• Fig. 3.2 Plan of the
  Figure 3.2
  
  Plan of the Stratum XIX Temple
  
  (after Loud 1948: Fig. 390; Courtesy of The Oriental Institute of The University of Chicago).
  
  Finkelstein et al. (2000)
  Stratum XIX Temple from Finkelstein et al. (2000)
• Fig. 3.1 Aerial view of
  Figure 3.1
  
  Aerial view of Area J at the end of the 1996 season; looking east.
  
  Finkelstein et al. (2000)
  f Area J at the end of the 1996 season from Finkelstein et al. (2000)
• Fig. 31.3D - Fractured
  Figure 31.3D
  
  In Area J, the monumental walls of the Level J-4 temple are fractured in several places along their strike (Fig. 31.3d) as well as perpendicular to the strike (Figs. 31.3e-f). The overlying walls of the EB III temple 4050 are not fractured
  
  Marco et. al. (2006)
  Temple Walls from Marco et. al. (2006)
• Fig. 31.3E - Fractured
  Figure 31.3E
  
  In Area J, the monumental walls of the Level J-4 temple are fractured in several places along their strike (Fig. 31.3d) as well as perpendicular to the strike (Figs. 31.3e-f). The overlying walls of the EB III temple 4050 are not fractured
  
  Marco et. al. (2006)
  Temple Walls from Marco et. al. (2006)
• Fig. 31.3F - Fractured
  Figure 31.3F
  
  In Area J, the monumental walls of the Level J-4 temple are fractured in several places along their strike (Fig. 31.3d) as well as perpendicular to the strike (Figs. 31.3e-f). The overlying walls of the EB III temple 4050 are not fractured
  
  Marco et. al. (2006)
  Temple Walls from Marco et. al. (2006)
• Fig. 40.1 - Extension
  Figure 40.1
  
  Wall 96/J/1 in Square H-10, looking east. Note fracture along the strike on the edge of the wall.
  
  Finkelstein et al. (2013 Vol. 3)
  Fractures in Stone Wall of Level J-4 Temple from Finkelstein et al. (2013 Vol. 3)
• Fig. 2.28 - J-4 Temple
  Figure 2.28
  
  Square H/7 east section. Striated remains near the base of the section are Level J-4a squatter’s debris and temple collapse (see also Fig. 2.38). At left, Probe 06/J/099/PT003 reveals the bottom of Wall 00/J/21. Note phytolith level 08/J/185 in low baulk at right.
  
  Adams in Finkelstein et al. (2013)
  Collapse Layer from Adams in Finkelstein et al. (2013)
• Fig. 2.47 - Strike Fractures
  Figure 2.47
  
  Strike fractures in Temple 4040 Wall 08/J/38.
  
  Adams in Finkelstein et al. (2013)
  in Temple 4040 from Adams in Finkelstein et al. (2013)

• Fig. 3.2 Plan of the
  Figure 3.2
  
  Plan of the Stratum XIX Temple
  
  (after Loud 1948: Fig. 390; Courtesy of The Oriental Institute of The University of Chicago).
  
  Finkelstein et al. (2000)
  Stratum XIX Temple from Finkelstein et al. (2000)
• Fig. 3.1 Aerial view of
  Figure 3.1
  
  Aerial view of Area J at the end of the 1996 season; looking east.
  
  Finkelstein et al. (2000)
  Area J at the end of the 1996 season from Finkelstein et al. (2000)
• Fig. 31.3D - Fractured
  Figure 31.3D
  
  In Area J, the monumental walls of the Level J-4 temple are fractured in several places along their strike (Fig. 31.3d) as well as perpendicular to the strike (Figs. 31.3e-f). The overlying walls of the EB III temple 4050 are not fractured
  
  Marco et. al. (2006)
  Temple Walls from Marco et. al. (2006)
• Fig. 31.3E - Fractured
  Figure 31.3E
  
  In Area J, the monumental walls of the Level J-4 temple are fractured in several places along their strike (Fig. 31.3d) as well as perpendicular to the strike (Figs. 31.3e-f). The overlying walls of the EB III temple 4050 are not fractured
  
  Marco et. al. (2006)
  Temple Walls from Marco et. al. (2006)
• Fig. 31.3F - Fractured
  Figure 31.3F
  
  In Area J, the monumental walls of the Level J-4 temple are fractured in several places along their strike (Fig. 31.3d) as well as perpendicular to the strike (Figs. 31.3e-f). The overlying walls of the EB III temple 4050 are not fractured
  
  Marco et. al. (2006)
  Temple Walls from Marco et. al. (2006)
• Fig. 40.1 - Extension
  Figure 40.1
  
  Wall 96/J/1 in Square H-10, looking east. Note fracture along the strike on the edge of the wall.
  
  Finkelstein et al. (2013 Vol. 3)
  Fractures in Stone Wall of Level J-4 Temple from Finkelstein et al. (2013 Vol. 3)
• Fig. 2.28 - J-4 Temple
  Figure 2.28
  
  Square H/7 east section. Striated remains near the base of the section are Level J-4a squatter’s debris and temple collapse (see also Fig. 2.38). At left, Probe 06/J/099/PT003 reveals the bottom of Wall 00/J/21. Note phytolith level 08/J/185 in low baulk at right.
  
  Adams in Finkelstein et al. (2013)
  Collapse Layer from Adams in Finkelstein et al. (2013)
• Fig. 2.47 - Strike Fractures
  Figure 2.47
  
  Strike fractures in Temple 4040 Wall 08/J/38.
  
  Adams in Finkelstein et al. (2013)
  in Temple 4040 from Adams in Finkelstein et al. (2013)

• from Kleiman et al. (2023)

• from Kleiman et al. (2023)

Due to Megiddo’s location in the Carmel fault zone it is particularly prone to seismic activity. It is therefore reasonable to suggest that Stratum VIA was destroyed in a massive earthquake, and in 2006 Shmuel Marco et al. published a report documenting the dates and probabilities of seismic events at Megiddo. Attributing a destruction level to an earthquake, however, is difficult to prove, as “Earthquake-related damage often resembles that deliberately caused by humans, geotechnical failure, or slow deterioration over the ages” (Marco et al. 2006: 569). Investigators thus looked for certain criteria that could suggest seismic activity, the main criterion being “time-constrained widespread damage. The temporal bounds should be tight enough to indicate that the damage occurred in a single catastrophic event” (Marco et al. 2006: 569). Other significant pieces of evidence include the absence or scarcity of weapons, historical records that document seismic activity, “deformation of coeval natural sediments, and the existence of certain types of damage that are uniquely associated with earthquakes” (Marco et al. 2006: 569).

Ultimately, the study was inconclusive regarding the occurrence of a major earthquake in Stratum VIA. Only two earthquakes were confirmed beyond doubt: “one at the end of the fourth millennium BCE and another in the 9th century BCE (which caused the damage in Stratum VA-IVB)” (Marco et al. 2006: 572). This study did not eliminate the possibility of an earthquake in Stratum VIA, however there was no conclusive evidence to support such an event (Marco et al. 2006: 572).

Despite the inconclusive results of the archaeoseismic study regarding Stratum VIA, other considerations have supported the view that this city was not destroyed in an earthquake. First of all, the remains from this stratum show extensive burning, suggesting that the city was intentionally burned down by a hostile force. “It is difficult to imagine that all the stone and brick-built houses could have simultaneously caught fire due to an earthquake as might happen in a modern city” (Finkelstein, Ussishkin and Halpern 2006: 850). Secondly, the following stratum showed discontinuity in material culture, suggesting that a new population inhabited the later city and that the inhabitants of the Stratum VIA city did not return following its destruction. “Had the Stratum VIA settlement been destroyed by fire resulting from an earthquake we would have expected the inhabitants to reconstruct their ruined houses without delay. The fact that they did not return indicates that they were forced to abandon their settlement forever by a human agent” (Finkelstein, Ussishkin and Halpern 2006: 850). Finally, as previously stated, the radiocarbon results from the region suggest that there were two main destruction events, further strengthening the theory that this destruction was part of an ongoing process such as the gradual conquest of the region by an outside power rather than a single time-restricted event.

In the search for the military power responsible for the destruction of this city, the Egyptian army of Pharaoh Shoshenq I has been a popular explanation. This campaign, known from 1 Kgs 14:25-28 (cf. 2 Chr 12:1-12) as well as from the temple of Amun at Karnak (see Simons 1937: 95-102), has traditionally been dated to around 926 BCE by the biblical text (Finkelstein 2002: 109), which states that “In the fifth year of King Rehoboam, King Shishak of Egypt came up against Jerusalem (1 Kgs 14:25, NRSV).

A late 10th century BCE date for Shoshenq’s campaign has proven troublesome for the archaeologist. Egyptian records do not corroborate the precise year of his campaign, thus some have suggested a wider range of possible dates for this event. Finkelstein raises four main problems regarding the date of this campaign:

1. the complicated chronology of the 21st and 22nd Dynasties in Egypt allow for the change of several years back or forth of Shoshenq I’s reign
2. it is unknown whether his campaign occurred early or late in his rule
3. the biblical references of the length of reigns of the early Davidides are completely schematized
4. the fifth year of Rehoboam datum may have been altered to fit the theology of the Deuteronomistic Historian (Finkelstein 2002: 110).

Thus after a critical review of the available evidence, “the Shoshenq campaign could have taken place almost any time in the mid- to late-10th century BCE” (Finkelstein 2002: 110; see also Ash 1999: 27-34).

Due to the uncertainty of the date of Shoshenq I’s campaign, identification of destruction layers to his campaign have been problematic. Indeed, “we do not have even one destruction layer which can safely be assigned to this campaign. For instance, in several sites there are two destruction horizons (e.g. and Megiddo, Stratum VIA and Stratum VA-IVB); both can be attributed to Shoshenq’s campaign” (Finkelstein 1996: 180). Furthermore, the destruction layer of Stratum IVA has also been attributed to this campaign (Guy 1931: 48).

Unfortunately, the most significant piece of evidence that could contribute to this debate was found out of context during Megiddo excavations. Found during Fisher’s excavations in one of Schumacher’s dumps, a fragment of a large stone stele of Pharaoh Shoshenq was discovered near the eastern edge of the mound (Ussishkin 1990: 71). This fragment appears to be part of a stele that would have measured approximately 3.30 m high, 1.50 m wide, and 50 cm thick (Ussishkin 1990: 72). Four criteria typify this type of monument:

First, they were erected by a monarch in a foreign land; second, their erection was carried out following, or in association with, a military campaign or submission by the local ruler to the conquering leader; third, they were erected in a public, central position inside a capital city; fourth, the city in question was inhabited at the time the monument was erected and dominated by the monarch who erected the stele (Ussishkin 1990: 72).

As a result of this find many scholars do not attribute the destruction of Stratum VIA to Shoshenq I’s campaign. The erection of a victory stele in the city implies that the city would have had inhabitants to appreciate such an impressive work, yet the total destruction of this stratum and the settlement gap following it would imply that had Shoshenq I been responsible for this destruction, he would have erected his stele in an uninhabited city.

Therefore, it would seem that Shoshenq I’s campaign occurred in the Iron Age IIA rather than during the Iron Age I. Stratum VIA is ruled out due to four main considerations:

1. At least some of these destructions are radiocarbon dated to before the highest possible date for his reign
2. the radiocarbon evidence indicates a gradual demise of these cities and not a single destructive event
3. there was no reason for a pharaoh who was probably interested in re- establishing Egyptian rule in the area to devastate a fertile valley that was the bread basket of the entire country
4. it is illogical that Sheshonq I would establish a stele in a deserted Megiddo (Finkelstein 2011: 232).

Shoshenq I’s campaign was intended to reestablish Egyptian hegemony in the region, made evident from the stele fragment. This being the case, he likely would have only destroyed the elite quarter of the city, or perhaps taken it peacefully (Finkelstein 2009: 121; regarding the possibility of a peaceful takeover see Finkelstein 2002: 122). Thus, Shoshenq I’s campaign to the north should not be attributed to Megiddo Stratum VIA, but rather to a later city in the Iron Age IIA (Ussishkin 1990: 73; Finkelstein 2011: 232).

After ruling out destruction by both earthquake and the military campaign of Pharaoh Shoshenq I, one must investigate the possibility of the Israelite expansion to the region from the northern highlands. Indeed, this suggestion best fits the current set of data available by both textual sources and archaeological excavations.

First of all, the data from radiocarbon studies “renders the earthquake and single military campaign theories invalid,” as the dates provided by these samples suggest two waves of destruction in the region rather than one (Finkelstein 2013: 1337). A series of raids or the gradual expansion of the Israelites from the highlands better explains these waves of destructions (Na’aman 2007: 402; Finkelstein 2011: 229; 2013: 1337).

Second, the later settlements at Megiddo support this reconstruction. The break in material culture between the Iron Age I-IIA suggests that the population that resettled in the city following Stratum VIA was a different people, most likely the Israelites who later on established the Omride Kingdom in the region (Megiddo VA-IVB and its contemporaries) (Finkelstein 2009: 122-123). This reconstruction still leaves room for the campaign of Shoshenq I in the decades following the Israelite expansion, thus allowing for a unified and harmonious historical reconstruction inclusive of all known historical events from this period.

• from Adams in Finkelstein et al. (2013:79-80)

The process described above of the Phase J-4a crisis of the Great Temple is opposed to the earthquake hypothesis proffered in Megiddo IV for the ‘destruction’ and abandonment of the site at the end of the EB IB (Finkelstein and Ussishkin 2003; 2006; Marco et al. 2006). Let us briefly review the evidence previously presented:

1. Monumental walls of the Level J-4 temple are fractured along their strike. It is noted (erroneously, see below) that the overlying walls of Temple 4040 are not fractured, suggesting that the shockwave that caused the fractures occurred between the construction of the Level J-4 temple and the Level J-7 temples (Marco et al. 2006: Table 31.1, No. 1; Finkelstein, Ussishkin and Peersmann 2006: 49). If the upper temple were fractured, it would suggest either: a) that if there had been an earthquake it happened later, affecting both strata, or b) the damage is related to something else (see suggestion below).


2. The sequence of two floors within the temple sanctuary separated by ca. 15 cm of collapsed mudbrick debris was interpreted as abandonment of the temple, collapse from the walls and reoccupation of the temple (Finkelstein, Ussishkin and Peersmann 2006: 38).


3. The north face of Wall 96/J/1, in Square D/10, has a vertical crack, and the wall east of the crack has sunk slightly toward the slope (Finkelstein, Ussishkin and Peersmann 2006: 49). Likewise, the north face of Wall 96/J/7 in Square H/9 displays a pulling-apart of the wall and a sinking of the wall west of the separated interstices (Fig. 2.45; Marco and Finkelstein, observations in the field, 2008).


4. When the inhabitants returned to the temple after the earthquake (Phase J-4a) they attempted to shore-up some of the crumbling walls within the corridors (Finkelstein, Ussishkin and Peersmann 2006: 50–51). This included the construction of two flimsy mudbrick Walls 96/J/11 and 98/J/12, perpendicularly across the corridors, presumably to keep the towering 3.5-m-thick walls from collapsing (ibid.: Fig. 3.20).


5. Basalt ritual Table 08/J/205 has clearly sunk. The western section of Square H/7 that runs through the table shows that this sinking caused a vertical fracture in the strata above (Fig. 2.46; Marco and Finkelstein, observations in the field, 2008).


6. Wall 96/J/1 (Squares B–C/11) was ‘repaired’ by laying flat stone slabs atop the stone foundation to receive new mudbricks (ibid. 2006: 51).

The 2004–2006 excavations in Area J provided new material that forces us to reconsider the evidence above. I offer the following data to refute the interpretation of the evidence above (numbers below correspond to the numbers above):

1. Level J-7 Temple 4040 above the Level J-4 temple does display numerous examples of fractures along wall strikes. I observed and photographed no fewer than eight separate examples of fracture lines across numerous stones along their strike in Walls 08/J/37, 08/J/38, 08/J/39 and 08/J/40 (Fig. 2.56). For the sake of brevity, one example is provided in Fig. 2.47.¹⁷ The strike fractures in the Level J-4 temple, therefore, provide no evidence for an earthquake at the end of the EB IB.


2. The evidence for two floors within the sanctuary has been corrected above. There was one floor, followed by slow collapse and irregular activity.


3. The vertical cracking and sinking of wall segments should be expected at these locations. These points are closer to the east and west edges of the upper terrace. The Level J-4 construction would have required enormous fills to support the temple, which would have been weakest along the deeper edges, thus more prone to natural sinking.¹⁸


4. The shoring up of the corridor walls is a very important component of Phase J-4a as presented in Megiddo IV. As we shall see in more detail below, these two ‘walls’ are artefacts of excavation, and are, in fact, part of a laid-mudbrick fill placed between and along the entire length of the corridors that was excavated through. This fill was laid in Level J-5 as part of the dismantling of the Level J-4 temple and the preparation of the site for construction at that time.


5. The fracturing of the accumulation near basalt ritual Table 08/J/205 is clear. There certainly was settling. As in number 3 of my refutations, above, this settling is expected. However, this settling occurred after the deposition of the Phase J-4a activity, which is represented in the striations in the section (Fig. 2.46). While much later settling is my preferred explanation, if one must hypothesize an earthquake, it will have to have taken place after the accumulation of Phase J-4a debris and the collapse of the Level J-4 roof, and therefore cannot be responsible for the abandonment of the site in the EB IB.


6. The technique of laying flat stones at the top of the socle was almost certainly part of the construction technique for the Level J-4 temple. It is an attested EB architectural feature at qiryat Ata, for example (Golani 2003: 75). Additionally, the flat stones in Wall 96/J/1 can be none other than those robbed from the Levels J-2 and J-3 Picture Pavement. These stones are distinct in their size, shape and thickness, and a good portion of stones at the top of the wall are identical.¹⁹ As Level J-3 was almost certainly covered in the course of construction of Wall 4045 and the terrace fill, it is most likely that these stones were robbed and used in the Level J-4 construction project. The proximity of this end of the wall to the location of the earlier pavement explains why it is only here that the builders used these particular stones. Note also the foundation deposit stone within the Level J-4 sanctuary, which almost certainly comes from the Picture Pavement as well (see above).

In conclusion, all of the evidence for an earthquake in the EB IB at Megiddo has been refuted by new data. Socio-economic and political factors must underlie the Phase J-4a crisis of the Level J-4 Great Temple, just as they underlie the striking change in settlement patterns throughout the Jezreel Valley at this time (Finkelstein et al. 2006: 763; Finkelstein and Ussishkin 2000b: 585–588; Adams, Finkelstein and Ussishkin forthcoming). While activity continued at the site irregularly, the site was nearly vacant during the succeeding EB II period (Finkelstein and Ussishkin 2000b: 585–586; Greenberg 2003). When the site was reoccupied in the EB III (Level J-5/J-6), the roof had caved in, but the mudbrick walls, although weathered by years of exposure, were probably standing to at least half if not most of their original height. The Level J-5/J-6 inhabitants of the site dismantled portions of the temple, robbed stone from the walls and filled in the sanctuary to create space for their construction activity

• from Ussishkin in Finkelstein et al. (2013:1317)

the termination of the 2010 excavation season brought to an end 18 years of intensive work in area BB, which had previously been excavated by the oriental institute expedition directed by gordon loud (1948). our goal was to elucidate the various remains of the cultic compound, which date in the main to the early bronze age. our work here – our area J – was reported in Megiddo III and Iv and also in this volume. there are naturally divergent views regarding the interpretation and history of the various remains. in this chapter i will summarize the evidence and history of the cultic compound as i understand them. my response will call special attention to the papers by adams and Finkelstein in the current report (chapters 2, 40).

the oriental institute expedition worked in area BB on a large scale, clearing the debris of later strata that had accumulated here for several metres above the early bronze remains. on the eastern, sloping side of area BB, the excavation reached bedrock, while on the western side it reached the earlier levels of the cultic compound. This situation determined the mode of our work: first, we were limited by the borders of Area BB, as extending our dig beyond the edges of Area BB meant first clearing the huge accumulation of later strata. Second, many structural remains of the early bronze cultic compound had been exposed above ground by loud and it was now fairly easy to follow in his footsteps and investigate them afresh. third, all efforts were taken by us to preserve the structural remains of the exposed temples, and this, naturally, limited the space available for further digging.

Following our 18 years of continuous, systematic excavation, it seems that a fairly clear picture of the history of the unique cultic compound can be drawn. however, the picture is incomplete, interpretations differ widely and there is room and justification for conducting further excavations in the future.

• from Ussishkin in Finkelstein et al. (2013:1320-1321)

For some unknown reasons, the EB IB religious authorities of Megiddo decided for the second time during this period to replace the existing temple with a new sanctuary facing the northern side of the site. the level J-3 temple and its courtyard were apparently abandoned and the entire complex left to deteriorate. The newly built magnificent Level J-4 temple – nicknamed by us the Great Temple – has been described in detail elsewhere (Adams, Chapter 2, Part III; Adams, Finkelstein And Ussishkin forthcoming). i will comment on a few aspects of this edifice below.

Significantly, the architects reverted to the model of the earlier Level J-2 temple rather than following that of the level J-3 temple. also here, as explained above, the broad-room sanctuary had two parallel wings, each having six offering tables arranged in two rows, with the entrance and the altar situated in the centre of the hall.

naturally, we can only speculate on the reasons behind the decision of the eb ib religious authorities of Megiddo to twice replace an existing and functioning sanctuary. i can pose three possibilities for the second change, that is the replacement of the level J-3 temple with that of level J-4: 1) the wish to revert to the plan and concept of the level J-2 temple, which was not adopted in the level J-3 temple; 2) the desire and the organizational ability of the prosperous community to build a larger, more impressive sanctuary; 3) the wish to have the sanctuary oriented towards ‘Ein El-Kubbi.

the question of the temple’s orientation has been addressed previously (Finkelstein, ussishkin and Peersmann 2006: 48–50). Peersmann and ussishkin, but not Finkelstein, argued that the position and orientation of the temple is not accidental, and have implications regarding the function and cult of the sanctuary. the orientation of the entrance of the temple hall, that is the orientation of the shorter axis of the building, is in line with ‘ein el-Kubbi, the spring at the bottom of the mound. it would follow that the temple and its cult were associated with water and with the spring that it faced. it can be assumed that – similar to the earlier temples of levels J-2 and J-3 – the façade and entrance of the great temple adjoined the upper edge of the northern slope of the site at that time. We can safely assume that here, too, as in the two earlier temples, the entrance opened to a courtyard or a kind of Picture Pavement extending down the slope in the direction of the spring. Finally, we must also mention the analogous case of the ghassulian temple of ‘ein gedi whose main entrance, built at the upper edge of the slope, faced the ‘ein gedi spring situated further down the slope (ussishkin 1980).

the walls in the three temples of levels J-2, J-3 and J-4 are all based on the surface of the time and do not have any subterranean foundations. Adams and Finkelstein believe that the Level J-4 edifice, being of such huge dimensions, must have been built on an artificial podium supported by terraces. However, no remains of such a podium and its terraces were uncovered. the so-called terraces on the eastern slope, to be discussed below, are later in date. the so-called terraces on the southern and western sides (Walls 96/J/23 and 00/J/19) are in fact proper walls that form part of the building complex or kind of a fence to enclose the building proper on those sides. the situation on the northern side, where we could not excavate, as it is located outside area BB, is unknown. however, it is clear that also here at least the inner side of the northern wall of the temple (Wall 00/J/21) was established on remains of the earlier levels (see Adams, chapter 2, Part III)

• from Ussishkin in Finkelstein et al. (2013:1321-1323)

According to the view of Marco, Agnon, Finkelstein and Ussishkin (2006; Finkelstein, Ussishkin and Peersmann 2006: 49–50) the main phase in the history of the great temple met its end in an earthquake. Strong arguments against this hypothesis are presented by Adams (chapter 2), but Finkelstein (chapter 40) and i hold to our view.

in the previous reports, based on the interpretation of the stratigraphical evidence, we concluded that following the assumed earthquake the damaged temple was partly repaired and reused (Finkelstein and Ussishkin 2000a: 65–67²; Finkelstein, Ussishkin and Peersmann 2006: 50–52). Adams (chapter 2), now supported by Finkelstein (Adams, Finkelstein and Ussishkin forthcoming), presents strong arguments against this hypothesis. Adams concludes that the temple was not repaired and that only one floor could be detected in the temple hall. He defines the late period of the temple (Phase J-4a) as a ‘period of abandonment’ rather than a ‘period of reuse’. relying on the study of Friesem and Shahack-gross (chapter 2, Part v), he states that “the later floor turned out to be a phytolith lens representing collapse and sporadic Phase J-4a activity within the temple” such as “ephemeral hearths scattered around the sanctuary”.

however, i believe that the evidence for a repair, at least in the corridors, is strong and cannot be ignored. The mudbrick superstructure of the southeast corner of the edifice, at the eastern end of Wall 96/J/1, was now rebuilt. the mudbricks were laid on a course of stone slabs placed on the top of the damaged stone substructure. A new, higher floor (94/J/39), demarcated by a row of stones (Wall 94/J/15), was laid at that corner.³ Some mudbrick walls were added in the corridors. Adams’ assertion that these are mudbrick fills dumped in Level J-5 is not convincing as the mud-bricks were laid in an orderly manner (see section in Finkelstein, Ussishkin and Peersmann 2006: Fig. 3.13).

obviously, the sanctuary started to fall apart at this time, as also evidenced by the discovery of owl pellets, but we all agree that cultic activities continued in the temple during Phase J-4a, as indicated in the main by ‘ephemeral hearths’ and animal bones. Whether the surface level in the temple hall on which these activities took place should be termed a ‘later floor’ or a ‘phytolith lens’ is of secondary importance. according to Adams, supported by Finkelstein, the entrance to the temple was intentionally damaged, apparently at the beginning of the ‘abandonment period’, as part of a religious rite. the sole factual support for this theory comes from a broken basalt slab, apparently part of the paneling decorating the door jamb, which was found lying in the entrance. however, the entire stone-built jamb and the continuing wall were robbed, at a later date, of their stones down to the foundation course, and most likely the basalt panels were ruined at that time. in any case, the basalt slab in question could have fallen at any time and no importance, particularly ritual importance, should be attributed to it.

Finally, I have to refer briefly to the cache of Egyptianized pottery. We concluded at the time that from the stratigraphical perspective the cache can be assigned either to Phase J-4a of the eb ib period or to the overlying level J-5 of the eb iii period (Finkelstein and Ussishkin 2000b: 586–587; Finkelstein, Ussishkin and Peersmann 2006: 50–52). new stratigraphical data related to the cache have not been added in recent seasons. based on the typological perspective, it was assigned by Joffe (2000: 170–174) to Phase J-4a of the eb i period. Adams now ascribes it to level J-7 of the ib period and considers the cache as foundation deposit of ‘megaron’ temple 4040 (m. Adams, personal communication).

• from Ussishkin in Finkelstein et al. (2013:1331)

adams (chapter 2, Part iii) now argues against our previous interpretation (Finkelstein et al. 2006c: 846), based on the work of marco et al. (2006), according to which the great temple of level J-4 had been damaged by an earthquake and was consequently abandoned for several centuries (in the eb ii). instead, adams relates the abandonment of the temple in particular and megiddo in general to socio-political change. Yet, none of Adams’ arguments is sufficiently strong to change the earthquake theory. Table 39.1 summarizes the arguments pro and con.



to sum up this issue, the two main pieces of evidence for the earthquake theory – nos. 1 and 3 in the table – have not been challenged and no factual evidence for an alternative theory has thus far been presented. Taking into consideration the sensitive location of Megiddo on the Carmel fault, and the difficult-to-explain abandonment of such a strategic site for several centuries, the earthquake theory should be maintained.

• from Ussishkin in Finkelstein et al. (2013:1320-1321)

For some unknown reasons, the EB IB religious authorities of Megiddo decided for the second time during this period to replace the existing temple with a new sanctuary facing the northern side of the site. the level J-3 temple and its courtyard were apparently abandoned and the entire complex left to deteriorate. The newly built magnificent Level J-4 temple – nicknamed by us the Great Temple – has been described in detail elsewhere (Adams, Chapter 2, Part III; Adams, Finkelstein And Ussishkin forthcoming). i will comment on a few aspects of this edifice below.

Significantly, the architects reverted to the model of the earlier Level J-2 temple rather than following that of the level J-3 temple. also here, as explained above, the broad-room sanctuary had two parallel wings, each having six offering tables arranged in two rows, with the entrance and the altar situated in the centre of the hall.

naturally, we can only speculate on the reasons behind the decision of the eb ib religious authorities of Megiddo to twice replace an existing and functioning sanctuary. i can pose three possibilities for the second change, that is the replacement of the level J-3 temple with that of level J-4: 1) the wish to revert to the plan and concept of the level J-2 temple, which was not adopted in the level J-3 temple; 2) the desire and the organizational ability of the prosperous community to build a larger, more impressive sanctuary; 3) the wish to have the sanctuary oriented towards ‘Ein El-Kubbi.

the question of the temple’s orientation has been addressed previously (Finkelstein, ussishkin and Peersmann 2006: 48–50). Peersmann and ussishkin, but not Finkelstein, argued that the position and orientation of the temple is not accidental, and have implications regarding the function and cult of the sanctuary. the orientation of the entrance of the temple hall, that is the orientation of the shorter axis of the building, is in line with ‘ein el-Kubbi, the spring at the bottom of the mound. it would follow that the temple and its cult were associated with water and with the spring that it faced. it can be assumed that – similar to the earlier temples of levels J-2 and J-3 – the façade and entrance of the great temple adjoined the upper edge of the northern slope of the site at that time. We can safely assume that here, too, as in the two earlier temples, the entrance opened to a courtyard or a kind of Picture Pavement extending down the slope in the direction of the spring. Finally, we must also mention the analogous case of the ghassulian temple of ‘ein gedi whose main entrance, built at the upper edge of the slope, faced the ‘ein gedi spring situated further down the slope (ussishkin 1980).

the walls in the three temples of levels J-2, J-3 and J-4 are all based on the surface of the time and do not have any subterranean foundations. Adams and Finkelstein believe that the Level J-4 edifice, being of such huge dimensions, must have been built on an artificial podium supported by terraces. However, no remains of such a podium and its terraces were uncovered. the so-called terraces on the eastern slope, to be discussed below, are later in date. the so-called terraces on the southern and western sides (Walls 96/J/23 and 00/J/19) are in fact proper walls that form part of the building complex or kind of a fence to enclose the building proper on those sides. the situation on the northern side, where we could not excavate, as it is located outside area BB, is unknown. however, it is clear that also here at least the inner side of the northern wall of the temple (Wall 00/J/21) was established on remains of the earlier levels (see Adams, chapter 2, Part III)

• from Ussishkin in Finkelstein et al. (2013:1321-1323)

According to the view of Marco, Agnon, Finkelstein and Ussishkin (2006; Finkelstein, Ussishkin and Peersmann 2006: 49–50) the main phase in the history of the great temple met its end in an earthquake. Strong arguments against this hypothesis are presented by Adams (chapter 2), but Finkelstein (chapter 40) and i hold to our view.

in the previous reports, based on the interpretation of the stratigraphical evidence, we concluded that following the assumed earthquake the damaged temple was partly repaired and reused (Finkelstein and Ussishkin 2000a: 65–67²; Finkelstein, Ussishkin and Peersmann 2006: 50–52). Adams (chapter 2), now supported by Finkelstein (Adams, Finkelstein and Ussishkin forthcoming), presents strong arguments against this hypothesis. Adams concludes that the temple was not repaired and that only one floor could be detected in the temple hall. He defines the late period of the temple (Phase J-4a) as a ‘period of abandonment’ rather than a ‘period of reuse’. relying on the study of Friesem and Shahack-gross (chapter 2, Part v), he states that “the later floor turned out to be a phytolith lens representing collapse and sporadic Phase J-4a activity within the temple” such as “ephemeral hearths scattered around the sanctuary”.

however, i believe that the evidence for a repair, at least in the corridors, is strong and cannot be ignored. The mudbrick superstructure of the southeast corner of the edifice, at the eastern end of Wall 96/J/1, was now rebuilt. the mudbricks were laid on a course of stone slabs placed on the top of the damaged stone substructure. A new, higher floor (94/J/39), demarcated by a row of stones (Wall 94/J/15), was laid at that corner.³ Some mudbrick walls were added in the corridors. Adams’ assertion that these are mudbrick fills dumped in Level J-5 is not convincing as the mud-bricks were laid in an orderly manner (see section in Finkelstein, Ussishkin and Peersmann 2006: Fig. 3.13).

obviously, the sanctuary started to fall apart at this time, as also evidenced by the discovery of owl pellets, but we all agree that cultic activities continued in the temple during Phase J-4a, as indicated in the main by ‘ephemeral hearths’ and animal bones. Whether the surface level in the temple hall on which these activities took place should be termed a ‘later floor’ or a ‘phytolith lens’ is of secondary importance. according to Adams, supported by Finkelstein, the entrance to the temple was intentionally damaged, apparently at the beginning of the ‘abandonment period’, as part of a religious rite. the sole factual support for this theory comes from a broken basalt slab, apparently part of the paneling decorating the door jamb, which was found lying in the entrance. however, the entire stone-built jamb and the continuing wall were robbed, at a later date, of their stones down to the foundation course, and most likely the basalt panels were ruined at that time. in any case, the basalt slab in question could have fallen at any time and no importance, particularly ritual importance, should be attributed to it.

Finally, I have to refer briefly to the cache of Egyptianized pottery. We concluded at the time that from the stratigraphical perspective the cache can be assigned either to Phase J-4a of the eb ib period or to the overlying level J-5 of the eb iii period (Finkelstein and Ussishkin 2000b: 586–587; Finkelstein, Ussishkin and Peersmann 2006: 50–52). new stratigraphical data related to the cache have not been added in recent seasons. based on the typological perspective, it was assigned by Joffe (2000: 170–174) to Phase J-4a of the eb i period. Adams now ascribes it to level J-7 of the ib period and considers the cache as foundation deposit of ‘megaron’ temple 4040 (m. Adams, personal communication).

• from Nur and Burgess (2008:189-199)

Figures Discussion

A modest mound southeast of the Carmel Ridge in Israel, the tell of Megiddo rises 50 meters above the surrounding Jezreel Plain and covers some 6 hectares (15 acres) of land (Figure 7.1). This tell was first identified as the legendary Armageddon [the Greek corruption of the Hebrew “Har Megiddo” or Mount Megiddo] by a Jewish writer of the fourteenth century, Esthori Haparchi, and then was rediscovered by the British army officer H. H. Kitchener five hundred years later. Extensive excavations were conducted at the site by C. Fisher, P. Guy, and G. Loud (Yadin 1975). These archaeological studies revealed physical evidence of the historical development of Megiddo.

Megiddo’s strategic importance, belied by its unimposing appearance today, stems from its unique topography (Figure 7.2). The land between the Mediterranean and the Jordan River served as a bridge between the civilizations in the South, in Egypt and Arabia, and those in the North, in Syria, Mesopotamia, and Anatolia. It was also a continuation of sea routes from the Mediterranean and the Gulf of Suez. The ruggedness of the region, however, crossed by several ranges of mountains and hills as well as the lowest valley on Earth, limited the possible routes for overland shipping or wheeled travel. The Carmel-Gilboa mountain range was a particular obstacle, and traffic from the Mediterranean to Syria and Jordan was funneled through a few gaps in the range. In fact, both the passes and the mountain range that obstructed traffic were products of tectonic motion along the seismically active Carmel Gilboa fault system, which branches off from the Dead Sea Transform (see Figure 4.5).

The mound of Megiddo stands guard over one of the most important of these mountain passes, the Nahal Iron Pass, a traffic bottleneck on the main route between Egypt and Syria. Until the advent of more elaborate road construction by the Roman Empire in the first and second centuries AD, the gap at Megiddo was the only one that permitted the passage of chariots, though it was not an easy passage. A description from the latter part of the thirteenth century BC, found in the Egyptian Papyrus Anastasi I, gives some idea of the difficulty of the route:

Thy path is filled with boulders and pebbles, without a toe-hold for passing by, overgrown with reeds, thorns, brambles, and wolf’s paw. The ravine is on one side of thee, and the mountain rises on the other. Thou goest on jolting, with thy chariot on its side, afraid to press thy horse too hard. . . . The horse is played out by the time thou findest night quarters. (Hori [Egyptian Royal Official], Papyrus Anastasi I)

This narrow and difficult route made the pass from Megiddo particularly easy to guard, so whoever held power in Megiddo controlled not only the course of trade in the Fertile Crescent but that of war as well. Thus, the site figured prominently in some of the greatest ancient battles fought in this region. As Pharaoh Thutmose III expressed it, “The capture of Megiddo is the capture of 1,000 towns.” Indeed, fortifications were built and rebuilt there for close to five millennia, until around 500 BC (Finkelstein and Ussishkin 1994).

Four levels of destruction in the mound of Tel Megiddo are consistent with earthquake destruction, the lowest one attributed to the conquest by Thutmose III in 1468 BC. Why, however, would the Pharaoh have destroyed the place if his goal were to occupy the site and exact tribute? Although it is clear that Thutmose III conquered Megiddo, there is no more reason to assume that he ordered its physical destruction than to believe it was caused by an earthquake.

The second massive destruction at Megiddo, which occurred around 1250 BC, has variously been attributed to the Israelites or the Philistines, although historical evidence supports neither candidate. Again, however, the excavation of collapsed walls in Megiddo, and similar contemporaneous destruction in many nearby sites (Davies 1986), make the earthquake hypothesis a likely candidate.

The strongest evidence for earthquake destruction at Megiddo is probably the layer dating to between 1130 and 1000 BC, which some scholars attribute to conquest by King David’s army. There is no historical mention, however, of David capturing Megiddo, much less leveling it, and, given the importance of the place at the time, it seems unlikely that such a conquest would go unheralded. More likely, Megiddo was destroyed by a massive earthquake, perhaps the same one that, according to the Bible, occurred during the battle of Michmash (1 Samuel 14:15). In the 1930s, excavators found collapsed walls from this period and, under the walls, smashed jars and several human skeletons, including the one shown in Figure 7.3. This layer is possibly contemporaneous with destruction layers from Dor, Gezer, and at least a dozen other sites in Israel and Jordan, all of them consistent with earthquake destruction.

The fourth layer of destruction occurred sometime after the conquest of Megiddo by Pharaoh Sheshonq in 925 BC. Although this layer is sometimes attributed to him, most documentary and archaeological evidence indicates that Sheshonq did not destroy the city but rather had a monument erected in his honor there and exacted tribute from the residents. Yet, on the other hand, no definitive evidence has yet been excavated to indicate that an earthquake caused this destruction. We know that the massive earthquake during the reign of King Uzziah, around 760 BC, was important in this area—so important, in fact, that it is mentioned in the prophesies in the book of Zechariah (14:4–5):

And his feet shall stand in that day upon the mount of Olives, which is before Jerusalem on the east, and the mount of Olives shall cleave in the midst thereof toward the east and toward the west, and there shall be a very great valley; and half of the mountain shall remove toward the north, and half of it toward the south. And ye shall flee to the valley of the mountains . . . yea, ye shall flee, like as ye fled from before the earthquake in the days of Uzziah king of Judah.

Other sources claim that the earthquake Zechariah cites occurred around 760 BC. In any case, clearly it was a major enough disaster to be used as a temporal reference in Zechariah’s prophecy. In fact, the prophecy includes such vivid details that it may actually describe the ground motion that occurred in the 760 BC earthquake, perhaps motion across a strike-slip fault.

Another example of an earthquake prophecy appears in the book of Revelation, where a mighty earthquake is prophesied to occur during the battle of Armageddon:

And they assembled them at the place that in Hebrew is called Armageddon . . . and there came . . . a violent earthquake, such as had not occurred since people were upon the earth, so violent was that earthquake: And the great city was split into three parts, and the cities of the nations fell . . . And every island fled away, and no mountains were to be found.

This account and the one in Zechariah may be examples of retrospective prophecies, a common feature in ancient literature where the author dramatizes a historic event as a kind of warning after the fact. This may indicate that, in the past, earthquakes were known to have struck Megiddo. Most important is that excavators keep in mind in future excavations that the area is subject to severe earthquake hazards, and always consider earthquakes as a possible cause for unexplained destruction there.

Topography is what made Megiddo so important militarily for so long, but other factors also determined the location of cities and forts in the ancient Mediterranean region. Another consideration in this arid land was the availability of water. The previous chapter discussed the importance of water in Qumran: dependent on cisterns to capture seasonal rainfall, Qumran was particularly vulnerable to earthquake damage. Cisterns were essential in many Israeli regions, and the collection, storage, and distribution of water was important throughout Israel’s history. Any place with a year-round supply of fresh water, particularly clean, filtered water from a dependable spring, was prime real estate. One example was Jericho, described in chapter 3.

Tel Jericho’s extraordinary stack of at least twenty-two layers of destruction has already been noted. With no modern city built on top of the tell—modern Jericho is a few miles from the remnants of its ancient namesake—archaeological excavation has progressed relatively unhindered, despite the usual impediment of political unrest in the area. This has been a liability in some ways, since excavations using early archaeological methods clearly destroyed at least as much information as they revealed. Still, because of these excavations, the repeated destruction and rebuilding of Jericho has been acknowledged for some time. With its long record and repeated rebuilding, Jericho would be a great candidate for a sort of key site, a place where evidence of repeated earthquake damage could be correlated with destruction in neighboring sites where the record might be less continuous. If we could systematically examine the written evidence of earthquakes and try to tie together the archaeological stratigraphy among the sites scattered around Israel and Jordan, we might be able to piece together a much more informative record than we could hope to gain from any one site alone. A great deal of archaeological work has been done to try to correlate the remains at Jericho with other sites throughout the Holy Land. Kathleen Kenyon (1978) was particularly comprehensive in her synthesis of the archaeology of the region. Although she does mention the general likelihood of earthquakes in this region and specifically ascribes certain damage layers to earthquakes, she makes no explicit effort to determine the range of such earthquake damage; she does not mention whether the earthquake in question could have damaged nearby towns.

One of the most suggestive layers in Jericho is one Kenyon dated, based on pottery styles, to between 1600 and 1550 BC, where the walls of Jericho collapsed, burying storage jars full of grain (Figure 7.4). The grain was carbonized by a fire concurrent with the collapse, and carbon dates of the grain have placed it remarkably close to Kenyon’s original estimate (Bruins and van der Plicht 1996). According to Kenyon (1979, 177–178), many towns in the region suffered total destruction at the same time (though she does not implicate an earthquake), including the sites of Tel Beit Mirsim, Hazor, Shechem, and Megiddo. Among these sites, only Megiddo was immediately rebuilt, with no apparent change in culture; most of the others were abandoned for more than a century.

In chapter 3, I speculated that an earthquake may have figured prominently in the biblical story of Joshua’s destruction of Jericho. Does the archaeological record preserve any evidence at all of Joshua’s Jericho? One of the early expeditions reported the discovery of a city wall from Joshua’s time, but Kenyon (1957) indicated that the dating of that wall was incorrect. As described in chapter 3, she found only one small remnant of Jericho that could have coincided with the traditional estimate of 1400 to 1250 BC for Joshua’s conquest of the site. If a larger habitation did exist at the site, the remains of that culture have all but vanished, eroded away during a long period of abandonment.

Recently, however, archaeologists have reexamined both the archaeological site at Jericho and radiocarbon dates for other events in the ancient world. One particularly notable theory links the plague of darkness in Egypt, before the Exodus, to the explosive eruption of the island of Thera in the Aegean Sea, an event dated to the end of the seventeenth century BC. If this date is correct, the nearly missing layer Kenyon attributed to Joshua’s time would be far too recent, and therefore the destruction from ca. 1600 would be a more likely candidate.

The debate over the assignment of this layer, based on samples of locally made pottery unearthed at Jericho, and on radiocarbon dates from the charred grain and from bits of timber found in the destruction layer, has at times been heated, almost vicious. (For an example of two diametrically opposed interpretations of the same evidence, presented side-by-side in one issue of Biblical Archaeology Review, complete with ad hominem attacks, see Bienkowski 1990 and Wood 1990.) It will be some time before the dust settles on this issue. For now, the archaeology is the source of more questions than answers, and until some of those questions can be resolved, using Jericho as a key site will remain difficult.

Although the mound of ancient Megiddo was eventually abandoned, and the modern city of Jericho has abandoned the old tell, the city of Jerusalem has been continuously inhabited for at least four thousand years (Cline 2004). The reason for its continued importance reaches beyond favorable topography or water supply to that least tangible reason for the persistence of cities: religious significance. King David’s choice of Jerusalem as the capital city of his Hebrew nation, and, probably more important, his decision to move the Ark of the Covenant into the city, sealed its fate as an enduring human habitation into modern times.

Despite its long history, most of the city’s archaeological secrets are nearly impossible to reveal. The whole city has been densely inhabited for millennia, with almost no abandonment in any quarter. Modern and ancient homes, churches, and official buildings block access to the layers of rubble beneath them. When a scrap of real estate is not actively occupied, it is usually because it is imbued with some intense political or religious importance and digging into it would touch off a firestorm of controversy.

As a result, excavations carried out in Jerusalem are often quite limited and piecemeal, undertaken only when an opportunity happens to present itself. This occurred in 1948, for example, when, during the Arab-Israeli war, a mortar hit a building that, according to tradition, housed the tomb of King David. In the process of repairing the damage inflicted by the explosion, the archaeologist Jacob Pinkerfeld (1990) also did a cursory excavation of one portion of the structure. Beneath the floor, he found over half a meter of debris, including three earlier floors. In fact, further investigations have suggested that the walls of this building were part of the Church of the Apostles, built in the first century AD at the place where the Last Supper was thought to have occurred. The walls have been destroyed and rebuilt many times, but because of the site’s importance to those who believe it was the tomb of David, the excavation was hindered and no further investigation was possible.

Still, although the city cannot be systematically excavated, modern instruments can probe beneath the surface to tell us not how old the layers are or what artifacts are in them but at least whether specific parts of the city are founded on archaeological debris or on solid ground. A government study by Israel’s Geological Survey was released in 2004, confirming what archaeologists had long suspected: the Old City is mostly founded on the rubble of previous constructions. As discussed in chapter 2, this makes the Old City particularly vulnerable to earthquake damage.

The historical record bears this out. In nearly every historical report of earthquakes affecting Jerusalem, damage is reported either to the Temple, the Old City walls, or both. Of course, the importance of this site and its role as the focus of religious pilgrimages for the world’s three major religions guarantees a record that is unparalleled in any other region. We have written records from many sources for earthquakes in Jerusalem, most of which also affected large regions of the countryside.

A recent geological investigation of sediments from the Dead Sea (Ken-Tor et al. 2001), confirms the validity of many of the historical reports mentioned in this chapter. As described in chapter 2, when shaking of sea-floor sediments is severe enough, the loose, water-saturated sediments lose their strength and flow like a liquid. When the shaking stops, the sediments settle and resolidify, leaving behind a chaotic, mixed layer that is readily identified, a seismite. These layers frequently contain organic material that scientists can date by Carbon-14 dating.

The 2001 study examined layers from the Ze’elim Terrace on the shore of the Dead Sea, a stack of sediments exposed by rapid modern drops in Dead Sea water levels (caused by diversion of the fresh water sources that feed the Dead Sea). There are gaps in this sequence, marking drought periods where the ancient water level in the Dead Sea was lower than usual and there was no sediment deposited at the Ze’elim Terrace site. However, during the periods when sediments accumulated there, every major historical earthquake on the Dead Sea Transform has been correlated to a seismite. This is a great new resource and one that can be extended in the future. The periods for which the Ze’elim Terrace contains no data could be examined by drilling into the deeper sediments in the Dead Sea floor. Thus, we can confirm the historical record at Jerusalem, and use it, along with the archaeological record in more accessible sites nearby, to build a physical earthquake stratigraphy for the region.

• from Israel Finkelstein in Adams et al. (2013 Vol. 3:1336-1337)

In contrast to the destruction at the end of the LB III, as well as the destructions in the late Iron IIA and Iron IIB, which were partial and encompassed only certain sectors of the settlement, the destruction of the late Iron I city was total. Evidence for this annihilation is now evident in every area excavated by our team and by our predecessors. The 2004-2008 seasons added the evidence from Levels H-9 and M-4.

For the moment, large numbers of radiocarbon results are available only for the late Iron I (Levels K-4, H-9 and M-4), while the number of determinations for Levels K-5, H-7 and H-5 does not allow the reaching of clear conclusions (many more samples for these and other Iron Age layers are now being processed and will be published in the future). Boaretto (Chapter 25) puts the destruction of late Iron I Megiddo in the first half of the 10th century or ca. 950 BCE (see also Finkelstein and Piasetzky 2007; 2009).

Many other late Iron I settlements in the northern valleys and the northern coastal plain came to an end in a violent conflagration. As is well known, scholars have suggested a single event for the end of this phase in the north: a major earthquake (Lamon and Shipton 1939:7; Marco et al. 2006; Cline 2011), King David's conquests (Yadin 1970: 95; Harrison 2004: 108), or Pharaoh Sheshonq I's campaign (Watzinger 1929: 58, 91; Finkelstein 2002).

Seven sites, five of which were destroyed by fire, supply 88 ¹⁴C determinations (for the data, see Finkelstein and Piasetzky 2007; 2009). When an uncalibrated date is established for each of these layers separately (by averaging its determinations — a legitimate procedure in the case of destruction layers, because one can assume that all samples came from the last years before the destruction), it becomes clear that the results do not provide similar dates. Rather, the uncalibrated dates fall into two groups that show geographical logic: the western Jezreel Valley-Acco Plain sites and the eastern Jezreel Valley-Sea of Galilee sites. The two groups are separated by 58 uncalibrated years (2852±13 and 2794±10 respectively), that is, 3.5 σ; the probability that they represent a single date is therefore low.

Most likely, then, the late Iron I horizon in northern Israel came to an end in two main events, or two clusters of events, in 1047-996 BCE and 974-915 BCE according to the `uncalibrated weighted average' method (Finkelstein and Piasetzky, 2007); 1017-984 and 969—898 BC according to a Bayesian model constructed for this purpose (idem. 2009).⁷

This conclusion renders the earthquake and single military campaign theories invalid. The most reasonable historical reconstruction would explain this gradual destruction in the north as representing early steps in the rise of a north Israelite territorial entity — the expansion of the highlanders into the northern valleys.

• Fig. 5.2 - Plan of Level H-9
  Figure 5.2
  
  Plan of Level H-9
  
  Finkelstein et al. (2013 Vol. 1)
  from Finkelstein et al. (2013 Vol. 1)
• Fig. 5.3 - Destruction Layers H-5 and H-9
  Figure 5.3
  
  Southern section of Area H, looking south. The figure is standing on the floor of Level H-9. From bottom: the thick accumulation debris of Level H-9, the floors of Level H-7 and the destruction of Level H-5. (Note the differences in the nature of destruction between Levels H-9 and H-5.)
  
  Finkelstein et al. (2013 Vol. 1)
  from Finkelstein et al. (2013 Vol. 1)
• Fig. 5.4 - South section of
  Figure 5.4
  
  South section of Area H (Squares E—F/6).
  
  Finkelstein et al. (2013 Vol. 1)
  Area H from Finkelstein et al. (2013 Vol. 1)
• Fig. 5.5 - East section of
  Figure 5.5
  
  East section of Area H (Squares F/6-9).
  
  Finkelstein et al. (2013 Vol. 1)
  Area H from Finkelstein et al. (2013 Vol. 1)
• Fig. 5.6 - West section of
  Figure 5.6
  
  West section of Area H (Squares E/6-9).
  
  Finkelstein et al. (2013 Vol. 1)
  Area H from Finkelstein et al. (2013 Vol. 1)
• Fig. 5.7 - Eastern section of
  Figure 5.7
  
  Eastern section of Squares E/6-9.
  
  Finkelstein et al. (2013 Vol. 1)
  Area H from Finkelstein et al. (2013 Vol. 1)
• Fig. 5.8 - Photo of Level H-9
  Figure 5.8
  
  General view of Level H-9, looking north.
  
  Finkelstein et al. (2013 Vol. 1)
  from Finkelstein et al. (2013 Vol. 1)
• Fig. 5.9 - Smashed storage jar
  Figure 5.9
  
  Smashed storage jar lying on the floor of Courtyard 08/H/38 of Level H-9 (note charred beam between mudbricks), looking east.
  
  Finkelstein et al. (2013 Vol. 1)
  and charred beam in Level H-9 from Finkelstein et al. (2013 Vol. 1)
• Fig. 5.10 - Complete storage jar in Level H-9
  Figure 5.10
  
  A complete storage jar on Paved Floor 06/H/51 in the northwestern part of Courtyard 08/H/38 of Level H-9 (note the depth of City Wall 325), looking east.
  
  Finkelstein et al. (2013 Vol. 1)
  from Finkelstein et al. (2013 Vol. 1)
• Fig. 5.11 - Smashed pottery vessels in Level H-9
  Figure 5.11
  
  Smashed pottery vessels in the eastern part of Courtyard 08/H/38 with Partition wall 06/H/8, Level H-9, looking east.
  
  Finkelstein et al. (2013 Vol. 1)
  from Finkelstein et al. (2013 Vol. 1)

• Table 4.1 - The Stratigraphy of Area M
  Table 4.1
  
  The Stratigraphy of Area M
  
  Finkelstein et al. (2013 Vol. 1)
  from Finkelstein et al. (2013 Vol. 1)
• Fig. 4.20 - Plan of Level M-4
  Figure 4.20
  
  Plan of Level M-4.
  
  Finkelstein et al. (2013 Vol. 1)
  from Finkelstein et al. (2013 Vol. 1)

• from Israel Finkelstein, David Ussishkin, and Robert Deutsch in Finkelstein et al. (2006:87)

The previous publication of the 1996 season in Area K (Lehmann et al. 2000) focused on the remains of Levels K-3 and K-2, with a short reference to Level K-4, which at that time was still under excavation. The current report focuses on Levels K-5 and K-4, with a brief reference to Level K-6. A full description of Level K-6 will be published in the next Megiddo report.

In the 1998 season the excavation of Area K was directed by Yuval Gadot, with the assistance of Assaf Yasur-Landau and Yulia Gottlieb. The square supervisors were Jonathan David, Sylvere Guirec, Hal Bonnette, Jesse Halpern and Christian Rata. Carolyn Higginbotham assisted in coordinating the work in the fine-grid squares. The main goal of the 1998 season was to continue working in the nine squares excavated in 1996 (M-O/9-11), in order to reach the floors of Level K-4. This goal was a direct result of the 1996 season, in which most of the remains belonging to Levels K-3 and K-2 were recorded and then excavated. By 1996 it was already clear that Level K-4 had been violently destroyed. In some places the accumulation of the destruction layer, including the collapse, was over 1 m thick. Following the 1998 season, after the Level K-4 floors were reached in most squares (Fig. 7.1), the inside baulks were excavated in an orderly manner under the supervision of Jennifer Peersmann, making it possible to obtain a complete plan of the Level K-4 remains.

The 2000 season in Area K was directed by Yuval Gadot with the assistance of Noga Blockman and Julye Bidmead. The square supervisors were Bill Baillie, Hal Bonnette, Jonathan David, Carolyn Higginbotham and Jennifer Koosed. Emma Frigeri assisted in drawing top plans. During this season three more squares (P/9-11) were added to the east,¹ right above the slope and inside the artificial railway terrace built by the University of Chicago excavators. After exposing the Level K-4 floors in these squares the new baulks were removed, thus connecting the new squares to the rest of the area and turning Area K into a wide 3 × 4 squares sectional trench on the edge of the mound (Fig. 7.2). In the remaining nine squares work continued in order to clarify the subtle stratigraphic problems related to Level K-4.

The 2002 season in Area K was directed by Noga Blockman and Mario Martin, with the assistance of Eran Arie. The main targets of this short, four-week season were to finish excavating the remains of Level K-4 and reach earlier strata. Work was undertaken in all twelve squares. Two architectural phases were exposed below Level K-4, termed K-5 and K-6 (Fig. 7.3). The former is rather dull in nature and both are domestic in character. Level K-5 dates to the early Iron I and should probably be equated with the University of Chicago’s Stratum VIB. The excavation of Level K-6 has not been concluded; the finds seem to date to latest phase of the Late Bronze Age.

• from Israel Finkelstein, David Ussishkin, and Robert Deutsch in Finkelstein et al. (2006:102-103)

In Megiddo III Levels K-3, K-2 and K-1 were equated with Strata VB, VA-IVB and IVA respectively of the previous excavations (Lehmann et al. 2000:137). While the overall stratigraphy of the University of Chicago team has been accepted, Levels K-3 and K-2 were separated into sub-phases not previously noticed.

Levels K-5 to K-4 also fit the general stratigraphic scheme established by the University of Chicago team and should be equated with Strata VIB and VIA respectively. The latter was destroyed by a violent fire, leaving typical red-coloured burnt bricks all over the mound (Loud 1948:33). The division between Strata VIB and VIA was established only in Areas AA and DD. In the former, a large edifice (2072) of Stratum VIA replaced more modest domestic structures of Stratum VIB, with no destruction layer between the two. Area K features the same pattern: The more elaborate Building 00/K/10 of Level K-4 (University of Chicago Stratum VIA) replaced the less developed Building 02/K/36 of Level K-5 (Stratum VIB).

The pottery of Levels K-5 and K-4 can be classified as early and late Iron I respectively (Chapter 13). Judging from the pottery, Level K-6 can apparently be equated with Stratum VIIA. Another possibility is to equate Level K-6 with the post Stratum VIIA transitional phase which was possibly detected in the excavation of the Late Bronze gate (Finkelstein and Zimhoni 2000:242-243). The final verdict on this matter will have to wait until the full exposure of this layer and the excavation of the remains below it in the coming seasons.

The stratigraphy in Area K is summarized in Table 7.1.

Table 7.1: Summary of Stratigraphy in Area K

Area K Level	The University of Chicago stratum	Period	Notes
K-6	‘VIC’ or VIIA	End phase of Late Bronze	No wholesale destruction
K-5	VIB	Early Iron I	Locally made Myc. IIIC vessel
K-4	VIA	Late Iron I	Destroyed in violent fire
K-3	VB	Iron IIA	Two main phases
K-2	VA-IVB	Iron IIA	Two main phases
K-1	IVA	Iron II	City Wall 325

• Fig. 1 Map of the site
  Fig. 1
  
  Aerial view of Megiddo, looking north, indicating the excavation areas discussed in the article
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  with excavation areas from Kleiman et al. (2023)
• Fig. 2 Aerial view of
  Fig. 2
  
  Aerial view of Squares H–I/4–5 in Area Q at the end of the 2016 season, looking north
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  Squares H–I/4–5 in Area Q from Kleiman et al. (2023)
• Fig. 4 Plan of Level
  Fig. 4
  
  Plan of Level Q-9 of the Late Bronze III (this phase was revealed only in limited probes below the floors of Level Q-8)
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  Q-9 of the Late Bronze III from Kleiman et al. (2023)
• Fig. 6 Plan of Level
  Fig. 6
  
  Plan of Level Q-8; shaded walls signify elements built in a previous occupational phase and reused
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  Q-8 from Kleiman et al. (2023)
• Fig. 8 Plan of Level
  Fig. 8
  
  Plan of Level Q-7b; shaded walls signify elements built in a previous occupational phase and reused
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  Q-7b from Kleiman et al. (2023)
• Fig. 10 Plan of Level
  Fig. 10
  
  Plan of Level Q-7a; shaded walls signify elements built in a previous occupational phase and reused
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  Q-7a from Kleiman et al. (2023)

• Fig. 1 Map of the site
  Fig. 1
  
  Aerial view of Megiddo, looking north, indicating the excavation areas discussed in the article
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  with excavation areas from Kleiman et al. (2023)
• Fig. 2 Aerial view of
  Fig. 2
  
  Aerial view of Squares H–I/4–5 in Area Q at the end of the 2016 season, looking north
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  Squares H–I/4–5 in Area Q from Kleiman et al. (2023)
• Fig. 4 Plan of Level
  Fig. 4
  
  Plan of Level Q-9 of the Late Bronze III (this phase was revealed only in limited probes below the floors of Level Q-8)
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  Q-9 of the Late Bronze III from Kleiman et al. (2023)
• Fig. 6 Plan of Level
  Fig. 6
  
  Plan of Level Q-8; shaded walls signify elements built in a previous occupational phase and reused
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  Q-8 from Kleiman et al. (2023)
• Fig. 8 Plan of Level
  Fig. 8
  
  Plan of Level Q-7b; shaded walls signify elements built in a previous occupational phase and reused
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  Q-7b from Kleiman et al. (2023)
• Fig. 10 Plan of Level
  Fig. 10
  
  Plan of Level Q-7a; shaded walls signify elements built in a previous occupational phase and reused
  
  (courtesy of the Megiddo Expedition).
  
  Kleiman et al. (2023)
  Q-7a from Kleiman et al. (2023)

• from Kleiman et al. (2023:3)

Area Q features a stratigraphic sequence of more than ten layers and sub-phases, covering the era from the end of the Late Bronze Age to the Iron IIB (Figs 1– 2; Table 2). Previous studies of the finds concentrated on the Iron II strata, mainly from the northern part of this area (see list of references in Homsher and Kleiman 2022). In 2012, the excavation was extended to the southern part of the area (Squares H–I/1–4), where a heavy rainstorm had exposed the ruins of the Iron I city. The original objectives were to:

1. re-study the destruction of the city using a variety of micro-archaeological approaches (Forget and Shahack-Gross 2016; Forget et al. 2015; Shahack-Gross et al. 2018)
2. re-examine the date of Schumacher’s Südliches Burgtor (Homsher and Finkelstein 2018), a monumental structure erroneously interpreted as a gate complex (Schumacher 1908: 77–80).

These excavations also revealed five short-term phases, radiocarbon dated from the late 12th century to the 10th century BCE, providing unique snapshots of the events leading up to the destruction of the Iron I city

• from Kleiman et al. (2023:3-4)

The events associated with the end of the Late Bronze Age at Megiddo have been discussed extensively (e.g., Ussishkin 1995). New evidence from the current excavations in Areas H and K (Arie 2022; Martin 2022, respectively), as well as a re-evaluation of the data from Area AA (Finkelstein et al. 2017: 263–64), shed light on the events. These studies demonstrate that the destruction of the Late Bronze III city (Stratum VIIA) in the second half of the 12th century BCE was partial and did not lead to radical changes in either the spatial organization of the settlement or its material culture (as already hinted by Engberg 1940; see also Esse 1992: 84, n. 21; Ussishkin 1995: 260–61). To date, signs of destruction were documented mainly in the area of the palace in Area AA and partially in Area M (Level M-6, see Finkelstein 2013a: 234), as well as in the domestic quarter in Area K (Level K-6, see Arie and Nativ 2013).

In Area Q, current excavations reached the Late Bronze III (Level Q-9) only in limited probes dug below the floors of Level Q-8 in Squares I/1 and I/2 (Fig. 4). A few complete vessels found on beaten-earth surfaces to the east of the Südliches Burgtor suggest a small-scale disturbance (Fig. 5). Due to the limited assemblage and exposure, it is difficult to securely date this phase. Based on a preliminary review of the pottery, Homsher and Finkelstein (2018: 305) suggested correlating Level Q-9 with Level H-11 of the Late Bronze III, or the beginning of the Iron I (for Level H-11, see Arie 2022: 97–109; Finkelstein et al. 2017: 267–69).

• from Kleiman et al. (2023:4-5)

More apparent, but still restricted in extent, are the remains of Level Q-8 (Fig. 6). In Squares I/1–3, the finds associated with this phase were limited to two ashy surfaces between the floors of Levels Q-9 and Q-7b (Fig. 7). One of these surfaces abuts the eastern wall of the Südliches Burgtor, indicating the continuing function of this monumental building at the beginning of the Iron I, corroborating the results of the OI excavations in nearby Area CC (see below). Fragmentary remains of domestic structures belonging to this phase were exposed in Squares H/4–5, c. 15–20 m to the east of the Südliches Burgtor. The floors associated with these buildings include patchy phytolith surfaces and dark grey ash layers found c. 45 cm below the floors of Level Q-7b. A few restorable jars were found in these structures, testifying to another small-scale disturbance. The pottery of Level Q-8 is generally similar to that of Levels Q-7b and Q-7a, suggesting its possible correlation with Stratum VIB of the OI, which was allegedly absent from the nearby Area CC (Harrison 2004: 19; Loud 1948: 113).

• from Kleiman et al. (2023:5-6)

Level Q-7b portrays a substantial change in Area Q. It is represented by the construction of Building 14/Q/ 53, apparently an eastern expansion of the Südliches Burgtor (Figs 8–9) and Building 14/Q/145 further to the east. The remains of the former structure suggest the existence of a sizable unit that measures about 7.5 × 8.0 m. It consists of three large and at least partially paved rooms, constructed in lesser quality compared to the massive walls of the Südliches Burgtor. The new addition nearly doubled the size of the previous structure and could be considered part of an ad-hoc defence system. The remains of Building 14/Q/145 in Squares H–I/4 represent a small structure, of which only one wall, a semi-circular fire installation and a beaten-earth floor survived. Exceptional finds in this context were three carefully cut, small octagonal stone pillars, arranged as an equilateral triangle; the tops of all three were cut intentionally (Kleiman et al. 2017: 26; see Fig. 9: a–b). Possible parallels to these pillars can be found in the Level VI temple, in Area P, at Lachish, where they are assumed to have functioned as part of two adjoining cult niches (Arie 2016; Ussishkin 2004: 231–38, pl. 6.29). At Megiddo too, these finds seem to be part of a small shrine, although no other strong indication of cult-related activity was traced here (but for broken basalt stelae reused in Level Q-7a, see below).

The separation of the ceramic assemblages of Level Q-7b and Q-7a was difficult as some of the structures associated with the former phase continued to function through both levels, until the destruction of the city (e.g., Building 14/Q/53). Still, there is enough evidence to assign Level Q-7b to the late Iron I, probably a short while before the destruction of Stratum VIA.

• from Kleiman et al. (2023:6-9)

In the last phase before the destruction of the city, designated Level Q-7a, a large stone-paved structure (Building 16/Q/48) was constructed above the remains of the building with the octagonal pillars (Building 14/Q/145), an action that represents a significant change in the area (Fig. 10). The new structure includes a c. 80-cm-wide wall, which was preserved to a length of almost eight metres. On both sides of this wall, stone pavements were laid. The eastern one included broken basalt slabs laid over a 30–50-cm-thick fill (e.g., Fig. 11:a).¹ These slabs may represent parts of broken stelae originating from a small Level Q-7b shrine (most likely Building 14/Q/145). Large quantities of mudbricks were found in Squares H/4–5, arranged in piles along what could be the northern wall of the structure. They may be understood as building materials collected in preparation for new construction. Both the construction technique and orientation of Building 16/Q/48 differ from the remains associated with Building 14/Q/53 of Level Q-7b. The eastern expansion of Schumacher’s Südliches Burgtor continued to function, but the pavement in its south-eastern room was replaced by a beaten-earth floor on which a small installation was built (Locus 12/Q/210). Some of the features associated with the previous phase were still visible in Level Q-7a, for example, the upper part of at least two of the three octagonal stone pillars.

Level Q-7a ended in the massive conflagration of Stratum VIA (Fig. 12). It includes burnt reddish mudbricks, charred beams and charcoal, and collapsed architecture. In Square I/2, the destruction debris reached more than a metre. Preliminary geoarchaeological analysis of the burnt bricks from the destruction indicates that they were exposed to up to 700–800°C (Forget et al. 2015; see also Forget and Shahack-Gross 2016).

• from Kleiman et al. (2023:9)

Following the destruction of Stratum VIA, the settlement may have been abandoned for a few decades at the very end of the Iron I. Two pieces of evidence support this possibility: 1) radiocarbon determinations, which indicate that sites in the eastern sector of the Jezreel/Beth-she’an Valley were destroyed later than Megiddo VIA and Yokne‘am XVII, but still within the late Iron I (Finkelstein and Piasetzky 2009: 266–67); and 2) the pottery of Yokne‘am XVI and possibly other sites in the north (e.g., Tel Kinrot), which hint at a post-destruction phase still within the Iron I (Arie 2011: 275; Münger et al. 2011: 87).

At any rate, Megiddo recovered in the early Iron IIA (Level Q-6b). The earliest activity that can be attributed to this period is represented by meagre architectural remains constructed directly above the ruins, with little continuity from the previous settlement (Homsher and Kleiman 2022: 122–27; Kleiman 2022: 937, e.g., fig. 23.41: 5, 9). The only exception here was a stone monolith that towered above the ruins and may indicate evidence of a ruin cult (Kleiman et al. 2017: 26; Fig. 13).

• from Kleiman et al. (2023:9)

Beyond architectural remains, several additional finds illuminate the ongoing processes leading up to the destruction of Stratum VIA:

1. human remains documented in Building 16/Q/48
2. a hoard of metal items near the eastern wall of the Südliches Burgtor, within Building 14/Q/45
3. faunal remains associated with Levels Q-7b and Q-7a, which reveal the subsistence economy of the local residents shortly before the destruction

• from Kleiman et al. (2023:9-11)

Fragmentary remains of a young adult, probable female (Fig. 14:a; for additional details, see Supplemental Material 1), were exposed in Baulk H–I/4 on a floor to the north-east of Building 16/Q/ 48 of Level Q-7a, directly below the destruction debris (Locus 16/Q/79). The individual was found oriented on a south–north axis, with the head in the north, but the exact body position was difficult to determine due to fragmentation.

All skeletal elements show evidence of burning; the range of colour being grey, white, blue and pink. The cranium was mostly grey (medium-high temperature), the bones of the axial skeleton were yellow-white (low temperature), while the bones of the appendicular skeleton (limbs) were yellow-white to grey (medium/high temperature) (Delvin and Herrmann 2015; Symes et al. 2015). One warp-fractured long-bone fragment was blue (high temperature), and the right temporal showed portions of pink discolouration (medium temperature) near the zygomatic process (Delvin and Herrmann 2015) (Fig. 15). A fragment of an un-sideable humeral head had a stark colour differentiation on its articular surface (Fig. 15), and colour differences were seen throughout the articulated proximal femur and right os coxa (Fig. 14:b), reflecting burn patterns consistent with the ‘tissue-shielding’ of a fleshed human body (Symes et al. 2015: 36, 54–55). The signs of warping, shrinkage and surface cracking were most pronounced on the right skeletal elements (Fig. 14:c–d), and indicate that the bone experienced prolonged heat damage in excess of 700 °C (Ubelaker 2009). Warping, irregular splitting and fractures most commonly occur in fleshed bodies (Baby 1954; Ubelaker 2009; Whyte 2001). Bone shrinks when heat denatures proteins in bone tissue (Symes et al. 2015: 46), and the degree to which bone shrinks increases exponentially at around 650 °C (Shipman et al. 1984).

Bones from the individual and sediments above and below the skeleton were analysed by Fourier Transform Infrared (FTIR) spectrometry to refine burning temperature estimates and confirm the depositional relationship between the destruction and the individual. FTIR is well established as a means of estimating firing temperature based on changes to clay minerals (Berna et al. 2007; Forget et al. 2015: 81 and references therein) and calcite (Regev et al. 2010), and has been used extensively at Megiddo to study destruction processes (Forget et al. 2015; Regev et al. 2015; Shahack-Gross et al. 2018). FTIR is also commonly used to explore bone properties, including diagenesis, preservation and burning (Ellingham et al. 2015; Weiner 2010). Bone samples for FTIR analysis were assigned burn codes following Stiner et al. (1995), and both bone and sediment samples were analysed between 4000 and 400 cm-1 at a 4 cm-1 resolution using a Thermo Scientific Nicolet iS5 spectrometer at the Laboratory for Sedimentary Archaeology, University of Haifa (details in Supplemental Material 2). The bone mineral spectra corroborate the macroscopic analysis, indicating that some bones were unburnt (or heated to temperatures below 200 °C) and others were exposed to higher temperatures (>600 °C) (Fig. 16, more details in Supplemental Material 2). This is consistent with the tissue shielding identified macroscopically. The clay minerals in sediments above and below the skeleton were exposed to temperatures of 700–800°C and ∼600 °C respectively, similar temperatures to the bones, demonstrating that the body was burnt in situ. This provides strong evidence that this individual was burnt — and likely died — in the destruction event.

• from Kleiman et al. (2023:11-12)

In 2012, a hoard of metal objects was discovered near the eastern wall of the Südliches Burgtor (Level Q-7a; Fig. 17). It included two stacked bronze bowls, beside which eight or nine iron blades were found (some of them are bimetallic; see details in Hall 2021). Bronze scale pans, jewellery and other items were found inside the bronze bowls, the outer of which contained remains of textile on its exterior. It appears that both the bowls and the blades were wrapped in textiles before final deposition. Was this hoard buried in a ritual deposit, intended for permanent deposition, or as a cache for temporary storage with the intent of future retrieval? There are several reasons to suggest that this was a foundation/building deposit (see also Bjorkman 1994: 7–8):

1. the hoard itself contained a bird’s beak, which is best explained as an apotropaic offering
2. bimetallic knives, such as those uncovered in the hoard, are often found in association with ritual contexts (Dothan 2002: 14– 17; Mazar 1985: 6–8, fig. 2). However, the fact that the items were arranged neatly and buried in a textile may point to the non-ritual status of the hoard, i.e., its burial with the intention of future retrieval.

One way or another, a substantial number of examples of non-ritual hoarding activities in Stratum VIA exist at Megiddo (Hall 2016: table 5.1; 2021: table 1). The rise in hoarding at Megiddo before destruction events seems to represent ‘crisis behaviour’ (more below), which was spurred by the instability of the inhabitants’ socio-political and economic surroundings. Crisis hoards are found elsewhere in the Eastern Mediterranean, especially prior to the Late Bronze Age collapse, and are representative of a response to the failure of the socio-political systems (Knapp et al. 1988)

• from Kleiman et al. (2023:12-15)

A total of 280 animal bones (macrofauna) were retrieved from the floors of Levels Q-7b and Q-7a. Identification of fragments to skeletal elements and the lowest taxonomic level was achieved using the comparative collections stored at the Zooarchaeology Laboratory and at The Steinhardt Museum of Natural History of Tel Aviv University. All skeletal fragments were recorded, i.e., epiphysis as well as diaphysis. Long-bone fragments were coded according to the completeness of five morphological zones (proximal and distal epiphysis, proximal and distal diaphysis and mid-shaft diaphysis). Other bone fragments were coded according to their percentage of total completeness of element. Quantifying species and body parts frequencies were based on NISP (Number of Identified Specimens).

The assemblage is dominated by caprines (sheep (Ovis aries) and goats (Capra hircus); 80% of the assemblage; identification based on morphological criteria following Zeder and Lapham 2010), followed by cattle (Bos taurus; 18%) and supplemented with pig (Sus scrofa) and game animals (NISPs in Fig. 18). Unlike the situation in Levels H-9 and K-4 of the same late Iron I horizon (Sapir-Hen et al. 2016: table 1), the caprine assemblages include a higher representation of goats to sheep (a ratio of 3:1). The age profile of these caprines, based on epiphyseal fusion (following epiphyseal fusion sequences of Zeder 2006), shows a focus on young and young-adult animals; very few animals were kept beyond three years (table 4 in Supplemental Material 3). This low frequency of older animals implies that caprines were probably mainly exploited for their meat, and not kept for their secondary products. Skeletal elements frequencies of the caprines (table 5 in Supplemental Material 3) indicate that the upper parts of the fore and hind limbs are highly represented, more than the lower parts of these limbs and extremities, suggesting a higher presence of parts that are rich in meat, although all body parts are represented to some extent. Unfortunately, although cattle constitute 18% of the assemblage, the small sample did not allow for the analysis of age profiles and body parts. Still, there is a representation of young-adult and adult animals (table 6 in Supplemental Material 3).

The profile of caprines and cattle exploitation for Level Q-7 resembles the contemporaneous assemblage from Area H (Level H-9) and differs from the one found in Area K (Level K-4), located only a few metres away from Area Q. A previous study demonstrated that the people of Areas H (near the palace and city-gate) and K (in the south-eastern sector of the site) had differential access to animals and their products (Sapir-Hen et al. 2016). The inhabitants of Area H were provisioned with good cuts of caprine meat (meat-rich parts of young animals) and were not engaged in agriculture, while the lower-status inhabitants of Area K raised and consumed their own livestock. The observation that the access to animals in Area Q is similar to that in Area H is supported by other finds, showing a similar public-related function of the two areas at the end of Iron I. However, a notable difference between Areas Q and H is the preference for goats over sheep in Area Q. Among the two caprine species, sheep are considered more expensive (Redding 1981; 2010) and were preferred by the elites of the southern Levant (Sapir-Hen forthcoming; Sapir-Hen et al. 2022). Possibly, the higher frequency of goats, the less expensive caprine species that is also more reliable and less sensitive to disease (Redding 1981; 1984), fits other clues for ‘crisis behaviour’ observed in the south-eastern sector of the mound. It is acknowledged, nonetheless, that the interpretation of this specific pattern is more difficult than other lines of evidence presented here.

• from Kleiman et al. (2023:16)

An overview of the remains associated with Stratum VIA across the mound is provided elsewhere (Arie 2011: 89–95; Esse and Harrison 2004; Finkelstein 2009: 115–16; Finkelstein et al. 2017; Homsher and Finkelstein 2018: 297–99). Here only remains exposed immediately to the east and west of Area Q are discussed.

• from Kleiman et al. (2023:16)

In Area CC of the OI’s expedition, finds associated with the Iron I were exposed under the residential quarter associated with Stratum V of the Iron IIA (Harrison 2004; Lamon and Shipton 1939: 3–4, fig. 5). No remains were assigned to Stratum VIB, although small-scale changes in the domestic architecture of Stratum VIA were noted (Harrison 2004: 19; Loud 1948: 113). This area is also the location of the Südliches Burgtor mentioned above. Some of the massive walls of this structure were preserved to a height of more than a half-metre. Homsher and Finkelstein (2018) discussed the history of the Südliches Burgtor in detail and concluded, primarily based on the new evidence from Area Q, that the earliest possible date for its construction was around the Late Bronze III, and that it functioned until the destruction of Stratum VIA. The latter observation corresponds to Schumacher’s description of a thick accumulation of yellowish-red burnt bricks with burnt wood and ash (1908: 80; see also the Iron I ceramic finds in Taf. XXII), as well as to the attribution of the structure to Strata VII and VI in Megiddo II (Loud 1948: 409–10; see also Harrison 2004: 9). As most of the structure was excavated at the beginning of the 20th century, it is difficult to reassess its function, but it may have been another temple in the city (details in Homsher and Finkelstein 2018).

Similar to other areas at Megiddo, this residential quarter was utterly destroyed at the end of the Iron I. This is evident in the wealth of finds, the high number of skeletons found within the ruins and the piles of burnt debris (Harrison 2004: 20, 23; for the accumulation of debris in this area, see also Sections P–Q and R–S in Lamon and Shipton 1939: fig. 35). An exceptional phenomenon observed here is the burials of individuals prior to the destruction of the city (see, e.g., Fig. 21; and description in Harrison (2004: 20); for a full list and re-evaluation, see Table 4). These are the first intramural burials observed in the city since the beginning of the Late Bronze Age when this characteristic Middle Bronze Age practice ceased (Martin et al. 2022a: 32; 2022b; for earlier intramural burials in Area CC, see Cradic 2018: 196–97, table 3). P. L. O. Guy (in a letter from Guy to H. Breasted quoted in Harrison 2004: 8–9) and Finkelstein et al. (2000: 260), raised the possibility that the burials in Area CC were either cut into the destruction debris of Stratum VIA, or belonged to an earlier phase in the Area CC sequence (see also Cline 2011: 60; Gadot and Yasur-Landau 2006: 586; Kreimerman 2021: 236). A fresh look at photos published by Harrison (2004) after the publication of Megiddo III refutes this interpretation and indicates that these burials predate the destruction of Stratum VIA (see more below). The sudden reintroduction of intramural burials at the site, nearly half a millennium after they had decidedly gone out of fashion, is an important clue as to what was happening in the city prior to its destruction.

• from Kleiman et al. (2023:16-17)

Remains of a domestic courtyard house belonging to Level K-4 were exposed in Area K, situated c. 30 m to the east of Area Q (Arie 2006; Gadot and Yasur-Landau 2006: 586; Gadot et al. 2006). The finds included large quantities of restorable vessels within and below a thick layer of burnt destruction debris. Essentials for the current discussion are the following finds:

• a human skeleton in the western courtyard of the house (Square M/9; Locus 98/K/40). A nearly complete krater covered the skull of this individual (Arie 2006: 196–97; Gadot and Yasur-Landau 2006: 586; Gadot et al. 2006: 101; see also Harrison 2004: 8–9).


• remains of a child in the south-western room of the house (Baulk N–O/9; Locus 98/K/100). The skull of this individual was encircled by three small stones (Sameora and Adams 2022: table 14.1).


• a child burial in the northern part of the area (Baulk N–O/11; Locus 04/K/38; Fig. 22). In the original report, this find was associated with Level K-5 (Gadot et al. 2006: 92), as it was found on top of one of the walls of Level K-6 (Wall 04/K/7; for the association of this burial with Level K-4, see below).

Previous studies assumed that the first individual mentioned above (Burial 98/K/40) was interred after the destruction of the settlement, presumably by people who returned to the site (Arie 2006: 196–97; Gadot and Yasur-Landau 2006: 586; Gadot et al. 2006: 101). In retrospect, no evidence of a pit cut through the destruction debris, or any other clue of post-destruction activity was found here, and thus, the possibility that this individual was buried before the event must be considered.² In contrast, Burial 98/K/ 100, the second burial listed above, was clearly found below the collapse, meaning that the child must have been buried before the destruction of Stratum VIA.³ As for Burial 04/K/38, it is reasonable to argue that it was dug from the floor of Level K-4 down, an option which was already raised by the excavators (Gadot et al. 2006: 92). Other skeletal remains that likely belong to pre-destruction burials were found in additional contexts in Area K (e.g., Loci 98/K/64 and 00/K/11; see details in Table 4), but were not well preserved. All of this means that, along with individuals trapped in the destruction of Stratum VIA, several burials preceding the event can be discerned in Area K, a trend that aligns with the evidence from Area CC described above.

• from Kleiman et al. (2023:17-19)

The picture which emerges from the south-eastern sector of the mound fits well with the finds from other parts of the site (e.g., Areas AA and H), but also presents several distinct local trends. Most important is the phenomenon of pre-destruction burials, documented in Areas CC (15 individuals) and K (3– 5 individuals), which has not been observed in the northern sector of the mound (details in Table 4).⁴ Most of the buried individuals were placed in an extended supine position in shallow pits (see especially, Harrison 2004: fig. 73); the position of the clavicles and the tight placing of the ankles suggest that some of them were buried in shrouds (Duday et al. 2009). Similar to earlier Bronze Age mortuary traditions in the southern Levant (Martin et al. 2022b: 244), the pre-destruction burials in Stratum VIA were interred along the axes of wall foundations (Harrison 2004: fig. 10), indicating purposeful placement prior to the coverage of the architecture by debris. This evidence makes a strong case against identifying these burials as dug into the ruins after the destruction. All this suggests a unique behaviour, distinctive to the south-eastern sector of Megiddo.

Another distinct phenomenon documented in the south-eastern sector of the mound is that of individuals, including children and women (i.e., families), who were trapped beneath the collapse: two or three skeletons in Area CC, up to five in Area K and one in Area Q (Fig. 23). Each individual is characterized by atypical body position, similar to those found in association with the Late Bronze III destruction at Tel Azekah in the Judean Shephelah (Berendt et al. 2021; see also the review in Kreimerman 2017). At least in the case of the Area Q individual, the bones are unequivocally burnt, as indicated by colour, morphology and FTIR analysis. In contrast, excavations of Iron I contexts in the northern sector of the mound resulted in only a single bone fragment, probably of a child (Area H, see Sameora and Adams 2022: table 14.1). However, this fragment may be a relic related to earlier burial activities (e.g., Cradic 2018; Martin et al. 2022b) and/or later disturbances. In theory, one could argue that the exposure of Stratum VIA in the south-eastern sector of the mound was more extensive than in the north (especially in light of the large extent of Area CC), and that is why clear evidence for trapped individuals was only found there. However, the consistent discovery of skeletal remains in all three areas excavated in this sector vis-à-vis the absence of such finds elsewhere in the city, especially in the widely-exposed Area AA,⁵ is significant and important for the reconstruction of the events preceding the destruction of the Iron I city. According to one possible scenario, in the final days of Stratum VIA, the city’s population concentrated in the south-east, perhaps because this was the highest point in the town (Homsher and Kleiman 2022: 119, table 5.1).

Excavations in Area Q also suggest that the crisis was not limited to the final days of Stratum VIA, but may have begun a few decades earlier, as evidenced, for instance, by rapid architectural changes in the vicinity of monumental buildings. In particular, two crucial architectural transformations are noted: 1) the expansion of the Südliches Burgtor to the east (Building 14/Q/53); and 2) the establishment of a small shrine near it (Building 14/Q/145) and its quick decommission in favour of a larger structure (Building 16/Q/48), the construction of which was never completed; no other structure in Stratum VIA was discovered in such a state (Fig. 24). Another piece of evidence that might be related to the city’s crisis is the peculiar preference for goats over sheep, as observed in the faunal remains of Area Q.

• from Kleiman et al. (2023:19-24)

Over the years, scholars have remained undecided regarding the cause of Megiddo VIA’s destruction. Some scholars theorized devastation by a natural disaster, i.e., an earthquake (e.g., Gadot and YasurLandau 2006: 583; Kempinski 1989: 89–90; Lamon and Shipton 1939: 7; Marco et al. 2006: 572; Mazar 2007: 85; Cline 2011 with references to earlier studies). One of the main arguments in favour of this scenario has been the alleged absence of evidence for weapons (e.g., arrowheads or slingstones) in Stratum VIA (Cline 2011: 65; Fiaccavento 2014: 222), similar to the cases of Stratum III at Lachish (Gottlieb 2004) or Stratum II at Beersheba (Gottlieb 2015). A review of the registration records published in the final reports reveals, nonetheless, that more than 20 arrowheads were found in Stratum VIA (Blockman and Sass 2013: 889, fig. 15.8: no. 405; Harrison 2004: 85–86; Gadot and Finkelstein 2000: 626, without illustrations) along with more than 100 slingstones (Sass and Cinamon 2006: 398, nos 688– 690; Blockman and Finkelstein 2006: 449–54, without illustrations). Furthermore, the majority of destructions in the southern Levant during the Bronze and Iron Ages, including those unanimously assumed to be destroyed by a human agent, did not produce similar evidence to that of Lachish and Beersheba (Kreimerman 2016: 234–35), for instance, Stratum XIII at Hazor (Zuckerman 2007) or Stratum IV at Tel Rehov (Mazar 2020: 126). Lastly, it can be emphasized here that Marco et al. (2006), who noted several possible indications of an earthquake in the finds from Stratum VIA (e.g., titled walls in Area K), ultimately defined the physical evidence as inconclusive (see also P. L. O. Guy cited in Harrison 2004: 9; Mazar 2022: 10).

Other scholars argued that the destruction of Stratum VIA was caused by a human agent (e.g., Albright 1936: 28–29; Arie 2011: 388–90; Finkelstein 2003: 78–89; 2013b: 34–36; Harrison 2004: 108; Watzinger 1929; Ussishkin 2018: 309–315, who stressed the total devastation of the city, something that is unlikely to happen in the case of an earthquake or an incident). The main argument against this model has frequently been the absence of clear signs of ‘crisis behaviour’ in the archaeological record of Stratum VIA (e.g., Arie 2011: 388). This term refers, inter alia, to a situation in which people react to a ‘deeply felt frustration or basic problem with which routine methods, secular or sacred, cannot cope’ (La Barre 1971: 11; see also Driessen 1995: 65; Zuckerman 2007). Crisis behaviour may be detected in architectural changes (Driessen 1995; Ikehara 2021), anomalous burial practices (Tamorri 2019) and hoarding (Fokkens 1997; Hall 2016; Knapp et al. 1988). In terms of architecture, unfinished building projects may signal socio-political breakdowns (Ikehara 2021). Monuments abandoned in the process of construction likewise hint at social and political unrest, or simply, the loss of control over the workforce. Abandonment may also reflect economic instability, with leaders unable to organize labour efficiently. In terms of priorities, in the lead-up to a crisis, leaders would likely abandon monumental building projects in favour of the construction of defensive works. Quick architectural changes and poor construction techniques may also point to crisis behaviour, wherein buildings are constructed hastily, signalling a lack of support from the ruling authorities (Driessen 1995: 67–76; Zuckerman 2007: 11). In the case of burials, deviant/ anomalous practices include the deposition of bodies outside of formal cemeteries, sometimes due to the lack of physical access (Tamorri 2019: 92). A dearth of grave goods associated with burials may also be a sign of economic depression or socio-political crisis. Regarding hoarding, some scholars see the increase in the number of hoards as a ‘failure of confidence’ in the ruling systems (Knapp et al. 1988: 258). At Megiddo, most non-ritual hoards are dated to the Late Bronze III and late Iron I, periods that saw partial or total destruction events (Hall 2016: 112).

All the above-mentioned characteristics of ‘crisis behaviour’ can be observed in Stratum VIA in the south-eastern sector of the mound:

1. Rapid changes in public architecture, including the enlargement of a monumental building (Schumacher’s Südliches Burgtor) using inferior construction techniques, the possible establishment of a small shrine next to it, and the quick decommissioning of the structure in favour of a much larger building, the construction of which was never completed (a distinct case in the architecture of Stratum VIA).


2. Abandonment by most of the local residents, as evident in the small number of people caught in the collapse (no more than ten individuals in all excavated areas across the site) and the possible concentration of the population in the inner, highest part of the settlement.


3. Pre-destruction intramural burials, suggesting the necropolis was inaccessible (for the finds from the necropolis, see Guy and Engberg (1938); and Arie (2011: 103–05)).


4. Caching of luxurious objects, e.g., the hoard in Area Q, possibly reflecting the abandonment of the town by individuals/families with economic means (as evidenced by the lack of grave goods discovered in the burials).


5. Clues for preference of goats in the faunal remains from Area Q, which are considered to be cheaper, more reliable and less sensitive to disease than sheep.

These points demonstrate that the crisis that befell Megiddo in the early 10th century BCE was anticipated by the local residents and that the agent of destruction of the city was human. Moreover, the available evidence hints that this was probably the culmination of a process which included a siege, rather than a sudden and unexpected military attack

• from Kleiman et al. (2023:24-25)

The devastation of Megiddo VIA is ultimately one of many destruction episodes which characterized the Iron I/II transition in the southern Levant (Mazar 2022: 10–11) and were often linked in the literature to historical or allegedly historical events (Fig. 25). Over the years, several historical scenarios have been offered for the Megiddo destruction, among them a long-term conflict with the neighbouring city of Ta‘anach (e.g., Knauf 2000, but see Arie 2011: 388), the military campaign of Shoshenq I (e.g., Finkelstein 2002: 120–22; Watzinger 1929), an early Israelite conquest (e.g., Harrison 2004: 108; Kochavi 1989: 15; Maisler 1951: 23; Mazar 1980: 46–47), or a Philistine territorial expansion (e.g., Albright 1936: 28–29). According to a more nuanced model, which underscored the radical changes that occurred in the city following the destruction, Megiddo was destroyed during the expansion of the highlanders into the valley, a development that soon brought about the emergence of the Kingdom of Israel (Arie 2011: 388–90; Finkelstein and Piasetzky 2007: 257–58). Memories of these events may have been embedded in pre-Deuteronomistic biblical texts, first and foremost the saviours’ story (and song) in Judges 4–5 and the Saul narrative in 1 Samuel (e.g., Arie 2011: 391–90; Finkelstein 2017; see also Engberg 1940: 6–7).

• from Nur and Burgess (2008:73-74)

The belief that one or more supernatural beings were responsible for natural cataclysms such as earthquakes, floods, volcanic eruptions, and hurricanes was intrinsic to every known society. Today rationalists often scoff at the notion, as does David Webster (2002), an anthropologist at Penn State University: “One can imagine that unpredictable catastrophic events like these caused a collective psychological panic of ‘the gods are angry’ kind, even in the absence of serious damage. To me such suppositions are fanciful to the point of absurdity.” Nevertheless, the recurring response of societies to large historical earthquakes has been a search for their meaning and an attempt to understand natural disasters in general. Even today, when the natural causes of earthquakes are well known, many people can not resist the urge to find greater meaning in natural calamity:

Such occurrences are, on the one hand, natural phenomena like hurricanes, flood, and tornadoes. On the other hand, however, such explanations, while rational, fail to provide us with an answer as to the ultimate cause. Why should the earth’s crust not remain completely static and stable? Why do some earthquakes, such as those mentioned in the Bible, occur with such pinpoint accuracy and timing so as to defy human explanation? Why would a loving God allow such disasters to happen? (Fast 1997)

The author of that passage goes on to propose that earthquakes are the result of man’s general rebellion against God.

Most biblical references to earthquakes suggest that the calamities are more directly the result of God’s pleasure or displeasure with man. Whether the hand of God was sending help or dealing retribution depended on the results of the quake, that is, who was hurt and who benefited. The following account of God aiding the people of King Saul in their battle against the Philistines at Michmach, ca. 1020 BC, appears in 1 Samuel 14:15–23:

And there was trembling in the host, in the field, and among all the people: the garrison, and the spoilers, they also trembled, and the earth quaked: so it was a very great trembling . . . and, behold, the multitude melted away, and they went on beating down one another. And there was a very great discomfiture . . . So the LORD saved Israel that day.

God could also send earthquakes as punishment for some short coming; even the famous story of the destruction of Sodom and Gomorrah may describe an earthquake, where the Lord “overthrew those cities, and all the plain, and all the inhabitants of the cities, and that which grew upon the ground (Genesis 20:24–25). So buried in antiquity is this story that no record exists of where these cities were located, although one possible site is the plain north of the Dead Sea, east of Jericho. Accompanying the description is an account of fire and brimstone raining from heaven, and smoke rising from the whole countryside; this could be simple poetic elaboration, but it could describe fires that started during the earthquake or the rising clouds of dust that always accompany earthquakes in such dry areas.

The religious language and ambiguity in nearly every biblical earthquake account may make us uncomfortable, when what we desire is a dispassionate, scientific account. However, our discomfort should not make us disregard these references as completely unreliable. We should not be surprised that the people affected by an earthquake, just like people today, try to find a way to give it some higher meaning in their lives.

• from Nur and Burgess (2008:189-199)

Figures Discussion

A modest mound southeast of the Carmel Ridge in Israel, the tell of Megiddo rises 50 meters above the surrounding Jezreel Plain and covers some 6 hectares (15 acres) of land (Figure 7.1). This tell was first identified as the legendary Armageddon [the Greek corruption of the Hebrew “Har Megiddo” or Mount Megiddo] by a Jewish writer of the fourteenth century, Esthori Haparchi, and then was rediscovered by the British army officer H. H. Kitchener five hundred years later. Extensive excavations were conducted at the site by C. Fisher, P. Guy, and G. Loud (Yadin 1975). These archaeological studies revealed physical evidence of the historical development of Megiddo.

Megiddo’s strategic importance, belied by its unimposing appearance today, stems from its unique topography (Figure 7.2). The land between the Mediterranean and the Jordan River served as a bridge between the civilizations in the South, in Egypt and Arabia, and those in the North, in Syria, Mesopotamia, and Anatolia. It was also a continuation of sea routes from the Mediterranean and the Gulf of Suez. The ruggedness of the region, however, crossed by several ranges of mountains and hills as well as the lowest valley on Earth, limited the possible routes for overland shipping or wheeled travel. The Carmel-Gilboa mountain range was a particular obstacle, and traffic from the Mediterranean to Syria and Jordan was funneled through a few gaps in the range. In fact, both the passes and the mountain range that obstructed traffic were products of tectonic motion along the seismically active Carmel Gilboa fault system, which branches off from the Dead Sea Transform (see Figure 4.5).

The mound of Megiddo stands guard over one of the most important of these mountain passes, the Nahal Iron Pass, a traffic bottleneck on the main route between Egypt and Syria. Until the advent of more elaborate road construction by the Roman Empire in the first and second centuries AD, the gap at Megiddo was the only one that permitted the passage of chariots, though it was not an easy passage. A description from the latter part of the thirteenth century BC, found in the Egyptian Papyrus Anastasi I, gives some idea of the difficulty of the route:

Thy path is filled with boulders and pebbles, without a toe-hold for passing by, overgrown with reeds, thorns, brambles, and wolf’s paw. The ravine is on one side of thee, and the mountain rises on the other. Thou goest on jolting, with thy chariot on its side, afraid to press thy horse too hard. . . . The horse is played out by the time thou findest night quarters. (Hori [Egyptian Royal Official], Papyrus Anastasi I)

This narrow and difficult route made the pass from Megiddo particularly easy to guard, so whoever held power in Megiddo controlled not only the course of trade in the Fertile Crescent but that of war as well. Thus, the site figured prominently in some of the greatest ancient battles fought in this region. As Pharaoh Thutmose III expressed it, “The capture of Megiddo is the capture of 1,000 towns.” Indeed, fortifications were built and rebuilt there for close to five millennia, until around 500 BC (Finkelstein and Ussishkin 1994).

Four levels of destruction in the mound of Tel Megiddo are consistent with earthquake destruction, the lowest one attributed to the conquest by Thutmose III in 1468 BC. Why, however, would the Pharaoh have destroyed the place if his goal were to occupy the site and exact tribute? Although it is clear that Thutmose III conquered Megiddo, there is no more reason to assume that he ordered its physical destruction than to believe it was caused by an earthquake.

The second massive destruction at Megiddo, which occurred around 1250 BC, has variously been attributed to the Israelites or the Philistines, although historical evidence supports neither candidate. Again, however, the excavation of collapsed walls in Megiddo, and similar contemporaneous destruction in many nearby sites (Davies 1986), make the earthquake hypothesis a likely candidate.

The strongest evidence for earthquake destruction at Megiddo is probably the layer dating to between 1130 and 1000 BC [JW: Based on the Caption in Fig. 7.3, this refers to Stratum VI], which some scholars attribute to conquest by King David’s army. There is no historical mention, however, of David capturing Megiddo, much less leveling it, and, given the importance of the place at the time, it seems unlikely that such a conquest would go unheralded. More likely, Megiddo was destroyed by a massive earthquake, perhaps the same one that, according to the Bible, occurred during the battle of Michmash (1 Samuel 14:15). In the 1930s, excavators found collapsed walls from this period and, under the walls, smashed jars and several human skeletons, including the one shown in Figure 7.3. This layer is possibly contemporaneous with destruction layers from Dor, Gezer, and at least a dozen other sites in Israel and Jordan, all of them consistent with earthquake destruction.

The fourth layer of destruction occurred sometime after the conquest of Megiddo by Pharaoh Sheshonq in 925 BC. Although this layer is sometimes attributed to him, most documentary and archaeological evidence indicates that Sheshonq did not destroy the city but rather had a monument erected in his honor there and exacted tribute from the residents. Yet, on the other hand, no definitive evidence has yet been excavated to indicate that an earthquake caused this destruction. We know that the massive earthquake during the reign of King Uzziah, around 760 BC, was important in this area—so important, in fact, that it is mentioned in the prophesies in the book of Zechariah (14:4–5):

And his feet shall stand in that day upon the mount of Olives, which is before Jerusalem on the east, and the mount of Olives shall cleave in the midst thereof toward the east and toward the west, and there shall be a very great valley; and half of the mountain shall remove toward the north, and half of it toward the south. And ye shall flee to the valley of the mountains . . . yea, ye shall flee, like as ye fled from before the earthquake in the days of Uzziah king of Judah.

Other sources claim that the earthquake Zechariah cites occurred around 760 BC. In any case, clearly it was a major enough disaster to be used as a temporal reference in Zechariah’s prophecy. In fact, the prophecy includes such vivid details that it may actually describe the ground motion that occurred in the 760 BC earthquake, perhaps motion across a strike-slip fault.

Another example of an earthquake prophecy appears in the book of Revelation, where a mighty earthquake is prophesied to occur during the battle of Armageddon:

And they assembled them at the place that in Hebrew is called Armageddon . . . and there came . . . a violent earthquake, such as had not occurred since people were upon the earth, so violent was that earthquake: And the great city was split into three parts, and the cities of the nations fell . . . And every island fled away, and no mountains were to be found.

This account and the one in Zechariah may be examples of retrospective prophecies, a common feature in ancient literature where the author dramatizes a historic event as a kind of warning after the fact. This may indicate that, in the past, earthquakes were known to have struck Megiddo. Most important is that excavators keep in mind in future excavations that the area is subject to severe earthquake hazards, and always consider earthquakes as a possible cause for unexplained destruction there.

Topography is what made Megiddo so important militarily for so long, but other factors also determined the location of cities and forts in the ancient Mediterranean region. Another consideration in this arid land was the availability of water. The previous chapter discussed the importance of water in Qumran: dependent on cisterns to capture seasonal rainfall, Qumran was particularly vulnerable to earthquake damage. Cisterns were essential in many Israeli regions, and the collection, storage, and distribution of water was important throughout Israel’s history. Any place with a year-round supply of fresh water, particularly clean, filtered water from a dependable spring, was prime real estate. One example was Jericho, described in chapter 3.

Tel Jericho’s extraordinary stack of at least twenty-two layers of destruction has already been noted. With no modern city built on top of the tell—modern Jericho is a few miles from the remnants of its ancient namesake—archaeological excavation has progressed relatively unhindered, despite the usual impediment of political unrest in the area. This has been a liability in some ways, since excavations using early archaeological methods clearly destroyed at least as much information as they revealed. Still, because of these excavations, the repeated destruction and rebuilding of Jericho has been acknowledged for some time. With its long record and repeated rebuilding, Jericho would be a great candidate for a sort of key site, a place where evidence of repeated earthquake damage could be correlated with destruction in neighboring sites where the record might be less continuous. If we could systematically examine the written evidence of earthquakes and try to tie together the archaeological stratigraphy among the sites scattered around Israel and Jordan, we might be able to piece together a much more informative record than we could hope to gain from any one site alone. A great deal of archaeological work has been done to try to correlate the remains at Jericho with other sites throughout the Holy Land. Kathleen Kenyon (1978) was particularly comprehensive in her synthesis of the archaeology of the region. Although she does mention the general likelihood of earthquakes in this region and specifically ascribes certain damage layers to earthquakes, she makes no explicit effort to determine the range of such earthquake damage; she does not mention whether the earthquake in question could have damaged nearby towns.

One of the most suggestive layers in Jericho is one Kenyon dated, based on pottery styles, to between 1600 and 1550 BC, where the walls of Jericho collapsed, burying storage jars full of grain (Figure 7.4). The grain was carbonized by a fire concurrent with the collapse, and carbon dates of the grain have placed it remarkably close to Kenyon’s original estimate (Bruins and van der Plicht 1996). According to Kenyon (1979, 177–178), many towns in the region suffered total destruction at the same time (though she does not implicate an earthquake), including the sites of Tel Beit Mirsim, Hazor, Shechem, and Megiddo. Among these sites, only Megiddo was immediately rebuilt, with no apparent change in culture; most of the others were abandoned for more than a century.

In chapter 3, I speculated that an earthquake may have figured prominently in the biblical story of Joshua’s destruction of Jericho. Does the archaeological record preserve any evidence at all of Joshua’s Jericho? One of the early expeditions reported the discovery of a city wall from Joshua’s time, but Kenyon (1957) indicated that the dating of that wall was incorrect. As described in chapter 3, she found only one small remnant of Jericho that could have coincided with the traditional estimate of 1400 to 1250 BC for Joshua’s conquest of the site. If a larger habitation did exist at the site, the remains of that culture have all but vanished, eroded away during a long period of abandonment.

Recently, however, archaeologists have reexamined both the archaeological site at Jericho and radiocarbon dates for other events in the ancient world. One particularly notable theory links the plague of darkness in Egypt, before the Exodus, to the explosive eruption of the island of Thera in the Aegean Sea, an event dated to the end of the seventeenth century BC. If this date is correct, the nearly missing layer Kenyon attributed to Joshua’s time would be far too recent, and therefore the destruction from ca. 1600 would be a more likely candidate.

The debate over the assignment of this layer, based on samples of locally made pottery unearthed at Jericho, and on radiocarbon dates from the charred grain and from bits of timber found in the destruction layer, has at times been heated, almost vicious. (For an example of two diametrically opposed interpretations of the same evidence, presented side-by-side in one issue of Biblical Archaeology Review, complete with ad hominem attacks, see Bienkowski 1990 and Wood 1990.) It will be some time before the dust settles on this issue. For now, the archaeology is the source of more questions than answers, and until some of those questions can be resolved, using Jericho as a key site will remain difficult.

Although the mound of ancient Megiddo was eventually abandoned, and the modern city of Jericho has abandoned the old tell, the city of Jerusalem has been continuously inhabited for at least four thousand years (Cline 2004). The reason for its continued importance reaches beyond favorable topography or water supply to that least tangible reason for the persistence of cities: religious significance. King David’s choice of Jerusalem as the capital city of his Hebrew nation, and, probably more important, his decision to move the Ark of the Covenant into the city, sealed its fate as an enduring human habitation into modern times.

Despite its long history, most of the city’s archaeological secrets are nearly impossible to reveal. The whole city has been densely inhabited for millennia, with almost no abandonment in any quarter. Modern and ancient homes, churches, and official buildings block access to the layers of rubble beneath them. When a scrap of real estate is not actively occupied, it is usually because it is imbued with some intense political or religious importance and digging into it would touch off a firestorm of controversy.

As a result, excavations carried out in Jerusalem are often quite limited and piecemeal, undertaken only when an opportunity happens to present itself. This occurred in 1948, for example, when, during the Arab-Israeli war, a mortar hit a building that, according to tradition, housed the tomb of King David. In the process of repairing the damage inflicted by the explosion, the archaeologist Jacob Pinkerfeld (1990) also did a cursory excavation of one portion of the structure. Beneath the floor, he found over half a meter of debris, including three earlier floors. In fact, further investigations have suggested that the walls of this building were part of the Church of the Apostles, built in the first century AD at the place where the Last Supper was thought to have occurred. The walls have been destroyed and rebuilt many times, but because of the site’s importance to those who believe it was the tomb of David, the excavation was hindered and no further investigation was possible.

Still, although the city cannot be systematically excavated, modern instruments can probe beneath the surface to tell us not how old the layers are or what artifacts are in them but at least whether specific parts of the city are founded on archaeological debris or on solid ground. A government study by Israel’s Geological Survey was released in 2004, confirming what archaeologists had long suspected: the Old City is mostly founded on the rubble of previous constructions. As discussed in chapter 2, this makes the Old City particularly vulnerable to earthquake damage.

The historical record bears this out. In nearly every historical report of earthquakes affecting Jerusalem, damage is reported either to the Temple, the Old City walls, or both. Of course, the importance of this site and its role as the focus of religious pilgrimages for the world’s three major religions guarantees a record that is unparalleled in any other region. We have written records from many sources for earthquakes in Jerusalem, most of which also affected large regions of the countryside.

A recent geological investigation of sediments from the Dead Sea (Ken-Tor et al. 2001), confirms the validity of many of the historical reports mentioned in this chapter. As described in chapter 2, when shaking of sea-floor sediments is severe enough, the loose, water-saturated sediments lose their strength and flow like a liquid. When the shaking stops, the sediments settle and resolidify, leaving behind a chaotic, mixed layer that is readily identified, a seismite. These layers frequently contain organic material that scientists can date by Carbon-14 dating.

The 2001 study examined layers from the Ze’elim Terrace on the shore of the Dead Sea, a stack of sediments exposed by rapid modern drops in Dead Sea water levels (caused by diversion of the fresh water sources that feed the Dead Sea). There are gaps in this sequence, marking drought periods where the ancient water level in the Dead Sea was lower than usual and there was no sediment deposited at the Ze’elim Terrace site. However, during the periods when sediments accumulated there, every major historical earthquake on the Dead Sea Transform has been correlated to a seismite. This is a great new resource and one that can be extended in the future. The periods for which the Ze’elim Terrace contains no data could be examined by drilling into the deeper sediments in the Dead Sea floor. Thus, we can confirm the historical record at Jerusalem, and use it, along with the archaeological record in more accessible sites nearby, to build a physical earthquake stratigraphy for the region.

Due to Megiddo’s location in the Carmel fault zone it is particularly prone to seismic activity. It is therefore reasonable to suggest that Stratum VIA was destroyed in a massive earthquake, and in 2006 Shmuel Marco et al. published a report documenting the dates and probabilities of seismic events at Megiddo. Attributing a destruction level to an earthquake, however, is difficult to prove, as “Earthquake-related damage often resembles that deliberately caused by humans, geotechnical failure, or slow deterioration over the ages” (Marco et al. 2006: 569). Investigators thus looked for certain criteria that could suggest seismic activity, the main criterion being “time-constrained widespread damage. The temporal bounds should be tight enough to indicate that the damage occurred in a single catastrophic event” (Marco et al. 2006: 569). Other significant pieces of evidence include the absence or scarcity of weapons, historical records that document seismic activity, “deformation of coeval natural sediments, and the existence of certain types of damage that are uniquely associated with earthquakes” (Marco et al. 2006: 569).

Ultimately, the study was inconclusive regarding the occurrence of a major earthquake in Stratum VIA. Only two earthquakes were confirmed beyond doubt: “one at the end of the fourth millennium BCE and another in the 9th century BCE (which caused the damage in Stratum VA-IVB)” (Marco et al. 2006: 572). This study did not eliminate the possibility of an earthquake in Stratum VIA, however there was no conclusive evidence to support such an event (Marco et al. 2006: 572).

Despite the inconclusive results of the archaeoseismic study regarding Stratum VIA, other considerations have supported the view that this city was not destroyed in an earthquake. First of all, the remains from this stratum show extensive burning, suggesting that the city was intentionally burned down by a hostile force. “It is difficult to imagine that all the stone and brick-built houses could have simultaneously caught fire due to an earthquake as might happen in a modern city” (Finkelstein, Ussishkin and Halpern 2006: 850). Secondly, the following stratum showed discontinuity in material culture, suggesting that a new population inhabited the later city and that the inhabitants of the Stratum VIA city did not return following its destruction. “Had the Stratum VIA settlement been destroyed by fire resulting from an earthquake we would have expected the inhabitants to reconstruct their ruined houses without delay. The fact that they did not return indicates that they were forced to abandon their settlement forever by a human agent” (Finkelstein, Ussishkin and Halpern 2006: 850). Finally, as previously stated, the radiocarbon results from the region suggest that there were two main destruction events, further strengthening the theory that this destruction was part of an ongoing process such as the gradual conquest of the region by an outside power rather than a single time-restricted event.

In the search for the military power responsible for the destruction of this city, the Egyptian army of Pharaoh Shoshenq I has been a popular explanation. This campaign, known from 1 Kgs 14:25-28 (cf. 2 Chr 12:1-12) as well as from the temple of Amun at Karnak (see Simons 1937: 95-102), has traditionally been dated to around 926 BCE by the biblical text (Finkelstein 2002: 109), which states that “In the fifth year of King Rehoboam, King Shishak of Egypt came up against Jerusalem (1 Kgs 14:25, NRSV).

A late 10th century BCE date for Shoshenq’s campaign has proven troublesome for the archaeologist. Egyptian records do not corroborate the precise year of his campaign, thus some have suggested a wider range of possible dates for this event. Finkelstein raises four main problems regarding the date of this campaign:

1. the complicated chronology of the 21st and 22nd Dynasties in Egypt allow for the change of several years back or forth of Shoshenq I’s reign
2. it is unknown whether his campaign occurred early or late in his rule
3. the biblical references of the length of reigns of the early Davidides are completely schematized
4. the fifth year of Rehoboam datum may have been altered to fit the theology of the Deuteronomistic Historian (Finkelstein 2002: 110).

Thus after a critical review of the available evidence, “the Shoshenq campaign could have taken place almost any time in the mid- to late-10th century BCE” (Finkelstein 2002: 110; see also Ash 1999: 27-34).

Due to the uncertainty of the date of Shoshenq I’s campaign, identification of destruction layers to his campaign have been problematic. Indeed, “we do not have even one destruction layer which can safely be assigned to this campaign. For instance, in several sites there are two destruction horizons (e.g. and Megiddo, Stratum VIA and Stratum VA-IVB); both can be attributed to Shoshenq’s campaign” (Finkelstein 1996: 180). Furthermore, the destruction layer of Stratum IVA has also been attributed to this campaign (Guy 1931: 48).

Unfortunately, the most significant piece of evidence that could contribute to this debate was found out of context during Megiddo excavations. Found during Fisher’s excavations in one of Schumacher’s dumps, a fragment of a large stone stele of Pharaoh Shoshenq was discovered near the eastern edge of the mound (Ussishkin 1990: 71). This fragment appears to be part of a stele that would have measured approximately 3.30 m high, 1.50 m wide, and 50 cm thick (Ussishkin 1990: 72). Four criteria typify this type of monument:

First, they were erected by a monarch in a foreign land; second, their erection was carried out following, or in association with, a military campaign or submission by the local ruler to the conquering leader; third, they were erected in a public, central position inside a capital city; fourth, the city in question was inhabited at the time the monument was erected and dominated by the monarch who erected the stele (Ussishkin 1990: 72).

As a result of this find many scholars do not attribute the destruction of Stratum VIA to Shoshenq I’s campaign. The erection of a victory stele in the city implies that the city would have had inhabitants to appreciate such an impressive work, yet the total destruction of this stratum and the settlement gap following it would imply that had Shoshenq I been responsible for this destruction, he would have erected his stele in an uninhabited city.

Therefore, it would seem that Shoshenq I’s campaign occurred in the Iron Age IIA rather than during the Iron Age I. Stratum VIA is ruled out due to four main considerations:

1. At least some of these destructions are radiocarbon dated to before the highest possible date for his reign
2. the radiocarbon evidence indicates a gradual demise of these cities and not a single destructive event
3. there was no reason for a pharaoh who was probably interested in re- establishing Egyptian rule in the area to devastate a fertile valley that was the bread basket of the entire country
4. it is illogical that Sheshonq I would establish a stele in a deserted Megiddo (Finkelstein 2011: 232).

Shoshenq I’s campaign was intended to reestablish Egyptian hegemony in the region, made evident from the stele fragment. This being the case, he likely would have only destroyed the elite quarter of the city, or perhaps taken it peacefully (Finkelstein 2009: 121; regarding the possibility of a peaceful takeover see Finkelstein 2002: 122). Thus, Shoshenq I’s campaign to the north should not be attributed to Megiddo Stratum VIA, but rather to a later city in the Iron Age IIA (Ussishkin 1990: 73; Finkelstein 2011: 232).

After ruling out destruction by both earthquake and the military campaign of Pharaoh Shoshenq I, one must investigate the possibility of the Israelite expansion to the region from the northern highlands. Indeed, this suggestion best fits the current set of data available by both textual sources and archaeological excavations.

First of all, the data from radiocarbon studies “renders the earthquake and single military campaign theories invalid,” as the dates provided by these samples suggest two waves of destruction in the region rather than one (Finkelstein 2013: 1337). A series of raids or the gradual expansion of the Israelites from the highlands better explains these waves of destructions (Na’aman 2007: 402; Finkelstein 2011: 229; 2013: 1337).

Second, the later settlements at Megiddo support this reconstruction. The break in material culture between the Iron Age I-IIA suggests that the population that resettled in the city following Stratum VIA was a different people, most likely the Israelites who later on established the Omride Kingdom in the region (Megiddo VA-IVB and its contemporaries) (Finkelstein 2009: 122-123). This reconstruction still leaves room for the campaign of Shoshenq I in the decades following the Israelite expansion, thus allowing for a unified and harmonious historical reconstruction inclusive of all known historical events from this period.

• from Nur and Burgess (2008:189-199)

Figures Discussion

A modest mound southeast of the Carmel Ridge in Israel, the tell of Megiddo rises 50 meters above the surrounding Jezreel Plain and covers some 6 hectares (15 acres) of land (Figure 7.1). This tell was first identified as the legendary Armageddon [the Greek corruption of the Hebrew “Har Megiddo” or Mount Megiddo] by a Jewish writer of the fourteenth century, Esthori Haparchi, and then was rediscovered by the British army officer H. H. Kitchener five hundred years later. Extensive excavations were conducted at the site by C. Fisher, P. Guy, and G. Loud (Yadin 1975). These archaeological studies revealed physical evidence of the historical development of Megiddo.

Megiddo’s strategic importance, belied by its unimposing appearance today, stems from its unique topography (Figure 7.2). The land between the Mediterranean and the Jordan River served as a bridge between the civilizations in the South, in Egypt and Arabia, and those in the North, in Syria, Mesopotamia, and Anatolia. It was also a continuation of sea routes from the Mediterranean and the Gulf of Suez. The ruggedness of the region, however, crossed by several ranges of mountains and hills as well as the lowest valley on Earth, limited the possible routes for overland shipping or wheeled travel. The Carmel-Gilboa mountain range was a particular obstacle, and traffic from the Mediterranean to Syria and Jordan was funneled through a few gaps in the range. In fact, both the passes and the mountain range that obstructed traffic were products of tectonic motion along the seismically active Carmel Gilboa fault system, which branches off from the Dead Sea Transform (see Figure 4.5).

The mound of Megiddo stands guard over one of the most important of these mountain passes, the Nahal Iron Pass, a traffic bottleneck on the main route between Egypt and Syria. Until the advent of more elaborate road construction by the Roman Empire in the first and second centuries AD, the gap at Megiddo was the only one that permitted the passage of chariots, though it was not an easy passage. A description from the latter part of the thirteenth century BC, found in the Egyptian Papyrus Anastasi I, gives some idea of the difficulty of the route:

Thy path is filled with boulders and pebbles, without a toe-hold for passing by, overgrown with reeds, thorns, brambles, and wolf’s paw. The ravine is on one side of thee, and the mountain rises on the other. Thou goest on jolting, with thy chariot on its side, afraid to press thy horse too hard. . . . The horse is played out by the time thou findest night quarters. (Hori [Egyptian Royal Official], Papyrus Anastasi I)

This narrow and difficult route made the pass from Megiddo particularly easy to guard, so whoever held power in Megiddo controlled not only the course of trade in the Fertile Crescent but that of war as well. Thus, the site figured prominently in some of the greatest ancient battles fought in this region. As Pharaoh Thutmose III expressed it, “The capture of Megiddo is the capture of 1,000 towns.” Indeed, fortifications were built and rebuilt there for close to five millennia, until around 500 BC (Finkelstein and Ussishkin 1994).

Four levels of destruction in the mound of Tel Megiddo are consistent with earthquake destruction, the lowest one attributed to the conquest by Thutmose III in 1468 BC. Why, however, would the Pharaoh have destroyed the place if his goal were to occupy the site and exact tribute? Although it is clear that Thutmose III conquered Megiddo, there is no more reason to assume that he ordered its physical destruction than to believe it was caused by an earthquake.

The second massive destruction at Megiddo, which occurred around 1250 BC, has variously been attributed to the Israelites or the Philistines, although historical evidence supports neither candidate. Again, however, the excavation of collapsed walls in Megiddo, and similar contemporaneous destruction in many nearby sites (Davies 1986), make the earthquake hypothesis a likely candidate.

The strongest evidence for earthquake destruction at Megiddo is probably the layer dating to between 1130 and 1000 BC [JW: Based on the Caption in Fig. 7.3, this refers to Stratum VI], which some scholars attribute to conquest by King David’s army. There is no historical mention, however, of David capturing Megiddo, much less leveling it, and, given the importance of the place at the time, it seems unlikely that such a conquest would go unheralded. More likely, Megiddo was destroyed by a massive earthquake, perhaps the same one that, according to the Bible, occurred during the battle of Michmash (1 Samuel 14:15). In the 1930s, excavators found collapsed walls from this period and, under the walls, smashed jars and several human skeletons, including the one shown in Figure 7.3. This layer is possibly contemporaneous with destruction layers from Dor, Gezer, and at least a dozen other sites in Israel and Jordan, all of them consistent with earthquake destruction.

The fourth layer of destruction occurred sometime after the conquest of Megiddo by Pharaoh Sheshonq in 925 BC. Although this layer is sometimes attributed to him, most documentary and archaeological evidence indicates that Sheshonq did not destroy the city but rather had a monument erected in his honor there and exacted tribute from the residents. Yet, on the other hand, no definitive evidence has yet been excavated to indicate that an earthquake caused this destruction. We know that the massive earthquake during the reign of King Uzziah, around 760 BC, was important in this area—so important, in fact, that it is mentioned in the prophesies in the book of Zechariah (14:4–5):

And his feet shall stand in that day upon the mount of Olives, which is before Jerusalem on the east, and the mount of Olives shall cleave in the midst thereof toward the east and toward the west, and there shall be a very great valley; and half of the mountain shall remove toward the north, and half of it toward the south. And ye shall flee to the valley of the mountains . . . yea, ye shall flee, like as ye fled from before the earthquake in the days of Uzziah king of Judah.

Other sources claim that the earthquake Zechariah cites occurred around 760 BC. In any case, clearly it was a major enough disaster to be used as a temporal reference in Zechariah’s prophecy. In fact, the prophecy includes such vivid details that it may actually describe the ground motion that occurred in the 760 BC earthquake, perhaps motion across a strike-slip fault.

Another example of an earthquake prophecy appears in the book of Revelation, where a mighty earthquake is prophesied to occur during the battle of Armageddon:

And they assembled them at the place that in Hebrew is called Armageddon . . . and there came . . . a violent earthquake, such as had not occurred since people were upon the earth, so violent was that earthquake: And the great city was split into three parts, and the cities of the nations fell . . . And every island fled away, and no mountains were to be found.

This account and the one in Zechariah may be examples of retrospective prophecies, a common feature in ancient literature where the author dramatizes a historic event as a kind of warning after the fact. This may indicate that, in the past, earthquakes were known to have struck Megiddo. Most important is that excavators keep in mind in future excavations that the area is subject to severe earthquake hazards, and always consider earthquakes as a possible cause for unexplained destruction there.

Topography is what made Megiddo so important militarily for so long, but other factors also determined the location of cities and forts in the ancient Mediterranean region. Another consideration in this arid land was the availability of water. The previous chapter discussed the importance of water in Qumran: dependent on cisterns to capture seasonal rainfall, Qumran was particularly vulnerable to earthquake damage. Cisterns were essential in many Israeli regions, and the collection, storage, and distribution of water was important throughout Israel’s history. Any place with a year-round supply of fresh water, particularly clean, filtered water from a dependable spring, was prime real estate. One example was Jericho, described in chapter 3.

Tel Jericho’s extraordinary stack of at least twenty-two layers of destruction has already been noted. With no modern city built on top of the tell—modern Jericho is a few miles from the remnants of its ancient namesake—archaeological excavation has progressed relatively unhindered, despite the usual impediment of political unrest in the area. This has been a liability in some ways, since excavations using early archaeological methods clearly destroyed at least as much information as they revealed. Still, because of these excavations, the repeated destruction and rebuilding of Jericho has been acknowledged for some time. With its long record and repeated rebuilding, Jericho would be a great candidate for a sort of key site, a place where evidence of repeated earthquake damage could be correlated with destruction in neighboring sites where the record might be less continuous. If we could systematically examine the written evidence of earthquakes and try to tie together the archaeological stratigraphy among the sites scattered around Israel and Jordan, we might be able to piece together a much more informative record than we could hope to gain from any one site alone. A great deal of archaeological work has been done to try to correlate the remains at Jericho with other sites throughout the Holy Land. Kathleen Kenyon (1978) was particularly comprehensive in her synthesis of the archaeology of the region. Although she does mention the general likelihood of earthquakes in this region and specifically ascribes certain damage layers to earthquakes, she makes no explicit effort to determine the range of such earthquake damage; she does not mention whether the earthquake in question could have damaged nearby towns.

One of the most suggestive layers in Jericho is one Kenyon dated, based on pottery styles, to between 1600 and 1550 BC, where the walls of Jericho collapsed, burying storage jars full of grain (Figure 7.4). The grain was carbonized by a fire concurrent with the collapse, and carbon dates of the grain have placed it remarkably close to Kenyon’s original estimate (Bruins and van der Plicht 1996). According to Kenyon (1979, 177–178), many towns in the region suffered total destruction at the same time (though she does not implicate an earthquake), including the sites of Tel Beit Mirsim, Hazor, Shechem, and Megiddo. Among these sites, only Megiddo was immediately rebuilt, with no apparent change in culture; most of the others were abandoned for more than a century.

In chapter 3, I speculated that an earthquake may have figured prominently in the biblical story of Joshua’s destruction of Jericho. Does the archaeological record preserve any evidence at all of Joshua’s Jericho? One of the early expeditions reported the discovery of a city wall from Joshua’s time, but Kenyon (1957) indicated that the dating of that wall was incorrect. As described in chapter 3, she found only one small remnant of Jericho that could have coincided with the traditional estimate of 1400 to 1250 BC for Joshua’s conquest of the site. If a larger habitation did exist at the site, the remains of that culture have all but vanished, eroded away during a long period of abandonment.

Recently, however, archaeologists have reexamined both the archaeological site at Jericho and radiocarbon dates for other events in the ancient world. One particularly notable theory links the plague of darkness in Egypt, before the Exodus, to the explosive eruption of the island of Thera in the Aegean Sea, an event dated to the end of the seventeenth century BC. If this date is correct, the nearly missing layer Kenyon attributed to Joshua’s time would be far too recent, and therefore the destruction from ca. 1600 would be a more likely candidate.

The debate over the assignment of this layer, based on samples of locally made pottery unearthed at Jericho, and on radiocarbon dates from the charred grain and from bits of timber found in the destruction layer, has at times been heated, almost vicious. (For an example of two diametrically opposed interpretations of the same evidence, presented side-by-side in one issue of Biblical Archaeology Review, complete with ad hominem attacks, see Bienkowski 1990 and Wood 1990.) It will be some time before the dust settles on this issue. For now, the archaeology is the source of more questions than answers, and until some of those questions can be resolved, using Jericho as a key site will remain difficult.

Although the mound of ancient Megiddo was eventually abandoned, and the modern city of Jericho has abandoned the old tell, the city of Jerusalem has been continuously inhabited for at least four thousand years (Cline 2004). The reason for its continued importance reaches beyond favorable topography or water supply to that least tangible reason for the persistence of cities: religious significance. King David’s choice of Jerusalem as the capital city of his Hebrew nation, and, probably more important, his decision to move the Ark of the Covenant into the city, sealed its fate as an enduring human habitation into modern times.

Despite its long history, most of the city’s archaeological secrets are nearly impossible to reveal. The whole city has been densely inhabited for millennia, with almost no abandonment in any quarter. Modern and ancient homes, churches, and official buildings block access to the layers of rubble beneath them. When a scrap of real estate is not actively occupied, it is usually because it is imbued with some intense political or religious importance and digging into it would touch off a firestorm of controversy.

As a result, excavations carried out in Jerusalem are often quite limited and piecemeal, undertaken only when an opportunity happens to present itself. This occurred in 1948, for example, when, during the Arab-Israeli war, a mortar hit a building that, according to tradition, housed the tomb of King David. In the process of repairing the damage inflicted by the explosion, the archaeologist Jacob Pinkerfeld (1990) also did a cursory excavation of one portion of the structure. Beneath the floor, he found over half a meter of debris, including three earlier floors. In fact, further investigations have suggested that the walls of this building were part of the Church of the Apostles, built in the first century AD at the place where the Last Supper was thought to have occurred. The walls have been destroyed and rebuilt many times, but because of the site’s importance to those who believe it was the tomb of David, the excavation was hindered and no further investigation was possible.

Still, although the city cannot be systematically excavated, modern instruments can probe beneath the surface to tell us not how old the layers are or what artifacts are in them but at least whether specific parts of the city are founded on archaeological debris or on solid ground. A government study by Israel’s Geological Survey was released in 2004, confirming what archaeologists had long suspected: the Old City is mostly founded on the rubble of previous constructions. As discussed in chapter 2, this makes the Old City particularly vulnerable to earthquake damage.

The historical record bears this out. In nearly every historical report of earthquakes affecting Jerusalem, damage is reported either to the Temple, the Old City walls, or both. Of course, the importance of this site and its role as the focus of religious pilgrimages for the world’s three major religions guarantees a record that is unparalleled in any other region. We have written records from many sources for earthquakes in Jerusalem, most of which also affected large regions of the countryside.

A recent geological investigation of sediments from the Dead Sea (Ken-Tor et al. 2001), confirms the validity of many of the historical reports mentioned in this chapter. As described in chapter 2, when shaking of sea-floor sediments is severe enough, the loose, water-saturated sediments lose their strength and flow like a liquid. When the shaking stops, the sediments settle and resolidify, leaving behind a chaotic, mixed layer that is readily identified, a seismite. These layers frequently contain organic material that scientists can date by Carbon-14 dating.

The 2001 study examined layers from the Ze’elim Terrace on the shore of the Dead Sea, a stack of sediments exposed by rapid modern drops in Dead Sea water levels (caused by diversion of the fresh water sources that feed the Dead Sea). There are gaps in this sequence, marking drought periods where the ancient water level in the Dead Sea was lower than usual and there was no sediment deposited at the Ze’elim Terrace site. However, during the periods when sediments accumulated there, every major historical earthquake on the Dead Sea Transform has been correlated to a seismite. This is a great new resource and one that can be extended in the future. The periods for which the Ze’elim Terrace contains no data could be examined by drilling into the deeper sediments in the Dead Sea floor. Thus, we can confirm the historical record at Jerusalem, and use it, along with the archaeological record in more accessible sites nearby, to build a physical earthquake stratigraphy for the region.

• from Nur and Burgess (2008:189-199)

Figures Discussion

A modest mound southeast of the Carmel Ridge in Israel, the tell of Megiddo rises 50 meters above the surrounding Jezreel Plain and covers some 6 hectares (15 acres) of land (Figure 7.1). This tell was first identified as the legendary Armageddon [the Greek corruption of the Hebrew “Har Megiddo” or Mount Megiddo] by a Jewish writer of the fourteenth century, Esthori Haparchi, and then was rediscovered by the British army officer H. H. Kitchener five hundred years later. Extensive excavations were conducted at the site by C. Fisher, P. Guy, and G. Loud (Yadin 1975). These archaeological studies revealed physical evidence of the historical development of Megiddo.

Megiddo’s strategic importance, belied by its unimposing appearance today, stems from its unique topography (Figure 7.2). The land between the Mediterranean and the Jordan River served as a bridge between the civilizations in the South, in Egypt and Arabia, and those in the North, in Syria, Mesopotamia, and Anatolia. It was also a continuation of sea routes from the Mediterranean and the Gulf of Suez. The ruggedness of the region, however, crossed by several ranges of mountains and hills as well as the lowest valley on Earth, limited the possible routes for overland shipping or wheeled travel. The Carmel-Gilboa mountain range was a particular obstacle, and traffic from the Mediterranean to Syria and Jordan was funneled through a few gaps in the range. In fact, both the passes and the mountain range that obstructed traffic were products of tectonic motion along the seismically active Carmel Gilboa fault system, which branches off from the Dead Sea Transform (see Figure 4.5).

The mound of Megiddo stands guard over one of the most important of these mountain passes, the Nahal Iron Pass, a traffic bottleneck on the main route between Egypt and Syria. Until the advent of more elaborate road construction by the Roman Empire in the first and second centuries AD, the gap at Megiddo was the only one that permitted the passage of chariots, though it was not an easy passage. A description from the latter part of the thirteenth century BC, found in the Egyptian Papyrus Anastasi I, gives some idea of the difficulty of the route:

Thy path is filled with boulders and pebbles, without a toe-hold for passing by, overgrown with reeds, thorns, brambles, and wolf’s paw. The ravine is on one side of thee, and the mountain rises on the other. Thou goest on jolting, with thy chariot on its side, afraid to press thy horse too hard. . . . The horse is played out by the time thou findest night quarters. (Hori [Egyptian Royal Official], Papyrus Anastasi I)

This narrow and difficult route made the pass from Megiddo particularly easy to guard, so whoever held power in Megiddo controlled not only the course of trade in the Fertile Crescent but that of war as well. Thus, the site figured prominently in some of the greatest ancient battles fought in this region. As Pharaoh Thutmose III expressed it, “The capture of Megiddo is the capture of 1,000 towns.” Indeed, fortifications were built and rebuilt there for close to five millennia, until around 500 BC (Finkelstein and Ussishkin 1994).

Four levels of destruction in the mound of Tel Megiddo are consistent with earthquake destruction, the lowest one attributed to the conquest by Thutmose III in 1468 BC. Why, however, would the Pharaoh have destroyed the place if his goal were to occupy the site and exact tribute? Although it is clear that Thutmose III conquered Megiddo, there is no more reason to assume that he ordered its physical destruction than to believe it was caused by an earthquake.

The second massive destruction at Megiddo, which occurred around 1250 BC, has variously been attributed to the Israelites or the Philistines, although historical evidence supports neither candidate. Again, however, the excavation of collapsed walls in Megiddo, and similar contemporaneous destruction in many nearby sites (Davies 1986), make the earthquake hypothesis a likely candidate.

The strongest evidence for earthquake destruction at Megiddo is probably the layer dating to between 1130 and 1000 BC [JW: Based on the Caption in Fig. 7.3, this refers to Stratum VI], which some scholars attribute to conquest by King David’s army. There is no historical mention, however, of David capturing Megiddo, much less leveling it, and, given the importance of the place at the time, it seems unlikely that such a conquest would go unheralded. More likely, Megiddo was destroyed by a massive earthquake, perhaps the same one that, according to the Bible, occurred during the battle of Michmash (1 Samuel 14:15). In the 1930s, excavators found collapsed walls from this period and, under the walls, smashed jars and several human skeletons, including the one shown in Figure 7.3. This layer is possibly contemporaneous with destruction layers from Dor, Gezer, and at least a dozen other sites in Israel and Jordan, all of them consistent with earthquake destruction.

The fourth layer of destruction occurred sometime after the conquest of Megiddo by Pharaoh Sheshonq in 925 BC. Although this layer is sometimes attributed to him, most documentary and archaeological evidence indicates that Sheshonq did not destroy the city but rather had a monument erected in his honor there and exacted tribute from the residents. Yet, on the other hand, no definitive evidence has yet been excavated to indicate that an earthquake caused this destruction. We know that the massive earthquake during the reign of King Uzziah, around 760 BC, was important in this area—so important, in fact, that it is mentioned in the prophesies in the book of Zechariah (14:4–5):

And his feet shall stand in that day upon the mount of Olives, which is before Jerusalem on the east, and the mount of Olives shall cleave in the midst thereof toward the east and toward the west, and there shall be a very great valley; and half of the mountain shall remove toward the north, and half of it toward the south. And ye shall flee to the valley of the mountains . . . yea, ye shall flee, like as ye fled from before the earthquake in the days of Uzziah king of Judah.

Other sources claim that the earthquake Zechariah cites occurred around 760 BC. In any case, clearly it was a major enough disaster to be used as a temporal reference in Zechariah’s prophecy. In fact, the prophecy includes such vivid details that it may actually describe the ground motion that occurred in the 760 BC earthquake, perhaps motion across a strike-slip fault.

Another example of an earthquake prophecy appears in the book of Revelation, where a mighty earthquake is prophesied to occur during the battle of Armageddon:

And they assembled them at the place that in Hebrew is called Armageddon . . . and there came . . . a violent earthquake, such as had not occurred since people were upon the earth, so violent was that earthquake: And the great city was split into three parts, and the cities of the nations fell . . . And every island fled away, and no mountains were to be found.

This account and the one in Zechariah may be examples of retrospective prophecies, a common feature in ancient literature where the author dramatizes a historic event as a kind of warning after the fact. This may indicate that, in the past, earthquakes were known to have struck Megiddo. Most important is that excavators keep in mind in future excavations that the area is subject to severe earthquake hazards, and always consider earthquakes as a possible cause for unexplained destruction there.

Topography is what made Megiddo so important militarily for so long, but other factors also determined the location of cities and forts in the ancient Mediterranean region. Another consideration in this arid land was the availability of water. The previous chapter discussed the importance of water in Qumran: dependent on cisterns to capture seasonal rainfall, Qumran was particularly vulnerable to earthquake damage. Cisterns were essential in many Israeli regions, and the collection, storage, and distribution of water was important throughout Israel’s history. Any place with a year-round supply of fresh water, particularly clean, filtered water from a dependable spring, was prime real estate. One example was Jericho, described in chapter 3.

Tel Jericho’s extraordinary stack of at least twenty-two layers of destruction has already been noted. With no modern city built on top of the tell—modern Jericho is a few miles from the remnants of its ancient namesake—archaeological excavation has progressed relatively unhindered, despite the usual impediment of political unrest in the area. This has been a liability in some ways, since excavations using early archaeological methods clearly destroyed at least as much information as they revealed. Still, because of these excavations, the repeated destruction and rebuilding of Jericho has been acknowledged for some time. With its long record and repeated rebuilding, Jericho would be a great candidate for a sort of key site, a place where evidence of repeated earthquake damage could be correlated with destruction in neighboring sites where the record might be less continuous. If we could systematically examine the written evidence of earthquakes and try to tie together the archaeological stratigraphy among the sites scattered around Israel and Jordan, we might be able to piece together a much more informative record than we could hope to gain from any one site alone. A great deal of archaeological work has been done to try to correlate the remains at Jericho with other sites throughout the Holy Land. Kathleen Kenyon (1978) was particularly comprehensive in her synthesis of the archaeology of the region. Although she does mention the general likelihood of earthquakes in this region and specifically ascribes certain damage layers to earthquakes, she makes no explicit effort to determine the range of such earthquake damage; she does not mention whether the earthquake in question could have damaged nearby towns.

One of the most suggestive layers in Jericho is one Kenyon dated, based on pottery styles, to between 1600 and 1550 BC, where the walls of Jericho collapsed, burying storage jars full of grain (Figure 7.4). The grain was carbonized by a fire concurrent with the collapse, and carbon dates of the grain have placed it remarkably close to Kenyon’s original estimate (Bruins and van der Plicht 1996). According to Kenyon (1979, 177–178), many towns in the region suffered total destruction at the same time (though she does not implicate an earthquake), including the sites of Tel Beit Mirsim, Hazor, Shechem, and Megiddo. Among these sites, only Megiddo was immediately rebuilt, with no apparent change in culture; most of the others were abandoned for more than a century.

In chapter 3, I speculated that an earthquake may have figured prominently in the biblical story of Joshua’s destruction of Jericho. Does the archaeological record preserve any evidence at all of Joshua’s Jericho? One of the early expeditions reported the discovery of a city wall from Joshua’s time, but Kenyon (1957) indicated that the dating of that wall was incorrect. As described in chapter 3, she found only one small remnant of Jericho that could have coincided with the traditional estimate of 1400 to 1250 BC for Joshua’s conquest of the site. If a larger habitation did exist at the site, the remains of that culture have all but vanished, eroded away during a long period of abandonment.

Recently, however, archaeologists have reexamined both the archaeological site at Jericho and radiocarbon dates for other events in the ancient world. One particularly notable theory links the plague of darkness in Egypt, before the Exodus, to the explosive eruption of the island of Thera in the Aegean Sea, an event dated to the end of the seventeenth century BC. If this date is correct, the nearly missing layer Kenyon attributed to Joshua’s time would be far too recent, and therefore the destruction from ca. 1600 would be a more likely candidate.

The debate over the assignment of this layer, based on samples of locally made pottery unearthed at Jericho, and on radiocarbon dates from the charred grain and from bits of timber found in the destruction layer, has at times been heated, almost vicious. (For an example of two diametrically opposed interpretations of the same evidence, presented side-by-side in one issue of Biblical Archaeology Review, complete with ad hominem attacks, see Bienkowski 1990 and Wood 1990.) It will be some time before the dust settles on this issue. For now, the archaeology is the source of more questions than answers, and until some of those questions can be resolved, using Jericho as a key site will remain difficult.

Although the mound of ancient Megiddo was eventually abandoned, and the modern city of Jericho has abandoned the old tell, the city of Jerusalem has been continuously inhabited for at least four thousand years (Cline 2004). The reason for its continued importance reaches beyond favorable topography or water supply to that least tangible reason for the persistence of cities: religious significance. King David’s choice of Jerusalem as the capital city of his Hebrew nation, and, probably more important, his decision to move the Ark of the Covenant into the city, sealed its fate as an enduring human habitation into modern times.

Despite its long history, most of the city’s archaeological secrets are nearly impossible to reveal. The whole city has been densely inhabited for millennia, with almost no abandonment in any quarter. Modern and ancient homes, churches, and official buildings block access to the layers of rubble beneath them. When a scrap of real estate is not actively occupied, it is usually because it is imbued with some intense political or religious importance and digging into it would touch off a firestorm of controversy.

As a result, excavations carried out in Jerusalem are often quite limited and piecemeal, undertaken only when an opportunity happens to present itself. This occurred in 1948, for example, when, during the Arab-Israeli war, a mortar hit a building that, according to tradition, housed the tomb of King David. In the process of repairing the damage inflicted by the explosion, the archaeologist Jacob Pinkerfeld (1990) also did a cursory excavation of one portion of the structure. Beneath the floor, he found over half a meter of debris, including three earlier floors. In fact, further investigations have suggested that the walls of this building were part of the Church of the Apostles, built in the first century AD at the place where the Last Supper was thought to have occurred. The walls have been destroyed and rebuilt many times, but because of the site’s importance to those who believe it was the tomb of David, the excavation was hindered and no further investigation was possible.

Still, although the city cannot be systematically excavated, modern instruments can probe beneath the surface to tell us not how old the layers are or what artifacts are in them but at least whether specific parts of the city are founded on archaeological debris or on solid ground. A government study by Israel’s Geological Survey was released in 2004, confirming what archaeologists had long suspected: the Old City is mostly founded on the rubble of previous constructions. As discussed in chapter 2, this makes the Old City particularly vulnerable to earthquake damage.

The historical record bears this out. In nearly every historical report of earthquakes affecting Jerusalem, damage is reported either to the Temple, the Old City walls, or both. Of course, the importance of this site and its role as the focus of religious pilgrimages for the world’s three major religions guarantees a record that is unparalleled in any other region. We have written records from many sources for earthquakes in Jerusalem, most of which also affected large regions of the countryside.

A recent geological investigation of sediments from the Dead Sea (Ken-Tor et al. 2001), confirms the validity of many of the historical reports mentioned in this chapter. As described in chapter 2, when shaking of sea-floor sediments is severe enough, the loose, water-saturated sediments lose their strength and flow like a liquid. When the shaking stops, the sediments settle and resolidify, leaving behind a chaotic, mixed layer that is readily identified, a seismite. These layers frequently contain organic material that scientists can date by Carbon-14 dating.

The 2001 study examined layers from the Ze’elim Terrace on the shore of the Dead Sea, a stack of sediments exposed by rapid modern drops in Dead Sea water levels (caused by diversion of the fresh water sources that feed the Dead Sea). There are gaps in this sequence, marking drought periods where the ancient water level in the Dead Sea was lower than usual and there was no sediment deposited at the Ze’elim Terrace site. However, during the periods when sediments accumulated there, every major historical earthquake on the Dead Sea Transform has been correlated to a seismite. This is a great new resource and one that can be extended in the future. The periods for which the Ze’elim Terrace contains no data could be examined by drilling into the deeper sediments in the Dead Sea floor. Thus, we can confirm the historical record at Jerusalem, and use it, along with the archaeological record in more accessible sites nearby, to build a physical earthquake stratigraphy for the region.