Solomon’s Negev Defense Line Contained Three Fewer Fortresses

040

In the May/June 1985 BAR,^a I reported on a large number of Iron Age fortresses in the central Negev desert. I argued that these fortresses (more than 40), formed Solomon’s defense line on the south. Together with their associated settlements, they revealed a uniform fortification effort involving the systematic construction of substantial strongholds. Such a building program, I asserted, had to have been directed by a strong central authority, like the one that existed in Solomon’s time (tenth century B.C.). The plans of these fortresses were oval for the most part, but some were rectangular with unequal sides, and four were square.

As a result of excavations conducted in the spring of 1985, it is now clear that three of the four square fortresses were misdated.^b They should be dated to the Persian period (sixth to fourth centuries B.C.) rather than to the tenth century B.C., as I previously dated them.

I hasten to add that redating these three square fortresses does not materially affect the thesis presented in my article. There are simply three fewer fortresses to support it. That is all. Dozens of fortresses in the area still unquestionably date to the Iron Age (tenth century B.C.).

Yet BAR readers will surely ask, how did such an error occur? They are entitled to an explanation, which will illustrate the uncertainties that still characterize our efforts to date structures and will perhaps suggest more caution in reaching our conclusions.

My earlier—and incorrect—dating of the three square fortresses was based in part on a small rescue excavation at one of these fortresses, Horvat Mesora, that I conducted in 1975 on behalf of the Israel Department of Antiquities. We originally surveyed two principal structures. One, the square fort, 70.5 feet (21.5 m) on each side, consisted of a casemate wall divided into nine rooms that enclosed a central courtyard. About 65 feet to the west was a second structure, similar in plan, though rectangular instead of square. This second structure consisted of casemate rooms arranged about a central courtyard, but there were seven casemate rooms instead of nine; the structure itself measured 56 feet (17 m) by 36 feet (11 m). I excavated two of the seven casemate rooms in this rectangular structure and found on the floor complete pottery vessels from the tenth century B.C.—the time of King Solomon.

I assumed—incorrectly as it turned out—that the square fortress, like the nearby, similar rectangular structure to the west, was from the same period. True, a probe in one of the casemate rooms of the square fortress 042revealed a few pottery sherds of the Persian period (sixth to fourth centuries B.C.), but these I interpreted as indicating a reuse of the building during the Persian period.

In the spring of 1985, I returned to the site of Horvat Mesora and excavated two of the casemate rooms in its square fortress. On the floors of these rooms I found pottery vessels typical of the Persian period. Most of them were large so-called Persian bowls and storejars. Moreover, the clay ovens for cooking and baking found in the rooms indicated only one occupation phase. Thus the square fortress at Horvat Mesora must be dated to the Persian period, not to King Solomon’s time, although the rectangular structure 65 feet west of this square structure does date to King Solomon’s time and probably served as the fortress for the settlement during that period.

As a result of my redating of the square fortress at Horvat Mesora, I was forced to reconsider my previous dating of another site, Horvat Ritma. Horvat Ritma was surveyed in the 1960s by Yohanan Aharoni, the father of the archaeology of the Negev. Aharoni dated the square fortress at Horvat Ritma to the Iron Age (a period that includes the tenth century B.C.). Aharoni’s dating was generally accepted and we, perhaps too easily, assumed it was correct.

In 1967 I surveyed the site and in 1970 carried out excavations there with Ze’ev Meshel as co-director. The site is spread over a large area and includes many structures in addition to the square fortress. We excavated at five of the Horvat Ritma structures, including the square fortress. Within the fortress we completely excavated one casemate room and partially excavated another. Meshel published the results of the excavation in 1977,¹ noting that most of the pottery from these two rooms in the square fortress of Horvat Ritma was from the Nabataean period, “although two sherds of Israelite bowls were found on or in the bedding of the floor.” Nevertheless, Meshel concluded (and I agreed at the time) that the square fortress at Horvat Ritma had been built in the Iron Age (tenth century B.C.). This conclusion was based primarily on the similarity in groundplan and dimensions to other fortresses that had been dated to the Iron Age.² For example, both the Horvat Ritma and the Horvat Mesora fortresses were not only square, but they were also nearly identical in dimensions and even contained almost the same number of casemate rooms, as if they had been built from the same architectural plan. In addition, the four other structures we excavated at Horvat Ritma contained a fair number of pottery vessels datable to the tenth century B.C. In his published report of the 1970 Horvat Ritma excavation, Meshel assumed, as I did, that the later pottery sherds we found in the buildings around the square fortress came from the later reuse of these buildings. As Meshel stated, the buildings surrounding the square fortress “were frequently cleared out and reused in later periods.” In fairness, it should be noted that Meshel also added that only further excavations could confirm his hypothesis.

Now, after we redated the square fortress at Horvat Mesora to the Persian period, it was necessary to redate the nearly identical square fortress at Horvat Ritma to the Persian period as well.

The third square fortress I incorrectly dated in my BAR article³ to the tenth century B.C.—King Solomon’s time—was the square fortress opposite Atar Haro‘a located on a ridge opposite the important Israelite oval fortress and settlement of Atar Haro‘a. This square fortress is almost identical in plan and dimensions (67.3 feet [20 m] on a side) to the other two square fortresses and, like Mesora, contains nine casemate rooms.

I had made a small probe of this fortress in the 1960s but did not find enough pottery to form a sound basis for dating the fortress. Accordingly, on the basis of the similarity of its plan to the fortresses at Horvat Mesora and Horvat Ritma, I dated the fortress opposite Atar Haro‘a to the tenth century B.C. Obviously this dating became untenable as soon as I redated the fortresses at Horvat Mesora and Horvat Ritma. But the redating of the square fortress opposite Atar Haro‘a became even more secure when I returned to the site in 1985.

In the spring of 1985, I excavated eight casemate rooms in this fortress. This time, unlike my probes at the site in the 1960s, I found easily datable pottery. On the floor of these rooms I found pottery vessels typical of the Persian period. Most of them were large so-called Persian bowls and store jars. Moreover, like the fortress at Horvat Mesora, only one phase of occupation was indicated by the clay ovens for cooking and baking found in the rooms. Thus the square fortress at Mesad Nahal Haro‘a, like the square fortresses at Horvat Mesora and Horvat Ritma, must be dated to the Persian period, not to King Solomon’s time.

As I previously stated, these redatings do not significantly detract from the thesis of my BAR article concerning the many remaining Solomonic fortresses in the central Negev.

The new datings of these three square fortresses to the Persian period do, however, add an important new element to the historical reconstruction of the area during the sixth to fourth centuries B.C. They may well indicate that the Biblical “Way of the Spies” (Numbers 21:1) continued in use after the fall of the kingdoms of Israel and Judah and into the Persian period. This is the route from Arad in the northern Negev to Kadesh-Barnea on 043the Negev-Sinai border, where Caleb and Joshua traveled to spy out the Promised Land.

Both Arad and Kadesh-Barnea have yielded significant Persian period remains. At Arad, 100 Aramaic ostraca from this period were found.⁴ At Kadesh-Barnea there are indications that some of the casemate rooms of the upper fortress (dated to the seventh and sixth centuries B.C.), particularly along its east side, were used as temporary dwelling quarters in the Persian period.⁵ In the northern part of the site a small room, apparently built by these Persian period settlers, was excavated. Finds from this period also come from pits dug into levels of the earlier fortresses at Kadesh-Barnea. The finds in these pits include imported Greek vessels, “Persian” large bowls, juglets and storage jars. We also found at Kadesh-Barnea an Aramaic ostracon from this period.

Now, we have three additional fortresses from the Persian period along the “Way of the Spies,” a route that I had previously dated to the Solomonic period.

Ephraim Stern has suggested that it is possible to determine the borders of the Persian province of Judah by the distribution of Yhd seal impressions.⁶ If that is true—and I believe such a possibility is likely—the other three square fortresses, between Arad and Kadesh-Barnea could 045have been strongholds along the southern border of the province of Judah.⁷ For at Kadesh-Barnea we found a seal impression stamped Yhd on a jar handle, the common designation of Yehud, the Persian satrapy of Judah. This is the first time that a Yhd seal impression has been discovered so far south. But this may alternatively suggest that there were large numbers of Jews living to the south of the province of Judah, as can be surmised from Nehemiah 11:25–30, which mentions many settlements from the Negev outside the borders of the province. It is stated that “they dwelt from Beer-sheba unto the valley of Hinnom” (Nehemiah 11:30).

Another possibility is that our three square Persian period fortresses were under Arabian control, although subject to Persian overlords. From the Arabian peninsula, these Arabs controlled the spice and frankincense routes from southern Arabia north to Transjordan and west to Gaza across the central Negev. We know about this from Herodotus,⁸ but until now we were unaware of the possibility that this trade route was protected by a series of fortresses. Perhaps in the future, other fortresses from this period may be found either in Israel or Jordan. These early Arabs were thus performing the same functions that their descendants the Nabateans performed 200 years later. This route later became a section of the famous Nabataean “incense road” from Petra to Gaza.⁹

The photographs in this article were taken by Nachshon Sneh of Kibbutz Gat.

In the May/June 1985 BAR,a I reported on a large number of Iron Age fortresses in the central Negev desert. I argued that these fortresses (more than 40), formed Solomon’s defense line on the south. Together with their associated settlements, they revealed a uniform fortification effort involving the systematic construction of substantial strongholds. Such a building program, I asserted, had to have been directed by a strong central authority, like the one that existed in Solomon’s time (tenth century B.C.). The plans of these fortresses were oval for the most part, but some were rectangular with unequal sides, and […]