Enclosed Settlements in the Negeb and the Wilderness of Beer-sheba
Ze'ev Herzog, 1983 AD
Bulletin of the American Schools of Oriental Research, No. 250. (Spring, 1983), pp. 41-49.
Tel Aviv University, Tel Aviv, Israel
(Enclosed Settlements in the Negeb and the Wilderness of Beer-sheba, Ze'ev Herzog, 1983 AD)
The regional approach to archaeological research is gradually assuming a predominant role in Israel. In the Negev this trend may be traced back to the early 1960s when Y. Aharoni began his regional research in the Arad-Beersheva Valley, a project that has continued even after his untimely death in 1976. This project, preceded by a survey of the Negeb of Judah (Y. Aharoni 1958), initiated the largescale excavations at Arad (Y. Aharoni 1975; M. Aharoni 1981), Tel Malhata (Kochavi 1977), Tel Masos (Kempinski et al. 1981), Beersheba and Tel 'Ira (Beit Arieh 1981; cf. Y. Aharoni 1973; Herzog, forthcoming). An overall picture of settlement in the Beersheva Valley was presented by Aharoni in 1976.
Somewhat later, Cohen and Meshel began their systematic exploration of the central Negev high-lands, concentrating mainly on what they referred to as Iron Age or Israelite fortresses.
While these two projects conform to the basic requirements of a regional study, i.e., a well defined geographical zone, adequate coverage of the area, and comparative analysis of the sites, they exhibit one of the major drawbacks of this approach—a tendency to overemphasize the regional data at the expense of interregional synthesis. This limitation is clearly seen when we consider the geographical horizons of the areas under discussion: in Cohen's comprehensive study, Tel Beersheba and Tel Arad are beyond the limits of the accompanying map (Cohen 1980: fig. 1); and in Aharoni's summary (1976) the central Negev sites are hardly mentioned.
The objectives of the present study are to provide a comparison of Early Iron Age data from both regions and to offer a synthesis of the material.
The modern term "Negev" designates a much broader area than that represented by the biblical Negeb. We may distinguish three distinct geo-graphical zones within the Negev of today and suggest their possible biblical equivalents. The Negeb (or "Negeb of Judah") was apparently confined to the Beersheva Valley, drained by Wadi el-Milh/ Wadi el-Mashash, and the rolling plain encompassed by Wadi es-Saba`/ Wadi el-Ghazzeh (Aharoni 1958). The "Wilderness of Beersheba" was most likely the region just south of Beersheba and Natal Besor, while the modern term "Negev highlands," to the south of the Wilderness of Beersheba, may have been subsumed under the appellation "Wilderness of Zin" (Rainey, forth-coming; see fig. 1 here).
The Negev "Fortresses"
More than 40 sites in the Wilderness of Beer-sheba and the Wilderness of Zin have been surveyed, and 12 of them have been partially excavated to date. They have been dealt with mainly by Cohen (1980) and Meshel (1979), who employed different classification systems for these "fortresses" and arrived at different conclusions regarding their dating. Only on one group do they agree: the rectangular fortresses with towers found at Kadesh-barnea and Horvat `Uza, which are dated by their distinctive features and pottery assemblages to the 8th-7th centuries B.c.1
Fig. 1. Sites in the Negeb of Judah, the Wilderness of Beersheba, and the Wilderness of Zin.
The other "fortresses" are classified by Cohen according to shape and by Meshel according to size and topographical conformity. Cohen's categories are: (1) roughly oval, (2) rectangular, and (3) square(Cohen 1980: 63; figs. 3, 5, 7, respectively). This typology, which ignores the factor of size, lumps radically different types of structures into a single category. For example, the fortress at Horvat Rahba, which is some 75 m. in length, is included in the same group as Horvat Haluqim, which is only 23 m. in diameter. The differences are even more outstanding when the areas of the two sites are compared: 3100 m2 for Horvat Rahba and 380 m2 for Horvat Haluqim. Cohen's second type is no less problematic, since Mesad Hatira and Horvat Ramat Boger (1980: fig. 5:2, 4) can hardly be considered rectangular. Nor is size taken into account for this group either.
Meshel, aware of this incongruity, classified these fortresses into two main categories—large and small—and further subdivided the structures of each size according to whether they were adapted to the contours of the site or whether they gave the appearance of having been preplanned according to a standard model (Meshel 1979: 17).
The pottery from these structures consists of two types: handmade Negev ware (sometimes called "Negebite ware") and wheel-made vessels of the types common in Judah at the time. Since recent discoveries have shown that the Negebite ware had an extremely wide chronological range—from the 13th century B.C. at Timna (Rothenberg 1972: 180-82) to the end of the Iron Age at Kadesh-barnea (Cohen 1980: 77)—the presence of this ware cannot, of course, support any particular date within the Iron Age for these sites. Cohen's suggestion to relate this ware to the Kenites (1980: 77) is apparently contradicted by the evidence from the only site so far attributed to this tribe, namely Stratum XII at Arad (Mazar 1965: 303), since no Negebite ware was uncovered in this stratum (M. Aharoni 1981).
The wheel-turned vessels found in the Negev fortresses are dated by Cohen exclusively to the 10th century B.C., and he thus attributes these sites to the reign of Solomon (Cohen 1980: 77-78). According to Meshel, however, this pottery cannot be dated any more precisely than the 11 th- 10th centuries B.C. On historical grounds, he prefers to associate these structures with "one of the kings who defeated the Edomites and Amalekites," Saul or David being the most likely candidates in his opinion (Meshel and Cohen 1980: 80).
Regarding the function of these fortresses, Meshel and Cohen do not see eye to eye either. While both agree that the network of Negev strong-holds must have been conceived under a central authority, Cohen believes that they were intended to create a strong defensive line along the southern boundary of the monarchy in addition to their role in guarding the roads crossing the central Negev (Cohen 1980: 77). Meshel (1977: 133), on the other hand, feels that they represent the initial conquest of the desert and a "show of force" toward the local inhabitants on the part of the central government.
Finally, Cohen (1980: 78) attributes the disappearance of the "fortresses" to the havoc wrought during Pharaoh Shishak's campaign, while Meshel considers the burnt patches found occasionally in these structures to be insufficient evidence for a military destruction and concludes that the sites were abandoned when the "local inhabitants" recognized the permanency of the central authority and hence determined that there was no longer any need for fortresses (Meshel 1977: 133).
The above survey of opinions may demonstrate how inconclusive is the present state of research regarding the typology, pottery date, historical attribution, and function of the Negev "fortresses." But even these views do not represent all the ideas on the subject. About fifteen years ago Rothenberg (1967: 92-97) suggested that the central Negev sites were settlements of the local Negev tribes, mainly Amalekites, who settled them as early as the 13th century B.C. However, while we accept Kochavi's (1979) recent identification of Tel Masos with "a city of Amalek" (1 Sam 15:5, RSV = "the city of Amalek"), Rothenberg's dating is incompatible with the Iron Age wheel-made pottery found in these sites (most of which, we should point out, came to light after the publication of Rothenberg's study).
Two of Rothenberg's basic arguments—namely that the Negev fortresses were actually agricultural settlements and that they were founded not as royal enterprises but by the semi-nomadic bearers of the Negev ware—were adopted by Etam in a short response to Cohen's Hebrew version of his summarizing article on the Negev fortresses (Cohen 1979; Etam 1980). But, unlike Rothenberg, Etam dates this settlement to the 11 th century B.C. and attributes the entire network of sites in the Negev highlands to an autochthonous population (he clearly avoids naming this population in more specific ethnic terms) that used this Negev ware. The termination of habitation was, according to Etam, caused by Saul, who destroyed the city of Amalek. Before considering this theory, I must point out a fundamental contradiction in Etam's hypothesis: he tries to prove that the settlements were not of defensive character, while at the same time he denies their attribution to Saul on the grounds that it is difficult to imagine that this king built defensive systems to fortify the southern border of the kingdom while it was still in its infancy.
At this point, having presented the wide range of views on the subject, I should like to stress what in my opinion is the basic drawback of all of these studies: the assumption that the Negev "fortresses" were homogeneous in nature and that a single historical and functional interpretation can be attributed to all of them. Rather, they should be viewed as the result of a developing process. This conclusion was reached after an intensive study of the finds from the Early Iron Age strata at Tel Beersheba (Herzog, forthcoming), material that was not available to the authors of the earlier studies.
Early Iron Age Strata at Tel Beersheba
Four strata (IX-VI), dating from the mid-12th to the early 10th century B.C., were unearthed at Beersheba, mainly on the southeastern slope of the mound (Herzog 1980). The architectural, chrono-logical, and historical sequence of these strata may be summarized as follows:
Stratum IX. This was a pit settlement accom-modating a semisedentary population of about 20 families who used the pits partly for dwelling and partly (the deeper ones) for storage. The pot-tery repertoire ranges from ca. 1150-1050 B.c. This long occupation is also attested by considerable accumulations of occupational debris in the pits. Historically, this might be the locale of the patri-archal stories of Beersheba (Genesis 21, 26). To this phase may also belong the astonishingly deep well that was hewn into the same slope, next to one of the dwelling pits.
Stratum VIII. This stratum is characterized by the reuse of many of the pits, but is also witness to the first stone-built dwelling, apparently of thefour-room house type, constructed near the well. There is an increase in the number of hand-burnished vessels, which dates this stratum to the second half of the 11th century B.C. These remains may represent the village in which Samuel's sons were appointed to serve as judges (1 Sam 8:1).
Stratum VII. An entirely new occupational pattern emerges in this stratum. A belt of houses encircling a large open courtyard was erected on the slope above the well. The individual houses, all of the four-room type, were built according to a preplanned scheme to create a defensive perimeter, resulting in what I term an "enclosed settlement." The finds in this stratum, as well as its relative position in the stratigraphical sequence, date it to the late 11th century B.C., thereby suggesting that it was founded by the royal mandate of King Saul.
Stratum VI. Surprisingly, the previous enclosed settlement appears to have been partially dis-mantled. Parts of the former houses were reused, mainly by dividing their long rooms into small compartments. One nicely built new house appears. The pottery dates to the early 10th century B.C., and the remains are interpreted as a workers' camp erected during the reign of David to accommodate the foremen, chief inspector, and the like, who were responsible for the construction of the first fortified city to be built on top of the hill (Stratum V).
The starting point for this analysis is an architectural comparison between the enclosed settlement of Stratum VII at Beersheba and a group of Negev fortresses represented by Cohen's "roughly oval" and "rectangular" categories and Meshel's "large fortresses adapted to the topography." These fortresses are Horvat Rahba (Cohen 1980: 67; fig. 2:3 here), Mesad Refed (Cohen 1980: fig. 2:2 here) and Mesad Hatira (Cohen 1980: fig. 2:1 here).
These sites have the following features in common:
(1) All of them lack an imposed concept of planned fortresses in being neither square nor rectangular nor circular. Their overall shape is amorphous and adapted to the local topography.
(2) They are considerably larger than any of the other Negev sites under discussion. Horvat Rahba covers an area of about 3100 m2 ( = 3/4 acre), Masad Refed about 2200 m2 and Mesad Hatira 2300 m2 (both more than 1/ 2 acre). These areas may be compared to those of the well-planned forts: Horvat Haluqim, a circular fortress (Cohen 1980: fig. 3:3), measures 380 m2, and the square fortresses at Natal Raviv, Horvat Ritma, 'Atar Haroca, and Horvat Mesora (Cohen 1980: fig. 7) cover only about 400 m2 (or 1/ 10 acre).
(3) The building units comprising the large "fortresses" differ from those of the smaller structures. While the small forts are surrounded by two parallel walls forming typical casemates, this clear-cut casemate arrangement is missing at the larger sites. Some of these large sites were excavated on only a limited scale, and their ground plans were extensively reconstructed; in doing the reconstructions, the excavators apparently were strongly influenced by their concept of these structures as casemate fortresses. That the actual situation may be interpreted differently can be demonstrated by examining the plan of the better preserved remains at Mesad Hatira (fig. 2:1). Here it is clear that the remains do not belong to an independent casemate wall. Instead, what we have is a series of dwellings, arranged as a belt around the circumference of the hilltop. Most of these dwellings have a front unit, which probably served as an inner courtyard for domestic activities, and a broadroom at the rear for sleeping. Actually, this is a forerunner of the four-room house, subsequently developed by sub-division of the front courtyard. The fact that no free-standing casemate walls were constructed may be deduced from the disjuncted orientation of the walls of the rear rooms. Careful observation of the northern side of the settlement shows that each dwelling unit was built separately: although the two rear walls are parallel, they do not form a straight line from one house to its neighbor. Moreover, there are no houses in the center of the northern side of the enclosure, the outer boundary of the site here being closed not by a double casemate wall but by a single wall. The inner wall is also missing in two houses on the northwestern side. Consequently, we may conclude that the architectural principle of this plan is not that of casemates adjoined by additional rooms (Cohen 1980: 67) but rather a circumvallation of houses, the broadrooms at the back of each creating the illusion of casemate walls. Since this is the same architectural layout that is found in the enclosed settlement of Stratum VII at Beersheba (fig. 2:4), the "fortresses" at Horvat Rahba and Mesad Refed should be similarly reconstructed, particularly in view of the cross-walls that extend beyond the basic two-room arrangement into their centralcourtyards.2 At Mesad Refed the disjunctive nature of the inner line of walls is striking—even in spite of the fragmentary state of the remains—as seen in the small corners at the junctions of the two inner segments (fig. 2:2). It is therefore very likely that at Horvat Rahba and Mesad Refed these are belts of interconnected dwellings and not free-standing case-mate walls. This structural difference between the large enclosed settlements and the small casemate fortresses reflects different functions: the multi-roomed dwelling units of the enclosed settlements were suitable to accommodate whole families, while the single-room casemates fit the requirements of a military garrison.
(4) The entrance into the true casemate fortress was simply a gap between two casemate rooms. In the settlements encircled by a belt of houses, however, the entry was protected by two "towers," which project either inside the settlement (Mesad Hatira) or outside (Beersheba Stratum VII).
The architectural characteristics of the group of sites discussed above correspond perfectly with those of the enclosed settlement of Stratum VII at Beersheba (fig. 2:4). We may therefore conclude that these three sites should be dated to the same period and defined likewise as enclosed settlements.
When the locations of these three enclosed settlements were marked on the map (fig. 1), it became clear that all three appear in a circumscribed area southeast of Beersheba, remote from the large concentrations of the other "central Negev fortresses."
This combined architectural and geographic data is apparently the key to unraveling the Gordian knot between two disparate types of settlements that for several decades were erroneously considered to be a single phenomenon.
The enclosed settlements of Horvat Rahba, Mesad Refed, and Mesad Hatira are all situated in the Wilderness of Beersheba. Since only at Tel Beersheba are the earlier phases of occupation pres-ent, we may assume that this site was the spring-board of occupation in the biblical Negeb, from which a group of similar enclosed settlements spread into the southeast during the 11 th century B.C.
To the same wave of settlement should be attributed the circle of houses at Tel Esdar Stratum III, also dated to the 11 th century B.C. (Kochavi 1969). Here the presence of individual dwellings is particularly clear, and this stratum may represent one of the early stages in the development of this process.
Another settlement might well be attributed to this group, a small site located on Nahal Beersheva about 10 km. west of Tel Beersheba (map ref. 1242 x 0707). This site was surveyed by Gophna, who also conducted a small probe (Gophna 1964: 241-46, Site E).3 The remains of the walls scattered around the perimeter of the partly eroded hill may easily be reconstructed as another enclosed settle-ment (fig. 2:5). Three additional small settlements were surveyed by Gophna on the tributaries of Nahal Pattish northwest of Beersheva (Tel Manopat, Beer-Mano'ati, and Beer-Halmut; Gophna 1964: 242; for location see fig. 1 here).
It is tempting to relate to this 11th century phase of settlement the recently found site at el-Quseima (Meshel 1981: 361-62; fig. 4) in the vicinity of Kadesh-barnea, which has all the appearances of an enclosed settlement. However, the oval fortress of Kadesh-barnea (i.e., "earliest fortress" of Cohen 1981: 107) and 'Ain Qudeis (Cohen 1980: fig. 3:1), probably belong to the 10th century. Hence, this region might have been another nucleus of 11 th century occupation—as might be expected from the prominent role of Kadesh-barnea in the formative stage of Israelite history.
All the rest of the "Negev fortresses" are characterized by two basic features: small square forts and small open villages, usually in close proximity to each other. They are found scattered over the Negev highlands (i.e., the Wilderness of Zin) between the two nuclei of enclosed settlements described above. Their establishment was undoubtedly connected with later geopolitical developments in the region involving royal initiative.
This functional-architectural study of the "Negev fortresses" enables us to distinguish between two types of sites: enclosed settlements and casemate forts, the stratigraphy of Beersheba providing dating evidence for the first type. This is a rare example of the use of architectural analysis in solving a chronological problem for which the ceramic data were inconclusive.
Once the 11 th century phase of occupation of enclosed settlements in the Negeb and the Wilderness of Beersheba is recognized, a history of the settlement process in the southern part of Judah may be outlined.
The earliest settlement in the late 13th century B.C. at Tel Masos (Stratum MB) is characterized by storage pits and huts. This settlement could have been occupied by an early Israelite population, driven out and replaced by the Amalekites in the early 12th century (Strata IIIA, II). This conflict is apparently reflected in the biblical tradition of the defeat of the Israelites by the Amalekites (Num 14:44-45; cf. 1 Sam 15:2).
In the 12th century B.C. the pit dwellings at Beersheba (Stratum IX) served the semi-sedentary Israelite patriarchal families, while at Arad Stratum XII a small Kenite village was built around an open courtyard that may have had a cultic function according to Judg 1: 16-17 (Mazar 1965). In the 12th century these three settlements represent the three ethnic entities inhabiting the biblical Negeb: Israelites, Amalekites, and Kenites, living peace-fully side by side.
It was only during the 11 th century that this harmony was disturbed when the Israelite settlers began to penetrate into the region in larger numbers. This new wave of settlement was apparently the result of Philistine pressure in the Shephela of Judah that drove the ever-increasing Israelite population into the more arid zones in the south. The new inhabitants spread into the area west and south of Tel Masos: on the northwest they built Site E on Nahal Beersheba and the sites on Nahal Pattish; on the west Stratum VII at Beer-sheba; and on the south Tel Esdar. Further south they expanded into the Wilderness of Beer-sheba at Horvat Rahba, Mesad Refed, and Mesad Hatira.
This wave of settlement must have led to a conflict of interests between the Israelites and the population of the "city of Amalek" (Tel Masos), eventually resulting in the campaign of Saul against the Amalekites (1 Sam 15:5, 7), which was apparently responsible for the destruction of Tel Masos Stratum II, while the destruction of Tel Esdar could have been the result of a counterattack by the Amalekites. All the rest of the sites continued to exist into the 10th century, when a new phase began in the history of the region with the ascension of David to the throne.
At first an administrative center was erected at Beersheba (Stratum V), probably under the initiative of this monarch [David] (Aharoni 1974), followed by construction of the large military fortresses at Arad Stratum XI, Kadesh-barnea ("earliest fortress," Cohen 1981: 107), and Tell el-Kheleifeh Period I (Glueck 1965: 71-82), all of which we believe are dated to Solomon's reign.
Fig. 3. Isometric reconstruction of an enclosed settlement based on plan of Beersheba Stratum VII.
Undoubtedly, the small forts and accompanying settlements were founded in connection with Solomon's trade operations in the Red Sea. In my opinion, the forts were intended to defend the caravans plying the routes of the Way of the Spies, the Way of the Hill Country of the Amorites, and the Way of the Red Sea (Meshel 1974: fig. 17; 1981: fig. 1). These royal enterprises enabled a network of open settlements to be interwoven with the forts, as an expansion of the earlier clusters of enclosed settlements in the north.4 Presumably there was mutual cooperation between the forts and the civilian settlements, the former providing protection and the latter logistical support in the form of agricultural products and perhaps even manpower. It was apparently Shishak's invasion that brought this network of settlements to an end. Some time afterwards, probably following an occupational gap in the 9th century B.C., attempts were made to rebuild this system, but that subject is beyond the scope of the present study.
'A forerunner of this type from the 10th century B.C. may be seen in the fortress of Stratum XI at Arad, which had towers at each of its four corners and two more along each of its sides (M. Aharoni 1981).
'This analysis may throw some light on the debate regarding the nature of the eastern casemates in Megiddo Stratum VA. The presence of cross-walls that continue beyond the inner casemate rooms there may point to an architectural concept identical to that of the Negeb settlements, thereby supporting Aharoni's interpretation (1972) contra Yadin (1970).
'This project is the third example of regional research in the northern Negev. Like the other two, however, its results were never integrated with the rest of the Negev material.
4It is suggested here that the Israelites settled Tel Masos Stratum I on top of the ruins of the Amalekite city (Stratum II). The square fort and open villages are another example of the type of settlement prevailing in the second half of the 10th century B.C. The drastic change in the layout and nature of the site in Stratum I can best be explained by assuming that this stratum was founded by an entirely different ethnic group.
1981 The Pottery of Strata 12-11 of the Iron Age Citadel at Arad. Eretz-Israel (Aharoni volume): 181-204 (Hebrew).
1958 The Negeb of 'Judah. Israel Exploration Journal 8: 26-38.
1972 The Stratification of Israelite Megiddo. Journal of Near Eastern Studies 31: 302-11.
1974 The Building Activities of David and Solomon. Israel Exploration Journal 24: 13-16.
1975 Arad Inscriptions. Jerusalem: Bailik Institute and Israel Exploration Society.
1976 Nothing Early and Nothing Late: Re-Writing Israel's Conquest. Biblical Archaeologist 39:55-78.
Aharoni, Y, ed.
1973 Beer-sheba I: Excavations at Tel Beer-Sheba, 1969-
1971 Season. Tel Aviv: Tel Aviv University.
Beit Arieh, I.
1981 Notes and News: Tel Ira, 1980. Israel Exploration Journal 31: 243 -45.
1979 The Israelite Fortresses in the Negev Highlands. Qadmoniot 12: 38-50 (Hebrew).
1980 The Iron Age Fortresses in the Central Negev. Bulletin of the American Schools of Oriental Research 236: 61-78.
1980 The "Fortresses" of the Negev Uplands—Settlement Sites. Qadmoniot 13:56-57 (Hebrew).
1965 Ezion-geber. The Biblical Archaeologist 28: 70-83. Gophna, R.
1964 Sites from the Iron Age between Beer-sheba and
Tell el-Fat-ca. Yediot 28: 236-46 (Hebrew).
1976 The City-Gate in Eretz-Israel and the Neighboring Countries. Unpublished doctoral dissertation, Tel Aviv University (Hebrew).
1980 Beer-sheba of the Patriarchs. Biblical Archaeology Review 6: 12-28.
forth- Summary and Conclusions, in Herzog, Z., et al. coming Beersheba II Early Iron Age Settlements. Tel Aviv. Kempinski, A., et al.
1981 Excavations at Tel-Masos: 1972,1974,1975. Eretz-Israel 15 (Aharoni volume): 154-80 (Hebrew). Kochavi, M.
1969 Excavations at Tel Esdar. `Atigot 5: 14-48 (Hebrew).
1977 Malhata. Tel. Pp. 771-75 in M. Avi-Yonah and E. Stern, eds., Encyclopedia of Archaeological Excavations in the Holy Land, vol. 3.
1979 The Beginning of Israelite Material Culture. Paper presented at the Sixth Archaeological Conference in Israel, Tel Aviv.
1965 The Sanctuary of Arad and the Family of Hobab the Kenite. Journal of Near Eastern Studies. 24: 297-303.
1974 History of the Negev in the Time of the Kings of Judah. Unpublished doctoral dissertation, Tel Aviv University (Hebrew).
1977 Horvat Ritma: An Iron Age Fortress in the Negev Highlands. Tel Aviv 4: 110-35.
1979 Who Built the 'Israelite Fortresses' in the Negev
Highlands. Cathedra 11: 4-28 (Hebrew).
1981 The History of "Darb el-Ghaza"—the Ancient Road to Eilat and Southern Sinai. Eretz-Israel, 15 (Aharoni volume): 358-71 (Hebrew).
Meshel, Z; and Cohen, R.
1980 Refed and Hatira: Two Iron Age Fortresses in the Northern Negev. Tel Aviv 7: 70-81.
Rainey, A. F.
forth- Early Historical Geography of the Negeb. In coming Herzog, Z., et al., Beer-sheba II: Early Iron Age
Settlements. Tel Aviv: Tel Aviv University.
1967 Negeb, Archaeology in the Negeb and the Aravah. Ramat Gan (Hebrew).
1972 Timna. London: Thames and Hudson.
1970 Megiddo of the Kings of Israel. The Biblical Archaeologist 33: 66-96.
By Steve Rudd: Contact the author for comments, input or corrections.
Go To Start: WWW.BIBLE.CA