Bulletin of the American Schools of Oriental Research, No. 236. (Autumn, 1979), pp. 61-79.

(The Iron Age Fortresses in the Central Negev, Rudolph Cohen, 1979 AD)

The Discovery of the Fortresses

The remains of several fortresses (Qasr er-R ukieibeh, Bir Birein, and Tell 'Ain el-Qudeirat) were first observed in 1914 by C. L. Woolley and T. E. Lawrence during their archaeological survey of the "Wilderness of Zin" (1914-15: 27-28,40-43,64- 67). On the basis of the potsherds found in the debris they dated the fortresses to sometime in the late second to early first millennium B.C. In addition, they identified the pottery as "Syrian" in type, concluding from this that the fortresses had not been Egyptian outposts (1914-15: 27-28). Concerning their specific historical background, they alternatively suggested that these fortresses may have belonged to the Patriarchal age, or that they were connected with the Red Sea "adventures" of one of the "Jewish" kings (1914-15: 41). According to the Bible Solomon established a maritime base at Ezion-Geber, near Eilat, in order to carry out trade with Ophir (1 Kgs 9:26), and Jehoshaphat and Uzziah both attempted later to repeat his success (1 Kgs 22:49; 2 Chr 20:36*). Thus, in the view of Woolley and Lawrence, the fortresses had probably served as "military police stations" designed to guard the caravan routes (1914-15: 28). Although in many respects their phraseology and descriptions are now obsolete, their discoveries and conclusions have provided a solid basis for subsequent study.

The Surveys of Glueck and Aharoni

Research on the fortresses resumed in the late 1950's and 60's. Extensive archaeological surveys, directed chiefly by Nelson Glueck (1959c: 146-86; 1961: 12-14) and Yohanan Aharoni (1967: 1-17), located numerous similar fortress remains, the majority of which were also Iron Age in origin. It rapidly became evident that a veritable fortress network had once existed in the Central Negev.

These fortresses consisted of a casemate wall around an inner courtyard but apart from this there was substantial variation in groundplan and architectural detail. Aharoni accordingly divided the fortresses into four principal types (1) fortresses with projecting towers, including those of Kadesh-barnea, Uza, and Tel Arad; (2) rectangular fortresses without towers, including those of Nahal

Raviv, Qasr er-Ruheibeh, Be'er Har Boger, Mishor Ha-Ruah, and the fortress north of Kadesh-barnea; (3) irregular fortresses, including those of `Ain Qudeis, Givcat Refed, Nahal Lacana, and Yotvata; (4) fortresses surrounded by a polygonal wall, including the fortress above En-gedi, and the one at Har Hesron (Aharoni 1967: 2-11).

Glueck and Aharoni maintained that the fortresses had been erected by the kings of Israel between the 10th and the 7th centuries B.C. in order to extend royal control over the Negev, the Aravah, and Eilat, and to protect their caravan routes (Aharoni 1967: 11-13). This southern road system was one of the Monarchy's chief economic assets, and thus it should be considered in greater detail.

The Negev Highway System

The ancient and extensive highway system in the Negev was undoubtedly determined by the harsh geographical conditions of the region. A number of its arteries are named and described in the Hebrew Bible. The "Way of the Spies," for instance, was the course taken by the Israelites who, spurning Moses' admonition, attempted to enter the Promised Land from the south, but were routed and repelled by the king of Arad (Num 21:1). From this account it follows that the "Way of the Spies" was the principal route from Kadesh-barnea to Arad, and during the time of the Monarchy fortresses and settlements were planted at strategic points all along its length, the remains of which can still be seen at Tel Arad, Givcat Refed, in the Sede Boger area, at Nahal Lecana, Beer Hafir, and Kadesh-barnea.

Fig. 1. Map showing principal Iron forts in the Central Negev.

A branch of this highway, turning south in the area of Avdat, continues towards the region of Mishor ha-Ruah, and from there via Be'erot `Oded and Har Rimon to Yotvata. This is the route which leads, in opposite directions, to the Dead Sea and Eilat.

The "Way of Shur" (Gen 16:7; 20:1) proceededfrom Beersheba through the areas of Halusa, Nisana, and Kadesh-barnea, and from there into the Sinai interior in the direction of Ismailiya. During the Monarchy it was guarded by the fortresses of Ruheibeh and Be'erotayim.

The Bible mentions additional roads belonging to the same complex: the "Way of Mount Seir (Deut 1:1-2); the "Way of the Mount of the Amorites" (Deut 1:19); and the "Way of the Red Sea" (Exod 113:18; Num 21:4; 14:25; Deut 1:40; 2:1). The exact course of these highways, however, is difficult to establish at present, and it is not known if fortresses were likewise erected along their lengths. Aharoni also distinguished another main road in the eastern Negev, the "Way of Edom," which led from Arad to H. `Uza, and from there descended along Nahal Qeni towards the Dead Sea and Aravah (1967: 11-12). But since this road is basically in the region north of the Central Negev, neither it nor its fortresses will be discussed in this article.

Recent Surveys

In summarizing the latest data in 1967, Aharoni predicted that more fortresses would surely be discovered in the future (1967: 2) He was absolutely right. Benno Rothenberg's survey in the same year supplied additional sites for the vicinities of Nahal Horsha and Kadesh-barnea (1967: 71-78). Between 1965-70, several fortresses were found in the area of Sede Boger by a team of the Archaeological Survey of Israel, led by the author (Cohen et al. 1965: 263-64; Cohen 1970: 6-24). In 1971, during the author's survey on behalf of the Department of Antiquities, a number of fortresses were located in the Ketef Shivta and Ramat Nafha vicinities (Cohen 1971: 29-30). In 1975, a fortress was found near Nahal Sarpad during Z. Meshel's survey of this area (1975: 30).

Excavations

Knowledge about the fortresses of the Central Negev, limited formerly to the results of the surveys, received a new dimension with the partial excavation of 12 of the sites. The ceramic remains recovered in these excavations enable the fortresses now to be dated more accurately than was possible on the basis of the surveys alone. The first of these excavations was carried out by the author in 1965-67 at the fortress of 'Atar Haroca (1970: 6-24), north of Sede Boger. Between 1969-72, excavations were conducted at four fortresses, under the joint direction of Z. Meshel and the author: Mesad Refed, 1970; Mesad Hatira, 1970 (Cohen and Meshel 1970: 27-28); H. Ritma, 1971 (Cohen and Meshel 1971: 28-29); and H. Har Boger, 1972 (Cohen and Meshel 1972: 37). Rescue excavations were carried out by the author on behalf of the Department of Antiquities at the following sites: H. Haluqim, 1971/ 72 (Cohen 1976c: 34-50); H. Rahba, 1974 (Cohen 1975a: 171-72); H. Ramat Boger, 1975 (Cohen 1975b: 40); H. Mesora,1975/76 (Cohen 1977a: 170-71); H. Ketef Shivta, 1976 (Cohen 1976b: 53-54); Qasr er-Ruheibeh, 1976 (Cohen 1976b: 59-60); and the fortress near 'Ain Qudeis, 1976 (Cohen 1977b: 171). The author also conducted four seasons of excavations between 1976-78 at the important fortress site of Tel Kadesh-barnea (Tell 'Ain el-Qudeirat; Cohen 1976a: 201-2).

Implications of the New Material

The new information yielded by these recent surveys and excavations warrants a reappraisal of the views in general currency regarding the fortresses and their historical role. The author would first of all suggest a slight revision in the fourfold classification of the fortresses according to their groundplan, similar to Aharoni's in overall conception but differing in several specifics. The proposed reclassification is as follows:

I. Roughly oval fortresses

2. Rectangular fortresses

3. Square fortresses

4. Fortresses with towers

To this, the most common, type of fortress, having a roughly oval groundplan and comprising several casemate rooms around a central courtyard (cf. fig. 2), belong the following sites: 'Ain Qudeis, 'Atar Haroca, H. Haluqim, Ketef Shivta, Rahba, Mesudat Nahal Horsha, and H. Nahal Sarpad.

Fig. 2. Iron Age fortress at 'Ain Qudeis.

b. 'Atar &trace,. The similar fortress of 'Atar Haroca (Grid Reference 1359 X 0352) is situated on a low hill about 5 km. northeast of Sede Boger. The site was surveyed in 1965 by a team of the Archaeo-logical Survey of Israel, led by the author, who also carried out excavations in 1965 and 1967 on behalf of the Department of Antiquities (Cohen 1970: 624). The diameter of the fortress, like that at 'Ain Qudeis, is ca. 50 m., and the gate and four out of 17 casemate rooms were uncovered (fig. 3: 2). The walls, ca. 0.50 m. wide and preserved to aheight of ca 1.00 m., are of rough-hewn local silex blocks. The size of the casemate rooms varies: width ca. 2.50-3.00 m.; length ca. 5.50-10.00 m. The gate to the east consisted of an open space of ca. 3.00 m. between two casemate rooms, but this was narrowed to ca. 1.00 m. by two piers of particularly massive stones. This fortress, too, contained both wheel-made from the 10th century B.C. and crude hand-made pottery, as well as stone implements, such as querns—all found in a layer of ashes on the floors of the casemate rooms. It showed only one phase of occupation. The remains of a settlement are visible on the same hill. It consists of at least 12 mostly small buildings with one to four rooms and a courtyard. One unusual building had four rooms around a central courtyard (fig. 3: 6).

c. Haluqim. A smaller fortress of this type is the one at H. Haluqim (Grid Reference 1310 X 0335), situated 2 km. northwest of Sede Boger on the eastern slopes of Har Haluqim (Jebel Haleiqum). The site was surveyed in 1953 by Anati, in 1958 by Aharoni, and in 1965 by a team of the Archaeological Survey of Israel led by the author, who in 1971-72 carried out excavations on behalf of the Department of Antiquities (Cohen 1976c: 34-

Fig. 3. Oval fortresses in the Central Negev: 1. 'Ain Qudeis; 2. 'Atar Haroca; 3. Ijorvat I-jaluqim; 4. Irlorvat Rahba; 5. Horvat Ketef Shivta; 6. Ramat Matred Fort 146.

Fig. 4. Iron Age fortress at Horvat

50). The fortress has a roughly oval ground plan (ca. 23 X 21 m.) and comprises eight casemate rooms around a central courtyard (fig. 3: 3). The walls, ca. 0.80 m. wide, of which only the foundations are preserved, are of rough-hewn local limestone blocks, resting on bedrock. The size of the casemate rooms, all of which were excavated, varies: width ca. 1.50-2.00 m.; length ca. 5.50-8.00 m. A circular stone basin was found in one of the casemate rooms, and three rectangular stone basins in another (fig. 4). The gate, on the southeastern side of the fortress, was extensively damaged by the foundation trenches of a later structure built over it. The pottery remains, found in a layer of ashes ca. 10 cm. thick that covered the beaten-earth floor of the casemate rooms, conformed to the two main classes mentioned above. The wheel-made variety included jars, kraters, and juglets, but hand-made cooking-pots were the most common vessel. The remains of a nearby settlement can be seen along three tributaries of Nahal Haroca; it consists of 25 buildings, including seven of the so-called "four- room" pattern (of which two were excavated: cf. Shiloh 1970, 1978), four cisterns, and stone terraces (Evenari, Shanan, and Tadmor 1971: 94-119).

d. Ketef Shivta. The fortress of H. Ketef Shivta (Grid Reference 1185 X 0347) is situated on the eastern spur of the Ketef Shivta range, next to Nahal Mesora. The site was first surveyed in 1971 by the author, who also carried out a small-scale excavation in 1976 on behalf of the Department of Antiquities (Cohen 1971: 29-30). The fortress has a roughly oval groundplan (ca. 36 X 25 m.), and the gate and five of its casemate rooms were exposed (fig. 3: 5). The walls, ca. 0.60 m. wide and preserved to a height of c. 1.00 m., are of rough-hewn local limestone blocks. The casemate rooms are ca. 1.70 m. wide, but their lengths have not been deter-mined. The gate, in the eastern wall by the southern corner, consists of a space of ca. 2.50 m. wide in the line of the casemate wall, but it is narrowed quite considerably by piers extending from either side. The pottery, found in the ash layer that covered the floor of the casemate rooms, included the wheel- made variety characteristic of the 10th century B.C. as well as the hand-made "Negev" type. One level of occupation was detected in the fortress, but below the hill are traces of a settlement comprising a number of buildings.

e. Rahba. The fortress of H. Rahba (Grid Reference 1526 X 0509), the largest of this type, is located on an oblong hill about one kilometer southeast of Dimona in a notably strategic position. The site was surveyed in 1937 and 1938 by G. E. Kirk, who called it Kh. 'Umm er-Tin and identified it as Roman-Byzantine in origin (Kirk 1938: 220). It was resurveyed in 1955 by Glueck (1957: 22-23); and in 1965 by a team of the Archaeological Survey of Israel, led by the author, who also carried out in 1974 rescue excavations on behalf of the Department of Antiquities (Cohen 1975a: 171-72). The fortress is oval in plan, with a diameter of ca. 75 X 60 m., and apparently follows the shape of the hill (fig. 3:4). The walls, ca. 0.60 m. wide, and preserved to a height of ca. 0.50 m., are constructed of rough-hewn local limestone blocks and rest on the bedrock. Eight casemate rooms of various sizes were exposed: width ca. 2.10 m.; length ca. 5.00-10.00. The floors of the casemate rooms were mostly formed of beaten earth and were covered by a layer of ashes in which were found wheel-made pottery of the 10th century B.C., including jars and juglets, hand-made "Negev" pottery, and other utensils such as grinding stones. The exact number of casemate rooms, as well as the position of the gate, is unknown, as above the remains of the Iron Age fortress a large structure (ca. 57 X 33 m.) was erected in the Roman-Byzantine period. No remains of a nearby settlement have been perceived.

f. Nahal Horsha. Nahal Horsha (Grid Reference 1048 X 0142) is located near Nahal Horsha, southeast of its confluence with Nahal Lacana. It was first surveyed in 1965 during a field trip organized by the Sede Boger Field School, with the participation of Meshel, Tsafrir, and the author (Cohen et al. 1973: 40). Its plan is roughly oval (ca. 34.30 X 21.50 m.), following the flat hilltop on which it was erected. It consists of a casemate wall around a central courtyard, with some pottery from the 10th century B.C. No traces of a settlement were sighted by the survey.

g. Nahal Sarpad. Another roughly oval fortress (Grid Reference 1114 X 0073), on one of the hills overlooking Nahal Sarpad, was discovered and surveyed in 1975 by a team from the Sede Boger

Field School, headed by Meshel (1975: 30). The fortress measures ca. 37 X 25 m. and consists of casemate rooms around a central courtyard. The walls are preserved in places to a height of ca. 1.50 m. Both wheel-made Israelite and "Negev" sherds were collected, but no remains of a settlement have yet come to light.

Most of the fortresses of this generally rectangular type correspond in their groundplan to the topography of the hill on which they were founded, and because of this their four sides are seldom equal in length (cf. fig. 5:1-4). Seven fortresses of this type were either surveyed or excavated during the past decade: Mesad Refed, Mesad Hatira, Har Boger, H. Ramat Boger, Mesad Nafha, Mesad Har Sacad, and Mesad Mishor Ha-R uah.

a. Mesad Refed. The fortress of Mesad Refed (Grid Reference 1490 X 0467), located on the hill of Refed in the midst of the Yeruham plain, was sighted and surveyed by Glueck in 1956 (1957: 23, Site 304) and partially excavated by Meshel and the author in 1970 (Cohen and Meshel 1970: 27). The overall plan of the fotress is roughly rectangular (ca. 50.70 X 28 X 56 X 47.50 m.) and consists of casemate rooms around a central courtyard (fig. 5:1). The walls, ca. 0.40-0.50 m. wide and preserved to a height of 2-3 courses, are of rough-hewn local limestone blocks. Excavations were undertaken in 12 casemate rooms, but the gate was not discovered. The casemate rooms vary in size: width ca. 2 m.; length ca. 5-9 m. It appears that an inner wall ran across the courtyard, thus dividing the fortress into two adjacent wings. The floors of the casemate rooms were formed of beaten earth, and some were overlaid with a clear layer of ashes in which both wheel-made and "Negev" pottery was found. One level of occupation was discerned in the fortress. No remains of a settlement have yet been perceived.

b. Mesad Hatira. The fortress of Mesad Hatira (Grid Reference 1484 X 0454), less than 1.5 km. distant from Mesad Refed, is located on a narrow hill that rises ca. 20 m. over the edge of the Yeruham plain. It was discovered by Glueck in 1956, at the time of his Negev survey (1957: 22-23, Site 306), and partially excavated by Meshel and the author in 1970 (Cohen and Meshel 1970: 27). Here, too, the plan of the fortress conforms to the area of the hilltop, with the casemate rooms skirting its edge, enclosing an area of ca. 20 X 41 X 80 X 65 m. (fig. 5:2). A row of additional rooms is joined in several sections to the line of the casemate rooms fromwithin, and there also appear to be walls extending to the courtyard's center. The fortress walls, of rough-hewn local limestone, are ca. 0.50-0.60 m. wide and preserved to a height of ca. 0.50 m. The excavations focused on the eastern side of the fortress and uncovered the gate, already discerned by Glueck, and the adjacent casemate rooms. The gate (ca. 2.50 X 2.50 m.) is built into the courtyard from the inside wall of the casemate row, which leaves an open square at its front, to which a road climbs up the slope of the hill. The casemate rooms vary in size: width ca. 2 m.; length ca. 5-10 m. Of the rooms along the sides of the gate, one long room (ca. 3.50 X 9 m.), attached to the casemate row from within, was also uncovered. Its floor was a surface of natural stone, on which were found a large pithos, several jars, and a deep bowl, all broken. Only one phase of occupation was discerned, but no signs of conflagration have as yet come to light. The sherds include both the usual wheel-made and "Negev" types. Not too distant from the fortress are the traces of a settlement; one of it several buildings was exposed, but its groundplan was not determined exactly.

Fig. 5. Rectangular fortresses and outlying buildings in the Central Negev: I . Mesad Refed; 2. Mesad Hatira; 3. Horvat Har Boger; 4. 4. Horvat Ramat Boger; 5. rectangular structure near Horvat Mesora; 6. Building II, Area C, 'Atar Haro`a; 7. Structure A, Horvat Ritma.

c. Har Boger. The fortress of H. Har Boger (Grid Reference 1257 X 0332) is situated some 4 km. northwest of Kibbutz Sede Boger on a flat hill belonging to the northeastern spur of the Har Boger range. It was first discovered and surveyed by Anati in 1955, in the course of his regional survey on behalf of the Department of Antiquities. He called it Mesad Ha-Kochav and described it as an irregular structure with four corner towers. In 1958 the site was resurveyed by Aharoni, who noted the casemate rooms and the layout corresponding to the shape of the hill (1967c: 6). In 1969 a plan of the fortress was drawn up by the Israel Survey Team of the South, headed by the author. Rescue exca-vations, carried out in 1971 by the author and Meshel on behalf of the Department of Antiquities, resulted in the correction of numerous details in the previous descriptions (Cohen and Meshel 1972: 37). The fortress is roughly rectangular, without towers, and the length of the sides is irregular: ca. 32 X 26 X 19.30 X 17.50 m. It comprised 11 casemate rooms around a central courtyard (fig. 5:3). The walls of the fortress, of rough-hewn local limestone blocks, are ca. 0.70 m. wide, and preserved to a height of ca. 0.50-1.50 m. The gate and all of the casemate rooms were uncovered. The gate is situated on the southern side of the fortress and consists of a diminishing space between two casemate rooms; the outer width is ca. 4.50 m., and the inner width cathe outer width is ca. 4.50 m., and the inner width ca. 3.50 m. The side of the casemate rooms varies: width ca. 2-2.50 m.; length ca. 4.50-7 m. One of the rooms on the western side is unusual in size and plan; it measures ca. 6 X 3.5 m. and is crossed by two parallel rows of three pillars each. The pottery, found in a thin ash layer, included both wheel-made and "Negev" types. Two levels of occupation were discerned in some of the casemate rooms: the level of the fortress' founding, which on the basis of the pottery can be ascribed to the 10th century B.C.; and a later level, in which some of the casemate rooms were reused, belonging to the 8th-7th century B.C. On the opposite bank of Natal Besor are traces of a small settlement that may have been connected with the fortress.

d. H. Ramat Boger. The fortress of H. Ramat Boger (Grid Reference 1281 X 0360) is located on a hill belonging to the southeastern spur of the Ramat Boger range, overlooking a wide expanse of the Naha' Boger valley. It was discovered in 1969 by the Israel Survey Team of the south, headed by the author, and surveyed again by Meshel in 1972. In 1975 a rescue excavation was carried out by the author on behalf of the Department of Antiquities (on the above, Cohen 1975b: 40). The fortress groundplan follows the contours of the hill, the three surviving walls measuring ca. 31 X 33 X 28 m. (fig. 5:4). The fourth wall was destroyed by the foundations of a later structure built over it. The walls, ca. 0.60 m. wide and preserved to a height of ca. 1.10 m., were of rough-hewn local limestone blocks. The gate was on the sourthern side of the fortress and was ca. 4 m. across. The size of the casemate rooms varies: width ca. 2 m.; length ca. 5.50-6 m. There was also one small room of unusual dimensions: ca. 2 X 1.50 m. The pottery, found in a layer of ashes covering the floors of the casemate rooms, belonged to the same two types recovered elsewhere. One phase of occupation was discerned in the fortress, though a Nabatean structure was built over its northern corner. To the east of the hill there were traces of a settlement, where a building of the "four-room" pattern was exposed. Two rows of columns, with five columns to each row, separated the couryard from the two parallel rooms, and a further row of six columns divided the perpendicular room from the other two rooms and the courtyard. All the columns were monolithic. The walls of the building were ca. 0.50 m. wide, and the pottery found on the floor of the rooms was identical to that of the fortress.

e. Nahal Nalha. The fortress of Natal Nafha

(Grid Reference 12815 X 02035) is located some 2.50 km. south of Avdat on a low hill overlooking the valley of Nahal Zin. It was surveyed first by Anati in 1955 and again by the author in 1966 (Cohen 1966b: 23-25). The fortress is roughly rectangular (ca. 18 X 14 m.) and consists of casemate rooms around a central courtyard. The gate was situated on the western side. One of the casemate rooms was exposed: width ca. 1.70 m.; length ca. 5.50 m. The finds included wheel-made and "Negev" sherds, and the remains of a small settlement were sighted nearby.

f. Mesad Har Sacad. The fortress of Mesad Har Sacad (Grid Reference 1309 X 0131) was erected on a small hill on the western slopes of Har Sacad descending toward the valley of Nahal Zin. It was first surveyed by Glueck in 1957 (1957: 23, Site 314) and again by the author in 1971 (1972: 39-40). This roughly rectangular fortress (ca. 21 X 16 m.) also consists of casemate rooms around a central courtyard. The walls, ca. 0.50 m. wide, are constructed of rough-hewn local limestone blocks. The casemate rooms are ca. 2 m. wide, and the gate appears to have been in the northern side of the fortress near the western corner. Animal pens were evidently attached to the southern wall from without. To the north of the fortress were found remains of a settlement comprising several buildings.

g. Mesad Mishor Ha-Ruah. The fortress of Mesad Mishor Ha-Ruah (Grid Reference 1318 X 0077) is located northwest of Mizpe Ramon on a ridge overlooking the wide expanse of Mishor Ha-R uah It was first surveyed by Aharoni in 1958 (Aharoni et a!. 1958: 234-35) and a rescue excavation was carried out by the author in 1966. The fortress is roughly rectangular (ca. 20 X 19.50 X 15 X 19 m.) and consists of three casemate walls and one apparently solid wall in the east (ca. 1.50 m. wide). These walls, ca. 0.50 m. wide and preserved to a height of ca. 1.20 m., are of unhewn local limestone blocks. Seven casemate rooms of carying sizes were distinguished: width ca. 2 m.; length ca. 2.50-6 m. The gate, on the eastern side, comprises a guardroom built into the solid wall and is protected by inner and outer abutments. The fortress of Mesad Mishor Ha-Ruah is similar in its ground-plan to the fortress north of Kadesh-barnea (Aharoni 1967: 6). Its finds include both wheel-made and "Negev" pottery. The remains of a settlement can be seen below the hill and on its eastern slope. The houses vary in size and shape, but several belong to the familiar "four-room" pattern. Between the houses are stone-walled enclosures, which may have served as animal pens. Eight circular cisterns, dug into the marl and lined with large unhewn stones, can be found below the slope.

3. Square Fortresses.

Fortresses of this type were apparently built to a standard groundplan of ca. 20 X 20 m. and include the following sites: H. Mesora, H. Ritma, the fortress of Nahal Raviv, and the fortess opposite 'Atar Haroca.

a. H. Mesora. The fortress of H. Mesora (Grid Reference 1221 X 0365) was erected on an oblong hill on the western spur of the range of Mesora, next to Nahal Besor (fig. 6). The site was surveyed by Glueck and Y. Feldmann in 1958 (1959: 102-4) and in 1966 by the Israel Survey Team of the South, headed by the author (Cohen 1966a: 30), who also carried out rescue excavations on behalf of the Department of Antiquities in 1975 and 1976 (Cohen 1977a: 170-71). The fortress plan is square (ca. 20 X 20 m.) and consists of nine casemate rooms around a central courtyard (fig. 7:4), which contains a hewn-out cistern. The walls, ca. 0.80 m. wide and preserved to a height of ca. 1.20 m., are constructed of rough-hewn local limestone blocks. Two of the casemate rooms were exposed: the southern room was ca. 9.80 X 2.30 m., and the eastern room was ca. 4 X 2.80 m. The location of the gate was not discerned. In the layer of ashes that covered the floors of the casemate rooms the usual wheel-made and "Negev" pottery was found. One phase of occupation is indicated. In the vicinity of the fortress are a number of other structures. Among them is a rectangular building (ca. 17 X 11 m.) consisting of seven rooms around a central courtyard (fig. 5:5). Three of these rooms were exosed, including an entrance-room in the eastern side, a small room (ca. 3 X 2.70 m.) to its south, and a longer room (ca. 8 X 1.90 m.) in its western side. Along the southern edge of the hill, in the space between the fortress and this building, is a row of casemate rooms (length, ca. 15 m.). One of its rooms, ca. 2.30 m. in width, was excavated. The pottery found in these structures belongs to the same two types as that found in the fortress.

b. H. Ritma. The fortress of H. Ritma (Grid Reference 1283 X 0347), located northwest of Sede Boger on a prominent hill of the Haluqim range overlooking Nahal Boger, is similar in size and shape to that of H. Mesora (fig. 7:2; cf. 7:4). It was surveyed by Glueck in 1934 (1934-35: 117), by Anatikg, in 1953, and in 1967 by the Israel Survey Team of the South, headed by the author, who also carried out excavations with Meshel in 1970 (Cohen and Meshel 1971: 20-29). The fortress has a square groundplan (ca. 21 X 21 m.) and comprises casemate rooms around a central courtyard. Surrounding the fortress are various other struc-tures. On the summit of the hill is a house of the "four-room" pattern (cf. fig. 5:7; fig. 8) and on the eastern slope of the hill are found the remains of ten buildings, presumably private dwellings, three of which were exposed. The architectural plans and the pottery of this fortress have now been fully published by Meshel (1977: 110-35). The pottery of Period III (figs. 6, 7) is typical of the 10th century wheel-made and "Negev" wares of the other forts. Periods II-I are Persian and Roman.

Fig. 7. Square fortresses in the Central Negev: I. Nahal Raviv; 2. Horvat Ritma; 3. small fortress near 'Atar Haro`a; 4. Horvat Mesora.

c. Nahal Raviv. The fortress at Nahal Raviv (Grid Reference 1075 X 0280) was surveyed by Y. Feldmann, at that time of Kibbutz Revivim, in 1957 (1959: 102-4). This fortress has a sqare groundplan (ca. 19 X 20 m.) and comprises casemate rooms around a central courtyard (fig. 7:1). Each of these rooms is connected individually to the courtyard by an aperture of ca. 1.80 m. in height, with some of the lintels, constructed of a single stone, remaining still in place. The walls are preserved to a height of ca. 2.50 m. The fortress gate was near the southeastern corner, and a small room, attached to the casemate wall from without, probably served as a guardroom.

d. Another fortress of this square type is situated on a hill facing 'Atar Haroca (Grid Reference 1361 X 0355). It was surveyed in 1953 by Anati, who attributed it to the Roman period, but in 1965 it was resurveyed and examined by the author and its groundplan was found to be identical with that of H. Mesora (fig. 7:3; cf. 7:4).

Fig. 8."Four-room" house, Horvat Ritma.

Two examples of this type of fortress are known in the Central Negev: Kadesh-barnea and H. Uza.

Fig. 9. The oasis at Kadesh-barnea.

Fig. 10. The fortress, Kadesh-barnea.

74 RUDOLPH COHEN BASOR 236

Fig. 11. Kadesh-barnea fortress: a. Middle fortress plan, 8th-7th cent. B.C.; b. Upper fortress plan, 7th-6th cent. a.c.

Fig. 12. Pottery of the 7th-6th cent. B.C. on surface of upper fortress, Kadesh-barnea.

b. H. Uza. The fortress of H. `Uza (Grid Reference 1657 X 0687) is quite similar in plan to the upper fortress at Kadesh-barnea. It is situated above Naha! Qina, ca. 8 km. southeast of Tel Arad, on the edge of the Negev plateau, just before the terrain proceeds on its steep descent to the Dead Sea. The fortress was first described in 1908 by A. Musil, and A. Alt assigned it to Roman times (Musil 1908: 19-20; Alt 1931: 81). It was surveyed in 1956 by Y. Aharoni (1958: 33-35). It is rectangular in shape (ca. 53 X 41 m.) and comprises a casemate wall around a central courtyard, with towers projecting from the four corners as well as from the middle of each of the four sides. The walls, ca. 0.75-1.50 m. in width, are constructed of rough-hewn limestone and preserved to a height of ca. 1-2 m. The fortress is bisected by another line of casemate rooms, thus creating two coterminous courtyards. The pottery sherds found in the debris belong to the 8th-6th centuries B.C.

From the above survey of Kadesh-barnea and H. Uza it can be inferred that the towered type of fortress is substantially later in origin than the first three fortress types, all of which belong to the 10th century B.C.

The Pottery

In all the Negev fortresses surveyed or excavated so far, the findings consist basically of two main pottery types: wheel-made pottery, characteristic of Iron Age Sites throughout Israel, and primitive hand-made pottery, called "Negev" ware, characteristic of Iron Age sites in certain parts of the Negev.

1. The wheel-made pottery (cf. fig. 14:1-12). The pottery unearthed in the first three types of fortresses shows many distinct similarities. For example, the wheel-made cooking pots, especially those from 'Atar Haroca, Rahba, and H. Haluqim, clearly derive from the same overall assemblage. They are shallow, carinated forms with a round base and belong to a type customary in southern Israel and the Negev in the 1 1 th- 10th centuries B.C. (Aharoni et al. 1960: fig. 11:12, 14 [=fig. 14: 7 here; Ramat Matred]; Kochavi 1969: fig. 5: 6, 7 [Tel Esdar II]; Meshel 1977: 7 [= fig. 14:8 here], 9, 12 [H. Ritma III]). A wide carinated "krater" bowl, possessing a flat base, with parallels at Tell Beit-Mirsim, was also found in the fortresses, including complete examples from 'Atar Haroca and H. Haluqim (Cohen 1976c: fig. 9:8 [=fig. 14:12 here; H. Haluqim], 11-17; Meshel 1977: fig. 6:10 [=fig. 14:11 here; H Ritma]). The storage jars are of an ovoid type very frequent in sourthern Israel, and to which Megiddo V furnishes northern parallels (Cohen 1970: fig. 10:4; ['Mar Haroca]; Kochavi 1969: fig. 5:8 [Tell Esdar II]; cf. Aharoni et al. 1960: fig. 11:1-3 [Ramat Matred]). Another type of storage jar found commonly in the fortresses is the pithos, of which 'Atar Haroca provides the most complete instances (Cohen 1970: fig. 10: 1 [=fig. 14:1 here], 6-8). It possesses an everted rim and carinated shoulder and contrasts strikingly with the vessels of the the collared-rim type. Numerous fragments from pithoi of this kind were discovered at H. Ritma (fig. 14:2 here), Mesad Hatira, and H. Rahba. The pottery from the fortresses also includes other bowl types (cf. fig. 14:3-5), jugs, and juglets; and these, too, have parallels in 10th-century B.C. vessels from other parts of the country.

Fig. 13. Kadesh-barnea ostracon, 7th-6th cent. B.C.

The pottery found in the fourth, towered fortress type includes vessels typical of the final century of the Iron Age, most closely paralleled by the pottery of Lachish II and `En-Gedi V (7th century B.C.). 2. Primitive hand-made pottery (cf. fig. 14:13-17). At both Bir Birein and Tell el-`Ain el-Qudeirat (Kadesh-barnea), Woolley and Lawrence observed the presence of "rough hand-made wares" and were thus evidently the first to have identified the "Negev" pottery (Woolley and Lawrence 1913-15: 67). It was "rediscovered" by Glueck in his excavations at Tell el-Kheleifeh and subjected to a more formal study. Originally he conjectured that such vessels had been used as crucibles in the regional metal industry (Glueck 1938: 14; 1940: 17-18). However, as similar hand-made vessels subse-quently began to appear at other Iron Age sites, especially in the Central Negev and Timna-Eilat area, it became clear that they were ordinary household vessels (Glueck 1959a: 12).

Aharoni proposed that such pottery "was wrought by nomadic potters from the southern deserts, who were able to carry with them only the simplest of utensils. They made these pots with their hands from local material on straw matting, and baked them in temporary kilns dug into the ground.

Fig. 14. Typical 10th-cent. B.c. pottery of the early Israelite forts of the Central Negev: 'A tar Haro` a (nos. I, 6, 13, 14, 16, 17; after Cohen 1970);1!. Ritma (nos. 2, 5, 8, 9, 11; after Meshel 1977); Ramat Mat red (nos. 7, 15; after Aharoni et at 1960); Tel Esdar (nos. 3,4, 10; after Kochavi 1969); N. Ifaluqim (no. 12; after Cohen 1976c). Scale = 1:5.

They wandered from settlement to settlement, and supplied the inhabitants with their simple wares, rather than bringing them from the nearest Judean towns" (1962: 66). Glueck came to subscribe to the same idea and after attributed such pottery to the nomadic and semi-nomadic dwellers of the Negev, the Kenites, Rechabites, Calebites, and Yerah-meelites (1965: 76).

A Summary of the Findings

The archaeological findings reveal, first of all, that a network of fortresses, including the first three types, existed in the 10th century B.C., and that most of the sites, after a brief phase of occupation, were permanently abandoned. Second, at Kadesh-barnea in the 8th-7th centuries B.C. a solid-walled fortress was erected over its predecessor's remains. Finally, towards the end of the Iron Age the towered type of fortress was introduced in the Central Negev at two strategic sites at least, including Kadesh-barnea. These different phases naturally reflect different historical settings and accordingly will have to be treated separately.

The fundamental problem of the early fortress network is its date. Since there are no contemporary allusions in biblical or extra-biblical sources to its existence (with one possible and important exception, to be discussed below), it is crucial that its date be fixed as accurately as possible, for this will provide the overall historical background against which its function can be inferred. And, in the absence of documentary finds, the only chronological measure presently at the disposal of scholars is the wheel-made pottery unearthed in the excavations. As it happens, however, the archaeologists concerned have disagreed among themselves in assigning the ceramic complexes to the 11th-10th centuries B.C. or earlier still, and this has consequently led to differing opinions regarding the attribution and role of the fortress network. As noted above, Glueck and Aharoni believed that the fortresses had been erected by kings of Israel between the 10th and the 7th centuries B.C. for the purpose of defending the Negev road system (cf. Aharoni 1967: 11-13). Rothenberg, however, insists on an exceptionally early date for these wares, in the 13th century B.C. or before, and relates them to the inhabitants of the Negev in the period before the Exodus—the Amalekites—and attributes their destruction to David. He bases his earlier dating on the results of his excavations at the Timna sanctuary (1972: 153-54). Aharoni, incidentally, subsequently altered his position and later contended that the fortresses could be ascribed to the 11th century B.C. and that they had been established by Saul in the course of his wars with the Amalekites (1976: 55-76).

The 10th-century fortress network appears to have been destroyed in the course of Pharaoh Shishak's campaign into Palestine, several years after Solomon's death, following which the border of Judah retreated to its former line along the Beersheva Basin (Amiran 1953: 66). It is probable, therefore, as Mazar has proposed, that some of the fortresses may be included in the list of conquered sites and cities appearing on the victory stele that Shishak (Sheshonk I) erected in Karnak. Among the various place-names relating to the Negev is a group of seven composed with the base plmr or phgr, and Mazar has pointed out an "evident correlation between the hagarim in the list of Shishak and the haserim, the net of fortified settlements which is mentioned in biblical sources" (1957: 57-66). It is reasonable to assume that the destruction of these fortresses along the principal Negev routes was one of the objects of the pharaoh's campaign.

A number of problems still remain concerning the fortress network. It is possible, for instance, that the employment of three distinct fortress types reflects some yet unrecognized difference in their specific function. It would, furthermore, be interesting to know who exactly occupied these fortresses—whether mainly royal troops from the north, or possibly also the local semi-nomadic inhabitants of the Negev, as might be indicated by the common use of hand-made ware.

The Singular Case of Kadesh-Barnea

BIBLIOGRAPHY

Aharoni, Y. et al. Albright, W. F.

1958 The Ancient Desert Agriculture of the Negev. Israel

Exploration Journal 8: 231-68. 1932 The Excavations of Tell Beit Mirsim I: The Pottery

of the First Three Campaigns. Annual of the

1960 An Israelite Agricultural Settlement at Ramat American Schools of Oriental Research 12. New

Matred. Israel Exploration Journal 10: 23-26,

97-11 l. Alt, A. Haven, CT: American Schools of Oriental Research.

Aharoni, Y. 1931 Limes Palaestinae. Palastinajahrbuch 27: 75-84.

1958 The Negeb of Judah. Israel Exploration Journal8: Cohen, R.

26-38. 1966a Sede Boger Area. Hadashot Arkheologiyot 17:

1962 The Iron Age Pottery of the Timna` and 'Amran 30-31 (Hebrew).

Area. Palestine Exploration Quarterly 94: 66-67. 1966b Sede Boger Area. Hadashot Arkheologiyot 18-19:

1967 Forerunners of the Limes: Iron Age Fortresses in 23-25 (Hebrew).

the Negev. Israel Exploration Journal 17: 1-17. 1970 'Atar Haro`a. `Atigot (Hebrew Series) 6: 6-24.

1976 Nothing Early and Nothing Late. Biblical Arche-

ologist 39: 55-76. 1971 Katef Shivta. Hadashot Arkheologiyot 39: 29-30

(Hebrew).

1972 Surveys in the Negev. Hadashot Arkheologiyot 41-42: 39-40 (Hebrew).

I975a Notes and News: H. Rahba. Israel Exploration Journal 25: 171-72.

1975b H. Ramat Boger. Hadashot Arkheologiyot 57-58: 40 (Hebrew).

I976a Notes and News: Kadesh-barnea. Israel Explora-tion Journal 26: 201-202.

I976b H. Ketef Shivta. Hadashot Arkheologiyot 49-50: 53-54 (Hebrew).

1976c Excavations at Horvat Halugim. `Atigot (English Series) II: 34-50.

I977a Notes and News: H. Mesora. Israel Exploration Journal 27: 170-71.

1977b Notes and News: `Ein Qedeis. Israel Exploration Journal 27: 171.

Cohen, R., and Meshel, Z.

1970 Mesudat Refed and Mesudat Hatira. Hadashot Arkheologiyot 37: 27 (Hebrew).

1971 Kh. Ritma. Hadashot Arkheologiyot 39: 28-29 (Hebrew).

1972 Mesad Har Boger. Hadashot Arkheologiyot 40: 37 (Hebrew).

Dothan, M.

1965 The Fortress at Kadesh-Barnea. Israel Exploration Journal 15: 134-151.

Evenari, M.; Shanan, L.; and Tadmor, N.

1971 The Negev. Cambridge: Harvard University. Feldmann, J.

1959 Mibifnim 21: 102-104 (Hebrew).

Fensham, F. G.

1964 Did a Treaty Between the Israelites and the Kenites Exist? Bulletin of the American Schools of Oriental Research 175: 51-54.

Glueck, N.

1934-35 Explorations in Eastern Palestine, II. Annual of the American Schools of Oriental Research 15. New Haven: American Schools of Oriental Research.

1938 The First Season of Excavations at Tell el-Kheleifeh (Ezion-geber). Bulletin of the American Schools of Oriental Research 71: 3-17.

1940 The Third Season of Excavations at Tell el-Kheleifeh. Bulletin of the American Schools of Oriental Research 79: 2-18.

1956 The Fourth Season of Exploration in the Negeb. Bulletin of the American Schools of Oriental Research 142: 17-35.

1957 The Fifth Season of Exploration in the Negeb. Bulletin of the American Schools of Oriental Research 145: 11-25.

I959a An Aerial Reconnaissance of the Negev. Bulletin of the American Schools of Oriental Research 155: 2-13.

I959b The Negev. The Biblical Archaeologist 22: 82-97.

1960 Rivers in the Desert. New York: Farrar, Straus and Giroux.

1961 The Archaeological History of the Negev. Hebrew Union College Annual 32: 11-18.

1965 Ezion-Geber. The Biblical Archaeologist 28: 70-87.

Kirk, G. E.

1938 Archaeological Exploration in the Southern Desert.

Palestine Exploration Quarterly 70: 211-235. Kochavi, M.

1969 Excavations at Tel Esdar. 'Atiqot 5: 14-98 (Hebrew).

Lamon, R. S., and Shipton, G. M.

1939 Megiddo I. Oriental Institute Publications 42. Chicago: University of Chicago.

Malamat, A.

1968 The Last Kings of Judah and the Fall of Jerusalem.

Israel Exploration Journal 18: 137-56.

1975 The Twilight of Judah in the Egyptian-Babylonian Maelstrom. Supplement to Vetus Testamentum 28: 123-145.

Mazar, B.

1957 The Campaign of Pharaoh Shishak to Palestine.

Vetus Testamentum, Supplement 4: 57-66.

Mazar, B.; Dothan, T.; and Dunayevsky, E.

1966 `Ein-Gedi: Archaeological Excavations 1961-1962. `Atigot (English Series) 5.

Meyers, C.

1976 Kadesh Barnea: Judah's Last Outpost. Biblical Archeologist 39: 148-51.

Meshel, Z.

1973 Survey in the Negev. Hadashot Arkheologiyot 45: 39-40 (Hebrew).

1974 Notes and News: Yotvata. Israel Exploration Journal 24: 273-74.

1975 Har Ha-Negev Survey. Hadashot Arkheologiyot 54-55: 30 (Hebrew).

1977 Horvat Ritma-An Iron Age Fortress in the Negev Highlands. Tel-Aviv 4: 110-35.

Musil, A.

1908 Edom, II. Vienna: Kaiserliche Akademie der Wissenschaften.

Rahmani, L. Y.

1962 A Partial Survey of the Adulam Area. Yediot 28: 209-31 (Hebrew).

Rothenberg, B.

1967 Negeb, Archaeology in the Negeb and the Arabah. Ramat-Gan (Hebrew).

1972 Timna. London: Thames and Hudson.

Shiloh, Y.

1970 The Four-room House-Its Situation and Function in the Israelite City. Israel Exploration Journal 20: 180-90.

1978 Elements in the Development of Town Planning in the Israelite City. Israel Exploration Journal 28: 36-51.

Ussishkin, D.

I966a Building IV in Hamath and the Temples of Solomon and Tell Tayanat. Israel Exploration Jornal 16: 104-10.

I966b King Solomon's Palace and Building 1723 in Megiddo. Israel Exploration Journal 16: 174-86.

Woolley, C. L., and Lawrence, T. E.

1914-15 The Wilderness of Zin. Palestine Exploration Fund Annual 3. London: Palestine Exploration Fund. Yadin, Y.

1958 Solomon's City Wall and Gate at Gezer. Israel Exploration Journal 8: 80-86.

By Steve Rudd: Contact the author for comments, input or corrections.

Go To Start: WWW.BIBLE.CA