The article is dedicated to one of the less studied aspects of the Byzantine–Early Islamic period transition: the recycling of valuable materials—marble and bronze—as reflected in the archaeological findings discovered during the salvage excavations at the Givati Parking Lot site in Jerusalem. In the course of the work, a portion of one of the major streets of Byzantine Jerusalem was exposed, which once served as an important pilgrimage route of the city. During the Umayyad period the street was severely damaged, and the entire area was turned into an industrial zone. Of special interest are the raw materials used in the industrial installations discovered during the excavations. These include precious imported marble veneers and fragments of liturgical furniture, which had been used as raw material for lime production in a large limekiln constructed on the original course of the Byzantine street; and scraps of bronze fragments of liturgical vessels, some decorated with crosses, which were discovered together with metal slags, pottery crucibles, and fragments of production waste, all testifying to the presence of a small-scale metallurgical workshop at the site.
This article is dedicated to one of the less studied aspects of the Byzantine–Early Islamic period transition, as reflected in the archaeological findings discovered during the salvage excavations at the Givati Parking Lot site in Jerusalem.1 In the course of the work, a portion of one of the major streets of Byzantine Jerusalem was exposed, running from north to south along the Tyropoeon Valley toward the Pool of Siloam. During the Umayyad period the street was severely damaged, some of its slabs were removed, and a series of walls and plaster floors were built over it, turning the entire area into an industrial zone. Of special interest are the raw materials used in the industrial installations discovered during the excavations. These include precious imported marble veneers and fragments of liturgical furniture, which had been used as raw material for lime production in a large limekiln constructed on the original course of the Byzantine street.
A second category of finds includes scraps of bronze, as well as lead fragments of liturgical vessels, some decorated with crosses, which were discovered together with metal slags, pottery crucibles and fragments of production waste, testifying to the presence of a small-scale metallurgical workshop at the site. The phenomenon of reuse and spoliation of valuable materials and artifacts has become the subject of extensive study in recent decades, but it seems that the phenomenon of Early Islamic “recycling” is less well known and deserves special attention.
FINDS AND THEIR ARCHAEOLOGICAL CONTEXT
Excavations at the Givati Parking Lot (Fig. 1), located on the eastern slope of the Tyropoeon Valley in close proximity to the historical nucleus of ancient Jerusalem (the “City of David”), and some 30 m south of the Ottoman wall around Jerusalem's Old City, have revealed architectural remains from the Iron Age II (ninth century B.C.E.) to the Abbasid period (ninth–tenth centuries C.E.). The excavations have exposed a large area, totaling 0.5 hectares.2
Byzantine paved street
One of the most dominant layers exposed at the site dates to the Byzantine period, specifically to the fifth to early seventh centuries C.E. (Fig. 2). This layer, represented by three stratigraphic phases, includes an impressive complex, probably administrative in character, discovered in the northern part of the area, two additional buildings in its southern part, and agricultural terrain with a thick layer of garden soil retained by terrace walls.3 On the west, the Byzantine administrative building and the garden were flanked by the paved street which is the focus of this study (Fig. 3).
This wide north–south street measures ca. 5.5 m between the curbstones. Parts of this same street were exposed in the 1920s to the south,4 and recently also to the north.5 All told, 120 m of the street are known archaeologically today. Its continuation farther north supposedly runs parallel to the western wall of the Temple Mount, between the wall and the Eastern Cardo.6 Following the topography, the street slopes significantly southward: the elevation difference between the northern and southern excavated sections is nearly 5 m. The street is paved with flagstones of various sizes (0.4 x 0.4 m–0.8 x 1 m), smoothed from use, laid perpendicularly to the north–south axis of the street (Fig. 4). No traces of grooves left by wheels, no notches against slipping, and no game boards or other intentional marks were found. Curbstones delineated the sidewalks of the street, which were paved with rather smaller stones. Under the pavement a few sections of drainage channels were discovered that once served the buildings constructed east and west of the street, and connected to the major drainage running from north to south. Other sections of the street discovered to the south and to the north of the Givati excavations bear similar characteristics. The numismatic material discovered under the flagstones permits the pavement to be dated to the days of Justinian's rule (a coin generally dated to 527–567;7 a nummus of Justinian dated to 534–539;8 and fifth–sixth century coins).9
The sections of the street so far revealed represent one of the principal thoroughfares of Byzantine Jerusalem,10 which served as an important pilgrimage route connecting the sanctuaries of the city center in the north to the Pool of Siloam in the south. The Pool of Siloam as a venerated Christian site, the place of the Healing of the Blind (John 9.1-12), is first mentioned in pilgrimage descriptions of Jerusalem in the fourth century;11 however, the first to mention a church at the site is the Pilgrim of Piacenza, ca. 570.12
The remains of the church explored at the site have been dated to two periods: an early phase from the fifth century and a later phase to the sixth century C.E.13 It seems that during Justinian's reign, as part of the program of imperial patronage directed towards the city of Jerusalem, the Pool of Siloam was transformed into an important memorial shrine, one of the stations for major liturgical celebrations, and the whole area between the pool and the city center was turned into a thriving pilgrimage area.14 A series of shops were built along the paved street, discovered by Crowfoot and Fitzgerald,15 which supplied the needs of the numerous passersby, whose steps smoothed the hard limestones of the pavement. Additional testimony of pilgrim movement along the street was discovered in the Givati excavations, where a complete miniature box carved of bone was unearthed, adorned on the inside with two icon paintings.16
Using the classification recently proposed by Michele Voltaggio,17 we may safely include the Byzantine street running along the Tyropoeon Valley on a list of “liturgical streets” of the Holy City, built on the higher parts of the slope, linking the holy sites and covering certain segments of pilgrim routes.
End of the Byzantine period and Early Islamic changes
It seems difficult to propose a single, unequivocal explanation for the end of Byzantine structures built along the paved street. The large administrative building discovered in the Givati Parking Lot met a violent end, precisely dated by a large hoard of unused gold solidi of Emperor Heraclius, minted between 610 and 613 and sealed under the debris of a ruined structure.18 The hoard therefore provides a rare opportunity to determine an absolute dating of the building's destruction to the Persian invasion of Jerusalem in 614 CE.
It is possible that parts of the street went out of use in different times. The destruction layer locally exposed directly above the paved street (Fig. 5) should be dated to the second half of the seventh century.19 This impressive collapse, located at the southern end of the Givati excavation, contained ashlars and numerous profiled architectural fragments, probably originating from a monumental structure located east of the street.20 Hagbi and Uziel discovered a similar layer of collapse, although leveled off in a later period, above the northern fragment of the street.21 During this leveling operation, some of the ashlar stones were thrown into an underground cistern which went out of use at the end of the Byzantine period. To the south, Crowfoot and Fitzgerald's excavations presented no clear picture of destruction—neither debris, nor a clear collapse layer, nor significant complete pottery vessels were found. However, based on the historical narrative, the excavators ascribed the end of the Byzantine layer to the Persian conquest of 614 and, based on a number of Arab-Byzantine imitative coins (641–668), proposed that the area was briefly resettled thereafter.22
Sometime during the Early Islamic period the entire large area exposed in the Givati excavations witnessed dramatic changes and was turned into an industrial zone no longer devoted to domestic purposes.23 The Byzantine street running along the Tyropoeon Valley went completely out of use as a major city artery and was significantly narrowed. Some of the street's pavement slabs were damaged or entirely removed. No buildings of any kind were constructed along it other than a few carelessly built walls and fragmentary preserved plaster floors, mainly associated with a large lime kiln discovered in the northwestern part of the excavated area, or with a small-scale metallurgical workshop discovered in the southwest. It is important to stress the ephemeral nature of this new industrial enterprise: during the later stages of the Umayyad period, prior to the renewed Abbasid building activity at the site, the entire area was abandoned and was covered by thick layers (0.5–0.8 m) of alluvium.
A large rounded lime kiln, ca. 5 m in diameter and preserved to a height of ca. 2 m, was constructed against the wall of the destroyed Byzantine administrative building (Fig. 6). It was built of ashlars, probably taken from the ruined Byzantine building; due to exposure to the intense fire the stones had lost their original form and merged together. The inner face and floor of the kiln were covered with an orange clay coating to protect the walls from the intense heat.24 On the top, the kiln was covered by a heap of whitish limestone debris resting on a layer of dark ash, the result of the last use in the kiln. As noted, a few carelessly built walls associated with the kiln complex were also discovered.
Northwest of the kiln, the Byzantine street slabs were entirely removed and some of the nearby Byzantine building's walls were razed, leaving in their place heaps containing a large quantity of limestone and marble chunks (Fig. 7). These heaps no doubt served as raw material for lime production. This was where the largest concentration of marble finds was discovered in secondary deposition: the kiln itself and the associated piles of raw material—gravel and larger stone pieces intended to be burned—contained 160 marble fragments. Of these, only seven were of the colored marble sort; and the remaining 153 fragments are of the white and grayish sorts of Proconnesian marble. The analysis of the finds25 indicates that they included mainly fragments of opus sectile floors, wall veneers, and fragments of liturgical furniture taken from richly decorated structures.
Marble as a raw material for the production of high-quality lime was known to the Romans (Vitruvius, De arch. II.5), but it seems that such production became more common in Late Antiquity. A similar phenomenon has been documented in Rome,26 where during the ninth–tenth centuries the Forum Traiani was turned into a source of raw material for calcararii (lime burners) and sassori (masons). The discovery of marble use in Early Islamic and later lime kilns has been reported from various excavation sites in the region.27 The only report from the Jerusalem area of a concentration of raw material for lime production near a kiln, however, comes from the nearby excavations of B. Mazar, conducted at the foot of the Temple Mount and dated by the latter to the post-Fatimid phase.28
The proximity of both of these known marble concentrations in Jerusalem does not seem to be random. The large concentration of marble finds discovered in and around the lime kiln, built on the original course of the Byzantine street, may possibly be explained by the need for high-quality plaster for the large construction projects in Umayyad-period Jerusalem. The Early Islamic buildings on and around Haram al-Sharif reused large quantities of Roman and Byzantine marble decoration from the city of Jerusalem.29 It is plausible that some of the best marble in the city could have found secondary use in the nearby mosques and palatial structures, while the more poorly preserved or smaller fragments were simply burned in kilns or recycled as regular building material.
Small-scale metallurgical workshop
In the southern part of the Givati excavation, on the course of the Byzantine street, a long, narrow wall was constructed, built of well-dressed ashlars and clearly in secondary use. A series of plaster floors covering the original Byzantine pavement of the street and its eastern sidewalk was found abutting the wall on the west and east (Figs. 8a and 8b). Further north, Hagbi and Uziel observed a similar transformation of the Byzantine street during the Umayyad period.30 In the Givati excavations, the most interesting find from the level of the Umayyad plaster floors was a small-scale metallurgical workshop that had operated on that spot: numerous copper-alloy and lead fragments of Byzantine objects, some decorated with crosses, were discovered here, together with metal slags, production waste, and fragments of pottery crucibles31 (Figs. 9, 10). No proper furnace was found,32 indicating that production was by means of the simplest mode—clay casseroles and animal-hide bellows.33 On the plastered floors, on which large quantities of ash were found, fragments of liturgical implements were also discovered.34 These included swinging censers for incense burning or lighting polycandela with cross-adorned chains and calyx-shaped hooked hangers and a personal cross pendant (Fig. 11). Liturgical vessels of this type, generally dated to the sixth–seventh centuries, have been reported from a number of local archaeological sites35 and are mainly known from museum collections.36 The copper-alloy objects apparently served as scrap for local craftsmen, who recycled them for a new, unknown use.37
According to the stratigraphic position of the loci and the pottery and numismatic readings, the archaeological context of the finds should be dated to the late sixth to the first half of the seventh centuries. Additional dating may be provided by the personal cross pendant, for which nearly a dozen parallels from this region are known and generally dated to the sixth–seventh centuries (for example, crosses from urban contexts in Caesarea and Jericho;38 additional parallels have been reported from the Eastern Mediterranean and Central Europe).39 Based on a detailed typology of cross pendants from the necropoleis of the Crimea,40 our type can be precisely dated to the last quarter of the sixth to the first half of the seventh centuries.
Another, much larger assemblage of copper-alloy artifacts should be mentioned in association with metal recycling in Jerusalem during the period of Byzantine–Islamic transition. That assemblage was discovered by B. Mazar's excavation of a large Byzantine structure in Area XV of the Ophel, just few dozen meters northeast of the Givati site (Fig. 12).41 Based on an historical approach and archaeological deduction, that structure was dated to an unclear stage in the Byzantine period (fifth or sixth century); significant changes in its plan were dated to the middle of the sixth century, and its destruction was ascribed to the Persian invasion of 614. The metal hoard discovered there included various copper-alloy objects, among them metal parts of a wooden (?) box, a large processional or altar cross, and lamps.42
The metal artifacts discovered in the structure were initially interpreted by M. Ben-Dov as unconnected to the original complex, but rather to one of Jerusalem's magnificent churches, from which it had been brought to the site as a trophy in the post-Byzantine period.43 But during preparation of the final report on this building, its interpretation changed and it was identified as the Monastery of Virgins, known from Byzantine sources.
A critique of this identification is beyond the scope of this article. Still, it should be noted that the proposed new identification has led to reinterpretation of both the structure characteristics and finds.44 The finds of Christian character, i.e. liturgical marble furniture fragments and copper alloy crosses (“bronze,” according to the obsolete terminology of the publication), were assigned to a hypothetical second-story chapel. In fact, of the entire impressive assemblage of metal artifacts, only the large cross45 can be described securely as property belonging to a church or other ecclesiastic institution. All the rest—the clasps and lock for the wooden box,46 a lamp,47 doorknockers,48 and a dragon figure (lamp-holder?)49 may have belonged to any wealthy household, private or public.
Thus it seems that the metal finds originated from various structures, not necessarily from the building where they were discovered, and it was their diverse character that led the archaeologist Meir Ben-Dov to interpret them as trophies of some sort brought to the site.50 However, in light of recent finds from the Givati excavation, the Ophel metal hoard can be explained as the result of metal scavenging, dated to the post-Byzantine period, after the structure had already been abandoned.
The hypothesis of systematic re-melting of older metal materials in Late Antiquity was proposed a decade ago by Enrico Giannichedda,51 who listed the objective conditions favorable for such recycling (preserved technical knowledge, closure of government-owned mines, contacts with new people with different traditional techniques, the mobility of craftsmen, and the simplicity of their tools), but noted the lack of relevant and secure archaeological findings. Based on the medieval examples, Giannichedda assumed52 that the archaeological evidence for recycling should include evidence that metal objects had been robbed from ancient monuments and hoarded, as well as possibly broken and bent for reuse.53 It appears that for late antique Jerusalem, these signs of metal hoarding and recycling are now well attested.54
Who was to blame?
Although some of the marble revetment may have originated either from ecclesiastic or secular buildings, during the Byzantine period luxurious imported marble was known to have been mainly used for the decoration of churches and administrative buildings of high-ranking authorities.55 The quantity and variety of decoration and material suggests that these elements were not included in a single structure, but had been used at a number of different sites. Certain fragments may have originated from the Byzantine administrative building, or even in earlier, late Roman structures discovered at the Givati site, while others may have decorated nearby structures.
The presence of marble liturgical furniture and metal paraphernalia in the assemblages opens up the possibility for speculation regarding their origins. A few churches are known from the historical sources to have been located in the immediate vicinity of the Givati site: to the south, a church by the Pool of Siloam; to the west, the churches on Mount Zion—Holy Zion, Penitence of St. Peter, and the Nea; to the north, the Church of St. Sophia, or Pretoria, located not far from the Temple Mount. There were also numerous Christian structures to the east, on the slopes of Mount of Olives.56 The discoveries of some of these ecclesiastic complexes were reported from archaeological excavations.57 It should be stressed here that all the churches uncovered were found almost totally lacking their architectural décor or metal accessories—frames, pipes, clamps, grills, et cetera; indeed, they were most likely stripped and reused in later periods.
It is not easy to establish who was responsible for the massive displacement of the church décor. Byzantine law, both civil and canon, strictly prohibited alienation of church property, both real estate and movable. The Justinian codex repetitae praelectionis of 534 (1.2.21) strictly banned the sale or other disposition of sacred vessels or vestments, with the only exception being for the redemption of captives. A similar position was expressed by the early church fathers, although some insisted that vessels to be alienated must be broken and melted within the sacred precincts of the church.58 Formal adoption of this law was proclaimed by the Church only in the eighth century, under the growing pressure of Islamic conquests. The iconoclast Council of Hieria (the “Headless Council,” 754) prohibited the robbery of churches or the alteration of holy vessels, vestments, cloth, or anything dedicated to divine service under the pretext of destroying images. The decree of the Second Council of Nicea (787) threatened the excommunication of any bishops, other clerics, or laity who used sacred vessels for common purposes. The Eighth Ecumenical Council of Constantinople (869–870) declared that “no bishop is permitted to sell the sacred vessels or other valuables belonging to his Church except the case specified by the ancient canons, namely, when it is necessary for the redemption of captives.”59
It would be careless of us to presume that canon and civil law was always followed, and therefore to date Byzantine marble spolia solely to the Early Islamic period. In fact, in the Givati excavations, decorative Byzantine marble elements, including chancel screen fragments, were found in secondary use or simply discarded, not only in the Early Islamic period, but also in Byzantine layers.60 For example, the layers of Byzantine agricultural terraces revealed over 70 marble artifacts, and dozens more were discovered in Byzantine construction fills. Apparently the reason marble was neglected or reused was not necessarily for ideological or other significant reasons. Marble panels imported from afar may have reached their destination incomplete, or they may have been damaged during installation. Other examples of reuse of church marble, presumably pre-Islamic, were reported from the region.61 We may assume similar circumstances for metal reuse.
According to accumulated textual and archaeological data, the de-Christianization of Palestine was a gradual, slow process that took nearly two centuries.62 The final exclusion of the Palestinian Christian population from the Byzantine-oriented cultural sphere, the abandonment of churches, followed by their massive robbing, the prohibition of public display of crosses, and other restrictions against Christians63 would all develop only during the Abbasid period. The intentional, “ideological” destruction of church marble décor and liturgical vessels in seventh-century Jerusalem seems, therefore, most improbable. Rather, the archaeological evidence appears to be part of a phenomenon of “pragmatic recycling,” lacking any ideological motifs.
The archaeological picture of the Byzantine–Early Islamic transition in Jerusalem remains a highly debatable and complex issue. Once regarded as a definite break in the long development of the Roman and Byzantine city, it is now more commonly considered a prolonged, gradual process.64 Moreover, according to the archaeological record emerging from the excavations of Jerusalem, the metamorphosis of the urban settlement during this period of transition seems to be characterized by great diversity.65
In the case described here, one of the principal streets, once an important pilgrimage artery of Jerusalem, practically ceased to exist, temporarily giving way to essential industrial needs of the city. Elsewhere in Jerusalem, this phenomenon was also noted in the Byzantine colonnaded streets, mainly the two cardines crossing the city from north to south. Archaeologically, this process is best reflected in the Eastern Cardo, which was significantly narrowed during the Early Islamic period,66 and to a certain extent also in the Western Cardo, where the sidewalks were sealed and turned into shops.67 Similar processes have been documented in Caesarea,68 Beth Shean69 and other Decapolis cities,70 and other large urban centers in the region.71
The changes in urban development observed during the recent excavations in the Givati Parking Lot mark a definite turning point in the long continuity evident throughout the Late Roman and the Byzantine eras. These changes are characterized by rapid transformation of the public space into an industrial zone following the destruction and abandonment of residential and administrative structures once located in the area. Presumably, these changes must be viewed against the larger political and cultural background of Jerusalem in the second half of the seventh century C.E., with the construction of the Dome of the Rock and the Al-Aqsa Mosque, as well as the large Umayyad palatial complex south of Haram al-Sharif. According to Donald Whitcomb,72 the eastern and southeastern parts of Jerusalem underwent major changes, and a new Muslim city came into being there alongside the Christian city. The increase of industrial activity, first seen during the Umayyad period, would soon become a major characteristic of Early Islamic urban development.73 This phenomenon is well attested elsewhere in the region in the Abbasid city of Ramla.74
New archaeological data from the Givati excavations attests to the decline of an important pilgrimage route in early seventh-century Jerusalem and the recycling of church marble décor and liturgical metal vessels. It seems that now we can pose the same queries as in the past, recognizing that the answers are not as unequivocal as once believed. Did the political and religious changes affect the city's layout? How quickly did these changes occur? What was their impact on the Christian population of Jerusalem? Hopefully the evidence presented here will help refine our understanding of complex processes related to the Byzantine–Islamic transition.