It Is There: The Archaeological Evidence Proves It

034

Margreet Steiner makes three startling historical conclusions based on her analysis of the archaeological evidence from Jerusalem: (1) that during the Late Bronze Age (the period just before ancient Israel began to emerge in the central hill country) there was no town of Jerusalem but only a small pharaonic estate governed by a royal steward; (2) that during Iron Age I (the period of the Judges) there was still no town of Jerusalem; and (3) that only during early Iron Age II (the period of the United Monarchy, when King David and King Solomon are said to have reigned, and/or shortly thereafter) was a small town consisting primarily of administrative buildings established. My response to Steiner’s argument is based on my research as the member of Yigal Shiloh’s staff responsible for publishing the results of his excavations in Area G, the area in which he investigated the famous Stepped-Stone Structure and the soil- and stone-filled terraces.

Steiner observes that Jerusalem’s history in the Late Bronze and Iron Ages is usually based on analyses of written sources, such as the Bible and the Amarna letters, rather than archaeological evidence. Proposing a fresh approach, Steiner promises to analyze Jerusalem’s history “from the opposite direction, starting with the archaeological evidence.” Unfortunately, Steiner’s ensuing “analysis” ignores most of the published evidence. It also fails to present evidence from Kathleen Kenyon’s excavations that Steiner asserts supports her conclusions and that Steiner herself is preparing for scientific publication, and it fails to critically evaluate evidence that she does present. Consequently, Steiner offers her readers startling historical conclusions that are not substantiated by the archaeological record, that appear to contradict evidence she herself has published elsewhere and—worst of all—that perpetuate the propensity of some archaeologists to publish sensational claims while maintaining exclusive access to unpublished evidence so that other scholars cannot independently evaluate the evidence.

Steiner’s presentation of the archaeological evidence for the Late Bronze Age consists solely of two sentences: The first states that most of the Late Bronze Age pottery found in Jerusalem comes from tomb deposits; the second states that an Egyptian temple may have existed north of today’s Old City. Steiner then asserts, “No remains of a town, let alone a city, have ever been found: not a trace of an encircling wall, no gate, no houses. Not a single piece of architecture. Simply nothing!” In a footnote following this assertion, Steiner mentions that Kenyon dated a system of terraces excavated near the Gihon Spring to the Amarna period of the 14th century B.C.E., and that Shiloh suggested lowering its date to the 13th century B.C.E. (both Late Bronze Age), but that subsequent study of the pottery in Leiden and Jerusalem has “clearly shown that the terraces did not originate in the Late Bronze Age,” but rather in Iron Age I. Steiner then offers “three possible explanations as to why no Late Bronze Age remains have been discovered in Jerusalem: (1) Not enough has been excavated. (2) All remains have eroded or were dug away in antiquity. (3) There were no remains to begin with.” Citing large-scale excavations in the City of David conducted by Raymond Weill, J.W. Crowfoot, Kenyon and Shiloh, none of whom, she asserts, recovered either architecture or more than minuscule amounts of pottery ascribable to the Late Bronze Age, Steiner rejects the first two explanations in favor of the third: “These extensive excavations have not revealed any trace of the Late Bronze Age city because there was no city there in the Late Bronze Age.” Unable to ignore evidence presented by the Amarna letters completely, Steiner then suggests that Late Bronze Age Jerusalem was not a town or a city but “simply a royal domain of the pharaoh … [managed by a royal] steward.”

Contrary to Steiner’s assertions, architectural remains attributable to the Late Bronze Age have been reported from the northwest corner of Shiloh’s Area G,¹ and stratified remains containing pottery and other artifacts attributable to the Late Bronze Age have been reported from three additional areas excavated in the City of David: two dug by Kenyon (Trench A and Site P) and one dug by Shiloh (Area E1). Trench A: In 1990 Steiner herself published Late Bronze Age pottery found at the eastern edge of Trench A, including at least six of only eight sherds published from the first architectural phase identified in that vicinity.² Site P: On at least three 035occasions, Kenyon reported that the lowest levels of stratified remains found in her Site P—located immediately northwest of Shiloh’s Area G—dated to the Late Bronze Age and that in some places these remains reached half a meter (about 20 inches) in depth.³ Area E1: On at least two occasions, Shiloh reported sparse but stratified remains of the Late Bronze Age above remains of the Middle Bronze Age in Area E1.⁴

Steiner’s assertion that no more than minuscule amounts of pottery ascribable to the Late Bronze Age have been recovered by archaeologists conducting large-scale excavations in the City of David is both incorrect and disingenuous: incorrect because Late Bronze Age pottery, including imported Cypriot and Mycenean wares, has been found by several archaeologists, including R.A.S. Macalister and J.G. Duncan,⁵ Shiloh,⁶ and Kenyon;⁷ and disingenuous because archaeological reports generally treat pottery solely as a chronological indicator and, as such, describe only the latest pottery found in any given context. Earlier pottery deemed to be out of context is seldom if ever mentioned. In older excavations it was often discarded in the field.

In light of these published reports of stratified remains containing architecture, pottery and other artifacts attributable to the Late Bronze Age from at least four separate areas of the City of David, Steiner cannot cogently argue that “no Late Bronze Age remains have been discovered in Jerusalem.”

Based on ceramic studies performed in Leiden and Jerusalem, Steiner summarily rejects the Late Bronze Age dates offered by Kenyon and Shiloh for the terrace system excavated along the City of David’s eastern slope in favor of a date in Iron Age I. As evidence of the Leiden study, Steiner states that “in the terrace fill Kenyon recovered only 15 pottery sherds from the Late Bronze Age (some postdating the Amarna period), against several hundred sherds dating from Iron Age I.” Unfortunately, Steiner chooses neither to describe nor to illustrate the pottery in question. Moreover, because she additionally chooses not to describe the methods by which the pottery was gathered and/or quantified—and she has stated elsewhere that Kenyon’s habit was to save only rim sherds, to discard pottery after it had been classified and to omit details such as vessel diameter and surface treatment from descriptions of individual sherds⁸—her quantitative analysis does not appear to be either well founded or statistically valid.

As evidence of the Jerusalem study, Steiner cites the response that David Tarler and I presented to her lecture at the International Conference on Biblical Archaeology held in Jerusalem during June and July of 1990.⁹ In that response, Tarler and I dated the latest pottery recovered from the Stepped-Stone Structure’s rubble core to the 13th and/or 12th century B.C.E. and described it as identical to the pottery recovered from the terrace fills. Based on the identical character of pottery from these two ceramic assemblages and on structural connections evidenced between the two features, we concluded that the Stepped-Stone Structure and the terrace fills must have been contemporary and should be interpreted as the superstructure and substructure of a single architectural unit. We did not state that pottery attributable to Iron Age I dominated the ceramic assemblage from the terrace fills. On the contrary, the vast majority of sherds found in the terrace fills during Shiloh’s excavations derive from vessels that range in date from the Middle Bronze Age II to the Late Bronze Age II. Few if any sherds are definitely attributable to Iron Age I. Consequently, Tarler and I have consistently argued that the terrace fills date to the transitional period between Late Bronze Age II and Iron Age I, about the late 13th and/or early 12th century B.C.E.¹⁰—a date to which Steiner herself has elsewhere agreed.¹¹ Moreover, we have purposely refrained from speaking of the terrace fills in terms of a single cultural horizon.¹²

As noted by Steiner herself, attempts to reconstruct Late Bronze Age Jerusalem are hindered by the lack of completed excavation reports. They are additionally hindered by the sparse and fragmentary preservation of the remains. While these hindrances complicate our ability to reconstruct the Jerusalem of the Late Bronze Age, they are merely negative features of the archaeological record; they are not negative evidence, that is, they are not “evidence that there was no settlement.” Moreover, when analyzed within their spatial and temporal contexts, the Late Bronze Age remains from Jerusalem do not differ materially from those found at most other sites in the region.

While Steiner correctly states that no city wall or gate built during the Late Bronze Age has been identified in Jerusalem, the Late Bronze Age is not a period during which city walls and gates were constructed, but rather a period during which cities and towns either remained unfortified or reused fortification walls built during the Middle Bronze Age.¹³ As observed by Rivka Gonen: “A comprehensive examination of 77 sites and the evidence that keeps emerging in new excavations point to the absence of a city wall in the great majority of L[ate] B[ronze] settlements.”¹⁴ The absence of a city wall in Jerusalem at this time is not outstanding; it is consistent with the prevailing pattern.

Stratified remains of the Late Bronze Age have been documented in the City of David; they stretch from Shiloh’s Area E1 in the south to Kenyon’s Site P in the north, a distance greater than that documented for the Middle Bronze Age. Although sparse and 036fragmentarily preserved, remains from the Late Bronze Age are indisputably present in Jerusalem. Moreover, these remains are significant because they provide evidence of one of the few hill-country sites occupied during the Late Bronze Age,¹⁵ a period characterized by settlement decline and one better known for its cemeteries than for its cities and towns.¹⁶

As shown by Nadav Na’aman in his previous BAR article, the paucity of Late Bronze Age remains in Jerusalem is largely attributable to the local topography and to the building habits of the city’s later residents.¹⁷ Because stone was and is the most commonly available building material in the central hill country, Jerusalem’s builders regularly excavated to bedrock in order to secure both firm foundations and building stones. These building practices prevented the accumulation of superimposed archaeological strata characteristic of tell sites. As vividly described by Kenyon, they also caused irreparable damage to the archaeological record:

Evidence of early occupation on the summit area [of the City David] does not exist. This lacuna is mainly because Roman quarrying and Byzantine buildings have destroyed all earlier structures and earlier occupation. For all we know, the original height of the eastern ridge may have been appreciably above that of the surviving rock.¹⁸

Consequently, the best-preserved structures in Jerusalem are the last ones built, with earlier remains preserved only when reused or avoided by later builders. Because the City of David was heavily quarried as late as the Roman period, because its summit is now covered with modern buildings and because the last buildings constructed on its eastern slope were destroyed by the Babylonians in 586 B.C.E., archaeologists have not yet found—and may never find—rich deposits ascribable to the Late Bronze Age. For these reasons, we are not able to describe many physical features of the city’s Late Bronze Age occupation. This inability does not mean, as Steiner concludes, that there was no town in Jerusalem during the Late Bronze Age; it means only that the archaeological record has not been sufficiently preserved, that it has not yet been sufficiently developed and that we do not yet possess the expertise needed to interpret it.

In light of evidence from the Amarna letters, Steiner stops short of arguing that Jerusalem was wholly unoccupied during the Late Bronze Age, suggesting instead that at least during the Amarna period (14th century B.C.E.) it was the site of a small pharaonic estate managed by a royal steward. In support of this suggestion, Steiner cites Barry Gittlen, who proposed the concept of a “baronial estate” to explain the sparse and fragmentary nature of the Late Bronze Age remains unearthed at Tel Miqne-Ekron.¹⁹ To date, neither Gittlen nor Steiner has established criteria by which a “baronial estate,” or any other kind of small estate, can or should be distinguished from other types of settlements. Moreover, as Na’aman notes in this issue (“It Is There: Ancient Texts Prove It”), the Amarna letters themselves offer no evidence for the existence of 038estates managed by royal stewards. Because the difficulty of identifying archaeological remains ascribable to the 14th century B.C.E. in Jerusalem is mirrored at virtually every other Canaanite site mentioned in the Amarna letters,²⁰ Steiner’s suggestion finds no support in either the textual or the archaeological record.

Like her treatment of the Late Bronze Age, Steiner’s treatment of Iron Age I also suffers from the failure to present reported evidence and the failure to critically evaluate the evidence that she does present. Examples of evidence not presented include Shiloh’s report of stratified layers containing pottery and architecture ascribable to the Iron Age I in his Areas D1 and E1²¹ and Kenyon’s report of Iron Age I material in her Sites P and H.²² One example of Steiner’s failure to critically evaluate the evidence that she does present is the inherent contradiction between her assertion that “For Iron Age I, all that was found was a building with a complete collared-rim jar on its plastered floor, which allowed the building to be dated to Iron Age I”²³ and her assertion a few paragraphs later that “the terrace system [above the building with the collared rim jar] should be dated to the 12th century B.C.E., squarely in Iron Age I.” Thus, although she does not explicitly say so, Steiner identifies two strata ascribable to Iron Age I in the City of David: one represented by the building with the collared-rim jar and one represented by the terrace fills. Moreover, she offers no discussion concerning either the cultural significance of the terrace fills—a unique and imposing architectural feature that she herself describes as “the only fortified building known from Iron Age I in the hill country”²⁴—or the chronological significance of the collared-rim jar fragments.

Steiner uses the collared-rim jar fragments to date the building remains found beneath the terrace fills to Iron Age I without taking into account the many recent studies showing that these jars began to appear during the final phase of the Late Bronze Age (around the mid- to late 13th century B.C.E.)²⁵ and that they remained in use for a long period of time, extending at least to the tenth or ninth century B.C.E. west of the Jordan River²⁶ and perhaps even to the seventh century B.C.E. in Transjordan.²⁷ Accordingly, the mere presence of collared-rim jar fragments in a ceramic assemblage cannot be considered indicative of Iron Age I.²⁸ Indeed, collared-rim jars exhibit such a long range of use that isolated examples are not good chronological indicators of any period and must themselves be dated by the other pottery and artifacts with which they are found. Because the collared-rim jar fragments recovered from the building beneath the terrace fills were found together with cooking pots characterized by everted, triangular rims typical of the Late Bronze Age,²⁹ the building remains in which they were found should probably be ascribed to the final phase of the Late Bronze Age rather than to Iron Age I.³⁰

039

Steiner’s discussion of early Iron Age II (the period of the United Monarchy) focuses on the Stepped-Stone Structure, which she dates to the tenth century B.C.E. on the basis of pottery Kenyon found in its fill. As corroborating evidence, Steiner cites Kenyon’s recovery of pottery ascribable to the tenth century B.C.E. based on traces of habitation found at its foot, its connection to a casemate fortification wall and the discovery of a proto-aeolic capital, ashlars and an ashlar-built wall in its vicinity. None of this evidence supports Steiner’s date for the Stepped-Stone Structure.

In places where Shiloh removed stones belonging to the structure’s stepped mantle and he excavated sealed areas of the mantle’s fill, he recovered pottery that can only be ascribed to the transitional phase between Late Bronze Age II and early Iron Age I (late 13th-early 12th century B.C.E.). In contrast, where he excavated soil fills that covered the Stepped-Stone Structure, he recovered pottery that can only be ascribed to the final phase of Iron Age I or the early phase of Iron Age II (c. 11th or 10th century B.C.E.). Moreover, in places where he found that stones composing the structure’s stepped mantle had been removed in antiquity, he also found intrusions cut into its fill that contained pottery like that found on top of it, that is, pottery traditionally ascribable to the 11th and/or 10th century B.C.E. Finally, on the earliest floor surfaces built above the Stepped-Stone Structure, he found hand-burnished sherds with a dark red slip traditionally dated to the 10th century B.C.E., unslipped hand-burnished sherds, and fragments of a Phoenician bichrome flask generally datable to the 11th and/or 10th century B.C.E.³¹ Evidence from Shiloh’s excavation suggests, therefore, that the Stepped-Stone Structure was intentionally cut and partially dismantled sometime during the 10th century B.C.E. to accommodate new construction.

Neither Steiner nor Kenyon has indicated which, if any, areas of the Stepped-Stone Structure’s fill excavated by Kenyon were actually sealed by its mantle stones and which were not sealed. Nor have they indicated where the traces of habitation found at its foot were located, what their stratigraphic relation was to the Stepped-Stone Structure or how they differed from buildings constructed on top of it. The significance of these omissions is illustrated by Steiner’s assertion that Tarler and I have wrongly found the stepped mantle to be structurally connected to the terrace fills.

As proof that the stepped mantle is not structurally connected to the terrace fills, Steiner describes Kenyon’s findings in Square A XXIII, located immediately south of Area G. In this square, Steiner asserts, “there is clear evidence that the terrace system and the stepped mantle are two different architectural units.” There, she continues, “the Stepped-Stone Structure was not a mantle only (it could not be), but an impressive buildup of large stones. Kenyon excavated more than 13 (!) layers 040of large boulders on top of each other without reaching bedrock.” Unfortunately, the evidence from Square A XXIII is anything but clear, as is evidenced by Steiner’s own description of it in an earlier publication:

Southeast of the Maccabean tower [=Square A XXIII] she [Kenyon] discovered remains of a structure, dated by her to the 10th century B.C.E. No [architectural] plan has survived, but section drawings and photographs show clear evidence of its step-like construction. Iron II houses were built on top of it.³²

To date, none of the architectural plans or photographs published from Square A XXIII indicates that remains of the structure’s stepped mantle found there by Kenyon were preserved intact. And according to Steiner, she has no architectural plans to publish. Moreover, Steiner’s assertion that Kenyon failed to find remains of the substructural terraces beneath the remains of the stepped mantle excavated in Square A XXIII contradicts Kenyon’s own description of the stones pictured on page 32: “The large stones thought to be part of a retaining wall turned out to be a further terrace rebuild, this time in exceedingly massive stones. The stones were laid back against the sloping collapse of the earlier fill to the north, and were laid in regular horizontal courses which extended beyond the excavated area to the south … Moreover, since each course tended to overlap that below, the excavation area contracted rapidly and clearance had to be suspended at a height of 1.40 m above that at which bedrock was reached to the north. It was not therefore possible to expose the remains of the earlier terrace system that disappeared beneath it.”³³

Based on the evidence from Shiloh’s excavations, and on the lack of evidence indicating that any of the pottery Kenyon recovered from inside the Stepped-Stone Structure came from areas of fill sealed by its stepped mantle, the tenth-century pottery that Steiner uses to date the structure appears to be intrusive, that is, it appears to have been introduced into the Stepped-Stone Structure’s fill at the time it was cut and partially dismantled to accommodate the construction of new buildings. Consequently, the tenth century B.C.E. appears not to have been the century during which the Stepped-Stone Structure was built, but rather, the century during which it was superseded by new construction.

Finally, the architectural features that Steiner cites as corroboration for her dating of the Stepped-Stone Structure do not support her conclusion. The fragment of casemate wall identified by Kenyon is not—despite Steiner’s assertion—connected to the Stepped-Stone Structure; it is located more than 30 feet to its north in Site H.³⁴ The proto-aeolic capital and ashlar blocks found by Kenyon in destruction debris near the Stepped-Stone Structure were not—as intimated by Steiner—found in debris ascribable to the tenth century B.C.E.; they were found in debris resulting from the Babylonian destruction of 586 B.C.E.³⁵ Moreover, Yigal Shiloh, in his comprehensive study of Israelite ashlar masonry, found this capital stylistically comparable to those used in the Judahite palace at Ramat Rahel dated to the late seventh and early sixth centuries B.C.E. Consequently, Shiloh concluded that this capital could not be earlier in date than the ninth century B.C.E.³⁶ Excavators of the ashlar-built wall referenced by Steiner say that it cannot predate the eighth 041century B.C.E.³⁷ Because Jerusalem’s builders commonly used ashlar stones for the construction of both public and private buildings throughout Iron Age II, because the upper courses of the Stepped-Stone Structure remained exposed throughout the period, and because proto-aeolic capitals are often found in secondary use, the mere presence of these architectural features near the Stepped-Stone Structure indicates neither the character nor the date of the surrounding structures. Consequently, none of these architectural features support Steiner’s ascription of the Stepped-Stone Structure to the tenth century B.C.E.

The primary tasks of field archaeologists and their scholarly heirs, successors and assigns—such as Steiner and myself—are to identify, record, preserve and publish the archaeological record as objectively, comprehensively and quickly as possible. Absent completion of these primary tasks, historical interpretation of the archaeological record is premature and all too often unfounded. Steiner’s startling conclusions concerning the history of Jerusalem during the Late Bronze and early Iron Ages, which she says are not only supported but mandated by critical analysis of the archaeological record, are actually nothing more than the premature and unfounded musings of an archaeologist who has yet to complete the difficult task of critically analyzing either the published material or the unpublished material placed at her disposal. The community of archaeologists, historians and other parties interested in Jerusalem’s development during these periods expects more and deserves better.

Margreet Steiner makes three startling historical conclusions based on her analysis of the archaeological evidence from Jerusalem: (1) that during the Late Bronze Age (the period just before ancient Israel began to emerge in the central hill country) there was no town of Jerusalem but only a small pharaonic estate governed by a royal steward; (2) that during Iron Age I (the period of the Judges) there was still no town of Jerusalem; and (3) that only during early Iron Age II (the period of the United Monarchy, when King David and King Solomon are said to have reigned, […]