept with an equal illogicality. Vermes supported de Vaux’s ‘library’ concept asking, of the caves, ‘why should anyone choose a spot so difficult of access, unless for the purpose of keeping undesirable visitors at a distance from highly valued documents’ something that might well be considered desirable for a geniza (Vermes 1956: 11; and see Milik 1959: 20). The discovery by Bedouin of ancient scrolls in caves near the Dead Sea in 1947 led to scholastic and popular excitement. The contents of some of the more complete scrolls were rapidly made public and it was suggested they had a connection with the Essenes, a Jewish sect whose name does not appear in any of the scrolls but which is mentioned by three classical authors, Philo, Josephus and Pliny the Elder (Sukenik 1948, 1950). In 1950 and 1952 the French scholar Dupont-Sommer published books in which he, too, identified the scrolls as the product of Essenes, who, he suggested, had lived at the, as yet, unexcavated ruins of Khirbet Qumran near to the caves (Dupont-Sommer 1952, 1954). Father Roland de Vaux, director of the École Biblique in Jerusalem, was initially sceptical about any connection between the scrolls and Qumran because, from a surface survey, he considered it to be a Roman fort (de Vaux 1973: vii). The cave in which the first scrolls had been discovered, Cave 1, was identified in 1949, and investigated by de Vaux and G. Lankester Harding, a British archaeologist then in charge of the Jordanian Department of Antiquities. Jars in which it was assumed some of the scrolls had been placed were recovered together with a few other vessels. With the exception of some sherds identified as ‘Roman’, it was claimed that 95% of ‘the pottery is pre-Roman, and most probably of the second century BC’ (Harding 1949: 113), a dating agreed to by de Vaux (de Vaux 1949: 236). Thus the scrolls had also to date, on archaeological grounds, to the last two centuries BCE and could not relate to a supposed Roman fort. The leading authority on Jewish scripts of the time, Solomon Birnbaum, had independently and firmly palaeographically dated all of the Cave 1 scrolls to not later than mid-1st century BCE, in agreement with the dating of the archaeologists (Doudna 2006: 147-8 and see there more references to articles by various scholars published in the 1950s). Following the exploratory excavations of 1951, de Vaux radically revised his opinions, accepting the association of the site with a sectarian community who, he believed, deposited the scrolls in caves in the 1st century CE. It was also decided that ‘the place was, so to speak, a “closed settlement” having, perhaps, but little contact with the outside world’ (Harding 1952: 105) although there is no explanation as to how such a sweeping conclusion could have been drawn from the sparse archaeological evidence gleaned from a three week season, in which only seven loci were opened. It was more likely an inference drawn from one of the scrolls, the ‘Damascus Document’. Milik, a member of the excavation team, explained that ‘according to the Teacher a communal way of life in seclusion provided the conditions necessary to enable man to overcome his weakness in the face of the Evil One’ (Milik 1959: 78). The modern geo-political situation at the time of the excavations meant that Qumran was exaggeratedly isolated, being close to a border between hostile modern states. Although the actual border between Jordan and Israel was, until 1967, some kilometres to the south (to the north of Ein Gedi) the ruggedness of the terrain south of Ein Feshka meant that Qumran was very much a border post. When I hitchhiked to Qumran in 1964 the only vehicles in the vicinity of Qumran were military, the few people one saw were soldiers, and there were signs warning of imminent mine fields further south. Similar warnings could be seen, I later noticed, on the other side of the border, immediately north of Nahal David in Ein Gedi. This apparent isolation at the time of the excavations added artificial weight to the concept of a secluded community that had been gleaned from both the sectarian scrolls and the classical authors. However in a brief season of excavations in 1951 (24/11 – 12/12) at Khirbet Qumran, again directed by de Vaux and Harding, the discovery of a jar of the same type as had been recovered from Cave 1 indicated not only a connection between the ruins and the cave but also forced them to reconsider their dating of the pottery. The jar was sunk into the floor of L2 and beside it was found a coin of the Roman Procurators under Augustus, dated to 10 C.E. Harding admitted, at that time, ‘how little is really known about the accurate dating of pottery of the first centuries BC-AD’ (Harding 1952: 105). Most parallels had to be found from poorly stratified excavations, or from random tombs, in Jerusalem or further afield, even in mainland Greece. De Vaux, on his own from now on and without any moderating interpretations that might have come from Harding, started more extensive excavations which continued for four seasons between 1953 and 1956. Although many textual scholars pored over the contents of the available scrolls, and widely disseminated their differing conclusions, de Vaux, who as well as being the Sukenik, after the first scrolls were discovered in Cave 1, had suggested that they were from a geniza (a repository of damaged or outmoded Jewish documents that might contain the name of God and can not, according to Jewish tradition, be destroyed but must be disposed of in a 7 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS simply thrown outside the walls’ (de Vaux 1973: 25). The member of de Vaux’s team who had first noticed ‘Trench A’ was even more emphatic about the thoroughness of the cleaning operation; ‘most traces of the earlier occupation were obliterated’ (Milik 1959: 51) for ‘during the work of reconstruction the collapsed buildings were thoroughly cleaned out and swept, and the debris was thrown away in a small ravine to the north of the settlement’ (Milik 1959: 53). director of the École Biblique was also president of the trustees of the Palestine Archaeological Museum through whom the finances for the excavations came, soon established a virtual monopoly on the interpretation of the archaeological evidence, something that at least one scholar regretted at the time. ‘What is distressing... is the fact that in a matter of so much importance and of such tenuous evidence, virtually no specialist’s estimate was ever made altogether independently of de Vaux’s influence’ (North 1954: 434). It was generally maintained by the epigraphers, historians and theologians, including, after 1951, de Vaux himself, that the scrolls which, before the first season of excavations, had been dated to the second and early first centuries BCE, constituted a library belonging to people living at Qumran, who may have been Essenes, a Jewish sect who had left the iniquities of Jerusalem for a life of isolation in the desert (i.e. Qumran) where they practised a communal life, with communal meals, prayers and meetings (Dupont-Sommer 1952, 1954; Vermes 1956; Allegro 1956; Milik 1959; Cross 1958). Although de Vaux ‘sought to avoid drawing any premature conclusions’ (de Vaux 1973: viii), it would be surprising if he was not influenced by the weight of so much scholarly opinion. In the archaeological seasons from 1953 onwards he anticipated uncovering evidence of rooms used for communal gatherings, a concept with which he, as a member of a religious community - the Dominicans - was sympathetic and familiar. By 1968 he waxed lyrical ‘If you go to the Dead Sea, stand looking at the plateau, and then study the excavation structure by structure, you will know that this is the very site of the Essenes described by Pliny nineteen hundred years ago’ which, as Golb points out, is ‘more a declaration of faith than a statement of science’ (Golb 1995: 64). When a plastered pool L48/9 was found in the ‘main’ building, built partially over a pair of earlier pottery kilns, a large crack running through it was assumed to be the result of the earthquake of 31 BCE, mentioned by Josephus. Pottery found within it, and near by, was therefore assigned to a pre-31 BCE date (de Vaux 1954: Fig. 2). In the third excavation season in the 1954 (13/12 – 14/4) season de Vaux continued to work in the ‘main’ building and uncovered a large number of broken vessels in a room L89 (Humbert and Chambon 1994: pls 338-9) which, he believed, were smashed by the earthquake. Despite having previously concluded that meticulous post- earthquake housecleaning had taken place in loci immediately to the north, these sherds were not cleared away because, apparently, the room was ‘too much encumbered’ (de Vaux 1973: 25). These sherds were not, however, heavy rocks, and although they were the remnants of over a thousand vessels, most were from small bowls and could easily have been swept up and removed with, if we are to believe the excavators, all the rest of the debris from all other parts of the building. Nonetheless these vessels were identified as being from a pantry damaged in the earthquake, and the large room adjacent to it was identified as the ‘refectory’. In 1953 (9/2 - 24/4), further excavations conducted in what became known as the ‘main building,’ revealed evidence of occupation that seemed earlier than that which had previously, in the 1951 season, been dated on numismatic grounds only to the first century CE. The pottery corpus, as noted above, was not well enough known to provide secure dating assistance, rather the reverse, with the pottery being dated following the results of the excavation. By chance an ancient rubbish dump was stumbled upon, outside the northern wall, containing much broken pottery, which seemed to be somewhat earlier than that being found in the building itself. Moreover the coins recovered from this dump were, with one exception, from the second and first centuries BCE. Thus de Vaux had now to accept that what, following his first season’s work, he had considered solely a first century CE site must have had earlier origins in what he now called ‘Period I’, later to be still further sub-divided into Periods Ia and Ib. The excavators jumped to the unlikely conclusion that, following an earthquake in 31 BCE, ‘most of the rooms were cleared out, a circumstance which has deprived us of many pieces of evidence, which would have been valuable for Period Ib. Some of the debris was carried out to the north of the ruins and left on the slopes of the ravine where it has been brought back to light by one of our excavation trenches, Trench A... Other parts of the debris were In the fourth season of excavation in 1955 (2/2 – 6/4), attention turned to the ‘western quarter’ near the round cistern L110 and the water system which supplied it. Because Iron Age pottery and remnants of Iron Age walls had been found beneath the ‘main building’ it was logical to identify the round cistern as originally Iron Age, as such cisterns were known from Iron Age sites elsewhere, whereas second temple period sites have rectangular pools. De Vaux recognised that in the first post-Iron Age occupation (Period Ia, de Vaux 1973: Pl IV) a channel was built which drained water from an open area to the north and ran beneath a floor (L115/6) to feed the round cistern, and that two other rectangular cisterns, L117 and L118 were dug nearby together with ‘a decantation basin’ (L119) ‘where the silt was deposited, common to the three cisterns’ (de Vaux 1973: 4, Pl IV). A number of rooms were built surrounding them which, together with the pottery kilns1 beneath cistern L66 and a few poorly 1 Although de Vaux, during the course of excavation, justifiably queried whether ‘since both were earlier than the establishment of cistern 49 and of its associated area, might they be Israelite?’ (L66 1/4/54), he later wrote firmly that ‘there is nothing to indicate that these kilns were already in service during the Israelite period’ (de Vaux 1973: 4-5). Recent excavators have concluded that Iron Age Qumran was probably a refugee village, possibly only seasonally inhabited, with 8 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN EARLY CONCLUSIONS REACHED DURING THE EXCAVATIONS IN THE 1950S but were not registered, perhaps because they were ‘not restorable’ (L101 5/2/55). They would appear to be the same type of jar, identified as belonging to Period Ia, as had been found the previous season beneath the earthquake debris in the tower room L10A (de Vaux 1954: Fig. 1:2, 1973: 20 and Pls 19, 20). A ‘collapse ... due to the earthquake’ is mentioned in the room adjacent to L101 (L102 17/3/55) but it is very unclear what is meant, particularly as, later, a pit is reluctantly attributed to the earthquake (L102 21/3/55). In L111 it is pondered whether some alterations could be attributed to the earthquake but no debris is specified (L111 10/2/55) although a new floor was laid down above a deposit which was at least 40 cm deep, and thus suggestive of a destruction layer (L111 19/3/55), in marked contrast to the situation in the ‘main building’. Could it be that de Vaux had difficulty in accounting for the fact that debris was not tidied away here in the western quarter as it had been, supposedly, in the ‘main’ building? He certainly seemed to suggest that some debris from this western building was cleared and dumped in L130 (L130 2/3/55). Were these rooms of the western building with their layers of debris also ‘too much encumbered’, in keeping with de Vaux’s explanation for why L89 also was not cleared?4 determined walls to east and south, represented the earliest occupation, his ‘Period Ia’, which he dated with, as he admits, flimsy evidence, to the time of John Hyrcanus, 135-104 BCE (de Vaux 1973: 5). The structures in the ‘western quarter’ were less well built than the ‘impressive complex of buildings’ (ibid: 5) which the archaeologists had spent their first three seasons excavating. Indeed they were disappointingly unimpressive and did not, perhaps, live up to the by now firmly established notion that Qumran was an orderly home to a scholarly, ascetic ‘community’ (ibid: 10) where white robed sectarians dedicated much of their time to the writing of a library of scrolls that were stored in nearby caves. The buildings in the western quarter were generally dated to the second and first centuries BCE, thus predating the ‘main’ building whose predominantly first century CE occupation was now seen as contemporary and associated with the production of the scrolls. Whether or not this influenced de Vaux’s interpretation, the humble western buildings were dismissed as ‘less important’ and ‘probably... served as covered workshops’ (de Vaux 1973: 8), and, as such, were given little more thought. The ‘modest nature of the buildings’ in the western quarter was enough to condemn them as being ‘of short duration’ (ibid: 5). On the plans they were relegated to ‘outbuildings of the main structure’ (Humbert and Chambon 2003: Plan III). Regrettably none of the pottery from these deposits in the western building, whether resulting from an earthquake or not, has been published apart from that found in L114 (see below). One suspects that Hasmonean pottery was retrieved from the lowest levels within all the rooms of the workshop area, in marked contrast to the ‘main’ building where none was found. Declaring these rooms to be of little or no importance perhaps helped de Vaux overcome the discrepancy. One hopes that, having had over fifty years to prepare it, the pottery will soon be published by the École Biblique. However, extensive debris, including fallen mud-bricks, which, from the photographs and the field notes, seem likely to have been dislodged from an upper storey by an earthquake, was found on the floors of some of the rooms of the ‘western building’, in particular the first room that was excavated, L101. In places the debris rested on a layer of ashes (pls 281, 284-6;2 L101 4-5/2/55). Although de Vaux noted that ‘the north-west corner of the secondary building was ... damaged’ by the earthquake (de Vaux 1973: 20) and that it was ‘strengthened by a buttress’ (ibid: 25) (outside L123, Humbert and Chambon 2003: Plan IV), he does not mention the debris in L101, although its ‘collapsed bricks’ (L101 5/2/55)3 are reminiscent of the ‘collapsed bricks’ (L10A, complementary notes) acknowledged as earthquake debris in the tower. Moreover two storage jars, associated with the debris, whose rims and shoulders were preserved, are clearly visible in L101 in the photographs According to de Vaux, in Period Ib, which he dated between 100-31 BCE, the buildings were enlarged. ‘They consisted of a main building with a massive tower, a central courtyard, rooms for communal use, a large hall (which also served as a refectory)...’ (de Vaux 1993: 1236). The water system was replaced by one at a higher elevation and many new cisterns were dug. De Vaux realised that pool L48/9 had been fed by this new ‘main aqueduct’. As he had already identified the pool as having been damaged by the earthquake of 31 BCE the water system, of necessity, had, for him, to predate the earthquake. This led to some anomalous interpretations some of which still receive credence to this day. The dating of the ‘main aqueduct’ is a complex archaeological issue which I addressed in a paper published in Dead Sea Discoveries (Stacey 2007) in which I concluded that it could not have been built before the earthquake of 31 mud-brick huts built on stone foundations (Magen and Peleg 2007: 2228). De Vaux identified no Iron Age floors and it appears that most were eroded away. No definitely Iron Age walls survived more than a course or two whereas, in contrast, the kilns, by their nature, impermanent, stood c. 1 m above their operational surface. Although it is possible that Iron Age kilns had existed those that survived did so to a higher elevation than the top step of L48 (Humbert and Chambon 1994: pl 174). If these kilns were not operating in the Hasmonean period their remains would have been substantial obstacles and it seems inconceivable that they would not have been removed by the simple expedient of collapsing them into their own fire-pits; (how intrusive the kilns would be, if not operational, can be seen at http://www1.ku-eichstaett.de/KTF/qumran/bilder/abfrage/loc66.jpg ). 2 Where ‘pls’ are not otherwise designated they refer to the photographic plates in Humbert and Chambon 1994. 3 All references in this format refer to de Vaux’s field notes in Humbert and Chambon 2003. 4 For a more comprehensive account of the early excavations and the alacrity with which the Qumran-Essene hypothesis became a ‘certainty’ see the first chapter of Golb 1995. 9 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS BCE. Here I will attempt to simplify some of my arguments, add some important archaeological evidence that was not available at that time, and add some observations made in the field since, one of which has been overlooked by all previous scholars. 10 2. When was the ‘main’ aqueduct built? walls of the round cistern were heightened’ (de Vaux 1973: 9). He also realised, although he does not specifically mention it, that the sides of the two pools L117, L118 must also have been raised, because the water inlet was now at a higher elevation, and this necessitated the addition of three or four steps. In the plan of Period Ia (de Vaux 1973: Pl IV, Humbert and Chambon 1994: Plan III) both pools are shown with fewer steps than in Period Ib (cf Humbert and Chambon 1994: Plan IV). Moreover the original water inlet for L117 (cf ibid: Plan III and Plan IV), which received water both from the channel beneath L115/6 to its north via the sedimentation pool L119, and also water draining from the roofs above L117 and/or L101, which entered a channel beneath the wall to its south (de Vaux 1973: 4), can be seen to this day, in its west wall, considerably lower than the inlet connected to the ‘main’ aqueduct (Magen and Peleg 2007: Fig 33).6 De Vaux did not publish a final report before his death in 1971 so we have to rely on the text of a series of lectures summarising his conclusions (de Vaux 1973). It was not until over thirty years after his death that a compilation of photographs and field notes were published, supplying some of the evidence with which one can check his conclusions (Humbert and Chambon 1994; 2003). It can be discerned from the photographs (Humbert and Chambon 1994 pls 110, 113, 262, 264) that the field technique employed was fairly primitive, with little attempt to isolate layers and little use of sections as temporary records of layers5 and how they related to walls, and the recording was even worse. Some comparative floor levels in the ‘main building’, recording their depth below the top of a stone in a wall, appear on a couple of plans (de Vaux 1954: Plans V and VI), but no absolute elevations are published and may never have been established. In the notes no floor levels, nor elevations on recovered objects, appear, and this makes it difficult to reconstruct the dating and stratigraphy of the site. The following day it was decided that ‘the forms fit period Ib, but are more varied than that of the large deposit of 1954 in locus 89’ (L114 24/3/55). Although the ceramics were originally published as belonging to period II, or post earthquake (de Vaux 1956: Fig. 4), this was revised to period Ib (de Vaux 1973: 5 fn 1). It was conjectured that the vessels had been smashed by the earthquake but, because they were found considerably below the level of the top of the aqueduct, this interpretation would mean that the aqueduct would have had to have been built free-standing on the plaster floor, which belonged to Period Ia. It would also mean that the raised western wall of the round cistern must also have been free-standing, which is extremely unlikely as the outer face was not built with a regular external finish (Humbert and Chambon 1994: pl 226). Clearly the area between it and the wall to its west – the east wall of L111 - was backfilled to help withstand the weight of the water within the cistern. Both Magness (Magness 2007) and Mizzi (Mizzi 2009: 52 fn 133) dispute the precise location of the pottery but neither seriously addresses the fact that the deposit was below the level of the top of both the ‘main’ aqueduct and of the raised side of L110 which could not have been free-standing. Thus, no matter the exact find spot, the pottery pre-dates the raising of the cistern and the building of the aqueduct that fed it. As well as the paucity of knowledge concerning the ceramics of the period, already mentioned, de Vaux had few parallels with which to compare the way aqueducts were built and the manner they were integrated into an occupied area. Although some excavations in Jericho were published during the course of de Vaux’s work at Qumran (Kelso and Baramki 1955), and some more soon afterwards (Pritchard 1958), the soundings were limited and were confined to Herodian layers not closely connected to the extensive water systems of Jericho. Since then large scale excavations in Hasmonean and Herodian Jericho, and at other relevant sites near the Dead Sea such as Cypros, Masada, Callirhoe and Herodium, have taken place and their programme of final publication is well advanced so that we now have comprehensive knowledge of Second Temple period aqueducts, and far more reliable dating for the pottery. Jericho is a particularly important site because it was built on the same marl layer as Qumran, a layer of silt many metres thick that had accumulated on the bed of the prehistoric Lake Lisan, which existed in the rift valley until about 11,000 years BP. At Qumran, when excavating in a small area, L114, next to, but well below the upper elevation of, the new aqueduct some pottery vessels were discovered (Humbert and Chambon 1994: pls 222, 223). They were found close to a plaster floor laid almost on virgin soil and ‘we must decide if the deposit is contemporary with the first circular cistern or with our major period Ib’ (L114 23/3/55). De Vaux had recognised that because the new aqueduct ‘was at a higher level than that of Period Ia the Moreover the aqueduct arrived into L114 after passing through the door leading from L115 where it was built over the original sub-floor water channel of Period Ia. It took up the complete width of the door, so gaining access to L115 from the plaster floor in L114 which, according to de Vaux’s interpretation, would still be operational, would mean clambering up on to the aqueduct, which, together with the thickness of its capstones, would be about 75 cm in height, and then dropping down the other 5 A section drawing through L130 has been published (Donceel 2005: Fig 5) and through L86, 87, 89 (Humbert 2006: 33). Neither are mentioned in the field notes so it is possible that other drawings exist from other loci. One is mentioned for Trench A (Trench A, 6 & 9/3/53). 6 The earlier inlet is visible as the dark square hole. The later is higher and to its left, see Galor 2003: Fig 5, which regrettably does not include the earlier inlet. 11 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS have been found in the upper part of the locus and it is possible that pottery from two different periods was found but not recognised as such. ‘In the northwest corner several pottery forms and many sherd(s) appeared’ (L114 22/3/55). The next day ‘under the potsherds from yesterday, we discovered a deposit of pottery:’ (L114 23/3/55). The upper sherds, including the spatulate lamps, may have been a deposit associated with the ‘Main’ Aqueduct, whereas the sherds lower down were from before it was built.9 side. A similar obstacle course would have presented itself to get between L115 and L116.7 There were no signs of steps built against the aqueduct and, indeed, in L114, any supposed steps would have had to have been built in the area where the pottery was located. Clearly the aqueduct was built over, and later than, the pottery that had been smashed, most probably by the earthquake of 31 BCE. The published pottery supports this date with the possible exception of the three spatulate lamps (de Vaux 1956: Fig. 4.14). De Vaux regarded them as rougher and earlier than true ‘Herodian’ lamps (de Vaux 1973: 5 fn 1). Bar-Nathan says that this wheel-made form (her Type J-LP4) ‘had appeared by the Herodian 2 period (15BCE-6CE) alongside mold-made lamps, especially those with a spatulate nozzle, J-LP2C’ (Bar-Nathan 2002: 111).8 However another wheel-made lamp form (J-LP3) makes its first appearance at this time but has only been found in Jericho, Masada and Qumran (de Vaux 1956: Fig. 1:1-4 and, probably de Vaux 1954: Fig. 2:16). Thus it appears that potters in the Dead Sea region were experimenting with making lamps on a wheel soon after the earthquake. Two types of nozzle were experimented with, and the spatulate nozzle was ultimately found to be more satisfactory because it was more robust. These lamps were unlikely to have been ‘exported’ in any number to Jerusalem but a few may have diffused there over time. Local Jerusalem potters began copying these lamps but they only become de rigueur towards the end of Herod’s reign (Hershkovitz 2009: 275). As well as largely blocking access into L115 and L116 [the wall between these two rooms only abuts the western wall and is a secondary feature built together with the ‘main’ aqueduct (Humbert and Chambon 1994: pl 211)] a free standing aqueduct would have considerably hampered movement around the site, making access into, for example, L101, L102 and L104 and L85, very awkward. In a scale model of the ‘monastery’ (sic) ‘constructed to plans published by Pere Roland de Vaux’ it was necessary to reconstruct a bridge, for which there is no evidence, crossing the supposed free-standing aqueduct in L100 (Allegro 1959: Pl 101). De Vaux had to believe that the new or ‘main’ aqueduct as it ‘wound between the buildings’ (de Vaux 1973: 8) was free standing and built onto the Period Ia floor surfaces, as it is shown on all his Period Ib plans. Because it had become an article of faith that the pool L48/9 had been destroyed by the earthquake it was axiomatic that the water system that fed it also existed before 31 BCE. But for this to be true he had to ignore some of his own observations. He had noted that in L100 ‘the lower floor continues under the channel’ (i.e. the channel was not built on the floor). ‘The same floor reappears in the western part of L106, where it is clearly cut by the main channel’ (L100 20/3/55). This surface relates to the original plastering of the recess in the south west of the locus (pl 293). A secondary ‘whitish floor ... is associated with the upper plastered floor of the recesses’ (16/3/55 pls 295, 297). It is clear that both these floors were cut by the insertion of the aqueduct whose raggedy, external side can be seen (pls 293, 294) unplastered, and unplasterable, in marked contrast to the well-finished and plastered walls of the locus. The rough stone slab ‘upper’ floor, into which were integrated the upper surfaces of two column bases,, related to a floor of slabs, of which one survived, covering the aqueduct (contra Magness 2007). However it is also possible that the spatulate lamps came from a deposit later than the construction of the aqueduct. From their registration numbers these lamps appear to 7 Mizzi comments that ‘there was no need to cross over from west to east in the area of L115, L116 and L114’ (Mizzi 2009: 52 fn 133). The problematic direction of movement would have been from south to north. Even though the floor covering the aqueduct could theoretically be reached via L108, if the aqueduct was partially freestanding, as he proposes, access into L115 and L116 would still have been extremely awkward even with the unlikely insertion of wooden steps. He does not address the fact that the western face of the aqueduct running through L115 and L116 was neither regularly finished, nor plastered (pl 216) in stark contrast to the walls (pls 211, 215) and floor. Mizzi claims that the key issue is ‘the fissure which starts in L.111 and goes diagonally (in a southwest-northeast direction) through L.115, the water channel, and L.118’ (pls 211, 212, 215) which he believes shows that the aqueduct could only have existed at the same time as the plaster floor in L115. However the crack is shown on the plan continuing over L118 although there is no sign of a crack in the steps of that cistern (pls 220, 239). Clearly the fissure ran above the fill and occurred after L118 had gone out of use. The aqueduct could also no longer have been operational (there is no sign of a repair and the force of any flood water would have rapidly undermined the channel at the point of the fissure (in particular if it was, as Mizzi, believes, partially freestanding). As the aqueduct continued to feed water into L71 at least into Period III the fissure, which we can assume was a minor techtonic movement effecting an area vertically as well as horizontally , occurred long after the site had been abandoned. Mizzi claims that because de Vaux makes no mention of an upper floor there was therefore no upper floor flush with the water channel. There indubitably was, however, an upper floor from which the ‘silos’ in L115 and L116, which de Vaux recognised as being ‘late’, were dug. This surface had probably been eroded by flood water that was naturally channelled into this area long after its abandonment. 8 The numismatic evidence is not helpful: two coins possibly of Jannaeus and one of the Procurator under Tiberius. In L109 ‘over the entire locus we uncovered a lower floor, earlier than the main channel and undoubtedly contemporary with the first circular cistern’ (L109 9 Although Bar-Nathan has accepted a late date (1-60 CE) for the pottery from L114, apparently influenced by Humbert (Bar-Nathan 2006a: 8) she seems to have overlooked the two ovoid jars (KhQ 2656 & 2657, see Gunneweg and Balla 2003: Plan 4) that, in Jericho, were ‘common in the first century B.C.E.’ having appeared ‘during the Hasmonean period and continued into Herodian times’ (Bar-Nathan 2002: 27). Only one fragment of an ovoid jar was found, in a Zealot context, at Masada and it almost certainly originated in the Herodian storerooms. 12 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - WHEN WAS THE ‘MAIN’ AQUEDUCT BUILT? from the large cisterns quarried into the western face of the mountain on a path that traversed the northern end below the ‘hanging palace’. The channel, which pre-dated the building of the inner casemate wall in the south east, ran over open ground that, before the addition of the casemate wall, would have had very little traffic. Nonetheless the channel was sunk below the ground surface (Netzer 1991: 557-8, Ills 873-5). 23/3/55). This floor continued westwards into L113 where it ‘connects by a narrow opening with the channel feeding cistern L110’ (22/3/55). It was soon realised that this was was not a feeding channel but the beginning of a drain outlet whose continuation was in L103 (24/3/55). There was, however, no ‘trace of the drain’ below the plaster floor (L 109 & L113 24/3/55). This probably indicates that the drain dated to the Iron Age, (as shown in de Vaux 1973: Pl III) 10, and was filled in by the earliest Hasmonean builders who laid down the plastered floor which related to the well plastered walls visible, at least, in L113 (pls 227, 231) and must be the one on which a worker is standing (pls 232-3). From its elevation this would appear to be contemporary with the earliest floor in L101 by which it was connected by a door, and with the ‘first circular cistern’ before its sides were raised and could not have existed together with the aqueduct, whose thin unplastered walls could not have been freestanding. Thus the ‘main’ channel did not relate to the earliest (‘Period Ia’) floor. Nor does it to the secondary floor, (contra Magness 2007) who rather fancifully reconstructs a staircase11 going over the channel although no such staircase appears in the written or cartegraphical record. In reality such a staircase would lead directly into the western wall of L117 and not give useful access to anywhere else. The photographic record is poor and, as no elevations appear on the plans, it is impossible to know, for example, what was related to the oven mentioned in the north-east corner of L109. The same techniques would have been employed at Qumran. If the aqueduct running through the site had been free-standing its side walls would have been solidly built and their outer faces regularly finished and plastered. That this was not so can clearly be seen in the field to this day and in all the photographs taken at, or close to, the time of the excavation, where it can be seen that the side walls were of irregular thickness and had a haphazard, un-plastered outer face (Humbert and Chambon 1994: pls 215, 216, 299; Allegro 1996: D3:8, D5:8, D8:6, E1:6. http://www1.ku-eichstaett.de/KTF/ qumran/bilder/abfrage/loc100.jpg). As Magness so rightly says, ‘for archaeological interpretations to be valid, they must be based on a thorough consideration of all available evidence and on parallels with contemporary sites in the same geographic region’ (Magness 2002: 99). Thus, unless evidence is ever found of a substantial freestanding aqueduct with irregular, un-plastered outer wallfaces, at any of the ‘contemporary sites in the same geographic region’, there can be little doubt that the channel at Qumran ran below the floors to which it related, which are the various floors designated as period II, or post earthquake by de Vaux. Magness disregarded her own dictum when she wishfully wrote ‘the walls of the main aqueduct apparently did rise above the floors of some of the rooms at Qumran. This seems to be true mainly in the northwest sector’ (Magness 2007: 255, my italics), without producing any parallels. In Jericho many contemporary aqueduct systems have been thoroughly excavated over a ten year period, many of them under my supervision. The daily, hands-on experience of exposing such aqueducts over a wide area gave me a familiarity with the way their builders worked, a practical familiarity which both Magness and Mizzi lack. On open ground, where the aqueducts did not hinder free movement, they were built as free-standing channels between two solidly built and well- plastered side walls. All three of the exposed faces of the side walls would be well plastered (inter alia Netzer 2001: Ills 110, 111, 2004: Ills 23-25). Within occupied areas, however, channels were built conveniently beneath the floors, with the side walls being only retaining walls. Only the upper, above ground surfaces of these walls, together with the channel itself, were plastered (Netzer 2004: Ills 56-62). Although in a rather muddled note de Vaux suggested that a plaster floor in L115 was ‘contemporary with the upper channel’ (i.e. the ‘main’ aqueduct), he offered no specific justifications for this claim. If the function of the room was such that it warranted a plastered floor surface one would expect the walls to be plastered too. That at least the western and southern walls of L115 and L116 were originally lime plastered can be seen in photographs (pls 211, 212) so there is no doubt that the plastering would have continued up the supposed exposed external face of the channel. That it did not proves that the wall of the channel was buried below ground level and was not, despite de Vaux’s claim, contemporary with the plastered floor. At Masada the same can be seen. A channel was built starting near the snake-path gate which led water into a large cistern 1907. The water would have been carried up 10 The remnants of the drain are the only surviving Iron Age remains in the vicinity of cistern L110. The cistern, like many in the Negev, was probably unenclosed and had water directed into it by a shallow ditch. As flood water would have continued to flow after the Iron Age site was abandoned the surfaces surrounding the cistern, particularly those to the north, would have been eroded, together with, quite possibly, some of the upper course of stone lining. Although Magen and Peleg contend that the inhabitants of Iron Age Qumran were not ‘capable of digging such a huge pool’ (Magen and Peleg 2007: 28) it would not have been so taxing to dig such a pool into the marl. Its depth could have been extended in the Hasmonean period. 11 This is not the same ‘bridge’ as envisaged by Allegro (see above) further south in L100. De Vaux noted that ‘the north-west corner of the secondary building was ... damaged’ by the earthquake (de Vaux 1973: 20) and that it was ‘strengthened by a buttress’ (ibid: 25) (outside L123, Humbert and Chambon 2003: Plan IV).12 He realised that the ‘main’ aqueduct was originally fed from a basin, L132 to the north of the 12 Although it is possible that the buttressing occurred earlier, or even later, than 31 BCE the earthquake was its most likely cause. 13 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS to the east it grows progressively thinner’ (de Vaux 1973: 23). As he believed that this sediment was waterborne de Vaux rightly asked himself ‘why is the sediment not horizontal?’ (L130 3/3/55). This, justifiably, continued to worry him (see L130 7/3/55) but he managed to dismiss his qualms over time and convinced himself of his early hypothesis that ‘an earthquake broke down the water system’. buttress. The south and west walls of this basin were built to the buttress which means, if de Vaux’s dating of the buttress is correct, that the basin and the aqueduct it fed must post-date the earthquake. It would appear that he realised this discrepancy in his interpretation and removed a part of the buttress in the hope that he would find evidence that the basin existed before the buttress, i.e. that the original basin walls and its plaster floor would have been buried within the buttressing, and that they would run up to the wall north of L123/120. ‘We removed a part of the reinforcement of the northwest corner of the building in order to follow the plaster of the basin. The plaster disappears very quickly and the wall of the basin stops. We reached the original wall of the building’ (pl 260, L132 27/3/55). On the plan of Period Ib a tentative continuation of the basin wall up to the original wall of the building is shown, but there was no evidence whatsoever for this and, to this day, it can be seen that both walls of the basin were built abutting the sloping face of the buttress (pl 263; Donceel 2005: Fig. 9). De Vaux described the sediment in L130 and L132 as being of ‘yellow earth’ (L130 27/2/55), and presumably it had a silt like appearance to justify it being considered a water deposit. Rather than derive from a recent flood deposit, it is far more likely that it came from the Lisan marl, the natural ‘virgin soil’ at Qumran, and itself an ancient silt deposit, that had been excavated either from one of the deep pools, L58, that was dug in association with the main aqueduct, or from a cave/quarry, such as Caves 7-9 or other as yet unnoticed ‘caves’ on the plateau, which may have been dug at this time when mud-bricks were needed for the post-earthquake renovations and new buildings. In Jericho the marl, when exposed, was usually of a yellowish-green hue (elsewhere it could be reddish-brown). When pools were dug the spoil was generally spread around them to level up the surrounding floors and to reduce the depth that had to be actually dug out. This was noticeable beside the Hasmonean swimming pool complex where in places the fill was 1.70 m deep (Netzer 2001: 77); and south of the swimming pool in Herod’s Second Palace where the bath-house wing was largely constructed on back-fill, held behind a retaining wall and creating a level platform above an existing slope; ‘the fills were deliberate and fairly uniform of the type of local virgin soil’ (ibid: 198, 206). Thus this ‘sediment’ in L130, and all the material culture associated with it, was, in all probablility, contemporary with the construction of the aqueduct and not later than it. A corpus of pottery from this locus was published, which, together with that from L114, can suggest a date for the aqueduct. Both were assigned to Period Ib (de Vaux 1956: Figs 1, 3 and 4). Drawing on the far larger corpus from Jericho, Bar-Nathan suggests that the material from L130 must be dated to 31-15 BCE (Bar-Nathan 2002: 111, 204),14 so a date for the construction of the main aqueduct of soon after 31 BCE is most likely. In the vicinity of the buildings the main aqueduct was dug into an accumulation that, in nearby loci, appears to be earthquake debris. It approached the buildings from the north at an elevation some 75 cm above the opening of the Period Ia channel that ran beneath the floor of L115 (the opening can be seen in pl 269, the ‘main’ aqueduct wall at the extreme left of the picture). It is not clear how locally falling rain water was steered into the early channel but, for the ‘main’ aqueduct, water falling on the slopes north-west of the site was initially collected into a wide, shallow decantation basin, L132, which was built to the buttressing at the north-west corner of the western building. Its southern limit was the wall between itself and L130, which was a retaining wall and could not have been built as a free standing wall (pls 261-263) as it was ‘not so thick and seemingly without a foundation’ (L135 6/3/55). The surface associated with the aqueduct would have been one covering the soil, including a number of animal bone deposits, deposited in L130 as a back-fill, some of which may, according to De Vaux, have been ‘due to a clearing of the building after the earthquake’ (L130 2/3/55). The wall between L130 and l132 created a necessary barrier between the water collected in the basin and the north wall of the western building, thus preventing any damaging effects the water would otherwise have had.13 Despite the fact that the wall between L130 and L132 was built to the post-earthquake buttressing at its western end, de Vaux illogically believed that L132 was functioning as a decantation basin before the earthquake. He postulated that the aqueduct was damaged by the earthquake and that ‘sediment from this accumulated in the large decantation basin 132, overflowed it, and spread into loc. 130 as far as the wall of the building’ despite the fact that the division wall ‘was seemingly without foundation’ and thus unlikely to have been freestanding. In the north-west the accumulation ‘reaches a thickness of 75 cm. As it extends Further compelling evidence that the ‘main’ aqueduct post-dated the earthquake has been produced in recent years, during excavations conducted by Magen and Peleg. They excavated along the line of an outlet channel, already noted by de Vaux, that took water from the pool 14 The numismatic evidence is, as too often at Qumran, inconclusive. Although seven coins were identified as perhaps of Alexander Jannaeus, a coin of the Procurator under Tiberius clouds the issue. From its registration number, however, it came from early in the excavation and probably represents a coin from an upper level. The failure to record, or at least publish, the elevations of find spots causes confusions. In L120, for example, three hoards of coins were found ‘buried beneath the level of period II and above that of Period Ib’ (de Vaux 1973: 34) yet in the field notes they are said to come from the ‘lower level’ a term that probably refers only to the lower level on the day the field notes were written. 13 The raised surface in L130 may have been utilised as a garden, conveniently watered from L132. 14 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - WHEN WAS THE ‘MAIN’ AQUEDUCT BUILT? L117. As discussed above, this pool was dug in ‘Period Ia’ but had to have its sides raised when the ‘main’ aqueduct was built. The outlet is integrated into the three added steps so there is no doubt that the outlet must date to the same time as the main aqueduct (or possibly even later).15 Beneath the outlet channel was the ‘northern rubbish dump’ parts of which had been excavated by de Vaux as ‘Trench A’ (Magen and Peleg 2007: 7, 8 fn 8). In a photograph (ibid: Fig. 11), the layers of the rubbish dump can be seen running beneath the channel, (the large disturbance visible in the baulk to the left of the picture is probably a back-fill of part of ‘Trench A’), and there is a distinct possibility that the uppermost layers of the ‘dump’ were deposited specifically to support the drain. De Vaux regarded the contents of ‘Trench A’ as being from the clearing of the buildings following the earthquake i.e. the pottery was dated to pre 31 BCE. In their preliminary report, Magen and Peleg claim only that it post-dates that from a southern dump which, in turn, dates from the ‘first half of the first century BCE’ (ibid: 7) i.e., presumably, they date it to the second half of the first century BCE. Bar-Nathan, moreover, suggests that the material from Trench A should also be dated into the Herodian period (Bar-Nathan 2002: 107)16 which again indicates that the main aqueduct, and the pools and structures associated with it, which includes the final manifestation of the ‘main’ building, did not exist before the earthquake of 31 BCE. Buqe’ah plateau can send occasional floods thundering down Nahal Qumran. Neither source is reliable, or predictable, and, in some years, no water would have arrived at the site at all. However in an average year this run-off channel should have been able to gather what might be considered sufficient water for Qumran. The water supply for Hasmonean Cypros, the most modest of the Hasmonean forts in terms of its size, was obtained solely by collecting the rain water that fell on a small hill neighbouring the site. The water ran into four cisterns, with a total capacity estimated at 2000 cu m (Garbrecht and Peleg 1994: 168). As the total capacity of all the pools at Qumran is only about 1100 cu m (and they were not necessarily all active at the same time), it is clear that in an average year they could have been filled from rain falling locally on the slopes to the north-west of the site and collected in the broad channel.17 Nonetheless, a far more technically challenging aqueduct was constructed from the western extremity of the broad channel, which went around a steep bluff, in places having to run through tunnels quarried in the bedrock, and into the north side of a basin in Nahal Qumran (Ilan and Amit 2002: Fig. 1, nos. 3, 6-16). The channel starts, cut into the bedrock, (see arrow, Figs. 1, 2)18 close to the foot of a small gully (running down behind figure seated on outcrop above channel commencement, Fig. 1) which might flow with water when rain falls in the immediate hills (Ilan and Amit 2002: Fig. 1, no. 3; Humbert and Chambon 1994: Plans XXIX, XXX). Observation in the field, however, reveals few signs of water activity in this gully. It seems unlikely that so much work was carried out to construct a technically demanding aqueduct merely to gather what would have been an insubstantial extra quantity of water. De Vaux realised that ‘there had to be a basin at the point where the aqueduct started so as to regulate the flow of water’ (de Vaux 1973: 9) but, because there were no traces of masonry, he reckoned it had to be a natural basin. His plans, however, show a ‘barrage’ and Milik expressed no doubt that a dam was built (Milik 1959: 51). Ilan and Amit accept that a ‘substantial dam must have been built in order to overcome the difference between the levels’ (Ilan and Amit 2002: 381-2, fn 6), i.e. that between the wadi bed and the commencement of the aqueduct which is at a higher elevation. 2.1 The aqueduct and the dam north-west of the inhabited buildings. Two different building techniques are noticeable in the aqueduct north-west of the buildings, which almost certainly represent temporal stages in its construction. The earliest is a 1.1 m broad, shallow channel running down from below the cliffs to the site (Ilan and Amit 2002: 385, Fig. 1 no. 16 eastwards; Magen and Peleg 2007: Figs. 38-40). This was a run-off aqueduct that gathered rain water falling on the lower slopes near the site. It continues to do so to this day (Magen and Peleg 2007: Fig. 41). Although it may have been built to feed the pools L110, L117 and L118 (total capacity c. 210 cu m) already in the Hasmonean period, it is more likely to have been constructed during the early years of Herod’s reign. Whereas, during the Hasmonean period, rain falling on nearby roofs was utilised to help fill L117, when the sides of L117 were raised as part of the alterations associated with the construction of the ‘main aqueduct’ the capture of this limited quantity of water was abandoned indicating that a more extensive supply was available from elsewhere. Rain falls at Qumran, and/or in the foothills of the Buqe’ah close to it, on only two or three days most years. Heavy rains falling on the Moreover a rock-cut channel (Fig. 2, to left of figure in foreground, Fig. 5, arrow) can be seen in the bottom of a gully on the south side of the basin, opposite the beginning of the northern aqueduct (Ilan and Amit 2002: Fig. 1 no. 5; Humbert and Chambon 1994: Plans XXIX, XXX), which was probably designed to gather water falling in the hills to the south of the Wadi (Fig. 3). Its 17 Although Cypros is in the foothills it is doubtful if the rainfall is twice as heavy as that on the plains just below. 18 I made two visits to Qumran in 2009 specifically to investigate the area of the dam. A number of colleagues not only drove me to the site but generously shared their expertise in long and valuable discussions. I would like to thank Benny Arubas (October) and, variously, David Amit, Shimon Gibson, Yuval Peleg and Stephen Pfann (November). Special thanks to Shimon for making his photographs taken on that day (Figs. 1-6) available. The interpretations expressed are my own and are not necessarily shared by my colleagues. 15 Because the construction of the outlet channel necessitated breaking through the north wall of the ‘western’ quarter, it is most likely that this was done at the same time as the large expansion of the site contemporary with the introduction of the main aqueduct. 16 The numismatics: 18 coins of Jannaeus or earlier, but one of Archelaus. Although the latter was found at a depth of more than 75 cm below the surface it is anomalous and may have been in an unnoticed pit, or even kicked in by workers from the surrounding ground level. 15 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS northern extremity is at a lower elevation than the beginning of the northern aqueduct, so must have fed water into a pool behind a dam. If enough rain fell the pool would rise to overflow into the aqueduct but, if not, water could have been lifted by bucket. At the southern extremity of the southern gully a channel cut through the bedrock (Fig. 3, arrow, Fig. 4, from west), and debouched directly into the Nahal Qumran. It may have acted as a ‘safety valve’ in time of flood allowing excess water to escape. aqueduct is some 2 or more metres above the bedrock of the basin across which the dam must have been built. Even if a metre of bedrock has been eroded during the past 2000 years, which is unlikely, the dam would still have been substantial and quite unlike any of the other known desert dams from the Second Temple period, which were ‘low and unimposing; their function (being) simply to divert a very small part of the water flowing in the ravine, not to stop the water, since they are quite unable to control the full force of the floods’ (Ilan and Amit 2002: 382, fn 6). Such a dam was not only unique but would have ‘demanded both extensive hydrological knowledge and a large budget’ (Hirschfeld 2004b: 115) and would have required frequent and expensive maintenance. It was certainly not the sort of casual construction implied by Magness, who writes ‘the inhabitants dammed the pool at the foot of this waterfall’ (Magness 2002: 54), as though such dams were common throughout the country. When floods come down from the Buqe’ah the water is concentrated into a narrow, towering chute that falls into a basin behind a substantial bluff of hard rock (Fig. 5, right of picture). The force of the water has gouged out a hole in the bedrock, immediately below its fall, some 4-5 m deep and 4-5 m across. This defuses the force of the water which will then swirl against the hard rock bluff to its east and be sent northwards to break against the northern wall of the basin before eventually flooding eastward towards the location of the dam from where it could be steered into the aqueduct. It seems likely that the increase in water catchment, following the construction of the dam and this section of the aqueduct, was planned together with the construction of the ‘main’ aqueduct across the site. At the eastern end of this upper basin a ridge of hard rock (Figs. 1, 5) has resisted the flood water,19 but adjacent softer rock to the east has been partially worn away so that a low waterfall empties into a substantial lower basin (Fig. 3, looking south; Fig. 6, lower right, looking north). The dam was probably constructed on, and to the east, of the hard ridge as far as the start of the aqueduct channel c. 3 m east of the ridge (Fig. 1, Fig. 5). It must have stood at least three metres high at its central highest point, so the wall at its base would have been two or three metres thick.20 It would have been easier to shape the softer bedrock to secure the foundations and the ridge of hard rock would then have acted as a natural protection to the base of the dam. This protection would have been further increased if the sediment to its west (Fig. 5) was cleaned out, which coincidentally would have increased the capacity of the dam pond. The dam is certainly not Iron Age in origin.21 Cross and Milik noted dams found in Iron Age forts in the Buqe’ah. At Khirbet Abu Tabaq they describe a system of seven transverse dams, ‘the function of the works is to deflect and spread winter torrents, creating a broad, flat “bottom land” which will retain its moisture long after the rains are over’ (Cross and Milik 1956: 8-9). Stager drew a plan of these seven ‘dams’ which were between 0.50m and 1.00m high, and describes one ‘as a baffle breaking the velocity of the floods, but not impounding the water behind it as a solid dam of similar height (over 1.00m) might have done’ (Stager 1976: 155). At some sites, water was diverted into cisterns but not, it must be stressed, after been raised by several metres. No trace of the Qumran dam has survived the torrential floods of the past 2000 years, but its existence seems certain because the elevation of the opening of the Dams had been constructed in Mesopotamia from as early as c. 2000 BCE (Smith 1971: 9-12), but most were designed only to retard and deflect the flow of water. Containment dams had been ‘attempted in arid regions of the Near East as early as the Early Bronze Age, but seldom with much success. Even when structurally successful, containment dams provided a pool that was subject to evaporation and to pollution’ (Oleson 2007: 220). ‘Dams were small structures, and the vast majority were only temporary affairs intended for one season’s use’ (Smith 1971: 23). In Nabataea, during the last three centuries BCE, it has been claimed that ‘dams to create reservoirs were very rare and never large’ (ibid: 24), although irrigation projects, some with small dams, were widespread by the end of the period (Oleson 2007: 225; Kamash 2010: 23-30). The eastern approach to Petra, via 19 This hard rock is probably the ‘circle of rocks’ that de Vaux refers to (de Vaux 1973: 9). 20 There are two long dams associated with the low level aqueduct at Caesarea, both dated to the Byzantine period. One is 3.5 m high and has a thickness of between 2.1 and 2.5 m. the other is 7 m high and is 5.5 m thick at its base and 4.5 m at its crown. (Peleg 2002). 21 Although Ilan and Amit suggest that ‘it is quite possible that as early as the Iron age water was brought from levelled and tilled areas in the valley of Hyrcania’ (Ilan and Amit 2002: 385) this is very tenuous and does not specifically claim that the dam should be dated to the Iron age. Amit himself would not now attempt to support such a date (Amit pers. com.). The dam would appear to have been located approximately 25 m from the back wall of the ‘waterfall’ and the capacity of the ponded water would therefore have been c. 1500 cu m, which is more than the total capacity of all the plastered pools that existed in Qumran. Low, expendable walls may have been built within the basin behind the dam to break the full force of the water, but even if it was somewhat dissipated the dam would have to have been an exceptional construction technologically. 16 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - WHEN WAS THE ‘MAIN’ AQUEDUCT BUILT? the narrow canyon, ‘the Siq’, was always susceptible to danger from flashfloods. Various diversionary and containment dams, at least one of which had a concrete core, were built around the entrance to alleviate the problem. The dates of their construction are uncertain. McKenzie dated them to c. 50 CE (McKenzie 1990: 37) but more recently a date in the third quarter of the first century BCE has been suggested (Bellwald 2004: 81). city which he named Caesarea in honour of Augustus (Lichtenburger 2009: 46; Roller 1998: 133-144) which he had to supply with water. Agrippa, the ‘unsung waterman’ (Dembskey 2009: 156) would have been able to offer Herod technical advice from his own experience for he had been Curator Aqurum in Rome since 36 BCE (and remained so until his death in 12 BCE) and had built two aqueducts in Rome at his own expense (ibid: 98-99, 15659). He had also constructed a safe naval base at Cumae, which he named Portus Julius (Suetonius, Life of Augustus 16.1). Agrippa was thus the ideal person to aid Herod, powerful enough to supply the necessary raw materials and technical expertise (Burrell 2009: 220; Kamash 2010: 64), and with a personal interest in the sort of problems likely to be encountered (Richardson 1996: 263). Herod would no doubt have been aware of any that did exist in his day for his mother was Nabataean and he was a frequent visitor to Petra. However a dam at the entrance to Petra was of considerably more urgency, indeed at times it could be a matter of life and death, than was one in the wadi at Qumran where, as shown above, all the pools could, in a normal year, have been filled by rain falling locally. Even if a concrete cored dam did already exist at the entrance to the Siq in Hasmonean times it is extremely unlikely that it would have inspired the building of one at Qumran unless one could see evidence for a resulting luxurious building with at least a bathhouse which would have justified the effort and expense. Indeed most architectural influence went in the other direction with the ‘Nabataeans model(ling) their paradesios at Petra on the pleasure gardens that graced Herod’s palaces’ some two decades earlier (Bedal 2001: 37, 2002: 232), indeed ‘Herod was the dominant architectural and political influence throughout the region (Bedal 2004: 173). A ‘sign of specifically Herodian influence in Petran building projects is the apparent use of the Roman pes as a unit of design for the Pool Complex’ of the paradeisos and ‘these metric similarities suggest Herodian and Nabataean works may even have employed the same builders’ (Cloke 2007: 165-166). It is likely that it was decided that a concrete core would give a dam at Qumran more strength than the dams of Nabataea. Although few dams are known from the early Roman Empire which can be assigned to the first century BCE, most of those that are had a concrete core.22 Roman concrete was made utilising pozzolana or volcanic ash, a raw material not found in Palestine, which would have been imported from Italy. It has been estimated that between 100 and 150 large shiploads of pozzolana had to be imported for the harbour at Caesarea alone (Oleson 2005) so it is logical to suppose that most of the few ‘concrete’ structures known in Palestine, including the Qumran dam, were constructed during the same period as the Harbour, c. 20-15 BCE. ‘These typical Roman materials are very rare in the East of the Imperium Romanum and they can only be explained by conscious and expensive importation resulting from a close relationship with Rome’ (Lichtenburger 2009: 42, and see 50-52).23 The considerable cost and effort involved in the construction of a dam at Qumran I associate with Herod’s desire to bolster his prestigious building projects at, in particular, Hyrcania, Masada and Herodium. Not only did he have to supply them with building materials, e.g. timber from the Galilee, and soft, carveable stone from the quarries at Khirbet es-Samra (Conder and Kitchener 1883: 182, 212-3; Hirschfeld 2004: 65) near Ain Auja about ten kilometres north of Jericho, but the enormous work forces that laboured in these arid areas could not have relied on locally produced crops and the Galilee was the usual source of surplus food. I believe that Herod developed the seasonal site of Qumran into a depot that could support a small force of quartermasters throughout the year to distribute supplies to these projects so dear to his heart. When Agrippa visited Judea in 15 BCE Herod proudly showed him around Hyrcania and Herodium. No doubt Herod would have shown his gratitude to Agrippa had Agrippa expedited their successful completion. Once these palaces were completed, and their storehouses stocked with supplies worthy of the Royal family, the dam would have been superfluous and been abandoned to the flash floods. There would have been a large quantity of water retained behind the dam. In all the other major Hasmonean and Herodian sites in the rift valley (except Jericho which was fed by springs) rain water was stored in underground 22 One in France is tentatively dated to 20 BCE (Smith 1971: 34) and two or three in Spain were assigned to the reorganisation of Hispania by Augustus in 25 BCE (ibid: 43), although, more recently, the Spanish dams have been dated to the early first Century CE (Arenillas and Castillo 2003). 23 Safrai and Eshel note that ‘the aqueduct bringing water’ (from the Qumran dam) ‘to the site is short but highly sophisticated, and required the use of Roman measuring instruments and technology’ (Safrai and Eshel 2000: 232), but they use this observation to bolster their idea that the site was occupied by a wealthy sectarian community though they offer no other example of such instruments and technology being available to a private religious institution. Herod probably realised that, because of the ferocity of the flash floods, any dam built at Qumran had to be more substantial than any that may already hve existed in Nabataea. It was a challenge he may have discussed with Marcus Agrippa, the Emperor Augustus’ right hand man, with whom he had formed a good relationship. In 22 BCE Herod had travelled to Mytilene (Lesbos) to meet with Agrippa who was touring the eastern empire. At much the same time Herod commenced the construction, at ‘vast expense’ (Josephus Ant. 17.87), of a new harbour 17 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Jericho which Netzer placed ‘around 15 BCE’ (Netzer 2001: 9; see Burrell 2009: 220), and which Bar-Nathan accepted when dating the pottery. Thus her dating of ‘Herodian 1’ (HR1) to 31-15 BCE (Bar-Nathan 2002: 5) should be revised to 31-20 BCE (and HR2 to 20 BCE – 6 CE) a revision that she herself seems to advocate in a paper presented in 2002 (Bar-Nathan 2006b: 274).25 This irons out the possible discrepancy of the appearance of spatulate lamps in L114. If they are asigned to HR2 with a revised date of 20 BCE – 6 CE there is no problem with assigning the latest date for the pottery from L114 to 2015 BCE, which would have preceeded the building of the ‘main aqueduct’ over the fill of L114. cisterns hewn into the bedrock. From them water had to be transported, by hand or pack animal, up into other cisterns within the fort or palace, which were at a higher elevation than the initial collecting cisterns. No rock hewn cisterns existed at Qumran, although there is no reason why they could not have been cut into the cliff faces to the west of the site. However, as the dam at Qumran is at a higher elevation than the settlement, it would have been far simpler to store water in the dam pond and lift it into the channel by means of a shaduf, a bucket on a beam with a counter weight, from where it could run to the site by gravity. It is extremely unlikely that, having exerted so much effort and expertise into building the dam, the water gathered behind it would have been allowed to merely evaporate away. The large pools in Qumran, particularly L71 and L91, which have often been identified as miqva’ot (ritual baths) despite the fact that they ‘are considerably larger than most contemporary miqva’ot’ (Collins 2010: 189), were primarily cisterns, the pragmatic equivalent of rock hewn cisterns elsewhere but requiring far less effort to construct as they were dug into the marl. Their mistaken identification as miqva’ot is often used, in circular argumentation, to bolster the case for Qumran being a religious, sectarian settlement (Magness 2002: 158; Collins 2010: 196). 2.2 The Southern wall, ‘W1’, of the ‘Main Building’. There is another piece of evidence, visible in the field to this day, which suggests that the main aqueduct was built after the earthquake. De Vaux did not draw attention to it and, to the best of my knowledge, neither has anyone else in the years since. He identified the southern boundary of Period Ia as a wall, that I will refer to as ‘W1’ (see Plan 1), which later became the northern wall of L77. He suggested that it was Iron Age in origin (de Vaux 1973: 2-4), but was reused in Period Ia, although its Iron Age origin is ‘refuted’ by later excavators (Magen and Peleg 2007: 28).26 What is very noticeable is that either a part of this wall had, at some stage, collapsed and been rebuilt, or, as there is no sign of a rebuild on the southern face of the wall (Fig. 10), that its northern face had ‘peeled’ off. The ‘seam’ between the original wall and the unevenly finished rebuild is clear (Figs. 7 & 9),27 although neither de Vaux, Galor (Galor 2003), nor Magen and Peleg mention it. This localised damage was most likely caused by the earthquake of 31 BCE. However W1 clearly pre-existed the pools, Loci 55-58, which were integral with the earliest stage of the ‘main’ aqueduct, as the north-south walls of the pools, Loci 55 & 57, were built to the reconstructed W1. The western wall of L55 had a slight batter (Fig. 8) and was clearly a retaining wall supporting a considerable back-fill (c. 1.5 m in depth) to its west in the area of L54. [L54 has been wrongly identified as a pool (Galor 2003: 296-7) but, in contrast to Loci 55 & 57, there are no traces of plaster on any of its walls.] The original surface to which W1 was built, and onto which it must have collapsed, would have been just below the rebuild, i.e. the lowest level reached in L54, which was the same as that later used for the plastered floor of the pools Loci 55 and 57 (see remaining course of original wall in Fig. 9). It is worth noting that only two other dams are known in Second Temple Judea, both of which have been attributed to Herod. One, at Ceasarea itself, is conjectured to have been necessary to raise water to the elevation of the High Level aqueduct (Patrich 2002: 105) which was surely built at the time of that city’s foundation [(Levine 1975: 30-35; a date in the first century CE has been suggested (Patrich 2002: 125) but is not convincing (Netzer 2006: 102, 291)]. A dam, 4m high, in the course of the aqueduct from Wadi el-Biyar to Jerusalem is dated to the time of Herod although it is not known if it had a concrete core (Mazar 2002: 225; Amit 2002: 253-266). It so happens that both these aqueducts are, in part, arcaded. Kamash notes ‘As arcades were clearly a Roman technology, it is particularly interesting that two of the aqueducts with arcaded sections may be Herodian: ... This suggests that Herod may have been an agent of transmission of this technology to the east’ (Kamash 2010: 48). I would add the concrete dam as another of Herod’s imported technologies, and it is feasible that his dams inspired the larger, concrete-cored, dams which were built in Nabataea in the 1st – 3rd centuries CE, for example at Mampsis/Kurnub (Oleson 2007: 228; Rubin 1988: 234235; Kloner 1998: 10-16).24 25 Throughout the course of this work I quote her dating taken from the Jericho pottery volume (Bar-Nathan 2002) but the reader should bear in mind the slight revision. 26 Theoretically the lower part of the wall could have been Iron Age, and the rebuild a Hasmonean addition but I assume that if the ‘plaster floor,’ found beneath Iron Age pottery in L77 related to this wall the excavators would have informed us (Magen and Peleg 2007: 27). 27 These photographs (Figs. 7-11) were taken by the author during an enlightening visit in June 2007 accompanied by Benny Arubas and Dennis Mizzi. Although the observations were made by all three of us the conclusions reached here are my own and are not necessarily shared by my colleagues. If we accept that the Qumran dam be dated to c. 20-15 BCE and that the construction of the ‘main aqueduct’ was contemporary with it (see above) then the date of the ‘main aqueduct’ can be narrowed down from ‘soon after 31 BCE’ (above p 11) to between 20-15 BCE. I would also assign this date to the opus reticulatum building in 24 Magness makes no mention of the probable 2nd cent. CE dating of those at Mampsis when she compares them to the water system at Qumran (Magness 2002: 155-157). 18 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - WHEN WAS THE ‘MAIN’ AQUEDUCT BUILT? Fig 1 Fig 4 Fig 2 19 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Fig 3 Fig 5 Fig 6 20 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - WHEN WAS THE ‘MAIN’ AQUEDUCT BUILT? It is also noticeable that the wall running parallel to W1 to its north (which was the southern wall of the Hasmonean courtyard and against which the pools in L34 were built – see Plans 3, 8) was also damaged (and/or partially demolished) before the pools Loci 55-58 were constructed. As part of the floor raising process necessary for inserting these pools, a new wall was built against the southern face of this wall from the pre-existing floor level, probably along its whole length south of L36 and L34.28 Above the raised floor level associated with the top of the pools, the new wall continued above ground on the new foundations in the west but, in the east, was built based on the remains of the earlier wall although somewhat narrower than it. new floor on top of it, often by bringing a small amount of soil in from outside for levelling purposes. This process, after all, helped create the familiar tells of the ancient Near East. Occasionally, where a decorative floor existed or a space had a special function, debris might have been cleared. It must be remembered that an earthquake rarely demolishes walls down to the height at which they are found by archaeologists for whom it may be relatively simple to throw debris outside the walls. It is, of course, difficult to ascertain how widespread the destructive effect would be but in an earthquake with the reputed severity of that of 31 BCE it is unlikely that only one or two rooms would suffer damage; in nearby Jericho the Twin Palaces were completely destroyed by the earthquake of 31 BCE, according to Netzer (2001: 14774). A summary of the evidence that the main aqueduct system, and thus the main building, were built in the early Herodian period:- From my own archaeological experience I would point to Tiberias where there was an impressive Roman street laid with basalt flagstones in a decorative pattern (Stacey 2004a: Figs. 4.2, 4.3), which was still in use in the Umayyad period some 700 years later because one ‘Sulayman’ had hammered his name, in Kufic script, onto one of the stones (ibid: Fig. 4.11, corrected reading thanks to Katia Cytryn-Silverman, pers. com.). Following an earthquake of 749 CE the slabs were buried beneath about 30 cm of debris which was not cleared away but tamped down to form a dirt street. Its level gradually rose over time until the final street, laid down over debris from an earthquake of 1033 CE, was 1.80 m above the Roman slabs. This gradual rise is noticeable throughout Tiberias. Only in the Byzantine church on Mt Berenice, which had mosaic floors, was there evidence suggesting that the debris of 749 CE had been partially cleared, as ‘parts of the mosaic floor of the Byzantine-period church continued in the Abbasid period’ (Hirschfeld 2004a: 113). 1. The aqueduct in the occupied areas was not built in the manner of free-standing aqueducts of the Second Temple period, with plastered external faces. Thus it was built to run beneath the floors designated by de Vaux as ‘Period II’. 2. Access to L115 from L114 would only have been on a floor covering the capstones of the channel which completely filled the width of the door. The pottery found in L114, which dates to the early Herodian period, was from beneath that floor. Thus the channel can be no earlier than the early Herodian period. 3. L130 was back-filled as part of the construction of the aqueduct system. Its pottery dates to the early Herodian period. 4. An outlet channel in L117 was associated with the additional steps that had to be added to the pool when the ‘main’ aqueduct was built at a higher elevation than the pool’s previous water inlet. It ran over a rubbish dump, ‘Trench A’, whose pottery is dated up to the early Herodian period. 5. The ‘main’ aqueduct system was associated at some stage with the construction of a water raising dam. Such dams are only known from the Herodian period or later and would have been state enterprises. 6. W1 was reconstructed, probably following earthquake damage (of 31 BCE?), which occurred before the main water system was built. The wall that ran parallel to it in the north was also damaged demanding a partial demolition. Because de Vaux had already decided at the end of the second season that the crack in pool L48/9 and the broken pottery from L89 were the result of the earthquake (and that, thus, the aqueduct that fed the pool had to be predate the earthquake), we must consider these two loci in light of the fact that evidence from other parts of the site shows that the aqueduct was constructed after the earthquake. 2.3 Pool Locus 48/9. There is no doubt that there is an impressive crack through the floor of this pool. But was it the result of an earthquake and, if so, was that the earthquake of 31 BCE? A number of people with expertise in the field of earthquakes have questioned whether it was earthquake damage. ‘A British expert, loaned to the Jordanian Government under a technical aid agreement by the British Government, Mr Tom Zavislock, an architect with wide experience, inter alia, in reconstructing buildings destroyed or damaged by earthquakes in the Far East, was in charge of the actual reconstruction work at Qumran’... He is reported as saying ‘that there were no traces whatsoever of any earthquake damage to the Qumran buildings’ ... and that ‘the building or repair of these This makes unnecessary de Vaux’s concept of a thorough clearing of the main building of supposed earthquake debris, which could have been 30-40 cm deep, to account for the lack of any occupational layers dateable to before the earthquake which is, in any case, against human nature. It is far simpler, following a destruction, be it military or physical, to tamp down the debris and create a 28 The construction of the later channel to the north of L56/8 and of the small pool L67 obscures whether the new ‘skin’ wall continued south of L34. 21 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS cisterns was faulty and that the cracks were caused by the weight of the water introduced’ (Steckoll 1969: 33-34; de Vaux 1973: 20 fn 1). The eastern side of the pool, insufficiently supported by backfill, was weak and would have had difficulty withstanding the weight of the water once the pool was full. Eventually it slumped, either due entirely to structural weakness or, perhaps, aggravated by water seepage, sub-stratum instability, or an earth tremor. Two Israeli experts also doubted that the damage was caused by an earthquake, but was more likely the result of unstable sub-strata in the area (Karcz and Kafri 1978: 241-2), a conclusion supported more recently by the earthquake scientist Ambraseys (Ambraseys 2006: 1014). Alternatively, recent excavators have suggested that some large cracks at the site may be the result of the earthquake in 749 CE (Magen and Peleg 2007: 10). There were signs of earthquake damage at Qumran, but these should be seen in the thick mud-brick debris in L101, and L10A in the tower, in the buttressing of the tower and of the north-west corner of the western building, and in the rebuild of W1. In Jericho the ‘twin’ palaces were destroyed by the earthquake of 31 BCE. Its rooms were filled with mud-brick debris so compacted that the use of a powerful entrenching drill was necessary to break through it (Netzer 2001: colour plate VI). Beneath the debris the palace was well preserved; neither the stepped pools nor any of the floors had fissures in them (Netzer 2001: 147-171 and Ills. therein). Where cracks in floors were found elsewhere they were usually attributable to the settling of a sub-floor fill, a dramatic example of which was found at Cypros (Netzer 2004: Ills 288, 291). This phenomenon may have accounted for a number of small cracks found at Qumran and identified as earthquake damage (e.g. in L111 & L115). When did the pools L48/9 go out of use? It is difficult to say. Although the little pottery that was published was, inevitably, because of assumptions made, assigned to Period I, most of it is transitional Hasmonean/Herodian and difficult to date with any precision. A cooking pot (de Vaux 1954: Fig. 3:22), however, is certainly Herodian, as is a narrow necked flask (de Vaux 1954: Fig. 3:19) 30 from the neighbouring pool L50 which also showed signs of subsidence. An oil lamp from L52 (de Vaux 1954: Fig. 3:17) in the area east of the pool, is dateable to the second half of the first century CE (Barag and Hershkovitz 1994: 68)31 but probably dates to a time after the L48/9 pool had gone out of use. Although in Jericho most pools were filled in soon after they went out of use, both to prevent accidents and to make use of the space, this was not necessarily so for L48/9. Because its eastern wall stood proud of the adjacent floor there was little danger of falling into the empty pool, and any floor formed by back-filling would have been on a higher elevation than, and difficult to integrate with, the nearby floor to the east. The small amount of pottery that has been published suggests the L48/9 pool went out of use already in the time of Herod, but the numismatic evidence might indicate that it was not finally filled in until about 50 CE (a coin [866] possibly of Herod was found on the stairs of L48 and one [892] of Herod Agrippa (37-44 CE) from the ‘lower level’ of L49, the deepest end of the pool). Perhaps a weakened wall finally slumped during the earthquake of 48 CE (Kallner-Amiran 1950: 225)? One of the causes of the crack in L48/9 was the failure of the builders of the cistern to sufficiently consolidate the backfill that supported its eastern wall. The cistern ‘had been superimposed on the pottery kilns of the earlier period’ (de Vaux 1973: 7). We do not know the exact elevation of the undisturbed virgin soil into which the lower part of the pool was dug. However, the firing box of the western kiln was to its north in the area where the pool was dug, and, judging by the depth of the bottom of the kiln – about 1.5 m below the eventual top of the pool (see metre stick in pl 175) – one can estimate that at least the uppermost 1.5 m of the pool was built above the then existing ground level.29 To ensure stability, the area to its east should have been backfilled right up to the rim of the pool. (The western side of the pool was built hard up against an existing wall.) Many sizeable pools in Hasmonean and Herodian Jericho were dug into the Lisan marl. The excavated spoil was spread around the top of the pool to build up the sides, so that a maximum depth could be obtained with a minimum of digging, but the contemporary floor was always laid down flush with the top of the pool. This does not seem to have been done adjacent to L48 where at least a metre of the eastern side of the pool was free-standing above the neighbouring floor (pls 148, 152, 153, assuming that the floor east of the pool and contemporary with it was at the same elevation as the surfaces in L52, which, based on the material recovered from the surfaces in L52, has to be dated to the time of the First Revolt, though those surfaces had presumably been in use from earlier times). As discussed above, de Vaux immediately identified the earthquake as being responsible for the broken vessels in L89 although he does not explain why they were not tidied away as was, he claimed, all the rest of the earthquake debris in the ‘main’ building. The pottery thus had to be assigned to his pre-earthquake period Ib (de Vaux 1956: Fig. 2). But by comparing the L89 ceramics with pottery found during the widespread excavations at Jericho Bar-Nathan asserts that the L89 pottery must be attributed to the ‘first half of Herod’s reign’ between 3115 BCE (31-20 revised dating) (Bar-Nathan 2002: 2034).32 Humbert, following Milik (Humbert 2006: 31) wants 30 See Bar-Nathan 2002: 43 fn 2 where it is dated to HR3 (648CE). However at Masada (Bar-Nathan 2006a: 107) it has a wider date range (c. 28 BCE- 73 CE). 31 For a lamp from the same atelier found in a First Revolt destruction layer at Ein Gedi see http://eingedi.cjb.co.il/ go to season 5 (2007). 32 Magness attempts to dispute this (Magness 2003: 427) but, because she accepts unquestionably that L89 was destroyed by the 31 BCE earthquake (Magness 2004: 53), her arguments are circular and 29 Note that a little to the north ‘two floors were separated by more than a meter of fill’ (L39 9/4/53). 22 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - WHEN WAS THE ‘MAIN’ AQUEDUCT BUILT? to date the L89 ceramics in the first century CE, on the strength of an inscription on one of the bowls, but this seems to be putting too much reliance on the palaeography of one inscription (see Doudna’s section on palaeographic dating in the present volume). Moreover he sidesteps the issue of the dating of the introduction of the main water system (which, stratigraphically effects the dating of L89) preferring to hedge his bets by writing that ‘since everyone agrees that by the Herodian Period the Qumran water system was fully-developed, and since our main area of interest is the phase of the site from this period onwards, the dating of the water cisterns becomes less imperative’ (ibid: 53). Mizzi claims that ‘a pre-31 BCE existence remains the most likely timeframe for this locus’ (Mizzi 2009: 86 fn 32). However he offers no conclusive evidence. lack conviction. She fails to abide by her own admonition to supply ‘parallels with contemporary sites in the same geographic region’ for an assertion about the chronological development of ‘carinated plates’, and her claim that a jug from Jericho ‘differs significantly in the form of the rim and body’ from one in Qumran is exaggerated (cf de Vaux 1956: Fig 2:1 and Bar-Nathan 2002: Pl 8:58). 23 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Fig 7. The eastern end of L54, looking south east. Benny and Dennis are in L77. Fig 8. The retaining wall at the western end of L55, where it meets with W1. Not remnants of plaster on both walls and floor. 24 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - WHEN WAS THE ‘MAIN’ AQUEDUCT BUILT? Fig 9. L55 looking south east. Rebuild of W1, and outlet channel, clearly visible. Fig 10. Outlet from L55 running through W1 and beneath floor od L77. 25 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Plan 1 26 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - PLANS Plan 2 27 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Plan 3 28 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - PLANS Plan 4 29 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Plan 5 30 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - PLANS Plan 6 31 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Plan 7 32 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - PLANS Plan 8 Plan 9 Plan 11 Plan 10 33 3. Qumran in the Hasmonean and early Herodian Period 100-31/20BCE. (Plans 1, 2) (Hirschfeld 1998: 181). If, however, the stronghold was built hurriedly in response to the same threat from the east that, in Jericho, left Jannaeus feeling so insecure that he buried the existing palace building with the spoil from a 7m deep moat with which he surrounded it (Netzer 2001: 3-4), it may be that the salvaged building stones, including these architectural elements, were brought to Qumran from elsewhere. For safety, Janneus and the estate officials were reduced, in Jericho, to a small building, the ‘Fortified Palace,’ erected on top of the artificially created hill. Dressed ashlars were incongruously incorporated into some of the deep foundations of the Fortified Palace and occasionally into the retaining wall of the moat, and, elsewhere, I have suggested that these stones came from the demolition of A(B)103 nearby (Stacey 2006: 199-200). Surplus stones from this same demolition may have been hurriedly transported from Jericho to Qumran to construct another defensive position at the same time. Qumran in the Hasmonean and early Herodian period is equivalent stratigraphically to de Vaux’s Period Ia (de Vaux 1973: Pl IV) with the addition of the rooms (Loci 111, 121, 120 etc) that were annexed to the western building prior to the earthquake; and the tower, a northern room of which, L10A, was put out of use by the earthquake (pl 18, L10A 18/3/53, and see Mizzi 2010: 174 fn 2). The tower has rightly been added to the Period Ia plan (Humbert and Chambon 1994: Plan III), although, without explanation, no plan at all of Period Ia appears in the English edition of that volume which originally appeared in French. Besides the tower Hasmonean Qumran consisted of pottery kilns and some rooms around a central courtyard, and an industrial zone clustered round three cisterns. The water supply of Hasmonean Qumran was meagre, consisting only of the cisterns L110, L117 and L118, and would only have supported limited occupation of the site during the winter season, when what water there was was at its freshest and most abundant. This undermines de Vaux’s vision of a Qumran that ‘is not a village or a group of houses; it is the establishment of a community .... for the carrying on of certain communal activities’, which he never questioned were religiously inspired and took place year round. Consequently this phase was downplayed and considered to be of ‘short duration’ (de Vaux 1973: 10). However the evidence suggests that the utilisation of the limited facilities at the site in this phase lasted until the earthquake of 31 BCE. Chronologically, but not architecturally, Hasmonean Qumran covers both de Vaux’s Periods Ia and Ib. It is worth comparing the situation in L101 where there were four earlier floor surfaces before that on which the earthquake debris fell (L101 24/3/55) with that in L2 where there were no earlier floors (L2 28/3/55) below that from which, according to de Vaux’s unlikely conjecture, earthquake debris had been cleared. It is, however, one thing for Magness to revise the dating of a clearly discernible archaeological stratum but quite another to decide that it therefore ‘does not exist’ (Magness 2002: 68), thereby removing awkward impediments to an assumption that ‘the settlement was sectarian from the beginning (ibid: 66). The southern wall of the Hasmonean complex was W1, which had a return to the north, running east of L52, L51 etc. Parallel to the eastern end of W1 and about 10 m to the north was another wall against which were built two pools, L34, one of which appears to be a bathtub, with a resemblance to bathtubs in Jericho, particularly to the pair in the ‘Twin’ Palaces which had heating devices (Netzer 2001: Figs. 220-1, 225-6). Although de Vaux describes a ‘drain’ in the north-east ‘which passes under one of the steps of cistern 49’ (L34 22/2/54) this is unlikely, as it would have run through the area of the kiln firebox whose floor may even have been at a lower elevation (cf L64, pl 360; L74,33 pls 182, 368) than the supposed drain. It is possible that the channel through the wall was part of a heating device, similar to those in Jericho, whereby a fire in the area of the kiln firebox heated water in a container west of the wall. The heated water could then be poured into the bath (Netzer 2001: 159-162). If this was a heating installation then the drain that was uncovered beneath the floors in L60 (L60 28/2/54) and L61 (L61 27/2/54) must have related to the ceramic pipe in the eastern wall of L52 (L52 2/4/53). Although de Vaux dated the start of his Period Ia to the time of John Hyrcanus (135-104 BCE), I would agree with Magness that that is too early (Magness 2002: 65). I would put the beginning of Qumran in the early first century BCE in the time of Alexander Jannaeus. A number of architectural fragments found out of context, some associated with debris from the 31 B.C.E. earthquake and others rebuilt into later walls, are thought by some to have come from the ‘main’ building in Qumran (Hirschfeld 1998: 181; Magness 2002: 69, 124; Humbert 2003: Fig. 1). If, however, the columns, voussoirs, and consoles were, indeed, integral parts of the pre-31 BCE architecture, then that building was of considerably more sophistication than the evidence suggests. Moreover it is very difficult to reconstruct logical locations for these architectural elements within the excavated building. De Vaux and Magness have suggested various ground floor locations, but without conviction (Magness 2002: 69, 124), while Hirschfeld, even less convincingly, elevates them to an upper storey 33 L74 was sunk about 50cm below the surrounding area (pls 182, 368) to give access to the firebox of the eastern kiln. Ashes were dumped on the surrounding floor (http://www1.kueichstaett.de/KTF/qumran/bilder/abfrage/loc66.jpg) or carried outside and dumped against W1 (pl 368). It was entered down steps in the north east corner (pl 181) or via a door to the outside in the same corner. 34 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - QUMRAN IN THE HASMONEAN AND EARLY HERODIAN PERIOD 100-31/20BCE (PLANS 1, 2) The floor to which the L34 pools were built was at about the same elevation as the surfaces associated with the pre earthquake phase of W134 (see above). The wall to which the pools were built continued in a straight line to the west and there was a door connecting the courtyard, in which were the pools L34, and the area bound by W1 further south [Humbert and Chambon 1994: Plan X (‘Period Ib’, revised here as my Plan 8), pl 61]. This was not recognised by de Vaux, probably because he did not realise that the top of the pools L55-58 had been built substantially above the then existing ground surface which had been at the elevation of the threshold in a door which de Vaux called a ‘recess’, although in fact it was a door later blocked by the wall built to the south to which L56 was built (pl 61).35 De Vaux claimed that the western end of the original southern wall of L36 to which the pools in L34 were built was a recess whose western side was “cut into the virgin soil of the terrace” near to the line of the later wall of L30. Only the eastern side of the ‘recess’ had a face (L36 13/4/53). The southern wall did not extend further westward. Excavations on its projected line, below the level of the foundations of the eastern wall of the later room L30, found no extension of the wall but uncovered an oven, a plastered installation and a Hasmonean cooking pot (pls 122, 123; Humbert and Chambon 1994: Plan III). As there was no return to the north one must reconstruct a return to the south, evidence for which may be the ‘face’ noted on the east side of the supposed western ‘recess’ which may, in fact, not be a recess but the beginning of a return to the south which would have met W1 where pool L55 was later built, and which would have been destroyed when the face of W1 collapsed. It is likely that the ‘step’ in the floor of pool L55 (L55 25/2/54) and also the wide lowest step of L56 (Galor 2003: Fig. 8) were influenced by the remains of an earlier return wall over which they were built. The space between W1 and the parallel wall to its north could have served as a workshop (see Plan 2) in which the potter would have had his wheel and laid out his pots to dry prior to firing. Rain water gathered on its roof would have fed the pools in L34 and may have been used in the production process as well as, as suggested above, for bathing. It is noticeable that the proposed return wall to the south, which would have joined the western end of the early wall in L36 with W1, is on the same line as the western wall of L38/41, to the north of the courtyard; and the early door in L36 is on very much the same alignment as the doors in the northern and southern walls of L38/41 (Plan 8). A supposedly Hasmonean floor was uncovered in this room (L41 14/3/56; L38, coin 2667 ‘on the pavement from Period I’). Thus a degree of symmetry existed north and south of the courtyard in the Hasmonean period. It is difficult to differentiate the function of the range of rooms to the east of the courtyard. Clearly the kilns would have existed in an 34 Estimated by subtracting 1.50 m from the elevation given for the top step in L56 (de Vaux 1953 Plan V) which gives –3.64. The floor in L34 is –3.72. 35 He did realise that this wall existed before the main aqueduct (‘we excavated an earlier wall in front of the threshold of L36.... The wall founded on the ground was contemporary with the south wall of L36 (and of L34) and was earlier than the channel’ (L36 13/3/56)]. The pools were not fed from the channel that eventually ran north of Loci 56 & 58 [‘it is not from L67 that the basins of L34 were fed unless it was made by an outlet into the basin at the southwest corner of L34’ (L67 27/2/54) for which there is no evidence. 35 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS open, unroofed space at the southern end but there were rooms (L51, Loci 39 and 47, L6/40) at the north with L6/40 creating some symmetry with the tower. continued to be a thriving settlement. Qumran was thus ‘a veritable maelstrom of activity rather than an isolated ascetic site’ (Hirschfeld 2004b: 213). No clear Hasmonean levels can be detected in these rooms from the excavation notes, but they may not have been reached as few probes to virgin soil seem to have been dug.36 Recent analysis of earlier excavations, even further south, in the northern Negev, has shown that Jannaeus built a fortress at Horvat Ma’agurah, south-west of the Dead Sea and that ‘with the aid of this fortress he was able to halt any Nabataean activity ... and force them out of the Negev’38 (and see Taxel 2011: 399-401). The tower at Qumran would have played a small part in the establishment of control over the lower rift valley and the Negev. 3.1 The Tower. It is difficult to date the foundation of the tower from the published archaeological evidence which comes from the time of its partial destruction by the earthquake. However, historically, it was probably built by Alexander Jannaeus (103-76 BCE) who was, ‘during his reign of 27 years, almost continuously involved in foreign and internal wars, for the most part deliberately provoked by him’ (Schürer 1973: 1.220). He attempted to annex parts of Nabatea lying between the southern end of the Dead Sea and the northern Negev (100 BCE; Josephus, Antiquities 13.357; 14.101) but encountered serious opposition and was defeated by both Obodas I in c. 94 BCE (Antiquities 13.375) and, a few years later, by Aretas III, who briefly invaded Judea (Antiq 13.392). In Jericho, Jannaeus felt so insecure that he buried the existing palace beneath the spoil from a 7 m deep moat with which he surrounded it (Netzer 2002: 3-4). On top of this artificially created hill he built a much smaller ‘Fortified Palace’. The fragile political situation would have demanded the hurried erection of a watch-tower in Qumran from which warnings could be given of attempted excursions into the Buqe’ah particularly via the track that wound up the rift behind the site. It also overlooked the path to Ein Gedi (Porat 2006) and the whole of the northern end of the Dead Sea. It was not intended as a defendable fort – the track to the Buqe’ah could more easily be defended by a handful of troops judiciously placed on high ground, from where they could roll boulders and missiles down on the necessarily single file of any attackers toiling up the switch-back track. The control of water supplies, particularly in the desert, has always been a demonstration of political power. In the time of Jannaeus, the pressure on the limited available water supplies in both Jericho and Ein Gedi, for domestic use but especially for the irrigations of crops, the processing of which were so lucrative (Stacey 2006: 191202, Porath 2005), would have encouraged the exploitation of the alternative water resources of Qumran, which must be considered an integral, though outlying, part of the Jericho estate.39 With the expansion of Jannaeus’ sphere of control east of the Jordan the strategic military importance of Qumran was reduced, but the increased security allowed the industrial activities to expand. The tower was of two storeys (at least) and had no entrance at ground level. A door at the level of the first floor gave entrance from the south and, before the construction of the western wing in the Herodian period (see below), must have been accessed by ladder. The ground floor rooms would have been dark and airless and could only have been used for storage purposes. A stairway round a central pillar in L8A led down to give access to all the ground floor rooms except for L9A which must have been a sub-floor space reached only by its own ladder. The stairway, presumably, continued up to the roof. There are some unusual aspects which may give, indirectly, an indication of the original use of the first floor rooms. In the northern wall of L10 there were ‘two narrow loopholes or ventilation holes’ high in the northern wall, and apparently connecting to the upper floor (L10 complementary notes; de Vaux 1973: 6).40 These slits were covered by the buttressing on the outside of the tower following the earthquake of 31 BCE. The only serious damage was to the north-east corner of the tower and the lower room there, L10A, was filled with mud-brick debris (Chambon and Humbert 1994: pl 18. L10, 18 &19/3/53). None of the walls, however, were Qumran was only one of several sites along the western littoral of the Dead Sea to be developed during the Hasmonean period, most probably by Jannaeus, who, eventually, gained control of land to the north-east of the Dead Sea where he established a fortress at Machaerus, c. 90 BCE (War 7.6.2). Harbour installations were built at Rujm el Bahr and at Qasr el-Yehud/Khirbet Mazin (BarAdon 1989) and a large structure at En el-Ghuweir was built (Bar-Adon 1977).37 Further south Ein Gedi ‘A hole in the virgin soil’ was noted (L6 11-12/12/51) but not from what level it was dug. In L51 the ‘lower floor (was) established on the gravel’ (22/4/53) but this gravel only seems to have been excavated where a possible toilet had been dug which would have destroyed any earlier surfaces. The c. 30 cm depth of gravel (pl 150) is not noted elsewhere in Qumran and might represent a levelling fill above an earlier floor. 37 Although Magness claims that a Hasmonean presence at Ein el-Ghuweir is ‘not supported by the ceramic evidence’ (Magness 2002: 219) this is not certain. Four of the storage jar rims (Bar Adon 1977: Fig. 10:2, 4, 7 & 8) belong to type J-SJ4 ‘the most common and 36 characteristic storage jar in the Hasmonean palace complex’ (BarNathan 2002: 30). 38 http://www.mfa.gov.il/MFA/History/Early+History++Archaeology/Hasmonean-rule-extended-to-Negev-10-Dec-2009.htm 39 A broadly similar conclusion had been reached when the excavation of Second Temple Jericho had scarcely begun (Bar-Adon 1981). 40 Unfortunately there are no details as to the dimensions of these slits nor as to whether they related solely either to the upper storey or to the basement. 36 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - QUMRAN IN THE HASMONEAN AND EARLY HERODIAN PERIOD 100-31/20BCE (PLANS 1, 2) built of mud-brick41 so where did the debris originate? The north and west exterior walls of the upper storey of the tower were considerably narrower than those of the ground floor (L8 & L9 24/3/53, L10 12/3/53 and complementary notes. For the change in width in the NW(?) corner of L9(?) see Hirschfeld 2004b: Fig 27).42 De Vaux believed this decrease in width was to create a ledge to support the intermediate floor but does not explain why a similar decrease did not occur on the eastern and southern walls. It seems more likely that the debris found in L10A came from an internal ‘skin’ of mud-brick attached to, but not bonded with, the upper western and northern walls. Could it be that the upper floor originally served as a columbarium with nesting niches easily created by laying mud bricks with gaps between each one? The lack of a door at ground level is something the tower shares with other dovecotes (Negev and Gibson 2001: 125) and the concept of a columbarium being multi-purpose is not unusual (Netzer 1991: 637). The location of a dovecote confined to an upper storey is, however, without parallel, although this may be due to the shortage of excavated towers with well preserved upper storeys. A dovecote would have been a valuable source of food and of dung. An alternative explanation for the anomalies here may be that mud-brick alcoves or cupboards were built onto the western and northern walls. activity took place on it. There are no dedicated living quarters anywhere at Qumran in this period yet this phase lasted from c. 100 BCE until at least 31 BCE. The lack of living space and the limited water supply indicates that the site was only used seasonally for a few months in the winter. Qumran was not the ‘“closed settlement” having ... but little contact with the outside world’ (Harding 1952: 105) as envisaged by the early excavators. Although in a marginal area, for a few weeks in the winter, it played a small but important part in the regional economy being a centre for various industrial activities. 3.2 A note on the man made caves in the terraces near to Qumran. These were probably initially dug to obtain material for making mud-bricks which were used to build upper storeys (see L10A, L101, L16) and even perhaps the upper courses of ground floor rooms. When excavating graves dug into the Lisan marl near Tell es-Samarat near Jericho (Stacey 2004b: 226-8) a bell shaped chamber was cleared which showed no signs of ever having been used for burial and had no dateable diagnostics. The local Bedouin explained that, when they were still building with mud-brick, the marl in that area was ideal for mudbricks, and that the bell-shaped chamber was similar to their own marl ‘quarries’. The Lisan marl is not uniform; in some places it is clean silt, ideal for mud-brick, in others it can have an admixture of gravel and stones making it unsuitable. The caves (caves 4-10) around Qumran were almost certainly dug where the marl was most suitable for mud-bricks; which does not mean that secondary factors were not taken into consideration when deciding where to quarry them. Besides the watch-tower there were pottery kilns and an industrial area consisting of shallow pools (L121), tanks, fireplaces and beating surfaces which utilised in some way the water gathered at the site. The floor in L115/6 (all one floor in the Hasmonean period) was built on a solid foundation of cobble stones similar to that found in cisterns. The area however could not have held water so this reinforced floor indicates that some vigorous beating 41 Although the surviving walls today are as they were in Period III [by far the best stratigraphic consideration is that by Mizzi (Mizzi 2010: 174 Table 7 fn 2)] they are all of stone (pl 17) and it is unlikely that any major alterations in the tower occurred before period III as the ground floor rooms remained operational until 68 CE, and only in L10A was mud-brick debris recovered. 42 The plaster visible is of Period III. 37 4. The expansion of Qumran in the time of Herod, 20-15 BCE – 1 BCE. (Plans 3-7) gradient the ground level over which it ran had sometimes to be raised. This was particularly true at the southern boundary of the site where, as we have seen, the Hasmonean wall W1 was repaired and originally formed a barrier to any southward expansion. The floor level to the north of this wall (that which was utilised as the bottom of pools L55 and L57) was about 1.5 m lower than that in L100 only 12 or 15 m to the west where the settling tank L83 was constructed against the wall. To lessen the gradient towards the east the lowest 2 m of pool L58 was dug into the floor of what may previously have been a workshop and the spoil used to support the steps in L56 and to raise the surface in L79 between L83 and the top of those steps. It must be stressed that the lowest part of most of the pools in Qumran were dug into marl, not hard bedrock. The marl was not self-supporting which necessitated the building of a facing retaining wall at least 0.5 m thick. The backfill on which the upper steps were built (see Ritmeyer isometric reconstruction, below) was even less self-supporting so it was technically better to build broad steps across the whole width of the pools than it was to try to construct a narrow set of stairs against one side of the pool. If water volume was a concern it was easier to extend the length of the pool – the marl could be dug with a pick – than it was to try to build a narrow staircase whose essential retaining wall, together with those of the pool itself, in any case, took up space. Whether there was a total break in the seasonal occupation of the site following the earthquake of 31 BCE depends upon whether any run-off rain water could still reach the three extant cisterns. Although it is likely that the cisterns were partially filled with debris, which would need clearing, it is less certain whether the flow of water into them was totally impeded. The ‘broad’ channel collecting local rain-fall, if it existed already, would still have functioned and any blockage at the entrance to the under floor channel which fed the pools could have been cleared relatively easily to ensure that some water still flowed into them. The archaeological evidence is inconclusive. Whatever the case a new water system was begun soon after 31 BCE, which culminated a few years later, c. 20-15 BCE, in the construction of a dam and an aqueduct connecting it to the earlier rain run-off ‘broad’ channel. At the same time the ‘main building’ was modified and its western wing erected, although there were later subphases in which extensions were added particularly to the south. A number of people have been misled by what they see as the pleasing symmetry of the final plan into concluding that it must have been constructed together with the tower (although all the walls that meet with the tower merely abut it), thereby overestimating its architectural importance (inter alia Hirschfeld 2004b: Fig 21; Humbert 2003: Figs. 1, 3). They overlook that in Jericho there was also a large building constructed to an earlier tower (Netzer 2001: 26-27, Plan 11)43 as there was also in En Boqeq, where a tower integrated into a later building has been tentatively dated to the days of Herod (Fischer et al. 2000: 137), although a Hasmonean date is not totally ruled out. Although de Vaux did not differentiate any stages within the construction of the aqueduct and the pools it fed (de Vaux 1973: 9-10) he ignored some of his own observations. It would appear that, originally, the channel from L83 fed only L 56/58 with a branch to the south that penetrated W1 (Plan 3). The southern branch ‘belongs to Level I. At Level II (or before...) it was abandoned. Its point of departure from the main channel is closed off (it seems also that the main channel had then been raised). The passage through the wall was then blocked by stones’ (L54 25/3/54). At this early stage in the Herodian period it is possible that the Hasmonean kilns still survived partially determining the location of the east wall of L5658. However the loss of any workshop area would soon have determined the movement of the kiln to a more open area further east. 4.1 The Aqueduct and pools. As already noted, the ‘main’ aqueduct within the occupied area was at a higher elevation than the original water channel, necessitating the raising of the sides of the three pools L110, L117 and L118. It then continued to the south where it was gradually extended and new pools added over time. The engineers had to take into account that the ground sloped down quite steeply from west to east with a drop of at least 2.5 m over 35 m (de Vaux 1954: Plan V). To ensure that the aqueduct had a gentle The main channel was raised so that it could skirt the northern side of L56/58 to feed what must have been a newly built pool L48/9 (which demanded the demolition of the kilns in L66 and the establishment of a new kiln (L84) and working area east of the door in L74) (Plan 4). ‘We searched in Locus 54 for the arrival point of water to cistern 56. We found the entrance of channel I but not that of channel II. Inversely, it appears that during period I, there was no channel extending along the building to the south’ (he is referring here to the channel north of L56 and south of the main building) ‘but rather that it existed during level II’ (L54 28/3/54). From the meagre photographic evidence (pl 191) it does look as though the 43 Although in both Qumran and Jericho structures integrating a corner tower were built around a central courtyard and have some architectural elements in common (particularly the twin doors next to the tower; in Qumran see L12 and L13; in Jericho see A(A)45 and A(A)36) the Jericho structures are earlier and the similarities should not be taken as evidence that the structures at the two sites are contemporary. At Masada the similarities between the plan of certain buildings with that of the twin palaces in Jericho led Netzer to posit a Hasmonean date for those Masada buildings. Excavations in 1989, however, proved that they were early Herodian (Netzer 1991: 646-49). The Jericho tower must be dated earlier than the pottery of Alexander Jannaeus found in the attached building (Bar-Nathan 2002: 5-6). For the possibility that the tower was Seleucid see Stacey 2006: 198-200. 38 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN THE EXPANSION OF QUMRAN IN THE TIME OF HEROD, 20-15 BCE – 1 BCE (PLANS 3-7) inlet to the pool L56 was blocked, but at what time it is difficult to say because the abundant pottery found in the two pools (pls 187, 188) has yet to be published.44 De Vaux claimed that the two pools were filled in by the Romans after the First Jewish Revolt (de Vaux 1973: 43) but gave no evidence. The upper metre, at least, of the eastern wall of L58 was free-standing (pl 182, http://www1.ku-eichstaett.de/KTF/qumran/bilder/abfrage /loc66.jpg) and was only 50 cm wide. It is doubtful if this would have been able to long withstand the weight of water when the pool was full.45 The floor of the pool did have ‘a long fissure which fractured the cistern close to its eastern end’ (L58 2/3/54) and it is possible that the wall that was built between L56 and L58 (L56 22/2/54) was intended to completely isolate the structurally unsound eastern end of the pool. It does appear that pottery was deliberately dumped onto the accumulated silt layer after the pool L58 went out of use (L58 3/3/54, cf pl 188 and 304). A rough north-south wall in the fill dividing L58 in two (L58 25/2/54, 4/3/54) suggests that the partially filled-in pool was made use of for other purposes. L56 could have continued as a pool until such time as its water inlet was blocked. The numismatic evidence suggests that L56 went out of use and was filled in, perhaps gradually, by, at the earliest, 42-43 CE (coins 980 and 995 of Agrippa I, ‘upper level’). slightly (Plan 5) and, at the point where formerly it had terminated at L48/9, a new section was built turning south and hugging the outside of the eastern wall of L58 where, as de Vaux quite rightly said, it rested on a fill from the digging of cistern L71 (L66 2/3/54, 9/3/54). Whereas north of L56/8 ‘there are two stages to the channel’ (L42 13/4/53) only one exists in the section south of L48/9 which must, therefore, relate to the second, higher stage, which hints that water stopped running into L48/9 already at this stage (contra at the later stage shown in Humbert and Chambon 1994: Plans XII, XIII). This channel ran south until its further progress was blocked by W1. A small break had to be made in W1 to allow passage to the channel (L66 3/3/54)46 and at the same time a large room L77 was built south of W1. Its eastern wall was bonded into the rebuild of W1 after the channel had penetrated it. The channel was built together with the eastern wall of L77 and hugged its eastern face before turning east to feed the large pool L71, which was a major replacement for L48/9 and also perhaps L58. It should be noted that the eastern wall of L77 which was bonded with W1 when the channel was laid was bonded with its southern wall which, in turn, was bonded with its western wall. The western wall of L77 (‘W2’), however, only abuts, at its northern end, the pre-existing W1. Thus it is clear that L77 was not built together with the first manifestation of the main aqueduct but two sub-phases later. These three sub-phases can be summarised thus:Phase 1 (Plan 3) – the channel from L83 to L56/58 and a southern extension that penetrated W1. When pool L48/9 slumped and went out of use, the channel along the northern side of L56/58 was raised 44 Only one vessel, a lid, has so far been published (de Vaux 1956: Fig. 5:2). 45 The contemporary floors in L32, L36 to the north of the pool are more than a metre lower than the top of the pool (de Vaux 1954: Plan V). Although the north wall of the pool was built against, and at the same time as, another wall, creating a ‘double skin’, one would expect, from parallels in Jericho, to find a consolidating earthen fill between the two walls. 46 De Vaux noted the possibility that W1 ‘is from period I and the main channel from period II’ (L66 3/3/54). Pool 68 which was fed by the channel was ‘without a doubt from period II’ (L68 2/3/54) but he seems to ignore his own observations later and L68 appears on the period Ib plans. 39 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Phase 2 (Plan 4) – the addition of the channel that skirted L56/58 to feed the new pool L48/9. Probable isolation of L58. New kiln L84 working area in east, if not already in Phase 1. Phase 3 (Plan 5) – L48/9 cracks and goes out of use although it was never deliberately backfilled as it stood proud of the surrounding floors; the southern channel through W1 is blocked and the channel skirting L56/8 is raised slightly and a new channel is connected to it running south against the walls of L58 and L77 before feeding a new large pool L71. Room L77 is built together with this channel. 4.2(i) Two pools associated with L56/58. Locus 55 That this pool was built to the repair of W1 is clearly visible, as is an outlet channel, associated with the repair, which goes through the wall at the eastern end of the pool at a height of about 40 cm above the floor (Figure 9). The top of the channel can be seen emerging on the southern side of W1 (Figure 10), although to discover its destination would require some small excavations. Locus 57 Activity associated with Phase . No outlet channel, similar to that in L55, can be discerned in pool L57. However, a plastered niche was visible in the surviving upper course of W1. This was not mentioned by Galor (Galor 2003) but may be indicated on a plan by Magen and Peleg which may also show the outlet channel in L55 (Magen and Peleg 2007: Fig. 43; both the niche and the outlet from L55 can be seen in Fig. 64). As there is no evidence for any sort of floor covering the pool the niche would have been at about head height for anyone standing in the pool itself. One hopes that in Magen and Peleg’s final publication more evidence will be presented so that the function of this niche(s) may be better understood. As it is doubtful that it would have been possible to bore a channel through an already existing, substantial stone wall it is more likely that the penetration of W1 for the ‘southern channel’ was done together with the construction of the nearby doorway in W1. There are no clear signs of a rebuild of W1 anywhere along its southern face, which may indicate that either far more care was taken to marry the repair to the original on what was then an external surface of the complete building, or that only part of the wall’s northern face had peeled off. The doorway, in any case, lies to the west of the visible repair of the north face of W1. The Hasmonean floor level to the north of W1 can be reconstructed with some confidence and we have shown that when the main aqueduct was built the floor level near the area of the doorway was raised by about a metre. The situation to the south of W1 is less certain but it seems likely that W1 had served in part as a retaining wall for higher ground to its south. This would have been deposited since the Iron Age by erosion from the higher plateau to the west. The outlet of the channel through W1 is at the same elevation as the lowest step in the actual doorway (which is some 25 cm higher than the Phase 3 floor in L77). The channel may have been intended to irrigate a small garden in the area of the ‘southern refuse dump’ in which was found pottery from the ‘first half of the first century BCE’, and which was ‘abandoned already in the first century BCE, due to changes in the main building’ (Magen and Peleg 2007: 5-7) or, possibly, to have terminated at a pool L68 which appears to have existed before the Phase 3 channel which eventually fed it. Although a channel feeding L68, after running north of L72, is shown on the plan (Humbert and Chambon 1994: Plan IV), branching off the phase 3 channel, there is no mention of this junction in the field notes nor does it appear in the limited photographic record (pls 365, 366). Moreover there was a direct link to the channel in the south-west corner of the pool and it is unlikely that such a small pool would have had two feeder channels from the same water source. In one photograph (pl 365) it appears that the northern inlet was at a lower level than the main channel and may have pre-existed it. A handy supply of water in L68 would have greatly benefited the potters whose kilns were nearby. There were no signs of water inlets that might have fed the two pools, Loci 55 & 57, and the reconstruction of a channel (Galor 2003: Fig. 8; cf pls 186, 191) is as mistaken as the identification of L54 as a pool. Water to fill the pools must have been bucketed from the large reservoir L56/58 by standing on a path created on top of the wall which separated it from pools L55 and L57. 4.2 (ii) The implications. W1 clearly pre-existed the pools, Loci 55-58, which were built to its reconstruction. The outlet from L55 indicates that the pool may have been used for some ‘industrial’ process requiring frequent changes of water, such as scouring wool, processing hides or retting flax. The location of the outlet 40 cm above the floor suggests that a considerable sediment was expected that could cause blockage to the overflow channel and/or was useful in itself. A possible process would be the preparation of hides, the lees from which, a mixture of excrement and animal hair and fats, would not only block the outlet but, if collected, was a very nutritious garden ‘compost’. The outlet could have been controlled with a sluice immediately south of W1 later incorporated within the floor of L77. Activity associated with Phase 3 This was a major phase. For the first time development took place south of the originally Hasmonean wall W1. The quality of the hydraulic engineering is superior to that evident in the first two sub-phases. Instead of building large pools into the natural slope with their 40 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN THE EXPANSION OF QUMRAN IN THE TIME OF HEROD, 20-15 BCE – 1 BCE (PLANS 3-7) down-slope, unreinforced, eastern side-walls partially free-standing, a method doomed to failure, the large pool L71 was built in a relatively level area and its sides were built flush with the contemporary surface, which was in part created with spoil from the digging of the pool. The long wall running from east of the pool to the southern end of the plateau was built together with L71 and served as a retaining wall for the pool (between the wall and the pool was ‘simply fill’ L71 9/3/54 and note particularly the coin, 1536, of Herod the Great from within that fill which helps date the construction of L71 and the long retaining wall) and for a garden in the south. Can one detect here the expertise of the engineers who constructed the dam and the sophisticated aqueduct which brought water out of the Wadi Qumran? A paved square that was built to L77 and L86 had a water channel, probably for drainage, at its southern edge (Magen and Peleg 2007: 11, Figs. 4, 17). This channel was later partially filled in to allow the passage of an overflow channel from L91 the construction of which must have post-dated the paving. The room, L86, was an annexe to the main storeroom L77 and the broken vessels found in it were probably part of the production of the Qumran kilns awaiting sale (Donceel-Voûte 1992) and not from a ‘refectory’ pantry. A recent scholar, having admitted that ‘the interpretation’ (of L77 as a dining room) ‘is largely based upon the presence of the pantry..’, dismisses the likelihood that the vessels found here were pots made in the Qumran kilns awaiting sale with some illogicality. ‘Had the Locus 89 niche been a place to store pottery that was destined for sale in the regional market-place, one would expect the pottery to have been picked through and removed in an attempt to salvage undamaged pottery that could still be sold for profit. Because the pottery was left in place and buried, it was most likely not for resale, but rather pottery used domestically by the residents’ (Cargill 2009: 188). However, if these were indeed the bowls used by the residents for their daily meals, presumably the only ones they possessed, they would have had a far greater incentive to retrieve some on which to serve their next meal than they would to sort through a stack of plates awaiting sale for which they had no immediate use, and may not even have owned. When the large room L77 was built south of W1 the existing sloping ground surface was levelled by removing soil from the higher western end and using it to level up the lower eastern end.47 This levelling necessitated adding two rather awkward steps in front of the Phase 1 door, which protruded into the room itself (Humbert and Chambon 1994: pls 324, 335). The ‘southern channel’ through W1 from L54 had been blocked in Phase 2 and its continuation south of the wall may have been partially destroyed in the levelling process of Phase 3. De Vaux, influenced by the concept of Qumran as a communally-eating, sectarian settlement, extravagantly identified this room as a ‘refectory’ and ‘a place where the president of the assembly would have taken his stand’ (de Vaux 1973: 11), although evidence from Masada, already available to de Vaux before his death, shows that such long, comparatively narrow, rooms were used as storerooms (Yadin 1966: 86-105). Similar storerooms have since been found at Herodium (Netzer 1981: 21-25) and Jericho (Netzer 2001: 131-32, Plan 14). The channel that he claimed was used to clean the floor of the refectory had, in reality, gone out of use before this room was built and the slope to the floor which he averred was deliberate and ‘made it easy to be washed’ (de Vaux 1973: 11) was more likely caused by the settling of the backfill at the eastern end. An obvious anomaly is the number of bowls, over one thousand. If this was an active pantry are we to assume that there were a thousand residents each with his own bowl? Milik estimated the ‘community’ consisted of as many as 200 members (Milik 1959: 97). More recently Patrich suggests ‘perhaps no more than thirty’ (Patrich 2000: 726). Both consider Qumran as an Essene community and that this was their pantry, yet, if we are to accept Josephus’ observation that, when Essenes ate, a single bowl of food was set before each of them (War 2.131), the population would have to have been closer to a thousand, which is extremely unlikely. Neither recognized that the site was seasonal and that the population could fluctuate from zero in the summer months to, perhaps, as many as 75 on a very busy day in the winter.48 Phase 4 (Plan 6) In this sub-phase Room L86 was built adjoining L77 in the south-west. Their western walls, W2 and W3, can be seen in the field to be on a slightly different alignment (cf Humbert and Chambon 1994: Plan XXV) and, although the Period III aqueduct broke the connection between W5 (Plan 1) and the southern wall of L77, it can be seen that the walls had not been bonded (pl 348). Observations in the field seem to show that W3 and W5 were bonded with W4, although after 50 years exposure this is not certain. Both W2 and W3 were partially retaining walls, supporting higher ground to their west. 48 If we assume that all the industrial processes (see below) were being carried out simultaneously, which is unlikely, and that each process – shepherding, leather working, wool scouring, potting, flax retting, working with date-palms, herb gathering, lye making, bitumen and salt collecting, passing balsam caravans - had on average five workers, which is generous, that would be a total of fifty. Add a few general labourers, a couple of passing coracles bringing potting clay in and taking completed pots out, a few donkey drivers loading up with hides or wool for the Jericho market, and, perhaps ten permanent quartermasters in the Herodian period, at a stretch there may be as many as 75 people present. But half that would be a more realistic average number in the height of winter. 47 The necessity of levelling up the eastern end is echoed in a similar fill (L73 4/3/54) that had to be deposited immediately to the east of L77, and also one east of L58 (L66 9/3/54). 41 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Phase 5 (Plan 7) overflow from the pool more often than on the two or three days in the year when flooding occurred – further evidence for the use of a shaduf . In this, the final sub-phase connected with the development of the water system, the pool L91 was dug into ground that sloped down quite steeply from west to east with the top of the west side of the pool being at close to the original surface (Humbert and Chambon 1994: pl 302). The eastern wall of L91 had to be built partially above ground but, unlike earlier pools, a backfill was poured to raise the surrounding floor surface to the level of the top of the pool. This backfill, which would have consisted mainly of spoil from the digging of the lower depths of the pool, was supported in the east by W2 and W3, and, in the south by a retaining wall W6 which was not bonded with W3 and W4 and must therefore be technically later than them. Part of W1 had to be demolished and steps accessing the pool L85 were built north of W1 adjacent to the settling tank L83 which also supplied the pool with water. These were technical sub-phases, probably not separated by long periods of time, with water storage capacity being added pragmatically. The first phase should be dated to soon after 31 BCE. The superior hydraulic engineering exhibited in Phase 3, and the ambitious size of the pool L71 associated with it, perhaps indicates that this was work contemporary with the construction of the dam, which, as already suggested, might date to 20-15 BCE. The alteration to the water system north of the buildings, in which the pool L132 was filled in with a “fill of fine earth with few sherds” (L132 7/3/55), and a direct channel built along its eastern edge, probably occurred in Phase 5 with the fill originating in L91. As elsewhere this channel could not have been built freestanding and the surface relating to it would have been at the elevation of the channel’s capstones. The pool L138 was constructed to this surface which was at a higher elevation than the pool could actually be filled. The water inlet is at the level of the third step down (Galor 2003: Fig 3) so the top couple of steps would be superfluous unless the ground level to which it was built was at the elevation of the top step. How exactly this pool received water from the aqueduct is uncertain. Water must have been drawn from the aqueduct before it plunged down the short slope of basin L136 (pls 255, 256). Although a connection between the basin and the pool is shown on a plan (Humbert and Chambon 1994: Plan XX) this is a bold reconstruction and ‘what lies between the basin and L138 remains obscure’ (L137 9/3/55).50 De Vaux’s notes refer to two levels in the area east of the pool (in L81, north of the Period III channel that ran across the area), but the lower level was well below the top of the sidewall of L 91 and must thus predate it (pls 302, 315-318) and most likely represents the ground surface from which the pool was dug, and not a floor related to the pool. It is possible that all five of the jars found here, and interpreted by de Vaux as being from two levels, with two standing on the doubtful lower floor and three embedded into the ‘later’ floor, were in fact all associated with the ‘later’, in fact the only, floor, which was associated with L91. A ‘rectangular basin’ noted in L 81 and a bath-like pool in L88 (Humbert and Chambon 1994: Plan XXII, pls 302, 311-312; but not mentioned in field notes) both appear to rest above a loose back-fill and probably stood on the ‘later’ floor.49 They may have been fed by water collected on the roofs of L77 and L86, which may also have been channelled into three inlets noted in the top of the eastern wall of L 91 (Humbert and Chambon 1994: Plan XXII; Galor 2003: Fig. 7) to which should be added a fourth inlet (Magen and Peleg 2007: Fig. 42). The pool in L88 was not in existence for long as it was built over by a stone construction which may have been merely buttressing for W3, or, perhaps more likely, a staircase leading to the roof. From the little available evidence it is difficult to date the time of L91’s construction. From the ‘lower’ level in L81, which is probably from the backfill, apparently came a ‘Herodian’ lamp and a cylindrical jar, with an ovoid jar sunk into the floor sealing the fill. These, particularly the latter, give a probable date in the first half of Herod’s reign. Coins tentatively dated to the time of the procurators under Tiberius were perhaps introduced during the construction of the ‘Period III’ aqueduct. The whole area, L81/88, was probably unroofed and used for various processes requiring a ready supply of water. The outlet from L91 to the south indicate that water could be relied upon to 4.3 Overflows within the system. In the Hasmonean period there was an outlet from the round cistern L110 which could discharge excess flood water into the wadi via a short drain below L111 and L103 (pl 306). When the sides of the cistern were raised in the Herodian period this drain had to be blocked off. There are a number of ‘overflow’ outlets in the Herodian water system within the settlement which do not appear to be for draining unwanted, excess water at the height of a flood. Such water could more sensibly have been conducted directly into the wadi before it entered the built-up area and caused erosion and damage to the plastered pools and channels. The most likely location of an overflow sluice is close to the eastern end of the broad channel (Humbert and Chambon 1994: Plan XXIX ‘déversoir’), but there would have been many possible 50 What is certain is that L138 could not have co-existed with the decantation basin L132 as shown in Humbert and Chambon 1994 Plan XIX. The water in L132, even at its highest, would have been at a much lower elevation than the water in L138 as there were no free-standing, waterproof walls to either the east or south of L132 which would have been necessary to raise the water level that high. There is no hydraulically plastered retaining wall south of L138 which would have been necessary to support both it and the outlet pipe that exits the pool in the south before running towards the wadi, whilst at the same time retaining water in L132. 49 Assuming they were free-standing basins. If however they were sunken basins then the floor to which they related was not noted at all and may have eroded before the excavations. 42 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN THE EXPANSION OF QUMRAN IN THE TIME OF HEROD, 20-15 BCE – 1 BCE (PLANS 3-7) locations for an overflow before the aqueduct reached the built up area. of flood by overflow from L91, thereby eliminating the need for either trapdoors or bucketing. A number of these ‘outlet’ channels give every impression that they were built to transport water to unknown destinations with some regularity. Such a one is the channel from L117, discussed above, which relates to the late stage of the pool after its sides had been raised, together with that of the round cistern L110. Curiously de Vaux postulated that “the overflow from the cisterns and the water used in the various industries in this area was carried off by means of (this) drain which can be traced northwards to the point where it leaves the settlement” (de Vaux 1973: 9). This seems to imply that, having gone to the trouble of bringing fresh water into pool L117, it was then contaminated with waste water from industrial processes, which makes no sense. The channel was well built and covered with capstones and ran for a considerable distance, in part over an earlier refuse dump (Magen and Peleg 2007: Fig. 11). It seems likely that this ‘dump’ was deliberately filled, at least finally, to provide passage for the ‘channel’ possibly to create an irrigated, terraced garden to its east where the turnstile and restaurant now stand. Most of its length, however, has been excavated revealing no outlet sluices. Although today it is truncated by a small northern wadi it may, in antiquity, have terminated in a settling tank, since eroded, that could distribute water to irrigate small garden areas. So long as water survived in the dam pond, however, it would have been simple to replenish all, or most, of the pools in the built up area and to allow water to overflow in these channels to irrigate much needed seasonal gardens. Although the areas that could have been irrigated using these channels was small, an enterprising gardener could have encouraged limited crops of quickgrowing plants such as herbs, onions, pulses, gourds, or mallow with which to vary the diet, or of plants such as woad useful for dyeing. The addition of nutrient-rich lees from tanning vats53 may even have encouraged a few palms to grow, more for shade than for fruit; even two or three would have been welcome to anyone who had to survive a summer at Qumran. Recent excavations have shown that “a hole in the eastern wall of Pool L48 enabled water to be fed into a “basin, from the bottom of which a pipe led north” (Magen and Peleg 2007: 12). Although the excavators do not attempt an explanation it seems unlikely that so much effort would be expended if water would only have flowed in this system on the very occasional days when there were floods. Again it appears that the pool L48 could be more regularly refilled. 4.4 The main building. There is an outlet channel from the southern wall of L91, and, less certainly, from the south of L71,51 both of which probably lead to a garden area on the southern terrace. Although theoretically it would have been possible to feed these channels by bucketing water from the pools themselves this seems unlikely. The outlet channel connects directly to the pool and if water was being bucketed out such a connection would have been unnecessary. Moreover there would have been less water loss if bucketed water was poured into a wide, shallow basin situated at the start of the outlet, but no such basin existed. De Vaux proposed that ‘the cisterns were uncovered’ but realised that this would have resulted in a high loss to evaporation (de Vaux 1973: 9). It is more likely that they were roofed with something sufficiently substantial to withstand the vicious winds, usually accompanied by whipping sandstorms that can blow up during the winter months.52 Any roof for these two pools had no need for much superstructure and could have stood no more than 50cm above the maximum water level. If water was to be bucketed out of the pool, however, some sort of trapdoor would have had to be included in the roof. It would have been simpler to locate a pool further to the south of the terrace, where the water was clearly wanted, which could have been filled at time The southern, eastern and northern wings of the main building were adaptations of the earlier Hasmonean structures. The western wing, however, was an Herodian addition. There is no evidence of earlier Hasmonean layers beneath its earliest floor [in L2 “two soundings were made in order to verify that there was not a lower floor” (L2 28/3/55). A ‘plate’, 2617, was found in one of the probes but is unpublished]. It was built on a slope which would have necessitated a considerable degree of back-filling in order to create level floors (there was a difference of up to a metre between contemporaneous surfaces either side of the eastern wall of L30, see de Vaux 1954: Plan V). Such earth moving was not a characteristic of Hasmonean Qumran but has already been noted in relationship to the construction of L56/8, L71 and L91 and, partially, L77 in the Herodian period. Moreover, if the western wing had been built in the Hasmonean period it would, following any architectural logic, have connected to W1. The most rational explanation for its failure to do so - and there is no evidence that the north-south walls ever did meet W1 - is that it was conceived together with the introduction of the pool L56 and the channel that fed it which date to the Herodian period. The main rooms of the west wing (Loci 1, 2, 4, 19 etc) were built first at the top of the slope. Soil to level the floors could have come from the beginning of the It is not certain that the southern ‘outlet’ is at a lower elevation than the inlet. It is illustrated as an outlet by Galor (Galor 2003: Fig. 14), and was considered as such by de Vaux (de Vaux 1973: 10), but not by Magen and Peleg (Magen and Peleg 2007: Fig. 20). 52 Beams of palm trees would have been essential (Galor 2003: 302-3) but flimsy roofing of mats or skins would not have survived the windstorms. 51 53 Sedimentary waste from tanning was commonly used horticulturally in nineteenth century England (Cresswell 2006: 116) 43 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS quarrying of the lowest 2 m of pool L58, which were dug into the existing ground surface. There would have been free passage to carry spoil from L58 because L30 was technically built later. (The southern wall of L30 was not bonded with the wall east of L2). After the foundations had been built and backfilled the lower parts of the walls of L30 would have been built as a box into which a large quantity of soil had to be poured to create a level floor. This would still have come from L58.54 The upper parts of the walls of L30 could then be built, and the southern one would have supported the construction of the steps of L56 built, together with the raising of the ground to their west (L54, L79), on the remaining spoil from L58. The upper part of the north wall of L58 was supported by a new wall slightly south of the partially demolished earlier wall (that associated with L34).55 That the western end of this new wall abuts the wall of L30 indicates the technical phases here were designed to allow spoil from L58 to be distributed with as little hindrance as possible. 54 The part of the pool L58 that was dug into the existing ground was c. 2 x 6∙5 x 3∙5 = 45∙5 cu m. This was more than enough to supply spoil for levelling L30 c. 4 x 13 x 0∙5 (av.) = 26 cu m with some left for elsewhere, e.g. L1,2,4 etc., and for creating the steps in L56 which were built above the existing ground surface. 55 de Vaux was aware that ‘due to the slope of the marl terrace the southern walls are retaining walls’ (L32 2/3/54 and see L36 13/4/53) but does not seem to have thought through the implication that, for example, L56 was built almost entirely above the original ground surface. 44 5. Herod and Post Herod occupation. 5.1 The south-west of the ‘main’ building. paved floor in the two rooms formed either side of the wall and the wall could have been built into this accumulation, in which case a foundation trench should have been visible, although as the level 2 floor largely eluded de Vaux it is doubtful if he was looking for foundation trenches sealed beneath it. It was only in the threshold between L1 and L2, based on large stones (pl 146), that de Vaux could identify a second floor surface, ‘Level 2’,58 elsewhere it ‘eluded’ him. It is possible in a few loci to distinguish modifications made within the Herodian building. Although the field notes for the first loci excavated are reasonably comprehensive they soon became vaguer. It may be that de Vaux quickly felt that he generally understood the archaeology and did not need to record everything in great detail. Unfortunately even in loci where the notes are at their most comprehensive and some of the ceramics has been published, as with Loci 1 and 2, the limitations in the record keeping, and, one suspects, the standard of excavation, prevent a confident interpretation of the data. In 1951, the first season of excavation, all material culture from the seven loci dug, Loci 1-7, was numbered sequentially from 2-146.56 Objects, however, were not registered on the day they were found. It would appear that all the pottery from Loci 1-3, dug between 24/11 and 3/12, was not assigned numbers, 2- 47 before December 3rd because no. 27 was a jar buried in the floor of L2 which was only removed on that day (L2 1-3/12/51). Nos. 96-98 were assigned to three vessels found in L7 which was only excavated on 12/12/51 which was the last day of the season. As all the coins, from all loci, were numbered from 102-123 they could only have received these numbers on the last day of the three week season. The potential for error, particularly with the small coins and their context, is obvious. Resting on this accumulation, in L2, was a layer of black ash and debris from a fallen ceiling made of reeds and earth in which were found ‘quite a few potsherds’ (L1 & L2, 1-3/12/51). Of the ‘many sherds’ several forms were published (de Vaux 1953: Figs. 2,3,4) and, as de Vaux notoriously only saved largely complete vessels, it would appear that much of the pottery was complete or restorable. A substantial accumulation with restorable vessels is most likely to represent either a deliberate back-fill, a destruction, or a period of abandonment. As there is no mention of ashes or of fallen roof or mud-brick within the accumulation, it is more likely to be a back-fill. It was described as a ‘couche de marne’ or layer of marl (de Vaux 1953: 91) or as consisting of ‘argillaceous earth’ (Humbert and Chambon 2003 L2 1-3/12/51), i.e. resembling white potter’s clay, a word not used elsewhere for an accumulation.59 This description may indicate that the fill was of Lisan marl dug out during the construction of a cistern. However some anomaly between the pottery and the coins may indicate another source. In subsequent seasons registration does seem to have occurred more frequently, although several numbers seem to have been missed. Coins, certainly, were registered during the course of the dig so are more likely to be accurately recorded. The pottery from within these accumulations can be dated generally to c. 31 BCE – 50 CE.60 Whilst many forms which begin in the time of Herod continue with little change well into the first century CE there are a number of vessels that would better fit a terminus ante quem of 5 BCE particularly, an ovoid jar in L1 (de Vaux 1953: Fig. 2:7) and, in particular, the cooking pot found in L2 (ibid: Fig. 3:7)61 which is of a Hasmonean form which finally went out of use ‘toward the end of Herod’s reign’ (BarNathan 2002: 69). Loci 1 & 2. These would appear to have been originally one large room (Humbert and Chambon 1994: Plan XI) with a floor paved with small stones. As such it would have had two entrances from the north, one located centrally and a second only 3 m away in the north-west corner. The second door would make more sense if it gave access to a separate room, i.e. L1. In any case it was soon blocked57 (see L4 below). The wall that divides the area into two rooms is, however, most likely a later addition not being bonded with the other walls. It was based, without foundation on the earliest (‘Level 1’) paved floor which may have remained in use with it (as indicated in Humbert and Chambon 1994: Plan XI). Let into the paved floor, in L2, was a cylindrical jar (de Vaux 1953 Fig. 2:4. Humbert and Chambon 1994: pls 140142). There was a substantial accumulation of soil, about 55-65 cm in depth, with ‘many sherds’, covering the It may be that there was an early Herodian floor sealing in an occupational accumulation, which included the earlier Herodian pottery, that had built up on the paved Level 1 floor, and that it went unnoticed. The numismatic evidence, which as discussed above may not be totally These ‘Levels’ are local and must not be confused with de Vaux’s ‘Periods’. 59 It is used elsewhere (L56 23/2/54) but not to describe an accumulation. 60 As all the pottery was registered at the same time it is impossible to know if any of that published came from the level 2 floor. 61 It is possible that this cooking pot came from beneath the Level 1 floor in the large pit dug to remove the cylindrical jar (pl 142). ‘Many sherds’ were recovered at that time but the cooking pot is not specifically mentioned. An unpublished ‘plate’, 2617, was later found in a probe beneath the paved floor. 58 56 No. 1 was a Hasmonean storage jar found in Tomb 4 (de Vaux 1953: Fig. 2:5). Nos. 124 and 145 do not appear to have been used. 57 In pl 147 the blocking has been made artificially to look as though it only took place from 50 cm above the paved floor. In fact the blocking commenced at the paved floor as can be seen today on site. And see http://www1.ku-eichstaett.de/KTF/qumran/bilder/abfrage/ loc1.jpg 45 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS reliable, would not, however, support this for the latest of several post-Herod coins said to have been found on the original Level 1 floors were, in L1, of the procurator Antonius Felix [102] (dated to 54 CE), and in L2 of Agrippa I [112], i.e. some fifty years later than the likely date for at least two of the ceramic vessels. It may be that the description of the soil as ‘argillaceous’ can help account for this discrepancy. As well as being a word that could describe the Lisan marl it would also be apposite for the silt accumulating in the bottom of cisterns. It is noticeable that little pottery was recovered from the pools Loci 110, 117 and 118, and of the handful of coins very few date from before the time of Herod Agrippa. These pools would certainly have needed cleaning out following the earthquake of 31 BCE, but it would appear that they were cleaned again in the time of Agrippa who, as well as repairing the walls of Jerusalem (Antiquities 19.326), may well have renovated some of his border positions (Taxel 2011: 404). The silt from the pools, which in Jericho was generally rich in pottery, could conveniently have been used to backfill loci 1, 2 and 4 (see below). identify the Level 2 floor, a floor that he admits eluded him in L1, that must have existed to correspond with that in L1 and L2 some 50 cm above the original floor, otherwise there would have been an awkward step in the door between L4 and L2. There is no evidence of fallen roof in the accumulation on the Level 1 floor. De Vaux mentions a ‘plaster plinth (approx. 50 cm high)’ (L4 5/3/55) which ran round the room above the level of the bench (pl 132-134). This was less likely a plinth than the original wall plaster that had been preserved by being buried by a backfill about 50 cm deep over which a floor had been created at the same elevation as that in L2. This floor, at about -1.70, would have eliminated the need for the step down in the northern doorway. It is noticeable that some of the published pottery from L4, which probably came from between the top of the bench and the floor, is identical to that found in L1 (compare de Vaux 1953: Fig. 4; 1,2,5,9,14, 16, with Fig. 4; 3,4,6,8,11,13,15) and we can assume that the backfilling occurred at the same time. As with the majority of the ceramics from L1 and L2, it can be generally dated between 31 BCE and 50 CE. However the thin walled cup (de Vaux 1953: Fig. 3: 9) was a form which became squatter after the middle of Herod’s reign c. 15 BCE (Bar-Nathan 2002: 99-100).63 The fact that the level 2 floor, clearly of beaten earth, was not noticed at all in L4 and scarcely in L1 and L2 perhaps indicates that it was not in use for any great length of time. As with loci 1 and 2, the numismatics conflicts with a small proportion of the pottery, as a coin of the procurator Valerius Gratus [121] (15-26 CE) is said to have been found ‘on the (presumably Level 1) floor’.64 The debris that rests on this floor could represent the destruction of 68 CE, but the Level 3 floor is a metre higher which would suggest not only a destruction but a considerable period of abandonment during which roofs partially collapsed allowing wind bourn dust and sand to build up in the rooms. About 90 cm above the ‘Level 2’ floor in Loci 1 and 2, the lowest of three courses of a wall survived which was a rebuild of, and based on, the remnants of the north wall. There was no door in this wall giving access to the north (pls 139, 147) as there had been in the earlier wall. No corresponding floor was noticed. Indeed when Harding wrote ‘almost on the present surface was a floor level and cross walls which seem to belong to a period of medieval Arab occupation’ (Harding 1952: 105) he was certainly referring to this, and a similar wall north of L4 (de Vaux 1953: 92).62 The elevation of a newly constructed threshold in the eastern wall allows us to reconstruct a ‘Level 3’ floor at about – 65 (de Vaux 1953: Fig. 1) in contrast to that of Level 1 at – 2.30, and level 2 at – 1.70. In Level 3 the room was divided by a north-south wall to give a small room L149 (c. 2m x 4m) in the east and a larger room L150 (6m x 4m) in the west. It ran between the walls rebuilt above the north walls of L1 and L4 (pls 127-9, 133) and had a floor surface at about – 115 (de Vaux 1954: Pl VI). At the western end of the western room a bench, perhaps a sleeping platform, was built. Sealed in by this structure was a coin [118] of the First Revolt. How these two rooms were accessed is unclear; the door from L16 shown on the plans (Humbert and Chambon 1994: Plan XI) is unlikely. Locus 4. The room to the north of Loci 1 & 2. The room was entered from a central door in the north and gave access to L1 through a door in its south-west corner and into L2 through the central door in the south. It originally had a plastered floor and a low bench (20 cm high) running from the eastern side of the northern door around the north-eastern, eastern and south-eastern walls to the eastern side of the southern door. The floor was raised slightly and re-plastered, clearly at the time that the door into L1 was blocked because the re-plastering relates to a bench that was added along the western side of the room, in part in front of the blockage (L4 22-23/3/55). De Vaux notes that ‘much pottery was found between the level of the top of the bench and the floor; pottery also lay on the bench’ (L4 6-9/12/51). It would appear that he failed to As there is no mention of mud-brick, which could have fallen from a second storey, within the debris covering the Level 2 floors in either Loci 1, 2 or 4 it is probable that these rooms were single storey. 63 Seventy five of these cups were found in L89 (de Vaux 1956: Fig. 8,9) together with 204 of the wide, shallow bowls, which first appear early in Herod’s reign (Bar-Nathan 2002: 90). Both forms were found in L1 and L4 (cf de Vaux 1953: Fig. 4; 1-9, 11, 14 with de Vaux 1956: Fig. 2; 5, 6, 7). 64 One does wonder about the archaeological methodology. From the field notes one learns that the western part of L4 was excavated in 1951, the eastern part in 1953. On pls 131-2 it can be seen that a narrow tunnel was dug along the south east wall, presumably in 1953, from the top of the preserved wall down to the floor on which the bench was built, not the best technique for observing intermediate beaten earth floors. If the remaining soil was then removed in vertical, rather than horizontal sections there would be many opportunities for coins to be found far below their level of deposition. 62 The ‘Arabic sherds of the 9/10th century’ mentioned by de Vaux were probably Nabateaean cream ware. 46 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - HEROD AND POST HEROD OCCUPATION as being ‘on the floor’ came from the Level 2 surface.65 Thus the inkwells must be associated with the level 2 occupation of the ground floor rooms and not with the debris of the plastered elements which resulted from the destruction of a second storey at the end of Level 2.66 Unfortunately the presence of the plastered elements seems to have overwhelmed the excavators and there are few observations about the layer between the original Level 1 floor and the Level 2 surface above which the elements were found. It is not even clear which surface the ‘mat’ (pl 119; de Vaux 1973: 29, L16 18/3/53) at the southern end of the room was on although it was probably that of Level 2.67 Locus 30 (below Loci 15, 16 and 20). Although the notes from this locus, and the manner in which it was excavated, are muddled there are hints that a similar situation occurred here as in the three loci to its west. In L15 (a ‘Period III’ locus, above L30 which was divided into three rooms) there was a poorly paved floor equivalent to the Level 3 floor in L149 (L15 5/3/53). Below it, at an indeterminate depth, was an earlier floor upon which were ‘ashes, burned wood and much pottery, often burned’ (L15 18/3/53), similar to the debris on the Level 2 floor in L2, together with, slightly higher, ‘tumbled bricks’ (L15 17/3/53). Before this Level 2 floor had been reached elsewhere, the ‘plastered elements’, later claimed to be scriptorium benches, had begun to appear (L16 14/3/53) and became the complete focus of attention, and serious stratified excavation ceased as can be seen in the photographs (pls 110-113). It was noted that the plastered elements ‘were not directly on the floor’ (L30 24/3/53 Complementary notes) which, as de Vaux is writing about a ‘stage II’ presumably means a Level 2 floor, and that below them lay ‘charred wood’ and bricks, as had been found on the Level 2 floor in L15. In L16 the plastered elements were found immediately under the ‘upper’, presumably Period III, floor (L16 14/3/53) but somewhat lower down were ‘collapsed bricks’ (L16 17/3/53). Similarly ‘collapsed bricks’ were found in L20 (L20 17/3/53) so this burnt and bricky debris, clearly indicative of a second storey, existed throughout L30 fallen on to a Level 2 surface, mainly in the east, which initially led de Vaux to conjecture that only the upper part of the eastern wall had collapsed (entries for 17/3/53 for loci 15, 16 and 20) although a week later he considered that bricks had fallen from all around the room (L30 24/3/23, complementary notes). What is more difficult to explain is how the ‘plastered elements’ were consistently found at a higher elevation than the bricks which must have destroyed the ceiling through which they fell. It would appear that the ‘plastered elements’ could not have been standing above that ceiling or they would have been found intermixed with the fallen bricks. Could it be that the ‘plastered elements’ were attached in some way to the western face of the two storey western wall at the elevation of the flat roof covering the single storey rooms of the western wing (Loci 1, 2 and 4)? One can then envisage a delay between the collapse of the eastern wall and an eventual collapse of the western wall and its attached elements. No coins are recorded from the Level 1 floor in L30. However several coins apparently from the Level 2 floor were found, the latest of which was from the second year of the First Revolt. It is thus possible to suggest the following dating for the three levels in the western wing of the main building:- Level 1, 20 BCE - 43 CE . Level 2, 43 - 68 CE. Level 3, 73-125+ CE. This is, of course, a simplification as alterations would not necessarily have occurred all over the site at the same time. 5.2 Other Loci in the ‘Main’ Building Locus 36 (below Loci 22 and 31). Locus 36 was part of the Hasmonean structure relating to the pools in L34 and to the Hasmonean cooking pot and installation beneath the lowest floor of L30. At that time it was part of an open courtyard which could be accessed from the west and from the pottery kilns (L66) to the east, and from which there was a door through the southern wall into a building, possibly a storeroom, or the potter’s workshop, that occupied the area later taken up by the pools Loci 5558 (Plan 8). The Hasmonean wall south of L34 and L36 was probably damaged during the earthquake of 31 BCE as was W1 parallel to it and further south. It is difficult from the muddled notes, the sparse photographic record (pls 60-62) and the plans (Humbert and Chambon 1994: Plan X) to understand all the developments that occurred during the Herodian period. A wall which was part of the retaining wall for the construction of L55/58 was built against the southern face of the, now damaged, original southern wall, blocking the doorway in it. Immediately east of the doorway the new wall was built based on what survived of the original wall though narrower than it (Plan 9). Later de Vaux wrote that ‘two inkwells were found among the debris’ (de Vaux 1973: 29-30 – he is referring here to the debris of the plastered elements) but at the time of excavation he described them both as coming from ‘on the upper floor’ (L30, objects of L30 middle). Presumably he did not mean the Level 3 floor because L30 started below that floor. He must then have been referring to the Level 2 floor, noted quite decisively in L15 (which was above the ‘middle’ of L30), and we must also assume that the coins, including one [438] from the second year of the First Revolt, which are also described The wall that delineated L36 to the north abutted the eastern wall of L30, so must post date it at least technically. A simple lean-to shed was thus created Notice that they were registered together - 436, Inkwell (‘on the upper floor’); 438, coin of first revolt (‘locus 30 in the middle on the floor’) – and one hopes that they were not excavating two different floors in the same small locus at the same time! A cooking pot 477 (de Vaux 1954: Fig. 4:15), that can not be dated more precisely than Herod – end 1st Century CE, apparently came from this floor. 66 This was noted already by the Donceels (Donceel and DonceelVoûte 1994: 31). 67 From the photograph it looks more like debris in the lair of an animal such as a porcupine or possibly hyrax, which, if true, would indicate a period of human abandonment. 65 47 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS against the southern wall of the courtyard, with access from the reduced courtyard via a door that related to a wall, and some other structures (L35), built to the north of L34 and partially blocking the entrance in the east from where the pottery kilns had stood. The pools of L34 were adapted to contain the large cylindrical jar found in one of them (pls 69, 70). No clear floor levels associated with these changes were described, although it would be reasonable to assume that the Hasmonean floor level (3.40) was raised by 10-25cm (Plan 9). At some stage a wall was built to the east of L36 barring access from L32. It was related to the narrowing of the doorway leading into L32 from the north, contemporary with the construction of the mud-brick walls of L33 (L32 9/4/53). The only means of access to L36 would then have been from the south (Plan 10). This was a radical alteration. It indicates that the pool L56/58 was already filled in to allow access to this new door. Moreover the small room(s) created probably served as a unit together with a room situated in the partially filled in pool L58.The pottery from these pools has not been published, but the numismatic evidence would suggest that they were cancelled around the time of Herod Agrippa I (L56, coins 980, 995 of Agrippa I). As there were no signs of stairs leading down from the south into L36 the floor level must have been raised by at least 1 m to about -2.20. There is no mention of any destruction debris so presumably it was a deliberate back-fill similar to, and probably contemporary with, that in Loci 1, 2, and 4. The entrance door is shown on the plans to be in the south-east corner (Plan 10) and it may be that the plastered threshold, later blocked, is visible in pls 60-61. No floor was recorded as directly relating to this door but it must have been the lower of two floors, 20 cm apart, noted in L22 (L22 21/3/53). On this floor was based a narrow mud-brick wall that divided the space into L22, in the west, and L31 in the east and associated with it were found a coin [556] of the procurator under Tiberius and one [563] of Agrippa. again to L32. Just above this floor, which was close to the ground surface, were found four large coins (L22 9/3/53) of which two were of Caesarea under Nero and one from Dora dated 67-68. The depth of the accumulation above the ‘Period III’ surface may indicate that this floor relates to a 2nd century CE (Bar Kochba revolt?) or even later occupation. Locus 33, together with Locus 32. It was noted above that the construction of L33, with mud-brick walls on a single course of stones, was related to the addition of a wall east of L36 which was then back-filled i.e. to ‘Level 2’. Only two floor surfaces were recorded, about 25 cm apart (L33 11/4/53), but it is impossible from the paucity of data to determine whether both floors belong to ‘Level 2’. The lower floor (- 3.30), covering a layer of ash (L32 9/4/53) apparently related to a similar surface in L32 so that there was a two roomed unit. However the upper floor (- 3.05) seems to have related to a blocking of the door into L32 (at which time, if not from the beginning, there must have been an entrance to L33 via a door in its north-east corner). Following the blocking, L32 must have been back-filled to the same level as the ‘Level 2’ or ‘Level 3’ floors in L22/31 (L36). The walls separating L32 from L22/31 in the west and L34 in the east were demolished allowing access from L22/31 through into the area covering the pools in L34 which must also have been backfilled68 to create the ‘upper level’ from which came a coin [590] of Ashkelon, dated 72/73 CE (but also one [605] of the 4th century CE). It is not clear to which of the floor levels in L22/31 this relates but as a ‘Judea Capta’ coin [595] and one [596] of Dora (66-67 CE) came from an ‘upper’ level it is reasonable to suppose that there was a ‘Level 3’ surface in L32. Locus 35. This room was at the bottom of the slope over which the ‘main building’ was built. Its earliest recorded floor was 35 cm lower than that in L33 immediately to the west; (overall the contemporary floors were at, from west to east, L4 - 2.30, L30 - 2.40, L23/L33 – 3.30, L35 – 3.65). Although de Vaux noted only two floor levels (L35 31/3/53) the coins are allocated to an ‘upper’ and an ‘intermediate’ level so presumably there must have been a lower level as well. Both the upper and the intermediate level are dated numismatically to the time of the First Revolt or later. A rather flimsy pillar (Humbert and Chambon 1994: pls 76, 77) was added at some time to support the roof.69 It is far from clear whether the inkwell found in L31 was assosciated with the backfilling of Agrippa or with the post 68 CE occupation (cf Mizzi 2009: 144). The upper of the two floors, of beaten earth and chalk, relates to a new entrance in the south-west (Plan 11) with a threshold of ashlars (pls 56-57). There was a fireplace just inside the new door on the west, and in the west wall a rectangular cupboard in which was a small jar, perhaps 364 which was published as belonging to ‘Period III’ (de Vaux 1954: Fig. 6.4). It is not a common jar form and has a suggested date range of 15 BCE – 73 CE (Bar-Nathan 2006a: 70 M-SJ19A). Coins [496, 497] of Agrippa I were found with this floor but the insertion of the entirely new entrance indicates that it represents a ‘Period III’ reoccupation following the 68 CE destruction. 68 Although on de Vaux’s plans the wall to the north of L34 is shown to be demolished in Period III (de Vaux 1954: Plan VI) it must have remained to act as a retaining wall for the ‘upper’ floor which would have been at c. – 1.80 (c.f. the floors in L22 and L31) whereas the floor in L35 to the north is shown as – 3.15. A door in the south east of L34 is shown in Period II (Humbert and Chambon 1994: Plan X) but this must have been very close to the surface at the start of excavations (see pl 74) and is far from certain. If it did exist it would more likely relate to Period III than Period II. 69 This is unlikely to have been a support for a spiral staircase (contra Magness 2002: 122; and cf Netzer 2001: Ills 238-9). No stone steps were found and, as any steps would have to have been c. 1.20 m in width, such substantial treads in wood would have been hard to source. More than 75 cm above this floor (L22 12/3/53) was a further floor associated with a rebuild of the northern and southern walls (L22 11/3/53). Associated either with this, or the uppermost of the two previous floors, was the partial demolition of the east wall allowing access once 48 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - HEROD AND POST HEROD OCCUPATION Fig 11. Channel through wall in south west of L86, looking south west. 5.2(i) Loci in the northeast of the ‘Main’ Building. close to the ground surface when excavations began. A lamp, in which there was a coin [1438] of Tyre, dated 53 CE, and one of Caesarea under Nero (68 CE), was found in L40 (de Vaux 1973: 37 fn 3) at an elevation that probably relates to the platform in L6. The three levels noted in the south-west of this area are not so clear in these rooms, moreover these rooms were founded in the Hasmonean period. The earliest floor in L38/41, which may have continued in use into the early Herodian period, was of plaster (at least in the west, L41 14/3/56) on which was a ‘very hard layer of ashes’ indicating either that this was a cooking area or, possibly, that it was abandoned following a fire. Some 20 cm above the plastered surface was a second floor, associated with the blocking of the door in the north wall, and on which were three ovens (L38/41 9/4/53). Three coins [741, 742, 743] of the First Revolt were found on this floor. Any ‘Level 3’ floor was probably so close to the surface that it was not distinguished. Indeed what was identified as a ‘wall’ but was later decided to be nothing more than ‘collapsed stones’ (L38/41 4-5/4/53) may well have been the remnants of a Level 3 wall. At this ‘upper level’ was recovered a coin [714] possibly of the Second Revolt. The ‘Level 3’ floors and alterations equate to de Vaux’s Period III. It is clear that despite his claims that ‘the new occupants of Period III cleared out part of the ruins’ (de Vaux 1973: 37, 43) no such thing happened. There would have been considerable physical difficulty to ‘clear(ing) out the rooms on the southern side’ and throwing the debris ‘into neighbouring cisterns’. There are no signs that the wall separating the southern rooms from the cistern L56/58 was demolished to floor level and the floors to the north of that wall on which debris fell were, in any case, at a lower elevation than the top of the cistern. Moreover, the introduction of a doorway into the southern wall of L22 in the period before Period III (see above) indicates that the pools were already largely filled in, because access solely on the top of the aqueduct channel along the side of a deep pool would have been perilous, particularly at night by candlelight. The numismatic evidence, whilst far from conclusive, does support a filling in of the cistern after Agrippa (two coins [980, 995] of Agrippa in the ‘upper level’ of L56 and one [1029] in the ‘upper level of L55) but before the First Revolt; (coin [988] of the Revolt found close to the top of pool L55). The Level 3 floors are usually nearly 1 m above the previous surfaces and thus there are no signs of a general clearance which is as unlikely as one following the earthquake of 31 BCE. In rooms to the east (Loci 39, 47, 6/40) only two floors were recorded about 1 m apart which are not easy to relate to the floors in L38/41. A pair of installations, probably domestic, in the north-east corner of the upper floor in L39 seems to relate to the Level 2 floor in L38. Doors were blocked and others opened, especially to L51 in the south, but no floors relating to them were recorded there and the paucity of data makes it impossible to reconstruct the developments here. Against the east walls of both L6 and L47 were built what are referred to as ‘reinforcement walls’ (L47 16/3/54, and see Humbert 2003: 433 and Fig. 6) but which were probably sleeping platforms as found in L150 and L16 in ‘Period 3’ (cf Taylor 2006: 135). The platform in L6 was said to be based on a thick layer of ash. No floors are specifically related to these platforms and they were probably very 5.3 Some Loci outside the ‘Main’ Building. Loci 77 & 86/87/89. It is difficult from the sparse recording to reconstruct the developments here. As discussed above, L77 was built in the Herodian period as 49 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Perhaps too many had been stacked on to inadequate shelving which collapsed under the considerable weight. Rather than tidy up the mess it was buried beneath a low platform supported by a narrow retaining wall.72 a long east-west storeroom with two entrances in the southern wall from the southern plateaux. An annexe L86 was added at right angles centred on the western of the two southern doors. Unusually this room was built with two mud-brick pillars, one centrally placed, the other attached to the southern wall. Similar pillars were built at the eastern end of L77 but, according to de Vaux, these were built to a secondary floor (de Vaux 1973: 26. L77 24/3/54). To de Vaux these pillars suggested a shortage of long palm trunks for roofing (de Vaux 1973: 26). I suggest that they were inserted to support the roof weakened or damaged in some way. It may be, however, that the attached pilaster, at least in L86, had another function. The southern end of L86 does, however, appear to have become increasingly unsound structurally probably due to the construction of the large cistern L91 which required a considerable weight of back-fill to be poured against the western face of W3. The south and east walls, W4 and W5, and possible the western W3, were buttressed externally and the central pillar was inserted to support what had become an unstable roof. The southern end was eventually sealed off by a partition wall built to the central pillar (Humbert and Chambon 1994: Plan XXV, Periode II). In the field today, a socket can be seen in W3 some 30 cm above the floor in the south-west corner of L86 (Fig. 11). To the best of my knowledge it has not been recorded before.70 It may have been plastered over by the final phase excavated by de Vaux, although the lime plaster visible in pictures from the time does not seem to cover the area of the socket (Humbert and Chambon 1994: pl 336). Although this socket resembles the channel through the wall in L77 (ibid: pl 327) it is unlikely to be for water (contra Wagemakers and Taylor) as W3 predates the building of the large cistern L91(the retaining wall W6, which was an integral part of the construction of that pool, only abuts the junction of W 3 and W4).71 Moreover it would have been at a considerably lower depth than the top of that pool. At a stretch, it may have related to an otherwise unknown water system that predated L 91, but this seems extremely unlikely, particularly as it would mean that water debouched into the room. The northern part of L86 continued as a reduced annexe to L77. Humbert identifies a secondary floor (ibid: pl 332; Humbert 2006: 34) above the original plaster floor. No comparable floor is discernable in L77 so a mud-brick threshold was inserted in the doorway between the two rooms to retain the higher floor in L86. The main Level 2 floor would have been at the same elevation, and contemporary with, the upper floor in L77 which was about 40 cm above the original (see scale in pl 326) and might date to the time of Agrippa (L77 coin 1585). A likely time would be after the earthquake of 48 CE which may have weakened part of the roof of L77 necessitating the inserting of mud-brick supporting pillars and caused some of the 40 cm of deposit on the original floor. The contemporary floor in L86 is also 40 cm above the original floor (see scale in pl 323).73 During the life time of this floor the door between L77 and L86 was completely blocked,74 and the space between it and the partition wall in L86 was back-filled.75 A new two roomed unit was created, with an entrance from the south-east76 for which a floor was recorded in the south (L89 18 & 22/3/54) and to which level the backfilling in the north would have reached. A door was breached in the partition wall simply by removing the upper courses of the mud-bricks of the pillar incorporated in it leaving the rest as a threshold (pls 330 and 332).77 As three coins of Porcius Festus (L86 coins 1424-5, 1430) dateable to 59 It is possible that it was a socket for a beam, supporting low shelving across the southern end of the room, integrated in some way with the pilaster. Unfortunately the pilaster was ‘restored with a new coat of plaster’ upon excavation (Humbert 2006: 32) so we have no idea if originally it showed signs of supporting shelving. It is in the area of this proposed shelving that many vessels were found (Humbert and Chambon 1994: pls 336, 338-9). 70 It is not mentioned in a recent, comprehensive study of these loci (Humbert 2006), which is odd as it can clearly be seen in a photograph from over thirty years before http://www1.kueichstaett.de/KTF/qumran/bilder/abfrage/loc89.jpg. (Since this was written an article has appeared discussing the socket and these loci (Wagemakers and Taylor 2011). I will insert a few comments pertinent to it. 71 It is certain that L91 was technically built later than L86, but how much later is a moot point. For the ‘outlet’ in L86 to be fed by the short lived pool in L88 (thus Wagemakers and Taylor 2011: 146-8) they would have to have been planned together though the pool is technically later. I considered this possibility in 2007 but ultimately rejected it because the ‘outlet’ is located south of where the pool stood – in fact below the buttress - rather than directly beneath it which would appear more logical, and because no outlet was noted in the pool at the time of excavation. As all traces of the pool have disappeared it would be easy to do a small probe to see if an outlet ever existed. Just as the outlet in W1 predates the building of L77 and either fed L68 or irrigated an area south of W1 (or both) I would have preferred to see the ‘outlet’ in W3 predating the construction of L86 and serving a similar purpose. However it did appear that W3 and W4 were bonded and a connection between L83 and the ‘outlet’ seemed implausible 72 The supposed door at the western end of this wall (L89 22/3/54 and pl. 338) is illusorily. 73 There is no sign of three steps in pl 323 (contra Wagemaker and Taylor 2011: 150) only four courses of the raised mud-brick threshold, partially demolished by the excavators. 74 Although de Vaux thought this blockage belonged to ‘Period III’ (L86 18/3/54) the plastering on the north side of the blockage seems to relate to the floor of Agrippa (pl 322). 75 De Vaux thought this fill was a collapsed upper storey (L86 22/3/54) but there is no reason to suppose that either L77 or L86 had an upper storey. 76 Visible in http://www1.ku-eichstaett.de/KTF/qumran/bilder/ abfrage/loc89.jpg 77 Contra Humbert who prefers to see instead the, admittedly slight, evidence for a doorway as ‘proof that some partition wall stones ... were removed by workers to examine the top of the cube’ (Humbert 2006: 32). He goes on to use this as a plank in ‘a possible worship interpretation’. I prefer a more pragmatic interpretation but as all this was close below the surface at which excavations commenced, and, moreover, the partition wall was rapidly demolished, there will never be certainty. 50 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN - HEROD AND POST HEROD OCCUPATION CE were found in the fill it can be attributed to ‘refugees’ arriving in Qumran during the troubled times leading to the First Revolt. Just as, at Masada, the ‘Zealots’ carved out family units from the existing buildings so in Qumran, on a smaller scale, refugees created discreet units such as L86/89, Loci 1,2, and 4, L36 and above L58, and the various caves on the plateau. probes below floors were done haphazardly and with little attempt to understand how the floors related to the walls (ibid: pl 328). The designation ‘Locus 30’ was only assigned after excavation had proceeded some depth beneath the Period III floors of Loci 16 and 20, and, in particular, L15. The recording leads to some confusion. An inkwell was registered in L30 on 18/3/53 (Mizzi 2009: 70 fn 217) the day before L30 was, according to the field notes, opened. It was said to come from ‘the upper floor’. On the same day two sherds from L20 were registered as coming from ‘upper level, on the floor’. It might seem that this ‘upper floor’ in L20 must refer to the period III floor but this does not seem to be the case. The evidence from L15 indicates that, in at least part of the locus, a floor had been reached beneath the mudbrick debris that lay below the Period III floor. In my designation this would be a ‘Level 2’ surface sandwhiched between the original floor and that of Period III. It is even possible that the original ‘Level 1’ floor had been reached in a probe in what was still L15. Thus the ‘upper floor’ from which objects were recovered from ‘L20’ would appear to be the same ‘upper floor’ as that refered to in L30, which would, in fact, be the same ‘Level 2’ surface solely mentioned in the notes to L15. Any later Period III floors would have been eroded before excavation (ibid: pl 331). One suspects that such a floor would have survived at the western end of L77 but none was recorded. According to de Vaux a considerable alteration was made to the water system in Period III whereby the large cistern L91 was filled in and a ‘bypass’ channel was built over the fill and through the remains of L86 and south of L77 to reconnect with the earlier channel near L72 from where it filled the large cistern L71. However, if the numismatic evidence is correct, L91 was not filled in until the 4th century CE, as a coin [1563] dated to 330335 CE was found in the fill and one [1583] thought to be of the 3rd century CE was found ‘close to the floor’. It is noticeable that two 4th century CE coins were found near to L91; one [1462] from L88 to the east of L91 and one [1522] from L96 west of the pool. Another 4 th century coin [1147] was found in the small channel leading into L68 which suggest that it was operating together with the channel bypass. It is possible that, if indeed the original channel was no longer functioning in Period III,78 it was replaced (or augmented?) by the short stretch of covered channel running above L52, and not much discussed by de Vaux, which probably drained the roofs of the northern rooms of the ‘Main’ building into the cistern L71 during Period III. Excavation often seems to have progressed not horizontally from upper to lower levels but vertically from the side (pls 110, 329) and even following tunnelling along a wall thereby breaking any connection between floors and the wall (pls 131-2). This method of working could easily lead to material, particularly small objects such as coins, which really related to an upper surface, being recorded as coming from a lower one to which they had dropped during excavation. As only large sherds were saved it seems very likely that the gradual build-up of beaten earth floors, which normally will only have small fragmentary sherds associated with them, were not noticed or recorded. It would be expected that a floor build-up would have occurred, for example, on the stone floor, presumably originally plastered, of L1 and L2, and would have been noticeable before the 50 cm of backfill covered it. Should we accept that no such buildups occurred and deduce from that that occupation was short lived or sporadic; or did beaten earth floors exist but were not noticed by (or of no interest to) the excavators? 5.4 A brief note on the archaeological methodology, and its possible implications. It has been noted that beaten earth floors, such as those in L4 and L86, were not noticed during excavation. Retaining sections to better understand accumulations and floors and how they related to walls does not appear to have been a high priority. The whole of large loci such as L77 (pl 329), and L30 (pls 110, 112-3) seem to have been excavated with out leaving any sections at all and 78 The discovery of 26 coins of the First Revolt at the bottom of the settling tank L83 does not necessarily indicate that the tank was filled in in Period III as claimed by de Vaux (de Vaux 1973: 43). The coins could as well have been thrown into the water in extremis and settled into the silt at the bottom of a still active ‘decantation’ basin. 51 6. What was the character of the industrial activity at Qumran? The shepherds would have been well aware of the cistern at Qumran, at the foot of a path up to the Buqe’ah, from where they could conveniently trade dairy products and meat with the villagers of both Jericho and Ein Gedi. Two ostraca found at Ein Gedi, dated to the fourth century BCE, are evidence that trans-humance was certainly established by the Persian period (Cross 2007). They are tallies for the preparation of lamb hides. As sheep would not flourish in the intense heat of an Ein Gedi summer, it is probable that these lambs had been born to sheep coming down from the hill country. It is even conceivable that because, as noted by Cross, ‘the manufacture of leather was odorous and despised so that isolation was desirable’ the work on the skins had been carried out at Qumran. Winter temperatures near the Dead Sea are mild, so mild that, according to Josephus, ‘the people are clothed in linen even when snow covers the rest of Judea’ (War 4:473). The warmth encourages plants to grow rapidly after any significant fall of rain so that they can complete their life cycle before the searing heat of summer (for pictures see Roitman 1997: frontispiece; Netzer 2001: Pl XVI (top); 2004: Pl XIV; Yadin 1966: 34). The movement of animals, particularly sheep, down from the hill country to exploit the winter grazing on the Dead Sea littoral was an essential part of the local economy and had been since time immemorial. ‘This cycle assures the animals of a favourable climate and readily available grass, fodder, and water in each season of the year’ (HarEl 2000: 13). Moreover a short period during which their diet would consist largely of the sort of salty plants found near the Dead Sea was essential to combat a serious intestinal disorder common in flocks (Bailey and Danin 1981: 149). Although de Vaux saw flocks pass by during the course of his excavations in the winter months (de Vaux 1973: 84) and Milik was aware that the Buqe’ah ‘serves as marginal grazing land during certain seasons of the year’ (Cross and Milik 1956: 5), and that around Qumran ‘in the spring... a little vegetation appears and then the semi-nomadic Ta’amireh tribe brings its sheep and goats down into the valley’ (Milik 1959: 11), neither seem to have considered whether this was continuing a millennia-old exploitation of rare resources which would have been followed both before and after the Second Temple period. In Jericho, the Hasmoneans began to extend intensive, irrigated agriculture into areas that previously would have been waste land freely available for grazing by seasonal flocks. They, too, would have been aware of the cistern at Qumran and would have made sure that its water was under their control, so that flocks could flourish reasonably close to the centre of population but far from the fields. Flocks of sheep and, in particular, omnivorous goats would not have been welcome in areas where any crops, but in particular balsam, was intensively grown. Balsam was a low bush or shrub (Fischer et al. 2000: 121; Heppner and Taylor 2004: 37) and would have been as vulnerable to rampaging flocks of hungry animals as would crops of vegetables or grain. As a second grade, but still valuable, oil could be extracted from balsam’s leaves, bark and twigs, flocks would have been encouraged to forage well away from cultivated areas. After the construction of a water cistern in Qumran in the Iron Age, shepherds would have gravitated to it from the Buqe’ah during the seasonal trans-humance particularly for fresh water for themselves, and perhaps even for their flocks although they would have utilised the saline water of Ein Feshka. Even after the abandonment of the Iron Age fort there is no reason to suppose that the cistern ceased to collect water, or that the shepherds would not periodically have cleansed it of erosive dirt. Indeed once irrigated agriculture developed in Ein Gedi during the late Iron Age (Tel Goren, Stratum V),, for this first time in the historical period, flocks, which previously would have wandered freely in the oasis during the winter months, would have been encouraged to find grazing well away from the crops cultivated there, already vulnerable to predation from the wild ibex that frequent the region. It is probable that a number of small cells, scattered along a terrace about 200 m above the village, which were excavated by Hirschfeld and identified by him as Essene hermitages (Hirschfeld 2000, 2004b: 233-238), were, in fact, small shelters where the young men of the village would dwell during the growing season. As well as watering and weeding the crops on the upper terraces, they would endeavour to ward off the ibex (after whom the oasis is named) and the hyrax which could consume the potential harvest far quicker than villagers could race up the steep terraces to chase predatory animals away. Although normally the flocks ‘returned to the mountains in summer for the birthing and shearing seasons’ (Har-El 2000: 446) the expansion of labour intensive agriculture in both Jericho and Ein Gedi would have necessitated a considerable influx of labour which, in turn, would have greatly added to the potential market for whatever the shepherds could produce. It would have been in the interest of the Royal Palace at Jericho, with its dependent workforce, to ensure that some lambing and shearing took place nearby rather than in the hills because ‘the fleece provided the only fabric for winter clothing ... the animals’ wool was used to weave rugs, tents … and sacks for the transportation of grain; the hides were used to make harnesses and shoes while the skins of lambs and kids provided vessels for water ... oil and wine... as well as the parchment on which the Holy Scriptures were written...’ (Har-El 2000: 14). Sheep would not flourish in the heat of a Qumran summer – they suffer heat stress in temperatures above 25° C, a normal summer night time temperature at Qumran. The mature date plantations of Jericho, and the smaller ones established at Ein Feshka, Ein Ghazal and Ein Tannur, would have supplied some late grazing and shade for a limited number of flocks before they had to return to the hills. Moreover the available grazing could have been extended locally by 52 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN WHAT WAS THE CHARACTER OF THE INDUSTRIAL ACTIVITY AT QUMRAN? Eventually Jannaeus built a technically demanding aqueduct to bring spring water from Na’aran which was later, probably in Herod’s time, supplemented by a channel from Ein el-‘Auja some 13 km north of Jericho (Netzer and Garbrecht 2002; Netzer 2004: 6-8). Qumran, 13 km to the south of Jericho, already collected water in an Iron Age cistern so it would be simple to increase the quantity gathered and its utilisation would minimise the need to use the ‘expensive’ water supply in Jericho for other than domestic or lucrative agricultural purposes. selective burning of the reeds around Ein Feshka. Masterman noted that ‘places are cleared by burning the reeds, the young sprouting shoots that spring up affording excellent fodder for cattle’ (Masterman 1902: 165). He recorded the reeds being burnt at the end of August and showing ‘young green sprouts over a foot high’ which are ‘brilliantly green and fresh’ by late October (Masterman 1904a: 95; 1905: 159). In 1903 he noted that the reeds and succulent shrubs were in bloom by late October, in full bloom by the end of December and ‘yellow and dry’ by early February, in contrast to the year previous, which had been wetter, when ‘the plains of Jericho (were) still covered with flowers’, even if a good deal withered’ on 21 March (Masterman 1904a: 91). 6.1 Potential uses of the ‘industrial area’ and the water resources. A number of industries would have been able to source necessary raw materials at, or near to, Qumran in the winter season. That some processes took place at Qumran has been considered by various scholars, beginning as early as de Vaux, but, generally, they have been viewed within the assumption that the site was a permanent ‘sectarian’ settlement. Consequently they have been regarded as a minor activity and their unpleasant impact on the supposed ‘monastic’ environment underplayed or ignored. Magness gave little thought to the industrial zone and the processes that might have been conducted there, merely reiterating de Vaux in saying that the rooms ‘included storerooms, industrial installations and workshops’ and the plastered floor of L115/6 ‘seems to have been used for some industry requiring water’. Other functions ‘are unknown’ (Magness 2002: 53). Elsewhere she claims that ‘the types of the vessels found at the site ... reflect the activities carried out there’ (Magness 2004: 4), but makes no attempt to discern the activities carried out in the industrial area. Hirschfeld realized that Qumran was ‘a production centre for commodities of commercial value’, but saw the industrial area as largely restricted to processing ‘the valuable perfumes and ointments produced from balsam. (Hirschfeld 2004b: 138-42), a conclusion previously reached by the Donceels (Donceel and Donceel-Voûte 1994: 26-7). Zangenberg, in a perceptive article, realized the likely importance of various industries at Qumran but does not stress their seasonality (Zangenburg 2004). He had noticed that ‘early in the year, in January and February, Bedawin descend into this part of the plain (near ‘Ain el-Feshka) and flocks of goats and sheep and also camels may be seen on all hands. The Bedawin at this time inhabit caves in the hills around. The ‘Ain Feshka oasis itself has been tenanted for some eight months now by two men (natives of Abu Dis) who are in charge of a large herd of cattle, belonging to the Sultan, which thrive in the reeds. The men collect and dry the rushes, which are sold for basket work’ (Masterman 1902: 166); ‘I have seen long lines of these grasses drying but have only once come across the people themselves’ (Masterman 1904a: 91). In following years, he noted, on the way down from the Buqe’ah ‘a large flock of sheep and goats which had apparently been washed in the Pool’ at Ein Feshka (Masterman 1904b: 281) and Bedouin ‘descending the rocks above ‘Ain Feshka to water their flocks’ (Masterman 1905: 159). Masterman’s observations show that the local resources of the Dead Sea littoral were still being exploited in the early twentieth century, not by permanent residents, but by short seasonal visits from people and flocks more usually based in the hill country. ‘Two sheepfolds were built between Caves 11 and 1’ (Safrai and Eshel 2000: 231); these are not dated and could have been used over a long period of time. The concept that permanent inhabitants of Qumran ‘undoubtedly raised herds of sheep, goat and cattle’ (Magness 2002: 21) is mistaken. Whilst a handful of goats and camels might have survived year round, they would soon have decimated the sparse grazing available in the summer time, which would consist mainly of the spiny acacia bush. In reality it was the presence of the seasonal flocks that made it possible for seasonal workers, sustained on their dairy produce, to exploit the water that could be gathered at Qumran. 6.1(i) Industries associated with Sheep and Goats. Leather Production When the flocks first arrived in the winter, superannuated animals would have been slaughtered to conserve the limited grazing. The meat would have been consumed (most likely in the Royal Palaces at Jericho) and the hides processed. Following lambing, at the end of the season, surplus male lambs and kids would also have been slaughtered and finer leather made from their hides. The early Hasmoneans had been in need of a steady and reliable income – John Hyrcanus had had to plunder the tomb of King David to pay his mercenaries (Ant 13.249) – and the plains west of Jericho, once irrigated with water brought by aqueduct from Wadi Qelt, were able to produce high value crops of balsam and dates. Their agricultural endeavours flourished but, if they wished to expand their productivity and increase their income, they had to look for ways to bring more land into production. The production of leather was a malodorous and flyridden process. Not only was there the blood from the butchered beasts but the hair was removed from the hides by being soaked in dung and urine. The mid-second century CE author Artemidorus, a native of Ephesus, 53 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS wrote: ‘The tannery is an irritant to everyone. Since the tanner has to handle animal corpses, he has to live far out of town, and the vile odour points him out even when hiding... The vultures are companion to the potters and the tanners since they live far from towns and the latter handle dead bodies’ (Interpretation of Dreams I: 54; 2:20). In the Mishnah it was written that both pottery kilns and tanneries should be at least 50 cubits from a town (Baba Bathra 1:10, 2:9); Qumran was considerably further than that from the Palaces of Jericho! Although the tanner was viewed negatively because his handling of dead animals, and the use of urine and dung in the preparation of hides, would cause him to be ritually unclean (Jeremias 1969: 301-12), nevertheless his products, from water skins to sandals and phylacteries, were in demand and the specialist preparation of parchment on which sacred texts could be written was regarded as an honourable profession. would be to prevent putrefaction if the skins could not be further treated immediately. The hides would then need re-hydrating by soaking in a warm aqueous solution either containing faeces or else vegetable matter such as barley, flour or cereal husks, or even wine or beer, so that fermentation would soften the leather (Reed 1972: 60). The use of alum (potash – available either from the Dead Sea or by burning vegetable matter), introduced to this aqueous solution to produce a soft flexible ‘leather’ (Reed 1975: 32) may already have been known as it was in later periods. The ‘parchment’ was then hung over supports (the plastered elements from L30?) to dry in the air for some weeks before being stretched in individual wooden frames. As there was only a limited supply of suitable wood for such frames at Qumran the partially prepared parchment may have been taken elsewhere for this final stretching process. The soaking processes could have taken place in the shallow pools in L121 or, in the Herodian period in the enigmatic pools L55 and L57. Beating the hides could have been carried out on the reinforced floor of L115/6. It is possible that an ‘iron sickle’ found in L126 (object 2173) was in fact a de-hairing knife (Roitman 1997: 33 (Hebrew); Hirschfeld 2004b: 141, Fig 78). Another sickle, or possible de-hairing knife, was found together with several unidentified bronze and iron tools in L52. De Vaux did suggest that a tannery, perhaps for the preparation of parchment, may have existed at Ein Feshka (de Vaux 1959: 230-237; 1973: 70-82) but scientific analyses threw doubt on this (Zeuner 1960: 33-36; Poole and Read 1961: 114-123). As the installations at Ein Feshka are far more likely to have been a date wine press (Netzer 2005: 97-100) this is not surprising. The exact tanning processes performed in the Second Temple period are not certain, but later writings give some indications (Poole and Reed 1972; Reed 1972: 92-107, 1975; Forbes 1957: 36-9). Initially the hair and flesh had to be removed from the hide. It was first soaked in cold water with an admixture of urine or fermenting plant material, which both cleaned and hydrated it. The hair was then removed ‘manually using a very sharp, twohandled knife with the skin thrown over a wooden beam’ (Reed 1972: 53). At this stage the skin was often beaten to increase its malleability. The final softness and flexibility of the leather was improved by ‘puering’, i.e. immersing the hide in a warm infusion of dung in which bacterial enzymes would loosen the hair roots. Dung was usually sourced from dogs, pigeons or hens (Reed 1972: 55). In 1851 it was estimated that there were between two and three hundred ‘pure’ gatherers collecting dog excrement from the streets of London for use in tanneries, particularly for the processing of goat skins (Mayhew 1851: 142-45). Accumulating sufficient excrement around Qumran might have been more problematic. Seasonal workers probably brought a few chickens with them; pigeon dung could have been collected in the suggested columbarium in the tower, in that suggested in Ein Feshka (Hirschfeld 2004b: 185), and/or brought from that in Jericho (Netzer 2004: 32-34); bat dung was available in some of the nearby caves. Human excrement may have been utilised as a last resort. After puering or drenching, the hides were tanned to preserve them and make them resistant to bacterial attack. Many plants containing tannins were available locally: palm, acacia, sumac and tamarisk (Poole and Reed 1966: 181) to which should be added henna, the biblical ‘camphire’ growing in the vineyards of Ein Gedi (Song of Solomon 1:14). Although Poole and Reed contended that ‘in neither of the two “industrial” quarters has a tannery been recognised’, they accepted that ‘many of the constituent rooms have pits, vats or cobbled floors (suggesting that wet work was carried out there)’. They recognized that water was a ‘valuable commodity’, but concluded that it could not ‘have been spared for tanning purposes’ and that ‘the community would have been too strict to permit’ (Poole and Reed 1972: 151–2) tanning, a conclusion clearly reached not by scientific investigation, as they only did tests in Ein Feshka, but by accepting the prevailing theory of a permanent sectarian community. Perhaps they should have stuck to their scientific guns and concluded that the limited water supply and the nature of the tanning process meant that it was unlikely that any ‘sectarian community’ ever lived at the site, and insisted that analyses were carried out on sediments at Qumran. There is no reason to believe that any strict rules written in the scrolls actually applied to the occupants of Qumran, most of whom would have only been there seasonally precisely to utilize the water that was gathered there. That some of those seasonal workers were involved with the production of leather goods and, apparently, parchment is shown by the fact that in Cave 8, on the marl terrace ‘beaucoup de fines lanières et des languettes Parchment may have been produced by a slightly different process (Reed 1975: 32). Although in the medieval period parchment was not tanned some of those from the Dead Sea certainly were. It is possible that the skins were initial cured by application of salt (easily available by the Dead Sea) but the main benefit of this 54 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN WHAT WAS THE CHARACTER OF THE INDUSTRIAL ACTIVITY AT QUMRAN? de cuir’ (Baillet et al. 1962: 31; Roitman 1997: 13, 34; Broshi and Eshel 1999: 334) were found; and in three caves about two kilometres north of the site ‘leather in various stages of being worked’ was recovered. In one, named by the excavators ‘the Cave of Leather’, was found ‘a large quantity of tanned skins... The skins were in various stages of being worked and some were even being prepared as products such as thick pieces to be used as sandal parts and thin pieces to be used as parchment’. More such skins were found in a nearby cave, X42, and also in X35, in which previous excavators (Patrich and Arubas 1989) had found a balsam oil juglet. (Itah, Kam and Ben-Haim 2002: 169-73). In the IAA’s exhaustive investigation of caves from north of Jericho to south of Ein Feshka, such leather, undergoing processing, was only found in these three caves near Qumran. making, as they still are to this day’ (de Vaux 1973: 73). He went on to conjecture that (in Ein Feshka) ‘the depilation pits and the tanning pits are small and numerous. This makes it possible to treat simultaneously several lots of hides at different stages of preparation. Moreover, since the workers can put the hides in and take them out without going down into the pits it has the further effect of reducing danger from the corrosive fluids in the baths’ (de Vaux 1973: 80–1). How much more would this be true if the ‘pits’ were portable basins. Such basketry will be discussed more fully later. Glue manufacture Most of the meat from slaughtered animals probably finished up in the dining rooms of the Jericho Royal Palaces but some of the offal and poorer cuts may have reached the plates of the labouring men, even of the seasonal workers of Qumran. The bones and odd pieces of the hides (ears, off-cuts etc.) appear to have been boiled up for the production of glue (Hodges 1989: 163) and/or of gelatin. The residue of the glue-making process, sometimes still in the pots in which it had been boiled, was buried around the site to minimize attracting flies and vermin. De Vaux wanted these buried bones to be the remains of sacrificial meals (de Vaux 1973: 12–4), but more recently a number of people have recognized that the burial of animal bones was a matter of hygiene (Cansdale 1997: 160; Doudna 1999: 43–45; Magen and Peleg 2007: 42–4), although none of them considered the strong probability that many were remnants not of meals79 but of glue or gelatin making, a smelly process which turned the otherwise wasted by-products of animal slaughter into a valuable commodity. During excavations at En Boqeq, an oasis at the south-western end of the Dead Sea, ‘animal fat residues were found in our chemical analysis of soil samples taken from floors and the vicinity of ovens’ (Fischer, Gihon and Tal, 2000: 100). This fat was assumed to derive from animal bones and helped identify the site as a centre for the production of medicinal and aromatic unguents. Unfortunately such tests were not carried out at Qumran. It is worth bearing in mind that the intensive building programme of the Hasmoneans and, in particular, Herod, would have created an exceptional demand on leather for sandals and mason’s aprons. Those few of us who were volunteers for months on end during Yadin’s excavations at Masada often wore aprons improvised from sandbags to withstand some of the wear and tear caused by moving rocks and humping soil. One hopes that those who had to roughly dress large boulders, and to lift and place them into walls for years on end, at least had decent sandals to wear and were supplied with aprons of leather! In an urban environment, such as Pompeii or Rome, with permanent occupation and plentiful running water, one would expect more substantial pools and tanks in a tannery (Leguilloux 2004). In a very rural seasonal setting like Qumran, without running water, they would have been less prominent. At Masada there was evidence for leather working in the Zealot period. Tower L1206 ‘must have been used for leather working judging by the quantity of unworked leather found there’ (Sheffer and Granger-Taylor 1995: 239) and Yadin assumed ‘this tower housed a sandalmaker’s family’ because of the ‘great quantity of leather sandals, scissors (and) awls’ found there (Yadin 1966: 87). On the western side of Masada, Tower L1276 is often referred to as Tanner’s Tower and, indeed, it is the only candidate for the source of the leather found in Zealot Masada. Netzer, with suitable caution, says that in the Zealot period it ‘was converted into a workshop, apparently a tannery with installations in all its parts’ (Netzer 1991: 440). Included in the installations are four pools or tanks, although there are no indications as to how they were filled or drained. These pools are somewhat larger and deeper than the pools in L121 in Hasmonaean Qumran, but no larger than pools that may have been used for the production of leather in the Herodian period (L48, 49, 50, 55, 57). Moreover, it is possible that in Qumran, where the tanning season was limited, portable, ‘disposable’ containers were employed. Baskets woven from the reeds and/or palm fronds of Ein Feshka could have been rendered waterproof by lining them with leather or with bitumen. As de Vaux noted, ‘reeds grow abundantly and must have been turned to good account for basket In the Medieval period a thin application of glue was made to the surface of parchment to produce a better writing surface, although there is, as yet, no evidence that it was so used in the Second Temple period. Gelatin can be used for surface sizing of parchment but is also employed as a gelling agent in the preparation of perfumes and medicines. Some, particularly those found in deliberate fills such as in L130 and in the southern refuse dump, probably were meal residues. These fills are the result of Herodian building activity which would have been carried out by gangs of labourers drafted in from elsewhere, who all needed feeding. If Roman engineers did supervise the construction of the dam and the aqueduct leading from it, they would have felt entitled to more meat in their diet than would tanners or potters. If an area is being raised by back-filling it makes sense to bury any rubbish within the fill.and if, as I suggest, some of these areas were developed as small gardens, the bones would have been a useful additive to poor spoil. 79 55 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Fig 12. (L) Lant Jar from Yorkshire Mill, c. 1900, 19" high. (R) Cylindrical Jar from Qumran (de Vaux 1954 Fig. 5:4) c.48 cm high. (same scale). Nathan 2004: 24, J-SJ2A1, Ill 8).82 No dipper was found with it so it is unlikely that it contained drinking water. In Qumran several similar jars were found standing around the site. ‘The most distinctive feature of these jars... is their short neck and wide mouth’ which would have been ideal for the collection of urine, and their ‘ring or disc bases enabled the cylindrical and ovoid jars to stand on their own’ at convenient locations. They had ‘bowlshaped lids ... designed to be placed over the wide mouths’ which ‘could not be easily sealed’ (Magness 2002: 82). Such lids would have deterred insects, been easily removable to allow for another urine contribution, and would aid the breakdown of urea. Clearly those jars that were buried below floor level would not have been used for the collection of urine but it is noticeable that several that were free-standing were located in secluded corners near industrial areas, either Hasmonaean (L114, L133) or Herodian (L44, L45, L84). Large mouthed jars of similar form and size, known as ‘lant’ jars, were used for the collection of urine in factories in nineteenth and early twentieth century England (see Fig. 12). Wool preparation Shearing took place soon after lambing. A pair of shears was found at Qumran (Roitman 1997: 34; Hirschfeld 2004b: Fig. 78) and the initial processing of wool could have taken place in the western industrial area. The raw fleece had to be scoured to remove the natural oils (lanolin). ‘Urine is the only material that ought to be used for scouring wool... Urine that is fresh voided will not scour well... it should be kept in close vessels until it has completely undergone those changes by which its ammonia is developed....for scouring fine wool, use one bucket of urine to two of water... The urine should be old and the water the softest that can be procured’ (Patridge 1823: 32).80 In the Scottish Hebrides, home of Harris Tweed, ‘urine was a valuable commodity to be collected carefully and not squandered lightly’ (Storrier 2006: 300– 1)81 until well into the last century. Every weaver’s croft had a ‘large tub in the byre or some sheltered place where the rain would not dilute the collection too much and interrupt its maturing, Every household contributed to its own tub as did friendly visitors who were doubly welcome if they happened to be breaking their journey on the long road home after a night on the beer’ (MacDonald 1982: 20–21). In Yorkshire, centre of the English woollen industry in the nineteenth century, urine was routinely saved and collected. An account, written in the 1850s, says that hand-loom weavers were ‘compelled... to take care of their urine to scour pieces, and there used to be a large tub at the top of the stairs in every chamber, and both men and women used these tubs’ (Stead 1981: 17). In the Twin Palaces in Jericho a wide-mouthed jar was found standing in the corner of a small room just off an entrance corridor (Netzer 2001: 153, AE16. Plan 26; Bar- Fleeces could have been scoured in the shallow basins in L121 and the water at Qumran would have been the softest in the region. After scouring, the wool needed beating to detach the fibres and remove any lingering dirt. As noted by de Vaux, the floor of L115/116 (all one room in the Hasmonaean period) was laid on a very solid foundation of large cobbles and would have withstood much beating. According to Patridge it was normal in northern climes for wool to be scoured in wicker baskets so that it was simple to move the wool from scouring tank to scouring tank and then for rinsing in a nearby stream (Patridge 1823: 30). Qumran did not, of course, have the luxury of a stream of running water. Similar wicker baskets were 80 And see Pliny, Natural History 28: 26; 28: 91. For Vespasian’s imposition of a tax on the collection of urine see Suetonius, Vespasian 23. 81 I thank John Shaw of the School of Scottish Studies, Edinburgh University, for this reference. 82 For the routine collection of faeces and urine in Herodian palaces see Netzer 2006: 258. 56 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN WHAT WAS THE CHARACTER OF THE INDUSTRIAL ACTIVITY AT QUMRAN? to use fresh, soft (i.e. rain) water (Vogler 2002: 20).84 Thus the rain water collected in Qumran would have been better for dyeing purposes than the calcareous spring waters of Jericho and Ein Gedi. The plants used would also have had to grow nearby. Woad ‘was extracted from the plant Isatis, mentioned in the Mishnah, which grew in the vicinity of Jericho, Zoar and Beth- Shean’ (Har-El 2000: 14). It is a biennial plant that readily seeds itself at the end of its second year. The plant germinates quite rapidly and the leaves would probably have been ready for harvesting two months after the first rain. As it is easy to grow from seed it would have been possible to ‘cultivate’ limited quantities near to the site. In England, according to Patridge, the leaves could be harvested up to five times a year; but around Qumran perhaps two or three cuttings would have been possible before the heat of summer desiccated the plant. used to suspend wool in a boiling dye vat so that the wool was not scorched by being in direct contact with the heated base of the vat. As with leather production, it is possible that portable vats were employed at Qumran, particularly for the scouring of wool. An advantage of portable containers would be that the residues rich in organic nutrients, which remained after tanning, scouring wool, or cold dyeing, could readily be carried to irrigate and manure plants growing nearby. Dyeing In antiquity it was normal to dye wool before it was spun (Forbes 1956: 21). Urine was needed as a mordant to fix some dyes. Dyeing usually required the wool to be boiled together with the dyestuff and, where required, a mordant. The dye pot had to be kept continuously on the boil. If the installation in L125 had indeed had a large bronze vessel over a furnace (Hirschfeld 2004b: 129, Fig. 69) it would have been ideal for dyeing. Periodically, to ensure even dyeing, ‘it was usual to pummel the material thoroughly, either by hand or by treading underfoot’ (Hodges 1989: 150). The reinforced floor in L115/6 was nearby in the Hasmonaean period. A number of vegetable dyes could have been extracted from plants likely to have been growing in the vicinity of Qumran. Red dyes from madder, henna, or sumach; blue dyes from woad. The necessary parts of the plants (leaves, roots, berries etc.) could have been gathered by shepherds out grazing their flocks. Analysis of the dyes used on textiles found at Masada revealed that madder was the most widely used dye, followed by indigotin probably extracted from woad. A yellow dye from an unidentified source was used occasionally (Koren 1995: 262–263). Most dyeing processes produced strong, offensive smells.83 It was possible to concentrate some vegetable dyes, particularly woad, into powder form which could be added to paints which could have been used, inter alia, for adding tempura details to wall frescos so prominent in Herod’s palaces. As the fermentation process to produce powdered woad is best carried out below 52° C, it would have had to be carried out in the winter months because summer temperatures in Qumran can reach 40+° C without any additional heat caused by fermentation.85 Locally available plants and minerals would have been utilized for the extraction of tannins, perfumes and medicines. Henna, for example, could be used for both tanning and dyeing but also had anti-fungal properties; date palms are an important food source but also have tannic and many traditional medicinal properties, while ropes can be made from palm-frond fibres; sumac has both tannic and dyeing properties and could be added as a spice to food. The acacia tree could produce tannins, perfumes, an astringent medicine, and gum-arabic (which was used as a paint or ink base). Honey from bees that have fed on its flowers is considered a delicacy. Moreover, acacia is thought to be the source of ‘shittim’ wood used for the construction of the Ark of the Covenant (Exodus 25: 10; 30: 1). A black dye could be produced from bitumen, but it was also, according to Josephus, an ingredient in many medicines (War 4: 481). A black dye could have been extracted from the bitumen found floating on the Dead Sea and it is possible that ‘royal purple’ was extracted from murex shells brought from the Mediterranean coast. A number of shells were found near Herod’s Third Palace in Jericho, together with ‘shallow bowls bearing traces of different dyes’ (Netzer 2001: 276, room B116). As the prolonged boiling of the shellfish necessary for extracting the dye was particularly noxious, it is unlikely that the process took place near to Herod’s Palace but the discovery of the shells in Jericho indicates that some extraction was carried out somewhere nearby and Qumran is a likely venue. Strabo, when referring to the production of ‘Tyrian purple’ in Tyre, says that ‘the great number of dye-works make the city unpleasant to live in’ (Geography 16: 2: 23). A Masada textile revealed traces of murex dye (Koren 2011: 2). Spinning and weaving is unlikely to have taken place in Hasmonaean Qumran [‘although some few spindle whorls were found at Qumran it is not known to what period they should be assigned’ (Donceel and DonceelVoûte 1994: 14). An alabaster whorl is from Period III (Taylor 1999: 318–321; 2006: 141). A limestone spindle whorl from L18 in Ein Feshka can be no earlier than Period II]. However, woven cloth may have returned to Qumran for fulling, a process which involved a soaking in urine followed by treading. The area identified as a laundry (L52) by de Vaux may have been used for Dyeing can have variable results. When using vegetable sources it is best to use fresh picked leaves or flowers and 83 A modern British extractor of woad utilizing urine states, ‘The urine vat has a strong smell and is best used outside during the summer’ http://www.woad.org.uk/html/fermenting.html and see http://invention.smithsonian.org/centrepieces/whole_cloth/u3tc/u3mater ials/natDye.html 84 85 57 http://www.woad.org.uk/html/extraction.html#Top1 http://www.newcomen.com/excerpts/woad.htm QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS fulling86 in the Herodian period, as may the two circular installations associated with the late, or even post, Herodian floor in L115 -116.87 production without the loss or contamination of the only available drinking water. Recent petrographic analysis of some of the ceramics from Qumran concluded that ‘it is certain that the clays of the upper part of Wadi Qumran were not the raw material of which the examined ceramics were made’ and that if ‘the Wadi Qumran deposit was indeed used for pottery making, it should be stressed that this is not a raw material dominating among the Qumran vessels’ (Michniewicz 2009: 139-40).88 6.1(ii) Industries not associated with Sheep and Goats. Pottery Although the making of pottery was not directly associated with animals, the potters would have timed their arrival in Qumran to coincide with that of the flocks in the winter. Dairy products would have been their staple diet supplemented by whatever dry foodstuffs they brought with them. The winter temperatures were bearable, the water fresh, and driftwood on the Dead Sea brought down by the winter floods was at its most plentiful to help fuel the kilns. The dried dung of the flocks would also have been an important source of fuel, supplemented by date stones, dried reeds and palm fronds, bitumen and wood from local trees. More trees would have existed than have survived the ravages of man today. De Vaux noted man’s detrimental impact. He observed that ‘to bake their bread the workers used brushwood and roots torn up from the area round the settlement’, and noted that ‘bushes growing in the area watered by the springs actually provided wood for the refugees in the camps at Jericho to this day’ (de Vaux 1973: 85). The potters would have made a year’s supply of pottery for the local market in a month or two and would then have left Qumran to return to the hill country or to labouring on the Royal Estate in Jericho. Magen and Peleg also did not consider the seasonal nature of Qumran. If it was producing nothing but pottery throughout the whole year it would have overwhelmed the limited local market.89 That pottery kilns existed in Qumran is a certainty; the source of potting clay is more problematic. Although it has been claimed that some of the clay used in vessels found at Qumran came from Jerusalem (Yellin, Broshi and Eshel 2001), something that Magness has expanded to make far reaching conclusions about concerns for ritual purity (Magness 2002: 89; 2004: 161), Michniewicz casts doubt on the claim and the scientific methods used to reach it (Michniewicz 2009: 24-28). He shows that most of the vessels belonged to his petrographic groups II and III which ‘represent either the Cenomian Moza Formation or the top part of the Lower Cretaceous Kurnub Group outcropping on the eastern side of the Jordan, and in Eastern Samaria’ i.e. ‘between the northern Dead Sea and Zarga ... and ‘in Wadi Far’ah, Wadi el Malikh’ (Michniewicz 2009: 137-8, 141). This is the same general area as Khirbet el Samra, (Conder and Kitchener 1883 182, 212-3) and the large underground chamber recently discovered north of Jericho,90 from whose quarries the sandstone ashlars, column drums and capitals integrated into the architecture of, inter alia, Masada and Second Temple Jericho are thought to have come (Hirschfeld 2004: 65). During the 1950s and into the 1970s, a number of pottery kilns existed in the refugee camps north of Jericho, which may well have been exploiting this clay source. Although it has recently been claimed that ‘the main purpose of the entire water supply system, with its channels and large pools, was to provide potter’s clay’ (Magen and Peleg 2007: 13) this is extremely unlikely. If it was the main purpose to collect silt, it makes no sense to dig a large pool, such as L71 and L91, to a depth of 5 m to collect a 1 m thick layer of silt which would be inaccessible until the 4 m of water covering it either evaporated or was bucketed out. If the primary interest had been in silt a wide shallow pool no more than 1.5 m deep would have been dug. Layers of silt were found in the bottom of all the pools excavated in Second Temple Jericho yet it is inconceivable that its collection was the main purpose of these pools many of which were focal points within palace complexes. 88 It appears that clay samples were only taken within Wadi Qumran (ibid: 35). Silt from the Wadi would only have been deposited following floods. Another possible, but untested, source might have been silt deposited by local rain-water run-off to the northwest of the site, although it is clear that the majority of the pottery was not made from locally sourced clays. 89 ‘A case-study from fifteenth-century Malta may elucidate more on the matter: Malta is an island which also lacked abundant fuel resources (as it still does). At Ghajn Klieb, in the outskirts of the main city, Mdina, there was a kiln which was used for the baking of roofing tiles, which were used in large quantities on both domestic and ecclesiastical buildings within the city. Apparently this kiln was only lit during the week following the feast of the Assumption (on 15 August), and thus it seems that enough tiles were produced for the duration of a whole year’ (Mizzi 2009: 46). 90 http://news.nationalgeographic.com/news/2009/06/090624giant-christian-cave.html Moreover throughout most of the Hasmonean period, when two kilns existed in L66, the only water supply was in the refurbished Iron Age cistern L110, and the two nearby pools L117 and L118. It is difficult to see how silt in these pools could have been accessed for pottery 86 It is unlikely that a rural area used for fulling, and, perhaps, other things, only seasonally, would be as well defined as that in an urban location where fulling would have taken place year round (cf. Flohr 2005: 59–63). 87 Sheffer and Granger-Taylor suggest that the two circular ‘installations’ in the Masada ‘Tanners’ Tower could have been used for fulling (Sheffer and Granger-Taylor 1995: 234). 58 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN WHAT WAS THE CHARACTER OF THE INDUSTRIAL ACTIVITY AT QUMRAN? Medicines and perfumes Flax retting Many plants with medicinal and/or aromatic properties were to be found in the vicinity of Qumran. As these have been identified and discussed elsewhere (in inter alia Fischer et al. 2000: 97-100, Taylor 2009: 233-39, and see the exhaustive study of medicinal plants in the Mediterranean region in Lev and Amar 2008), there is no need to list them all here. However I would draw attention to the humble mallow, whose leaves and flowers were used for fomentations and poultices.91 In the nineteenth century mallow was regarded as an important food for the poor (Drake 1881: 312, 314), and for those who lived through the siege of Jerusalem in 1947 it was a fall-back food when little else was available. It certainly grows as a weed around Jericho following winter rain and it added seasonal variety to our breakfasts when we were excavating there in the 1970/80s. Like spinach or stinging nettles, an awful lot boils down to very little, but it would have helped sustain a seasonal workforce. Good-quality flax was grown as an irrigated crop around Jericho (Har-El 2000: 14). It is quick growing, being ready for harvesting about three months after sowing. It needs cool nights so, in Jericho, could only have been grown in the winter. Once harvested, the flax had to be threshed or trampled on to remove the seed bole. Separation of the fibrous core of the plant from the outer layer was done by retting; i.e. soaking in a shallow pool for several days (the warmer the water the quicker the process; the water may need changing once or twice). ‘Pure soft water, free from iron and other materials which might colour the fibre, is essential. Any water much impregnated with lime is also specially objectionable’.94 So again, as with scouring wool and dyeing, the rain water of Qumran would be better than the calcareous water of Jericho. Following retting, the fibres were more completely separated by beating. The retting process is a malodorous activity (Heinrich 1992: 19–32) but the unwanted residue could be dried and used as a fuel, a useful benefit if carried out at Qumran. 95 The linen ‘scroll wrappers’ found in Cave 1 were noted as being, in all probability, a local product (Crowfoot in Barthélemy and Milik 1955: 18, 22). The balsam plant, from which perfumed oil and medicines were extracted, would not have grown near to Qumran as it needs irrigation during the summer and there is no evidence for extensive irrigated agriculture at Qumran. However, it is possible that second-grade balsam, extracted by soaking leaves, twigs and trimmings and boiling the resulting liquid, was produced at Qumran (Patrich and Arubas 1989; Patrich 2006; Netzer 2004: 135–8; Hepper and Taylor 2004) using trimmings brought from Ein Gedi either by boat or by pack animals via the coastal road (Porat 2006). The final processing and sale almost certainly took place in Jericho as balsam was a royal monopoly (Broshi 2007: 27 and see Strabo Geography 17: 1:15). Basketry and Mat Making De Vaux realised that ‘reeds grow abundantly and must have been turned to good account for basket making, as they still are to this day’ (de Vaux 1973: 73 and see 85). ‘Mats’ were found in a number of rooms and caves but unfortunately of ‘objects made of wood and other vegetable substances, such as palm-leaf-work or wicker work, … most have no inventory number, some were photographed, few were described and many have disappeared ... as in the case of ... the large mats found in some rooms’ (Donceel and Donceel-Voûte 1994: 14). Because of their ready accessibility, reeds and palmleaves would have been used for many purposes, but the objects made would have had a short working life and, once past their ‘use-by date’ would have served as fuel for the fire. Rope making Rope could be made directly from palm frond fibres (Danin 1983: 127-9), but also from the fibrous sheath at the base of palm fronds.92 As with flax, hemp or coir, it is possible that the latter fibre could have been prepared by retting, i.e. soaking in water to loosen and cleanse the fibres. A piece of rope 4.5m long made of date-palm fibres was found in the Cave of the Balsam Jar (Itah, Kam and Ben-Haim 2002: 172), and its length may indicate that it was of local manufacture (although it is not entirely certain that it dates to the Early Roman Period).93 Palm fibre rope has been found attached to ‘Roman’ anchors near to Ein Gedi (Hadas et al. 2005). Considerable quantities of rope would have been essential for shipping on the Dead Sea. I have suggested that Qumran was primarily a seasonal industrial site and that some of the industrial activities, particularly the curing of skins, the scouring of wool and the retting of flax, were carried out in readily made, and ultimately disposable, containers made of local reeds and/or palm fronds made watertight with skins and/or bitumen collected off the Dead Sea. Several different processes were undertaken in which raw materials were soaked in various liquids which required regular dilution and/or a complete change of liquid over several stages. There are few permanent pools at Qumran suitable for these activities, but it would have been more convenient, particularly where, as at Qumran, there was no running 91 http://www.vitacost.com/Healthnotes/Herb/Mallow.aspx http://www.botanical.com/botanical/mgmh/m/mallow07.html 92 http://flora.huji.ac.il/browse.asp?action=content&keyword =useful%5Fplants%5Fa 93 Although three lamps found there are said to date to the Second Century CE, the parallels that are given place them firmly in the time of the First Revolt, as does the coin found with them. 94 www.1911encyclopedia.org/Flax For the identification of many cisterns, pools and beating surfaces in early Islamic Ramle, as an area for the preparation of flax see Tal and Taxel, 2008: 123–4. 95 59 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS coir cord to the basketry walls’ (Hornell 1938: 153-5 Pls III & IV).100 The outside of the finished ‘basket’ would then be waterproofed with asphalt. They were generally about 2 or 3 m in diameter but some as large as 5 m were reported.101 Very similar baskets, except waterproofed internally, would have been ideal containers for the various industrial processes at Qumran. water, to have many portable containers which could be placed near to available water. The processing of hides was a particularly polluting activity and it would have been advantageous to have ‘disposable’ containers that could be destroyed at the end of each season. Both wool and flax were probably processed at Qumran but, because it was forbidden for Jews to mix wool and linen in the same garment (Deuteronomy 22:11), it would have been simpler to process them in entirely separate containers. In India a coracle can be built, once all the raw materials are to hand, by a team of three in ten hours (McGrail 2003: 246). Due to the inferiority of the materials available in Mesopotamia, which meant that a mat had to be made before a framework could be added, the construction must have required more man-hours overall. The simplest and most obvious container would have been a spherical basket similar to the Mesopotamian ‘quffah’96 or coracle, depictions of which appear on Assyrian reliefs at Nineveh and Nimrud;97 a description of which was written by Herodotus (Book 1: 194); and which survived well into the 20th century. The raw materials, or close equivalents, utilised in Mesopotamia would have been freely available near the Dead Sea.98 There is no doubt that basket and mat making date back to very early times in Palestine, nor that quffat flourished in Mesopotamia at the time that the Assyrians invaded Palestine. As very similar raw materials used for making quffat on the Tigris-Euphrates were freely available on the shores of Lake Tiberias and the Dead Sea, and there was a similar scarcity of timber for planked boats, it seems extremely likely that similar craft were made there too. One could go further and say that it is extremely unlikely that such easily and cheaply constructed craft that were stable (in experienced hands) and able to take quite heavy loads, did not exist on these inland seas. Although I know of no historical or ethnographic evidence for their existence, both Strabo and Diodorus (possibly using the same source) mention that asphalt was collected on the Dead Sea using reed boats which appear to be simply bundles of reeds tied together (similar vessels existed together with quffat in Mesopotamia, and were also known in Egypt). ‘In construction a quffah is just a huge lidless basket, strengthened within by innumerable ribs radiating from around the centre of the floor’ (Hornell 1938: 153). In India, where similar shaped coracles are still made,99 the builders use split bamboo with which to build a strong interwoven framework which can then be covered, originally with hides, although today with plastic fertiliser sacks (McGrail et al. 2003). Smaller versions can be seen in the vegetable markets of India carried on the heads of porters. Bamboo was not available in Mesopotamia (or by the Dead Sea), nor were there any other plants growing nearby from which strong ribbing could be made. Consequently the first part of the Mesopotamian guffah to be built was a coiled ‘mat’ made of ‘a continuous and flattened spiral, formed of a stout bundle of parallel lengths of some fibrous material – straw, reeds or palm leaflets – bound by a parcelling or whipping into a rope-like cylinder. By concentric coiling of this “filled rope”, the shape required is gradually built up ... The gunwale consists of a bundle of numerous withies, usually of willow, forming a stout cylindrical hoop bound around and attached to the uppermost and last-formed coil by closely set series of coir lashings. The inner framework, giving strength and rigidity to the coiled walls of the quffah, is formed of a multitude of curved ribs, closely set; usually split branches of willow, poplar, tamarisk, juniper or pomegranate are employed; when these are not available the midribs of date-palm leaves are used, but these are less esteemed ... As each of these ribs and frames is placed in position, it is sewn with Diodorus, however, describes a sea battle on the Dead Sea in which a Nabataean force of 6000 men defeated the naval force of Antigonus Monophthalmus in 312 BCE by killing many of them with arrows fired from their reed boats (Diodorus Book 19: 98-100). As each boat was said to have two rowers and an archer, we can deduce that the Nabataean ‘navy’ consisted of 2000 boats. Even if we discount this number as exaggerated, the implication is that many boats were available to the Nabataeans, and, where wood was in very short supply, we can suppose that reeds were used. Quffat would have been more stable and more readily manoeuvrable than crude rafts so it may be that part of the Nabataean’s success was due to their having superior vessels with which they were familiar from regular use on the Dead Sea. Herodutus says that quffat were made in Armenia and used to float goods down to Baghdad, where the hide was stripped off the framework and taken back to Armenia for re-use, ‘for it is not by any means possible to go up 96 Arabic for basket, plural quffat. Both on display in the British Museum. At En Boqeq the excavators of what was identified as a perfume factory also noted the probable use of portable containers (Fischer et al. 2000: 138). Although the processing of balsam probably required some distillation over fire (ibid: 116-118) the concentration of some liquids by evaporation may have been achieved simply by putting them in wide, shallow basins thereby exposing the maximum surface area to the heat of the sun. Basins of reeds would have been light and cheap to produce. 99 http://www.telegraph.co.uk/news/picturegalleries/theweek inpictures/8246520/The-week-in-pictures-7-January-2011.html? image=28 97 98 100 For a photograph of Baghdad quffat c. 1935 see Hornell 1970: Pl. IVB. Note the weight of grain they are carrying. 101 Coracles as large as 6 m in diameter are recorded in 17 th cent India (McGrail 2003: 243) but they had the advantage of having split bamboo to make a more solid framework. 60 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN WHAT WAS THE CHARACTER OF THE INDUSTRIAL ACTIVITY AT QUMRAN? stream by water’ (Herodotus I.194). Similarly in 19 th century Tamil Nadu hide-boats, each carrying 4 tonnes of rice, were floated down river in time of flood. Once at their destination the ‘bamboo framing was sold off and the hide retained for further use’ (McGrail 2003: 244). Although Quffat could have been made near Lake Tiberias (possibly waterproofed with asphalt brought up from the Dead Sea), it is unlikely that they were used to carry agricultural produce from the Galilee down the Jordan when in flood because, as Lt Lynch of the US navy found to his consternation, in 1848, when he managed to navigate two metal boats from Lake Tiberias to the Dead Sea, ‘the Jordan was interrupted in its course by frequent rapids’ (Lynch 1849: 264) in fact by ‘twenty seven threatening rapids, besides a great many of lesser magnitude’ (ibid: 265). There are, however, no rapids in the lower reaches of the Jordan, and coracles would have been ideal for transporting building stones and potting clay from north of Jericho (see above) to Qumran. The craft would then remain on the Dead Sea, where they could be used, particularly in the Second Temple period, for carrying building and food supplies in short ‘hops’ along the shore, inter alia for transporting balsam from Ein Gedi to Qumran for onward movement to the royal estate in Jericho for final processing and sale. They would have been smaller, though more numerous than, wooden vessels that undoubtedly did exist on the Dead Sea over several centuries (Hadas 2008, Nissenbaum 1993, Oren 2007/8) although ‘until now anchors have been the only archaeological evidence of sailing on the lake’ (Hadas 2008: 31). A ship-dock has been identified at Qasr el-Yahud, and a possible jetty at Rujm el-Bahr and at Callirhoe, but no docking facilities have been discerned at sites where they would be expected, such as Ein Gedi and Masada (or Qumran). Quffat would have been ideal for lightering from larger vessels anchored offshore. The great advantage of these buoyant vessels is that they would have been cheap, readily available and required only the most rudimentary of landing stages, if any at all. A clearance of rocks may have been all that was necessary. It is to be hoped that with the rapid shrinkage in the shore line of the Dead Sea sharp eyes will look out, not only for anchors, but for surviving relics of more humble craft not built of wood. have extended by a few weeks the period that occupation was possible at the end of the season when the remaining water grew ever more unpalatable (for cereal beer, from Jewish sources, in the Roman Period see Amar, Lev and Yanir 2005: 3–5). Terracotta tallies? During the 1951 excavations an ‘irregular sphere of terracotta pierced with seven holes’ (L2 object 140) was found near to a coin dated 10 CE. Similar terracotta balls, with varying numbers of holes, were found subsequently in contexts ranging from the Hasmonean period (L10A 422/3, Trench ‘A’ 168) through to post-Herodian levels (L91 1525, 1569).103 These apparently unfired, and readily produced, balls may well have been tallies given out as acknowledgement for work done, similar to, but far simpler than, ostraca (c.f. Cross 2007). The frequency with which they were found indicates a busy workforce. 6.2 Seasonal activities and the character of habitation at Qumran. Of all these activities there is positive evidence for the ceramic industry, tanning and, less certainly, rope making. However, all of the industries considered required an ample supply of water, some (dyeing, the preparation of wool, flax and perfumes/medicines) would have greatly benefited from the use of the soft rain water collected at Qumran. A number (tanning, the preparation of wool and flax) would have required the sort of reinforced surfaces found in the ‘industrial zone’ on which beating could have taken place. All were malodorous, varying from the pungency of a corral of sheep and goats to the downright stench of tanning and the boiling up of glue. Besides the smell, a number (pottery making, hot dyeing, glue manufacture) would have added smoke and smuts to the generally unpleasant atmosphere. Most of these processes would have taken place in the winter months following rain. There would be fresh water in the cisterns and plants growing to supply grazing, and from which could be extracted tannins, dyes, aromatics and medicines. In the summer it is unendurably hot in Qumran,104 and the site could not even have benefited from the shade of date plantations; any water would have evaporated fast, become rapidly stagnant and infested with bacteria and mosquito larvae. Flocks, and with them the dairy products which could help support habitation, would have returned to upland grazing. 6.1(iii) Other seasonal activities Further seasonal activities at Qumran could have included the following: the collection of bitumen and salt 102 from the Dead Sea; the burning of indigenous plants to produce lye (Amar 1998), a substance noted for its ability to counteract oil stains and thus, potentially, of use in the preparation of wool; the manufacture of date wine. The presence of a grinding mill in L100 suggests that grain was brought down from the Buqe’ah or elsewhere. Some grain may have been used to produce beer which, in a site where the cultivation of grapes was unfeasible, could 103 L2 140; L10A 422/3; L28 428; L30 2520; L33 755; L44 1001; L66 1172, 1298-1301; L90 1515; L91 1525, 1569: L99 1657/8; L102 2458; L104 2036, 2067; L111 2075, 2110, 2134; L129 2236; L130 2272, 2300, 2280; L131 2574; Trench ‘A’ 168. 104 Inhabitants of the plains of, for example, Mesopotamia or central India, have to endure higher temperatures but they have little choice. From Qumran one has only to climb up into the Buqe’ah to escape the worst of the summer heat. 102 In shallow pools near Ein Gedi Masterman describes salt smugglers ‘having to stand in the water and plunge hands and arms into the saturated brine to seize crystallised masses from the bottom’ (Masterman 1904a: 91-92). 61 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Both the building of the watch-tower and the exploitation of the potential water resources in Qumran would have been instigated by the Royal Estate in Jericho. It was to its advantage to encourage the period of the annual visit of shepherds to be extended both for the dairy products they could supply and for the hides and fleeces that could be processed. This they did by digging more cisterns and building industrial facilities. This reduced the need to use the precious water and irrigable land in Jericho for other than the growing of lucrative cash-crops, and the fact that the inevitable stench and flies from the industrial processes was well away from the Royal Palaces was clearly an advantage. The Jericho estate was not only looking for extra water resources but also for land suitable for date and balsam plantations. Dates would flourish around the brackish springs of Ein Feshka although the salinity of the area would make it unsuitable for growing balsam (Patrich 2005: 250-1). Judging from the number and the size of pools found in Jericho which are thought to be associated with the production of date wine and date honey (Netzer 2004: 134-5),105 the Royal estate had probably acquired a near monopoly on these lucrative processes which required large quantities of dates, just as they had on the production and sale of balsam (Broshi 2007: 27; Taylor 2009: 234). The establishment of date plantations around Ein Feshka would have required a considerable initial investment, from which there would have been no return until the newly planted palms began to bear fruit after about ten years. Date propagation requires an already existing plantation from which to bring suckers. Plants raised from seed are of uncertain quality and would be half male, half female; only two or three male trees are wanted per hectare and to guarantee good quality plants of both sexes it is best to take suckers from plants of proven value. Such plants were available in Jericho but it is likely that the Royal Estate would hold a monopoly on their carefully nurtured plants, and on their eventual harvested products which they would want for manufacturing and marketing in Jericho. It may also be assumed that ‘the salt mines in the Dead Sea region ... were probably royal property’ (Safrai 2003: 111). The economic importance of the royal monopoly over industries in the region underscores the unlikelihood that the King would tolerate competition from a selfgoverning community in Qumran. The restricted water supply available in Hasmonean Qumran meant that the processing of balsam from clippings brought from Ein Gedi was probably negligible. Top grade, sap-derived balsam from Ein Gedi, however, would have regularly passed through Qumran on its way to the royal monopoly market in Jericho. The land route from Ein Gedi to Qumran was over difficult cliff paths (Porot 2006), which would have required sure-footed animals and knowledgeable caravan drivers. The existence of facilities at Qumran would have allowed the balsam to be off-loaded onto different animals for the onward journey to Jericho. What is certain is that Qumran was not a closed settlement with little contact with the outside world. A flaw made by almost all previous scholars is the assumption that any time that Qumran was occupied it was by a permanent year-round population, which encourages the tendency to see the site as having a single over riding function, either, for example, as a ‘monastery’, a villa rustica (Donceel-Voûte 1994), a fortified manor house (Hirschfeld 2004b), the villa of an agricultural estate (Humbert 1994), or a customs post (Cansdale 1997). Although it served as a watchtower it was scarcely a strongly defended fortress (Golb 1995). Hasmonean Qumran could not have been permanently inhabited; rather the site was in an ecologically marginal area which would have been exploited seasonally, following winter rains, for various functions under the impetus of the nearby Royal Estate. Because the work was seasonal there would have been little incentive to build permanent dwellings and indeed there are no signs of any purpose-built living quarters in Hasmonean Qumran. The hard, and none too pleasant, seasonal work would have been done by men. As the water supply was limited, unproductive hands would not have been encouraged. One or two young wives, unencumbered with children, might have been employed to cook and clean, but a female presence would have been minimal, out of hard necessity not because of the celibacy or otherwise of the workforce (contra Magness 2002: 163-85). Magness, en passant, suggests that ‘some of this habitation could have been seasonal—that is, perhaps, some of the members lived at Qumran on a temporary basis’ (Magness 2002: 70), but does not explore the implications of this suggestion. Patrich asks, ‘why should a community with a well-built centre and a sophisticated water supply system have most of its members live for more than a century in humble huts and fragile tents fit for nomadic and semi-nomadic societies, rather than in solid, well-constructed dwellings built of local stone or sun-dried brick?’ (Patrich 2000). It would, perhaps, have been better if he had considered whether, if most people lived in caves and humble huts for more than a century, was not that a strong indication that there was no large permanent community and that the sophisticated water system was exploited by a population of seasonal Although palms would eventually produce a valuable return, possibly the highest of any from the many industries of Qumran, an immediate return could be gained from exploiting the presence of transhumant sheep in the winter. Netzer identifies four date wine presses in Jericho:- the ‘Large Winepress’ (Netzer 2004: 25-9) was Hasmonean (although I believe it was for grape-wine because its location above the Na’aran aqueduct precluded it from having easy access to water). The ‘mosaic winepress’, (ibid: 30-32) was probably Hasmonean. The winepress in Area FD (ibid: 108-113) was almost certainly built in the Hasmonean period and continued to operate into the Herodian period. The wine press A(B)25 (Netzer 2001: 99, 127-8) was Herodian. 105 62 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN WHAT WAS THE CHARACTER OF THE INDUSTRIAL ACTIVITY AT QUMRAN? copies that strayed too widely from the newly accepted standard. Caves in the vicinity of the more populous Jerusalem were more likely to be pressed into use by the local population, for storage, for their animals, for dwelling, or for conversion into tombs, than were the caves around Qumran and, as the winter palaces in Jericho were a popular resort for the ruling classes, it would have been simple to include a few ‘decommissioned’ scrolls in the royal entourage moving between Jerusalem and Jericho. workers, most of whom spent no more than a couple of months at the site each year. Elsewhere he talks of a cave that is ‘large and habitable, but it is doubtful whether it ever served other than for temporary dwelling for shepherds or refugees’ (Patrich 1994: 90). Other caves he identifies as places of refuge at the time of the First Jewish Revolt, and he suggests one cave (Cave FQ37) as a place in which someone ‘got permission to live in seclusion’ (Patrich 1994: 92). It is far more likely that they were occupied by shepherds and other seasonal workers. As Ariel observes ‘considering the paucity of coins from the caves... We may conclude that the areas surveyed were not utilized as places of refuge during or shortly after the war’ (Ariel 2002: 296). Part of the sectarian literature refers to a rift between their authors and the Temple leadership, a conflict that, ultimately, they lost. The more serious that conflict was, the more necessary it would have been for the High Priest to confiscate the heterodox or seditious writings generated by any subversive sects to prevent them influencing others. He would have held on to them as evidence against the sectarians and stored them in the Temple or in the privacy of one of his own palaces. They remained, however, potentially subversive, could not be disposed of casually, and would ultimately have found their way to a geniza which served ‘the twofold purpose of preserving good things from harm and bad things from harming’ (Adler 1903). By the time they were ultimately relegated to the Qumran caves it is possible that the person who instigated their disposal had little idea of their contents but deposited them together with other Temple documents waiting to go into a geniza.106 The pottery found in caves associated with scrolls, of which, admittedly, there is little, dates mainly to the time of Herod or later. When Herod expanded the temple in Jerusalem new ‘library’ facilities would, no doubt, have been included in the development. The transference of documents to their new facility would have provided an opportune incentive to weed out damaged and outmoded or subversive documents and deposit them in genizot. As the industrial processes of Qumran were malodorous, it is unlikely that any scrolls were composed or copied in the polluted atmosphere where the slaughter of animals and the use of dung and urine in processing their byproducts would have rendered all present ritually impure. Writing scrolls was painstaking, skilled work and scribes would have required the best possible conditions in which to work to minimise mistakes. As members of an educated elite, it is unlikely that any would choose to move to Qumran for the short period that fresh water was available to work in the unpleasant atmosphere created by the industrial processes carried out there. Apart from the problems with ritual pollution associated with animal slaughter, the stench, the smoke, and the smuts which were unavoidable consequences of the processes, together with the swarms of flies and mosquitoes attracted by them, would not have been conducive to undisturbed concentration on writing. Ben–Sira, fragments of whose writings were recovered from Qumran Cave 2, wrote in the 2nd century BC that ‘the wisdom of the learned man/scribe depends on the opportunity of leisure, only he that is not preoccupied with labour shall become wise’, and goes on specifically to rule out the possibility of mixing learning with farming, carpentry, smithing or potting. (Ben-Sira 38:24-34). This no doubt more nearly reflects the attitude of scribes in the first century BCE than does the anachronistic transposition to Qumran of monastic ideals of combining work and study which did not even begin to be formulated before the 4th century CE. It is noticeable that all the scroll bearing caves not on the plateau itself were north of Qumran on the path from Jericho. Moreover two references from antiquity, one from the early 3rd Century CE and one from the end of the 8th, mention that ancient Hebrew books had been found in caves near Jericho at those times (Golb 1995: 105-8) which suggests a movement of scrolls from Jerusalem, via Jericho, with some penetrating as far as Qumran. Although ‘archaeology establishes a connection between the settlement and the scrolls in the caves’ (Magness 2002:43) it does not establish an intimate and direct connection, nor that the settlement was sectarian. It is not known when the practice of depositing old documents in special repositories, called genizot, began, but the discovery of two scrolls deliberately buried beneath a back-room of the synagogue at Masada (Yadin 1966: 187-89) indicates that it was operating at least before 73 CE. The high priest would have been ultimately responsible for finding ‘safe’ genizot for the disposal of old, damaged, outmoded or controversial scrolls from the Jerusalem Temple, which may have included, at a time when Biblical texts were being standardised, any existing At the time of the First Revolt a smaller number of scrolls, perhaps including the Copper Scroll found in cave 3, which itself refers to items hidden near to Jericho, may have been deposited by Jews, of whom a handful may have been Essenes, fleeing Jerusalem and heading towards Macaerus, Masada or refuge in Nabatea. I am a field archaeologist, however, and the scrolls are not my field of expertise; Doudna brings a more scholarly 106 Following Sukenik’s early identification of the Cave 1 scrolls as a geniza (Sukenik 1948-50) a French scholar argued that all came from genizot deposits (Del Medico 1957). I have long held this view (Stacey 2004c), and it is increasingly being accepted (Taylor 2012a). 63 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS approach to some of these questions in his part of this volume. would regard any sharing of vessels with castes not their own as defiling, meant that it was expedient to discard these cups, after a single use, in large heaps besides the tracks. It seemed that the Jericho vessels had accumulated for some similar reason. A number of the scrolls in Cave 1 were, according to the Bedu who found them, stored in cylindrical jars, (although no scrolls have ever been found in such a jar during scientific excavation, and the vast majority of the scrolls were not stored in jars). The jars were too tall and wide to be an ideal size for storing scrolls 107 and were originally made for a difference purpose. Similar jars have been found at the settlement (without scrolls) and clearly served a number of storage purposes; a few were buried up to their necks within floors (e.g. L2 27; L13 764/8; L61 1404, 1474; L80 1465), whilst others were free-standing in varied locations, some tucked away in fairly isolated places (e.g. L44 2989; L45 908; L84 1401; L114 2657; L133 2649) where, as already suggested, they may have served for collecting urine. As this type of jar has not been widely found outside Qumran108 it is probable that it was designed specifically for one of the specialised processes carried out there rather than because of a supposed ‘community’s’ ‘unique concern with purity (Magness 2002: 84). Netzer downplayed the huge quantity of tableware found in the miqva’ot writing ‘In our view, this phenomena has a simple and unambiguous explanation. During a prolonged period, perhaps even a decade, the pools were not cleaned. Apparently there were not too many stoppages of water flow (perhaps none at all), and the considerable height of the pools therefore made possible the regular function of the mikveh with constant replacement of its water. Over the years the silt carried by the influx was deposited on the bottom of the pools, and together with this accumulation of silt, the pools became a repository for ceramic vessels which fell into them from time to time. The actual value of such vessels, mainly small plates, was small, and people using the mikveh simply did not bother to retrieve them’ (Netzer 2001: 42). He did not explain what people were doing bringing the vessels into the miqva in the first place nor why they were so clumsy that, by his reckoning, a bowl must have been accidentally dropped in to the water every three days – 1000 vessels over a decade! Magness also claims that ‘the inhabitants produced their own pottery to ensure its purity. This is why the ceramic corpus from Qumran consists largely of the same, undecorated types of cups, bowls and plates, used for dining, which appear to have been manufactured at the site’ (Magness 2004: 91). She also says that ‘the great numbers of repetitive identical plates, cups, and bowls found at the site form a strong contrast with contemporary assemblages at other Judean sites, which are typologically much richer and more varied’ (ibid: 15). It was noticeable that some of the bowls and plates had been so poorly fired that, although they had been unbroken when they fell into the pool it was impossible to recover them because they disintegrated into thousands of tiny pieces no bigger than a match-head. There had, thus, existed more vessels than arrived on the desk of our ceramicist, Bar-Nathan, who believes they ‘indicat(e) their employment in Hasmonean purification rites’ (BarNathan 2002: 86), ‘a ritual of halakhic purpose or some unknown tenet of Sadducean sect, to which the Hasmonean king/priest adhered’ (Bar-Nathan 2006b: 272). The reality is that, in Jericho, exactly the same plates, cups and bowls were predominant particularly in the silt accumulated at the bottom of miqva’ot and other water installations (Netzer 2001: 42, Ill 60, Plate V; Bar-Nathan 2002: 6 , 86, Plate III). From one miqva alone over one thousand of these plates and bowls were recovered and ‘in the Hasmonean palaces, tableware constitutes more than 50% of the entire corpus of pottery’ (Bar-Nathan 2002: 266). Only in one miqva [A(B) 64] were more varied vessels found alongside the usual monotonous bowls and plates and they seem to have been placed there deliberately (Netzer 2001: 123, Ills. 170-1, Plate VII; Bar-Nathan 2002: 6, Plates IV & V). Surprisingly the commonest lamp in Hasmonean Jericho was an adaptation of some of these bowls or plates whereby a nick was broken out of the rim to hold a wick (Bar-Nathan 2002: 86, 104-5). It is not so easy to take a nick out of the rim of an earthenware plate without destroying the complete vessel and many must have been sacrificed for each successful adaptation into a lamp. Thus it appears that these vessels were made for a single use and were then discarded or converted into lamps. When we were excavating these water installations in the 1970s I was struck by the close similarity between the numerous cups we were removing from them and the simple earthenware cups I had seen, and used, at tea stalls, particularly at railway stations, whilst travelling in India a few years before. The many different religions, castes, and sub-castes of India, the members of which 107 Whatever the explanation what is certain is that these vessels must have been produced in large numbers, and, as there were kilns at Qumran, it is likely that some were manufactured there. The large numbers found at Qumran could as easily indicate, not that the occupants were sectarians concerned with their own purity, but that they were busy producing vessels to accommodate an otherwise unknown custom of the high priests living a few kilometres away in Jericho. http://www.uhl.ac/blog/?paged=13 entries for June 20 & 21, 2007. 108 In Jericho (Bar-Nathan 2002: 23-7), Masada (Bar Nathan 2006: 67-72) and Ein Gedi http://eingedi.cjb.co.il/ Season 4 (2006), and another (Hadas pers. com.). 64 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN WHAT WAS THE CHARACTER OF THE INDUSTRIAL ACTIVITY AT QUMRAN? The claim that, at Qumran, ‘the presence of numerous discarded ceramic wares, particularly table vessels’ (Mizzi 2009: 292) was evidence of a particular concern for ritual purity at Qumran is highly questionable. Mizzi admits that most of the Hasmonean pottery ‘evidence comes from dumps’ (ibid: 85). According to recent excavators a number of these dumps contained ‘large amounts of production waste’ (Magen and Peleg 2007: 49) which in certain cases contributed to deliberate constructional fills. Thus to claim that ‘no other site of the same size seems to have yielded such a large pottery assemblage’ (ibid: 154) would only be valid if those sites also had kilns producing pottery for a century and a half, and were built initially on sloping ground striated with, in places, quite deep gullies which needed levelling. The deposits of tableware in miqva’ot in Jericho did not continue into the Herodian period when, in fact, there were far fewer miqva’ot. Thus one might expect a reduction in the predominance of such vessels in Qumran in the post-Hasmonean occupation to be replaced with a far less specialised repertoire. Indeed Mizzi writes that ‘a significant difference is discernible between the Hasmonaean (and possibly the late 1st century BCE) assemblage and the 1st century CE assemblage’ (ibid: 153), a difference I would expect to be particularly noticeable from the time of Agrippa onwards. 65 7. The change of character of Qumran in the Herodian period. The Herodian additions, in particular the western wing of the main building, were built to a better standard than Hasmonean Qumran. Dressed ashlars were employed in the two doorways giving access to the ‘main’ building from the north-west, and its construction on a slope demanded that a considerable quantity of earth, probably from the digging of pool L58, had to be brought in to create a level floor. Whereas all of Hasmonean Qumran was utilitarian and served as an industrial suburb of Jericho, the Herodian western wing was probably an administrative area. scribes in the east until at least the 17th century CE (‘The Bearded Scribe’ Isfahan, dated 1626, Canby 2009: 199). The ancient leather specialists, Poole and Read, suggested that the tables were used for the final preparation of parchment (Poole and Read 1961: 115). They may even have been used as trestles to support skins during the dehairing process, a messy process most likely carried out in the open air. It is, however, far more probable that they were used during one of the several industrial processes suggested for Qumran than for the writing of documents.110 The first two Herodian phases were somewhat slapdash with pools constructed with some sides free-standing, a weakness that caused two of them to crack and become useless. At this point, phase 3, it seems likely that Herod, himself, sent engineers, quite possibly Roman, to oversee the construction of a dam and an aqueduct connecting it, via tunnels cut through bedrock, to the earlier rain run-off channel. This was a major change; a state enterprise that greatly increased the potential water supply and made some limited year-round occupation possible. Since the unearthing of the two inkwells, one of bronze and one of clay, several more pottery inkwells have been found (e.g. Magen and Peleg 2007: Pl 5.5) but this is not surprising as at least one (found at Ein Feshka) was ‘locally made at Qumran’ (Gunneweg and Balla 2003: 13). As ink wells were made in the Qumran kilns it is not surprising that several would be found near by. De Vaux, of course, saw the inkwells as proof that the scrolls were written here. However it is wrong to equate signs of literacy with the writing of religious scrolls rather than with normal administrative work. It is true that inkwells are unusual finds on excavations but it would be dangerous to assume that the lack of inkwells found in the Royal Palaces at Jericho, Herodium, or from Masada, was evidence for the illiteracy of the courts of the High Priests/King. More recently at least five inkwells have been found at a site, dated between the two Jewish Revolts, at Shu’afat, north of Jerusalem (Sklar-Parnes 2005),111 but it would be foolish to suggest that they were proof that scrolls were written there. The large pool L71 was probably constructed at the same time as the dam, as a reservoir for the potential increase in the water supply. It, and somewhat later, L91, were constructed more soundly without free-standing sides, back-filling where necessary between the pool and a retaining wall. The large storeroom L77 and the winepress L75 relate to this phase, as in all likelihood do the building and winepress at Ein Feshka (Hirschfeld 2004b: 189). De Vaux identified two of the new rooms using monastic terms, unfortunate anachronisms as Christian monasticism did not begin to develop before the fourth century CE. The long room L77 he identified as a ‘refectory’ although the many parallels that have been excavated since indicate that it was a typical storeroom.109 An upper floor above L30 became a ‘scriptorium’ similar to ‘rooms in monasteries of the Middle Ages’(!) (de Vaux 1973: 30), on the grounds that some plastered mud-brick structures were identified as tables on which scrolls were written, and the discovery of two inkwells. The mud-brick structures were reconstructed on wooden frames in the Rockefeller Museum to look like a modern picnic table. The reconstruction (de Vaux 1973: Pl XXIa) was inaccurate; the ‘table and bench as exhibited bear no true relation to the original structures’ (Clark 1963: 63). Moreover it was quickly pointed out that ancient scribes did not sit at tables but sat cross-legged on the floor resting their writing tablet on their knees (Metzger 1958-59). Several sculptures from Egypt show seated scribes cross-leged dating back as far as 2,500 BCE (e.g. ‘The Seated Scribe’ in the Louvre) and that posture remained typical for The water storage capacity was greatly increased, not only by the construction of two large pools but by the utilisation of water held behind the dam in Wadi Qumran, which could be raised into the water system by means of a shaduf. Although for many Jews the use of a shaduf to replenish pools would not necessarily negate the use of those pools as miqva’ot, providing 40 se’ah (c. 750 litres) of ritually pure water remained in them, for the authors of the Damascus Document, any water that had been contained in a vessel (which would include a shaduf) was not acceptable for purification (CD 10.11-14).112 The replacement of water in a pool by the use of a shaduf would probably have made them increasingly unacceptable for purification purposes for any but the least observant, as might the regular extraction of water from pools for industrial or other purposes, as inevitably some water would have dripped back in from a bucket or other vessel dropped into the pool. The pool L138, whose method of connection to the water system is uncertain (see above), could easily have been isolated from it to 110 Although it has been suggested that the tables were a triclinium for dining they clearly were not (cf Netzer 2001: 189-193). 111 http://www.hadashot-esi.org.il/report_detail_eng.asp?id=179 &mag_id=110 112 For a discussion on the halakah of miqva’ot see Magness 2002: 134-62; Galor 2003: 316-17). 109 Unless the similar c. twenty ‘storerooms’ at Masada were separate dining rooms for twenty different sects! 66 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN THE CHANGE OF CHARACTER OF QUMRAN IN THE HERODIAN PERIOD prevent the entry of ‘impure’ water and is thus the only pool likely to have been a miqva at all times (cf, although with different reasoning, ‘We believe that the only pool that may have served as a ritual bath is pool L138’, Magen and Peleg 2007: 37). Soon after the Battle of Actium early in September 31 BCE, Herod, who had supported Antony and the losing side, decided that he had to visit the victor, Octavian, in an attempt to salvage his position. Not being confident of his family’s complicated loyalties whilst he was away he banished his wife, Mariamne, and her mother to the desert fortress at Alexandrium, north of Jericho, but sent his two young sons, under the watchful eye of his mother, to Masada (Ant. 15.183-86), of which little had yet been built. Herod may have had only a limited interest in Qumran militarily but it was well located to facilitate the collection and distribution to Masada of building materials, and of food supplies, much of which would have come down the rift valley from the Galilee, suitable not only for his royal sons but for the hundreds of workers employed on its construction. Herod succeeded in making himself useful to Octavian to such an extent that Octavian returned to him the date and balsam groves around Jericho which Antony had given to Cleopatra. In 27/26 BCE Octavian sent Herod a shipment of wine perhaps for the inauguration of the Northern Palace at Masada (Bar-Nathan 2006a: 18, 314-5), the construction of which had required the import of timber for scaffolding and roofing, of dressable stones for its columns and capitals, of mosaic tesserae for its floors, and of colourants for its frescos, besides rations for its numerous workers at a site that could produce little food of its own. There are overflows in some pools113 that indicate that they could be refilled more regularly than by occasional floods, and they seem to direct water to garden areas which would have required a steady water supply. These factors indicate that, besides the seasonal industries, the demand for which would have continued, the feasible length of the season was increased and some limited year round occupation was made possible. The courtyard building at Ein Feshka (Hirschfeld 2004b: Fig. 113) has similarities with the row of nine buildings near the industrial zone in Jericho (Netzer 2004: Plans 14, 19-21, 24) which were assumed to be the living quarters of estate managers. Ein Feshka probably served the same purpose. Whereas in the Hasmonean period the date plantations were only being established and their limited harvest taken to Jericho for processing, by the Herodian period they would have been mature and producing heavy crops, the processing of which was probably the most lucrative of all the industrial activities of Qumran. It was worthwhile to introduce winepresses and process some of the dates on site. This not only saved water in Jericho but the date-stones were a valuable commodity; they could be used as fuel for the Qumran kilns (Magen and Peleg 2007: Figs. 9-10), either directly or following crushing to extract the kernels as fodder for sheep, particularly valuable in a season when grazing was poor, or to extend the period that sheep could stay in the area (Hawes 2008: 362). Processing the dates was a skilled job and the house in Ein Feshka would have supplied accommodation when necessary for specialists from Jericho who could make the date wine and maintain the plantations. It should be noted that, unlike in the houses in Jericho, there was no miqva in that in Ein Feshka (although the springs themselves could have been used for ritual bathing). At the same time Herod was constructing other palatial fortresses in the region, particularly those at Hyrcania, Machaerus and Herodium, and this would have needed a small staff of quartermasters, able to live permanently at Qumran, to supervise the distribution of the royal stores. Caravans would have travelled down the rift valley from the Galilee on regular pack animals, but at Qumran the goods would have needed transferring either onto surefooted animals used to mountainous terrain, or onto ships or coracles for ongoing distribution. He may have chosen his quartermasters from the ranks of the Essenes, as Josephus claims that Herod ‘continued to honour all Essenes’ (Ant. 15. 371-379), although this was on account of one educated Essene’s ‘knowledge of divine revelations.’ The story, associated with Herod’s childhood, is, however, tacked on to an account of his problems arising from his adult unpopularity and may be an apocryphal addendum indicating Josephus’ own interest in divination. To use it to ‘surmise that the gifts given by Herod to Essenes included a tract of land between En Gedi and Jericho’ (Taylor 2012b: 247) is a step too far. It would be wildly out of character for Herod to relinquish control over the Qumran water supply to a quasi-independent sectarian community particularly as he had such difficulty bringing sufficient water to his estate in Jericho (see introduction above). That he foresaw that, by the construction of a prestigious dam, the site could become a cog in the extensive building programme which was so important to him is far more likely, The seasonal activities would have continued to meet established demand but a limited year round occupancy could now occur at Qumran, made possible by the considerable engineering and building programme initiated by Herod. During 32-31 BCE Herod was engaged in conflict with the Nabateans during which he ‘began to fortify the eastern boundary of his kingdom by reconstructing older Hasmonean forts and fortresses and by establishing new ones on both sides of the Jordan river and the Dead Sea’ (Taxel 2011: 401). 113 Some claim that later halakha would rule that overflows would ‘disqualify (the pools) as installations for the function of ritual immersion’ (Galor 2003: 316). In Jericho, however, many pools which were definitely miqva’ot because they came in pairs, one functioning as an otzar (‘treasury’) of ritually pure water and the other with steps for immersion, did have overflows (Netzer 2001: 39-43, Ills 55, 56, 58; Netzer 2004: 149-150, Ill 167). Essenes were associated by classical authors with the area north-west of the Dead Sea (Pliny, Hist. Nat. 5:15 [73], 67 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Taylor 2009: 231-2). This may have come about because of their involvement in shepherding (Josephus, Ant. 18, 19, Philo Q.o.P. 78) whereby they would have wandered with flocks, mainly belonging to rich land-owners, through the Buqe’ah and down to the Dead Sea littoral in much the same way that, in more recent times, the Ta’amirah Bedouin gained some control over the region. The interdependence of Qumran and Jericho, however, coupled with the strategic location of Qumran and its water supply at the foot of a track leading up to the Buqe’ah, make it extremely unlikely that the Hasmonean kings, in particular, would have tolerated a selfgoverning, religious community there as envisaged by de Vaux, particularly not one hostile to them in their capacity as High Priest (Collins 2010: 10-11, 192). subjected to circumcision. Some may have been impressed to undertake hazardous jobs (such as building the northern palace at Masada, or the dam at Qumran) or those activities, such as the preparation of skins, which would have been ritually polluting for Jews. We know of one of Herod’s slaves by name. Following Herod’s death, Simon, a tall, handsome slave, who had served Herod in Jericho where, because, he was ‘far superior to others of his order ... had had great things committed to his care’, burnt down and looted the royal palace in Jericho, set a diadem on his head, and attempted to set himself up as king. To help finance his ambition he and his followers attempted to plunder other of the royal palaces, including, possibly, that in Jerusalem, before eventually retreating to Perea, apparently his place of origin, where he was pursued by Roman soldiers who cut off his head (Ant. 17. 273-277; War. 2.57; and Tacitus Histories 5. 9; and see Kokkinos 2007: 297). The details are sketchy but the following scenario is not impossible. Simon may have been familiar with Qumran, perhaps even had certain aspects of its daily life ‘committed to his care’. Part of the loot from the palace in Jericho included silver coins which Simon chose to bury for safe keeping in Qumran whilst he went off looking for more treasure. His unfortunate demise meant that he never returned to retrieve his loot, represented by the three pots of Tyrian silver coins found in L 120. The unlikelihood that these represented a collection for payment of the temple tax, as has been suggested (Magness 2002: 191-2), is pointed out by the Lonnqvists. They note that in the limited amount of the hoard that they have been able to study at least four of the coins came from the same die indicating that the coins represent a currency hoard, possibly belonging to ‘a single wealthy family’ (Lonnqvist 2006: 139-140). That wealthy family may have been that of Herod himself. Itinerant potters and leather and wool workers may have been attracted from as far away as Jerusalem, but it may well have been more difficult to find the unskilled workers to do the unpleasant and ritually defiling work such as the preparation of hides. Any shortfall in the labour force could have been made up from the large number of slaves who were available in Jericho, if we are to judge from the 500 that could be assembled, presumably within twenty four hours, to take part in Herod’s funeral procession (War. 1. 673, Ant. 17. 199). There are many references in the writings of Josephus to the slaves, even eunuchs, of the Hasmonean kings and of Herod. Many who are actually mentioned were in positions of some importance within the royal households (inter alia, Ant. 15.226: 16.230-3; 17. 55-6; War. 2.57). These slaves were presumably of Jewish origin, or were gentiles who had undergone circumcision and ritual immersion so that there was no danger that by their very touch they would have rendered food or wine impure. Although ‘Roman agricultural writers advised estate owners to employ day labourers rather than slaves for the most difficult and excruciating tasks to avoid damaging their slaves’ bodies’ (Hezser 2005: 249) it is not clear that such considerations would have concerned the Hasmonean Kings, or Herod. The rapid expansion of the royal estate in Jericho could only have been achieved with a large influx of labourers and some of these may well have been slaves who ‘were available throughout the year’ and ‘could be organized in work gangs which were easy to supervise’ (Hezser 2005: 250). However, as the day-to-day life of slaves was of no more concern to Josephus than were the lives and conditions of domestic animals, we have little idea as to how many slaves were owned by the royal estate, or to what work they all were put. If Qumran was a seasonal industrial suburb of Jericho and there was a shortage of labour there was no nearby settlement from which surplus day labourers could be drawn and slaves may have been drafted in from Jericho. Not all slaves were of Jewish origins (inter alia, Ant. 13. 319, 397; War. 1. 376), especially those who had been captured in some numbers during Herod’s military campaigns, nor is it likely that all would have been We know little about the daily life of slaves and even less about what happened to them once they were dead. It has been claimed that the cemetery at Qumran is exclusively male and therefore proof that the site was occupied by a sect of celibate Essenes. Even if the cemetery was indeed predominantly male (and too few graves have been excavated to be certain of this), this could be accounted for by the fact that the site was predominantly seasonal and seasonal workers would be predominantly male. It has been pointed out that some of the peculiarities of the graves (when compared to rock cut tombs known elsewhere which clearly belonged to wealthy families) are because they were for the burial of the poor (Taylor 1999: 312-3). A particular part of the poor would have been slaves who would have had no family tomb in which to be interred. A small area of poor Hellenistic burials, dug shallowly into the soil, was excavated near to Herod’s hippodrome in Jericho (Stacey 2004b: 226).114 If these were burials of the poor and/or of slaves then the 114 Because a photograph of one of the EBIV graves found during the excavations was inadvertently used to illustrate the skeleton in the Late Hellenistic grave (Stacey 2004b: Ill 263) I want to take this opportunity to publish correct pictures (Fig. 13), particularly because this grave is referred to in Avni’s chapter that follows on the Cemetery. 68 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN THE CHANGE OF CHARACTER OF QUMRAN IN THE HERODIAN PERIOD misfortune to be eunuchs they would have been exclusively so! encroachment of buildings and the introduction of irrigated agriculture onto previously uncultivated land would seriously have restricted the area available to them for burial and it is possible that the bodies of some slaves were transported to Qumran for burial instead. Slaves were not allowed to marry Israelite women (Hezser 2005: 197) and it may be that for some male slaves, particularly non-Jewish prisoners of war, the prospect of finding an acceptable partner, even if they were encouraged to do so, would be unlikely. Thus slaves buried in Qumran would have been predominantly men. And if they had the Many shaft graves similar to those in Qumran, have been found in a large cemetery at Qazone in Nabataea. They are dated a few centuries later and definitely contain burials of men, women and children. Could it be that the type of grave originated in Nabataea and that those in Qumran belonged primarily to slaves who had been captured during the various conflicts between the Nabataeans and both the Hasmoneans and Herod? Fig 13. Late Hellenistic military grave in Jericho; skeleton supine, head to east. NB – ledge about 2m from top of shaft and 1m from its bottom which supported a mesh of organic matter (palm fronds?) on which rested several large body sherds from MBIIB storage jars recovered from a tomb which the eastern end of the grave had penetrated (undercut visible in east at elevation of ledge). A sword stood on the northern ledge. 69 8. Qumran after Herod. After the death of Herod there would have been little incentive for year round occupancy at Qumran. His programme of building and stocking desert palace fortresses died with him and the buildings at Qumran went into decline as they were no longer supported from the Royal purse. The Qumran dam, which would always have had problems with silting, and the section of the aqueduct leading from it that clung to the face of the steep bluff, would no longer have been maintained. Nearly half of that aqueduct has since collapsed into the wadi and it would always have been vulnerable to erosion (Ilan and Amit 2002: Figs. 1 & 2). A return to seasonal occupation may account for the shortage of floor accumulations that can be confidently attributed to this period. Undoubtedly at the time of the destruction of Qumran by the Romans in 68 CE it would have been simple to destroy any surviving parts of the aqueduct running around the bluff, leaving just the original ‘broad’ aqueduct to collect local rainfall. improvements to the fortifications of Jerusalem, (Josephus, Wars 5.4.2) and he may well have attempted to renovate some of the defensive positions on his borders. The quite extensive earth-moving noticeable in certain loci at Qumran (L1, 2, 4, 36 etc) do appear to be part of a restoration project of a neglected structure. Political instability under the procurators Antonius Felix (52-58 CE) and Porcius Festus (59-62 CE), and rivalry between the Zealots and the Sicarii, created a situation when ‘Judea was afflicted by robbers’ and ‘all the villages were set on fire’ (Ant 20:185-6). A lack of security would have contributed to the abandonment of sites along the western side of the Dead Sea starting c. 50 CE with, perhaps, a concentration of some of their inhabitants at the recently renovated and more easily defended site of Qumran. But there could have been ecological reasons as well. The soil in this arid region has a naturally high salt content and it is poorly drained which means that water remains close to the surface where it evaporates rapidly in the strong sun (Bresler 1981: 65-66). The effect of 100-150 years of intensive irrigated agriculture, particularly if water was drawn from saline springs (as existed in Ein Feshka, and in one of the springs of En Boqeq) would have been a deleterious build up in soil salinity which would have eventually resulted in declining agricultural returns. There was no longer open hostility with the Nabataeans. Over 50% of the coins found at En Boqeq were Nabataean which reflects that a close trade relationship had developed between Nabataea and southern Judea during the first third of the first century CE. Despite this the perfume factory at En Boqeq was abandoned around 55 CE (Fischer et al. 2000: 137). The latest coin found at En el-Ghuweir was of Agrippa I, dated 42 CE, and occupation may well have ended within a decade of its minting (Bar-Adon 1977: 12). Although the site may have been destroyed by fire as part of a scorched earth campaign by the Romans at the time of the First Revolt there were no signs of either refugee inhabitation or of a military attack. In Jericho the Royal estate went into serious decline after the demise of Archelaus (4 BCE-6 CE). Many of the date and balsam plantations would have to have been abandoned once the aqueducts that had irrigated them no longer functioned. There would have been a reduction in the population which, together with the fall in the wealth being generated, would have lowered the demand for utilitarian products. Although the final deterioration of the estate, including the industrial area, is blamed on an earthquake in 48 CE (Netzer 2001: 10) this was probably only one of several contributory causes. At the height of the intensive agriculture exploitation, when every available spring was utilised for irrigation, seasonal flocks of transhumant sheep and goats would have been forced onto ever more marginal areas. Qumran, with its water cisterns but only very limited scope for irrigated agriculture, would have attracted more and more flocks and the resulting over grazing would have denuded and debased the land over an ever increasing area. However, once the formerly irrigated plantations in Jericho and En Boqeq reverted to marginal land the flocks could return to graze over them. Consequently Qumran’s importance in the flagging local economy would have deteriorated and it would have reverted to being a watering hole for transhumant flocks whose shepherds would have had little incentive to process skins or wool far from their traditional markets in the uplands, but who may have continued to bury some of their dead in what had become a traditional burying ground (see Avni, below, and Avni 2009). Agrippa I during his brief reign (41-44 CE) instigated 70 9. Could Qumran have been a sectarian settlement? ‘reasonable man on the Jericho donkey cart’ who was, no doubt, more concerned with finding his daily crust.115 The text sholars’ esoteric studies, which have generated thousands of learned articles, and will no doubt generate thousands more, have a tendency to place more importance on the scrolls’ contents in the way Qumran is viewed, than on the archaeological evidence from Qumran, or of the pragmatic economic realities of wresting a living from marginal land. We have already seen that the assumption was made early on that the ruins at Qumran had been the home of a sectarian community which had turned its back on the sins of the city to live in isolation, and settled there to write the scrolls found in nearby caves. Concepts found in the sectarian literature of the scrolls, and in references to Essenes by various classical authors, were freely used to interpret aspects of the archaeology of Qumran, and these interpretations then used, in a circular argument, to ‘prove’ that the site was ‘religious in character, with special ritual observances of its own’ (de Vaux 1973: 87). Throughout this study I have offered alternative, practical interpretations which, without being able to offer definite proof, would point to Qumran being, predominantly, a seasonally occupied, industrial suburb of Jericho. Unfortunately, when rare scientific tests were carried out during the excavations they were not conducted objectively. Analyses looking for residues from tanning were carried out at Ein Feshka but not at Qumran itself because, it was assumed, ‘the community would have been too strict to permit’ tanning there (Poole and Reed 1972: 151–152; de Vaux 1973: 78-82). An objective programme might have revealed the presence of some of the industries I have suggested. Today it is unlikely that, after more than half a century of exposure to the elements, and to the tramp of thousands of curious feet, any modern tests at the site would give conclusive evidence for or against intensive industrialisation at Qumran, any more than they could prove the existence of an isolated sectarian community. Ancient sources only place Essenes generally west of the Dead Sea, none specifically at Qumran. To interpret ‘statements about the Essenes in light of the DSS and then use the alleged parallels to prove the identity of the two groups’ is a circular argument (Mason 2009: 240). Those statements that do not accommodate so well to the theory are downplayed (Mason 2009: Chapter 8, particularly Pp. 244-8). Collins accepts that the location of a sectarian site ‘in an area dominated by Hasmonean fortresses is problematic’, although he adds ‘if multiple stepped pools were part of the complex from the beginning, then it was presumably a sectarian settlement from the start, despite the proximity of the Hasmonean fortresses’ (Collins 2010: 10-11). He prefers to relegate ‘the difficulty of dating the pools’ to a footnote (ibid: 190 fn 115) and argues that ‘the Qumran pools are considerably larger than most contemporary miqva’ot’ due to the ‘size of the community using them’ (ibid: 189). Perhaps because he has already accepted that there was a ‘sectarian community’ he does not consider whether, because Qumran, unusually, was at a lower elevation than its water sources, it was convenient to have its main cisterns dug into the easily quarried marl within the site itself. He comes to the conclusion that ‘the argument that Qumran was a religious, sectarian settlement is based especially on three features: the burial of animal bones, the cemetery, and the number of miqva’ot’ (ibid: 196). The animal bone deposits are assumed to be unusual because they are identified as the remains of ‘sacred’ or communal meals (de Vaux 1973: 12-15; Magness 2002: 117-21; Collins 2010: 196-7; Humbert 2003: 434) rather than because they are the inevitable by-product of animal slaughter. I hope to have shown here that none of these three factors need necessarily relate to sectarians. Nor will such proof necessarily come from a final publication of all the data from the excavations, although it is hoped that, one day, this will appear. It may consolidate my dating of the ‘main’ water system and help to determine the stratigraphic progression of the site, but the lack of controls, and the poor recording during the excavations, will always leave some room for doubt. Fundamentally people will continue to accept an interpretation of the site that best satisfies their own psyche, although I hope that they will take into account my redating of its development. The concept of a community of poor sectarians isolated in the desert and busily writing scrolls has some obvious appeal for scholars labouring in the ivory tower of academe, or for theologians sequestered within their own esoteric communities. Furthermore it is the romantic, mystical aura that has been generated around Qumran that sells semi-popular books, fills lecture halls, and brings in the tourist, not the unremarkable ruins themselves. Any indication that the site may have existed solely to play a small part in the local regional economy will be resisted as an altogether too mundane concept. Mizzi, whose methodical study of the material culture from Qumran led him to conclude that it ‘is generally indistinguishable from that of other sites’ (Mizzi 2009: 290, nonetheless claims that ‘one cannot disregard the fact that the archaeological evidence does indicate that the Qumranites were particularly concerned with ritual Textual scholars, who have delved into the arcane world of the scrolls, have, perhaps, allowed their own modern fascination with the glimpses that the scrolls give of rarefied philosophical debates of days long past, to exaggerate the importance they would have had to the 115 Classical authors claim that there were 4000 Essenes, or only c. 0.2% of the Jews of Second Temple Palestine, which historians estimate to be about half of the total population of 4 million. So if the scrolls were indeed the work of an Essene elite they directly concerned very few people, 71 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS purity, as attested by the evidence of the miqwa’ot, by the presence of numerous discarded ceramic wares, particularly table vessels, and particularly by the presence of numerous stone kraters’ (ibid: 292). This seems to be based more on the statements of various Classical authors than it is on the archaeology. Whilst he accepts that Qumran was an ‘agro-industrial’ site he limits the industries to pottery making, which he downplays (ibid: 48), and, primarily, to the processing of dates. It is true that the propogation of dates had, perhaps, the most potential for profit but for that very reason the Royal estate is likely to have maintained a monopoly over it. It would not have been in the interest of either the Hasmoneans, or Herod, to tolerate a ‘selfsufficient and self-sustainable community’ (ibid: 290) at Qumran even if it was feasible. Mizzi understates the smelly, ritually polluting and water-intensive processes such as the preparation of wool and hides from sheep, whose seasonal presence in the winter is a certainty, not because he can prove that they did not take place at Qumran - any more than I can prove that they did [although the partially processed leather found in the Cave of Leather (see above) is a tentative indicator] – but, seemingly, because it would reduce the number of cisterns interpreted as miqva’ot and thereby detract from the interpretation of concern for ritual purity of the site. A possible reason for the large number of table vessels in the Hasmonean period I suggest above, and I suspect that the presence of stone kraters will be seen to be less ‘particular’ at Qumran once the vessels from Jericho, Cypros and Masada are finally published. Winter was the season for Qumran. If the rains fell, water would be available. Transhumant flocks would supply sustenance; the River Jordan in flood would bring drift wood into the Dead Sea, and could be exploited to float potting clay and building stones from north of Jericho. Caravans bringing surplus foodstuffs from the Galilee would travel down the rift valley. Dates would be picked, and, at the end of winter, their flowers would be pollinated. Summer would have been unpleasant at Qumran with little shade from the incessant sun. Any surviving water would rapidly stagnate and become infested with insect larvae. Sheep could not survive the heat. Although a few goats could survive around Ein Feshka their rapaciousness would rapidly degrade the area. No agriculture would be possible and food would have to be imported (most likely from Jericho!). It is difficult to see any industries being active during the summer. What incentive would there be to live there all the year round? It seems that a yearning to believe in the romantic notion of a community of idealised, peaceful scholars quietly writing scrolls in austere conditions of extreme ritual purity has led to some exaggerated and often illogical claims being made, in an attempt to ‘prove’ that scrolls were written at Qumran. This was claimed when a fragment of parchment from the caves was shown to be derived from a Nubian Ibex, a desert dweller, frequently to be seen around Qumran to this day (Bar-Gal 2006). Although it is very plausible that the skin was prepared at Qumran along with those from sheep and goat, there is no reason to suppose that the prepared skins were written on on the spot, rather than being transported to scribes elsewhere. It is clear that not all stepped pools were miqva’ot. The large cisterns along the western face of Masada would have contained ritually pure water but it is extremely unlikely that they were used as miqva’ot. Many of the Qumran pools were more likely cisterns whose suitability for use for ritual purification would have been compromised by the extraction of water for industrial purposes, and their replenishing by use of a shaduf. A more far-fetched claim has been made that, because the ink used on one of the scrolls had a high bromine content, which was assumed to have been introduced when dry pellets of carbon ink were mixed with water from the Dead Sea, thus ‘we could directly link the fragment, and consequently, the production .... to the Qumran area’ (Rabin et al. 2009: 102). No explanation is offered as to why any scribe at Qumran would go all the way to the Dead Sea for a small jar of water to mix with his ink pellets when he was surrounded with cisterns full of rain water. Nor were tests made to determine whether a scroll, or indeed any porous material, would naturally absorb bromine after 2000 years in a cave near the Dead Sea. Such a possibility is dismissed on the curious grounds that the scroll which had been tested was amongst the first to be acquired by Sukenik in 1947 and ‘thus, we expect it to show minimal traces from the mineral deposits of the cave rocks’ (ibid: 98, my italics). And yet in the 1950s when linen cloths, thought to be scroll wrappers, and derived from the same cave as the scroll fragment, were studied ‘it was ... suspected that the textiles were contaminated with Dead Sea salts. Crowfoot therefore cleaned the larger textiles using ... detergent’ (Taylor et al. 2005: 163). Moreover, tests on the human bones from the cemetery at Qumran have shown a high It is usually overlooked that, throughout history, the area around Qumran was exploited by shepherds bringing their flocks down from the Buqe’ah for winter grazing, and that the nearest settlement of any consequence was the economic and political powerhouse of the Royal Estate in Jericho, which would have influenced whatever developments occurred in Qumran, particularly the establishment of date plantations, and the building of a water retaining dam. The question as to whether year round occupation was feasible (or desired) is rarely addressed but is of great importance. The site, and the part it played in the local economy, was developed during seventy years of Hasmonean exploitation before the Herodian expansion of the water system and of the number of cisterns and pools. The available water was therefore limited. Is it likely that it was eked out to try to support year round habitation? Or was the water that was available simply exploited, as long as it lasted, for waterintensive industries? 72 DAVID STACEY: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN COULD QUMRAN HAVE BEEN A SECTARIAN SETTLEMENT? subsidising by spells on building sites in Ein Gedi or En Boqeq). Thus I am familiar with seasonal work, have sympathy for those who carry it out, and am acquainted with buildings that were only used seasonally.116 My intimate knowledge of the regional archaeology tells me that Qumran, for a large part of its active life, could not have supported year-round occupation, and it satisfies my psyche to view Qumran as a predominantly seasonal site, producing, by processes that were unpleasant, articles that were necessary; a place bustling with activity for a few weeks in the year, but, during the harsh heat of summer, lying empty and abandoned to the thirsty, desiccating sun. I am sure the seasonal workforce preferred it that way too. bromine content (Rasmussen et al. 2003b), indicating that porous material does absorb bromine over time near the Dead Sea. I am not a scrolls scholar, but for many years I worked as a field archaeologist at, inter alia, Hasmonean/Herodian Jericho, Masada, Cypros and Herodium. Inevitably I do not look at Qumran in isolation, but as just one archaeological site amongst others of the same period and within the same region. In my youth I worked in several countries, either seasonally or as opportunity arose, and my earliest professional involvement in archaeology was little more than a fascinating but pitifully paid alternative to such itinerant work (which sometimes needed 116 For several seasons I lived and worked, during the hop harvest, in traditional Kentish oast houses which were in use for only six or eight weeks in the year. The remaining ten months of the year they gathered dust. Nowadays they have been converted into trendy houses for the wealthy and are, in all too many areas, the only visual reminder that hops were ever grown there. 73 QUMRAN REVISITED: A REASSESSMENT OF THE STRATIGRAPHY OF QUMRAN de Vau s phases th th Iron II Border outpost, occupied when necessary. 6 -2 Cent. BCE th nd --- Winter transhumance of sheep/goats. No permanent occupation. Use of cemetery?? Hasmonean 100-31/20 BCE Ia and part of Ib Transhumance of sheep/goats. Military border post. Seasonal industries. Specialised pottery production. Cemetery. Herod 31/20-1 BCE Parts of Ib and II Transhumance of sheep/goats. Military border post. Seasonal industries. More general pottery production. Distribution depot. Year round occupation by small number of quarter-masters. Mai ater s ste . Da . Cemetery. 1BCE-40 CE Part of II Transhumance of sheep/goats. Seasonal industries in decline in tandem with the declining royal estate in Jericho. Distribution depot in rapid decline as desert forts go out of use, end of year round occupation. Dam no longer maintained. Seasonal use of cemetery. Transhumance of sheep/goats. Some encouragement for processing of wool and hides. Cemetery 7 -6 Cent. BCE Herod Agrippa I 41-44 CE 50-68 CE II Minor restoration of border position and a return to some year round occupation. Transhumance of sheep/goats. Influx of refugees follo i g reakdo i rural se urit + re olutio aries 68-73 CE III Roman outpost. 117 73-132 CE Transhumance of sheep/goats. Garrison of pro-Roman Jews/Nabateans?? 132-135 CE Bar Kokhba revolutionaries. 135- s CE Cemetery. Transhumance of sheep/goats. Byzantine period Some transient occupation, particularly in Ein Feshka. 1947-1967 Jordanian military camps. 117 Cemetery. Cemetery in occasional use until at least the early Islamic period (e.g. grave in L118) For the possibility that this was manned by Jews re-settled by the Romans, see Doudna 2001: 744, and Taylor 2006. 74 ABBREVIATIONS AASOR Annual of the American Schools of Oriental Research BA Biblical Archaeologist BAR Int. S. British Archaeological Reports International Series BAIAS Bulletin of the Anglo-Israel Archaeological Society BASOR Bulletin of the American Schools of Oriental Research DJD Discoveries in the Judean Desert DSD Dead Sea Discoveries IAA Israel Antiquities Authority IAA Reports Israel Antiquities Authority Reports IEJ Israel Exploration Journal IES Israel Exploration Society JBL Journal of Biblical Literature JJS Journal of Jewish Studies JRA Supp. S. Journal of Roman Archaeology: Supplementary Series OUP Oxford University Press PEF Palestine Exploration Fund PEFQS Palestine Exploration Fund Quarterly Statement PEQ Palestine Exploration Quarterly QC The Qumran Chronicle RB Revue Biblique RevQ Revue de Qumran SBL Society of Biblical Literature STDJ Studies on the Texts of the Desert of Judah WUNT Wissenschaftliche Untersuchungen Zum Neuen Testament Bibliography (1959) The People of the Dead Sea Scrolls. Routledge & Kegan Paul, London. (1996) Allegro Qumran Collection: Supplement to the Dead Sea Scrolls on Microfiche.Vol. 2. G. J. Brooke (ed.), E. J. Brill & IDC, Manchester. Amar, Z. (1998) The Ash and the Red Material from Qumran. DSD 5: 1-15. Amar, Z., Lev, E., and Yanir, Z. (2005) Cereal Beer (Sheikhar) in Jewish Sources. Viennese Ethnomedicine Newsletter 8: 3-7. Ambraseys, N.N. (2006) Earthquakes and Archaeology. Journal of Archaeological Science 33: 10081016. Amit, D. (2002) Dating the High Level, Biyar and Herodium Aqueducts. Pp. 253-66 in D. Amit, J. Patrich and Y. Hirschfeld (eds.). Abegg, M. (1997) Who Ascended to Heaven? 4Q491, 4Q427, and the Teacher of Righteousness. Pp. 61-73 in C. Evans and P. Flint (eds.) Eschatology, Messianism, and the Dead Sea Scrolls. Eerdmans, Grand Rapids, MI. (2003) 1QSb and the Elusive High Priest. Pp. 316 in S. Paul et al. (eds.) Emanuel. Studies in Hebrew Bible, Septuagint, and Dead Sea Scrolls in Honor of Emanuel Tov. Brill, Leiden. Adler, E. (1903) Genizah. The Jewish Encyclopedia Vol. 5. New York. Adler, W. (1996) The Apocalyptic Survey of History Adapted by Christians: Daniel’s prophecy of 70 weeks. Pp. 201-38 in J.C. VanderKam and W. Adler (eds.) The Jewish Apocalyptic Heritage in Early Christianity. Assen, The Netherlands. Allegro, J. (1956) The Dead Sea Scrolls. Penguin, London. 75 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Avni, G., Dahari, U. & Kloner, A. (2008) The Necropolis of Bet Guvrin-Eleutheropolis. IAA Reports 36. Jerusalem. Avni, G. & Greenhut, Z. (1996) The Akeldama Tombs: Three Burial Caves in the Kidron Valley, Jerusalem. IAA Reports 1. Jerusalem. Bagatti, P.B. & Milik, J.T. (1958) Gli scavi del ‘Dominus Flevit’ I: La necropoli del periodo Romano. Publications of the Studium Biblicum 13. Jerusalem. Bailey, C. and Danin, A. (1981) Bedouin Plant Utilization in Sinai and the Negev. Economic Botany 35/2: 145-62. Baillet, M., Milik, J.T. and de Vaux, R. (1962) Les ‘petites grottes’ de Qumran. DJD 3. Clarendon Press, Oxford. Bar-Adon, P. (1977) Another Settlement of the Judean Desert Sect at ‘Ein el-Ghuweir on the Shores of the Dead Sea. BASOR 227: 1-26. (1981) The Hasmonean Fortesses and the status of Khirbet Qumran. Eretz Israel 15: 349-52 (Hebrew), 86* (English summary). (1981) In Desert Tents: From the Notes of a Hebrew Shepherd among the Bedouin Tribes (Hebrew). Jerusalem. (1989) Excavations in the Judean Desert. Atiqot 9 (Hebrew with English summary). IAA, Jerusalem. Barag, D. and Hershkovitz, M. (1994) Lamps from Masada. Pp. 7-78 in J. Aviram, G. Foerster, and E. Netzer (eds.) Masada IV: The Yigael Yadin Excavations 1963-1965. Final Reports. Jerusalem. Bar-Gal, G.K. (2006) Principles of the Recovery of Ancient DNA--What it Tells Us of Plant and Animal Domestication and the Origin of the Scroll Parchment. Pp. 41-50 in J. Gunneweg, C. Greenblatt and A. Adriaens (eds.) Bio- and material cultures at Qumran: papers from a COST Action G8 working group meeting held in Jerusalem, Israel on 22–23 May 2005. Fraunhofer IRB Verlag, Stuttgart. Bar-Kochva, B. (1999) The battle between Ptolemy Lathyrus and Alexander Jannaeus in the Jordan Valley and the dating of the Scroll of the War of the Sons of Light. Qatedrah le-toldot Eres Yisra’el el we-yissubah 93: 7-56. (English abstract of Hebrew) Bar-Nathan, R. (2002) Hasmonean and Herodian Palaces at Jericho Vol III: The Pottery. Jerusalem. (2006a) Masada VII: The Yigael Yadin Excavations 1963-1965. The Pottery. Jerusalem. (2006b) Qumran and the Hasmonean and Herodian Winter Palaces of Jericho: The Implications of the Pottery Finds on the Interpretation of the Settlement at Qumran. Pp. 263-77 in K. Galor, J.-B. Humbert and J. Zangenburg (eds.). Barthélemy, D. and Milik, J.T. (eds.) (1955) Qumran Cave 1. DJD I. Clarendon Press, Oxford. Amit, D., Patrich, J., and Hirschfeld, Y. (eds.) (2002) The Aqueducts of Israel. JRA Supp. S. Vol 46. Portsmouth, RI. First published in Hebrew 1989. Amusin (Amoussine), J. (1974) A propos de l’interprétation de 4Q161 (fragments 5-6 et 8). RevQ 31: 381-92. Angel, J. (2010) The Liturgical-Eschatological Priest of the Self-Glorification Hymn. RevQ 96: 585-605. Arenillas, M. and Castillo, J. (2003) Dams from the Roman Era in Spain. Analysis of Design Forms. Paper presented at the First International Congress on Construction History, Madrid, 20 th – 24th January, 2003. http://tinyurl.com/chfgpvt. Arensburg, B. & Belfer-Cohen, A. (1994) Preliminary Report on the Skeletal Remains for the ‘En Gedi Tombs’. Atiqot 24: 12*-14*. Ariel, D. (2002) The Coins from the Surveys and Excavations of Caves in the Northern Judean Desert. Atiqot 41/2: 281–304. Atkinson, K. (1999) On the Herodian Origin of Militant Davidic Messianism at Qumran: New Light from Psalm of Solomon 17. JBL 118: 435-60. (2004) Herod the Great as Antiochus Redidivus: Reading the Testament of Moses as an AntiHerodian Composition. Pp. 134-49 in C.A. Evans (ed.) Of Scribes and Sages: Early Jewish Interpretation and Transmission of Scripture. T & T Clark, Edinburgh. (2008) Review of K. Galor, J.-B. Humbert, and J. Zangenberg (eds.) Qumran, The Site of the Dead Sea Scrolls: Archaeological Interpretations and Debates (Proceedings of a Conference Held at Brown University, November 17-19, 2002). Review of Biblical Literature http://www.bookreviews.org/pdf/5905_6266.pdf (accessed – 13/11/11). Aviam, M. (2004) Jews, Pagans and Christians in the Galilee: 25 Years of Archaeological Excavations and Surveys, Hellenistic to Byzantine Periods. University of Rochester Press, Rochester, NY. Avigad, N. (1976) Beth She’arim - Report on the Excavations during 1953-1958, Vol. III. Catacombs 12-23. IAA, Jerusalem. Avner, R. & Gendelman, P. (2007) Caesarea. Hadashot Arkeologiyot 119. http://www.hadashot-esi.org.il /report_detail_eng.asp?id=470&mag_id=112 Avni, G. (1996) Nomad, Farmers and Town-Dwellers: Pastoralists-Sedentist Interaction in the Negev Highlands, Sixth-Eighth Centuries CE. Supplement to the Archaeological Survey of Israel, Jerusalem. (1997) The Necropoleis of Jerusalem and Beit Guvrin in the 4th – 7th Centuries CE as an Example of Urban Cemeteries in the Land of Israel in the Late Roman and Byzantine Period. PhD Thesis (Hebrew). Hebrew University. Jerusalem. (2009) Who were interred in the Qumran cemetery? The Ethnic Identity of Ancient Populations in light of Archaeological Findings in Burial Sites. Cathedra 131: 43-63 (Hebrew). 76 BIBLIOGRAPHY Cargill, R. (2009) Qumran through (real) time; a virtual reconstruction of Qumran and the Dead Sea Scrolls. Gorgias press, Piscataway, NJ. Carmi, I. (2002). Are the 14C Dates of the Dead Sea Scrolls Affected by Castor Oil Contamination? Radiocarbon 44: 213-16. Charlesworth, J. (2000) XJoshua. Pp. 231-39 in J. Charlesworth et al. (eds.) Miscellaneous Texts from the Judaean Desert. DJD 38. Clarendon Press, Oxford. Clamer, C. (2003) Jewellery Finds from the Cemetery. Pp. 171-179 in J.-B. Humbert & J. Gunneweg (eds.) Clark, K.W. (1963) The Posture of Ancient Scribes. BA 26/2: 63-72. Clermont-Ganneau, C. (1896) Archaeological Researches in Palestine during the Years 1873-1874 II. PEF, London. Cloke, C. (2007) Aqua Nabataea et Aqua Romana: Signs of Cultural Change in the Waterworks of Ancient Arabia. Unpublished MA thesis, Brown U. Providence, RI. Collins, J.J. (2009a) Beyond the Qumran Community: Social Origins. DSD 16: 351-69. (2009b) The Angelic Life. Pp. 291-310 in T. Seim and J. Økland (eds.) Metamorphoses: Resurrection, Body, and Transformative Practices in Early Christianity. Walter de Gruyter, Berlin. (2010) Beyond the Qumran Community: The Sectarian Movement of the Dead Sea Scrolls. Grand Rapids, MI, and Cambridge, UK. Conder, C.R. and Kitchener, H.H. (1883) The Survey of Western Palestine; Vol III, Judeah. London. Cook, E. (1994) Solving the Mysteries of the Dead Sea Scrolls. Zondervan, Grand Rapids, MI. Cotton, H.M. and Yardeni, A. (eds.) (1997) Aramaic, Hebrew, and Greek Documentary Texts from Nahal Hever and Other Sites, with an Appendix Containing Alleged Qumran Texts (The Seiyal Collection II). DJD 27. Clarendon Press, Oxford. Crawford, S.W. (2011a) The Pentateuch as Found in the Pre-Samaritan Texts and 4QReworked Pentateuch. Pp. 123-36 in H. von Weissenberg, J. Pakkala, and M. Marttila (eds.) Changes in Scripture. Rewriting and Interpreting Authoritative Traditions in the Second Temple Period. Berlin/New York. (2011b) Qumran: Caves, Scrolls, and Buildings. Pp. 253-73 in E. Mason (ed.) A Teacher for All Generations: Essays in Honor of James C. VanderKam. Brill, Leiden. Cresswell, W. (2006) Diary of a Victorian Gardener, William Cresswell and Audley End. London. Cross, F.M. (1958) The Ancient Library of Qumran and Modern Biblical Studies. New York. (1961) The development of the Jewish scripts. Pp. 133-202 in G.E. Wright (ed.) The Bible and the Ancient Near East: Essays in Honor of W.F. Albright. Doubleday, Garden City, NY. Bar-Tzvi, S., Abu-Rabia, A., and Kressel, G.M. (1998) The Charm of Graves: Mourning Rituals and Tomb Worshipping among the Negev Bedouin (Hebrew). Tel-Aviv. Baumgarten, J.M. (1987) The Calendars of the Book of Jubilees and the Temple Scroll. Vetus Testamentum 37: 71-8. (1996) (ed.) Qumran Cave 4. XIII: The Damascus Document. DJD 18. Clarendon Press, Oxford. Bedal, L.-A. (2001) A Pool Complex in Petra’s City Center. BASOR 324: 23-41. (2002) Desert Oasis: Water Consumption and Display in the Nabataean Capital. Near Eastern Archaeology 65/4: 225-35. (2004) The Petra Pool-complex: A Hellenistic Paradeisos in the Nabataean Capital. Gorgias Press, Piscataway, NJ. Beit-Aryeh, Y. (2003) Archaeological Survey of Israel: Map of Tel Malkhata. Jerusalem. Bellwald, U. (2004) Streets and Hydraulics: The Petra National Trusts Siq Project in Petra 1996–1999. The Archaeological Results. Pp. 73-94 in H.-D. Bienert and J. Häser (eds.) Men of Dikes and Canals: The Archaeology of Water in the Middle East, Petra, Wadi Musa, 15–20 June 1999. Verlag Marie Leidorf GmbH, Rahden, Germany. Bennett, C.M. (1965) Tombs of the Roman Period. Pp. 516-45 in K.M. Kenyon, Excavation in Jericho II: The Tombs excavated in 1955-58. London. Berrin, S. (2004) The Pesher Nahum Scroll from Qumran. Brill, Leiden. Bouzard, W.C. (2003) The Date of the Psalms Scroll from the Cave of Letters (5/6HevPs) Reconsidered. DSD 10: 319-37. Bresler, E. (1981) Irrigation and Soil Salinity. Pp. 65-102 in D. Yaron (ed.) Salinity in Irrigation and Water Resources. New York. Brooke, G. (1994) The Pesharim and the Origins of the Dead Sea Scrolls. Pp. 339-353 in M. Wise et al. (eds.). Broshi, M. (2004) The Dead Sea Scrolls, the Sciences and New Technologies. DSD 11: 133-42. (2007) Essenes at Qumran? A Rejoinder to Albert Baumgarten. DSD 14: 25-33. Broshi, M. and Eshel, H. (1999) Residential Caves at Qumran. DSD 6: 328-47. (2004) Three Seasons of Excavations at Qumran. Journal of Roman Archaeology 17: 321-31. Burrell, B. (2009) Caesarea on Sebastos: Urban Structures and Influences. Pp. 217-33 in D. Jacobson and N. Kokkinos (eds.). Caldararo, N. (1995) Storage Conditions and Physical Treatments Relating to the Dating of the Dead Sea Scrolls. Radiocarbon 37: 21-32. Canby, S.R. (2009) Shah ‘Abbas: the Remaking of Iran. British Museum, London. Cansdale, L. (1997) Qumran and the Essenes: A Reevaluation of the Evidence. Tübingen. 77 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS 34 in E. Chazon et al. (eds.) Reworking the Bible: Apocryphal and Related Texts at Qumran. Brill, Leiden. (2012) Israeli Scholarship on the Qumran Community. Pp. 237-80 in D. Dimant and I. Kottsieper (eds.) The Dead Sea Scrolls in Scholarly Perspective: A History of Research. Brill, Leiden. Dion, M.-F. (2011) L’identité d’Éphraïm et Manassé dans le Pésher de Nahum (4Q169). Pp. 405-27 in P. Flint et al. (eds.) Celebrating the Dead Sea Scrolls. A Canadian Collection. SBL, Atlanta. Donceel-Voûte, P. (1992) ‘Coenaculum’: La salle à l’étage du locus 30 à Khirbet Qumrân sur la Mer Morte. Pp. 61-84 in R. Gyselen with M. BernusTaylor et al. (eds.) Banquets d’Orient. Res Orientalis 4. Peeters, Leuven. (1994) Les ruines de Qumran reinterprétées. Archeologia 298: 24-35. Donceel, R. (1998) Poursuite des travaux de publication du matériel archéologique de Khirbet Qumrân. Les lampes en terre-cuite. Pp. 87-104 in Z.J. Kapera (ed.) Mogilany 1995. Papers on the Dead Sea Scrolls offered in memory of Aleksy Klawek. Qumranica Mogilanensia 15. Cracow. (2005) Khirbet Qûmran (Palestine) le Locus 130 et les ‘ossements sous jarre’. QC 13/1: 7-66. Donceel, R. and Donceel-Voûte, P. (1994) The Archaeology of Khirbet Qumran. Pp. 1-38 in M. Wise et al. (eds.). Doudna, G. (1998) Dating the Scrolls on the Basis of Radiocarbon Analysis. Pp. 430-71 in P. Flint and J. VanderKam (eds.) The Dead Sea Scrolls after Fifty Years. Vol. I. Brill, Leiden. (1999) Redating the Dead Sea Scrolls found at Qumran. QC 8/4: 2-96. (2001) 4Q Pesher Nahum. A Critical Edition. Sheffield Academic Press, Sheffield. (2002) 4Q Pesher Hoseab: Reconstruction of Fragments 4, 5, 18, and 24. DSD 10/3: 338-58. (2005) Who is the Lion of Wrath of Pesher Nahum? A Brief Analysis. Pp. 87-105 in M. Müller and T. Thompson (eds.) Historie og konstruktion. Festskrift til Niels Peter Lemche i anledgning af 60 års fødselsdagen den 6. September 2005. Museum Tusculanum Press, Copenhagen. (2006) The Legacy of an Error in Archaeological Interpretation: the Dating of the Qumran Cave Scroll Deposits. Pp. 147-57 in K. Galor, J.-B. Humbert and J. Zangenberg (eds.). (2007) Ostraca KhQ1 and KhQ2 from the Cemetery at Qumran: A New Edition. Pp. 59-116 in E. Ben Zvi (ed.) Perspectives on Hebrew Scriptures II. Gorgias Press, Piscataway, NJ. (2011) Allusions to the End of the Hasmonean Dynasty in Pesher Nahum (4Q169). Pp. 259-78 in G. Brooke and J. Høgenhaven (eds.) The Mermaid and the Partridge. Essays from the (1962) The Palaeographical Dating of the Copper Document. Pp. 217-21 in M. Baillet et al. Les ‘Petites Grottes’ de Qoumrân. DJD 3. Clarendon Press, Oxford. (1995) The Ancient Library of Qumran. 2nd ed. Fortress Press, Minneapolis, MN. (1998) Palaeography and the Dead Sea Scrolls. Pp. 379-402 in P.W. Flint and J.C. VanderKam (eds.) The Dead Sea Scrolls after Fifty Years. Vol. I. Brill, Leiden. (2007) Two Aramaic Ostraca from En-Gedi. Pp. 377-380 in E. Stern, En-Gedi Excavations I. Jerusalem. Cross, F.M. and Milik, J.T. (1956) Explorations in the Judean Buqe’ah. BASOR 142: 5-17. Crowfoot, G.M. (1955) Linen Textiles. Pp. 18-38 in D. Barthélemy and J.T. Milik (eds.) Qumran Cave 1. DJD I. Clarendon Press, Oxford. Daise, M. (2005) ‘The Days of Sukkot of the Month of Kislev’: The Festival of Dedication and the Delay of Feasts in 1QS 1.13-15. Pp. 119-28 in G. Boccaccini (ed.) Enoch and Qumran Origins. New Light on a Forgotten Connection. Eerdmans, Grand Rapids, MI. Danin, A. (1983) Desert Vegetation of Israel and Sinai. Jerusalem. Daube, D. (1990) On Acts 23: Sadducees and Angels. JBL 109/3: 493-97. Davies, P., Brooke, G., and Callaway, P. (2002) The Complete World of the Dead Sea Scrolls. Thames and Hudson Ltd, London. Davila, J. (1994) 4QGenb. Pp. 31-8 in E. Ulrich and F.M. Cross (eds.) Qumran Cave 4. VII. Genesis to Numbers. DJD 12. Clarendon Press, Oxford. Del Medico, H. (1957) L’Enigme des Manuscripts de la Mer Morte. Paris. Dembskey, E. (2009) The Aqueducts of Ancient Rome. Unpublished MA thesis, U. of South Africa. http://tinyurl.com/c9kok6m DeQueker, L. (1986) I Chronicles XXIV and the Royal Priesthood of the Hasmoneans. Pp. 94-106 in A.S. van der Woude (ed.) Crises and Perspectives. Studies in Ancient Near Eastern Polytheism, Biblical Theology, Palestinian Archaeology and Intertestamental Literature. Brill, Leiden. Diaz-Andreu, M. & Champion, T. (eds.) (1996) Nationalism and Archaeology in Europe. Boulder, CO. Dimant, D. (1995) The Qumran Manuscripts: Contents and Significance. Pp. 23-58 in D. Dimant and L.H. Schiffman (eds.) The Time to Prepare the Way in the Wilderness: Papers on the Qumran Scrolls by Fellows of the Institute for Advanced Studies of the Hebrew University. Jerusalem 1989-1990. STDJ 16. Brill, Leiden. (2001) Qumran Cave 4. XXI. Parabiblical Texts, Part 4: Pseudo-Prophetic Writings. DJD 30. Clarendon Press, Oxford. (2005) Between Sectarian and Non-Sectarian: The Case of the Apocryphon of Joshua. Pp. 10578 BIBLIOGRAPHY Eshel, H., Broshi, M., Freund, R. and Schultz, B. (2002) New Data on the East Cemetery of Khirbet Qumran. DSD 9: 135-65. Eshel, H. and Greenhut, Z. (1993) Hiam el-Sagha: A Cemetery of the Qumran Type, Judaean Desert. RB 100: 252-59. Falk, D. (1998) Daily, Sabbath, & Festival Prayers in the Dead Sea Scrolls. Brill, Leiden. Fischer, M., Gichon, M. and Tal, O. (2000) ‘En Boqeq: Excavations in an Oasis on the Dead Sea. Vol. II. Mainz am Rhine, Germany. Fitzmyer, J. (2000) 4QpapHistorical Text C and D. Pp. 281-86 in S. Pfann et al. (eds.) Qumran Cave 4: XXVI. Cryptic Texts and Miscellanea, Part I. DJD 36. Clarendon Press, Oxford. Fletcher-Louis, C.H.T. (2002) All the Glory of Adam. Liturgical Anthropology in the Dead Sea Scrolls. Brill, Leiden. Flint, P. (2000) 5/6HevPsalms. Pp. 141-66 in J. Charlesworth et al., Miscellaneous Texts from the Judaean Desert. DJD 38. Clarendon Press, Oxford. Flohr, M. (2005) Ars Fullonia. Interpreting and Contextualising Roman Fulling. Pp. 59–63 in C. Briault, J. Green, A. Kaldelis and A. Stellatou (eds.) Symposium on Mediterranean Archaeology. BAR Int. S. 1391. Oxford. Forbes, R.J. (1956) Studies in Ancient Technology, Vol. IV. Brill, Leiden. (1957) Studies in Ancient Technology, Vol. V. Brill, Leiden. Fraade, S. (2012) Review of Aharon Shemesh, Halakhah in the Making: The Development of Jewish Law from Qumran to the Rabbis. In Journal for the Study of Judaism in the Persian, Hellenistic, and Roman Period 43: 131-35. Galavo-Sobrinho, C.R. (1995) Funerary Epigraphy and the Spread of Christianity to the West. Anthenaeum 83: 431-62. Galor, K. (2003) Plastered Pools: A New Perspective. Pp. 291-319 in J.-B. Humbert and J. Gunneweg (eds.) Galor, K., Humbert, J.-B., Zangenberg, J. (2006) The Site of the Dead Sea Scrolls: Archaeological Interpretations and Debates. Proceedings of the Conference Held at Brown University, November 17-19, 2002. Brill, Leiden. Garbrecht, G. and Peleg, Y. (1994) The Water Supply of the Desert Fortresses in the Jordan Valley. BA 57/3: 161-70. García Martínez, F. (1996) 4QMMT in a Qumran Context. Pp. 15-27 in J. Kampen and M.J. Bernstein (eds.) Reading 4QMMT: New Perspectives on Qumran Law and History. Scholars Press, Atlanta, GA. Gazit, D. (1996) Archaeological Survey of Israel: Map of Urim. IAA, Jerusalem. Gillihan, Y. (2012) Civic Ideology, Organization, and Law in the Rule Scrolls. A Comparative Study of the Covenanters’ Sect and Contemporary Copenhagen Conference on Revising Texts from Cave Four. Brill, Leiden. Douglas, M. (1998) Power and Praise in the Hodayot: A Literary Critical Study of 1QH 9:1-18:14. PhD thesis, Divinity School, U. of Chicago. Chicago. (1999) The Teacher Hymn Hypothesis Revisited: New Data for an Old Crux. DSD 6: 239-66. Drake, C.F.T. (1881) The Survey of Western Palestine: Special Papers. PEF, London. Dupont-Sommer, A. (1952) The Dead Sea Scrolls: a Preliminary Survey. trans. from the 1950 publication in French by M. Rowley. Oxford. Dupont-Sommer, A. (1954) The Jewish Sect of Qumran and the Essenes. trans. from the 1952 publication in French by R.D. Barnett. London. Eakins, J.K. (1993) Tell el-Hesi: The Muslim Cemetery in Fields V and VI/LX (Stratum II). The Joint Archaeological Expedition to Tel el-Hesi, 5. Winona Lake, IN. Eisenman, R. (1983) Maccabees, Zadokites, Christians, and Qumran: A New Hypothesis of Christian Origins. Brill, Leiden. (1986) The ‘Three Nets of Belial’ in the Zadokite Document and Balla‘/Bela’ in the Temple Scroll. Pp. 87-94 in R. Eisenman, James the Just in the Habakkuk Pesher. Brill, Leiden. (1994) Theory of Judeo-Christian Origin: The Last Column of the Damascus Document. Pp. 355-70 in M. Wise et al. (eds.). Eisenman, R. and Wise, M. (1992) The Dead Sea Scrolls Uncovered. Element, Shaftesbury, Dorset. Elledge, C.D. (2006) Life and Death in Early Judaism: The Evidence of Josephus. Mohr Siebeck, Tübingen. (2007) The Prince of the Congregation: Qumran ‘Messianism’ in the Context of Milhāmâ. Pp. 178-207 in M. Davis and B. Strawn (eds.) Qumran Studies. New Approaches, New Questions. Eerdmans, Grand Rapids, MI. Eshel, H. (1996) A Note on a Recently Published Text: the ‘Joshua Apocryphon’. Pp. 89-93 in M. Porthuis and C. Safrai (eds.) The Centrality of Jerusalem: Historical Perspectives. Kok Pharos, Kampen, NL. (2001) The Kittim in the War Scroll and in the Pesharim. Pp. 29-44 in D. Goodblatt, A. Pinnick and D. Schwartz (eds.) Historical Perspectives: From the Hasmoneans to Bar Kokhba in Light of the Dead Sea Scrolls. Proceedings of the Fourth International Symposium of the Orion Center, 27-31 January 1999. Brill, Leiden. (2008) The Dead Sea Scrolls and the Hasmonean State. Eerdmans, Grand Rapids, MI. Eshel, E., Eshel, H. and Yardeni, A. (1992) A Qumran Composition Containing Part of Psalm 154 and a Prayer for the Welfare of King Jonathan and his Kingdom. IEJ 42: 199-229. Eshel, H. and Eshel, E. (2000) 4Q448, Psalms 154 (Syriac), Sirach 48:20, and 4QpIsaa. JBL 119/4: 645-59. 79 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Haas, H. and Nathan, H. (1968) Anthropological Survey on Human Skeletal Remains from Qumran. RevQ 6: 345-52. Haber, S. (2008) ‘They Shall Purify Themselves’: Essays on Purity in Early Judaism. A. Reinhartz (ed.). Society of Biblical Literature, Atlanta, GA. Hachlili, R. (1993) Burial Practices at Qumran. RevQ 16: 247-264. (2000) The Qumran Cemetery: A Reconsideration. Pp. 661-72 in L. Schiffman, E. Tov and J. VanderKam (eds.) The Dead Sea Scrolls: Fifty Years after their Discovery. Jerusalem. Hachlili, R. & Killebrew, A. (1999) Jericho: The Jewish Cemetery of the Second Temple Period. IAA Reports 7. Jerusalem. Hadas, G. (2008) Dead Sea Sailing Routes during the Herodian Period. BAIAS 26: 31-6. (2011) Hunting Traps around the Oasis of En Gedi. IEJ 61: 2-11. Hadas, G., Liphshitz N. and Bonani, G. (2005) Two Ancient Wooden Anchors from Ein Gedi, on the Dead Sea, Israel. The International Journal of Nautical Archaeology 34: 307-15. Harding, G.L. (1949) The Dead Sea Scrolls. PEQ 81: 112-16. (1952) Khirbet Qumran and Wady Muraba‛at. PEQ 84: 104-09. (1959) The Antiquities of Jordan. London. Har-El, M. (2000) Agriculture. Pp. 13-16 and Judea: Geography. Pp. 444-46 in L. Schiffman and J. Vanderkam (eds.) Encyclopaedia of the Dead Sea Scrolls. Oxford. Hawes, A. (2008) A Handful of Honey. London. Heinrich, L. (1992) The Magic of Linen: Flax Seed to Woven Cloth. Victoria, BC. Hempel, C. (2010) The Context of 4QMMT and Comfortable Theories. Pp. 275-92 in C. Hempel (ed.) The Dead Sea Scrolls: Texts and Context. STDJ 89. Brill, Leiden. Hengel, M. (1977) Crucifixion in the Ancient World and the Folly of the Message of the Cross. Philadelphia. Hepper, F.N. and Taylor, J. (2004) Date Palms and Opobalsam in the Madaba Mosaic Map. PEQ 136: 35-44. Hershkovitz, M. (2009) Herodian Pottery. Pp. 267-78 in D. Jacobson and N. Kokkinos (eds.). Hezser, C. (2005) Jewish Slavery in Antiquity. OUP, Oxford. Hirschfeld, Y. (2000) A Settlement of Hermits above ‘Ein Gedi. Tel Aviv 27: 103-55. (2004a) Excavations at Tiberias, 1989-1994. IAA Reports 22, Jerusalem. (2004b) Qumran in Context: Reassessing the Archaeological Evidence. Peabody, MA. Hjelm, I. (2000) The Samaritans and Early Judaism. A Literary Analysis. Sheffield Academic Press, Sheffield. Hodge, T. (1992) Roman Aqueducts and Water Supply. Duckworth, London. Voluntary Associations in Political Context. Brill, Leiden. Ginzberg, L. (1922) Eine unbekannte jüdische Sekte. Published in English in 1976 as An Unknown Jewish Sect. Jewish Theological Seminary of America, New York. Glessmer, U. (1999) Calendars in the Qumran Scrolls. Pp. 213-78 in P. Flint and J. VanderKam (eds.) The Dead Sea Scrolls after Fifty Years, Vol. II. Brill, Leiden. Gmirkin, R. (1996) The War Scroll and Roman Weaponry Reconsidered. DSD 3: 89-129. (1998) Historical Allusions in the War Scroll. DSD 5/2: 172-214. (2006) Berossus and Genesis, Manetho and Exodus. Hellenistic Histories and the Date of the Pentateuch. T & T Clark, New York. (2013 forthcoming) Greek Evidence for the Hebrew Bible. in P. Wajdenbaum and T.L. Thompson (eds.) When Japhet Dwelt in the Tents of Shem: a Hellenistic Bible. Acumen Publishing, Durham, UK. Golb, N. (1980) The Problem of Origin and Identification of the Dead Sea Scrolls. Proceedings of the American Philosophical Society 124/1: 1-24. (1995) Who wrote the Dead Sea Scrolls? Scribner, New York. Goodblatt, D. (1994) The Monarchic Principle. Tübingen. Goodman, M. (2010) Constructing Ancient Judaism from the Scrolls. Pp. 81-91 in T. Lim and J. Collins (eds.). Gorzalczany, A. (2007) The Kefar Saba Cemetery and Differences in Orientation of Late Islamic Burials from Israel/Palestine. Levant 39: 71-79. (2010) Baqa-el-Gharbiya Area: A Roman Period Cemetery and other finds. Atiqot 64: 105-36 (Hebrew) 161-2 (English summary). Grabbe, L. (2000) Judaic Religion in the Second Temple Period. Belief and Practice from the Exile to Yavneh. Routledge, New York. (2003) Were the Pre-Maccabean High Priests ‘Zadokites’? Pp. 205-15 in M. Hauge (ed.) Between Sheol and Temple. Sheffield. (2007) When Is a Sect a Sect—or Not? Groups and Movements in the Second Temple Period. Pp. 114-32 in D. Chalcraft (ed.) Sectarianism in Early Judaism: Sociological Advances. London. (2008) Hecataeus of Abdera and the Jewish Law: the Question of Authenticity. Pp. 613-26 in I. Kottsieper et al. (eds.) Berührungspunkte: Studien zur Sozial- und Religionsgeschichte Israels und seiner Umwelt. Festschrift für Rainer Albertz zu seinem 65. Ugarit-Verlag, Münster. Greenhut, Z. (1992) The ‘Caiaphas’ Tomb in North Talpiot, Jerusalem. Atiqot 21: 63-71. Gunneweg, J. and Balla, M. (2003) Neutron Activation Analysis: Scroll Jars and Common Ware. Pp. 353 in Humbert and Gunneweg (eds.). Guvrin, Y. (1992) Archaeological Survey of Israel: Map of Nachal Yatir. IAA, Jerusalem. 80 BIBLIOGRAPHY Scholarly Perspective: A History of Research. Brill, Leiden. Jeremias, J. (1969) Jerusalem in the Time of Jesus. (Trans. Cave and Cave.) Fortress Press, London. Johnson, M.J. (1997) Pagan-Christian Burial Practices of the Fourth Century: Shared Tombs? Journal of Early Christian Studies 5: 37-59. Jokiranta, J. (2001) ’Sectarianism’ of the Qumran ‘Sect’: Sociological Notes. RevQ 20/2: 223-39. (2010) Sociological Approaches to Qumran Sectarianism. Pp. 200-31 in T. Lim and J. Collins (eds.). Jones, S. (1997) The Archaeology of Ethnicity: Constructing Identities in the Past and Present. London & New York. Kallner-Amiran, D. (1950) A Revised Earthquake Catalogue of Palestine. IEJ 1: 223-46. Kamash, Z. (2010) Archaeologies of Water in the Roman Near East. Gorgias Press, NJ. Kane, S. (ed.) (2003) The Politics of Archaeology and Identity in Global Context. Boston, MA. Karcz, I. and Kafri, U. (1978) Evaluation of Supposed Archaeoseismic Damage in Israel. Journal of Archaeological Science 5: 237-53. Kartveit, M. (2009) The Origin of the Samaritans. Brill, Leiden. Kelso, J.L. and Baramki, D.C. (1955) Excavation at New Testament Jericho and Khirbet En Nitla. AASOR 29-30. Cambridge, MA. Klawans, J. (2007) Sadducees, Zadokites, and the Wisdom of Ben Sira. Pp. 261-75 in D. Capes et al., Israel’s God and Rebecca’s Children. Essays in Honor of Larry W. Hurtado and Alan F. Segal. Waco, TX. (2012) Josephus and the Theologies of Ancient Judaism. OUP, Oxford. Kletter, R. (2006) Can the Proto-Israelite Please Stand Up? Notes on the Ethnicity of Iron Age Israel and Judah. Pp. 573-86 in A.M. Meir & P. Miroschedji (eds.) ‘I Will Speak the Riddles of Ancient Times’: Archaeological and Historical Studies in Honor of Amihai Mazar, II. Winona Lake, IN. Klinghardt, M. (1994) The Manual of Discipline in the Light of Statutes of Hellenistic Association. Pp. 251-67 in M. Wise et al. (eds.). Kloner, A. (1998) Water-Retaining Dams in the Northeast Negev Highlands. BAIAS 16: 7-32 Kloner, A and Zissu, B. (2007) The Necropolis of Jerusalem in the Second Temple Period. Leuven. Kokkinos, N. (2007) The Royal Court of the Herods. Pp. 279-304 in N. Kokkinos (ed.) The World of the Herods: Vol. I of the International Conference ‘The World of the Herods and the Nabataeans’ held at the British Museum 17-19 April 2001. Knohl, I. (2000) The Messiah Before Jesus. The Suffering Servant of the Dead Sea Scrolls. University of California Press, Berkeley, CA. (2002) New Light on the Copper Scroll and 4QMMT. Pp. 233-56 in G.J. Brooke and P.R. Hodges, H. (1989) Artifacts. An Introduction to Early Materials and Technology. 3rd ed. John Baker, London. Hogeterp, A. (2009) Expectations of the End: A Comparative Traditio-Historical Study of Eschatological, Apocalyptic, and Messianic Ideas in the Dead Sea Scrolls and the New Testament. Brill, Leiden. Horbury, W. (2003) Messianism Among Jews and Christians. Biblical and Historical Studies. Continuum, London. Hornell, J. (1938) The Coracles of the Tigris and Euphrates. Pp. 153-59 in J. Hornell, British Coracles and Irish Curraghs; with a Note on the Quffah of Iraq. Hornell, London. (1970) Water Transport: Origins and Early Evolution. Hornell, London (orig. pub. 1946). Humbert, J.-B. (1994) L’espace sacré à Qumran. Propositions pour l’archéologie. RB 101: 161214. (2003) The Chronology During the First Century B.C. De Vaux and his Method: A Debate. Pp. 425-43 in J.-B. Humbert and J. Gunneweg (eds). (2006) Some Remarks on the Archaeology of Qumran. Pp. 19-39 in K. Galor, J.-B. Humbert and J. Zangenberg (eds.). Humbert, J.-B. and Chambon, A. (1994) Fouilles de Khirbet Qumrȃn et de ‘Ain Feshka. Vol. I. Fribourg. (2003) Excavations of Khirbet Qumran and ‘Ain Feshka. Vol Ia (English edition by Pfann). Fribourg. Humbert, J.-B. and Gunneweg J. (eds.) (2003) Khirbet Qumran et ‘Ain Feshkha II. Studies of Anthropology, Physics and Chemistry. Fribourg, Switzerland. Hutchesson, I. (1999) 63 BCE: A Revised Dating for the Depositation of the Dead Sea Scrolls. RevQ 8: 177-94. Ilan, Z. and Amit, D. (2002) The Aqueduct of Qumran. Pp. 380-86 in D. Amit, J. Patrich and Y. Hirschfeld (eds.). Israeli, R. (2008) Piracy in Qumran. The Battle over the Scrolls of the Pre-Christ Era. Transaction Publishers, New Brunswick, NJ. Itah, M., Kam, Y., and Ben Haim, R., (2002) Region X: Survey and Excavation of Caves along the Fault Escarpment South of Almog Junction and West of Qalya. Pp. 169-76 in Surveys and Excavations of Caves in the North Judean Desert 1993. Atiqot 41/2. Jacobson, D. and Kokkinos, N. (eds.) (2009) Herod and Augustus:Papers presented at the IJS Conference, 21st – 23rd June 2005. Brill, Leiden. Jacobus, H. (2010) 4Q318: A Jewish Zodiac Calendar at Qumran? Pp. 365-95 in C. Hempel (ed.) The Dead Sea Scrolls: Texts and Context. STDJ 90. Brill, Leiden. Jassen, A. (2012) American Scholarship on Jewish Law in the Dead Sea Scrolls. Pp. 101-54 in D. Dimant (ed.) and I. Kottsieper, The Dead Sea Scrolls in 81 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS (1993) Dead Sea Scrolls. http://www.manfredlehmann.com/news/news_de tail.cgi/41/0 (accessed 24/2/13) (1994) Scroll Study Safe in Jewish Hands. http://www.manfredlehmann.com/news/news_de tail.cgi/116/0 (accessed 24/2/13) Lemaire, A. (2000) Le roi Jonathan à Qoomrân (4Q448, B-C). Pp. 55-68 in E.M. Laperrousaz (ed.) Qoomrân et les manuscrits de la Mer Morte. Un cinquantenaire. Les Éditions du Cerf, Paris. Le Moyne, J. (1972) Les sadducéens. Librairie Lecoffre, Paris. Lev, E., and Amar, Z. (2008) Practical Materia Medica of the Medieval Eastern Mediterranean according to the Cairo Genizah. Brill, Boston/Leiden. Levine, L. (1975) Roman Caesarea: An ArchaeologicalTopographic Study. Qedem 2. Hebrew U., Jerusalem. Levine, Y. (1985) The Findings of Beit She’arim and their Importance in the Research of the Mishnah and Talmud Periods. Eretz-Israel 18: 277-81 (Hebrew). Lichtenburger, A. (2009) Was Romanisation a Goal of the Building Policy of Herod? Pp. 43-62 in Jacobson and Kokkinos (eds.). Lilley, J.M. et al. (1994) The Jewish Burial Ground at Jewbury. Archaeology of York, 12: The Medieval Cemeteries. York. Lim, T. and Collins, J. (eds.) (2010) The Oxford Handbook of the Dead Sea Scrolls. OUP, Oxford. Liver, J. (1967) The Sons of Zadok the Priest. RevQ 6: 330. Lonnqvist, K., and Lonnqvist, M. (2002) Archaeology of the Hidden Qumran: The New Paradigm. Helsinki. (2006) The Numismatic Chronology of Qumran. Fact and Fiction. The Numismatic Chronicle 166: 121-65. Lynch, W. Lt. (1849) Narrative of the U.S. Expedition in the River Jordan and the Dead Sea. Lea and Blanchard, PA. Maccoby, H. (2000) Jesus the Pharisee. SCM Press, London. MacDonald, F.J. (1982) Crowdie and Cream. London. Magen, Y. and Peleg Y. (2006) Back to Qumran: Ten Years of Excavation and Research, 1993-2004. Pp. 55-113 in K. Galor, J.-B. Humbert and J. Gunneweg (eds.). (2007) The Qumran Excavations 1993-2004: Preliminary Report. IAA, Jerusalem. Mandel, P. (1993) On the ‘Duplicate Copy’ of the Copper Scroll (3Q15). RevQ 16: 69-76. Magness, J. (2002a) The Archaeology of Qumran and the Dead Sea Scrolls. Grand Rapids, MI and Cambridge, UK. (2002b) Women at Qumran? Pp. 89-123 in. L.V. Rutgers (ed.) What Athens has to do with Jerusalem: Essays on Classical, Jewish and Davies (eds.) Copper Scroll Studies (2002). Continuum, London. (2009) Melchizedek: A Model for the Union of Kingship and Priesthood in the Hebrew Bible, 11QMelchizedek, and the Epistle to the Hebrews. Pp. 255-66 in R. Clements and D. Schwartz (eds.) Text, Thought, and Practice in Qumran and Early Christianity. Brill, Leiden. Kohl, P.L. (1998) Nationalism and Archaeology: On the Constructions of Nations and the Reconstruction of the Remote Past. Annual Review of Archaeology 27: 223-46. Kohl, P.L. and Fawcett, C. (eds.) (1996) Nationalism, Politics and the Practice of Archaeology. Cambridge, UK. Kohl, P.L., Kozelsky, M. and Ben Yehuda, N. (eds.) (2007) Selective Remembrances: Archaeology in the Construction, Commemoration, and Consecration of National Pasts. U. of Chicago Press, Chicago, IL. Koren, Z. (1995) Analysis of the Masada Textile Dyes. Pp. 257-64 in Masada: The Yigael Yadin Excavations 1963-63. The Final Reports. Vol. IV. IES, Jerusalem. (2011) Royal Purple Dye from Masada. PEQ 143/1: 2-4. Kraft, R.A. (2009) Pliny on Essenes, Pliny on Jews. Pp. 199-207 in R.A. Kraft, Exploring the Scripturesque: Jewish Texts and their Christian Contexts. Brill, Leiden. Kratz, R. (2011) Der Penal Code und das Verhältnis von Serekh ha-Yachad (S) und Damaskusschrift (D). RevQ 25/2: 199-227. Kutscher, E.Y. (1974) The Language and Linguistic Background of the Isaiah Scroll (1QIsaa). Brill, Leiden. Lange, A. (2003) The Parabiblical Literature of the Qumran Library and the Canonical History of the Hebrew Bible. Pp. 305-21 in S. Paul et al. (eds.) Emanuel. Studies in Hebrew Bible, Septuagint, and Dead Sea Scrolls in Honor of Emanuel Tov. Brill, Leiden. LaSor, W.S. (1956) Amazing Dead Sea Scrolls. Moody, Chicago, IL. Lefkovits, J.K. (2000) The Copper Scroll. 3Q15: A Reevaluation. A New Reading, Translation, and Commentary. Brill, Leiden. Leguilloux, M. (2004) L’identification des tanneries antiques par l’archéologie et l’archéozoologie. Pp. 38–48 in E. de Sena and H. Dessales (eds.) Archaeological Methods and Approaches: Industry and Commerce in Ancient Italy. BAR Int. S. 1262. Oxford. Lehmann, M. (1961) Ben Sira and the Qumran Literature. RevQ 3: 103-16. (1964) Identification of the Copper Scroll based on its technical terms. RevQ 5: 97-105. (1978) The Temple Scroll as a Source of Sectarian Halakhah. RevQ 9: 579-87. 82 BIBLIOGRAPHY Michniewicz, J. (2009) Qumran and Jericho Pottery: A Petrographic and Chemical Provenance Study. Poznan, Poland. Milik, J.T. (1957) Deux documents inédits de Désert de Juda. Biblica 38: 264-68. (1959) Ten Years of Discovery in the Wilderness of Judea. (trans. from the French (1957) by J. Strugnell). Studies in Biblical Theology 26. London. Mizzi, D. (2009) The Archaeology of Khirbet Qumran: Studies in a Contextual Approach. Unpublished PhD thesis, Oxford U. (2010) The glass from Khirbet Qumran: What does it tell us about the Qumran community? Pp. 99-198 in C. Hempel (ed.) The Dead Sea Scrolls: Texts and Contexts. STDJ 90. Brill, Leiden. Murphy, C. (2002) Wealth in the Dead Sea Scrolls and the Qumran Community. STDJ 40. Brill, Leiden. Nagar, Y. and Torgee, H. (2003) Biological Characteristics of Jewish Burial in the Hellenistic and Early Roman Periods. IEJ 53: 164-71. Naveh, J. (1968) Dated Coins of Alexander Jannaeus. IEJ 81: 20-5. Negev, A. and Gibson, S. (eds.) (2001) Archaeological Encyclopedia of the Holy Land. Hendrikson, NY. Netzer, E. (1981) Greater Herodium. Qedem 13. Hebrew University, Jerusalem. (1991) The Buildings, Stratigraphy and Architecture in J. Aviram, G. Foerster, and E. Netzer (eds.) Masada III: The Yigael Yadin Excavations 1963-1965. Final Report., IES, Jerusalem. (2001) Hasmonean and Herodian Palaces at Jericho., Vol. I. IES, Jerusalem. (2004) Hasmonean and Herodian Palaces at Jericho. Vol. II. IES, Jerusalem. (2005) Did any Perfume Industry exist at Ein Fashkhah? IEJ 55: 97-100. (2006) The Architecture of Herod, the Great Builder. Texts and Studies in Ancient Judaism 117. Tübingen. Netzer, E. and Garbrecht, G. (2002) Water Channels and a Royal estate of the late Hellenistic period in Jericho’s western plains. Pp. 367-79 in D. Amit, J. Patrich and Y. Hirschfeld (eds.). Newsom, C. (1990) ‘Sectually Explicit’ Literature from Qumran. Pp. 167-87 in W.H. Propp et al. (eds.) The Hebrew Bible and its Interpreters. Eisenbrauns, Winona Lake, IN. (1996) 4Q378 and 4Q379: An Apocryphon of Joshua. Pp. 35-85 in H.-J. Fabry et al. (eds.) Qumranstudien. Göttingen. Nissenbaum, A. (1993) The Dead Sea – an economic resource for 10,000 years. Hydrobiologia 267: 127-41. North, R. (1954) Qumrȃn and its Archaeology. Catholic Biblical Quarterly 16: 426-37. (1955) The Qumran ‘Sadducees’. Catholic Biblical Quarterly 17: 64-88. (1956) Qumran, Serek-a, and Related Fragments. Orientalia 25: 90-9. Early Christian Art and Archaeology in Honor of Gideon Foerster. Leuven. (2003) Review of R. Bar-Nathan, Hasmonean and Herodian Palaces at Jericho Vol. III: The Pottery. DSD 10/3: 420-28. (2004) Debating Qumran: Collected Essays on its Archaeology. Peeters, Leuven, Paris, Dudley, MA. (2007) A Response to D. Stacey, ‘Some Archaeological Observations on the Aqueducts of Qumran’. DSD 14/2: 223-44. Main, E. (1998) For King Jonathan or Against? The Use of Bible in 4Q448. Pp. 113-35 in M. Stone and E. Chazon (eds.) Biblical Perspectives: Early Use and Interpretation of the Bible in Light of the Dead Sea Scrolls. Brill, Leiden. Malkin, I. (1998) The Returns of Odysseus: Myth, Colonization and Ethnicity. U. of California Press, Berkeley and London. Mason, S. (2009) Josephus, Judea and Christian Origins: Methods and Categories. Peabody, MA. (2011) The Historical Problem of the Essenes. Pp. 201-54 in P. Flint et al. (eds.) Celebrating the Dead Sea Scrolls. A Canadian Collection. SBL, Atlanta, GA. Masterman, E.W.G. (1902) ‘Ain El-Feshkah, El-Hajar, El-Asbah and Khurbet Kumran. PEFQS 160-67. (1904a) Observations on the Dead Sea Level, Second Report; 1902-3. PEFQS 83-95. (1904b) Dead Sea Observations. PEFQS 280-81. (1905) Dead Sea Observations. PEFQS 158-59. Mayhew, H. (1851) London Labour and the London Poor. Vol. II. (repr. 1967) London. Mazar, A. (1992) Tombs from the Roman Period at Caesarea. Atiqot 21: 105-08. (2002) A Survey of the Aqueducts of Jerusalem. Pp. 211-44 in D. Amit, J. Patrich and. Y. Hirschfeld (eds.). Mazar, B. (1973) Beth She’arim I – Report on the Excavations during 1936-1940, Catacombs 1-4. Jerusalem. McCarter, K. (1994) The Copper Scroll Treasure as an Accumulation of Religious Offerings. Pp. 133-48 in M. Wise et al. (eds.). McGrail, S., Blue, L., and Palmer, C. (2003) Hide Boats of the River Kaveri, Tamil Nadu. Pp. 241-57 in S. McGrail, L. Blue, E. Kentley, and C. Palmer, Boats of South Asia. London. McKenzie, J. (1990) The Architecture of Petra. OUP, Oxford. McNamara, M. (2011) Targum and the New Testament: Collected Essays. WUNT 279. Mohr Siebeck, Tübingen. Meskell, L. (ed.) (1998) Archaeology under Fire: Nationalism, Politics and Heritage in the Eastern Mediterranean and the Middle East. Routledge, London and New York. Metzger, B.M. (1958-59) The Furniture in the Scriptorium at Qumran. RevQ 1: 509-15. Meyers, E.M. (2010) Khirbet Qumran and its Environs. Pp. 21-45 in T. Lim and J. Collins (eds.). 83 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Relative to Qumran. Pp. 213-19 in K. Galor, J.B. Humbert and J. Gunneweg (eds.). Poole, J.B. and Reed, R. (1961) The ‘Tannery’ at ‘Ain Feshka. PEQ 93: 114-23. (1966) A Study of Some Dead Sea Scrolls and Leather Fragments from Cave 4 at Qumran, Part II – Chemical Examination. Proceedings of the Leeds Philosophical and Literary Society. Scientific Section 9: 171-82. (1972) The Preparation of Leather and Parchment by the Dead Sea Scrolls Community. Pp. 143-68 in M. Kranzberg and W. Davenport (eds.) Technology and Culture: An Anthology. Doubleday, New York. Popović, M. (2009) Bones, Bodies, and Resurrection in the Dead Sea Scrolls. Pp. 221-64 in T. Niclas et al. (eds.) The Human Body in Death and Resurrection. (Deuterocanonical and Cognate Literature: Year book 2009) Walter de Gruyer, Berlin. (2012) Qumran as Scroll Storehouse in Times of Crisis? A Comparative Perspective on Judaean Desert Manuscript Collections. Journal for the Study of Judaism 43: 551-94. Porat, R. (2006) The Road along the Dead Sea Shore between Qumran and Ein-Gedi in the Second Temple Period. Cathedra 121: 5-22 (Hebrew). Porath, Y. (2005) Survey of the Agricultural Systems at the ‘Ein Gedi Oasis. Atiqot 50: 1-20 (Hebrew), 237-39 (English summary). Pritchard, J.B. (1958) The Excavations at Herodian Jericho. AASOR 32-33. New Haven, CT. Puech, E. (1992) Une apocalypse messianique (4Q521). RevQ 15: 475-519. (1996) Jonathan le prêtre impie et les débuts de la communauté de Qumrân. 4QJonathan (4Q523) et 4QPsAp (4Q448). RevQ 17: 241-70. (1998) The Necropolises of Khirbet Qumran and Ain el-Ghuweir and the Essene Belief in the Afterlife. BASOR 312: 21-36. Rabin, I., Hahn, O., Wolff, T., Masic, A. and Weiberg, G. (2009) On the Origin of the Ink of the Thanksgiving Scroll (1QHodayot). DSD 16/1: 197-206. Rasmussen, K., van der Plicht, J., Cryer, F.H., Doudna, G., Cross F.M. and Strugnell, J. (2001) The Effects of Possible Contamination on the Radiocarbon Dating of the Dead Sea Scrolls I: Castor Oil. Radiocarbon 43: 127-32. Rasmussen, K., van der Plicht, J., Doudna, G., Cross, F.M. and Strugnell, J. (2003a) Reply to Israel Carmi: ‘Are the 14C dates of the Dead Sea Scrolls Affected by Castor Oil Contamination?’ Radiocarbon 45: 497-9. Rasmussen, K., Gwozdz, R., Taylor, J., and Doudna, G. (2003b) Preliminary Data of Trace Element Concentrations in Human Bone Samples. In J.-B. Humbert and J. Gunneweg (eds.). Rasmussen, K.L., Gunneweg, J., Doudna, G., Taylor, J., Bélis, M., van der Plicht, J., Humbert, J.-B. and Egsgaard, H. (2006) Cleaning and Radiocarbon Norton, J. (2003) Reassessment of Controversial Studies on the Cemetery. Pp. 107-27 in J.-B. Humbert and J. Gunneweg (eds.). Oleson, J.P. (2005) Constructing the Harbour of Caesarea on the Sea: New Evidence from the ROMACONS Field Campaign of 2005. http://web.uvic.ca/~jpoleson/ROMACONS/Caes area2005.htm (2007) Nabataean Water Supply, Irrigation and Agriculture. Pp. 217-50 in K. Politis (ed.) The World of the Nabataeans. Volume 2 of the International Conference – The World of the Herods and the Nabataeans held at the British Museum 17-19 April 2001. Stuttgart. Oren, A. (2007/8) Marine Trade on the Dead Sea in the 12th Century – (accessed 18/2/2009) at http://maritime.co.il/db/ uploads/12th%20CenturyBoats%20on%20the%2 0Dead%20Sea.pdf Overman, J.A. (2009) Between Rome and Parthia: Galilee and the Implications of Empire. Pp. 279-99 in Z. Rodgers et al. (eds.) A Wandering Galilean: Essays in Honour of Seán Freyne. Brill, Leiden. Parker, F. (2003) The Terms ‘Angel’ and ‘Spirit’ in Acts 23, 8. Biblica 84: 344-65. Pasto, J. (2002) The Origin, Expansion and Impact of the Hasmoneans in Light of Comparative Ethnographic Studies. Pp. 166-201 in P.R. Davies and J.M. Halligan (eds.) Second Temple Studies. Vol. 3: Literature and Society in the Persian, Hellenistic and Roman Periods. Sheffield Academic Press, Sheffield. Patrich, J. (2000) Did Extra-Mural Dwelling Quarters Exist at Qumran? Pp. 722-27 in L. Schiffman, E. Tov and J. VanderKam (eds.) The Dead Sea Scrolls: Fifty Years After their Discovery; Proceedings of the Jerusalem Congress, July 2025, 1997. IES, Jerusalem. (2002) The Water Supply to Caesarea: a Reassessment. Pp. 105-29 in D. Amit, J. Patrich and Y. Hirschfeld (eds.). (2006) Agricultural Development in Antiquity: Improvements in the Cultivation and Production of Balsam. Pp. 241-47 in K. Galor, J.-B. Humbert and J. Gunneweg (eds.). Patrich, J. and Arubas, B. (1989) A Juglet Containing Balsam Oil(?) from a Cave near Qumran. IEJ 39: 43-59. Patridge, W. (1823) A Practical Treatise on Dying of Woolen, Cotton and Skein Silk. New York, reprinted Pasold Research Fund, Wiltshire, 1973. Peleg, Y. (2002) The dams of Caesarea’s Low-Level Aqueduct. Pp. 141-47 in D. Amit, J. Patrich and Y. Hirschfeld (eds.). Politis, K.D. (1998) Rescue Excavations in the Nabatean Cemetery at Khirbet Qazone 1996-1997. Annual of the Department of Antiquities of Jordan 62: 611-14. (2006) The Discovery and Excavations of the Khirbet Qazone Cemetery and its Significance 84 BIBLIOGRAPHY Dating of Material from Khirbet Qumran. Pp. 139-63 in Humbert and Gunneweg (eds.) Bioand Material Cultures at Qumran. Papers from a COST Action G8 working group meeting held in Jerusalem, Israel on 22-23 May 2005. Fraunhofer IRB Verlag, Stuttgart. Rasmussen, K., van der Plicht, J., Doudna, G., Nielsen, F., Hoejrup, P., Stenby, E.H., Pedersen, C.T. (2009) The Effects of Possible Contamination on the Radiocarbon Dating of the Dead Sea Scrolls II: Empirical Methods to Remove Castor Oil and Suggestions for Redating. Radiocarbon 51: 1005-22. Reed R. (1972) Ancient Skins, Parchments and Leathers. Seminar Press, New York. (1975) The Nature and Making of Parchment. Elmete Press, Leeds. Reeves, J.C. (2005) Complicating the Notion of an ‘Enochian Judaism’. Pp. 373-83 in G. Boccaccini (ed.) Enoch and Qumran Origins. Eerdmans, Grand Rapids, MI. Rengstorf, K.H. (1960) Hirbet Qumran and die Bibliothek vom Toten Meer. Studia Delitzschiana 5. Stuttgart. Reynolds, B. (2012) Between Symbolism and Realism. The Use of Symbolic and Non-Symbolic Language in Ancient Jewish Apocalypses. 33363 B.C.E. Voudenhoeck & Ruprecht, Göttingen. Richardson, P. (1996) Herod: King of the Jews and Friend of the Romans. Continuum, Edinburgh. Röhrer-Ertl, O. (2006) Facts and Results Based on Skeletal Remains from Qumran found in the Collectio Kurth – A Study and Methodology. Pp. 181-94 in K. Galor, J.-B. Humbert and J. Gunneweg (eds.). Roitman, A. (1997) A Day at Qumran. Israel Museum. Jerusalem. Roller, D. (1998) The Building Program of Herod the Great. U. of California Press, Berkeley, CA. Rosen, S. Avni, G. and Amitai-Press, N. (1997) The ‘Oded Sites: Investigation of Two Early Islamic Pastoral Camps South of the Ramon Crater. Ben-Gurion U. of the Negev Press. Beer-sheva. Rubin, R. (1988) Water Conservation Methods in Israel’s Negev Desert in Late Antiquity. Journal of Historical Geography 14/3: 229-44. Safrai, Z. (2003) The Agricultural Structure in Palestine in the Times of the Second Temple, Mishnah and Talmud. Pp. 105-126 in A. Maeir, S. Dor and Z. Safrai (eds.) The Rural Landscape of Ancient Israel. BAR Int. S. 21. Archaeopress, Oxford. Safrai, Z. and Eshel, H. (2000) Economic Life. Pp. 22833 in L. Schiffman and J. VanderKam (eds.) Encyclopedia of the Dead Sea Scrolls. OUP, Oxford. Saldarini, A. (2001) Pharisees, Scribes, and Sadducees in Palestinian Society. Eerdmans, Grand Rapids, MI. Sanders, E.P. (2000) The Dead Sea Sect and other Jews: Commonalities, Overlaps and Differences. Pp. 7- 43 in T. Lim (ed.) The Dead Sea Scrolls in their Historical Context. T&T Clark, New York. Säve-Söderbergh, T., Englund, G. and Nordström, A.H. (eds.) (1981) Late Nubian Cemeteries. The Scandinavian Joint Expedition to Sudanese Nubia, 6, Solna, Sweden. Schiffman, L. (1989) The Eschatological Community of the Dead Sea Scrolls: A Study of the Rule of the Congregation. Scholars Press, Atlanta, GA. (1992) The Sadducean Origins of the Dead Sea Scroll Sect. Pp. 35-49 in H. Shanks (ed.) Understanding the Dead Sea Scrolls. A Reader from the Biblical Archaeology Review. Random House, New York. (1994) Reclaiming the Dead Sea Scrolls. Jewish Publication Society, Philadelphia and Jerusalem. Schofield, A. and VanderKam, J. (2005) Were the Hasmoneans Zadokites? JBL 124: 73-87. Schülke, A. (1999) On Christianization and Grave Finds. European Journal of Archaeology 2/1: 79-106. Schuller, E. (1990) 4Q372 1: A Text About Joseph. Pp. 349-76 in F. García Martínez (ed.) The Texts of Qumran and the History of the Community. Vol. II. RevQ 1454. Gabalda, Paris. Schultz, B. (2006) The Cemetery of Qumran: 150 Years of Research. DSD 13: 194-228. Schürer, E. (1973) The History of the Jewish People in the Age of Jesus Christ. Vermes, G., Millar, F., and Black, M. (eds.). Edinburgh. Schwartz, J. (1988) On Priests and Jericho in the Second Temple Period. Jewish Quarterly Review 79: 2348. Schwartz, S. (1991) Israel and the Nations Roundabout: 1 Maccabees and the Hasmonean Expansion. JJS 42: 16-38. Shanks, H. (1999) The Mystery and Meaning of the Dead Sea Scrolls. Vintage Books, New York. Sheffer, A., and Granger-Taylor, H. (1995) Textiles from Masada. Pp. 153-250 in D. Barag et al. (eds.) Masada IV, The Yigael Yadin Excavations 19631965: The Final Reports. IES, Jerusalem. Shemesh, A. (2012) Trends and Themes in Israeli Research of the Halakhah in the Dead Sea Scrolls. Pp. 345-62 in D. Dimant and I. Kottsieper (eds.) The Dead Sea Scrolls in Scholarly Perspective: A History of Research. Brill, Leiden. Sheridan, S.G. (2002) Scholars, Soldiers, Craftsmen, Elites: Analysis of French Collection of Human Remains from Qumran. DSD 9/2: 199-248. Sheridan, S.G. and Ullinger, J. (2006) A Reconsideration of the Human Remains in the French Collection from Qumran. Pp. 195-212 in K. Galor, J.-B. Humbert and J. Zangenberg (eds.). Sklar-Parnes, D. (2005) Jerusalem, Shu’afat. In Hadashot Arkheologiyot 117 online at http://www.hadashot-esi.org.il/report_detail _eng.asp?id=179&mag_id=110 Silberman, N.A. (1989) Between Past and Present: Archaeology, Ideology and Nationalism in the Modern Middle East. New York. 85 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS qumraniens en hommage à Émile Puech. STDJ 61. Brill, Leiden. (2012) The Damascus Document (D) as a Rewriting of the Community Rule (S). RevQ 25/4: 605-20. Stone, M. (2006) Apocrypha, Pseudepigrapha, and Armenian Studies: Collected Papers, Volume 1. Leuven. Storrier, S. (ed.) (2006) Scottish Life and Society. A Compendium of Scottish Ethnography. Edinburgh. Sukenik, E. (1948, 1950) Megillot Genuzot. Vols. I and II (Hebrew). Jerusalem. Sussmann, Y. (1989-1990) The History of Halakha and the Dead Sea Scrolls: Preliminary Observations on Miksat maasei ha-torah (4QMMT). Tarbiz 59: 11-76 (Hebrew). (Quoted in trans. in E. Cook, Solving the Mysteries of the Dead Sea Scrolls 1994.) Talmon, S. (1951) Yom Hakkippurim in the Habakkuk Scroll. Biblica 32: 549-63. (1999) Hebrew Fragments from Masada. Pp. 1149 in S. Talmon et al. (eds.) Masada VI: Yigael Yadin Excavations 1963-1965. Final Reports: Hebrew Fragments from Masada. IES, Jerusalem. Taxel, I. (2011) Between Judea and Nabatea: A Historical, Archaeological and Socio-Economic Interpretation of Hellenistic and Early Roman ‘Aroer. Pp. 399-412 in Y. Thareani. Taylor, J. (1999) The Cemetery of Khirbet Qumran and Woman’s Presence at the Site. DSD 6: 285-323. (2006) Khirbet Qumran in Period III. Pp. 133-46 in K. Galor, J.-B. Humbert and J. Gunneweg (eds.) (2007) Qumran in Context: Reassessing the Archaeological Evidence. BAIAS 25: 171-83. (2007b) Philo of Alexandria on the Essenes: a case study on the use of Classical Sources in Discussions of the Qumran-Essene Hypothesis. The Studia Philonia Annual 19: 1-28. (2009) ‘Roots, Remedies and Properties of Stones’: The Essenes, Qumran and Dead Sea Pharmacology. JJS Vol. 60/2: 226-44. (2010) The Classical Sources on the Essenes and the Scrolls Communities. Pp. 173-99 in T. Lim and J. Collins (eds.). (2011a) Women, Children, and Celibate Men in the Serekh Texts. Harvard Theological Review 104/2: 171-90. (2011b) The Nazoraeans as a ‘Sect’ in ‘Sectarian’ Judaism? A Reconsideration of the Current View via the Narrative of Acts and the Meaning of Hairesis. Pp. 87-117 in S. Stern (ed.) Sects and Sectarianism in Jewish History. Brill, Leiden. (2012a) Buried Manuscripts and Empty Tombs: The Qumran Genizah Theory Revisited. Pp. 269315 in A. Maeir et al. (eds). ‘Go Out and Study the Land’ (Judges 18:2): Archaeological, (1997) Nationalism and Archaeology. Pp. 103-12 in E. Meyers (ed.) The Oxford Encyclopedia of Archaeology in the Near East. New York. Simpson, S.J. (1995) Death and Burial in the Late Islamic Near East: Some Insights from Archaeology and Ethnography. Pp. 240-51 in S. Campbell and A. Green (eds.) The Archaeology of Death in the Ancient Near East. Oxbow, Oxford. Smith, A.D. (1991) National Identity. Penguin, London. (2000) The Nation in History: historiographical debates about ethnicity and nationalism. Brandeis U. Press/Historical Society of Israel, Hanover, NH. Smith, N. (1971) A History of Dams. Peter Davies, London. Stacey, D. (2004a) Excavations at Tiberias, 1973-1974: The Early Islamic Periods. IAA Reports 21. Jerusalem. (2004b) Tombs in the Vicinity of the Hippodrome at Jericho. Pp. 226-8 in Netzer Hasmonean and Herodian Palaces at Jericho, Vol II. IES, Jerusalem. (2004c) Some Notes on the Archaeological Context of Qumran in the light of Recent Publications. http://www.bibleinterp.com/articles/Stacey_Qum ran_Light_of_Recent_Publications.shtm (2006) Hedonists or Pragmatic Agriculturalists? Reassessing Hasmonean Jericho. Levant 38: 191202. (2007) Some Archaeological Observations on the Aqueducts of Qumran. DSD 14/2: 222-43. (2008) Seasonal Industries at Qumran. BAIAS 26: 7-29. (2009) Three Notes on Qumran. http://www.bibleinterp.com/articles/stacey.shtml Stager, L. (1976) Farming in the Judean Desert during the Iron Age. BASOR 221: 145-58. Stauber, C.M. (2011) Determinism in the Rule of the Community (1QS): A New Perspective. Pp. 34558 in P. Flint et al. (eds.), Celebrating the Dead Sea Scrolls. A Canadian Collection. Atlanta, GA. Stead, J. (1981) Uses of Urine. Old West Riding 1/2: 1218. Steckoll, S.H. (1968) Preliminary Excavation Report in the Qumran Cemetery. RevQ 6: 323-36. (1969) Marginal Notes on the Qumran Excavations. RevQ 25: 33-44. Stemberger, G. (1995) Jewish Contemporaries of Jesus: Pharisees, Sadducees, Essenes. Minneapolis. Stern, S. (2000) Qumran Calendars: Theory and Practice. Pp. 179-86 in T. Lim et al. (eds.) The Dead Sea Scrolls in their Historical Context. T&T Clark, New York. (2010) Qumran Calendars and Sectarianism. Pp. 232-53 in T. Lim and J. Collins (eds.). Steudel, A. (1993) ‫ חרית י י‬in the Texts from Qumran. RevQ 16: 225-45. (2006) 4Q448 -- The Lost Beginning of MMT? Pp. 247-63 in F. García Martínez et al. (eds.) From MMT to Resurrection: Mélanges 86 BIBLIOGRAPHY Monarchy. Pp. 255-64 in T. Rajak et al. (eds.) Hellenistic Culture and Society 50. University of California Press, Berkeley, CA. de Vaux, R. (1949) Post scriptum. La Cachette Des Manuscripts Hébreux. RB 56: 234-38. (1953) Fouilles au Khirbet Qumran: Rapport préliminaire. RB 60: 83-106. (1954) Fouilles au Khirbet Qumrȃn: Rapport préliminaire sur la deuxième campagne. RB 61: 206-36. (1956) Fouillles au Khirbet Qumran: Rapport préliminaire sur les 3e, 4e, et 5e campagnes. RB 63: 533-77. (1959) Fouilles de Feshka. RB 66: 225-55. (1973) Archaeology and the Dead Sea Scrolls. OUP, Oxford. (1993) Qumran, Khirbet and ‘Ein Feshka. Pp. 1235-41 in Stern, E. (ed.) The New Encyclopaedia of Archaeological Excavations in the Holy Land. IES, Jerusalem. Vermes, G. (1956) Discovery in the Judean Desert. Desclee, New York. Vielhauer, R. (2001) Materielle Rekonstruktion und historische Einordnung der beiden Pescharim zum Hoseabuch (4QpHosa und 4QpHosb). RevQ 77: 39-91. Vila, F. (1978) La prospection archéologique de la valée du Nil, au sud de la cataracte de Dal (Nubie Sudanaise). Fascicule 9. Paris. Viviano, B. and Taylor, J. (1992) Sadducees, Angels, and Resurrection (Acts 23: 8-9). JBL 111: 496-98. Vogler, G. (2002) A Harris Way of Life. Isle of Harris. Wacholder, B.Z., and Abegg, M.G. (1991) A Preliminary Edition of the Unpublished Dead Sea Scrolls. Fascicule 1. Washington, D.C. Wagemakers, B., and Taylor, J. (2011) New Photographs of the Qumran Excavations from 1954 and Interpretations of L.77 and L.86. PEQ 143/2: 134-56. Ward, G. (1996) The Psalms of Solomon. A Philological Analysis of the Greek and the Syriac Texts. PhD Dissertation, Temple University, New York. Watson, F. (2004) Paul and the Hermeneutics of Faith. T & T Clark, London. Weis, Z. (1996) Foreign Influences on the Jewish Burial in the Galilee in the Mishnah and Talmud Period. Eretz Israel 25: 364-56. (Hebrew) Weinfeld, M. (1986) The Organizational Pattern and the Penal Code of the Qumran Sect: A Comparison with Guilds and Religious Associations of the Hellenistic-Roman Period. Novum Testamentum et Orbis Antiquus 2. Göttingen. Weissenberg, H. von (2009) 4QMMT: Reevaluating the Text, the Function, and the Meaning of the Epilogue. Brill, Leiden. Werman, C. (2006) Epochs and End-Time: the 490-Year Scheme in Second Temple Literature. DSD 13/2: 229-55. Wilson, E. (1986) The Fifties: from Notebooks and Diaries of the Period. Collins Publishing, Toronto. Historical and Textual Studies in Honor of Hanan Eshel. Brill, Leiden. (2012b) The Essenes, the Scrolls, and the Dead Sea. OUP, Oxford. Taylor, J., Rasmussen, K., Doudna, G., van der Plicht, J., and Egsgaard, H. (2005) Qumran Textiles in the Palestine Exploration Fund, London: Radiocarbon Dating Results. PEQ 137/2: 15967. Thareani, Y. (2011) Tel ‘Aroer. The Iron Age II Caravan Town and the Hellenistic-Early Roman Settlement. Hebrew Union College, Jerusalem. Thiering, B. (1989) The Date of the Composition of the Temple Scroll. Pp. 99-120 in G.J. Brooke (ed.) Temple Scroll Studies. JSOT Press, Sheffield. Tigchelaar, E. (2003) In Search of the Scribe of 1QS. Pp. 439-52 in S.M. Paul et al. (eds.) Emanuel: Studies in the Hebrew Bible, Septuagint, and Dead Sea Scrolls in Honor of Emanuel Tov. Vetus Testament Supplements 94. Brill, Leiden. Toombs, L. E. (1985) The Muslim Cemetery. Pp. 16-156 in K.G. O’Connell (ed.) Tell el-Hesi: Modern Military Trenching and Muslim Cemetery Field I, Strata I-II. The Joint Archaeological Expedition to Tel el-Hesi, Vol.II. Wilfrid Laurier University Press, Waterloo, Ontario. Tov, E. (1997) L’importance des textes du desert de Juda pour l’histoire du texte de la Bible hébraïque. Une nouvelle synthèse. Pp. 215-52 in E.-M. Laperrousaz (ed.) Qoumrân et les manuscrits de la Mer Morte. Les Éditions du Cerf, Paris. (2004) Scribal Practices and Approaches Reflected in the Texts Found in the Judean Desert. Brill, Leiden. (2011a) Some Thoughts at the Close of the Discoveries in the Judaean Desert Publication Project. Pp. 3-13 in A. Roitman et al. (eds.) The Dead Sea Scrolls and Contemporary Culture. Proceedings of the International Conference held at the Israel Museum, Jerusalem (July 6-8, 2008). Brill, Leiden. (2011b) The Sciences and the Reconstruction of the Ancient Scrolls: Possibilities and Impossibilities. Pp. 3-25 in A. Lange et al. (eds.) The Dead Sea Scrolls in Context (2 vols), I. Brill, Leiden. (2012) Textual Criticism of the Hebrew Bible. 3rd ed. Fortress Press, Minneapolis, MN. Tsafrir, Y. (ed.) (1993) Ancient Churches Revealed. IES, Jerusalem. VanderKam, J. (2004) From Joshua to Caiaphas. High Priests after the Exile. Augsburg Fortress, Minneapolis, MN. (2005) Response: Jubilees and Enoch. Pp. 16270 in G. Boccaccini (ed.), Enoch and Christian Origins. New Light on a Forgotten Connection. Eerdmans, Grand Rapids, MI. (2010) The Dead Sea Scrolls Today. 2nd ed. Eerdmans, Grand Rapids, MI. Van der Kooij, A. (2007) The Greek Bible and Jewish Concepts of Royal Priesthood and Priestly 87 QUMRAN REVISITED: A REASSESSMENT OF THE ARCHAEOLOGY OF THE SITE AND ITS TEXTS Wise, M.O. (1999) The First Messiah. Investigating the Savior Before Christ. HarperSanFrancisco, San Francisco, CA. (2000) “ ‫י‬ ‫ני‬ ‫ י‬. A study of 4Q491c, 4Q471b, 4Q427 7 and 1QHA 25:35-26:10*. DSD 7/2: 173-219. (2002) David J. Wilmot and the Copper Scroll. Pp. 291-310 in G. Brooke and P. Davies (eds.) Copper Scroll Studies. Sheffield Academic Press, Sheffield. (2003) Dating the Teacher of Righteousness and the Floruit of his Movement. JBL 122: 53-87. (2010) The Origins and History of the Teacher’s Movement. Pp. 92-122 in T. Lim and J. Collins (eds.). Wise, M.O., Pardee, D.G., Collins, J.J., Golb, N. (eds.) (1994) Methods of Investigation of the Dead Sea Scrolls and the Khirbet Qumran Site. New York Academy of Sciences, New York. Wise, M.O., Abegg, M., and Cook E. (1996) The Dead Sea Scrolls. A New Translation. HarperCollins, London. Wolters, A. (1998) The Copper Scroll. Pp. 302-23 in P. Flint and J. VanderKam (eds.) The Dead Sea Scrolls after Fifty Years, Vol. I. Brill, Leiden. Yadin, Y. (1966) Masada: Herod’s Fortress and the Zealots’ Last Stand. Weidenfeld and Nicolson, London. Yardeni, A. (2007) A Note on a Qumran Scribe. Pp. 28798 in M. Lubetski (ed.) New Seals and Inscriptions, Hebrew, Idumean and Cuneiform. Sheffield Phoenix Press, Sheffield. (2010) The Book of Hebrew Script. History, Palaeography, Script Styles, Calligraphy & Design. 3rd Eng. ed. Carta, Jerusalem. Yellin, J., Broshi, M. and Eshel, H. (2001) Pottery of Qumran and Ein Ghuweir: The First Chemical Exploration of Provenience. BASOR 321: 65-78. Young, I. (2002) The Stabilization of the Biblical Text in the Light of Qumran and Masada: A Challenge for Conventional Qumran Chronology? DSD 9: 364-90. (2005) The Biblical Scrolls from Qumran and the Masoretic Text: A Statistical Approach. Pp. 81139 in M. Dacy, J. Dowling, and S. Faigan (eds.) Feasts and Fasts. A Festschrift in Honour of Alan David Crown. Mandelbaum Publishing, University of Sydney, Sydney. (2013) The Contrast Between the Qumran and Masada Biblical Scrolls in the Light of New Data. A Note in Light of the Alan Crown Festschrift, forthcoming in S. Tzoref and I. Young (eds.) Keter Shem Tov: Collected Essays on the Dead Sea Scrolls in Memory of Alan Crown. Gorgias Press, Piscataway, NJ. Zangenburg, J. (2004) Opening up our View. Pp. 170-87 in Edwards, D. (ed.) Religion and Society in Roman Palestine: Old Questions, New Approaches. Routledge, London. Zeuner, F.E. (1960) Notes on Qumran. PEQ 92: 27-36. Zias, J.E. (2000) The Cemeteries of Qumran and Celibacy: Confusion Laid to Rest? DSD 7: 22053. Zissu, B. (1998) ‘Qumran Type’ Graves in Jerusalem: Archaeological Evidence for an Essene Community? DSD 5: 158-71. 88