Forts or agricultural estates? Persian period settlement in the territories of the former kingdom of Judah

ABSTRACT

The territories of the former kingdom of Judah were only sparsely settled during the Persian period, as exemplified by the extreme rarity of domestic structures unearthed in excavations. Viewed against this background, the large number of excavated forts and isolated administrative buildings from this period is remarkable, and they apparently outnumber the period's excavated dwellings. Not only is this an extremely unlikely situation, but various lines of evidence, pertaining to specific sites as well as to the phenomenon as a whole, render the possibility that all these structures were forts or administrative buildings re-examines implausible. Consequently, this article reexamines the phenomenon within the social landscape of the region in particular, and of the Achaemenid empire in general, in an attempt to embed those unique buildings within the broader demographic and political reality of this time. Given the location of many of the sites and the finds unearthed in them, and in light of the demographic reality in the region and of the broader Achaemenid imperial policy, the article suggests that most of the so-called forts were estates, created in the process of the resettlement of this previously devastated region.

KEYWORDS:

• Persian period
• Achaemenid empire
• Judah
• Yehud
• fortified structures
• settlements
• agricultural estates

ORCID

Avraham Faust http://orcid.org/0000-0001-7301-7135

Notes

1. Regardless of the question whether this concentration reflects an ancient reality, or results from biases that affect modern scholarship.

2. Fantalkin and Tal 2006, 182–83 are not explicit on this matter, but appear to accept Reich's suggestion, or at least the administrative nature of the site as a whole, see also Fantalkin and Tal 2012, 155-58.

3. The above list does not include all the sites that were identified in the past as Persian period forts, and is selective. Hoglund (1992, 165–205), who is partially responsible for the reinterest in the subject, attributed a Persian period date to additional structures, many of which were probably not settled at this time, e.g., el-Qul'ah [S]), or did not function as forts in the Persian period, e.g., Kh. Abu et-Twein. At el-Qul'ah (S) no Persian period whatsoever were reported (Hoglund 1992, 199–200; Kallai 1972, 165, site 9). At Kh. Abu et-Twein (Hoglund 1992, 191–97) only a limited amount of Persian pottery was unearthed, and following the excavator (Mazar 1982) it is quite clear that the building (if indeed a fort) was built in the Iron Age, and was only reused, perhaps in a different plan, in the Persian period (see extensive discussion in Faust 2012a, 237–40; see also Mazar and Wachtel 2015, 237, note 6). We will comment on this further below.

4. A recent publication (Alexandre 2015, 60) suggests that storage jars comprised some 75% of the assemblage, but does not supply an additional figures. The difference is immaterial for our purposes.

5. The site had both Iron Age and Persian period phases, but since the excavations did not reach the floors in most rooms, the excavators could not identify the plan of each phase. The overall size of the structure, however, was rougly 45 × 28 m (based on the plan on p. 161. The figures are rounded, as the plan is not completely regular, and exclude the towers). The excavators note the large percentage of Persian period imports, and the finding of 47 camel bones, indicating that the site was connected with trade.

6. Not only did Hoglund discuss sites in which no Persian period pottery was unearthed (above), but his redating of other sites to the fifth century was based only on historical reasoning, and is not supported by the archaeological data (see also Fantalkin and Tal 2006).

7. Bocher and Freund (2017) suggested recently (in a conference presentation and a published paper) that a few additional Iron Age sites continued to exist into the early Persian period. While the addition of a few tiny sites to the early Persian period list would not change the overall settlement picture, it must be stressed that their attempt to change the date of some of the sites is based on very problematic methodology, and should probably be rejected. Bocher and Freund based their suggestion on the observation that a small number of forms unearthed in a few Iron Age farmsteads are slightly different from the typical Iron Age forms of this period, mainly in material, and claimed that the difference is chronological. The differences, however, are not only minor, but refer to a very small percentage of the assemblage, and are therefore likely to reflect localized production centres. This is supported by the concentration of the discussed sites mainly to the south of Jerusalem. Furthermore, an understanding of archaeological formation processes casts strong doubts on the logic behind the new suggestion. As noted above, the last phase of occupation is the one that produces the vast majority of the finds. If the sites under discussion would have existed in the early Persian period, most of the forms unearthed were supposed to be from this period. That no typical Persian period forms were usually unearthed suggests that even if the minor differences should be interpreted as chronological, the discussed sites exceeded the other Iron Age sites by a few years only (otherwise these forms must have been far more differentiated and far more numerous). In addition, the underlining logic behind the restudy of the material is very problematic, since it takes an assemblage and breaks into different forms, hence artificially “spreading” it across a much longer period. Taking an assemblage and treating it as it was unearthed in a survey inevitably results with skewed periods of occupation. If, for some reason, one doesn't want to attribute the minute differences to localised production centres, and opt for a chronological difference, the most that can be suggested is that the discussed sites ceased to exist not in 586 but rather in 575 BCE or the like, but not much later. Notably, as an off-shoot of the Ramat Rahel project, Bocher and Freund's study suffers from relying on the assumption that this central site was a center for the collection of taxes and that many of the small Iron Age sites were its satellites. A systematic study of the data, however, reveals that the small sites were not connected with Ramat Rahel, which at least during the Iron Age was a center of production in its own right (see Moyal and Faust 2015, and references).

8. In the coastal plain more remains of buildings are known, for example at Shiqmona, Tel Yaoz, Megadim, Dor, and Tell Abu Hawam. Most of the structures, there, however, were only partially excavated, or remained unpublished (e.g., Elgavish 1994, 82; Fischer, Roll, and Tal 2008; Broshi 1993; Stern 1995; Balensi, Herrera and Artzi 1993, and references), hence limiting our ability to reach general conclusions even in this region.

9. Since Kh. Luzifar was only surveyed, it is likely that sherds from large vessels are somewhat better represented (compare Faust and Katz 2012, 174, 179, 181), and this might have somewhat increased their percentage in the collected assemblage (tough they must have been significant to start with). At Nahal Tut a large part of the site was not excavated, and it is possible that specialized activity in the excavated areas skewed the results. Still, even if the percentage of storage jars should be somewhat decreased, it would still be very high.

10. It is possible that this figure is slightly higher because the data refer to the entire site (and not only the fortified structure) but even so the percentage is still much lower than in “real” forts.

11. The author noted (Freud 2007, 119 note 2) that the statistics relates only to the published data (and not to the complete corpus), but there is no way to calibrate the data and (with caution) we refer to the data as is.

12. Note that in a few sites in Yehud some imported sherds were found, but those were mainly body sherds, which should not be included in any statistics (as local body sherds are not counted nor reported) and, furthermore, in many sites not a single imported sherd was found, exemplifying the rarity of this pottery in Yehud (the few body sherds, and even fewer rims, uncovered in the region, only exemplify the rarity of this pottery). Detailed discussion is beyond the scope of the present paper (see recently Sandhaus and Kreimerman 2015, 267, 269). For possible Iron Age roots of this tendency, see Faust 2006a, 49–64, 2006b and references; see also Gilboa, Sharon and Bloch-Smith 2015, regarding the Israelite phase at Dor, and Faust and Katz 2017, 13, for some of the qualities of pottery in this regard.

13. There is no consensus regarding a universal density coefficient (and even whether there is one), and figures vary (e.g., Naroll 1962; Brown 1987; Ember and Ember 1995; and references). Still, whatever coefficient one wishes to use we are discussing many dozens of peoples that resided in each structure, and sometimes perhaps even more. The comparison to Iron Age dwellings (below) only strengthen this conclusion.

14. The absence of imported pottery in sites in Yehud versus its presence in sites in the region that became Idumaea and in real forts, renders the biblical description of the returnees (to Yehud) as Judeans or Jews (the exact terminology is irrelevant for our purposes, see, e.g., Grabbe 2004; Rappoprt 2004; Berlin 2013), descendants of the exiled from Judah, very likely (i.e., even if among them there were also others, which is more than likely, the majority were probably such, and the others adopted this identity).

15. In addition to formation processes which slightly biased the data (especially this collected from surveys).