' Andrews University Seminary Studies, Season 1988, Vol. 26, No. 3, 217-252 Copyright © 1988 by Andrews University Press. i i THE JOINT MADABA PLAINS PROJECT A PRELIMINARY REPORT ON THE SECOND SEASON j AT TELL EL-`UMEIRI AND VICINITY (JUNE 18 TO AUGUST 6, 1987) i LAWRENCE T. GERATY LARRY G. HERR Atlantic Union College Canadian Union College South Lancaster, Massachusetts 01561 College Heights, Alberta TOC OZO OYSTEIN S. LABIANCA Andrews University l Berrien Springs, Michigan 49104-0030 During the summer of 1987 Andrews University sponsored a second archaeological expedition to Tell el-'Umeiri and its vicinity { in Jordan (see Plate 1). Continuing as the Madaba Plains Project, this second season of excavation (see Plates 2, 3), soundings, and survey was jointly sponsored by Atlantic Union College (South Lancaster, Massachusetts), Canadian Union College (Lacombe, Al- berta), and Southwestern Adventist College (Keene, Texas).' What 1 'The authors of this report are indebted to each member of the staff who helped r to make possible these results. In addition to the financial support of the consortium ? institutions, other funds were raised from private donations and volunteer participa- tion fees. Individuals who contributed generously to the general dig fund include d William Berecz, Jr., James M. Byers III, Ronald and Sheila Geraty, Thomas and Hazel Geraty, Robert Ibach, Salim Japas, Enid and John Leung, Gloria G. and John T. Martin, Charles A. Platt, Elizabeth Platt, C. Murray Robinson, Barbara Russell, Zorka Sandic, Stanley Squier, C. Erwin Syphers, Gary A. and Carolyn Waldron, and Ernest S. and Dorothy L. Zaire. We gratefully acknowledge the support offered by each of these friends of the project. Special thanks are due to Director-General of Antiquities Dr. Adrian Hadich, who eased several major problems during the course of the season; Amman Antiqui- ties Inspector Hefzi Haddad, who went out of his way numerous times to make our project a success; Department of Antiquities representative Nazmieh Rida, who served as a Square Supervisor and helped solve problems with workers: and business- man/scholar Raouf Abujaber, landowner of Tell el-`Umeiri, who was again gener- ous in allowing our research to proceed unhindered. The officers and staff of the American Schools of Oriental Research and its local affiliate, the American Center 217 t S4 ' G-' 218 GERATY, HERR, AND LABIANCA r /y0 _ ! I 1 f I I 1 I t 180- Wadi Zarga 170- 160- . Amman 150- /yo- • Tell el `Umeiri • Hesban 130 • Jerusalem ~~0 • Madaba • Bethlehem la0- , a //O- Q, W /00- Q i W Wadi Mujib 90 - q i 80 - 7p- • Kerak 60, , , , , I I I /1.o 170 180 190 boo Vo 2moo '3o ago 050 ZGo o 1o ~0 30 40 tiles SCALE I , t I ' I 0 10 .10 30 yo 50 6o kris r Plate 1. Map of Palestine with the location of Tell el-`Umeiri. f i } 1 SECOND SEASON AT TELL EL-`UMEIRI 219 p , Plate 2. Aerial view of Tell el-`Umeiri. I f DRAWN BY GLENN JOHNSON E MERLING ALOMIA GILLIAN GERATY RASCHEL BARTON \\ ROBERT L005 ~\ \ LARRY HERR 1 s \ \ I ~ \ \ C 1 j nB Av\\ \ ' ~ ~I I I, ,1A \ _~ \ I a ' \\\~ \\ D TELL EL-`UMEIRI Plate 3. Topographic map of Tell el-`Umeiri with Fields of excavation. 1 I r 220 GERATY, HERR, AND LABIANCA follows is a general preliminary report of the results achieved by the 100-member project team (see Plate 4). It should be read in the context of the first season's report, which also appeared in AUSS.2 There the overall goals of the project, a description of the site of Tell el- Umeiri, and the evidence for 'possible identifications are all documented. Once again the objectives of the project focused on cycles of intensification and abatement in settlement and land use in this frontier region: the Ammonite foothills on the northern edge of the Madaba Plains. Central to this focus was the study of the food systems employed by the inhabitants t1hrough time.' The implementation of these objectives was refined during the 1987 season by enlarging the regional survey to three teams, each with its own primary objectives; by conducting excavations at three hinterland sites, most notably Rujm Selim (a small agricultural settlement some 2 kin north of Tell el-`Umeiri); and by expanding excavation areas on the central tell itself. The following report will first siummarize the findings of the regional survey, next highlight the results of the most extensive hinterland excavation, and then discuss the excavation results on the central tell, field by field. of Oriental Research in Amman, provided invaluable assistance; the latter's director David McC:reerv, administrative director Glenn Peterinan, and Ibtisam Dababneh, administrative assistant at ACOR, most be particularly mentioned. Others within Jordan without whom the excavation would not have been possible were Prince Raad ibn Zeid, who has been a constant supporter, and Richard T. Krajczar, Superintendent of the American Community School in Amman, who provided generous logistical support. The Baptist School near Shmeisani, Amman, through its principal, Wilson "Datum, gave virtually all its very ample facilities to the dig for headquarters. It offered adequate space for sleeping, eating, working, meeting, and recreation. a. Lawrence '1-. Geraty, "A Preliminary Report on the First Season at hell el- Umeiri (June 18 to August 8, 198,1)," AL'SS 23 1 (1980: 85-110_ ;Our working definition of "food system'' is that of LaBianca from his 1987 Brandeis Ph.D. (issertation, soon to be published by Andrews University Po,) as Hesban I-Sedeitarz:atiou and Nonzadi:atiout at Hesbau and Cicim.ty: A Stu,!,., of Food System Tran.sitioiis in Transjordan; n0tnely, "a food system is a complex unity consisting of all of the purposive, patterned (institutionalized) and intercon- nected activities carried out by a group of individuals in their quest for food.'' SECOND SEASON AT TELL EL-`UMEIRI 221 i ~ Y L f k. S Ok ¢ ~; 54 Y. i Y -fir ~ql ~ iw ,, a~ c jr A4 ~i 1 222 GERATY, HERR, AND LABIANCA 1. The Regional Survey' In the preliminary report for the 1984 season the appearance, within the survey region, of a number of round or rectangular structures was noted.' Often constructed with "megalithic" foun- dations, these structures were thought to be "farmsteads." Further work during the 1987 season enabled the survey team to classify these round and rectangular structures into five basic types: (1) large agricultural complexes or "estates"; (2) small farmsteads; and several types which now appear to be best classified as something other than farmsteads; (3) field shelters; (4) forts; and (5) kilns. Large Agricultural Complexes Additional examination during the 1987 season indicated that many of the larger "megalithic" structures recorded in 1984 did not occur as isolated phenomena throughout the countryside but rather were regularly associated with a number of the other features, including various wall lines (perimeter, field, and terrace walls), bedrock winepresses, millstones, caves (cellars), cisterns, and a wide variety of cupholes. Together, these features lend support to the suggestion that these "megalithic" sites should be classified, with- out becoming overly specific, as "agricultural complexes" (see Plate 5). Excellent parallels for these complexes have been reported in the hill country around Jerusalem at sites such as Hurvat Ein Tutt, Khirbet er-Ras, and Ein Yalu.6 The only real difference between the "farm-units" reported around Jerusalem and the agricultural complexes near Amman is how the dominant building on the site was constructed: Near Jerusalem the central building was con- structed structed according to the "four-room" house plan, while the'rooms of those structures near Amman are divided differently and are of 1 "megalithic" construction. Rather than reflecting the function of 'The regional survey was directed by oystein S. LaBianca. Field Supervisors for t the three major survey operations (as described in the text) were Gary Christopherson, Jon Cole, and Randall Younker. They were assisted by Dorothy Irvin, ethnographer; Howard Krug, tomb surveyor; John Lee, lithicist; Doug Schnurrenberger, geologist; and Judy Christiansen, Raymond Pelto, John Podgore, Rhonda Sandic, and Tony Squier, volunteers. Translator for the survey team was Naji Tannous. 5Geraty, pp. 106-108. 6G. Edelstein and S. Gibson, "Ancient Jerusalem's Rural Food Basket," BA Rev ' 8/4 (1982): 46-54. 1 i s SECOND SEASON AT TELL EL-`UMEIRI 223 i 4 11 ~`_'l ~ t+ d 4. '.y Plate 5. Central building at Rujm Selim, a large agricultural com- plex near `Umeiri. the building, this latter construction technique may simply reflect what was locally available in terms of raw materials as well as what was absent in terms of skilled workers and financial resources. Small Farmsteads Many of the same features mentioned above may also be present at the small farmsteads, but they are usually smaller in number as well as in scale. Thus, size is the main determining factor in this analysis. Field Shelters Many smaller, more-isolated stone structures were found in the fields, away from the main agricultural complexes and their associ- ated features. They appear to be intended to provide shelter for families, farmers, or watchmen while they cared for the crops, f particularly grapes, at critical times of the year. In the Bible, these structures were probably included by the term migdal im,7 which ' can be divided into two groups: signal towers built by the state and t 'O• Borowski, Agriculture in Iron Age Israel (Winona Lake, IN: Eisenbrauns, 1987), p. 106. i 1 224 GERATY, HERR, AND LAIBIANCA those, such as ours, built by private individuals. Few, if any, of the structures found within the `Umeiri regional survey could be classi- fied as military or state-run signal towers. In general, our structures were not built on high hilltops and did not command strategic views of the surrounding region, at least from a military point of view, but were usually located on spurs of hills overlooking specific agricultural fields, often in association with terraces and field walls. In short, they appear to be small (albeit frequently of -megalithic" construction), privately-built agricultural watchtowers intended for long-term use associated with intensive agricultural production. I Forts There was at least one site that could properly be classified as a fort in the sense of the biblical biraniyot.8 This site was located on the summit of a high hill (see Plate 6).9 It was not only strategically located, providing an excellent view in any direction, but it also appeared more than adequate in both size and design to house a military garrison. The interior was divided into several rooms of varying site, undoubtedly serving different purposes such as storage, food preparation, and living. Caves and a large cistern were located nearby; but field walls, terraces, winepresses, and other features generally associated with the farmhouses appeared to be missing. All sherds collected by the survey were exclusively Iron I and II. I w Kilo is Virtually all of the structures described above-large and small farmsteads, field shelters, and forts-were rectangular in shape. However, as noted in the 1984 preliminary report, a large number of the structures were circular in shape. Additional examination in 1987 led to the realization that there was remarkable uniformity in size (5 in in diameter) and construction components (small field stones) for virtually all the circular structures. Initially it was thought that these structures represented variant forms of either farmsteads or field towers. However, the thick walls (1 m) and lack of any obvious entrance and the discovery of plaster oil the interior walls of the structures suggested some other function. Excavation of one of these circular structures revealed a large amount of intramural ceramic slag, suggesting that they were originally con- structed as lime kilns (see Plate 7). "A. "Iron .'Age Foruesses in the ,Juch4ean Hills'', PEQ July-De(ember 1982, 1). 10H6. 'Coorclinatcs 2i28.1398; 11m(h's Site 135: FoltIcI s Site D. I SECOND SEASON AT TELL EL-`UMEIRI 225 r a Plate 6. Remains of an Iron Age military fort as found by the regional survey. Plate 7. Lime Kiln located at `Umeiri Site 50. r r t I l 226 GERATY, HERR, ANTI LABIANCA 2. Pujm Selirn'o Regional Survey Site 34, Rujm Selim, was chosen for a sound- ing because of its several "agricultural'' characteristics identified during the 1984 survey. These included two cisterns, a 'tower''-like structure,_ and several cup holes and quarry marks in the vicinity. Furthermore, the pottery identified during the survey revealed a predominance of Iron Age sherds. ,ogether these characteristics suggested an agricultural settlement of the Iron Age (see Plate 8). Excavation in five Squares during the 1987 season revealed six, apparent phases of occupation. Altho_rgh there were no clear occu- pation loci from the earliest phase, Phase 6, the late Iron II period was evidenced by the main "tower" s ructure, with its four rooms, as well as the plastering of the structure and perhaps the cutting and plastering of the lower cistern. Phase 5, apparently late Iron 11' cLIrly Persian in date, produced a trilobate (Scythianj arrowhead found outside what appeared to be a perimeter wall. Inside this wall, a courtyard was discovered to have been leveled with fill dirt and then possibly cobbled and provided with a plaster installation. Inside the "tower" structure itself were found several interesting household remains, Including two ceramic loom weights, a spindle whorl, and sherds from a Persian water jug. Most of the evidence for Phase 4 ame from the interior of the square structure. Late Hellenistic ir~ date, it also provided the majority of the objects found at the site: a well-preserved coin of Ptolemy II (ca. 280-213 B.c.) (see Platte 9), a bronze pin clasp, two spindle whorls in situ, as well as a whole lamp. The interior of the "tower" structure appeared to have been remodeled at this time by collapsing the four original rooms of Phase 5 into two. In Phase 3, probably Roman in elate, the "tower" structure, which measured 9 x 9 m, was remodeled. One of the rooms con- tained the remains of a cobbled surfitce which was covered with smashed pottery of the Hellenistic,, Roman periods. During Phase 2, tentatively dated to the Ottoman period, there was evidence that the upper cistern was used as a dwelling, steps having been cut down into the cistern on the south side. "'Field Sulx•r%isur Inr Rujm Selim %%as Lurita Hshhard, assisted by square Sulxr~is,ns Jamrs \lillet, I,)dd Sanders, and 1,1()%(l Willis; volunteers inducted hristy I-lapses. Tam,ua Ho(Icr, Julio Ju,uez, Dons Straws. and Ronda Westman. i i SECOND SEASON AT TELL EL-`UMEIRI 227 I i r~. s -d P j i Plate 8. Aerial view of Rujm Selim, a large agricultural complex. ,`Yl sy Xy } },. Obverse (A) 4 Reverse (B) : Plate 9 (A and B). Coin of Ptolemy II found at Rujm Selim. t i 228 GERATY HERR AND LABIANCA Phase 1 evidenced seasonal occupation by modern nomads: t campfire remains, tent pegs, nails, and soda-bottle fragments. t It would appear that Rujm Selim, with its rather lengthy occupational history (it was abandoned in the late Persian to early Hellenistic centuries), fits the first category of structures analyzed above, that of the large agricultural complex. Only some two kilometers from Tell el-jUmeiri, it probably served the latter as a "daughter" agricultural site during the Iron Age and early Persian period, but was independently resettled in the Hellenistic period, when major sites in the region do not seem to have existed. 1 3. Stratigraphic Excavations at Tell el-`Umeiri Previous work by the Madaba Plains Project, during five seasons of excavation at Tell Hesban in the 1960s and 1970s and one at Tell el-cUmeiri in 1984, has suggested the hypothesis that a series of five broad cycles of settlement intensification and abate- ment took place in the frontier region of central Transjordan. Cycle 1: Prior to the Early Bronze Age a coherent picture of general regional intensification and abatement in settlement pat- terns is not available. From tine to time specific sites were settled intensively, but broad regional settlement patterns have not yet been documented. Beginning with the Early Bronze Age, however, surveys have shown large increases yin inhabited sites.'' EB III seems to have been the period when Tell el-cUmeiri was most extensively settled. In EB IV, however, the cycle seems to have begun the abatement process, with inhabited sites decreasing in quantity and quality until, by the Middle Bronze Age, very few sites have been located. Tell el-`Umeiri was, however, a glaring exception with regard to the Middle Bronze Age. Cycle 2: The period of abatementt continued through the Late Bronze Age, although Tell el=cUmeiri was still occupied, until the Iron I period, when settlements began to increase again. Intensi- fication continued through the Iron II period, when a climax seems "See especially the Hesban survey, torthcdrming as Hesban > and authored by Robert lbach, Jr.; also our 198-1 survey in Geraty, et al., "Nladaba Plains Project: A Prelirninarv Report of the 1981 Season at Tell1 el-Umeiri and Vicinity.'' B,-4SOR Supplement 24 (1986):125; and, among others, J. M. Miller, ";archaeological Survey of Central Moab, 1978,'' 13 1SOR 234 (1979): 41-52; and B. MacDonald, "The Wadi el-Hasa Survey 1979 and Pie ions Archaeological Work in Southern Jordan," 1L-1.SOR 2-15 (1982): 35-52. I 1 1 SECOND SEASON AT TELL EL-`UMEIRI 229 to have been reached during the seventh and sixth centuries B.C., a time when many major and minor sites have been located in the region. Cycle 3: Little is known of the late Persian and early Hellen- istic • istic periods, but, beginning with small late-Hellenistic settlements, during a time when Rujm Selim seems to have been flourishing, - the process of intensification began again, building slowly through the Roman centuries and reaching its zenith in the Byzantine era, when, except for the Modern Period, the region seems to have been the most heavily populated. Tell el-`Umeiri (East) was occupied during these periods, rather than the earlier Tell el-`Umeiri (West). The evidence is very strong that there was only a slight abatement during the initial years of Islamic rule, but when the caliphate moved to Baghdad with the Abbasids, the region seems to have been only lightly inhabited. Cycle 4: Perhaps due to the region's importance to the Islamic reconquest of the Holy Land from the Crusaders, settlement again increased during the Ayyubid and Mamluk periods, from which large numbers of sites, including Tell el-`Umeiri (North), have been found. But with Turkish control, intensification ceased and another period of abatement began. Cycle 5: Few settlements seem to have existed in the region until late Ottoman times, when cave villages, such as Tell el- `Umeiri (North), and fortified farmsteads began the fifth cycle of intensification-a cycle which has carried on unabated until the present. Excavations of the site had two goals related to this under- standing of cyclic intensification and abatement: (1) The hypothesis needed to be tested by excavation. This had been done initially by i the Madaba Plains Project for Cycles 3-5 and, to a lesser extent, Cycle 2 at Hesban, but Tell el-`Umeiri with its Bronze- and Iron- Age occupation allowed much more detailed testing from the earlier ,4 cycles. How did a major site reflect the cycles of intensification and abatement? (2) Exceptions to the hypothesis needed to be examined to understand how sites, occupied during periods of abatement, functioned. Indeed, this was a major reason for the choice of Tell el-cUmeiri for excavation, for preliminary surveys had suggested occupation during the abated Middle and Late Bronze Ages. How did a major site function during periods when sedentary regional support systems were not in evidence? 230 GERATY, HERR, AND LABIANCA In 1984, four fields of excavation had been opened on the site to examine these questions-Fields A, B, C, and D. In 1987, three of the four-A, B, and D-were expanded, while one-Field C- was diminished. In addition, two new fields were opened-Fields E and F. (Again see Plates 2, 3.) The Ammonite Citadel: Field A12 Examination of the last major period of occupation at the site was continued at the western edge of the acropolis, where four squares had been opened in 1984. Four new squares were laid out north of the 1984 excavation in order to outline the northern limits of the Ammonite Citadel discovered in 1984 and to connect with the expansion of Field B, the western defense system (see Plate 10). It was hoped that the detailed study of this major building in one of the most important parts of the summit would help answer the questions regarding the processes of abatement at the end of Cycle 2. How did the remains reflect the abatement processr When did it occur% Did it occur suddenly or over a length of time% Whereas the 1984 excavations had been inconclusive regarding these questions, the finds in 1987 suggested at least tentative answers. The two major phases discovered during the previous season were again encountered, but additional information was also dis- covered regarding: (1) the initial construction of the citadel; (2) the citadel's western and northern limits; and (3) occupation of the area after the citadel went into disuse, Although little is vet known of the Iron I settlement, excava- tions at the western edge of Field A; uncovered fragmentary Iron I walls (for a house%) and a deep debris deposit immediately to the west of the citadel and retained by one of the foundation walls that had been in use with the earliest please of the citadel. No other Iron I remains were found at similar levels inside the citadel. 1"'hhe staff was divided into four sections, responsible for excavation, regional survey, laboratories, anal camp logistics. In charge of planning and overall execution of the project were Lawrence -I . Gt'ntty, Lairy G. Hell, and Oystein S. LaBianca, co-directors of the project. hhe excavation stall, supervised bs Hetr, included six fields of excavation on the tell, and also one at Rujm Selim (a small agricultural site described above). Each field utilized (lilt' local workman per square. Field Supervisor for Field A, the Ammonitc Citadel, Was .John l,awlor, assisted by Squ;ure Supervisors Nicholas kronwall, Desmond Pons. 'I-horuas Pons, and Nazmieh Rida: volunteers included ,Dunes Beers. Chant's Castleber;g, Monique Escamilla. Sharon PertlcN. Nlalcolm Putts. Steven Russell, Junes Sawtell, and Dt'i a Look. ~` SECOND SEASON AT TELL EL-`UMEIRI 231 1 +}. rr # " ~r ty t~ - i 232 GERATY, HERR, AND LABIANCA l Plate 11. Foundation walls of the Ammonite citadel in Field A. Likewise, no foundation trenches have been found for any of the walls of the citadel. It would seem that the builders dug a large foundation area into the Iron I settlement in order to construct the foundations of the citadel (see Plate 11). Jirst how deep this founda- tion was dug is still unknown, because the bottoms of the citadel walls have not yet been found: but the lowest floors so far en- countered were ca. 1.4 m below the top' of the Iron I walls. The lack of foundation trenches indicated that this was a large-scale excavation intended to clear a large area for a series of basement rooms. It was thus a major alteration of the plan of the Iron-I settlement. It is here suggested that this construction was part of the overall intensification process taking place at Tell el-`ITmeiri and its region, III which a large public structure was built to meet the needs of a strengthening economic arid social order. I Only two of the exterior limits of the citadel have been un- covered: the west (where the foundation cut was discovered) and the north (where a pillared structure, possibly a house, has been found). Neither exterior wall was found to be sl5ecially strengthened, sug- gestiug that the citadel may have been a complex of buildings with an outer fortification wall. But the latter has yet to be discovered. ~~ 1 SECOND SEASON AT TELL EL- `UMEIRI 233 Unfortunately, no other data were found beyond those dis- covered in 1984 that would lend information concerning the specific function of the building: the thick walls and large size of the complex do not fit a domestic interpretation. The work of the 1987 season confirmed the monumental nature of the complex, which measured at least 17 m north to south and 12 m east to west, with no signs of the southern and eastern limits. Precisely when the citadel was constructed is not yet known, because excavations have not reached founding levels. The pottery from the earlier of the two phases so far isolated reflected the r corpus typical to central Transjordanian plateau sites in the late Iron II period. More can be said, however, about the citadel's end: Within one of the debris layers sealed below the surface of the upper phase of the citadel was an Attic sherd possibly dating to the fifth century B.C. In the rest of the earth layers and surfaces of the upper phase, other sherds of early-Persian date appeared together with those of the late Iron II corpus. It would thus appear that the citadel was in use well into the Persian period. The ceramic rhyton (see Plate 12) and pot found on the upper surface would support this conclusion. (A) (B) Plate 12 (A and B). Ceramic rhyton from Field A. s a 234 GERATY, HERR, AND LAB ANCA The abatement process does not seem to have been sudden. No signs of destruction were found in the upper levels of the last citadel. Instead, after the citadel had gone out of use, builders constructed two large semi-underground installations into the area north of the citadel, destroying much of a possible pillared house which seems to have been contemporary with the citadel. The first installation was a small plastered pool whose interior measured 2 x 2.75 to and was ca. 2 m seep (see Plate 13). The installation was buttressed strongly on all four sides by over a meter of well-laid stones. Five steps descer ded steeply into the pool from the north (the bottom two were narrower than the upper three), but the buttressing stones indicated that most likely one, and possibly two, others existed originally. The interior was covered with two layers of plaster, suggesting that it had been repaired and thus used over a period of time. The superstructure of the pool probably was constructed of finely-hewn ashlar blocks, because about fifteen such stones, not in evidence anywhere else on the site, were found in the fill inside the pool and in its immediate vicinity. Associated with this installation were two fragmentary surfaces that contained pottery from the late Iron II and early Persian periods, similar to the pottery found in tr the fill inside the pool. d ` si se st si tl ci st __ f tl ! hi T A 91 (, alt R Plate 13. Plastered pool in Field A. 1 I 1 i SECOND SEASON AT TELL EL-`UMEIRI 235 i } t i Plate 14. Stone-lined silo in Field A. The second installation was a narrow, stone-lined silo, measur- ing ca. 1.25 m in diameter at the lip and narrowing slightly as it descended to its bottom 2.80 m deep (see Plate 14). The fill from the silo contained nothing that suggested its function, but it may have served to hold jars which were lifted by means of rope or a hooked stick. It would thus seem that after the citadel went out of use the site continued to be occupied, probably on a less intensive level I than before; that is, the public structure of the citadel was appar- ently judged unnecessary, and no more large structures with walls sufficiently large to thwart destruction by farming activities were built. In this abated state, however, the site lasted long enough for the pool to receive a second coat of plaster. No evidence of further building activity was found in Field A. The Western Defensive System: Field B 13 A significant indicator for the processes of intensification and abatement at a settlement is the presence or absence of fortifications. f 13Field Supervisor for Field B, the Western Defensive System, was Douglas Clark, assisted by Square Supervisors Gillian Geraty, Gary Kent, David Merling, and Gotthard Reinhold; volunteers included Hans-Dieter Bienert, Caroline Cameron, Rafael Figueroa, Vanessa Martin, Kevin Nelson, Nora Peppers, Erwin Syphers, and Janelle Willis. a i 236 GERATY, HERR, AND LABIANCA The objectives for Field B on the western slope of the summit were to examine the changes which took place in the defenses of the site through time. For this reason, four squares and part of another had been excavated in 1984, and in 1987 the field was expanded to seven squares. In this way, the complete slope was excavated and con- nected with the northern squares of Field A. In one small area, a probe uncovered ashy destruction debris with pottery dating to the Late Bronze and early Iron Ages, includ- ing a Late Bronze biconical painted jug with handle on the shoulder. However, not enough is yet known of the Late Bronze settlement to be able to answer the question of just how much abatement had occurred between Cycles 1 and 2 at `Umeiri in the Middle Bronze Age and Late Bronze Age. By the late Iron I period (or earlier) it seems that the site was surrounded by a casemate wall and beaten-earth rampart system. i Such a development suggested that the intensification process for Cycle 2 was already well under way. An addition to this system, discovered in 1987, was a revetment wall, made of a single line of large boulders, at the bottom of the rampart (see Plate 15). Just inside the casemate wall, which seems to have continued through the Iron II period, were two phases of fragmentary walls and surfaces dating to the early Iron II Period. Farther to the east, in the two squares immediately to the north of Field A, two phases of fragmentary domestic architecture included stone walls; cobble, plaster, and beaten-earth surfaces; hearths; and pits. Unfortunately, not enough remained for a clear plan to emerge. Excavation of the foundation trench for the plastered pool, mostly in Field A, cut through these occupation levels, leaving portions of late Iron II holemouth pithoi in situ. One pithos base was sliced vertically through the middle. Activity patterns for these domestic areas thus included storage. A corner of the Field A plastered pool extended into Field B, where it seems to have been in use with the uppermost walls and surfaces in the northern part of the field, as evidenced from both the 1984 and the 1987 excavations. All walls and surfaces were, however, fragmentary and were difficult to interpret beyond obvious non-monumental characteristics. The masonry of the walls in- cluded a wide variety of stone cutting, suggesting that the occupants did not have the leisure and/or resources to construct fine structures. This confirmed the suggestion for Field A that this phase repre- sented the abatement process at the site, when, perhaps, the settle- ment was forced to adopt new, less-intensive strategies to exist. SECOND SEASON AT TELL EL-`UMEIRI 237 r v r Y t l h i Plate 15. Revetment wall and rampart in Field B (RW, revetment wall; CW, casemate wall).