Jerusalem in the Persian (and Early Hellenistic) Period and the Wall of Nehemiah

Israel Finkelstein, Institute of Archaeology, Tel Aviv University, Tel Aviv 69978, Israel  

Journal for the Study of the Old Testament Vol 32.4 (2008): 501-520 

Abstract

Knowledge of the archaeology of Jerusalem in the Persian (and Early Hellenistic) period — the size of the settlement and whether it was fortified — is crucial to understanding the history of the province of Yehud, the reality behind the book of Nehemiah and the process of compilation and redaction of certain biblical texts. It is therefore essential to look at the finds free of preconceptions (which may stem from the account in the book of Nehemiah) and only then attempt to merge archaeology and text.

The Current View

A considerable number of studies dealing with Jerusalem in the Persian period have been published in recent years (e.g. Carter 1999; Eshel 2000; Stern 2001: 434-36; Edelman 2005; Lipschits 2005, 2006; Ussishkin 2006). Although the authors were aware of the results of recent excavations, which have shown that the settlement was limited to the eastern ridge (the City of David), they continued to refer to a meaningful, fortified 'city' with a relatively large population.

Carter argued that Jerusalem grew from a built-up area of 30 dunams in the Persian I period to 60 dunams 'after the mission of Nehemiah' (1999: 200) and estimated the peak population to have been between 1250 and 1500 people (1999: 288). Based on detailed archaeological data from excavations and surveys, and using a density coefficient of 25 people per one built-up dunam (a number which may be somewhat too high—see below), Carter (1999: 195-205) reached a population estimate of c. 20,000 people for the entire province of Yehud in the Persian period (compared to c. 30,000 according to Lipschits 2003: 364, also using the 25 people per one built-up dunam coefficient). Carter rightly asked: If Yehud 'was this small and this poor, how could the social and religious elite sustain the literary activity attributed to the Persian period? ... How could such a small community have built a temple and/or refortified Jerusalem' (1999: 285; see the same line of thought in Schniedewind 2003; 2004: 165-78). Based on 'historical and sociological parallels'—in fact, almost solely on the biblical text—Carter (1999: 288) answered in the positive, arguing that the urban elite was large enough for both the production of a large body of texts and for the fortification of Jerusalem.

Eshel (2000) reconstructed the history of Jerusalem in the Persian period almost solely according to the biblical texts, arguing that the 'Jerusalem of Nehemiah was a small town...nevertheless it had eight gates. much more than the real need of the town at that time' (2000: 341). Eshel acknowledged that the population of Jerusalem was depleted (in line with Neh. 7.4), but in the same breath argued that it was populated by Levites and others, who were brought to Jerusalem by Nehemiah (11.1). He com­pared the demographic actions taken by Nehemiah to the synoikismos policy of Greek tyrants (also Weinberg 1992: 43; 2000: 308-309, 313-16). Regarding the rebuilding of the walls, following Nehemiah 3, Eshel envisioned a major operation, which involved many groups of builders.

Stern began the discussion of the archaeology of Jerusalem in the Persian period with a sentence based solely on the biblical text: 'Persian period Jerusalem was bounded by walls erected by Nehemiah' (2001: 434). At the same time, Stern acknowledged that 'only a few traces have survived of the city wall of Nehemiah along the course described in the Bible (2001: 435). Stern referred to a segment of a city-wall at the top of the eastern slope of the City of David, which was dated by Kenyon (1974: 183-84) to the Persian period (see below).

Edelman accepted Nehemiah 3 as accurately reflecting 'the names of individuals and settlements in Yehud at the time the walls of Jerusalem were constructed during the reign of Artaxerxes I' (2005: 222). Edelman, like Lipschits (below), saw the construction of the walls by Nehemiah as a turning point in the history of Yehud — marking the transfer of the capital from Mizpah to Jerusalem. The walls provided 'protection for the civilian population and government officials who would man the fort and carry out the administration of the province' (2005: 206). Edelman saw a major construction effort in Jerusalem under Persian auspices in the days of Artaxerxes I — an effort far greater than the reconstruction of the city-walls, that also included the Temple and a fort (2005: 344-48).

Ussishkin declared that 'the corpus of archaeological data should be the starting point for the study of Jerusalem... This source of information should take precedence, wherever possible, over the written sources, which are largely biased, incomplete, and open to different interpretations' (2006: 147-48). Reviewing the archaeological data, he rightly concluded that the description in Nehemiah 3 must relate to the maximal length of the city-walls, including the western hill. But then, solely according to the textual evidence in Nehemiah 3, he accepted that the Persian-period settlement was indeed fortified: 'When Nehemiah restored the city wall destroyed by the Babylonians in 586 BCE, it is clear ...that he restored the city wall that encompassed the Southwestern Hill, as sug­gested by the "maximalists" (Ussishkin 2006: 159; also 160).

Lipschits' reconstruction of the history of Jerusalem in the Persian period (2006) revolved around the rebuilding of the city-wall by Nehemiah. Though 'there are no architectural or other finds that attest to Jerusalem as an urban center during the Persian Period' (2006: 31), 'the real change in the history of Jerusalem occurred in the middle of the fifth century BCE, when the fortifications of Jerusalem were rebuilt. Along with scanty archaeological evidence, we have a clear description of this event in the Nehemiah narrative.' (2006: 34). Lipschits saw the construction of the city-wall as the turning point in the history of Jerusalem — when it became the capital of Yehud: 'The agreement of the Persians to build fortifications in Jerusalem and to alter the status of the city to the capital of the province was the most dramatic change in the history of the city after the Babylonian destruction in 586' (2006: 40). Lipschits described Jerusalem as a 'city' of 60 dunams, with a population of c. 1500 inhabitants (2006: 32; also 2003: 330-31; 2005: 212; see a different number, 3000 people, in 2005: 271).

Obviously, all the scholars who dealt with the nature of Jerusalem in the Persian period based their discussion on the biblical text, mainly on the description of the reconstruction of the city-wall in Nehemiah 3.

The Finds

Intensive archaeological research in Jerusalem in the past forty years has shown that:

1.  The southwestern hill was part of the fortified city in the Late Iron II and the Late Hellenistic periods (for the Iron II, see Geva 1979; 2003a: 505-18; 2003b; Avigad 1983: 31-60; Reich and Shukron 2003; for the Late Hellenistic period, see Geva 1983, 1994, 2003a: 526-34; Broshi and Gibson 1994; Chen, Margalit and Pixner 1994; Sivan and Solar 1994; Wightman 1993: 111-57).

2.  The southwestern hill was not inhabited in the Persian and Early Hellenistic periods. This has been demonstrated by excavations in the Jewish Quarter (Avigad 1983: 61-63; Geva 2000: 24; 2003a: 524; 2003b: 208), the Armenian Garden (Gibson 1987; Geva 2003 a: 524-25), the Citadel (Amiran and Eitan 1970) and Mt. Zion (Broshi 1976: 82-83). Apart from a few possible isolated finds (Geva 2003a: 525), there is no evidence of any activity in any part of the southwestern hill between the early sixth century and the second century BCE. The Persian and Early Hellenistic settlement should therefore be sought on the southeastern ridge — the City of David.

In the City of David, too, the evidence is fragmentary. Most finds from the Persian and Early Hellenistic periods were retrieved from the central part of the ridge, between Areas G and D of the Shiloh excavations (Shiloh 1984: 4). The Persian period is represented by Stratum 9, which fully appears, according to Shiloh (1984: 4, Table 2) in Areas D1 (Ariel, Hirschfeld and Savir 2000: 59-62), D2 and G (Shiloh 1984: 20), and which is partially represented in Area E1. According to De Groot (2001), the most significant finds were retrieved from Area E. But even in these areas the finds were meager and poor; most of them came from fills and quarrying refuse (see the difficulty of distinguishing the 'limestone chops layer' in Ariel, Hirschfeld and Savir 2000: 59). De Groot (2001) describes a possible reuse in one Late Iron II building in Area E. Persian-period sherds and a few seal impressions were found in Reich and Shukron's Areas A and B, located in the Kidron Valley and mid-slope respectively, c. 200-250 m south of the Gihon Spring; they seem to have originated in the settlement located on the ridge (Reich and Shukron 2007).

Stratum 8 stands for the Early Hellenistic period. It is fully represented only in Area E2, partially represented in Areas E1 and E3, and scarcely represented in Areas D1 and D2 (Shiloh 1984: 4, Table 2). In this case, too, the finds are meager. They are comprised of three columbaria (De Groot 2004) and a structure that yielded the only assemblage of Early Hellenistic pottery from Jerusalem (in Area E1 — Shiloh 1984: 15).

In the case of the City of David, too, the negative evidence is as important as the positive. No Persian or Early Hellenistic finds were found in Area A on the southern tip of the ridge. It is significant to note that in Area A1 Early Roman remains were found over Iron II remains (De Groot, Cohen and Caspi 1992). In Kenyon's Site K, located on the southwestern side of the City of David, c. 50 m to the north of the Siloam Pool, Iron II sherds were found on bedrock, superimposed by Late Hellenistic finds (Kenyon 1966: 84).

As for the northern part of the ridge, the Persian and Early Hellenistic periods were not represented in B. and E. Mazar's excavations to the south of the southern wall of the Temple Mount, which yielded Late Hellenistic and mainly Early Roman finds superimposed over Iron II buildings (E. Mazar and B. Mazar 1989: xv-xvi). It is also significant that Persian and Early Hellenistic finds were not reported from B. Mazar's excavations near the southwestern corner of the Temple Mount (B. Mazar 1971). A few finds, but no architectural remains or in-situ assemblages of pottery, were retrieved by Crowfoot in the excavation of the 'Western Gate' (Crowfoot and Fitzgerald 1929) and by Macalister and Duncan (1926) in the excavation immediately to the west of Shiloh's Area G. The 8-10 m thick dump-debris removed by Reich and Shukron on the eastern slope of the City of David, near the Gihon Spring (Reich and Shukron 2007; also 2004), yielded ceramic material from the Iron II and 'late Second Temple period', but no Persian and Early Hellenistic pottery. Reich and Shukron interpret this as evidence that Area G, located upslope from their dig, was uninhabited at that time. Finally, it is noteworthy that sifting of debris from the Temple Mount recovered almost no Persian- period finds (compared to a significant number of finds from the Iron II and from the Hellenistic-Early Roman periods — Barkay and Zweig 2006).

Reich and Shukron (2007) also noted that 75 of the 85 Yehud seal impressions from the Shiloh excavations published by Ariel and Shoham (2000) originated from Areas B, D and E. They concluded that the settlement of the Persian and Early Hellenistic periods was restricted to the top of the ridge, to the south of Area G (see a somewhat similar view in Ariel and Shoham 2000: 138).

All this seems to indicate that:

1.  In the Persian and Early Hellenistic periods activity on the Temple Mount was not strong (compare the Iron II finds to the south of the southern wall of the Temple Mount to the negative evidence for the Persian and Early Hellenistic periods and see Barkay and Zweig 2006), and in any event did not include intensively inhabited areas.

2.  The northern part of the ridge of the City of David was uninhabited.

3.  The southern part of the ridge was probably uninhabited as well.

The Persian and Early Hellenistic settlement was confined to the central part of the ridge, between Shiloh's Area G (which seems to be located on the margin of the inhabited area) and Shiloh's Areas D and E. The settlement was located on the ridge, with the eastern slope outside the built-up area. Even in this restricted area, a century of excavations, by a number of archaeologists, failed to yield even a single(!) house or proper floor from the Persian period, and only one structure from the Early Hellenistic period was found. The idea that the settlement was eradicated because of later activity and erosion (e.g. De Groot 2004: 67) must be rejected in the light of the reasonable preservation of the Late Hellenistic and Iron II remains.

The maximal size of the Persian and Early Hellenistic settlement was therefore c. 240 (N-S) x 120 (E-W) m, that is, c. 20-25 dunams (contra to the idea of a 60-dunam settlement [excluding the Temple Mount] in Carter 1999: 200; Lipschits 2006: 32; and a 30-acre settlement [possibly including the Temple Mount] in Avigad 1993: 720). Calculating the population according to the broadly accepted density coefficient of 20 people per one built-up dunam (Finkelstein 1992 and bibliography) — a number which may be too high for what seems to have been a sparsely settled ridge (on this problem too see Finkelstein 1992) — one reaches an estimated population of 400-500 people, that is, c. 100 adult men. This stands in sharp contrast to previous, even minimal estimates of 1250, 1500 or 3000 inhabitants (Carter 1999: 288; Lipschits 2005: 271; 2006: 32; 'a few thousands' in Avigad 1993: 720), estimates which call for a large settlement of 75-150 dunams — more than the entire area of the City of David. These data fit well the situation in the immediate environs of Jerusalem, where the number of spots with archaeological remains dropped from 140 in the Iron II to 14 in the Persian period (Kloner 2001: 92; 2003: 28*; 2003: 30* for the Early Hellenistic period). They also fit the general demographic depletion in the entire area of the province of Yehud — a maximum of 20,000-30,000 people in the Persian period according to Carter (1999: 195-205) and Lipschits (2003: 364), c. 15,000 according to my own calculations—about a third or a fourth of the population of that area in the Late Iron II (Carter 1999: 247 based on Broshi and Finkelstein 1992; Ofer 1993).

Nehemiah's Wall

Archaeologists have accepted the description of the reconstruction of the wall in Nehemiah 3 as an historical fact, and have been divided only about the course of the fortifications. The minimalists restricted them to the City of David, and the maximalists argued that the description included the southwestern hill (see the summary in Ussishkin 2006). Two finds in the field have been perceived as indications for the course of Nehemiah's city-wall: one on the crest above the eastern slope of the City of David and the other on the western side of that ridge.

Kenyon (1967: 111; 1974: 183-84) argued that because of the collapse of the Late Iron II city-wall and buildings on the eastern slope of the ridge as a result of the Babylonian destruction, the city-wall of Nehemiah was built higher up, at the top of the slope. In her Square A XVIII (adjacent to Shiloh's Area G), Kenyon identified a short segment in the city-wall that had first been uncovered by Macalister and Duncan (1926) — a wall that was later unanimously dated to the Late Hellenistic period (see the literature on the First Wall above) — as the city-wall built by Nehemiah. Her dating of this segment of the wall was based on pottery found in a layer dumped against its outer face; this pottery was dated by Kenyon (1974: 183) to the fifth-early third centuries BCE (the sixth-fifth centuries BCE in 1974: caption to Pl. 79). Shiloh, too, argued —without any archaeological evidence — that the city wall was built 'on the bedrock at the top of the eastern slope' (1984: 29; also Avigad 1993: 720; De Groot 2001: 78). Stern (2001: 435) accepted Kenyon's identification and dating of this segment as Nehemiah's wall. Ussishkin (2006: 160), on the other hand, suggested that Nehemiah reconstructed the Iron II wall, which runs on the lower part of the eastern slope of the City of David.

The only piece of information from the western side of the City of David comes from Crowfoot's 1927 excavations. A massive structure that had been founded on bedrock, under thick layers of later occupations and debris, was identified as a Bronze Age gatehouse that continued to be in use until Roman times (Crowfoot and Fitzgerald 1929: 12-23). Albright (1930-31: 167) identified Crowfoot's 'gatehouse' with the Dung Gate of Neh. 3.13, while Alt (1928) proposed equating it with the Valley Gate of Neh. 3.13. Alt's proposal has been accepted by most authorities (e.g. Avi Yonah 1954: 244-45; Tzafrir 1977: 39; Williamson 1984; Eshel 2000: 333).

Yet, both finds — the wall uncovered by Kenyon and the structure unearthed by Crowfoot—cannot be dated to the Persian period.

Kenyon's identification of Nehemiah's wall was based on (yet unpub­lished) pottery found in a small sounding, in a fill or a dump thrown against the outer face of the wall (1974: Pl. 79). As rightly argued by De Groot (2001: 78), such a layer cannot be used for dating a city-wall. This material could have been taken from any dump on the slope and put there in order to support the wall (for the same situation in the Outer Wall of Gezer, see Finkelstein 1994: 278). Shiloh re-examined this segment of the city wall and found Late Bronze material on the bedrock, close to its inner face; he therefore suggested that this part of the wall may have originated from a pre-Persian period (Cahill and Tarler 1994: 41). Excavations immediately to the west of this spot by Macalister and Duncan (1926) and E. Mazar (2007) did not unearth architectural remains of the Persian and Early Hellenistic periods, but they made clear that this segment is part of the Late Hellenistic city-wall, first uncovered by Macalister and Duncan (1926; see for details Finkelstein et al. 2007). Had it not been for Nehemiah 3, I doubt very much whether Kenyon would have dated a short segment in the well-preserved Late Hellenistic wall to the Persian period.

Ussishkin (2006) has recently dealt in detail with the structure excavated by Crowfoot and identified by him as a gatehouse. Ussishkin has cast doubt on the identification of the structure as a gate, and convincingly argued that it probably dates to the Late Hellenistic or Early Roman period (2006: 159; see also Kenyon 1964: 13).

To sum up this issue, there is no archaeological evidence for the citywall of Nehemiah. The wall in the east dates to the Late Hellenistic period and the structure in the west — regardless of its function — also post-dates the Persian period. Had it not been for the Nehemiah 3 account, no scholar would have argued for a Persian-period city-wall in Jerusalem. Three early city-walls are known in the City of David, dating to the Middle Bronze Age, the Late Iron II and the Late-Hellenistic period. All three have been easy to trace and have been found relatively well-preserved. No other city-wall has ever been found and I doubt if this situation will change as a result of future excavations.

One could take a different course and argue, with Ussishkin (2006), that Nehemiah merely rebuilt the ruined Iron II wall. Yet, in the many sections of the Iron II wall that have been uncovered — on both the southwestern hill and the southeastern ridge — there is no clue whatsoever for a renovation or reconstruction in the Persian period. In the parts of the Late Iron II city wall uncovered on the southwestern hill, the first changes and additions date to the Late Hellenistic period (Avigad 1983: 65-72; Geva 2000: 24; 2003a: 529-32). No such reconstruction has been traced in the long line of the Iron II wall uncovered in several excavations along the eastern slope of the City of David south of the Gihon Spring. Archaeologically, Nehemiah's wall is a mirage.

This should come as no surprise, judging from what we do know about the Persian-period settlement systems in Yehud in particular and the entire country in general. To differ from the construction of the Iron II and Late-Hellenistic fortifications in Jerusalem—which represent a well-organized territorio-political entity with significant wealth and population, evidence for high-level bureaucracy and clear ideology of sovereignty — the small community of several hundred inhabitants of Persian-period Jerusalem (that is, not many more than 100 adult men), with a depleted hinterland and no economic base, could not have possibly engaged in the reconstruction of the c. 3.5 km long(!) Iron II city-wall with many gates (accepting Ussishkin's 2006 reconstruction). And why should the Persian authorities allow the reconstruction of the old, ruined fortifications and make Jerusalem the only fortified town in the hill country? The explanations of scholars who have dealt with this issue— that this was made possible because of the pressure of the Delian League on the Mediterranean coast, revolt in Egypt and so on (summaries in Hoglund 1992: 61-64, 127-28; Edelman 2005: 334-40; Lipschits 2006: 35-38) —seem far-fetched, given the location of Jerusalem, distant from Egypt, international roads, coastal ports or other strategic locations (Lipschits, ibid). Indeed, Persian-period fortifications are known only along the coastal plain (Stern 2001: 464-68).

The Reality Behind Nehemiah 3

So what is the historical reality behind the description of Nehemiah's rebuilding of the walls of Jerusalem?

Scholars have noted the independent nature of the list in Nehemiah 3 as compared to the rest of the 'Nehemiah Memoir' (Torrey 1910: 225; Michaeli 1967: 318-19; Williamson 1985: 200; Blenkinsopp 1988: 231; Throntveit 1992: 74, 75; Grabbe 1998: 157), but are divided on the question of whether Nehemiah used an earlier or a contemporary source that was kept in the Temple archives (Michaeli 1967: 319; Kellermaan: 1967: 14-17; Williamson 1985: 201; Throntveit 1992: 75; Blenkinsopp 1988: 231), or whether a later editor inserted the text into the book of Nehemiah (e.g. Torrey 1896; 37-38; 1910: 249, who identified the editor with the Chronist; Mowinckel 1964: 109-16, who opted for a post-Chronist redactor). Taking into consideration the archaeological evidence presented in the present study, an existing source from the Persian period, which described a genuine construction effort at that time, is not a viable option. We are left, therefore, with the following possibilities:

1.  That the description in Nehemiah 3 is utopian; it was based on the geographical reality of the ruined Iron II city-wall but does not reflect actual work on the wall. The text may describe a symbolic act rather than an actual work, similar to symbolic acts connected to the founding of Etruscan and Roman cities. And it may corre­spond to an ascriptive, ideal-type of a city that ought to include a wall (cf. Odyssey 6.6-10).

2.  That a Persian-period author used an early source, which described the late eighth-century construction or a pre-586 renovation of the Iron II city-wall and incorporated it into the Nehemiah text.

3.  That the description was inspired by the construction of the Late Hellenistic, Hasmonean city-wall.

The first possibility is difficult to accept. The detailed description of the construction of the city-wall and the prominence of the story of the wall throughout the Nehemiah Memoirs (Neh. 1.3; 2.4, 8, 13, 17; 3.33, 38; 4.5, 9; 5.16; 6.1, 6, 15; 7.1; 12.27) renders it highly unlikely. Moreover, the description in Nehemiah 3—which includes reference to many gates, towers, pools and houses — seems to refer to a true reality of a big city; in the light of what has already been said, the Late Iron II and Hasmonean periods are the only options.

The second possibility should probably be put aside. First, there is no evidence — historical or archaeological — of major work on the Iron II city-wall in the late seventh or early sixth centuries, and it is doubtful if a source from the late eighth century would have survived until the fifth or fourth centuries without being mentioned in any late-monarchic biblical source. Second, most names of gates, towers and pools in the list do not correspond to the many such names in late-monarchic biblical texts.

The third option would put Nehemiah 3 with what scholars see as late redactions in Ezra and Nehemiah, which can be dated as late as the Hasmonean period (Williamson 1985: xxxv; Wright 2004, 2007). Bohler (1997: 382-97) explicitly put the rebuilding of Jerusalem story in Nehe­miah on Hasmonean background. The usage of words such as the province Beyond the River (mm Neh. 3.7), pelekh and nns (Neh. 3.11) does not present difficulty for such a late dating, as they appear in late Jewish sources (for "IHDH see 1 Macc. 7.8 — Rappaport 2004: 281; for pelekh in the rabbinical literature [without entering the discussion on the mean­ing of the word —Demsky 1983; Weinfeld 2000; Edelman 2005: 213-14], see Kohut 1926: 346; Demsky 1983: 243; for nns, see Dan. 3.27).

Dating the insertion of this text to the Hasmonean period may corre­spond to the importance given to the figure of Nehemiah in the first two chapters of 2 Maccabees (as the builder of the Temple!), which Bergren (1997; also 1998) interpreted as an attempt to bolster the figure of Judas Maccabeus, the hero of 2 Maccabees, by comparing him to Nehemiah—a prominent figure in the restoration, a builder, a political leader, a zealot for the law and a paradigm of piety (1997: 261-62). Nehemiah could have been chosen as such a model for the Hasmoneans because he repre­sented a non-Davidide, non-Zadokite leadership.

Clues that Nehemiah 3 does not reflect Persian-period realities may be found in the archaeology of two of the three well-identified and excavated (rather than surveyed) sites mentioned in the list — Beth-zur and Gibeon.

The archaeology of Beth-zur (Neh. 3.16) in the Persian period has been debated. Funk (1993: 261), Paul and Nancy Lapp (1968: 70; P. Lapp 1968: 29) and Carter (1999: 157) argued that the site was very sparsely, in fact, insignificantly inhabited in the Persian and Early Hellenistic periods. Funk noted that the 'interpretation of the Persian-Hellenistic remains at Beth-zur is dependent in large measure on the extant literary references (1968: 9). Based on a single locus(!), Stern (2001: 437-38; see also 1982: 36) adhered to the notion of a significant activity at the site in the Persian period. Reich (1992) argued in the same line according to an architectural analysis. The published material from the excavations (Sellers 1933; Sellers et al. 1968) includes only a limited number of finds — sherds, vessels and coins — that can safely be dated to the Persian period (Stern 2001: 437), while most forms belonging to the Persian-period repertoire are missing altogether. Hence, though archaeology may have revealed traces of some Persian-period activity at the site, it is clear that it was an important place only in the Late Iron II and the Late Hellenistic periods. It should be noted that Beth-zur — supposedly the headquarters of half a district in the province of Yehud — did not yield even a single Yehud seal impression (over 530 have so far been recorded — Lipschits and Vanderhooft 2007: 3).

Gibeon (Neh. 3.7) did not yield unambiguous Persian-period finds either. Without going into the debate over the dating of the Gibeon winery and inscriptions—late monarchic or sixth century (see summaries in Stern 1982: 32-33; 2001: 433; Lipschits 1999: 287-91) — the mwsh seal impressions and wedge-shaped and reed impressed sherds found at the site (Pritchard 1964: Figs. 32.7, 48.17) attest to a certain activity in the Babylonian or Babylonian/early Persian period. Yet, typical Persian- period pottery and Yehud seal impressions were not found (for the latter, see Lipschits 2005: 180). Late Hellenistic pottery and coins are attested. According to Pritchard, there is 'only scant evidence of occupation from the end of the sixth century until the beginning of the first century BCE' at Gibeon (1993: 513). Still, in an attempt to provide evidence for the Gibeon of Neh. 3.7, he argued that 'scattered and sporadic settlements' did exist there during the Persian and Hellenistic periods (Pritchard 1962: 163). Stern rightly interpreted the Gibeon finds as evidence for only sixth- century and possibly early Persian-period activity at the site (1982: 32-33; 2001: 433; Lipschits 2005: 243-45 — sixth century).

Yet, there are several problems regarding the Hasmonean option for the background of Nehemiah 3. First, the toponyms in the description of the First Wall in Josephus's Wars 5.4.2 — especially the 'gate of the Essens' (as well as names of gates mentioned by Josephus elsewhere) — are different from the toponyms in Nehemiah 3. But the change may be assigned to post-Hasmonean, mainly Herodian times. A more severe problem is the prominence of the story on the construction of the city-wall throughout the Nehemiah Memoirs. Accepting a Hasmonean reality behind the city-wall account in Nehemiah would therefore call for a drastic new approach to the entire book of Nehemiah (on this see Bohler 1997).

Conclusion

The Persian-period finds in Jerusalem and the search for Nehemiah's wall are additional cases in which archaeologists have given up archaeology in favor of an uncritical reading of the biblical text. The dearth of archaeological finds and the lack of extra-biblical texts on Persian-period Yehud open the way to circular reasoning in reconstructing the history of this period.

The finds indicate that in the Persian and Early Hellenistic periods Jerusalem was a small unfortified village that stretched over an area of c. 20 dunams, with a population of a few hundred people — that is, not much more than 100 adult men. This population — and the depleted population of the Jerusalem countryside in particular and the entire territory of Yehud in general — could not have supported a major reconstruction effort of the ruined Iron II fortifications of the city. In addition, there is no archaeological evidence whatsoever for any reconstruction or renovation of fortifications in the Persian period. Taking these data into consideration, there are three ways to explain Nehemiah 3: (1) that it is a utopian list; (2) that it preserves a memory of an Iron Age construction or renovation of the city-wall; (3) that the list is influenced by the construction of the First Wall in the Hasmonean period. All three options pose significant difficulties — the first two more than the third. In any event, the archaeology of Jerusalem in the Persian period — as presented above — must be the starting point for any future discussion.

On a broader issue, the archaeological evidence from Jerusalem casts severe doubt on the notion that much of the biblical material was com­posed in the Persian and Early Hellenistic periods. But this crucial issue is beyond the scope of this study and will be discussed elsewhere.