Locating Jerusalem’s Royal Palace in the Second Millennium BCE in Light of the Glyptic and Cuneiform Material Unearthed in the Ophel

Abstract

This article argues, on the basis of indirect evidence drawn from the results of the excavations of the Ophel, that during the second millennium BCE, the governing centre of Jerusalem was located on the Temple Mount. The conclusion rests mainly upon a numerical comparison between the glyptic material uncovered in the Ophel vis-à-vis that unearthed in the Southeastern Hill (the City of David) and upon the discovery of two fragmented cuneiform tablets in this area. It is postulated that the findings uncovered in the Ophel might serve as a litmus test for the early urban life in the Temple Mount above it. The city at the time included two distinct quarters: the Temple Mount, the seat of the king and his court, and the Southeastern Hill, consisting of the summit of the hill and the Gihon Spring. This two-part division of the city persisted from its foundation in the MB II down to the early first millennium BCE and continued until the 8th century BCE.

Keywords:

• Temple Mount
• Ophel
• Southeastern Hill
• Gihon Spring
• Scarabs
• Bullae
• Fragmented tablets

Introduction

Two fragments of cuneiform clay tablets (‘Jerusalem 1’ and ‘Jerusalem 2’) have been discovered in the wet sifting of the earth assembled in the excavations of the Ophel, the area located at the foot of the Temple Mount.¹ In what follows, I first analyse these fragments, each in its own right, and examine their significance for locating Jerusalem’s royal palace whence they must have been swept. I then proceed to discuss the assemblage of glyptic material discovered in the wet sifting of the earth assembled in the Ophel excavations and to evaluate its contribution to the debate regarding the location of Jerusalem’s centre of government in the second millennium BCE. Finally, I present the overall picture of second-millennium Jerusalem, emphasising the difference between its two main quarters—the Temple Mount and the Southeastern Hill—and the way the two-part city was administered by the king and his officials.

The cuneiform fragments uncovered in the Ophel

The ‘Jerusalem 1’ fragment

The fragmented cuneiform tablet unearthed in the Ophel excavations in the 2009 season contains the beginning of five lines on the obverse and four on the reverse. In light of the similarity of its script to that of the Jerusalem Amarna letters (EA 285–291), the publishers suggested that it is roughly contemporaneous to the latter and should be dated to the 14th century BCE (Mazar et al. 2010: 6–11). They further posited that, like the Amarna tablets, the fragmented tablet is a letter. Provenance study of the clay established that the tablet is made of terra rosa soil, common in the central hill country (ibid.: 12–18). This soil is quite widespread in the Jerusalem anticline, and although all five Amarna letters that were sent from Jerusalem are derived from the Moza and ʻAmminadav formation (Goren, Finkelstein and Na'aman 2004: 265–269),² the publishers nevertheless suggested that the tablet was locally made in the Jerusalem region (Mazar et al. 2010: 18).

As the fragmented tablet was discovered in Jerusalem, they suggested that it is an archival copy of the original letter sent by the king of Jerusalem to the Pharaoh (Mazar et al. 2010: 11–12, 18). They posited that the quality of writing indicates that the copy faithfully reflects the original letter, as it was written in a ductus appropriate for international correspondence (ibid.: 12).

Shortly after the publication of that article, Christopher Rollston (2010) criticised its conclusions, claiming that the small number of cuneiform signs that survived in the tablet is inadequate for establishing its genre. Examining the text in detail, he emphasized the uncertain nature of the interpretation of this small fragment as a letter (ibid.: 13–17), and suggested that the tablet could have been a letter, a legal document, or an administrative tablet. In this light, he concluded that ‘because there is such a dearth of preserved text on this tablet, it is best not to attempt to posit a historical context or even a genre’ (ibid.: 20).

Rollston’s warning to be cautious in the interpretation of the fragmented text is certainly warranted. Crucial to the discussion, however, is the question where the tablet was produced. The publishers assumed that it was written in Jerusalem and hence concluded that it is a copy of a letter. But is the local-origin hypothesis the most reasonable option? First, terra rosa is widespread throughout the central hill country. Second, whereas the five Amarna letters from Jerusalem were made of clay of the Moza and ʻAmminadav formations, one of the letters of Lab'ayu, ruler of Shechem (EA 253), contains a mixture of rendzina and terra rosa (Goren, Finkelstein and Na'aman 2004: 264). Third, Seth Sanders (2016) has demonstrated that the writer of the Jerusalem fragment was not the same as that of the Jerusalem Amarna letters on the grounds that he switched the common initial e- of verbs in the Amarna letters to i-, as indicated by the form i-pé-ša (obv. line 4). This observation suggests that the tablet was written in a city other than Jerusalem.

Finally, the publishers’ hypothesis that the fragmented tablet is a copy of an original letter sent to the Egyptian court is quite unlikely. Hundreds of tablets were sent in the Late Bronze Age from Canaan to the Egyptian court, but not a single copy was discovered in the many excavations of the major Canaanite sites of this period. The cuneiform tablets from this period unearthed to date in Canaan include internal correspondence between city-states, economic and administrative tablets, and a few scholarly tablets.³ Copies of letters that served as teaching models were discovered in Amarna (Na'aman 2002: 80–81) and possibly in the Egyptian administrative centre of Aphek (Na'aman and Goren 2009: 463–465), but not in Canaanite city-states.

Scholars might respond by claiming that no letters sent from the Egyptian court to Canaan were discovered either, except for a single fragment recently unearthed in the Ophel excavation (‘Jerusalem 2’, see below). Yet the number of letters that the royal court sent to the rulers of Canaan was far less than the number sent by the vassals to the Egyptian court. If we examine the Amarna archive, most of the references to the receipt of royal letters appear in texts referring to the organisation of the Egyptian campaign to Canaan (Na'aman 1990). At that time, letters were sent to many rulers, and their reception is attested in their response to the Pharaoh (EA 141:8–9, 142:6–7, 192:10–11, 193:22–24, 207:4–5, 292:17–19, 294:6–9, 304:15–16, 305:15–17 and 362:5–6). Otherwise, the number of references to the reception of royal letters is minimal (EA 100:24–25, 243:8–9, 253:7–9 and 254:6–7). Most of the Egyptian responses must have been verbal, made by the Egyptian officials who served in Canaan.

A unique source for investigating this problem is Papyrus Anastasi III, which records the passage of messengers, some bearing letters, from Egypt to Canaan and vice versa (Caminos 1954: 108–113; Wilson 1969: 258b–259a; Helck 1971: 231). Among the references to the dispatch of letters from Egypt to Canaan, only one missive was sent to the King of Tyre; the rest were sent to Egyptian officials who served in Canaan.

Evidently, far fewer letters were sent from the Egyptian court to Canaan than from the Canaanite vassals to Egypt. The fact that not a single copy of a letter sent from Canaan to Egypt was ever discovered raises doubts regarding the hypothesis that the Jerusalem 1 fragment was from a copy of a letter. Instead, the tablet must have arrived from another place, most likely the city of Shechem, whose soil fits well the petrographic composition of the fragmented tablet. This would undermine Rollston’s claim that the genre of the tablet is unknown—if the tablet was sent to Jerusalem, it must have been a letter.

In sum, the assumption that the tablet is a letter dispatched from Shechem to Jerusalem is most in keeping with the available evidence—namely, the composition of the clay, which differs from that of the Jerusalem Amarna letters, and the initial i- of the verb epēšu (ipēšu), as opposed to the initial e- of verbs in the Jerusalem Amarna letters.

The ‘Jerusalem 2’ fragment

The second fragment discovered in the Ophel excavations in 2013 contains only four cuneiform signs. The importance of the fragment is revealed by its provenance from ‘typical Egyptian Nile sediments as known from the vast literature’ (Mazar et al. 2014: 133). It indicates that the fragment is part of a royal letter sent from the Egyptian royal court to the King of Jerusalem (ibid.: 138). As noted above, this is the first time that a fragment of a letter sent by the royal court of Egypt has been discovered in Canaan.

Line 2 of the obverse is transcribed [… U]G-GU[R], which the publishers rendered as uqqur and interpreted as the D stative from aqāru, ‘to be scarce, expensive, precious’ (Mazar et al. 2014: 135). However, an examination of the Amarna letters sent by the Pharaoh to his Canaanite vassals shows the most prominent characteristic of the texts to be their commanding tone (see EA 367, 369–370), suggesting that the verbal form uqqur should be interpreted as an imperative derived from the verb naqāru, in the sense of ‘demolish, raze, tear down’. What the Pharaoh ordered the King of Jerusalem to destroy remains unknown.

How are we to explain the discovery of the two fragmented letters—one probably sent from Shechem and the other from Egypt—in the Ophel area? The summit of the Southeastern Hill, where scholars usually located ʻAbdi-Ḫeba’s royal palace, is far south from the Ophel, and it is unlikely that letters received at this remote summit would have reached there. It is more likely that the tablets were swept from the Temple Mount, located above the Ophel. Consequently, it is suggested that the palace was situated on the Tenple Mount.⁴ This conclusion leads me to discuss the controversial issue of the ancient location of Jerusalem’s royal palace, a topic that has recently come under fierce scholarly debate.

The debate over the location of early Jerusalem’s royal palace

For many years, scholars assumed that since the city’s foundation in the early second millennium BCE and until the early monarchic period, Jerusalem’s governing centre was on the summit of the Southeastern Hill.⁵ According to this assumption, the hill’s proximity to the Gihon Spring, the main source of water in this region, well explains the development of the urban centre near the spring. According to the conventional reconstruction, the transfer of the governing centres—namely, the palace and the temple—from the Southeastern Hill to the Temple Mount took place under King Solomon around the mid-10th century BCE. This assumption is based, of course, on the biblical histories of David and Solomon.

The discovery of Middle Bronze Age fortifications and extensive building activity on the Southeastern Hill and near the Gihon Spring became the main pillars of the claim that the city was first established in these areas.⁶ The paucity of remains from the Late Bronze Age in these areas, in contrast to the evidence of the Amarna letters, was explained by the fragmentary quality of the buildings constructed in these periods. According to this explanation, these fragile buildings were destroyed and dispersed by the extensive building operations conducted later in these areas, which left only scanty remains on the ground (Na'aman 1992; 2011: 165–173; De Groot 2012: 149–154).

The discovery of the two tablet fragments in the Ophel area has changed the balance of evidence. As observed above, the fragments indicate that the royal palace, from where they must have swept, was probably on the Temple Mount. This would account for the scanty architectural remains and the paucity of objects dated to the Late Bronze Age uncovered in the excavations conducted in the Southeastern Hill. According to this logic, the centre of the city at the time was located on the Temple Mount, north of the Southeastern Hill, with the latter having been a peripheral, poorly inhabited area in its vicinity.

This hypothesis brings to the fore the question when the royal palace was first built on the Temple Mount. In a groundbreaking article, Axel Knauf (2000: 75–81, 87) posited that the core area of Jerusalem was always on the Temple Mount, whereas the Southeastern Hill was peripheral and was settled only in periods of great economic activity and prosperity. The selection of the Temple Mount for the seat of the royal house was mainly strategic, as it is raised and naturally defended on all sides and it controls the surrounding regions and dominates the lower hill to its south. Conversely, the Southeastern Hill was always vulnerable to the possibility of an enemy occupying the commanding mount to its north, so that the occupation of the latter was necessary for defending the former. In light of these considerations, Knauf posited that from the foundation of the city onward, the Temple Mount was the centre of the city, whereas the Southeastern Hill was a secondary quarter under its domination.

With this postulate in mind, Knauf examined the archaeological evidence of the Late Bronze and Iron Ages and concluded that ʻAbdi-Khepa’s and David’s Jerusalem lies buried under the Herodian-through-Islamic structure of the Temple Mount’ (Knauf 2000: 87). Surprisingly, he limited his discussion to only the Late Bronze Age and the early Iron Age, ignoring the prominent Middle Bronze Age remains unearthed in the Southeastern Hill, which ostensibly run counter to his postulate.⁷

Knauf ‘s hypothesis was further developed by Finkelstein, Koch and Lipschits (2011), who thoroughly examined the results of the archaeological excavations held on the Southeastern Hill and—like Knauf—posited that this part of the city was settled only in periods of growth of the urban culture and economic prosperity. In contrast, in periods of decline, such as the Late Bronze Age, the Iron I–Early Iron IIA and the Persian and Early Hellenistic periods, the southern quarter of the city was poorly settled.

In support of their appraisal of Jerusalem’s urban development, they observed that like Jerusalem, other second-millennium BCE cities in the central hill country (i.e., Shechem, Bethel and Hebron) were also located on projected mounds. They calculated the assumed measurements of the Temple Mount and suggested that its area fits the scope of the other major highlands cities (Finkelstein, Koch and Lipschits 2011: 7–9). In this light, they concluded that ‘during the second millennium and early first millennium B.C.E.—until the great territorial expansion of Jerusalem in the Iron IIB … Jerusalem had been located on a mound which was later leveled and boxed-in under the Herodian platform’ (ibid.: 9).

Their discussion of the archaeological remains in the Southeastern Hill did not, however, explain how the Temple Mount was connected to the fortified Middle Bronze Age quarter constructed in the Southeastern Hill. In fact, they admitted that ‘the nature of at least some of the Middle Bronze remains in the summit of the City of David, as well as the extent of the Middle Bronze city, remain a riddle’ (Finkelstein, Koch and Lipschits 2001: 11).

As the ‘mound on the mount’ hypothesis is wholly conjectural and lacks concrete evidence, it soon encountered severe criticism. Geva and De Groot (2017) wrote a detailed rejoinder, its conclusion given in the title: ‘The City of David is not on the Temple Mount after all’. In this article, they dismissed the idea that Jerusalem was first established on the Temple Mount and supported the classical hypothesis according to which the city was first built in the Southeastern Hill, near the Gihon Spring, the city’s main source of water. Geva and De Groot did not discuss the circumstances under which the governing centre was transferred from the Southeastern Hill to the Temple Mount and only suggested that the transfer did not take place before the first millennium BCE.

The arguments put forward by Geva and De Groot in support of their thesis that in the early second millennium BCE the city centre was located in the Southeastern Hill might be summarised as follows:

1. They first examined the archaeological data collected in the surveys of the Temple Mount and its eastern slope (Geva and De Groot 2017: 34–39). Although the Temple Mount could not be excavated archaeologically, sporadic works held on its summit and eastern slope supplied rich data that sheds light on its history. After examining all the available data, Geva and De Groot concluded that although many small digs, surveys and sifting projects were conducted on the Temple Mount and its eastern slope, Middle Bronze Age pottery has not been found there.⁸ This conclusion is significant, since the Middle Bronze Age was a period of flourishment, during which settlement expanded in Jerusalem and the nearby highlands. Ostensibly, the absence of Middle Bronze Age pottery on the Temple Mount runs counter to the suggestion that it was the centre of the city in this period.

2. Geva and De Groot (2017: 39–42) demonstrated that the wide north–south MB II wall (2–4 m in width), located on top of the eastern slope of the Southeastern Hill, was part of the fortification that surrounded the hill and was constructed to defend the city (for details, see Keel 2007: 82–85, with earlier literature; De Groot 2012: 144–149; cf. Maeir 2017). The lower wall of this fortification passed in the middle of the eastern slope, and the city’s northern wall was probably located in Area G or north of it (De Groot 2012: 147–148). A defended corridor, composed of two broad walls (W108 and W109), connected the spring to the Southeastern Hill. Recently, Reich (2019: 512–514) posited that the northern and wider of these two walls (W108; 3.5–3.7 m in width) formed the northern wall of the Middle Bronze Age city and must have joined the city wall constructed on the summit of the Southeastern Hill. The fortified corridor guaranteed the supply of water in times of siege.

Excavations held in the area north of Wall 108 indicate that no similar corridor had led from the spring to the Temple Mount. Evidently, the spring was an integral part of the Middle Bronze Age enclosure that encompassed the Southeastern Hill and was detached from the Temple Mount located north of it.

1. The antiquity of the Spring Tower constructed to defend the Gihon Spring is hotly debated among scholars. In the past, scholars had assumed that the tower was first fortified in the MB II and renovated in the Iron II (Reich and Shukron 2010; De Groot 2012: 147; Uziel, Shukron and Szanton 2013). However, recent radiocarbon dates extracted from the lowest stones of the Spring Tower’s northeastern corner indicate that it was fortified no earlier than the 9th century BCE (Regev et al. 2017).⁹ Moreover, recent excavations below the lowest course of the tower’s eastern wall exposed Middle Bronze Age remains that are dated to the 19th–18th centuries BCE. Furthermore, radiocarbon dates extracted from three sites in the Southeastern Hill, the Gihon Tower included, indicate that the site suffered great decline in the 17th century BCE. Hence, the tower could not have been constructed before the 16th century BCE (Regev et al. 2021). Evidently, the Gihon Spring had remained unfortified in the Middle Bronze Age and was fortified for the first time in the late 9th century BCE.

In sum, Geva and De Groot’s claims that the Middle Bronze Age city on the Southeastern Hill was fortified and that the Gihon Spring—albeit not fortified—was an integral part of it are well founded. Notably, however, no Middle Bronze Age public building was unearthed in the Southeastern Hill excavations. Although one might argue that absence of evidence is not evidence of absence, the lack of a large public building on the hill might support the conclusion that it was established elsewhere.

New arguments in support of the ‘mound on the mount’ hypothesis

As noted above, hardly any remains of the Middle Bronze Age were discovered on the eastern slope of the Temple Mount. Investigation of the southern slope, however, reveals a different picture. I have already mentioned the discovery of the two fragmented cuneiform tablets in the excavation of the Ophel and suggested that their origin was on the mount located above it. No less significant, however, is the discovery in the Ophel excavations of dozens of bullae and scarabs bearing Egyptian iconographic motifs. Othmar Keel (2015a: 475–529; 2017: 458–499, Nos. 400–493) published this collection of Egyptianised artefacts, and I draw on his analysis to evaluate the historical significance of this material.

Ninety-four artefacts with Egyptian iconographic motifs were discovered in the Ophel excavations. Among them, 29 scarabs and fragmented bullae are from the Late Middle Kingdom and the Second Intermediate Period (the first half of the second millennium BCE). This assemblage includes four scarabs (Keel 2015a: 477–479, Nos. 5–6; 488–489, Nos. 27–28) and 25 fragmented bullae. The bullae had originally sealed various kinds of artefacts, and when these artefacts were opened, their seals were torn and the bullae disposed of. Hence, the bullae reflect economic or administrative activity that took place either in the Ophel or on the Temple Mount, whence they might have been swept to the Ophel area.

Unfortunately, the bullae were not analysed petrographically, and their place of sealing has not been established. They might have been made from either Egyptian or local Canaanite scarabs. Based on their iconographic style, Keel (2015a: 525) suggested that a few of them originated in Egypt and that the rest were of local Levantine manufacture. As is well known, many of the locally made Canaanite scarabs imitate the Egyptian originals and were used in Canaan in the second millennium BCE. In this period, relations between Egypt and Canaan had intensified, and the Canaanites adopted motifs of Egyptian iconography and incorporated them into their culture. This is reflected in the large-scale local production of scarabs and their widespread use in Canaan (for details, see Ben-Tor 1997; 2003; 2007: 185–192; 2009).

In this light, it is possible that a few of the scarabs and many of the bullae designed in the style of the Late Middle Kingdom and the Second Intermediate Period, which were discovered in the Ophel, were produced in the first half of the second millennium BCE. As the Ophel area was first occupied in the Early Iron IIA,¹⁰ these iconographic artefacts must have arrived from nearby. The Southeastern Hill was located far south of the Ophel and could not have been the source of the glyptic material. Rather, these artefacts must have swept from the Temple Mount, located above the Ophel. Thus, those bullae and scarabs that belong to the first half of the second millennium are the only remnants of the extensive economic and administrative operations that took place in the public buildings on the Temple Mount.

In addition to the Middle Bronze Age glyptic material, a few scarabs and bullae dated to the Eighteenth and Nineteenth Egyptian Dynasties were also uncovered in the Ophel (Keel 2015a: 491–493, Nos. 31–35). It remains unknown whether they were carried to Jerusalem in the early Iron Age or earlier, in the Late Bronze Age. If the latter possibility is viable, they must have joined the two fragmented Late Bronze Age cuneiform tablets unearthed in the Ophel.

For further evaluation of the place of the Temple Mount in the economy and administration of ancient Jerusalem, the group of Egyptianised iconographic artefacts unearthed in the Ophel excavations should be compared to those unearthed in the excavations of the Southeastern Hill.

1. Keel (2015b; 2017: 438–449, Nos. 355–377) published the assemblage of 24 scarabs and bullae unearthed on the ‘Summit of the City of David’. These include nine bullae with late Middle Kingdom impressions (Keel 2015b: 422–430, Nos. 5–13) and one scarab dated to Amenhotep III (ibid.: No. 1).

2. Keel (2012: 317–323; 2017: 340–426) also published the assemblage of 134 scarabs and bullae unearthed in the ‘Rock-cut Pool’ near the Gihon Spring, including 13 bullae and scarabs of the Middle Bronze II (Nos. 138–150) and eight of the Late Bronze Age (Nos. 151–158).

3. A third assemblage published by Keel (2017: 500–509, Nos. 496–518) was excavated in the area south of the Gihon Spring. It includes five bullae dated to the Middle Bronze Age (Nos. 496–500) and three to the Late Bronze Age (Nos. 501–503).

4. In addition, two scarabs, an impression on a handle and a possible stamped conoid seal¹¹ were unearthed in the excavations of Area E of the City of David (Brandl 2012: 379–382, 388, 390; Keel 2017: 300, Nos. 48–49; 324, No. 107).¹² Another bulla was unearthed in the excavations of the Southeastern Hill, bearing a private-name impression (Ben-Dor Evian et al. 2019).

Evidently, the number of second-millennium BCE scarabs and bullae unearthed in the area near the Temple Mount is greater than the number of those unearthed in the excavations in the Southeastern Hill. When one takes into account that the artefacts unearthed in the Ophel area are only a segment of the overall number of artefacts originally located on the Temple Mount, they indicate the centrality of the Temple Mount in the economy and administration of Jerusalem at that time.

In addition to the scarabs and bullae uncovered in the Ophel excavations, other imported artefacts have been discovered in the area. In light of the date of occupation of this quarter, the scholars who published these artefacts dated them no earlier than the Early Iron IIA (Winderbaum 2015; Lang 2015). However, some of these artefacts might have been imported to the city earlier and later have swept from the Temple Mount to the Ophel. Among these artefacts are a few figurines and amulets (Winderbaum 2015: 531–532, 534–535, 537), a dagger (Lang 2015) and a few ivories (Naeh 2015: 581–584). They might be included in the list of second-millennium BCE artefacts that were swept to the Ophel, which together reflect the economy and culture of Jerusalem’s royal palace and temple in the second millennium BCE.

The two quarters of Jerusalem in the second millennium BCE

The picture that emerges from the discussion is perplexing. On the one hand, the Temple Mount, where the royal palace and temple were located, was probably the economic and administrative centre of the city as early as the second millennium BCE. On the other hand, the fortifications that encompassed the Southeastern Hill formed a separate urban entity, detached from the Temple Mount. No architectural remains dated to the Middle or Late Bronze Ages were discovered in the excavations of the Ophel quarter, near the Temple Mount, and no corridor connecting the two parts of the city was uncovered in the excavations of the Southeastern Hill. It is possible, of course, that such a corridor was built in the Middle Bronze Age but was razed to the ground by later building activity. This possibility, though logical, cannot, however, be verified. Hence, the way the palace officials—whose seat was on the Temple Mount—administered the lower city in the second millennium BCE remains unclear.

It seems that the natural topography of the Jerusalem region and the location of the spring dictated the development of the city. As it was difficult to build fortifications that encompassed the city’s two discrete quarters, the royal authorities decided to treat each quarter as a separate unit. The northern part of the city was built on the high mound, enjoyed natural defence and did not require massive fortifications.¹³ The lower city to its south, in contrast, required stronger fortifications, which were constructed at a certain stage in the Middle Bronze Age.

The above discussion suggests that the findings uncovered in the Ophel, which antedated the construction of buildings in its area in the Early Iron IIA, might serve as a litmus test for the early urban life that took place in the Temple Mount above it. We may further suggest that the public buildings erected on the mound in the Middle Bronze Age were located above the Ophel area, so that small artefacts were swept on this slope and discovered in the Ophel excavations.

There is a remarkable difference in the number of glyptic artefacts unearthed in the Ophel excavations that were dated to the Middle Bronze Age and in the number dated to the Late Bronze Age. This difference reflects the disparity in the intensity of the economic activity that took place in the two periods. Of course, this comes as no surprise—it is in keeping with all that we know from the archaeological research of the Jerusalem area and the central hill country regarding the difference in settlement intensity between the first and second halves of the second millennium BCE.

No further details regarding the reality on the Temple Mount in the Middle Bronze Age are available; the glyptic material and a few other artefacts uncovered in the Ophel excavations are the only clue to that reality. The situation in the Late Bronze Age palace is better documented due to the Amarna letters and the two fragmented letters unearthed in the Ophel excavations. In fact, one of ʻAbdi-Ḫeba’s letters includes a passage that refers directly to the royal palace. In this letter (EA 287), the King of Jerusalem complains that the Egyptian garrison, consisting of Nubian (‘Kushite’) soldiers, which was established at that time in the city, broke by force into his palace, threatened his life and took the products (probably food, oil and cloth) that he refused to give them.

Two passages relate this episode. The first (lines 33–42) is slightly damaged, but might be reconstructed as follows:¹⁴

With regard to the Kushites, may the king make inquiry of the commissioners whether the house (i.e., the palace) is well fortified. And they attempted (to do) a crime. They (the Kushites) took their tools and broke (bat-qú-ú)¹⁵ the support (e-med) of the roof (ga-ag-gi).

The second passage (lines 71–75) is shorter and is directed to the Egyptian royal scribe:

And please make the Kushites responsible for the evil deed. I was almost killed by the Kushites in my own house.

The letter indicates that in the 14th century BCE (or possibly earlier), a strong palace was constructed in Jerusalem and that it served as the seat of the king and his officials. The two fragmented letters and the few scarabs and bullae dated to the Eighteenth and Nineteenth Egyptian Dynasties unearthed in the Ophel excavations must have swept from this palace, apparently located on the Temple Mount.

About ten imported artefacts dated to the late 11th–10th centuries BCE have been discovered in the Ophel excavations (Keel 2015a: 494–499, Nos. 36–45). Among them are a frog scaraboid, two scarabs, two bullae and various other objects, all (or most) of which are of Egyptian origin.¹⁶ No similar glyptic artefacts dated to the Late Bronze or early Iron Age have been discovered on the Summit of the Southeastern Hill. As the Ophel was first constructed later, in the Early Iron IIA, these findings, which swept from the Temple Mount, indicate that the mount retained its central economic and administrative role throughout the 11th–10th centuries BCE. (2 Sam 5:7) was located at the site of the Late Bronze Age palace. Moreover, the building of the palace and temple on the Temple Mount, as related in the history of Solomon, was not a revolutionary step in the city’s history (as was postulated by generations of scholars), but rather, constituted a natural continuity in the history of the city, in which the palace, the seat of the ruling dynasty, and the temple were always located on the Temple Mount.

In conclusion, during the second millennium BCE the governing centre of Jerusalem was located on the Temple Mount. The evidence supporting this conclusion is indirect, drawn from the results of the excavations of the Ophel. It rests mainly upon a numerical comparison between the glyptic material uncovered in this area and that unearthed on the Southeastern Hill, as well as upon the discovery of the two fragmented cuneiform tablets in the Ophel area. At the time, the city had included two distinct quarters: 1) the Temple Mount, the seat of the king and his court, and 2) the Southeastern Hill, consisting of the summit of the hill and the Gihon Spring. The way the city operated as an urban unit remains unresolved, as the architectural elements linking the two regions have not yet been discovered. This two-part division of the city persisted through the second millennium BCE, from the Middle Bronze II down to the early first millennium BCE and throughout the monarchic period.

Contributors

Nadav Na'aman: Professor Emeritus, Department of Jewish History, Tel Aviv University; email: [email protected]; ORCID: https://orcid.org/0000-0001-8153-6197

Disclosure Statement

The author reports that there are no competing interests to declare.

Additional information

Notes on contributors

Nadav Na'aman

Nadav Na'aman: Department of Jewish History, Tel Aviv University

Notes

1 For the excavations of the Ophel, see Mazar 2015; 2018.

2 Letter EA 285 was sent from the central Jordan Valley and EA 291 from the area of Gezer.

3 For a compilation of almost all the tablets discovered in Canaan, see Horowitz, Oshima and Sanders 2006.

4 Similarly, a fragment of an Akkadian cuneiform tablet that relates the epic of Gilgamesh was discovered below the large dump of the Chicago excavations at Megiddo and must have been swept from one of the palace rooms located on the mound (see Horowitz, Oshima and Sanders 2006: 102–105, with earlier literature). Ussishkin’s hypothesis (2021: 477–478) that the tablet belonged to some functionary attached to the Canaanite army that fought against the army of Thutmose III ca. 1475 BCE and that he dropped it in the battlefield is highly unlikely, as it stands in contrast to all that we know of the maintenance of literary tablets in the Ancient Near East.

5 Some scholars posited that Solomon’s temple was built on top of an earlier Canaanite/Jebusite temple (see Gressmann 1921: 148–149; Rupprecht 1977: 1–4, 13–17, 100–103; Gerhards 2008: 356–357). However, they did not develop this hypothesis beyond the limited scope of the story of David and Araunah the Jebusite (2 Sam 24:18–25).

6 In light of the results of the recent archaeological excavations in the Southeastern Hill, I withdraw my earlier suggestion that Jerusalem was not mentioned in the two groups of Egyptian Execration Texts (Na'aman 1992: 278–279). The toponym Ꜣwšmm should be transcribed ru-š[a]l-m-m, i.e., (Je)rusalem (with mimation), as rendered by the majority of scholars.

7 For further support for Knauf’s hypothesis, see Gerhards 2008: 356 n. 43.

8 For details, see, e.g., Barkay and Zweig 2006; 2007; Zweig 2008; Dvira, Sigdon and Shilov 2011; Barkay and Dvira 2012; Baruch, Reich and Sandhaus 2018.

9 For criticism of this suggestion, see Reich 2018; Finkelstein 2018: 191–192; Mazar 2020: 147.

10 For detailed discussions of the date of the first occupation of the Ophel and its gradual expansion, see Winderbaum 2022; Finkelstein 2022, both with earlier literature.

11 Brandl (2012: 378–379, No. 2) dated the stamped conoid seal to the Early Bronze Age, whereas Keel (2017: 324, No. 107) dated it to the Middle Bronze Age.

12 For the excavations of Area E, located on the summit of the Southeastern Hill, see De Groot and Bernick-Greenberg 2012.

13 It remains unknown whether fortifications were built around the Temple Mount in the second millennium BCE.

14 For slightly different translations of the text, see Moran 1992: 328–329; Liverani 1998: 92–93; Rainey 2015: 112–113.

15 Vita (1999: 621, note on line 36), who examined the original tablet, observed that the first sign of the verb is BE. Hence, it is preferable to render the verbal form bat-qú-ú (‘broke’) (see also Liverani 1998: 92 and n. 106). Moran (1992: 329 n. 9) restored it áš-ru-ú and translated ‘and I had to seek shelter by a support of the roof’ (similarly to Yoder and Lauinger 2022). Rainey’s transcription of the verb (páṭ-ru-ú) and his translation (‘they took their implements and mutinied to [se]ize the roof’) (2015: 112–113, 1593) are, in my opinion, untenable.

16 Remarkable is a conoid of the Cypro-Geometric I period (Keel 2015a: 496, No. 40).