The Fortresses King Solomon Built to Protect His Southern Border

056

An enormous number of Iron Age fortresses have been uncovered in the Central Negev, especially in recent years. The question is, what are they doing here? The answer depends in large part on who built them. And to determine who built them, we need to know as precisely as possible when they were built.

The remains of a few of these fortresses were found even before World War I. In their famous archaeological survey for the London-based Palestine Exploration Fund, Sir Charles Leonard Woolley and T. E. Lawrence (“Lawrence of Arabia”) discovered Qast er-Ruheileh, Bir Birein and Tell Ain el Qudeirat in what they identified as the Biblical “Wilderness of Zin.”¹

Nothing else happened until the 1950s and early 1960s, when some additional fortresses were identified in surveys of the area conducted by the American rabbi-archaeologist Nelson Glueck and, later, by one of Israel’s leading archaeologists, Yohanan Aharoni.²

In 1961, Beno Rothenberg also surveyed the area of Nahal Horsha and Nahal Kadesh Barnea.³

In 1965, I began my own work in this area. In the second half of the 1960s, I led a survey on behalf of the Archaeological Survey of Israel, and in the 1970s, another on behalf of Israel’s Department of Antiquities. In 1975, Ze’ev Meshel conducted a survey in the area of Nahal Sirpad, with the help of the staff of the field school at Sde Boqer. As a result of all these archaeological surveys, more than 40 Iron Age fortresses have now been identified in the Central Negev.

A new phase in our understanding of these fortresses was opened up, however, by a number of excavations, 058rather than by surveys. I excavated my first fortress, Atar Haro‘a, between 1965 and 1967. Then, between 1969 and 1972 Ze’ev Meshel and I excavated four additional fortresses.^a After the 1978 Camp David peace accords, which led to the transfer of important military airfields from Sinai to the Central Negev, I directed a number of emergency rescue excavations of these fortresses on behalf of the Department of Antiquities, the most recent in 1983. In all, 20 of these fortresses have now been at least partially excavated.⁴

As a result of all this work, we have identified four different styles of fortresses—perhaps, more accurately, I should say we have identified four different types of architectural plans. Some of the fortresses are (1) roughly oval in plan; others are (2) rectangular but with unequal sides; still others are (3) square; finally, two of the fortresses are (4) rectangular but with outcropping towers at the corners and sides.

Let us put aside the fourth category because the two fortresses^b with outcropping towers at their corners and sides date from the eighth to the sixth centuries B.C., considerably later in the Iron Age than the other three fortress plans; these other three are not only earlier, but contemporaneous. In this article, I will concentrate on the fortresses with the first three kinds of plans—oval, rectangular with unequal sides, and square. All three plans, while they differ in shape, consist of casemate walls enclosing a central courtyard.

The most common plan is the oval. We now know of 11 oval fortresses. Let me describe in more detail one of these, which I excavated in the spring of 1983.

060

We call it Mesudat Nahal Aqrav.⁵ The first word means “fortress” in Hebrew. Aqrav, as I shall call it here, is surrounded by a casemate wall, which is a wall formed by two parallel walls subdivided into rooms by transverse walls. The inner and outer casemate walls of the fortress are two feet wide and were constructed with rough-hewn blocks of local limestone. They are preserved to a height of over six feet. The entire casemate wall is approximately 130 feet in diameter and 406 feet in circumference, and is divided into 24 casemate rooms. The 8,712-square-foot central courtyard was open to the sky. The casemate rooms, however, were roofed and were used as lodgings, armories and storerooms. These rooms are each about eight feet wide, but they vary in length from about 10 to 25 feet. In three of these casemate rooms, we found stone columns in situ. That is why we know the casemate rooms were roofed. The columns once supported the roof, which had long since fallen in. In some of the doorways connecting the rooms with the courtyard, we found the lintels in situ.

In the southern part of the courtyard we found the remains of a large structure, about 30 feet wide and 80 feet long, attached to the inner wall of the casemate. We don’t know exactly what it was used for, but it probably served some administrative function.

The principal entrance to the fortress, on the south, was formed rather simply by a gap of about six feet in the casemate wall between two of the casemate rooms. A second gateway, almost eight feet wide, provided access through the northern side of the fortress, but at some stage it was sealed up with massive stones. West of this northern gateway was a square room, about 13 feet on a side, that had evidently served as a lookout tower.

The floors of the various rooms were beaten earth. When we found them, they were covered with a layer of ashes, indicating that the fortress had met with a violent end. Amid the ashes were the smashed remains of wheel-made pottery (that, as we shall see, enabled us to date the fortress), as well as some handmade pottery.

We also found the ruins of a contemporaneous 061settlement north of the fortress. The settlement included a number of structures. The largest and central structure had two wings of similar shape. Overall it was about 32 feet by 23 feet. Its walls, like those of the fortress, were built of rough-hewn limestone blocks. In the eastern wall were two stone pillars of five cylindrical drums each.

Aqrav is only one of 11 oval fortresses of Iron Age origin that have been identified in the Central Negev.⁶ Although all 11 are oval, some are more nearly circular than others. Their sizes vary considerably. The smallest, Horvat Haluqim, is about 70 feet in diameter, the largest, Horvat Rahba, about 215 feet in diameter. Sometimes the shape follows the topography of the hill on which it is built. These oval fortresses contain between eight and 24 casemate rooms, although the full number has not been determined in all cases. Generally, the gateway is simply a gap in the line of the casemate rooms, although sometimes the passage is narrowed by massive stone piers. Some of the fortresses, like Aqrav, have additional internal structures. But the basic groundplan is the same—casemate rooms around a central courtyard. Each of these oval fortresses, with perhaps one exception, has an associated settlement.

The second type of fortress is rectangular in plan. The walls are usually unequal in length and correspond to the topography of the hill on which they were built. Sometimes the line of casemate rooms simply skirts the edge of the hilltop. Seven of these rectangular fortresses have been identified.⁷

The third type of fortress is square and was evidently built in accordance with a standard plan, almost 65 feet on a side. There are four of these square fortresses.⁸

All three types of fortresses contained similar pottery, which, although not abundant, was homogeneous. These fortresses also shared many other characteristics, including casemate walls built of rough-hewn limestone blocks around a central courtyard and usually a beaten-earth floor with a layer of ashes on top indicating a fiery destruction. There were slight differences in addition to shape. Some had buildings in the courtyard. Some had outbuildings attached to the outside casemate wall. Some had guardrooms attached to the gateway. At some we found cisterns for water storage and at others animal pens; sometimes the courtyard was one big, open area, 063and sometimes it was subdivided by walls. But these variations were minor.

Associated settlements were found in the immediate vicinity of most of the fortresses. Many of the structures in the settlements consisted of so-called “four-room houses.” The four-room house is closely associated with the sedentarization (the change from a semi-nomadic life to a settled life) of the Israelites when they occupied Canaan after the Exodus. A typical four-room house contains three parallel long rooms (one of which may be an open courtyard) with a long, broad room perpendicular to the three long rooms. Any of the rooms may be subdivided without changing its basic structure or plan. The walls separating the long rooms from the courtyard often consist of monolithic stone pillars (three in each row) alternating with dry stone construction. Some four-room houses have five and six pillars in a row, and instead of being monolithic these pillars sometimes consist of stacked drums. Sometimes a building clearly seems to be part of a unified farm complex that includes nearby terraced fields, walls and cisterns. Cisterns were often located at the bottom of a small wadi in order to collect the rainwater. At Horvat Haluqim, a typical settlement, we recorded the remains of 25 structures, although not all were four-room houses.

While it is obvious that the settlements were associated with the fortresses, the settlement’s precise relationship to the fortress remains unclear. Were the inhabitants of the settlements civilian farmers who enjoyed the protection of the nearby military stronghold? Or, on the model of the Roman limitanei, were the inhabitants soldier-cultivators who took refuge in the fortress in time of danger?

Clearly, the three fortress types—oval, rectangular and square—were contemporary with one another, and all belonged to the same overall defensive network in the Central Negev. This is shown by the pottery and, to a lesser extent, by the other similarities, including the similarities in building technique.

The question naturally arises as to why three distinct fortress types were employed. I have tried to establish some correlation between the respective groundplans and function or terrain, but have not reached any satisfactory conclusion. One might suppose that two or three different architects were active in the area, each responsible for different projects, and each utilizing the “blueprint” he preferred or with which he was familiar. Or it may be that the fortresses were not all erected simultaneously, but 064came into being in short stages. Even if no more than a decade or so intervened between different types, one can easily imagine that new defensive groundplans came into vogue, replacing earlier models. However, this is all simply conjecture. But the existence of three basic plans does not contradict my basic thesis that the network in its entirety was notably short-lived, not enduring more than a half-century at most.

The three different shapes may be reflected in different words for fortresses used in the Bible. Professor Benjamin Mazar has suggested a relationship between the oval fortresses and the Biblical expression atarot. This word refers to a fortified city or encampment and appears either alone or as part of a place name. The verbal root of the word connotes the sense of “surrounding” or “closing in,” and occurs in words for “crown” and “wreath.” In a place name, by analogy, it would imply a round enclosure. In Numbers 32:34–35, among the sites established in Transjordan by the tribe of Gad, Ataroth and Atroth-shophan are mentioned. In Joshua 16:5, Atroth-addar is listed as one of the border cities of Ephraim. In 1 Chronicles 2:54, Atroth-beth-joab is cited as one of the eponymous sons of Salma. Even though they are rather general examples, these names containing the element atarot do suggest that a round or oval fortification had an accepted nomenclature in Biblical usage. Perhaps these 065names referred to an oval fortress of the type we have been discussing.

Elsewhere in scripture other words are used for fortresses. These include mesudah (plural, mesudoth) and mesadah (plural, mesadim). Sometimes two of these terms are used together, as in Genesis 25:16, where the Ishmaelites are enumerated “by their haserim and by their tiroth.” These two words are translated variously as towns and castles, villages and encampments, hamlets and fortresses, settlements and strongholds. The reference may well be to a fortress like those we have been discussing (more specifically the oval plan) and its associated settlement.

Later the Israelites fled from the Midianites and built for themselves refuges in caves and mesadim (fortresses, Judges 6:2).

When David fled from Saul, seeking refuge in the wilderness of Judah, he stayed in mesadim (fortresses) (1 Samuel 23:14). This same kind of fortress or stronghold is later referred to in 1 Samuel 24:22 as a mesudah.

The numerous fortresses that have been discussed in the Central Negev are altogether suitable to these Biblical terms and their contexts. On the other hand, these Central Negev fortresses and their associated settlements can be placed in a much narrower time frame. Once this has been done, we will be able to place these Central Negev fortresses in a more precise historical context. One 066of the most significant contributions of our recent work on these fortresses has been to re-date them—although not all scholars agree with our new dating.

The principal method for dating these fortresses is to date the pottery found lying in the ashes on the floor. Although never found in large quantities in any of the fortresses, it is nevertheless homogeneous pottery and clearly datable, in our view, to the tenth century B.C. Some scholars, however, have assigned the pottery to the 11th century, which radically changes the historical context in which the functions of the fortresses must be interpreted.

Two distinct types of pottery were found in the fortresses, hand-made and wheel-made. The particular kind of hand-made pottery found is restricted to the Negev, so scholars call it “Negbite” ware. Negbite ware was first observed by Woolley and Lawrence in their pre-World War I probe of Kadesh-Barnea. They noted fragments of “rough, handmade wares, thin-walled, of gritty clay burnt very hard in an open hearth.”⁹ This pottery was “rediscovered” by Glueck in his excavations at Tell el-Kheleifeh (identified by him with Biblical Ezion-Geber). Here he found “large quantities of crude, handmade, friable, smoke-blackened pots, many of which were built up on a mat, and most of which have various simple types of horn- or ledge-handles, or combinations of both.”¹⁰ Subsequently, similar hand-made pottery was found at numerous Iron Age sites in the Central Negev and Timna-Eilat area. Aharoni related it to the semi-nomadic inhabitants of the desert: “This latter handmade pottery was no doubt made locally by the most primitive methods, i.e., on a mat and with very bad firing. It may be conjectured that these vessels were the work of nomad potters, who, being constantly on the move from settlement to settlement in the Negev and Aravah, could not make use of the more highly developed instruments of their craft, such as a potter’s wheel and a permanent clay oven. These simple and cheap utensils largely satisfied the daily needs of the local population, especially as cooking-pots. At the same time, a certain amount of the usual pottery of the period was imported from further north.”¹¹

Interesting as this Negbite pottery is, it cannot be used at the present time to date the Central Negev fortresses because it has such a wide chronological range. True, at one time, the pottery was assigned to the tenth century B.C., but that was before my excavation at Kadesh-Barnea, where it clearly remained in use until the end of the Iron Age in the sixth century B.C.¹² Moreover, Beno Rothenberg, on the basis of his research in the Timna-Eilat area, argues that this pottery originated hundreds of years before the tenth century B.C.¹³ Although I have 067some reservations on this point, it is nevertheless clear that Negbite ware is of limited usefulness as a chronological indicator. This is unfortunately the case even though within its long time-span Negbite ware appears to have undergone some changes. For example, in the Central Negev fortresses, which I date to the tenth century, virtually the only forms of Negbite ware were cooking pots and bowls. By the eighth–seventh centuries B.C., as seen in the so-called middle fortress at Kadesh-Barnea, we found a wide variety of shapes, many of them quite clearly modeled on contemporaneous wheel-made types.

To securely date the Central Negev fortresses, however, we must turn to the wheel-made pottery that was imported from the north, the heartland of the United Monarchy and the culturally developed part of the country. The wheelmade pottery unearthed in all the fortresses shows many distinct similarities. Cooking pots from several sites, for example, clearly derive from the same overall assemblage. They are shallow, carinated (in profile, they have an angle), and have a round base.

In my view, the pottery associated with the oval fortresses throughout the Central Negev can be confidently assigned to the tenth century B.C. It is difficult to demonstrate this conclusion in non-technical language and it is even more difficult for the non-expert to assess the arguments.

If my arguments as to the dating of this pottery are correct, then a re-dating of a number of excavation levels is required at several important sites somewhat north of the Central Negev fortresses, in the Beer-Sheva Basin (Tel Beer-Sheva, level VII; Tel Masos, level I; and Tel Esdar, levels II–III). The excavators of these sites date these strata earlier than the tenth century B.C. It seems to me that the firm chronology established for the Central Negev fortresses necessitates a re-evaluation of their views. My own examination of the pottery from these levels has convinced me that they date to the tenth century B.C.^c

An accurate date for these Central Negev fortresses is not a merely academic issue. In the period embraced by the 11th and tenth centuries, Israel underwent a major transformation: It developed from a loose confederation of tribes into a unified state. (David’s reign began in about 1000 B.C.) The earlier or later dating of these fortresses is, therefore, crucial. It determines the historical background against which the role of the fortress and settlement network will have to be interpreted. Those who favor the earlier date are inclined to view the fortifications in the context of King Saul’s campaign against the desert nomads, such as the Amalekites. In my view, the fortresses were erected during the reign of King Solomon. Solomon’s reign was undoubtedly a period of expansion and royal planning par excellence, and the establishment of a fortress and settlement network in the Negev would have been of vital importance for the strengthening of his kingdom’s southern region. (On King Solomon’s building of fortifications, see 1 Kings 9:15.) It is in this context that we must understand the Central Negev fortresses and associated settlements.

In addition to the technical dating evidence based on the pottery, there are other arguments, historical and 069archaeological, that support my conclusion.

Let us consider another archaeological fact. Many of the fortresses and the associated settlements contained an ash-layer that convincingly proves these fortresses and their nearby settlements suffered the same sorry end. What or who was the destroyer? I believe there is only one reasonable candidate: Pharaoh Shishak (Sheshonq I).

About five years after Solomon’s death (ca. 924 B.C.), Pharaoh Shishak launched an attack against the kingdoms of Israel and Judah. There are cursory references to this Egyptian campaign in 1 Kings and 2 Chronicles. In 1 Kings 14:25–26, we read “In the fifth year of King Rehoboam, King Shishak of Egypt marched against Jerusalem and carried off the treasures of the House of the Lord and the treasures of the royal palace. He carried off everything; he even carried off all the golden shields that Solomon had made.” In 2 Chronicles 12:1–12, we are told that Shishak came with 12,000 chariots, 60,000 horsemen and innumerable troops. In the swath of destruction, he took the fortified towns of Judah. But Shishak “did not destroy [Rehoboam] entirely.”

A more extensive description of the campaign was inscribed by Shishak himself on the walls of the Temple of Amon in Karnak. In this victory inscription, which is one of the most important historical documents of its time, Shishak lists the names of the cities, villages and settlements he conquered. The first part of the inscription contains a long list of sites in the northern and central sections of Israel. The second part of the inscription, containing over 10 names, is apparently devoted to the Negev. Only a few of the names can be identified with cities known from the Bible. These include Arad, Yurza, Sharnhen, and the proposed identification of Ezion-Geber, which is doubtful.¹⁴

Of particular interest to us here are nine place-names formed with the component p.h-q-r, which can be readily associated with the Semitic root h-g-r (“fort”). The list includes, for example, the fortress of Great Arad, which clearly refers to the site still bearing that name. Identifying the other fortresses referred to by Shishak is more problematic, but it may well be that a number of the names refer to the fortresses we have been discussing.

This suggestion was first put forward by Professor Benjamin Mazar.¹⁵ Interestingly enough, the late Yohanan Aharoni, in his fundamental historico-geographical study, The Land of Israel, agreed with Mazar’s suggestion connecting the Central Negev fortress and settlement sites with the list of conquered sites on Shishak’s victory inscription.¹⁶ Indeed, Aharoni wrote that: “At least some of the sites in the network of forts along the Negev roads date to the United Monarchy. The Israelite agricultural settlement built on Ramat Matred … was founded in the first half of the tenth century and evidently destroyed by Shishak’s expedition shortly after Solomon’s death.”¹⁷

Later Aharoni argued for an 11th-century date for the Central Negev fortresses, largely because he dated level VII at Beer-Sheva, a major excavation he directed, to the 11th century B.C. But he never systematically refuted the reasons he and Mazar had adduced for connecting them with Shishak’s campaign of about 925 B.C.

Aharoni, when contending for an 11th-century date, argued that the settlement and fortress complex in the Central Negev represented a natural extension of the ongoing process of sedentarization commenced by the Hebrew tribes in the 13th century: “We have before us an instructive illustration of settlement pressure and the diffusion of the excess population into the most remote and difficult regions. Israelite settlement in the northern Negev began towards the end of the 13th century, and by the end of the 11th century it had reached the Negev’s southernmost corners.”¹⁸ I believe this argument is applicable to the Israelite expansion into the northern part of the area, but not the area in which these Central Negev fortresses are found. In considering the history of the Israelites in the Negev, one must distinguish between the northern part of the area, with the Beer-Sheva Basin at its focus, and the southern part, which features the limestone highlands of our immediate concern. The excavators of Beer-Sheva and Masos attribute the beginnings of Israelite sedentarization in the Beer-Sheva Basin to the beginning of the 13th and start of the 12th centuries B.C. While I am convinced that they have set the dates of their lowest (earliest) levels too far back, I basically concur with their description of the 11th century as a kind of efflorescence in that region. This corresponds with what is known as the Period of the Judges or (less Biblically) the Settlement period in Palestine. My disagreement with some of my colleagues concerns the end of this phase. Both historical and strictly archaeological considerations have led me to the conclusion that the Central Negev fortresses and settlements were not in existence in the Period of the Judges; their brief occupation can be dated to the tenth century B.C.

There can be no doubt that the Central Negev fortresses and settlements constitute the principal evidence of Israelite settlement in the area south of Beer-Sheva. The reality of Shishak’s devastating campaign to the Central Negev is accepted by everyone. Moreover, if the Central Negev fortresses and settlements were not among the sites enumerated by Shishak, we are confronted with an acute historico-archaeological problem: Where are the settlements destroyed by Shishak? There simply are no other possibilities. Following the destruction of these 070fortresses, the Central Negev was virtually abandoned. The conclusion is inescapable: Shishak demolished the Central Negev fortresses and settlements in about 924 B.C.

The excavations of these sites have all indicated that they were occupied only for a brief period, 50 years at most. They could not have been erected long before their demise. Therefore, they must have been constructed some time in the tenth century. Other archaeological evidence, such as the presence of red-burnished pottery, supports this contention.

Another reason I believe these fortresses were constructed in King Solomon’s reign is that they reflect a unified effort involving the systematic construction of dozens of remote but at the same time substantial strongholds. Such an effort clearly implies an initiative by a strong central authority, an authority not evident in the 11th century, the days of the Judges. In the tenth century, by contrast, there was a strong central authority in the person of King Solomon himself. In my view, the fortresses were erected during the reign of King Solomon, a vigorous and powerful ruler, whose numerous public works included the fortification of cities, the construction of storehouses, and the founding of distant trading-posts. Solomon’s fortress network provided a firm defensive line against attack from the south, and accordingly can be understood as the southern border of his kingdom.¹⁹

Following Shishak’s campaign, Judah’s southern border retreated to its former line along the Beer-Sheva Basin. The Israelites continued to display a definite interest in their country’s southern region, as witnessed by the succession of ambitious fortresses at Kadesh-Barnea. But the Central Negev, by and large, was abandoned, and no attempt at serious resettlement was undertaken for many centuries, until the coming of the Nabateans in the third and second centuries B.C.

An enormous number of Iron Age fortresses have been uncovered in the Central Negev, especially in recent years. The question is, what are they doing here? The answer depends in large part on who built them. And to determine who built them, we need to know as precisely as possible when they were built. The remains of a few of these fortresses were found even before World War I. In their famous archaeological survey for the London-based Palestine Exploration Fund, Sir Charles Leonard Woolley and T. E. Lawrence (“Lawrence of Arabia”) discovered Qast er-Ruheileh, Bir Birein and Tell Ain […]